feat: 根据激光找出图片中心点坐标

.idea/.gitignore

@@ -0,0 +1,8 @@
+# 默认忽略的文件
+/shelf/
+/workspace.xml
+# 基于编辑器的 HTTP 客户端请求
+/httpRequests/
+# Datasource local storage ignored files
+/dataSources/
+/dataSources.local.xml

.idea/.name

@@ -0,0 +1 @@
+network.py

.idea/archery.iml

@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<module version="4">
+  <component name="PyDocumentationSettings">
+    <option name="format" value="PLAIN" />
+    <option name="myDocStringFormat" value="Plain" />
+  </component>
+</module>

.idea/inspectionProfiles/profiles_settings.xml

@@ -0,0 +1,6 @@
+<component name="InspectionProjectProfileManager">
+  <settings>
+    <option name="USE_PROJECT_PROFILE" value="false" />
+    <version value="1.0" />
+  </settings>
+</component>

.idea/misc.xml

@@ -0,0 +1,7 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="Black">
+    <option name="sdkName" value="Python 3.13 virtualenv at H:\iot\racingiot_v1\.venv" />
+  </component>
+  <component name="ProjectRootManager" version="2" project-jdk-name="maixcam" project-jdk-type="Python SDK" />
+</project>

.idea/vcs.xml

@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="" vcs="Git" />
+  </component>
+</project>

Untitled

@@ -0,0 +1 @@
+v1.2.15.1] [ERROR] main.py:416 - [MAIN] 显示异常: 'LaserManager' object has no attribute 'remote_detect_tick'

adc.py

@@ -6,12 +6,11 @@ a = adc.ADC(0, adc.RES_BIT_12)
 while True:
     raw_data = a.read()
     print(f"ADC raw data:{raw_data}")
-    # if raw_data > 2450:
-    #     print(f"ADC raw data:{raw_data}")
-    # elif raw_data < 2000:
-    #     print(f"ADC raw data:{raw_data}")
-    # time.sleep_ms(50)
+    if raw_data > 2450:
+        print(f"ADC raw data:{raw_data}")
+    elif raw_data < 2000:
+        print(f"ADC raw data:{raw_data}")
+    time.sleep_ms(1)
 
-    # vol = a.read_vol()
-
-    # print(f"ADC vol:{vol}")
+    vol = int(a.read_vol() * 10) / 10
+    print(f"ADC vol:{vol:.1f}, {time.time():.4f}")

app.yaml

@@ -1,11 +1,12 @@
 id: t11
 name: t11
-version: 1.2.11
+version: 2.1.1
 author: t11
 icon: ''
 desc: t11
 files:
   - 4g_download_manager.py
+  - 4g_upload_manager.py
   - app.yaml
   - archery_netcore.cpython-311-riscv64-linux-gnu.so
   - at_client.py
@@ -18,12 +19,11 @@ files:
   - main.py
   - model_270139.cvimodel
   - model_270139.mud
-  - model_270820.cvimodel
-  - model_270820.mud
   - network.py
   - ota_manager.py
   - power.py
   - server.pem
+  - set_autostart.py
   - shoot_manager.py
   - shot_id_generator.py
   - target_roi_yolo.py
@@ -34,3 +34,4 @@ files:
   - vision.py
   - wifi_config_httpd.py
   - wifi.py
+  - wpa_supplicant_conf.py

config.py

@@ -34,10 +34,10 @@ WIFI_QUALITY_RSSI_BAD_DBM = -80.0  # 低于此 dBm（更负更差）视为信号
 WIFI_QUALITY_USE_RSSI = True  # 是否把 RSSI 纳入综合判定
 
 # WiFi 热点配网（手机连设备 AP，浏览器提交路由器 SSID/密码；仅 GET/POST，标准库 socket）
-WIFI_CONFIG_AP_FALLBACK = True # # WiFi 配网失败时，是否退回热点模式，并等待重新配网
+WIFI_CONFIG_AP_FALLBACK = False # # WiFi 配网失败时，是否退回热点模式，并等待重新配网
 WIFI_AP_FALLBACK_WAIT_SEC = 5 # 等待5秒后再检测STA/4G
 WIFI_CONFIG_AP_TIMEOUT = 5  # 热点模式超时时间（秒）
-WIFI_CONFIG_AP_ENABLED = True  # True=启动时开热点并起迷你 HTTP 配网服务
+WIFI_CONFIG_AP_ENABLED = False  # True=启动时开热点并起迷你 HTTP 配网服务
 WIFI_CONFIG_AP_SSID = "ArcherySetup"       # 设备发出的热点名称
 WIFI_CONFIG_AP_PASSWORD = "12345678"       # 热点密码（WPA2 通常至少 8 位）
 WIFI_CONFIG_HTTP_HOST = "0.0.0.0"          # HTTP 监听地址
@@ -106,6 +106,14 @@ DEFAULT_LASER_POINT =  (320, 245)   # 默认激光中心点
 HARDCODE_LASER_POINT = True  # 是否使用硬编码的激光点（True=使用硬编码值，False=使用校准值）
 HARDCODE_LASER_POINT_VALUE = (320, 296) # 硬编码的激光点坐标(315, 245) #   # 硬编码的激光点坐标 (x, y)
 
+# 远程激光点识别（TCP cmd=200）：画面内找红点，稳定 N 秒且无明显跳动后上报坐标
+LASER_REMOTE_DETECT_STABLE_SEC = 3.0       # 连续稳定时长（秒）
+LASER_REMOTE_DETECT_MAX_MOVE_PX = 12.0     # 窗口内最大位移超过此值视为大幅移动，重新计时
+LASER_REMOTE_DETECT_SAMPLE_MS = 80         # 采样间隔
+LASER_REMOTE_DETECT_MIN_SAMPLES = 8        # 判定稳定前窗口内最少样本数
+LASER_REMOTE_DETECT_WARMUP_MS = 500        # cmd=200 开激光后等待稳定再采样
+# 远程识别会话无总超时：cmd=200 启动后持续检测并上报，直至 cmd=201 停止
+
 # 激光点检测配置
 LASER_DETECTION_THRESHOLD = 140  # 红色通道阈值（默认120，可调整，范围建议：100-150）
 LASER_RED_RATIO = 1.5  # 红色相对于绿色/蓝色的倍数要求（默认1.5，可调整，范围建议：1.3-2.0）
@@ -144,6 +152,13 @@ TRIANGLE_SIZE_RANGE = (8, 500)
 # 如果射箭距离很固定，可设具体范围（如 min=2.5, max=6.0）作为额外保险
 TRIANGLE_DISTANCE_MIN_M = 0.0   # 0=不启用下限检查
 TRIANGLE_DISTANCE_MAX_M = 0.0   # 0=不启用上限检查
+# 三角形方向校验：四角黑三角应为 ◤ ◥ / ◣ ◢，即三角形从外角指向靶心；用于过滤相邻靶混入/跨靶组合
+TRIANGLE_DIRECTION_VALIDATE_ENABLE = False
+TRIANGLE_DIRECTION_MIN_PASS = 3  # 至少多少个真实三角方向正确才认为该组有效；3点补全时推荐3，误检多可设2
+TRIANGLE_DIRECTION_DOT_MIN = 0.0  # 方向点积阈值；0=只要求同向半平面，0.35≈夹角<70°，0.5≈夹角<60°
+TRIANGLE_DIRECTION_TO_CENTER_DOT_MIN = 0.35  # 必须指向候选靶心；0.35≈夹角<70°，用于过滤相邻靶混入
+TRIANGLE_CENTER_DISTANCE_VALIDATE_ENABLE = True  # 四角三角到候选靶心距离需近似一致，过滤跨靶组合
+TRIANGLE_CENTER_DISTANCE_RATIO_TOL = 0.45  # (max_dist-min_dist)/mean_dist 最大允许值；越小越严格
 # 三角形检测兜底增强：CLAHE（更鲁棒但更慢）。颜色阈值修复后通常不需要，保持关闭以优先速度。
 TRIANGLE_ENABLE_CLAHE_FALLBACK = False
 # 三角形检测调试：保存 Otsu 二值化图像（临时调试用，定位后关闭）
@@ -169,6 +184,7 @@ TRIANGLE_SHAPE_COS_TOLERANCE = 0.25    # 直角余弦绝对值上限（原 0.20
 # 建议设为实测最坏耗时的 1.2 倍；超时后圆心检测仍会并行跑完，跑完后若三角形已结束则优先用三角形。
 TRIANGLE_TIMEOUT_MS = 1000
 # True=打印各阶段耗时(ms)，用于定位瓶颈；稳定后可 False 减少日志
+ARCHERY_TIMING_ENABLE = False  # 总开关：False 关闭所有算法耗时统计（shoot_manager + triangle_target + vision）
 TRIANGLE_TIMING_LOG = True
 # True=Stage2 每个子框内传统三角失败时打一条统计（Otsu/Adaptive 下轮廓数与各拒绝原因计数）
 TRIANGLE_LOG_STAGE2_PATCH_REJECT = True
@@ -255,6 +271,14 @@ TRIANGLE_YOLO_REJECT_BAD_ROI = True
 TRIANGLE_CROP_ROI_MIN_SIDE_PX = 64
 # 射箭保存图 / 预览上绘制 YOLO 靶环 ROI 矩形 (x0,y0,x1,y1)，核对是否裁准；不需要时改 False
 TRIANGLE_YOLO_DRAW_ROI_ON_SHOT = True
+# 物方采样调试：以靶心为中心，取半径 15cm 的圆周样本点，用于黑/白颜色对比
+TRIANGLE_SAMPLE_ENABLE = True
+TRIANGLE_SAMPLE_TIMING_ENABLE = True  # 仅统计物方采样耗时（其他 timing 可关）
+TRIANGLE_SAMPLE_RADIUS_CM = 15.0
+TRIANGLE_SAMPLE_ANGLES_DEG = (0, 90, 180, 270)
+TRIANGLE_SAMPLE_PATCH_HALF_PX = 2
+# 物方采样判断黑白阈值（R/G/B 均小于此值视为黑）；40cm 黑靶在靶面位置全黑，20cm 白靶则 R/G/B 偏高
+TRIANGLE_SAMPLE_BLACK_THRESH = 30.0
 # 开机阶段预加载 YOLO detector；detect 使用 dual_buff=False，避免返回上一帧结果。
 TRIANGLE_YOLO_PRELOAD_ON_BOOT = True
 
@@ -306,12 +330,14 @@ LASER_LENGTH = 2
 
 # ==================== 图像保存配置 ====================
 SAVE_IMAGE_ENABLED = True  # 是否保存图像（True=保存，False=不保存）
+SAVE_RAW_SHOT_IMAGE_ENABLED = False  # 是否额外保存射箭原图；可通过 TCP cmd=46 动态开关
+VISION_TIMING_ENABLE = True  # 视觉圆检测耗时统计（detect_circle_v3 内部各步骤耗时）
 PHOTO_DIR = "/root/phot"    # 照片存储目录
 MAX_IMAGES = 1000
 # Stage2 调试目录（默认 PHOTO_DIR/stage2_roi）内 JPEG 最多保留张数；None 表示与 MAX_IMAGES 相同
 TRIANGLE_BLACK_YOLO_STAGE2_ROI_MAX_IMAGES = None
 
-SHOW_CAMERA_PHOTO_WHILE_SHOOTING = False  # 是否在拍摄时显示摄像头图像（True=显示，False=不显示），建议在连着USB测试过程中打开
+SHOW_CAMERA_PHOTO_WHILE_SHOOTING = True  # 是否在拍摄时显示摄像头图像（True=显示，False=不显示），建议在连着USB测试过程中打开
 
 # ==================== OTA配置 ====================
 MAX_BACKUPS = 5

cpp_ext/decrypt_ota_file.cpp

@@ -1,14 +1,11 @@
-#include <pybind11/pybind11.h>
-#include <pybind11/stl.h>  // 支持 std::vector, std::map 等
-#include <nlohmann/json.hpp>
 #include <cstring>
 #include <cstdint>
 #include <vector>
 #include <string>
 #include <fstream>
 #include <array>
-#include <openssl/evp.h>
 #include <algorithm>
+#include <openssl/evp.h>
 #include "native_logger.hpp"
 
 namespace netcore{
@@ -18,11 +15,11 @@ namespace netcore{
     constexpr size_t kOtaMagicLen = 7;
     constexpr size_t kGcmNonceLen = 12;
     constexpr size_t kGcmTagLen = 16;
+    constexpr size_t kHeaderLen = kOtaMagicLen + kGcmNonceLen;
+    // 分块解密，避免整包读入导致 RAM 峰值约为「文件大小×2」（小内存设备易 OOM）
+    constexpr size_t kDecryptChunk = 65536;
 
-    // 固定 32-byte AES-256-GCM key（提高被直接查看的成本；不是绝对安全）
-    // 注意：需要与打包端传入的 --aead-key-hex 保持一致。
     static std::array<uint8_t, 32> ota_key_bytes() {
-        // 简单拆分混淆：key = a XOR b
         static const std::array<uint8_t, 32> a = {
             0x92,0x99,0x4d,0x06,0x6f,0xb6,0xa6,0x3d,0x85,0x08,0xbe,0x73,0x5e,0x73,0x4d,0x8a,
             0x53,0x88,0xe6,0x99,0xfc,0x10,0x29,0xb9,0x16,0x9b,0xe7,0x0c,0x65,0x21,0x1c,0xce
@@ -36,56 +33,45 @@ namespace netcore{
         return k;
     }
 
-    static bool read_file_all(const std::string& path, std::vector<uint8_t>& out) {
-        std::ifstream ifs(path, std::ios::binary);
-        if (!ifs) return false;
-        ifs.seekg(0, std::ios::end);
-        std::streampos size = ifs.tellg();
-        if (size <= 0) return false;
-        ifs.seekg(0, std::ios::beg);
-        out.resize(static_cast<size_t>(size));
-        if (!ifs.read(reinterpret_cast<char*>(out.data()), size)) return false;
-        return true;
-    }
-
-    static bool write_file_all(const std::string& path, const uint8_t* data, size_t len) {
-        std::ofstream ofs(path, std::ios::binary | std::ios::trunc);
-        if (!ofs) return false;
-        ofs.write(reinterpret_cast<const char*>(data), static_cast<std::streamsize>(len));
-        return static_cast<bool>(ofs);
-    }
-
     bool decrypt_ota_file_impl(const std::string& input_path, const std::string& output_zip_path) {
-        std::vector<uint8_t> in;
-        if (!netcore::read_file_all(input_path, in)) {
-            netcore::log_error(std::string("decrypt_ota_file: read failed: ") + input_path);
+        std::ifstream ifs(input_path, std::ios::binary);
+        if (!ifs) {
+            netcore::log_error(std::string("decrypt_ota_file: open in failed: ") + input_path);
             return false;
         }
-
-        const size_t min_len = kOtaMagicLen + kGcmNonceLen + kGcmTagLen + 1;
-        if (in.size() < min_len) {
+        ifs.seekg(0, std::ios::end);
+        const std::streampos szp = ifs.tellg();
+        if (szp <= 0) {
+            netcore::log_error("decrypt_ota_file: empty input");
+            return false;
+        }
+        const uint64_t file_size = static_cast<uint64_t>(szp);
+        const size_t min_len = kHeaderLen + kGcmTagLen + 1;
+        if (file_size < min_len) {
             netcore::log_error("decrypt_ota_file: too short");
             return false;
         }
-        if (!std::equal(in.begin(), in.begin() + kOtaMagicLen, reinterpret_cast<const uint8_t*>(kOtaMagic))) {
+        const uint64_t ciphertext_len = file_size - kHeaderLen - kGcmTagLen;
+
+        ifs.seekg(0, std::ios::beg);
+        std::array<uint8_t, kHeaderLen> header{};
+        ifs.read(reinterpret_cast<char*>(header.data()), static_cast<std::streamsize>(kHeaderLen));
+        if (ifs.gcount() != static_cast<std::streamsize>(kHeaderLen)) {
+            netcore::log_error("decrypt_ota_file: read header failed");
+            return false;
+        }
+        if (!std::equal(header.begin(), header.begin() + kOtaMagicLen,
+                         reinterpret_cast<const uint8_t*>(kOtaMagic))) {
             netcore::log_error("decrypt_ota_file: bad magic");
             return false;
         }
+        const uint8_t* nonce = header.data() + kOtaMagicLen;
 
-        const uint8_t* nonce = in.data() + kOtaMagicLen;
-        const uint8_t* ct_and_tag = in.data() + kOtaMagicLen + kGcmNonceLen;
-        const size_t ct_and_tag_len = in.size() - (kOtaMagicLen + kGcmNonceLen);
-        if (ct_and_tag_len <= kGcmTagLen) {
-            netcore::log_error("decrypt_ota_file: no ciphertext");
+        std::ofstream ofs(output_zip_path, std::ios::binary | std::ios::trunc);
+        if (!ofs) {
+            netcore::log_error(std::string("decrypt_ota_file: open out failed: ") + output_zip_path);
             return false;
         }
-        const size_t ciphertext_len = ct_and_tag_len - kGcmTagLen;
-        const uint8_t* ciphertext = ct_and_tag;
-        const uint8_t* tag = ct_and_tag + ciphertext_len;
-
-        std::vector<uint8_t> plain(ciphertext_len);
-        int out_len1 = 0;
-        int out_len2 = 0;
 
         EVP_CIPHER_CTX* ctx = EVP_CIPHER_CTX_new();
         if (!ctx) {
@@ -95,6 +81,8 @@ namespace netcore{
 
         bool ok = false;
         auto key = ota_key_bytes();
+        std::vector<uint8_t> chunk_in(kDecryptChunk);
+        std::vector<uint8_t> chunk_out(kDecryptChunk + EVP_MAX_BLOCK_LENGTH);
 
         do {
             if (1 != EVP_DecryptInit_ex(ctx, EVP_aes_256_gcm(), nullptr, nullptr, nullptr)) {
@@ -109,27 +97,59 @@ namespace netcore{
                 netcore::log_error("decrypt_ota_file: set key/iv failed");
                 break;
             }
-            if (1 != EVP_DecryptUpdate(ctx, plain.data(), &out_len1, ciphertext, static_cast<int>(ciphertext_len))) {
-                netcore::log_error("decrypt_ota_file: update failed");
+
+            uint64_t remaining = ciphertext_len;
+            while (remaining > 0) {
+                const size_t n = static_cast<size_t>(std::min<uint64_t>(remaining, kDecryptChunk));
+                ifs.read(reinterpret_cast<char*>(chunk_in.data()), static_cast<std::streamsize>(n));
+                if (ifs.gcount() != static_cast<std::streamsize>(n)) {
+                    netcore::log_error("decrypt_ota_file: read ciphertext chunk failed");
+                    goto cleanup_ctx;
+                }
+                int outl = 0;
+                if (1 != EVP_DecryptUpdate(ctx, chunk_out.data(), &outl,
+                                          chunk_in.data(), static_cast<int>(n))) {
+                    netcore::log_error("decrypt_ota_file: update failed");
+                    goto cleanup_ctx;
+                }
+                if (outl > 0) {
+                    ofs.write(reinterpret_cast<const char*>(chunk_out.data()), outl);
+                    if (!ofs) {
+                        netcore::log_error("decrypt_ota_file: write plaintext failed");
+                        goto cleanup_ctx;
+                    }
+                }
+                remaining -= n;
+            }
+
+            std::array<uint8_t, kGcmTagLen> tag{};
+            ifs.read(reinterpret_cast<char*>(tag.data()), static_cast<std::streamsize>(kGcmTagLen));
+            if (ifs.gcount() != static_cast<std::streamsize>(kGcmTagLen)) {
+                netcore::log_error("decrypt_ota_file: read tag failed");
                 break;
             }
-            if (1 != EVP_CIPHER_CTX_ctrl(ctx, EVP_CTRL_GCM_SET_TAG, static_cast<int>(kGcmTagLen), const_cast<uint8_t*>(tag))) {
+            if (1 != EVP_CIPHER_CTX_ctrl(ctx, EVP_CTRL_GCM_SET_TAG, static_cast<int>(kGcmTagLen), tag.data())) {
                 netcore::log_error("decrypt_ota_file: set tag failed");
                 break;
             }
-            if (1 != EVP_DecryptFinal_ex(ctx, plain.data() + out_len1, &out_len2)) {
+
+            int outl2 = 0;
+            if (1 != EVP_DecryptFinal_ex(ctx, chunk_out.data(), &outl2)) {
                 netcore::log_error("decrypt_ota_file: final failed (auth tag mismatch?)");
                 break;
             }
-            const size_t plain_len = static_cast<size_t>(out_len1 + out_len2);
-            if (!netcore::write_file_all(output_zip_path, plain.data(), plain_len)) {
-                netcore::log_error(std::string("decrypt_ota_file: write failed: ") + output_zip_path);
-                break;
+            if (outl2 > 0) {
+                ofs.write(reinterpret_cast<const char*>(chunk_out.data()), outl2);
+                if (!ofs) {
+                    netcore::log_error("decrypt_ota_file: write final failed");
+                    break;
+                }
             }
             ok = true;
         } while (false);
 
+cleanup_ctx:
         EVP_CIPHER_CTX_free(ctx);
         return ok;
     }
-}
+}  // namespace netcore

cpp_ext/tcp_ssl_password.cpp

@@ -0,0 +1,33 @@
+#include "tcp_ssl_password.hpp"
+
+#include <openssl/md5.h>
+#include <sstream>
+#include <iomanip>
+
+namespace netcore {
+
+static std::string md5_hex(const std::string& input) {
+    MD5_CTX ctx;
+    MD5_Init(&ctx);
+    MD5_Update(&ctx, input.data(), input.size());
+
+    unsigned char digest[MD5_DIGEST_LENGTH];
+    MD5_Final(digest, &ctx);
+
+    std::ostringstream oss;
+    oss << std::hex << std::setfill('0');
+    for (int i = 0; i < MD5_DIGEST_LENGTH; ++i) {
+        oss << std::setw(2) << static_cast<unsigned int>(digest[i]);
+    }
+    return oss.str();
+}
+
+std::string calculate_tcp_ssl_password(const std::string& device_id, const std::string& iccid) {
+    std::string md5_device_hex = md5_hex(device_id);
+    if (!iccid.empty()) {
+        md5_device_hex += iccid;
+    }
+    return md5_hex(md5_device_hex);
+}
+
+} // namespace netcore

cpp_ext/tcp_ssl_password.hpp

@@ -0,0 +1,7 @@
+#pragma once
+
+#include <string>
+
+namespace netcore {
+    std::string calculate_tcp_ssl_password(const std::string& device_id, const std::string& iccid);
+}

design_doc/command_record.md

@@ -1,12 +1,12 @@
 
-1. CPP构建命令：
+1. CPP构建命令：在docker环境下执行以下命令
 
-cd /mnt/d/code/archery/cpp_ext
+cd /data/cpp_ext
 rm -rf build && mkdir build && cd build
 
-TOOLCHAIN_BIN=/mnt/d/code/MaixCDK/dl/extracted/toolchains/maixcam/host-tools/gcc/riscv64-linux-musl-x86_64/bin
-PYDEV=/mnt/d/code/shooting/python3_lib_maixcam_musl_3.11.6
-MAIXCDK=/mnt/d/code/MaixCDK
+TOOLCHAIN_BIN=/data/MaixCDK-main/dl/extracted/toolchains/maixcam/host-tools/gcc/riscv64-linux-musl-x86_64/bin
+PYDEV=/data/python3_lib_maixcam_musl_3.11.6
+MAIXCDK=/data/MaixCDK-main
 
 cmake .. -G Ninja \
   -DCMAKE_C_COMPILER="${TOOLCHAIN_BIN}/riscv64-unknown-linux-musl-gcc" \
@@ -90,5 +90,13 @@ opencv_interactive-calibration -t=chessboard -w=9 -h=6 -sz=0.025 -v="http://192.
  D:\data\test_target_photo 是用来叠加的背景图
 
  7.1 生成靶纸及黑色三角形的截图的图片，带动动，但1.12的外框
+bak
 python .\synth_compose_yolo.py  --perspective 0.04 --perspective-prob 0.8 --color-jitter 0.6  --bg-dir D:\data\test_target_photo --fg D:\code\shooting\target_photo\write.png --out ./synth_out --class-name triangle --zip ./maix_dataset.zip --num 60 --triangles-json archery_triangles_default.json --format voc  --stage2-crop  --stage2-pad-min 0.03  --stage2-pad-max 0.18 --motion-prob 0.9 --motion-kernel-max 8 --blur-max 0 --triangle-bbox-pad-frac 0.12
 
+bak_2
+python synth_keypoints_right_angle.py   --bg-dir D:\data\test_target_photo --fg D:\code\shooting\target_photo\write.png  --triangles-json archery_triangles_default.json --out ./synth_out --num 1000  --offscreen-shift-prob 0.3  --offscreen-shift-frac 0.4   --offscreen-min-visible 1  --stage2-crop  --stage2-pad-min 0.03  --stage2-pad-max 0.18 --motion-prob 0.9 --motion-kernel-max 8 --blur-max 0   --perspective-mode planar   --yaw-max-deg 10 --pitch-max-deg 8   --roll-max-deg 4   --planar-focal-frac 1.45   --perspective-prob 0.4
+
+python synth_keypoints_right_angle.py   --bg-dir D:\data\test_target_photo --fg D:\code\shooting\target_photo\write.png  --triangles-json archery_triangles_default.json --out ./synth_out --num 1000  --offscreen-shift-prob 0.3  --offscreen-shift-frac 0.4   --offscreen-min-visible 1  --stage2-crop  --stage2-pad-min 0.03  --stage2-pad-max 0.18 --motion-prob 1.0 --motion-kernel-max 8 --blur-max 0   --perspective-mode planar   --yaw-max-deg 10 --pitch-max-deg 8   --roll-max-deg 4   --planar-focal-frac 1.45   --perspective-prob 0.4
+
+
+ python pose_pixel_metrics.py --model D:\code\archery\runs\pose\runs\pose\target_pose_train\weights\best.pt --data D:\code\archery\datasets\dataset_pose.yaml --imgsz 640

laser_manager.py

@@ -6,6 +6,7 @@
 """
 import _thread
 import json
+import math
 import os
 import binascii
 from maix import time
@@ -34,9 +35,13 @@ class LaserManager:
         self._calibration_active = False
         self._calibration_result = None
         self._calibration_lock = threading.Lock()
+        self._remote_detect_active = False
+        self._remote_detect_lock = threading.Lock()
+        self._remote_detect_result = None
         self._laser_point = None
         self._laser_turned_on = False
         self._last_frame_with_ellipse = None  # 保存绘制了椭圆的图像（用于调试/显示）
+        self._remote_detect_last_pos = None
         self._initialized = True
     
     # ==================== 状态访问（只读属性）====================
@@ -113,7 +118,7 @@ class LaserManager:
         
         # 正常模式：从配置文件加载
         try:
-            if "laser_config.json" in os.listdir("/root"):
+            if os.path.exists(config.CONFIG_FILE):
                 with open(config.CONFIG_FILE, "r") as f:
                     data = json.load(f)
                     if isinstance(data, list) and len(data) == 2:
@@ -124,11 +129,240 @@ class LaserManager:
                         raise ValueError
             else:
                 self._laser_point = config.DEFAULT_LASER_POINT
-        except:
+        except Exception as e:
+            if self.logger:
+                self.logger.warning(f"[LASER] 加载激光点失败，使用默认值: {e}")
             self._laser_point = config.DEFAULT_LASER_POINT
         
         return self._laser_point
     
+    @property
+    def remote_detect_active(self):
+        with self._remote_detect_lock:
+            return self._remote_detect_active
+
+    def get_remote_detect_result(self):
+        """获取并清除远程激光识别结果 (x, y) 或 None。"""
+        with self._remote_detect_lock:
+            result = self._remote_detect_result
+            self._remote_detect_result = None
+            return result
+
+    def remote_detect_tick(self, frame):
+        """
+        主循环显示路径调用的轻量 tick。
+        兼容旧调用点：当前远程识别由后台线程处理，这里不做重计算，
+        仅保留接口避免 AttributeError。
+        """
+        return None
+
+    def overlay_remote_detect_preview(self, frame):
+        """
+        在预览画面叠加远程识别点与坐标文本。
+        """
+        try:
+            import cv2
+            from maix import image
+            with self._remote_detect_lock:
+                pos = self._remote_detect_last_pos
+            if not pos:
+                return frame
+
+            img_cv = image.image2cv(frame, False, False)
+            if img_cv is None or img_cv.size == 0:
+                return frame
+
+            x, y = int(pos[0]), int(pos[1])
+            h, w = img_cv.shape[:2]
+            if x < 0 or y < 0 or x >= w or y >= h:
+                return frame
+
+            color = (255, 0, 0)  # RGB
+            cv2.circle(img_cv, (x, y), 8, color, 2)
+            cv2.line(img_cv, (x - 12, y), (x + 12, y), color, 1)
+            cv2.line(img_cv, (x, y - 12), (x, y + 12), color, 1)
+            cv2.putText(img_cv, f"laser=({x},{y})", (max(5, x + 10), max(20, y - 10)),
+                        cv2.FONT_HERSHEY_SIMPLEX, 0.55, color, 1, cv2.LINE_AA)
+
+            return image.cv2image(img_cv, False, False)
+        except Exception as e:
+            if self.logger:
+                self.logger.debug(f"[LASER-REMOTE] overlay 绘制失败: {e}")
+            return frame
+
+    def _set_remote_detect_result(self, result):
+        with self._remote_detect_lock:
+            self._remote_detect_result = result
+
+    def set_hardcoded_laser_point(self, x, y):
+        """更新 config.HARDCODE_LASER_POINT_VALUE（TCP cmd=201）。"""
+        try:
+            ix = int(round(float(x)))
+            iy = int(round(float(y)))
+        except (TypeError, ValueError) as e:
+            raise ValueError(f"invalid laser point ({x!r}, {y!r})") from e
+        config.HARDCODE_LASER_POINT = True
+        config.HARDCODE_LASER_POINT_VALUE = (ix, iy)
+        self._laser_point = (ix, iy)
+        try:
+            with open(config.CONFIG_FILE, "w") as f:
+                json.dump([ix, iy], f)
+        except Exception as e:
+            if self.logger:
+                self.logger.warning(f"[LASER] 保存硬编码激光点到本地失败: {e}")
+            raise
+        if self.logger:
+            self.logger.info(
+                f"[LASER] 已设置硬编码激光点 HARDCODE_LASER_POINT_VALUE=({ix}, {iy}) 并已保存到 {config.CONFIG_FILE}"
+            )
+        return ix, iy
+
+    def start_remote_laser_detect(self):
+        """
+        启动远程激光识别会话（TCP cmd=200）：开激光后持续检测。
+        每次稳定 3s 上报一次坐标，外循环直到 cmd=201 调用 stop_remote_laser_detect()。
+        Returns:
+            True 已启动；False 会话已在运行
+        """
+        with self._remote_detect_lock:
+            if self._remote_detect_active:
+                return False
+            self._remote_detect_active = True
+            self._remote_detect_result = None
+            self._remote_detect_last_pos = None
+        _thread.start_new_thread(self._remote_laser_detect_worker, ())
+        if self.logger:
+            self.logger.info("[LASER] 远程激光识别已启动 (cmd=200)")
+        return True
+
+    def stop_remote_laser_detect(self):
+        with self._remote_detect_lock:
+            self._remote_detect_active = False
+
+    def _remote_laser_detect_worker(self):
+        from camera_manager import camera_manager
+
+        stable_sec = float(getattr(config, "LASER_REMOTE_DETECT_STABLE_SEC", 3.0))
+        max_move = float(getattr(config, "LASER_REMOTE_DETECT_MAX_MOVE_PX", 12.0))
+        sample_ms = int(getattr(config, "LASER_REMOTE_DETECT_SAMPLE_MS", 80))
+        min_samples = int(getattr(config, "LASER_REMOTE_DETECT_MIN_SAMPLES", 8))
+        warmup_ms = int(getattr(config, "LASER_REMOTE_DETECT_WARMUP_MS", 500))
+        stable_ms = int(max(500, stable_sec * 1000))
+
+        samples = []
+        miss_count = 0
+        stable_hit_count = 0
+        reported = False
+
+        try:
+            if not self._laser_turned_on:
+                try:
+                    self.turn_on_laser()
+                except Exception as e:
+                    if self.logger:
+                        self.logger.warning(f"[LASER] cmd200 worker 开激光失败: {e}")
+            if warmup_ms > 0:
+                if self.logger:
+                    self.logger.info(f"[LASER] cmd200 激光预热 {warmup_ms}ms …")
+                time.sleep_ms(warmup_ms)
+
+            if self.logger:
+                self.logger.info("[LASER] 远程识别外循环已启动，直至 cmd=201 停止")
+
+            while True:
+                with self._remote_detect_lock:
+                    if not self._remote_detect_active:
+                        if self.logger:
+                            self.logger.info("[LASER] 远程识别会话结束 (cmd=201 或取消)")
+                        return
+
+                try:
+                    frame = camera_manager.read_frame()
+                    pos = self.find_red_laser_remote(frame)
+                except Exception as e:
+                    if self.logger:
+                        self.logger.warning(f"[LASER] 远程识别帧异常: {e}")
+                    pos = None
+                    time.sleep_ms(sample_ms)
+                    continue
+
+                now_ms = time.ticks_ms()
+                if pos is None:
+                    miss_count += 1
+                    samples.clear()
+                    stable_hit_count = 0
+                    if miss_count == 1 or miss_count % 40 == 0:
+                        if self.logger:
+                            self.logger.info(
+                                f"[LASER-REMOTE] 本帧未检出激光点（累计 {miss_count} 帧），"
+                                f"全图多策略搜索中…"
+                            )
+                    time.sleep_ms(sample_ms)
+                    continue
+
+                miss_count = 0
+                x, y = float(pos[0]), float(pos[1])
+                samples.append((now_ms, x, y))
+                cutoff = now_ms - stable_ms
+                samples = [(t, px, py) for t, px, py in samples if t >= cutoff]
+
+                if len(samples) < 2:
+                    time.sleep_ms(sample_ms)
+                    continue
+
+                xs = [s[1] for s in samples]
+                ys = [s[2] for s in samples]
+                span = max(
+                    max(xs) - min(xs),
+                    max(ys) - min(ys),
+                )
+                for i in range(len(samples)):
+                    for j in range(i + 1, len(samples)):
+                        d = math.hypot(
+                            samples[i][1] - samples[j][1],
+                            samples[i][2] - samples[j][2],
+                        )
+                        span = max(span, d)
+
+                if span > max_move:
+                    if self.logger:
+                        self.logger.debug(
+                            f"[LASER] 检测到大幅位移 span={span:.1f}px>{max_move}，重新计时"
+                        )
+                    samples.clear()
+                    stable_hit_count = 0
+                    time.sleep_ms(sample_ms)
+                    continue
+
+                window_ms = samples[-1][0] - samples[0][0]
+                if window_ms >= stable_ms and len(samples) >= min_samples:
+                    fx = int(round(sum(xs) / len(xs)))
+                    fy = int(round(sum(ys) / len(ys)))
+                    stable_hit_count += 1
+                    if self.logger:
+                        self.logger.info(
+                            f"[LASER] 远程识别稳定命中 {stable_hit_count}/3 span={span:.1f}px → ({fx}, {fy})"
+                        )
+                    samples.clear()
+                    if stable_hit_count >= 3 and not reported:
+                        reported = True
+                        self._set_remote_detect_result(
+                            {"result":"laser_detect_ok", "x": fx, "y": fy}
+                        )
+                        if self.logger:
+                            self.logger.info(
+                                f"[LASER] 已连续3次坐标稳定，完成上报，继续等待 cmd=201 关闭会话"
+                            )
+                    time.sleep_ms(sample_ms)
+                    continue
+
+                time.sleep_ms(sample_ms)
+        finally:
+            if self.logger:
+                self.logger.info("[LASER] 远程识别线程退出，等待下一次 cmd=200")
+            with self._remote_detect_lock:
+                self._remote_detect_active = False
+
     def save_laser_point(self, point):
         """保存激光中心点到配置文件
         如果启用硬编码模式，则不保存（直接返回 True）
@@ -831,6 +1065,172 @@ class LaserManager:
             # 使用原来的最亮点方法
             return self._find_red_laser_brightest(frame, threshold, search_radius, ellipse_params)
 
+    def find_red_laser_remote(self, frame):
+        """
+        cmd=200 远程识别专用：全图搜索、多策略、放宽阈值，不限距画面中心距离。
+        常规 find_red_laser 仅搜中心 ±LASER_SEARCH_RADIUS 且距中心 >50px 会丢弃。
+        """
+        import cv2
+        import numpy as np
+        from maix import image
+
+        img_cv = image.image2cv(frame, False, False)
+        if img_cv is None or img_cv.size == 0:
+            return None
+        h, w = img_cv.shape[:2]
+
+        r = img_cv[:, :, 0].astype(np.int32)
+        g = img_cv[:, :, 1].astype(np.int32)
+        b = img_cv[:, :, 2].astype(np.int32)
+        brightness = r + g + b
+        red_ratio = float(getattr(config, "LASER_RED_RATIO", 1.5))
+        ratio_lo = max(1.15, red_ratio - 0.35)
+
+        strategies = []
+        base_th = int(getattr(config, "LASER_DETECTION_THRESHOLD", 140))
+        for th in (base_th, 120, 100, 80, 60):
+            mask = (
+                (r > th)
+                & (r > g * ratio_lo)
+                & (r > b * ratio_lo)
+            )
+            strategies.append(("rgb", th, mask))
+
+        oe_th = int(getattr(config, "LASER_OVEREXPOSED_THRESHOLD", 200))
+        oe_diff = int(getattr(config, "LASER_OVEREXPOSED_DIFF", 10))
+        mask_oe = (
+            (r > oe_th - 30)
+            & (g > oe_th - 40)
+            & (b > oe_th - 40)
+            & (r >= g)
+            & (r >= b)
+            & ((r - g) > max(5, oe_diff - 5))
+            & ((r - b) > max(5, oe_diff - 5))
+        )
+        strategies.append(("overexposed", oe_th, mask_oe))
+
+        mask_bright = (brightness > 380) & (r >= g) & (r >= b) & ((r - g) > 3)
+        strategies.append(("bright", 0, mask_bright))
+
+        hsv = cv2.cvtColor(img_cv, cv2.COLOR_RGB2HSV)
+        hc, sc, vc = cv2.split(hsv)
+        mask_hsv = ((hc <= 18) | (hc >= 162)) & (sc >= 60) & (vc >= 60)
+        strategies.append(("hsv", 0, mask_hsv))
+
+        best_pos = None
+        best_score = -1.0
+        best_tag = None
+
+        max_area = float(getattr(config, "LASER_REMOTE_MAX_AREA", 300.0))
+        min_circularity = float(getattr(config, "LASER_REMOTE_MIN_CIRCULARITY", 0.25))
+
+        for name, th, mask in strategies:
+            m = (mask.astype(np.uint8)) * 255
+            if cv2.countNonZero(m) == 0:
+                continue
+            contours, _ = cv2.findContours(m, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+            if not contours:
+                continue
+            for cnt in contours:
+                area = cv2.contourArea(cnt)
+                if area < 1.5 or area > max_area:
+                    continue
+                peri = cv2.arcLength(cnt, True)
+                if peri <= 0:
+                    continue
+                circularity = float(4.0 * math.pi * area / (peri * peri))
+                if circularity < min_circularity:
+                    continue
+                M = cv2.moments(cnt)
+                if M["m00"] <= 0:
+                    continue
+                cx = float(M["m10"] / M["m00"])
+                cy = float(M["m01"] / M["m00"])
+                ix, iy = int(round(cx)), int(round(cy))
+                if ix < 0 or iy < 0 or ix >= w or iy >= h:
+                    continue
+                local_r = float(r[iy, ix])
+                score = area * local_r * (1.0 + local_r / 255.0) * (0.5 + circularity)
+                if score > best_score:
+                    best_score = score
+                    best_pos = (ix, iy)
+                    best_tag = (name, th, area)
+
+        if best_pos is not None:
+            with self._remote_detect_lock:
+                self._remote_detect_last_pos = best_pos
+            self._save_remote_detect_debug_image(frame, best_pos, best_tag)
+            if self.logger:
+                self.logger.info(
+                    f"[LASER-REMOTE] 检测到激光点 {best_pos} "
+                    f"strategy={best_tag[0]} th={best_tag[1]} area={best_tag[2]:.1f}"
+                )
+        elif self.logger:
+            self.logger.debug("[LASER-REMOTE] 未通过面积/圆度过滤")
+        return best_pos
+
+    def _save_remote_detect_debug_image(self, frame, pos, tag=None):
+        """保存远程识别调试图：叠加激光坐标并落盘。"""
+        try:
+            if not bool(getattr(config, "SAVE_IMAGE_ENABLED", True)):
+                return
+            import cv2
+            from maix import image
+
+            img_cv = image.image2cv(frame, False, False)
+            if img_cv is None or img_cv.size == 0:
+                return
+
+            x, y = int(pos[0]), int(pos[1])
+            h, w = img_cv.shape[:2]
+            if x < 0 or y < 0 or x >= w or y >= h:
+                return
+
+            cv2.circle(img_cv, (x, y), 8, (255, 0, 0), 2)
+            cv2.line(img_cv, (x - 12, y), (x + 12, y), (255, 0, 0), 1)
+            cv2.line(img_cv, (x, y - 12), (x, y + 12), (255, 0, 0), 1)
+
+            desc = ""
+            if tag:
+                desc = f" {tag[0]} th={tag[1]} area={tag[2]:.1f}"
+            cv2.putText(
+                img_cv,
+                f"laser=({x},{y}){desc}",
+                (10, 24),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.55,
+                (255, 0, 0),
+                1,
+                cv2.LINE_AA,
+            )
+
+            base_dir = getattr(config, "PHOTO_DIR", "/root/phot")
+            debug_dir = f"{base_dir}/laser_remote"
+            try:
+                if debug_dir not in os.listdir("/root") and "/" not in debug_dir.replace("/root/", ""):
+                    os.mkdir(debug_dir)
+                else:
+                    try:
+                        os.makedirs(debug_dir, exist_ok=True)
+                    except Exception:
+                        pass
+            except Exception:
+                try:
+                    os.makedirs(debug_dir, exist_ok=True)
+                except Exception:
+                    return
+
+            ts = int(time.ticks_ms())
+            filename = f"{debug_dir}/remote_{x}_{y}_{ts}.jpg"
+            out = image.cv2image(img_cv, False, False)
+            out.save(filename)
+
+            if self.logger:
+                self.logger.info(f"[LASER-REMOTE] 调试图已保存: {filename}")
+        except Exception as e:
+            if self.logger:
+                self.logger.warning(f"[LASER-REMOTE] 保存调试图失败: {e}")
+
     def calibrate_laser_position(self, timeout_ms=8000, check_sharpness=True):
         """
         执行激光校准：循环拍照 → 检测靶心 → 检查激光点清晰度 → 找红点 → 保存坐标

main.py

@@ -402,7 +402,14 @@ def cmd_str():
             else:
                 if config.SHOW_CAMERA_PHOTO_WHILE_SHOOTING:
                     try:
-                        camera_manager.show(camera_manager.read_frame())
+                        frame = camera_manager.read_frame()
+                        laser_manager.remote_detect_tick(frame)
+                        if (
+                            laser_manager.remote_detect_active
+                            and getattr(config, "LASER_REMOTE_DETECT_DRAW_PREVIEW", False)
+                        ):
+                            frame = laser_manager.overlay_remote_detect_preview(frame)
+                        camera_manager.show(frame)
                     except Exception as e:
                         logger = logger_manager.logger
                         if logger:

model_270820.mud

@@ -1,13 +0,0 @@
-
-[basic]
-type = cvimodel
-model = model_270820.cvimodel
-
-[extra]
-model_type = yolov5
-input_type = rgb
-mean = 0, 0, 0
-scale = 0.00392156862745098, 0.00392156862745098, 0.00392156862745098
-anchors = 10, 13, 16, 30, 33, 23, 30, 61, 62, 45, 59, 119, 116, 90, 156, 198, 373, 326
-labels = triangle
-

network.py

@@ -403,7 +403,7 @@ class NetworkManager:
         """
         marker_path = os.path.join(config.APP_DIR, "iccid")
         # if os.path.exists(marker_path):
-        #     return False #
+        #     return False #让iccid一直提交
         iccid_val = self.get_4g_mccid()
         login_data["iccid"] = iccid_val if iccid_val is not None else ""
         return True
@@ -1879,6 +1879,7 @@ class NetworkManager:
                                         from laser_manager import laser_manager
                                         laser_manager.turn_off_laser()
                                         laser_manager.stop_calibration()
+                                        laser_manager.stop_remote_laser_detect()
                                         hardware_manager.start_idle_timer() # 开表
                                         self.safe_enqueue({"result": "laser_off"}, 2)
                                     elif inner_cmd == 4:  # 上报电量
@@ -1955,6 +1956,81 @@ class NetworkManager:
                                         mccid = self.get_4g_mccid()
                                         self.logger.info(f"4G MCCID: {mccid}")
                                         self.safe_enqueue({"result": "mccid", "mccid": mccid if mccid is not None else ""}, 2)
+                                    elif inner_cmd == 200:  # 远程激光点识别：稳定 3s 后上报 (x,y)
+                                        from laser_manager import laser_manager
+                                        # 远程激光识别期间不能停掉心跳/空闲计时，否则会影响数据上报与连接保持
+                                        # 这里仅启动远程识别，不停止网络侧心跳。
+                                        try:
+                                            laser_manager.turn_on_laser()
+                                            if self.logger:
+                                                self.logger.info("[LASER] cmd200 已发送开激光指令")
+                                        except Exception as e:
+                                            if self.logger:
+                                                self.logger.warning(
+                                                    f"[LASER] cmd200 开激光异常: {e}"
+                                                )
+                                        if not laser_manager.start_remote_laser_detect():
+                                            self.safe_enqueue(
+                                                {
+                                                    "cmd": 200,
+                                                    "result": "laser_detect_busy",
+                                                },
+                                                2,
+                                            )
+                                        else:
+                                            self.safe_enqueue(
+                                                {
+                                                    "cmd": 200,
+                                                    "result": "laser_detect_started",
+                                                },
+                                                2,
+                                            )
+                                    elif inner_cmd == 201:  # 设置硬编码激光点并结束远程识别会话
+                                        from laser_manager import laser_manager
+                                        laser_manager.stop_remote_laser_detect()
+                                        inner_data = (
+                                            data_obj.get("data", {})
+                                            if isinstance(data_obj.get("data"), dict)
+                                            else {}
+                                        )
+                                        raw_x = data_obj.get("x", inner_data.get("x"))
+                                        raw_y = data_obj.get("y", inner_data.get("y"))
+                                        try:
+                                            ix, iy = laser_manager.set_hardcoded_laser_point(
+                                                raw_x, raw_y
+                                            )
+                                            self.safe_enqueue(
+                                                {
+                                                    "cmd": 201,
+                                                    "result": "laser_point_set",
+                                                    "x": ix,
+                                                    "y": iy,
+                                                },
+                                                2,
+                                            )
+                                            self.logger.info(
+                                                f"[LASER] cmd201 硬编码激光点=({ix}, {iy})"
+                                            )
+                                        except Exception as e:
+                                            self.logger.error(f"[LASER] cmd201 失败: {e}")
+                                            self.safe_enqueue(
+                                                {
+                                                    "cmd": 201,
+                                                    "result": "laser_point_set_failed",
+                                                    "reason": str(e),
+                                                },
+                                                2,
+                                            )
+                                        hardware_manager.start_idle_timer()
+                                    elif inner_cmd == 46:  # 开关射箭原图保存
+                                        inner_data = data_obj.get("data", {}) if isinstance(data_obj, dict) else {}
+                                        enabled = True
+                                        if isinstance(inner_data, dict) and "enable" in inner_data:
+                                            enabled = bool(inner_data.get("enable"))
+                                        config.SAVE_RAW_SHOT_IMAGE_ENABLED = enabled
+                                        self.logger.info(f"[RAW_IMAGE] 射箭原图保存开关: {enabled}")
+                                        self.safe_enqueue({"result": "raw_image_save", "enabled": enabled}, 2)
+                                        hardware_manager.start_idle_timer()  # 重新计时
                                     elif inner_cmd == 41:
                                         self.logger.info(f"[TEST] 收到TCP射箭触发命令, {time.time()}")
                                         self._manual_trigger_flag = True
@@ -2042,7 +2118,7 @@ class NetworkManager:
                                     pass
                                 break
 
-                    # 发送激光校准结果
+                    # 发送激光校准结果（cmd=2 等传统校准）
                     if logged_in:
                         from laser_manager import laser_manager
                         result = laser_manager.get_calibration_result()
@@ -2050,6 +2126,21 @@ class NetworkManager:
                             x, y = result
                             self.safe_enqueue({"result": "ok", "x": x, "y": y}, 2)
 
+                    # 发送远程激光识别结果（cmd=200，会话持续至 cmd=201）
+                    if logged_in:
+                        from laser_manager import laser_manager
+                        rd = laser_manager.get_remote_detect_result()
+                        if rd and isinstance(rd, dict) and rd.get("status") == "ok":
+                            self.safe_enqueue(
+                                {
+                                    "cmd": 200,
+                                    "result": "laser_detect_ok",
+                                    "x": rd.get("x"),
+                                    "y": rd.get("y"),
+                                },
+                                2,
+                            )
+
                     # 定期发送心跳
                     current_time = time.ticks_ms()
                     if logged_in and current_time - last_heartbeat_send_time > config.HEARTBEAT_INTERVAL * 1000:

ota_manager.py

@@ -770,12 +770,12 @@ class OTAManager:
         # 很多 ML307R 的 MHTTP 对 https 不稳定；对已知域名做降级
 
         if isinstance(url, str) and url.startswith("https://static.shelingxingqiu.com/"):
-            base_url = "https://static.shelingxingqiu.com"
-            # TODO：使用https，看看是否能成功
-            self._is_https = True
+            base_url = "http://static.shelingxingqiu.com"
+            self._is_https = False
         else:
             base_url = f"http://{host}"
             self._is_https = False
+        self.logger.info(f"base_url: {base_url}, self._is_https: {self._is_https}")
         # logger removed - use self.logger instead
 
         def _log(*a):
@@ -1160,8 +1160,8 @@ class OTAManager:
                 self.logger.error(f"[OTA-4G][PWR] before_urc read_failed: {e}")
 
             t_dl0 = time.ticks_ms()
-            success, msg = self.download_file_via_4g(ota_url, downloaded_filename, debug=False)
-            t_dl_cost = time.ticks_diff(t_dl0, time.ticks_ms())
+            success, msg = self.download_file_via_4g(ota_url, downloaded_filename, debug=True)
+            t_dl_cost = time.ticks_diff(time.ticks_ms(), t_dl0)
             self.logger.info(f"[OTA-4G] {msg}")
             self.logger.info(f"[OTA-4G] download_cost_ms={t_dl_cost}")
 

shoot_manager.py

@@ -8,7 +8,7 @@ from laser_manager import laser_manager
 from logger_manager import logger_manager
 from network import network_manager
 from triangle_target import load_camera_from_xml, load_triangle_positions, try_triangle_scoring
-from vision import estimate_distance, detect_circle_v3, enqueue_save_shot
+from vision import estimate_distance, detect_circle_v3, enqueue_save_shot, enqueue_save_raw_shot
 from maix import image, time
 
 # 缓存相机标定与三角形位置，避免每次射箭重复读磁盘
@@ -58,6 +58,7 @@ def analyze_shot(frame, laser_point=None):
     # ── Step 1: 确定激光点 ────────────────────────────────────────────────────
     laser_point_method = None
     distance_m_first = None
+    best_radius1_temp = None
 
     if config.HARDCODE_LASER_POINT:
         laser_point = laser_manager.laser_point
@@ -102,9 +103,22 @@ def analyze_shot(frame, laser_point=None):
         r_img, center, radius, method, best_radius1, ellipse_params = cdata
         dx, dy = None, None
         d_m = distance_m_first
+        tri_h = None
         if center and radius:
             dx, dy = laser_manager.compute_laser_position(center, (x, y), radius, method)
             d_m = estimate_distance(best_radius1) if best_radius1 else distance_m_first
+            try:
+                import numpy as _np
+                px_per_cm = float(radius) / 10.0
+                if px_per_cm > 1e-6:
+                    cxp, cyp = float(center[0]), float(center[1])
+                    tri_h = _np.array([
+                        [1.0 / px_per_cm, 0.0, -cxp / px_per_cm],
+                        [0.0, 1.0 / px_per_cm, -cyp / px_per_cm],
+                        [0.0, 0.0, 1.0],
+                    ], dtype=float)
+            except Exception:
+                tri_h = None
         out = {
             "success": True,
             "result_img": r_img,
@@ -114,6 +128,7 @@ def analyze_shot(frame, laser_point=None):
             "laser_point": laser_point, "laser_point_method": laser_point_method,
             "offset_method": "yellow_ellipse" if ellipse_params else "yellow_circle",
             "distance_method": "yellow_radius",
+            "tri_homography": tri_h,
         }
         if yolo_roi_xyxy is not None:
             out["yolo_roi_xyxy"] = yolo_roi_xyxy
@@ -129,8 +144,10 @@ def analyze_shot(frame, laser_point=None):
     roi_xyxy = None
     yolo_ring_ms = 0.0
     yolo_black_ms = 0.0
+    _timing_on = bool(getattr(config, "ARCHERY_TIMING_ENABLE", True))
+    _sample_on = bool(getattr(config, "TRIANGLE_SAMPLE_ENABLE", False))
     if getattr(config, "TRIANGLE_YOLO_ROI_ENABLE", False):
-        _t_yolo_ring = time_std.perf_counter()
+        _t_yolo_ring = time_std.perf_counter() if _timing_on else None
         try:
             from target_roi_yolo import try_get_triangle_roi_from_yolo
             roi_xyxy = try_get_triangle_roi_from_yolo(
@@ -140,7 +157,8 @@ def analyze_shot(frame, laser_point=None):
             if logger:
                 logger.warning(f"[YOLO-ROI] {e}")
         finally:
-            yolo_ring_ms = (time_std.perf_counter() - _t_yolo_ring) * 1000.0
+            if _timing_on and _t_yolo_ring is not None:
+                yolo_ring_ms = (time_std.perf_counter() - _t_yolo_ring) * 1000.0
 
     _loc_mode = str(
         getattr(config, "TRIANGLE_BLACK_TRIANGLE_LOCATE_MODE", "yolo")
@@ -155,7 +173,7 @@ def analyze_shot(frame, laser_point=None):
         and roi_xyxy is not None
     )
     if _run_stage2_black_yolo:
-        _t_yolo_black = time_std.perf_counter()
+        _t_yolo_black = time_std.perf_counter() if _timing_on else None
         try:
             from target_roi_yolo import try_black_triangle_boxes_work
 
@@ -166,7 +184,8 @@ def analyze_shot(frame, laser_point=None):
             if logger:
                 logger.warning(f"[YOLO-BLACK] {e}")
         finally:
-            yolo_black_ms = (time_std.perf_counter() - _t_yolo_black) * 1000.0
+            if _timing_on and _t_yolo_black is not None:
+                yolo_black_ms = (time_std.perf_counter() - _t_yolo_black) * 1000.0
     elif (
         logger
         and _loc_mode == "traditional"
@@ -184,7 +203,7 @@ def analyze_shot(frame, laser_point=None):
         try:
             logger.info(f"[TRI] begin {datetime.now()}")
             logger.info(f"[TRI] K: {K}, dist: {dist_coef}, pos: {pos},  {datetime.now()}")
-            _t_wall_try = time_std.perf_counter()
+            _t_wall_try = time_std.perf_counter() if _timing_on else None
             tri = try_triangle_scoring(
                 img_cv, (x, y), pos, K, dist_coef,
                 size_range=getattr(config, "TRIANGLE_SIZE_RANGE", (8, 500)),
@@ -193,8 +212,8 @@ def analyze_shot(frame, laser_point=None):
                 yolo_ring_ms=yolo_ring_ms,
                 yolo_black_ms=yolo_black_ms,
             )
-            _wall_try_ms = (time_std.perf_counter() - _t_wall_try) * 1000.0
-            if logger and bool(getattr(config, "TRIANGLE_LOG_E2E_TIMING", True)):
+            _wall_try_ms = (time_std.perf_counter() - _t_wall_try) * 1000.0 if _timing_on else 0.0
+            if logger and bool(getattr(config, "TRIANGLE_LOG_E2E_TIMING", True)) and _timing_on:
                 _e2e = float(yolo_ring_ms) + float(yolo_black_ms) + float(_wall_try_ms)
                 logger.info(
                     f"[TRI] timing_e2e_triangle_ms={_e2e:.1f} "
@@ -280,6 +299,16 @@ def analyze_shot(frame, laser_point=None):
             "tri_markers_completed": tri.get("markers_completed", []),
             "tri_homography": tri.get("homography"),
         }
+        try:
+            import numpy as _np
+            _H = tri.get("homography")
+            if _H is not None and _np.all(_np.isfinite(_H)):
+                _H_inv = _np.linalg.inv(_H)
+                _pt = _np.array([[[0.0, 0.0]]], dtype=_np.float32)
+                _center_pt = cv2.perspectiveTransform(_pt, _H_inv)[0][0]
+                out["tri_center_px"] = [float(_center_pt[0]), float(_center_pt[1])]
+        except Exception:
+            pass
         if yolo_roi_xyxy is not None:
             out["yolo_roi_xyxy"] = yolo_roi_xyxy
         return out
@@ -318,11 +347,22 @@ def process_shot(adc_val):
     :return: None
     """
     logger = logger_manager.logger
+    _timing_on = bool(getattr(config, "ARCHERY_TIMING_ENABLE", True))
 
     try:
         network_manager.safe_enqueue({"shoot_event": "start"}, msg_type=2, high=True)
         frame = camera_manager.read_frame()
 
+        from shot_id_generator import shot_id_generator
+        shot_id = shot_id_generator.generate_id()
+
+        if getattr(config, "SAVE_RAW_SHOT_IMAGE_ENABLED", False):
+            enqueue_save_raw_shot(
+                frame,
+                shot_id=shot_id,
+                photo_dir=config.PHOTO_DIR if config.SAVE_IMAGE_ENABLED else None,
+            )
+
         # 调用算法分析
         analysis_result = analyze_shot(frame)
 
@@ -356,6 +396,107 @@ def process_shot(adc_val):
         )
         x, y = laser_point
 
+        # 物方采样调试（config.TRIANGLE_SAMPLE_ENABLE）：靶心为原点，取两个对称点判断黑白来区分 40/20 标靶
+        # 逻辑：若两个采样点 RGB 均 < 阈值 → 全黑 → 40cm 标靶；否则 → 20cm 标靶
+        sample_target_type = None
+        _t_sample = time_std.perf_counter() if _timing_on else None
+        _t_sample_ms = 0.0
+        sample_points = []
+        sample_patch_half = 2
+        if bool(getattr(config, "TRIANGLE_SAMPLE_ENABLE", False)):
+            sample_obj_radius_cm = float(getattr(config, "TRIANGLE_SAMPLE_RADIUS_CM", 15.0))
+            sample_obj_angles_deg = (0, 180)  # 只取两个对称点：+X 和 -X
+            sample_patch_half = int(getattr(config, "TRIANGLE_SAMPLE_PATCH_HALF_PX", 2))
+            sample_black_thresh = float(getattr(config, "TRIANGLE_SAMPLE_BLACK_THRESH", 30.0))
+            try:
+                import math as _math
+                import numpy as _np
+                import cv2 as _cv2
+
+                if tri_homography is not None:
+                    _H_inv = _np.linalg.inv(tri_homography)
+                    for _ang in sample_obj_angles_deg:
+                        _rad = _math.radians(float(_ang))
+                        _pt_obj = _np.array([
+                            [[sample_obj_radius_cm * _math.cos(_rad), sample_obj_radius_cm * _math.sin(_rad)]]
+                        ], dtype=_np.float32)
+                        _pt_img = _cv2.perspectiveTransform(_pt_obj, _H_inv)[0][0]
+                        _px, _py = float(_pt_img[0]), float(_pt_img[1])
+                        sample_points.append({
+                            "angle_deg": float(_ang),
+                            "obj_cm": (float(sample_obj_radius_cm * _math.cos(_rad)), float(sample_obj_radius_cm * _math.sin(_rad))),
+                            "img_px": (int(round(_px)), int(round(_py))),
+                        })
+                elif center and radius:
+                    _px_per_cm = float(radius) / 10.0
+                    for _ang in sample_obj_angles_deg:
+                        _rad = _math.radians(float(_ang))
+                        _px = float(center[0]) + sample_obj_radius_cm * _math.cos(_rad) * _px_per_cm
+                        _py = float(center[1]) + sample_obj_radius_cm * _math.sin(_rad) * _px_per_cm
+                        sample_points.append({
+                            "angle_deg": float(_ang),
+                            "obj_cm": (float(sample_obj_radius_cm * _math.cos(_rad)), float(sample_obj_radius_cm * _math.sin(_rad))),
+                            "img_px": (int(round(_px)), int(round(_py))),
+                        })
+
+                # 取样后立即读像素并判断黑白：三角成功用 H_inv；三角失败但圆心成功用 center/radius 近似物方半径
+                _all_black = False
+                _sample_infos = []
+                if sample_points:
+                    _img_cv_for_sample = image.image2cv(result_img, False, False)
+                    _all_black = True
+                    for _sp in sample_points:
+                        _sx, _sy = _sp["img_px"]
+                        _hh = max(1, sample_patch_half)
+                        _patch = []
+                        for _yy in range(_sy - _hh, _sy + _hh + 1):
+                            if _yy < 0 or _yy >= _img_cv_for_sample.shape[0]:
+                                continue
+                            for _xx in range(_sx - _hh, _sx + _hh + 1):
+                                if _xx < 0 or _xx >= _img_cv_for_sample.shape[1]:
+                                    continue
+                                _patch.append(_img_cv_for_sample[_yy, _xx].astype(float))
+                        if _patch:
+                            _mean_rgb = _np.mean(_patch, axis=0)
+                            _is_black = bool(_mean_rgb[0] < sample_black_thresh
+                                           and _mean_rgb[1] < sample_black_thresh
+                                           and _mean_rgb[2] < sample_black_thresh)
+                            if not _is_black:
+                                _all_black = False
+                            _sample_infos.append(
+                                f"{int(_sp['angle_deg'])}°@{_sx},{_sy} rgb=({int(_mean_rgb[0])},{int(_mean_rgb[1])},{int(_mean_rgb[2])})"
+                            )
+                    sample_target_type = "40cm_black" if _all_black else "20cm"
+                if _sample_infos:
+                    logger.info("[采样] " + " | ".join(_sample_infos) + f" → {sample_target_type}")
+            except Exception as _e_sample:
+                sample_points = []
+                if logger:
+                    logger.warning(f"[采样] 标靶类型判断失败: {_e_sample}")
+        if _timing_on and _t_sample is not None:
+            _t_sample_ms = (time_std.perf_counter() - _t_sample) * 1000.0
+
+        # 采样提前完成后，先确定靶型对应的物理半径，供后续距离/偏移/上报使用。
+        # 40cm_black 表示直径40cm，半径20cm；20cm 表示直径20cm，半径10cm。
+        target_radius_cm = 20.0 if sample_target_type == "40cm_black" else (10.0 if sample_target_type == "20cm" else 20.0)
+        target_type_value = 40 if sample_target_type == "40cm_black" else (20 if sample_target_type == "20cm" else None)
+
+        # 圆心分支原算法默认按40cm靶半径20cm换算；若采样判定为20cm靶，在上报前修正距离和偏移。
+        # 三角分支使用 triangle_positions.json 的物方坐标，不在这里二次缩放，避免影响三角单应性结果。
+        if sample_target_type == "20cm" and center and radius and not tri_markers:
+            try:
+                distance_m = (target_radius_cm * config.FOCAL_LENGTH_PIX) / float(radius) / 100.0
+                _scale = target_radius_cm / 20.0
+                if dx is not None:
+                    dx = float(dx) * _scale
+                if dy is not None:
+                    dy = float(dy) * _scale
+                if logger:
+                    logger.info(f"[采样] 20cm靶修正圆心测距/偏移: distance={distance_m:.2f}m scale={_scale:.2f}")
+            except Exception as _e_fix:
+                if logger:
+                    logger.warning(f"[采样] 20cm靶修正失败: {_e_fix}")
+
         # 三角形路径成功时 center/radius 为空是正常的；此时用 triangle 方法名用于保存文件名与上报字段 m
         if (not method) and tri_markers:
             method = "triangle_homography"
@@ -366,10 +507,6 @@ def process_shot(adc_val):
         if dx is None and dy is None and logger:
             logger.warning("[MAIN] 未检测到偏移量（三角形与圆形均失败），但会保存图像")
 
-        # 生成射箭ID
-        from shot_id_generator import shot_id_generator
-        shot_id = shot_id_generator.generate_id()
-
         if logger:
             logger.info(f"[MAIN] 射箭ID: {shot_id}")
 
@@ -386,7 +523,7 @@ def process_shot(adc_val):
             "shot_id": shot_id,
             "x": srv_x,
             "y": srv_y,
-            "r": 20.0,  # 保留字段（服务端当前忽略，物理外环半径 cm）
+            "r": target_radius_cm,  # 物理靶半径 cm：40cm靶=20，20cm靶=10
             "d": round((distance_m or 0.0) * 100),
             "d_laser": round((laser_distance_m or 0.0) * 100),
             "d_laser_quality": laser_signal_quality,
@@ -397,6 +534,7 @@ def process_shot(adc_val):
             "target_y": float(y),
             "offset_method": offset_method,
             "distance_method": distance_method,
+            "target_type": target_type_value,
         }
 
         if ellipse_params:
@@ -471,6 +609,11 @@ def process_shot(adc_val):
                     except Exception:
                         pass
 
+                # 物方采样标靶类型判断耗时（合并在上面采样块内，单独统计）
+                if _timing_on and bool(getattr(config, "TRIANGLE_SAMPLE_ENABLE", False)) and sample_target_type is not None:
+                    logger.info(f"[采样] 标靶类型: {sample_target_type}  耗时: {_t_sample_ms:.2f}ms")
+
+
                 # 叠加信息：落点-圆心距离 / 相机-靶距离等
                 try:
                     import math as _math

test/test_algo_preview_live.py

@@ -0,0 +1,403 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+实时摄像头预览：叠加与射箭存图相同的算法标注（YOLO ROI、三角/圆心、激光十字等），默认不写盘。
+
+在 MaixCAM 上从项目根目录运行：
+  python3 test/test_algo_preview_live.py
+  python3 test/test_algo_preview_live.py --interval 1.5
+  python3 test/test_algo_preview_live.py --every-frame
+
+说明：
+  - 完整算法走 shoot_manager.analyze_shot（与 process_shot 一致，含 YOLO + 三角/圆心）。
+  - 画面标注对齐 process_shot 存图前绘制 + vision._draw_yolo_roi_on_rgb_numpy / 圆心存图线。
+  - 预览模式会关闭 Stage2 裁切 JPEG 落盘，避免写满 /root/phot。
+"""
+
+from __future__ import annotations
+
+import argparse
+import math
+import os
+import sys
+import threading
+import time
+
+_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+if _ROOT not in sys.path:
+    sys.path.insert(0, _ROOT)
+
+import cv2
+import numpy as np
+from maix import image, time as maix_time
+
+import config
+from camera_manager import camera_manager
+from laser_manager import laser_manager
+from shoot_manager import analyze_shot, preload_triangle_calib
+from target_roi_yolo import preload_yolo_detector
+from vision import _draw_yolo_roi_on_rgb_numpy
+
+
+def _copy_maix_frame(frame):
+    """相机下一帧可能复用缓冲区，异步分析前先复制。"""
+    img_cv = image.image2cv(frame, False, False)
+    return image.cv2image(np.ascontiguousarray(img_cv), False, False)
+
+
+def _patch_preview_config():
+    """预览不写调试 JPEG，避免刷屏占存储。"""
+    config.TRIANGLE_BLACK_YOLO_SAVE_ROI_CROP = False
+    config.TRIANGLE_SAVE_DEBUG_IMAGE = False
+
+
+def _annotate_like_saved_shot(analysis: dict):
+    """
+    将 analyze_shot 结果绘制成与 process_shot -> enqueue_save_shot 存盘前一致的 Maix 图。
+    """
+    result_img = analysis.get("result_img")
+    if result_img is None:
+        return None
+
+    center = analysis.get("center")
+    radius = analysis.get("radius")
+    method = analysis.get("method")
+    ellipse_params = analysis.get("ellipse_params")
+    laser_point = analysis.get("laser_point")
+    dx = analysis.get("dx")
+    dy = analysis.get("dy")
+    distance_m = analysis.get("distance_m")
+    offset_method = analysis.get("offset_method", "")
+    distance_method = analysis.get("distance_method", "")
+    tri_markers = analysis.get("tri_markers") or []
+    tri_markers_completed = analysis.get("tri_markers_completed") or []
+    tri_homography = analysis.get("tri_homography")
+    yolo_roi_xyxy = analysis.get("yolo_roi_xyxy")
+
+    if laser_point is None:
+        return result_img
+
+    x, y = laser_point
+    draw_yolo_roi = (
+        yolo_roi_xyxy is not None
+        and getattr(config, "TRIANGLE_YOLO_DRAW_ROI_ON_SHOT", True)
+    )
+
+    if tri_markers:
+        img_cv = image.image2cv(result_img, False, False).copy()
+
+        if draw_yolo_roi:
+            _draw_yolo_roi_on_rgb_numpy(img_cv, yolo_roi_xyxy)
+
+        for m in tri_markers:
+            corners = np.array(m["corners"], dtype=np.int32)
+            cv2.polylines(img_cv, [corners], True, (0, 255, 0), 2)
+            cx, cy = int(m["center"][0]), int(m["center"][1])
+            cv2.circle(img_cv, (cx, cy), 4, (0, 0, 255), -1)
+            cv2.putText(
+                img_cv,
+                f"T{m['id']}",
+                (cx - 18, cy - 12),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.55,
+                (0, 255, 0),
+                1,
+            )
+
+        for m in tri_markers_completed:
+            if not m.get("is_virtual"):
+                continue
+            cx, cy = int(m["center"][0]), int(m["center"][1])
+            cv2.circle(img_cv, (cx, cy), 6, (255, 0, 255), 2)
+            cv2.putText(
+                img_cv,
+                f"VT{m['id']}",
+                (cx - 22, cy - 12),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.55,
+                (255, 0, 255),
+                1,
+            )
+
+        if tri_homography is not None:
+            try:
+                H_inv = np.linalg.inv(tri_homography)
+                c_img = cv2.perspectiveTransform(
+                    np.array([[[0.0, 0.0]]], dtype=np.float32), H_inv
+                )[0][0]
+                ocx, ocy = int(c_img[0]), int(c_img[1])
+                cv2.circle(img_cv, (ocx, ocy), 5, (0, 0, 255), -1)
+                cv2.circle(img_cv, (ocx, ocy), 9, (0, 0, 255), 1)
+            except Exception:
+                pass
+
+        lines = []
+        if dx is not None and dy is not None:
+            r_cm = math.hypot(float(dx), float(dy))
+            lines.append(f"offset=({float(dx):.2f},{float(dy):.2f})cm |r|={r_cm:.2f}cm")
+        if distance_m is not None:
+            lines.append(f"cam_dist={float(distance_m):.2f}m ({distance_method})")
+        if method:
+            lines.append(f"method={method} ({offset_method})")
+        y0 = 22
+        for i, t in enumerate(lines):
+            cv2.putText(
+                img_cv,
+                t,
+                (10, y0 + i * 18),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.5,
+                (0, 255, 0),
+                1,
+            )
+
+        out = image.cv2image(img_cv, False, False)
+    else:
+        img_cv = image.image2cv(result_img, False, False).copy()
+        if draw_yolo_roi:
+            _draw_yolo_roi_on_rgb_numpy(img_cv, yolo_roi_xyxy)
+
+        if center and radius:
+            cx, cy = center
+            if ellipse_params:
+                (ell_center, (width, height), angle) = ellipse_params
+                cx_ell, cy_ell = int(ell_center[0]), int(ell_center[1])
+                cv2.ellipse(
+                    img_cv,
+                    (cx_ell, cy_ell),
+                    (int(width / 2), int(height / 2)),
+                    angle,
+                    0,
+                    360,
+                    (0, 255, 0),
+                    2,
+                )
+                cv2.circle(img_cv, (cx_ell, cy_ell), 3, (255, 0, 0), -1)
+                minor_length = min(width, height) / 2
+                minor_angle = angle + 90 if width >= height else angle
+                minor_angle_rad = math.radians(minor_angle)
+                dx_minor = minor_length * math.cos(minor_angle_rad)
+                dy_minor = minor_length * math.sin(minor_angle_rad)
+                pt1 = (int(cx_ell - dx_minor), int(cy_ell - dy_minor))
+                pt2 = (int(cx_ell + dx_minor), int(cy_ell + dy_minor))
+                cv2.line(img_cv, pt1, pt2, (0, 0, 255), 2)
+            else:
+                cv2.circle(img_cv, (int(cx), int(cy)), int(radius), (0, 0, 255), 2)
+                cv2.circle(img_cv, (int(cx), int(cy)), 2, (0, 0, 255), -1)
+            cv2.line(img_cv, (int(x), int(y)), (int(cx), int(cy)), (255, 255, 0), 1)
+
+        lines = []
+        if dx is not None and dy is not None:
+            lines.append(f"offset=({float(dx):.2f},{float(dy):.2f})cm")
+        if distance_m is not None:
+            lines.append(f"dist={float(distance_m):.2f}m ({distance_method})")
+        if method:
+            lines.append(f"method={method}")
+        for i, t in enumerate(lines):
+            cv2.putText(
+                img_cv,
+                t,
+                (10, 22 + i * 18),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.5,
+                (0, 255, 0),
+                1,
+            )
+        out = image.cv2image(img_cv, False, False)
+
+    lc = image.Color(config.LASER_COLOR[0], config.LASER_COLOR[1], config.LASER_COLOR[2])
+    out.draw_line(
+        int(x - config.LASER_LENGTH),
+        int(y),
+        int(x + config.LASER_LENGTH),
+        int(y),
+        lc,
+        config.LASER_THICKNESS,
+    )
+    out.draw_line(
+        int(x),
+        int(y - config.LASER_LENGTH),
+        int(x),
+        int(y + config.LASER_LENGTH),
+        lc,
+        config.LASER_THICKNESS,
+    )
+    out.draw_circle(int(x), int(y), 1, lc, config.LASER_THICKNESS)
+    return out
+
+
+class _AlgoWorker:
+    def __init__(self):
+        self._lock = threading.Lock()
+        self._busy = False
+        self._latest_preview = None
+        self._latest_meta = ""
+        self._last_ms = 0.0
+
+    @property
+    def busy(self):
+        with self._lock:
+            return self._busy
+
+    @property
+    def last_ms(self):
+        with self._lock:
+            return self._last_ms
+
+    def get_preview(self):
+        with self._lock:
+            return self._latest_preview, self._latest_meta
+
+    def run_async(self, frame):
+        with self._lock:
+            if self._busy:
+                return False
+            self._busy = True
+
+        def _job():
+            t0 = time.perf_counter()
+            meta = ""
+            preview = None
+            try:
+                analysis = analyze_shot(frame)
+                if not analysis.get("success"):
+                    reason = analysis.get("reason", "unknown")
+                    meta = f"fail:{reason}"
+                else:
+                    preview = _annotate_like_saved_shot(analysis)
+                    dx, dy = analysis.get("dx"), analysis.get("dy")
+                    method = analysis.get("method") or "?"
+                    if dx is not None and dy is not None:
+                        meta = f"ok {method} ({dx:.2f},{dy:.2f})cm"
+                    else:
+                        meta = f"ok {method} no_offset"
+            except Exception as e:
+                meta = f"err:{e}"
+            elapsed = (time.perf_counter() - t0) * 1000.0
+            with self._lock:
+                self._latest_preview = preview
+                self._latest_meta = f"{meta} {elapsed:.0f}ms"
+                self._last_ms = elapsed
+                self._busy = False
+
+        threading.Thread(target=_job, daemon=True).start()
+        return True
+
+
+def _draw_status(frame, lines, color=None):
+    if color is None:
+        color = image.COLOR_YELLOW
+    y = 4
+    for line in lines:
+        frame.draw_string(4, y, line, color=color)
+        y += 16
+
+
+def _save_preview_jpeg(maix_img, out_dir):
+    os.makedirs(out_dir, exist_ok=True)
+    fn = os.path.join(out_dir, f"preview_{int(time.time() * 1000)}.jpg")
+    maix_img.save(fn)
+    return fn
+
+
+def main():
+    parser = argparse.ArgumentParser(description="实时预览射箭算法存图效果")
+    parser.add_argument(
+        "--interval",
+        type=float,
+        default=2.0,
+        help="两次完整 analyze_shot 的最小间隔（秒）；--every-frame 时忽略",
+    )
+    parser.add_argument(
+        "--every-frame",
+        action="store_true",
+        help="每帧都触发算法（很慢，仅调试用）",
+    )
+    parser.add_argument(
+        "--width",
+        type=int,
+        default=getattr(config, "CAMERA_WIDTH", 640),
+    )
+    parser.add_argument(
+        "--height",
+        type=int,
+        default=getattr(config, "CAMERA_HEIGHT", 480),
+    )
+    parser.add_argument(
+        "--save-dir",
+        default=config.PHOTO_DIR,
+        help="按板子按键无；用 --save-every N 每 N 次成功分析存一张",
+    )
+    parser.add_argument(
+        "--save-every",
+        type=int,
+        default=0,
+        help="每成功分析 N 次自动存一张到 --save-dir（0=不自动存）",
+    )
+    args = parser.parse_args()
+
+    _patch_preview_config()
+    print("[INFO] 预览模式：已关闭 TRIANGLE_BLACK_YOLO_SAVE_ROI_CROP / TRIANGLE_SAVE_DEBUG_IMAGE")
+
+    laser_manager.load_laser_point()
+    preload_triangle_calib()
+    if getattr(config, "TRIANGLE_YOLO_PRELOAD_ON_BOOT", False) or getattr(
+        config, "TRIANGLE_BLACK_YOLO_PRELOAD_ON_BOOT", False
+    ):
+        print("[INFO] 预加载 YOLO …")
+        preload_yolo_detector()
+
+    camera_manager.init_camera(args.width, args.height)
+    camera_manager.init_display()
+    worker = _AlgoWorker()
+
+    interval_s = 0.0 if args.every_frame else max(0.3, float(args.interval))
+    last_trigger = 0.0
+    ok_count = 0
+    frame_idx = 0
+
+    print(
+        f"[INFO] 摄像头 {args.width}x{args.height} "
+        f"interval={'每帧' if args.every_frame else f'{interval_s}s'}"
+    )
+    print("[INFO] 退出：Ctrl+C")
+
+    try:
+        while True:
+            frame = camera_manager.read_frame()
+            frame_idx += 1
+            now = time.perf_counter()
+
+            due = args.every_frame or (now - last_trigger >= interval_s)
+            if due and not worker.busy:
+                last_trigger = now
+                worker.run_async(_copy_maix_frame(frame))
+
+            preview, meta = worker.get_preview()
+            if preview is not None:
+                show_img = preview
+                status = [f"#{frame_idx}", meta]
+                if args.save_every > 0 and meta.startswith("ok"):
+                    ok_count += 1
+                    if ok_count % args.save_every == 0:
+                        try:
+                            fn = _save_preview_jpeg(preview, args.save_dir)
+                            status.append(f"saved:{fn}")
+                        except Exception as e:
+                            status.append(f"save_err:{e}")
+            else:
+                show_img = frame
+                if worker.busy:
+                    status = [f"#{frame_idx}", "analyzing…"]
+                else:
+                    status = [f"#{frame_idx}", "waiting…"]
+
+            _draw_status(show_img, status)
+            camera_manager.show(show_img)
+            maix_time.sleep_ms(1)
+    except KeyboardInterrupt:
+        print("[INFO] 已退出")
+
+
+if __name__ == "__main__":
+    main()

test/test_audio.py

@@ -0,0 +1,50 @@
+# test_audio.pyx
+from maix import audio, time, app, gpio
+
+def run_player_loop():
+    """
+    播放控制主循环函数
+    """
+    # 初始化音频播放器
+    p = audio.Player("/root/gun.wav")
+    p.volume(40)
+    
+    # 初始化 GPIO 引脚为输出
+    led = gpio.GPIO("A25", gpio.Mode.OUT)
+    # 设置低电平
+    led.value(0)
+    
+    # 主循环
+    while not app.need_exit():
+        led.value(1)           # 点亮 LED
+        time.sleep_ms(200)     # 保持 200ms
+        led.value(0)           # 熄灭 LED
+        p.play()               # 播放音频
+        time.sleep_ms(1000)    # 等待 1 秒
+    
+    print("play finish!")
+
+
+# 可选：添加一个简单的测试函数
+def hello():
+    return "Hello from test_audio!"
+
+
+# 可选：添加一个初始化函数
+def init_led():
+    """单独测试 GPIO"""
+    led = gpio.GPIO("A25", gpio.Mode.OUT)
+    led.value(0)
+    return "LED initialized"
+
+
+# 可选：添加一个播放测试函数
+def test_play():
+    """单独测试音频播放"""
+    p = audio.Player("/root/gun.wav")
+    p.volume(50)
+    p.play()
+    return "Playing..."
+
+
+run_player_loop()

test/test_button.py

@@ -0,0 +1,25 @@
+from maix import audio, time, app,gpio
+
+
+# button1 = gpio.GPIO("ADC", gpio.Mode.IN)
+button3 = gpio.GPIO("A26", gpio.Mode.IN) # 可用
+button2 = gpio.GPIO("A16", gpio.Mode.IN)
+#设置低电平
+from maix.peripheral import adc
+channel = 0
+res_bit = adc.RES_BIT_12
+_adc_obj = adc.ADC(channel, res_bit)
+
+
+while not app.need_exit():
+    # print(f"b1: {button1.value()}")
+
+    print(f"b2: {button2.value()}")
+
+    # print(_adc_obj.read_vol())
+    print(f"b3: {button3.value()}")
+    time.sleep_ms(50)
+
+    # time.sleep_ms(1000)
+
+

test/test_cammera.py

@@ -3,7 +3,10 @@ from maix import camera, display, time
 
 try:
     print("Initializing camera...")
-    cam = camera.Camera(640, 480)
+    cam = camera.Camera(640,480)
+    # cam = camera.Camera(1280,720)
+    # cam.get_exposure_us()
+    # print("Camera exposure: ", cam.get_exposure_us())
     print("Camera initialized successfully!")
     
     disp = display.Display()

test/test_laser.py

@@ -1,172 +1,246 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 """
-激光模块测试脚本
-用于诊断激光开关问题
-
-使用方法：
-python test_laser.py
-
-功能：
-1. 初始化串口
-2. 循环测试激光开/关
-3. 打印详细调试信息
+M01激光测距模块测试脚本 - 修正版
+基于文档中的完整命令示例
 """
 
 from maix import uart, pinmap, time
+import binascii
 
 # ==================== 配置 ====================
-UART_PORT = "/dev/ttyS1"   # 激光模块连接的串口（UART1）
-BAUDRATE = 9600            # 波特率
-
-# 引脚映射（确保与硬件连接一致）
-print("=" * 50)
-print("🔧 步骤1: 配置引脚映射")
-print("=" * 50)
+UART_PORT = "/dev/ttyS1"
+BAUDRATE = 9600
 
+# 初始化串口
 try:
     pinmap.set_pin_function("A18", "UART1_RX")
-    print("✅ A18 -> UART1_RX")
-except Exception as e:
-    print(f"❌ A18 配置失败: {e}")
-
-try:
     pinmap.set_pin_function("A19", "UART1_TX")
-    print("✅ A19 -> UART1_TX")
-except Exception as e:
-    print(f"❌ A19 配置失败: {e}")
-
-# ==================== 激光控制指令 ====================
-MODULE_ADDR = 0x00
-
-# 原始命令
-LASER_ON_CMD = bytes([0xAA, MODULE_ADDR, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x01, 0xC1])
-LASER_OFF_CMD = bytes([0xAA, MODULE_ADDR, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x00, 0xC0])
-
-# 备用命令格式（如果原始命令不工作，可以尝试这些）
-# 格式1: 简化命令
-LASER_ON_CMD_ALT1 = bytes([0xAA, 0x01, 0x01])
-LASER_OFF_CMD_ALT1 = bytes([0xAA, 0x01, 0x00])
-
-# 格式2: 不同的协议头
-LASER_ON_CMD_ALT2 = bytes([0x55, 0xAA, 0x01])
-LASER_OFF_CMD_ALT2 = bytes([0x55, 0xAA, 0x00])
-
-print("\n" + "=" * 50)
-print("🔧 步骤2: 初始化串口")
-print("=" * 50)
-print(f"设备: {UART_PORT}")
-print(f"波特率: {BAUDRATE}")
-
-try:
     laser_uart = uart.UART(UART_PORT, BAUDRATE)
-    print(f"✅ 串口初始化成功: {laser_uart}")
+    print("✅ 硬件初始化完成")
 except Exception as e:
-    print(f"❌ 串口初始化失败: {e}")
+    print(f"❌ 初始化失败: {e}")
     exit(1)
 
-# ==================== 测试函数 ====================
-def send_and_check(cmd, name):
-    """发送命令并检查回包"""
-    print(f"\n📤 发送: {name}")
-    print(f"   命令字节: {cmd.hex()}")
-    print(f"   命令长度: {len(cmd)} 字节")
+# ==================== 根据文档的完整命令集 ====================
+# 1. 激光开关（文档2.3.10，已验证可用）
+LASER_ON_CMD = bytes([0xAA, 0x00, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x01, 0xC1])
+LASER_OFF_CMD = bytes([0xAA, 0x00, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x00, 0xC0])
 
-    # 清空接收缓冲区
+# 2. 尝试不同的测距命令格式
+TEST_COMMANDS = [
+    # 格式1：文档2.3.12的单次测量（您测试失败的）
+    {
+        "name": "单次测量 (0x0020)",
+        "cmd": bytes([0xAA, 0x00, 0x00, 0x20, 0x00, 0x01, 0x00, 0x00, 0x21]),
+        "desc": "文档2.3.12 示例命令"
+    },
+    # 格式2：文档2.3.7的读取测量结果
+    {
+        "name": "读取测量结果 (0x0022)",
+        "cmd": bytes([0xAA, 0x80, 0x00, 0x22, 0xA2]),
+        "desc": "文档2.3.7 读取测量结果"
+    },
+    # 格式3：文档2.3.13的快速测量
+    {
+        "name": "快速测量 (0x0022带数据)",
+        "cmd": bytes([0xAA, 0x00, 0x00, 0x22, 0x00, 0x01, 0x00, 0x00, 0x23]),
+        "desc": "文档2.3.13 快速测量"
+    },
+    # 格式4：连续测量模式
+    {
+        "name": "连续测量模式 (0x0021)",
+        "cmd": bytes([0xAA, 0x00, 0x00, 0x21, 0x00, 0x01, 0x00, 0x00, 0x22]),
+        "desc": "文档2.3.14 连续测量"
+    }
+]
+
+def clear_buffer():
+    """清空串口缓冲区"""
     try:
-        old_data = laser_uart.read(-1)
-        if old_data:
-            print(f"   清空缓冲区: {len(old_data)} 字节")
+        data = laser_uart.read(-1)
+        if data:
+            print(f"清空: {len(data)}字节")
     except:
         pass
 
-    # 发送命令
+def send_and_wait(cmd, name, wait_time=2000):
+    """发送命令并等待响应"""
+    print(f"\n📤 发送: {name}")
+    print(f"   命令: {cmd.hex()}")
+    
+    clear_buffer()
+    
     try:
-        written = laser_uart.write(cmd)
-        print(f"   写入字节数: {written}")
+        laser_uart.write(cmd)
+        print(f"   已发送 {len(cmd)} 字节")
     except Exception as e:
-        print(f"   ❌ 写入失败: {e}")
+        print(f"   ❌ 发送失败: {e}")
         return None
     
     # 等待响应
-    time.sleep_ms(100)
+    start_time = time.ticks_ms()
+    response = b""
     
-    # 读取回包
-    try:
-        resp = laser_uart.read(50)
-        if resp:
-            print(f"   📥 收到回包: {resp.hex()} ({len(resp)} 字节)")
-            return resp
-        else:
-            print(f"   ⚠️ 无回包")
-            return None
-    except Exception as e:
-        print(f"   ❌ 读取失败: {e}")
+    while time.ticks_ms() - start_time < wait_time:
+        try:
+            chunk = laser_uart.read(1)
+            if chunk:
+                response += chunk
+                # 完整响应通常是9或13字节
+                if len(response) >= 9:
+                    # 检查是否完整帧
+                    if response[0] in [0xAA, 0xEE]:
+                        if len(response) >= 13:  # 测距完整响应
+                            break
+                        elif response[0] == 0xEE:  # 错误响应
+                            break
+        except:
+            break
+        
+        time.sleep_ms(10)
+    
+    if response:
+        print(f"   📥 响应: {response.hex()}")
+        print(f"   长度: {len(response)} 字节")
+        
+        # 解析错误码
+        if response[0] == 0xEE and len(response) >= 9:
+            err_code = (response[7] << 8) | response[8]
+            error_mapping = {
+                0x0000: "无错误",
+                0x0001: "硬件错误",
+                0x0002: "无输出数据", 
+                0x0003: "反射信号太弱",
+                0x0004: "反射信号太强",
+                0x0005: "温度太高(>40℃)",
+                0x0006: "温度太低(<-10℃)",
+                0x0007: "电源电压低(<2.5V)",
+                0x0008: "超出量程",
+                0x0009: "读通讯错误",
+                0x000A: "写通讯错误",
+                0x000B: "地址错误"
+            }
+            err_msg = error_mapping.get(err_code, f"未知错误: 0x{err_code:04X}")
+            print(f"   ❌ 模块错误: {err_msg}")
+    else:
+        print("   ⚠️ 无响应")
+    
+    return response
+
+def parse_distance_data(response):
+    """解析距离数据"""
+    if not response or len(response) < 13:
         return None
     
-def test_laser_cycle(on_cmd, off_cmd, cmd_name="标准命令"):
-    """测试一个开关周期"""
-    print(f"\n{'='*50}")
-    print(f"🧪 测试 {cmd_name}")
-    print(f"{'='*50}")
+    if response[0] != 0xAA or response[3] not in [0x20, 0x21, 0x22]:
+        return None
     
-    print("\n>>> 测试开启激光")
-    send_and_check(on_cmd, f"{cmd_name} - 开启")
-    print("   ⏱️ 等待 2 秒观察激光是否亮起...")
-    time.sleep(2)
+    # 解析4字节BCD码
+    bcd_bytes = response[6:10]
+    distance_int = 0
     
-    print("\n>>> 测试关闭激光")
-    send_and_check(off_cmd, f"{cmd_name} - 关闭")
-    print("   ⏱️ 等待 2 秒观察激光是否熄灭...")
-    time.sleep(2)
+    for byte in bcd_bytes:
+        high = (byte >> 4) & 0x0F
+        low = byte & 0x0F
+        
+        if high > 9 or low > 9:
+            return None
+        
+        distance_int = distance_int * 100 + high * 10 + low
+    
+    distance_m = distance_int / 1000.0
+    
+    # 信号质量
+    signal = 0
+    if len(response) >= 12:
+        signal = (response[10] << 8) | response[11]
+    
+    return {
+        'meters': distance_m,
+        'millimeters': distance_m * 1000,
+        'signal': signal,
+        'raw': response.hex()
+    }
 
 # ==================== 主测试 ====================
-print("\n" + "=" * 50)
-print("🚀 开始激光测试")
-print("=" * 50)
-print("\n请观察激光模块的状态变化...")
-print("测试将依次尝试不同的命令格式\n")
+print("\n" + "="*50)
+print("M01激光测距模块详细测试")
+print("="*50)
 
 try:
-    # 测试1: 标准命令
-    test_laser_cycle(LASER_ON_CMD, LASER_OFF_CMD, "标准命令")
+    # 1. 测试基本连接
+    print("\n1. 测试模块连接...")
+    version_cmd = bytes([0xAA, 0x80, 0x00, 0x0A, 0x8A])
+    resp = send_and_wait(version_cmd, "读取硬件版本")
     
-    input("\n按回车继续测试备用命令1...")
+    if resp and resp[0] == 0xAA and resp[3] == 0x0A:
+        print(f"✅ 模块正常，版本: {resp[6]:02X}{resp[7]:02X}")
+    else:
+        print("❌ 模块连接测试失败")
+        exit(1)
     
-    # 测试2: 备用命令格式1
-    test_laser_cycle(LASER_ON_CMD_ALT1, LASER_OFF_CMD_ALT1, "备用命令1 (简化)")
+    # 2. 开启激光
+    print("\n2. 开启激光...")
+    resp = send_and_wait(LASER_ON_CMD, "开启激光", 1000)
+    if resp and resp.hex() == "aa0001be00010001c1":
+        print("✅ 激光已开启")
     
-    input("\n按回车继续测试备用命令2...")
+    print("   等待激光稳定...")
+    time.sleep(2)  # 重要等待时间
     
-    # 测试3: 备用命令格式2
-    test_laser_cycle(LASER_ON_CMD_ALT2, LASER_OFF_CMD_ALT2, "备用命令2 (0x55AA头)")
+    # 3. 尝试不同的测距命令
+    print("\n3. 测试不同测距命令...")
     
-    print("\n" + "=" * 50)
+    for i, test_cmd in enumerate(TEST_COMMANDS):
+        print(f"\n{'='*30}")
+        print(f"测试 {i+1}: {test_cmd['name']}")
+        print(f"{test_cmd['desc']}")
+        print(f"{'='*30}")
+        
+        resp = send_and_wait(test_cmd['cmd'], test_cmd['name'], 3000)
+        
+        if resp:
+            if resp[0] == 0xAA and len(resp) >= 13:
+                result = parse_distance_data(resp)
+                if result:
+                    print(f"✅ 测距成功!")
+                    print(f"   距离: {result['meters']:.3f} m")
+                    print(f"   距离: {result['millimeters']:.1f} mm")
+                    print(f"   信号质量: {result['signal']}")
+                    break
+                else:
+                    print("❌ 无法解析距离数据")
+            elif resp[0] == 0xEE:
+                print("❌ 命令执行错误")
+            else:
+                print("❌ 无效响应格式")
+        else:
+            print("❌ 无响应")
+        
+        time.sleep(1)  # 命令间间隔
+    
+    # 4. 关闭激光
+    print("\n4. 关闭激光...")
+    send_and_wait(LASER_OFF_CMD, "关闭激光", 1000)
+    
+    print("\n" + "="*50)
     print("🏁 测试完成")
-    print("=" * 50)
-    print("\n诊断建议：")
-    print("1. 如果激光始终不亮/始终亮：")
-    print("   - 检查激光模块的电源连接")
-    print("   - 检查串口TX/RX是否接反")
-    print("   - 尝试不同的波特率 (4800/19200)")
-    print("")
-    print("2. 如果有回包但激光无反应：")
-    print("   - 命令格式可能正确但激光硬件问题")
-    print("")
-    print("3. 如果某个备用命令有效：")
-    print("   - 需要更新 config.py 中的命令格式")
+    print("="*50)
+    
+    print("\n📋 测试总结：")
+    print("1. 模块通信: ✅ 正常")
+    print("2. 激光控制: ✅ 正常")
+    print("3. 测距功能: ❌ 有问题")
+    print("\n建议：")
+    print("1. 检查激光是否实际发光（在暗处观察红点）")
+    print("2. 确保测量目标在有效范围内（0.2-60米）")
+    print("3. 确保目标有足够反射率（白色平面最佳）")
+    print("4. 如果所有测距命令都返回ERR_ADDR，可能是固件版本问题")
     
 except KeyboardInterrupt:
-    print("\n\n🛑 测试被中断")
-    # 确保激光关闭
+    print("\n\n🛑 用户中断")
     laser_uart.write(LASER_OFF_CMD)
     print("✅ 已发送关闭指令")
 except Exception as e:
     print(f"\n❌ 测试出错: {e}")
-    import traceback
-    traceback.print_exc()
-
-
-
-
-

test/test_laser_center_point.py

@@ -0,0 +1,541 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+激光中心点检测单元测试（单文件，无项目依赖）
+直接使用 maix 标准库，实现红色激光点坐标检测
+
+运行方式：
+  python3 test/test_laser_center_point.py
+
+Ctrl+C 退出，按 s 保存截图
+"""
+
+from maix import camera, display, image, time, app, uart, pinmap
+import os
+import struct
+import select
+
+_USE_CV = False
+try:
+    import cv2
+    import numpy as np
+    _USE_CV = True
+except ImportError:
+    pass
+
+WIDTH = 640
+HEIGHT = 480
+THRESHOLD = 140
+SEARCH_RADIUS = 50
+
+
+def read_key_ev():
+    """非阻塞读取 /dev/input/event0 按键（返回 key_code 或 -1）"""
+    try:
+        r, _, _ = select.select([_key_fd], [], [], 0)
+        if r:
+            event = _key_fd.read(16)
+            if len(event) == 16:
+                _, _, etype, code, value = struct.unpack("IIHHI", event)
+                if etype == 1 and value == 1:
+                    return code
+    except Exception:
+        pass
+    return -1
+
+
+def find_ellipse(img_cv, cx, cy, roi_r, th):
+    x1 = max(0, cx - roi_r)
+    x2 = min(WIDTH, cx + roi_r)
+    y1 = max(0, cy - roi_r)
+    y2 = min(HEIGHT, cy + roi_r)
+    roi = img_cv[y1:y2, x1:x2]
+    if roi.size == 0:
+        return None
+    r = roi[:, :, 0].astype(np.int32)
+    g = roi[:, :, 1].astype(np.int32)
+    b = roi[:, :, 2].astype(np.int32)
+    mask = (r > th) & (r > g * 1.5) & (r > b * 1.5)
+    oe = (r > 200) & (g > 200) & (b > 200) & (r >= g) & (r >= b) & ((r - g) > 10) & ((r - b) > 10)
+    combined = (mask | oe).astype(np.uint8) * 255
+    contours, _ = cv2.findContours(combined, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+    if not contours:
+        return None
+    largest = max(contours, key=cv2.contourArea)
+    if cv2.contourArea(largest) < 5:
+        return None
+    cnt = largest.copy()
+    for pt in cnt:
+        pt[0][0] += x1
+        pt[0][1] += y1
+    if len(cnt) >= 5:
+        (ex, ey), (ew, eh), ang = cv2.fitEllipse(cnt)
+        mask_ellipse = np.zeros((HEIGHT, WIDTH), dtype=np.uint8)
+        cv2.ellipse(mask_ellipse, (int(ex), int(ey)), (int(ew / 2), int(eh / 2)), ang, 0, 360, 255, -1)
+        brightness = img_cv[:, :, 0].astype(np.int32) + img_cv[:, :, 1].astype(np.int32) + img_cv[:, :, 2].astype(np.int32)
+        masked = np.where(mask_ellipse > 0, brightness, 0)
+        vals = masked[masked > 0]
+        if len(vals) > 0:
+            bth = np.percentile(vals, 90)
+            bmask = (masked >= bth).astype(np.uint8) * 255
+            bcontours, _ = cv2.findContours(bmask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+            if bcontours:
+                blargest = max(bcontours, key=cv2.contourArea)
+                if cv2.contourArea(blargest) >= 3 and len(blargest) >= 5:
+                    (ix, iy), _, _ = cv2.fitEllipse(blargest)
+                    return (float(ix), float(iy))
+                M = cv2.moments(blargest)
+                if M["m00"] > 0:
+                    return (float(M["m10"] / M["m00"]), float(M["m01"] / M["m00"]))
+        return (float(ex), float(ey))
+    M = cv2.moments(cnt)
+    if M["m00"] > 0:
+        return (float(M["m10"] / M["m00"]), float(M["m01"] / M["m00"]))
+    return None
+
+
+def find_brightest(img_cv, cx, cy, roi_r, th):
+    x1 = max(0, cx - roi_r)
+    x2 = min(WIDTH, cx + roi_r)
+    y1 = max(0, cy - roi_r)
+    y2 = min(HEIGHT, cy + roi_r)
+    best_score = 0
+    best_pos = None
+    for y in range(y1, y2):
+        for x in range(x1, x2):
+            r, g, b = int(img_cv[y, x, 0]), int(img_cv[y, x, 1]), int(img_cv[y, x, 2])
+            is_red = (r > th and r > g * 1.5 and r > b * 1.5)
+            is_oe = (r > 200 and g > 200 and b > 200 and r >= g and r >= b and (r - g) > 10 and (r - b) > 10)
+            if is_red or is_oe:
+                score = r + g + b
+                dx, dy = x - cx, y - cy
+                dist = (dx * dx + dy * dy) ** 0.5
+                score *= max(0.5, 1.0 - (dist / roi_r) * 0.5)
+                if score > best_score:
+                    best_score = score
+                    best_pos = (float(x), float(y))
+    return best_pos
+
+
+# 打开键盘输入设备
+_key_fd = None
+try:
+    _key_fd = open("/dev/input/event0", "rb")
+except Exception:
+    try:
+        _key_fd = open("/dev/input/event1", "rb")
+    except Exception:
+        _key_fd = None
+
+print("=" * 50)
+print("激光中心点检测单元测试")
+print("=" * 50)
+print()
+
+cam = camera.Camera(WIDTH, HEIGHT)
+disp = display.Display()
+print("[OK] 摄像头和显示初始化完成")
+
+# 初始化激光串口
+_laser_on = False
+_laser_uart = None
+try:
+    pinmap.set_pin_function("A18", "UART1_RX")
+    pinmap.set_pin_function("A19", "UART1_TX")
+    _laser_uart = uart.UART("/dev/ttyS1", 9600)
+    _laser_uart.read(-1)
+    print("[OK] 激光串口初始化完成")
+except Exception as e:
+    print(f"[WARN] 激光串口初始化失败: {e}")
+
+LASER_ON = bytes([0xAA, 0x00, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x01, 0xC1])
+LASER_OFF = bytes([0xAA, 0x00, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x00, 0xC0])
+
+# 默认开启激光
+if _laser_uart:
+    try:
+        _laser_uart.write(LASER_ON)
+        time.sleep_ms(50)
+        _laser_uart.read(-1)
+        _laser_on = True
+        print("[OK] 激光已开启")
+    except Exception as e:
+        print(f"[WARN] 开启激光失败: {e}")
+print()
+
+pos_ellipse = None
+pos_bright = None
+frame_count = 0
+use_ellipse = True
+
+while not app.need_exit():
+    frame = cam.read()
+    if frame is None:
+        time.sleep_ms(10)
+        continue
+
+    frame_count += 1
+
+    if _USE_CV:
+        img_cv = image.image2cv(frame, False, False)
+        cx, cy = WIDTH // 2, HEIGHT // 2
+
+        t0 = time.ticks_ms()
+        pos_ellipse = find_ellipse(img_cv, cx, cy, SEARCH_RADIUS, THRESHOLD)
+        t1 = time.ticks_ms()
+        pos_bright = find_brightest(img_cv, cx, cy, SEARCH_RADIUS, THRESHOLD)
+        t2 = time.ticks_ms()
+
+        dt_e = abs(time.ticks_diff(t0, t1))
+        dt_b = abs(time.ticks_diff(t1, t2))
+
+        if frame_count % 5 == 0:
+            e_str = f"({pos_ellipse[0]:.1f},{pos_ellipse[1]:.1f})" if pos_ellipse else "None"
+            b_str = f"({pos_bright[0]:.1f},{pos_bright[1]:.1f})" if pos_bright else "None"
+            print(f"[LASER] ellipse={e_str} ({dt_e}ms)  brightest={b_str} ({dt_b}ms)  "
+                  f"th={THRESHOLD}  radius={SEARCH_RADIUS}")
+
+        # 叠加显示
+        pos = pos_ellipse if use_ellipse else pos_bright
+        h, w = img_cv.shape[:2]
+        cv2.circle(img_cv, (cx, cy), SEARCH_RADIUS, (0, 255, 0), 1)
+        cv2.circle(img_cv, (cx, cy), 2, (0, 255, 0), -1)
+        if pos:
+            x, y = int(pos[0]), int(pos[1])
+            cv2.circle(img_cv, (x, y), 6, (0, 0, 255), 2)
+            cv2.line(img_cv, (x - 14, y), (x + 14, y), (0, 0, 255), 1)
+            cv2.line(img_cv, (x, y - 14), (x, y + 14), (0, 0, 255), 1)
+            cv2.putText(img_cv, f"({x},{y})", (x + 10, y - 10),
+                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)
+        info = [
+            f"pos={pos if pos else 'None'}",
+            f"method={'ellipse' if use_ellipse else 'brightest'}  th={THRESHOLD}",
+            f"laser={'ON' if _laser_on else 'OFF'}",
+        ]
+        for i, line in enumerate(info):
+            cv2.putText(img_cv, line, (8, 20 + i * 22),
+                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 1, cv2.LINE_AA)
+
+        display_frame = image.cv2image(img_cv, False, False)
+    else:
+        display_frame = frame
+
+    disp.show(display_frame)
+
+    # 按键处理（非阻塞）
+    key = read_key_ev()
+    if key > 0:
+        c = chr(key & 0xFF) if key < 256 else ""
+        if key == 113 or key == 81 or key == 0x1b:  # q/Q/ESC
+            break
+        if c == "e" or key == 18:  # e
+            use_ellipse = not use_ellipse
+            print(f"[KEY] Method: {'ellipse' if use_ellipse else 'brightest'}")
+        if c == "l" or key == 12:  # l
+            _laser_on = not _laser_on
+            if _laser_uart:
+                try:
+                    _laser_uart.write(LASER_ON if _laser_on else LASER_OFF)
+                    time.sleep_ms(30)
+                    _laser_uart.read(-1)
+                    print(f"[KEY] Laser: {'ON' if _laser_on else 'OFF'}")
+                except Exception as e:
+                    print(f"[KEY] Laser error: {e}")
+            else:
+                print("[KEY] Laser UART not available")
+    time.sleep_ms(30)
+
+# 关闭激光
+if _laser_on and _laser_uart:
+    try:
+        _laser_uart.write(LASER_OFF)
+        _laser_uart.read(-1)
+        print("[EXIT] 激光已关闭")
+    except Exception:
+        pass
+print("[EXIT] 测试结束")
+if _key_fd:
+    _key_fd.close()
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+激光中心点检测单元测试（单文件，无项目依赖）
+直接使用 maix 标准库，实现红色激光点坐标检测
+
+运行方式：
+  python3 test/test_laser_center_point.py
+
+Ctrl+C 退出，按 s 保存截图
+"""
+
+from maix import camera, display, image, time, app, uart, pinmap
+import os
+import struct
+import select
+
+_USE_CV = False
+try:
+    import cv2
+    import numpy as np
+    _USE_CV = True
+except ImportError:
+    pass
+
+WIDTH = 640
+HEIGHT = 480
+THRESHOLD = 120
+RED_RATIO = 1.3
+SEARCH_RADIUS = 60
+
+
+def read_key_ev():
+    """非阻塞读取 /dev/input/event0 按键（返回 key_code 或 -1）"""
+    try:
+        r, _, _ = select.select([_key_fd], [], [], 0)
+        if r:
+            event = _key_fd.read(16)
+            if len(event) == 16:
+                _, _, etype, code, value = struct.unpack("IIHHI", event)
+                if etype == 1 and value == 1:
+                    return code
+    except Exception:
+        pass
+    return -1
+
+
+def find_ellipse(img_cv, cx, cy, roi_r, th, ratio):
+    x1 = max(0, cx - roi_r)
+    x2 = min(WIDTH, cx + roi_r)
+    y1 = max(0, cy - roi_r)
+    y2 = min(HEIGHT, cy + roi_r)
+    roi = img_cv[y1:y2, x1:x2]
+    if roi.size == 0:
+        return None
+    r = roi[:, :, 0].astype(np.int32)
+    g = roi[:, :, 1].astype(np.int32)
+    b = roi[:, :, 2].astype(np.int32)
+    mask = (r > th) & (r > g * ratio) & (r > b * ratio)
+    oe = (r > 200) & (g > 200) & (b > 200) & (r >= g) & (r >= b) & ((r - g) > 10) & ((r - b) > 10)
+    combined = (mask | oe).astype(np.uint8) * 255
+    contours, _ = cv2.findContours(combined, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+    if not contours:
+        return None
+    largest = max(contours, key=cv2.contourArea)
+    if cv2.contourArea(largest) < 5:
+        return None
+    cnt = largest.copy()
+    for pt in cnt:
+        pt[0][0] += x1
+        pt[0][1] += y1
+    if len(cnt) >= 5:
+        (ex, ey), (ew, eh), ang = cv2.fitEllipse(cnt)
+        mask_ellipse = np.zeros((HEIGHT, WIDTH), dtype=np.uint8)
+        cv2.ellipse(mask_ellipse, (int(ex), int(ey)), (int(ew / 2), int(eh / 2)), ang, 0, 360, 255, -1)
+        brightness = img_cv[:, :, 0].astype(np.int32) + img_cv[:, :, 1].astype(np.int32) + img_cv[:, :, 2].astype(np.int32)
+        masked = np.where(mask_ellipse > 0, brightness, 0)
+        vals = masked[masked > 0]
+        if len(vals) > 0:
+            bth = np.percentile(vals, 90)
+            bmask = (masked >= bth).astype(np.uint8) * 255
+            bcontours, _ = cv2.findContours(bmask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+            if bcontours:
+                blargest = max(bcontours, key=cv2.contourArea)
+                if cv2.contourArea(blargest) >= 3 and len(blargest) >= 5:
+                    (ix, iy), _, _ = cv2.fitEllipse(blargest)
+                    return (float(ix), float(iy))
+                M = cv2.moments(blargest)
+                if M["m00"] > 0:
+                    return (float(M["m10"] / M["m00"]), float(M["m01"] / M["m00"]))
+        return (float(ex), float(ey))
+    M = cv2.moments(cnt)
+    if M["m00"] > 0:
+        return (float(M["m10"] / M["m00"]), float(M["m01"] / M["m00"]))
+    return None
+
+
+def find_brightest_bytes(frame, cx, cy, roi_r, th, ratio):
+    """使用 frame.to_bytes() 两阶段搜索，避免 cv2 转换"""
+    x1 = max(0, cx - roi_r)
+    x2 = min(WIDTH, cx + roi_r)
+    y1 = max(0, cy - roi_r)
+    y2 = min(HEIGHT, cy + roi_r)
+    data = frame.to_bytes()
+    best_score = 0
+    best_pos = None
+    # 第一阶段：隔点粗搜
+    for y in range(y1, y2, 2):
+        for x in range(x1, x2, 2):
+            idx = (y * WIDTH + x) * 3
+            r = data[idx]; g = data[idx+1]; b = data[idx+2]
+            if (r > th and r > g * ratio and r > b * ratio) or \
+               (r > 200 and g > 200 and b > 200 and r >= g and r >= b and (r - g) > 10 and (r - b) > 10):
+                score = r + g + b
+                dx = x - cx; dy = y - cy
+                score *= max(0.5, 1.0 - ((dx*dx + dy*dy) ** 0.5 / roi_r) * 0.5)
+                if score > best_score:
+                    best_score = score
+                    best_pos = (x, y)
+    if best_pos is None:
+        return None
+    # 第二阶段：候选点 7x7 精细搜索
+    fx, fy = best_pos
+    x1f = max(0, fx - 3); x2f = min(WIDTH, fx + 4)
+    y1f = max(0, fy - 3); y2f = min(HEIGHT, fy + 4)
+    best_bright = 0
+    final_pos = best_pos
+    for y in range(y1f, y2f):
+        for x in range(x1f, x2f):
+            idx = (y * WIDTH + x) * 3
+            r = data[idx]; g = data[idx+1]; b = data[idx+2]
+            if (r > th and r > g * ratio and r > b * ratio) or \
+               (r > 200 and g > 200 and b > 200 and r >= g and r >= b and (r - g) > 10 and (r - b) > 10):
+                rgb_sum = r + g + b
+                if rgb_sum > best_bright:
+                    best_bright = rgb_sum
+                    final_pos = (float(x), float(y))
+    return final_pos
+
+
+# 打开键盘输入设备
+_key_fd = None
+try:
+    _key_fd = open("/dev/input/event0", "rb")
+except Exception:
+    try:
+        _key_fd = open("/dev/input/event1", "rb")
+    except Exception:
+        _key_fd = None
+
+print("=" * 50)
+print("激光中心点检测单元测试")
+print("=" * 50)
+print()
+
+cam = camera.Camera(WIDTH, HEIGHT)
+disp = display.Display()
+print("[OK] 摄像头和显示初始化完成")
+
+# 初始化激光串口
+_laser_on = False
+_laser_uart = None
+try:
+    pinmap.set_pin_function("A18", "UART1_RX")
+    pinmap.set_pin_function("A19", "UART1_TX")
+    _laser_uart = uart.UART("/dev/ttyS1", 9600)
+    _laser_uart.read(-1)
+    print("[OK] 激光串口初始化完成")
+except Exception as e:
+    print(f"[WARN] 激光串口初始化失败: {e}")
+
+LASER_ON = bytes([0xAA, 0x00, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x01, 0xC1])
+LASER_OFF = bytes([0xAA, 0x00, 0x01, 0xBE, 0x00, 0x01, 0x00, 0x00, 0xC0])
+
+# 默认开启激光
+if _laser_uart:
+    try:
+        _laser_uart.write(LASER_ON)
+        time.sleep_ms(50)
+        _laser_uart.read(-1)
+        _laser_on = True
+        print("[OK] 激光已开启")
+    except Exception as e:
+        print(f"[WARN] 开启激光失败: {e}")
+print()
+
+pos_ellipse = None
+pos_bright = None
+frame_count = 0
+use_ellipse = True
+
+while not app.need_exit():
+    frame = cam.read()
+    if frame is None:
+        time.sleep_ms(10)
+        continue
+
+    frame_count += 1
+
+    cx, cy = WIDTH // 2, HEIGHT // 2
+
+    t0 = time.ticks_ms()
+    pos_bright = find_brightest_bytes(frame, cx, cy, SEARCH_RADIUS, THRESHOLD, RED_RATIO)
+    t1 = time.ticks_ms()
+
+    pos_ellipse = None
+    if _USE_CV:
+        img_cv = image.image2cv(frame, False, False)
+        t2 = time.ticks_ms()
+        pos_ellipse = find_ellipse(img_cv, cx, cy, SEARCH_RADIUS, THRESHOLD, RED_RATIO)
+        t3 = time.ticks_ms()
+    else:
+        img_cv = None
+        t3 = t2 = t1
+
+    dt_b = abs(time.ticks_diff(t0, t1))
+    dt_e = abs(time.ticks_diff(t2, t3))
+
+    if frame_count % 5 == 0:
+        e_str = f"({pos_ellipse[0]:.1f},{pos_ellipse[1]:.1f})" if pos_ellipse else "None"
+        b_str = f"({pos_bright[0]:.1f},{pos_bright[1]:.1f})" if pos_bright else "None"
+        print(f"[LASER] ellipse={e_str} ({dt_e}ms)  brightest={b_str} ({dt_b}ms)  "
+              f"th={THRESHOLD}  ratio={RED_RATIO}  radius={SEARCH_RADIUS}")
+
+    pos = pos_ellipse if use_ellipse else pos_bright
+    if img_cv is not None:
+        cv2.circle(img_cv, (cx, cy), SEARCH_RADIUS, (0, 255, 0), 1)
+        cv2.circle(img_cv, (cx, cy), 2, (0, 255, 0), -1)
+        if pos:
+            x, y = int(pos[0]), int(pos[1])
+            cv2.circle(img_cv, (x, y), 6, (0, 0, 255), 2)
+            cv2.line(img_cv, (x - 14, y), (x + 14, y), (0, 0, 255), 1)
+            cv2.line(img_cv, (x, y - 14), (x, y + 14), (0, 0, 255), 1)
+            cv2.putText(img_cv, f"({x},{y})", (x + 10, y - 10),
+                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)
+        info = [
+            f"pos={pos if pos else 'None'}",
+            f"method={'ellipse' if use_ellipse else 'brightest'}  th={THRESHOLD}  ratio={RED_RATIO}",
+            f"laser={'ON' if _laser_on else 'OFF'}",
+        ]
+        for i, line in enumerate(info):
+            cv2.putText(img_cv, line, (8, 20 + i * 22),
+                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 1, cv2.LINE_AA)
+        display_frame = image.cv2image(img_cv, False, False)
+    else:
+        display_frame = frame
+
+    disp.show(display_frame)
+
+    # 按键处理（非阻塞）
+    key = read_key_ev()
+    if key > 0:
+        c = chr(key & 0xFF) if key < 256 else ""
+        if key == 113 or key == 81 or key == 0x1b:  # q/Q/ESC
+            break
+        if c == "e" or key == 18:  # e
+            use_ellipse = not use_ellipse
+            print(f"[KEY] Method: {'ellipse' if use_ellipse else 'brightest'}")
+        if c == "l" or key == 12:  # l
+            _laser_on = not _laser_on
+            if _laser_uart:
+                try:
+                    _laser_uart.write(LASER_ON if _laser_on else LASER_OFF)
+                    time.sleep_ms(30)
+                    _laser_uart.read(-1)
+                    print(f"[KEY] Laser: {'ON' if _laser_on else 'OFF'}")
+                except Exception as e:
+                    print(f"[KEY] Laser error: {e}")
+            else:
+                print("[KEY] Laser UART not available")
+    time.sleep_ms(30)
+
+# 关闭激光
+if _laser_on and _laser_uart:
+    try:
+        _laser_uart.write(LASER_OFF)
+        _laser_uart.read(-1)
+        print("[EXIT] 激光已关闭")
+    except Exception:
+        pass
+print("[EXIT] 测试结束")
+if _key_fd:
+    _key_fd.close()

test/test_motor.py

@@ -0,0 +1,16 @@
+from maix import gpio, pinmap, time
+ 
+
+#设置引脚为输出
+led = gpio.GPIO("A25", gpio.Mode.OUT)
+#设置低电平
+led.value(0)
+ 
+while 1:
+    # time.sleep_ms(1000)
+    #对该引脚的电平进行取反（原高-》现低）
+    # led.toggle()
+    led.value(1)
+    #延时
+    time.sleep_ms(5000)
+    led.value(0)

test/test_yolo26.py

@@ -0,0 +1,18 @@
+from maix import camera, display, image, nn, app
+
+# 1. 初始化模型 (请确保模型文件 .mud 路径正确)
+detector = nn.YOLOv5(model="/root/model_279350.mud", dual_buff=True)
+
+# 2. 初始化摄像头，分辨率与模型输入匹配
+cam = camera.Camera(detector.input_width(), detector.input_height(), detector.input_format())
+disp = display.Display()
+
+# 3. 主循环：实时检测与显示
+while not app.need_exit():
+    img = cam.read()                         # 从摄像头读取一帧
+    objs = detector.detect(img, conf_th=0.5, iou_th=0.45) # 执行YOLO11推理
+    for obj in objs:                         # 绘制所有检测到的目标
+        img.draw_rect(obj.x, obj.y, obj.w, obj.h, color=image.COLOR_RED)
+        msg = f'{detector.labels[obj.class_id]}: {obj.score:.2f}'
+        img.draw_string(obj.x, obj.y, msg, color=image.COLOR_RED)
+    disp.show(img)                           # 更新屏幕显示

test/test_yolo_camera_simple.py

@@ -0,0 +1,209 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+摄像头实时 YOLOv5 简易测试脚本。
+
+特点：
+- 完全独立脚本，直接 python test/test_yolo_camera_simple.py 运行，不需要传参。
+- 不 import config，不依赖项目模块。
+- 直接调用 maix.nn.YOLOv5(model=..., dual_buff=False)。
+- camera.read() 得到的 Maix image 直接送 det.detect()。
+- 在画面上画检测框、类别、置信度，并显示到屏幕。
+
+运行环境：MaixCAM / MaixPy。
+"""
+
+import os
+
+
+CAMERA_WIDTH = 640
+CAMERA_HEIGHT = 480
+# 默认与主项目 config.TRIANGLE_YOLO_MODEL_PATH 一致（勿用 /root/yolo26_int8.mud，那是占位路径）
+_MODEL_DEFAULT = "/maixapp/apps/t11/model_270139.mud"
+try:
+    import config as _cfg
+
+    MODEL_PATH = getattr(_cfg, "TRIANGLE_YOLO_MODEL_PATH", _MODEL_DEFAULT) or _MODEL_DEFAULT
+except Exception:
+    MODEL_PATH = _MODEL_DEFAULT
+CONF_TH = 0.7
+IOU_TH = 0.45
+# native: Maix detect 返回框已映射到 camera.read() 图像坐标；letterbox: 需要从网络输入坐标反算
+COORD_MODE = "native"
+# 只用于 DRAW_ONLY_CLASS_IDS=True 时过滤显示；默认画所有框
+CLASS_IDS = (0,)
+DRAW_ONLY_CLASS_IDS = False  # True=只画 CLASS_IDS 里的类别；False=画所有 YOLO 返回框
+
+
+def _det_obj_class_id(o):
+    for key in ("class_id", "cls", "label", "category", "cat_id", "id"):
+        if hasattr(o, key):
+            v = getattr(o, key)
+            if v is None:
+                continue
+            try:
+                return int(float(v))
+            except (TypeError, ValueError):
+                continue
+    return None
+
+
+def _det_obj_from_seq(t):
+    if not isinstance(t, (list, tuple)) or len(t) < 6:
+        return None
+
+    class Box:
+        pass
+
+    b = Box()
+    b.x = float(t[0])
+    b.y = float(t[1])
+    b.w = float(t[2])
+    b.h = float(t[3])
+    b.score = float(t[4])
+    b.class_id = int(float(t[5]))
+    return b
+
+
+def _normalize_objs(objs):
+    out = []
+    for o in objs or []:
+        if isinstance(o, (list, tuple)):
+            m = _det_obj_from_seq(o)
+            if m is not None:
+                out.append(m)
+        else:
+            out.append(o)
+    return out
+
+
+def _letterbox_net_to_src_xyxy(x, y, w, h, src_w, src_h, net_w, net_h):
+    scale = min(net_w / float(src_w), net_h / float(src_h))
+    new_w = src_w * scale
+    new_h = src_h * scale
+    pad_x = (net_w - new_w) * 0.5
+    pad_y = (net_h - new_h) * 0.5
+    x0 = (x - pad_x) / scale
+    y0 = (y - pad_y) / scale
+    x1 = (x + w - pad_x) / scale
+    y1 = (y + h - pad_y) / scale
+    return x0, y0, x1, y1
+
+
+def _det_to_src_xyxy(o, coord_mode, src_w, src_h, net_w, net_h):
+    x = float(getattr(o, "x", 0.0))
+    y = float(getattr(o, "y", 0.0))
+    w = float(getattr(o, "w", 0.0))
+    h = float(getattr(o, "h", 0.0))
+    if coord_mode in ("native", "source", "camera", "full"):
+        return x, y, x + w, y + h
+    return _letterbox_net_to_src_xyxy(x, y, w, h, src_w, src_h, net_w, net_h)
+
+
+def _clip_xywh(x0, y0, x1, y1, src_w, src_h):
+    x0 = max(0, min(int(round(x0)), src_w - 1))
+    y0 = max(0, min(int(round(y0)), src_h - 1))
+    x1 = max(x0 + 1, min(int(round(x1)), src_w))
+    y1 = max(y0 + 1, min(int(round(y1)), src_h))
+    return x0, y0, x1 - x0, y1 - y0
+
+
+def _label(det, cid):
+    labels = getattr(det, "labels", None)
+    if labels is None:
+        return str(cid)
+    try:
+        return str(labels[int(cid)])
+    except Exception:
+        return str(cid)
+
+
+def main():
+    from maix import camera, display, nn, time, image
+
+    if not MODEL_PATH or not os.path.isfile(MODEL_PATH):
+        print("[ERR] 模型文件不存在:", MODEL_PATH)
+        return
+
+    print("[INFO] 初始化 YOLO 模型:", MODEL_PATH)
+    det = nn.YOLOv26(model=MODEL_PATH, dual_buff=False)
+    net_w = int(det.input_width())
+    net_h = int(det.input_height())
+    print(
+        "[INFO] net_in=%dx%d conf=%.2f iou=%.2f coord=%s class_ids=%s"
+        % (net_w, net_h, CONF_TH, IOU_TH, COORD_MODE, str(CLASS_IDS))
+    )
+
+    print("[INFO] 初始化摄像头: %dx%d" % (CAMERA_WIDTH, CAMERA_HEIGHT))
+    cam = camera.Camera(CAMERA_WIDTH, CAMERA_HEIGHT)
+    disp = display.Display()
+
+    color_cycle = []
+    for name in ("RED", "GREEN", "BLUE", "ORANGE", "YELLOW", "CYAN", "MAGENTA"):
+        c = getattr(image, "COLOR_" + name, None)
+        if c is not None:
+            color_cycle.append(c)
+    if not color_cycle:
+        color_cycle = [getattr(image, "COLOR_RED", 0)]
+
+    frame_idx = 0
+    last_log_ms = time.ticks_ms()
+    fps_count = 0
+
+    while True:
+        frame = cam.read()
+        src_w = frame.width()
+        src_h = frame.height()
+
+        t0 = time.ticks_ms()
+        raw = det.detect(frame, conf_th=CONF_TH, iou_th=IOU_TH)
+        detect_ms = time.ticks_ms() - t0
+        objs = _normalize_objs(raw if raw is not None else [])
+
+        draw_count = 0
+        for i, o in enumerate(objs):
+            cid = _det_obj_class_id(o)
+            if cid is None:
+                cid = -1
+            if DRAW_ONLY_CLASS_IDS and cid not in CLASS_IDS:
+                continue
+            try:
+                score = float(getattr(o, "score", 0.0))
+            except Exception:
+                score = 0.0
+
+            x0, y0, x1, y1 = _det_to_src_xyxy(o, COORD_MODE, src_w, src_h, net_w, net_h)
+            ix, iy, iw, ih = _clip_xywh(x0, y0, x1, y1, src_w, src_h)
+            col = color_cycle[cid % len(color_cycle)] if cid >= 0 else color_cycle[0]
+            frame.draw_rect(ix, iy, iw, ih, color=col)
+            frame.draw_string(ix, max(0, iy - 16), "%s %.2f" % (_label(det, cid), score), color=col)
+            draw_count += 1
+
+        frame.draw_string(4, 4, "YOLO boxes:%d draw:%d %dms" % (len(objs), draw_count, detect_ms), color=color_cycle[0])
+        disp.show(frame)
+
+        frame_idx += 1
+        fps_count += 1
+        now = time.ticks_ms()
+        if now - last_log_ms >= 1000:
+            print(
+                "[INFO] frame=%d fps=%d raw_boxes=%d draw_boxes=%d detect_ms=%d"
+                % (frame_idx, fps_count, len(objs), draw_count, detect_ms)
+            )
+            fps_count = 0
+            last_log_ms = now
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except KeyboardInterrupt:
+        print("[INFO] exit")
+    except Exception as e:
+        print("[ERR]", e)
+        try:
+            import traceback
+
+            traceback.print_exc()
+        except Exception:
+            pass

test/test_yolov8.py

@@ -0,0 +1,29 @@
+from maix import image, nn, display
+
+# 1. 加载模型
+detector = nn.YOLOv8(model="/root/279350.mud", dual_buff=False)
+# 2. 加载指定图片（根据模型输入尺寸自动缩放宽高）
+img = image.load("/root/tes.jpg")
+if img is None:
+    raise FileNotFoundError("图片加载失败，请检查路径")
+
+# 3. 调整图片尺寸到模型输入要求（可选，detect内部会处理，但提前缩放可提高速度）
+# img = img.resize(detector.input_width(), detector.input_height())
+
+# 4. 检测
+objs = detector.detect(img, conf_th=0.5, iou_th=0.45)
+
+# 5. 在图片上绘制结果
+for obj in objs:
+    img.draw_rect(obj.x, obj.y, obj.w, obj.h, color=image.COLOR_RED)
+    msg = f'{detector.labels[obj.class_id]}: {obj.score:.2f}'
+    img.draw_string(obj.x, obj.y, msg, color=image.COLOR_RED)
+
+# 6. 显示结果（如果设备有屏幕）
+disp = display.Display()
+disp.show(img)
+
+# 7. 保存结果（可选）
+img.save("/root/result.jpg")
+
+print("识别完成，结果已显示并保存为 result.jpg")

train_yolo/synth_compose_yolo.py

@@ -28,6 +28,11 @@ Stage2 ROI（对齐「先检整靶再裁小图」的第二步输入）：
   - --motion-prob：施加概率；--motion-kernel-min/max：模糊 streak 长度（奇数核，越大越糊）。
   - 可与 --blur-max 高斯模糊叠加；Stage2 建议：--motion-prob 0.5~0.7 --motion-kernel-max 35 --blur-max 1.2
 
+透视有两种模式（见 --perspective-mode）：
+  - corner：四角独立随机抖动（旧版），组合后易出现「不物理」的夸张梯形；--perspective 为 jitter 强度。
+  - planar：把靶当作平面，按 yaw/pitch/roll（度）随机旋转再投影，便于限定在例如「偏航 ±45°、俯仰 ±30°」内，
+    更接近 5m 外拍 40cm 靶等场景（焦距用 --planar-focal-frac 控制视场）。
+
 依赖：OpenCV + NumPy（PC 上跑即可；Maix 上若内存够也可试）。
 
 示例：
@@ -266,6 +271,74 @@ def _paste_fg_on_bg(bg_bgr, x, y, fg_scaled_bgra):
     roi_bg[:] = blended.astype(np.uint8)
 
 
+def _rotation_matrix_ypr(yaw_deg: float, pitch_deg: float, roll_deg: float, np) -> np.ndarray:
+    """靶平面绕相机坐标原点旋转：先 yaw(Y)，再 pitch(X)，再 roll(Z)。单位：度。"""
+    y, p, r = np.radians([yaw_deg, pitch_deg, roll_deg])
+    cy, sy = np.cos(y), np.sin(y)
+    cp, sp = np.cos(p), np.sin(p)
+    cr, sr = np.cos(r), np.sin(r)
+    Ry = np.array([[cy, 0.0, sy], [0.0, 1.0, 0.0], [-sy, 0.0, cy]], dtype=np.float64)
+    Rx = np.array([[1.0, 0.0, 0.0], [0.0, cp, -sp], [0.0, sp, cp]], dtype=np.float64)
+    Rz = np.array([[cr, -sr, 0.0], [sr, cr, 0.0], [0.0, 0.0, 1.0]], dtype=np.float64)
+    return Rz @ Rx @ Ry
+
+
+def _perspective_warp_planar_rgba(
+    img_bgra,
+    yaw_deg: float,
+    pitch_deg: float,
+    roll_deg: float,
+    focal_frac: float,
+    np,
+    cv2,
+):
+    """
+    平面靶透视：将 frontal 图像视作 z=1 平面上的针孔投影，再旋转三维射线方向，等价于靶相对相机倾斜。
+    focal_frac：焦距 fx=fy=max(w,h)*focal_frac，越大视场越窄、透视越温和。
+    返回 (warped BGRA, M)；退化时返回 (copy, None)。
+    """
+    h, w = img_bgra.shape[:2]
+    if min(w, h) < 16:
+        return img_bgra.copy(), None
+    fx = fy = float(max(w, h) * max(0.25, focal_frac))
+    cx, cy = w * 0.5, h * 0.5
+    R = _rotation_matrix_ypr(yaw_deg, pitch_deg, roll_deg, np)
+
+    uv_dst: list[list[float]] = []
+    z_min = 0.08
+    for u, v in ((0.0, 0.0), (float(w), 0.0), (float(w), float(h)), (0.0, float(h))):
+        x = (u - cx) / fx
+        y = (v - cy) / fy
+        z = 1.0
+        vec = R @ np.array([x, y, z], dtype=np.float64)
+        if float(vec[2]) < z_min:
+            return img_bgra.copy(), None
+        uu = fx * (vec[0] / vec[2]) + cx
+        vv = fy * (vec[1] / vec[2]) + cy
+        uv_dst.append([uu, vv])
+
+    pts_dst = np.float32(uv_dst)
+    xmin = float(pts_dst[:, 0].min())
+    ymin = float(pts_dst[:, 1].min())
+    pts_shift = pts_dst.copy()
+    pts_shift[:, 0] -= xmin
+    pts_shift[:, 1] -= ymin
+    out_w = max(4, int(np.ceil(float(pts_shift[:, 0].max()))) + 2)
+    out_h = max(4, int(np.ceil(float(pts_shift[:, 1].max()))) + 2)
+
+    pts_src = np.float32([[0, 0], [w, 0], [w, h], [0, h]])
+    M = cv2.getPerspectiveTransform(pts_src, pts_shift)
+    warped = cv2.warpPerspective(
+        img_bgra,
+        M,
+        (out_w, out_h),
+        flags=cv2.INTER_LINEAR,
+        borderMode=cv2.BORDER_CONSTANT,
+        borderValue=(0, 0, 0, 0),
+    )
+    return warped, M
+
+
 def _perspective_warp_rgba(img_bgra, jitter_frac: float, rng: random.Random, np, cv2):
     """
     对前景做轻微透视（四角微移），返回 (warped BGRA, M)。
@@ -311,6 +384,26 @@ def _perspective_warp_rgba(img_bgra, jitter_frac: float, rng: random.Random, np,
     return warped, M
 
 
+def _apply_perspective_warp_fg(fg_bgra, rng: random.Random, np, cv2, args) -> tuple:
+    """在已由 perspective_prob 抽中的样本上做一次透视。返回 (fg_out, M|None)。"""
+    mode = getattr(args, "perspective_mode", "corner")
+    if mode == "planar":
+        
+        x = rng.betavariate(2,2)
+        yaw = -float(args.yaw_max_deg) + x * (float(args.yaw_max_deg) +float(args.yaw_max_deg))
+        x = rng.betavariate(2,2)
+        pitch = -float(args.pitch_max_deg) + x * (float(args.pitch_max_deg) +float(args.pitch_max_deg))
+        x = rng.betavariate(2,2)
+        roll = -float(args.roll_max_deg) + x * (float(args.roll_max_deg) +float(args.roll_max_deg))
+     
+        focal_frac = float(getattr(args, "planar_focal_frac", 1.25))
+        return _perspective_warp_planar_rgba(fg_bgra, yaw, pitch, roll, focal_frac, np, cv2)
+    jitter = float(getattr(args, "perspective", 0.0))
+    if jitter <= 0:
+        return fg_bgra.copy(), None
+    return _perspective_warp_rgba(fg_bgra, jitter, rng, np, cv2)
+
+
 def _color_jitter_bgr(comp_bgr, strength: float, rng: random.Random, np, cv2):
     """整图 HSV 抖动：strength∈[0,1] 越大越强。"""
     if strength <= 1e-6:
@@ -519,11 +612,17 @@ def main():
         help="运动模糊 streak 长度上限，越大越像长曝光/手抖",
     )
     ap.add_argument("--jpeg-quality", type=int, default=92)
+    ap.add_argument(
+        "--perspective-mode",
+        choices=("corner", "planar"),
+        default="corner",
+        help="corner=四角随机抖动（易夸张）；planar=yaw/pitch/roll 平面投影（易限定视姿）",
+    )
     ap.add_argument(
         "--perspective",
         type=float,
         default=0.0,
-        help="轻微透视：四角扰动约为 min(靶宽,靶高)×该系数，0 关闭（建议 0.02~0.06）",
+        help="仅 perspective-mode=corner：四角扰动≈min(靶宽,靶高)×该系数，0 关闭（温和可用 0.015~0.03）",
     )
     ap.add_argument(
         "--perspective-prob",
@@ -531,6 +630,30 @@ def main():
         default=0.75,
         help="每张图应用透视的概率 0~1",
     )
+    ap.add_argument(
+        "--yaw-max-deg",
+        type=float,
+        default=45.0,
+        help="planar：偏航角均匀采样上界（度），实际 ∈[-max,max]",
+    )
+    ap.add_argument(
+        "--pitch-max-deg",
+        type=float,
+        default=30.0,
+        help="planar：俯仰角均匀采样上界（度）",
+    )
+    ap.add_argument(
+        "--roll-max-deg",
+        type=float,
+        default=8.0,
+        help="planar：滚转角均匀采样上界（度），手持可略大",
+    )
+    ap.add_argument(
+        "--planar-focal-frac",
+        type=float,
+        default=1.25,
+        help="planar：fx=fy=max(w,h)×该值，越大透视越温和（建议 1.1~1.8）",
+    )
     ap.add_argument(
         "--color-jitter",
         type=float,
@@ -662,8 +785,15 @@ def main():
         fg_s = cv2.resize(fg_crop, (new_w, new_h), interpolation=cv2.INTER_AREA)
 
         persp_M = None
-        if args.perspective > 0 and rng.random() < args.perspective_prob:
-            fg_s, persp_M = _perspective_warp_rgba(fg_s, args.perspective, rng, np, cv2)
+        want_p = (
+            rng.random() < args.perspective_prob
+            and (
+                args.perspective_mode == "planar"
+                or (args.perspective_mode == "corner" and args.perspective > 0)
+            )
+        )
+        if want_p:
+            fg_s, persp_M = _apply_perspective_warp_fg(fg_s, rng, np, cv2, args)
 
         fw2, fh2 = fg_s.shape[1], fg_s.shape[0]
         tx0, ty0, tw, th = _fg_bbox_from_alpha(fg_s)

train_yolo/train_black_triangle_pos.py

@@ -0,0 +1,506 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+YOLO11 关键点检测训练脚本（靶纸四角）。
+
+设备优先级（--device auto）：Intel XPU > NVIDIA CUDA > CPU。
+默认 imgsz=960；批大小默认 4（大图显存紧张时可再降）。
+
+关于「业务像素误差」：
+  Ultralytics 没有在 yaml 里设定「像素阈值」的选项；反向传播仍由 pose/kobj/box 等内部 loss 驱动。
+  - 监控：--pixel-metrics-every N（每 N 个 epoch 打印 mean/p95，并合并进 runs/.../results.csv，见 pose_pixel_metrics.py）。
+  - 选 best.pt / early stopping：加 --best-by-pixel，用验证集 mean 像素误差（与 pose_pixel_metrics
+    同一口径）代替 mAP 合成 fitness（fitness = -mean_px，越小越好）。
+    多卡 DDP（world_size>1）时会自动退回默认 mAP fitness。
+
+  XPU：Ultralytics BaseTrainer._get_memory / _clear_memory 把非 MPS、非 CPU 一律当 CUDA，
+  会在验证前调用 torch.cuda 而报错；本脚本在选用 XPU 时自动打补丁（见 _patch_ultralytics_trainer_for_xpu）。
+
+  务必使用 pose 任务：YOLO(...) 与 model.train(...) 均指定 task='pose'。若误用默认 detect，
+  会把 17 列 Pose 标注当成检测/分割解析，校验时出现「coordinates > 1」或 [2.] 等假象。
+"""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import gc
+import glob
+import os
+import tempfile
+from copy import deepcopy
+from pathlib import Path
+
+import torch
+from ultralytics import YOLO
+
+from pose_pixel_metrics import eval_val_pixel_error
+import warnings
+warnings.filterwarnings('ignore', 
+    message=".*scatter_add_kernel does not have a deterministic implementation.*")
+
+
+def _clear_ultralytics_label_caches(data_yaml_path: str) -> int:
+    """删除 data.yaml 的 path 下 labels/*.cache。
+
+    Ultralytics 的校验缓存 hash 仅依赖「标签/图片路径字符串 + 各文件 size 之和」，不含文件内容；
+    修正 *.txt 后若总和巧合不变，可能继续加载旧 cache 并重播旧的 corrupt 日志，训练前应删掉。"""
+    from ultralytics.utils import YAML
+
+    try:
+        cfg = YAML.load(data_yaml_path)
+    except Exception:
+        return 0
+    root = cfg.get("path")
+    if not root:
+        return 0
+    root = os.path.abspath(os.path.expanduser(str(root)))
+    pattern = os.path.join(root, "labels", "*.cache")
+    n = 0
+    for p in glob.glob(pattern):
+        try:
+            os.unlink(p)
+            n += 1
+        except OSError:
+            pass
+    return n
+
+
+def _pick_device(explicit: str | None):
+    """返回 ultralytics train/predict 可用的 device。"""
+    if explicit and explicit != "auto":
+        e = explicit.lower()
+        if e == "xpu":
+            if getattr(torch, "xpu", None) is None or not torch.xpu.is_available():
+                raise RuntimeError("指定了 --device xpu 但当前环境不可用")
+            return torch.device("xpu")
+        if e in ("0", "cuda", "gpu"):
+            if not torch.cuda.is_available():
+                raise RuntimeError("指定了 CUDA 但不可用")
+            return 0
+        if e == "cpu":
+            return "cpu"
+        return explicit
+    if getattr(torch, "xpu", None) is not None and torch.xpu.is_available():
+        return torch.device("xpu")
+    if torch.cuda.is_available():
+        return 0
+    return "cpu"
+
+
+def _default_amp(device) -> bool:
+    if isinstance(device, torch.device) and device.type == "xpu":
+        return False
+    if device == "cpu":
+        return False
+    return True
+
+
+def _patch_ultralytics_for_xpu():
+    """为 Ultralytics 打补丁，使其能在 XPU 环境下正常训练和验证。"""
+    import ultralytics.engine.trainer as ut_trainer
+    import ultralytics.engine.validator as ut_validator
+    from ultralytics.utils.torch_utils import select_device as _original_select_device
+
+    # 1. 覆盖 select_device：Trainer 初始化传入 torch.device("xpu") 会走原版早返回；
+    #    初始化后 args.device 会变成字符串 "xpu"，中期 val 用 trainer.device，不调用 select_device；
+    #    训练结束 final_eval 里 Validator 会 select_device("xpu")，且 validator 在 import 时已绑定原函数，
+    #    只改 torch_utils 无效，必须同时修补 trainer/validator 模块内的引用。
+    def _patched_select_device(device="", *args, **kwargs):
+        # Ultralytics 8.4.x: select_device(device="", newline=False, verbose=True)
+        # Older forks sometimes passed extra positional args; forward everything.
+        if isinstance(device, str):
+            d = device.strip().lower()
+            if d == "xpu" or d.startswith("xpu:"):
+                return torch.device(device.strip())
+        return _original_select_device(device, *args, **kwargs)
+
+    import ultralytics.utils.torch_utils
+
+    ultralytics.utils.torch_utils.select_device = _patched_select_device
+    ut_trainer.select_device = _patched_select_device
+    ut_validator.select_device = _patched_select_device
+
+    # 2. 修补 Trainer 的内存函数
+    BT = ut_trainer.BaseTrainer
+    if not getattr(BT, "_archery_xpu_memory_patched", False):
+        _orig_get_memory = BT._get_memory
+        _orig_clear_memory = BT._clear_memory
+
+        def _get_memory(self, fraction=False):
+            if self.device.type != "xpu":
+                return _orig_get_memory(self, fraction)
+            # ... (原有的 XPU 内存获取逻辑保持不变) ...
+            memory, total = 0, 0
+            try:
+                idx = self.device.index
+                if idx is None:
+                    idx = torch.xpu.current_device()
+                memory = int(torch.xpu.memory_allocated(idx))
+                if fraction:
+                    total = int(torch.xpu.get_device_properties(idx).total_memory)
+            except Exception:
+                pass
+            return (memory / total) if fraction and total > 0 else (memory / 2**30)
+
+        def _clear_memory(self, threshold=None):
+            if self.device.type != "xpu":
+                return _orig_clear_memory(self, threshold)
+            if threshold is not None:
+                assert 0 <= threshold <= 1, "Threshold must be between 0 and 1."
+                if self._get_memory(fraction=True) <= threshold:
+                    return
+            gc.collect()
+            if hasattr(torch.xpu, "empty_cache"):
+                torch.xpu.empty_cache()
+
+        BT._get_memory = _get_memory
+        BT._clear_memory = _clear_memory
+        BT._archery_xpu_memory_patched = True
+
+    # 3. 修补 Validator 的内存函数 (关键是添加这部分)
+    BV = ut_validator.BaseValidator
+    if not getattr(BV, "_archery_xpu_memory_patched", False):
+        # 为 Validator 添加同样的内存处理方法
+        BV._get_memory = _get_memory
+        BV._clear_memory = _clear_memory
+        BV._archery_xpu_memory_patched = True
+
+
+def _install_best_by_pixel_validate(data_yaml: str, imgsz: int, conf: float) -> None:
+    """用验证集关键点像素 mean 替代 mAP fitness，驱动 best.pt 与 patience early stopping。"""
+    import ultralytics.engine.trainer as ut
+    from ultralytics.utils import RANK
+
+    BT = ut.BaseTrainer
+    if getattr(BT, "_archery_best_by_pixel_installed", False):
+        return
+
+    _orig_validate = BT.validate
+
+    def validate(self):
+        import torch.distributed as dist
+
+        if self.ema and self.world_size > 1:
+            for buffer in self.ema.ema.buffers():
+                dist.broadcast(buffer, src=0)
+        metrics = self.validator(self)
+        if metrics is None:
+            return None, None
+        orig_fitness = metrics.pop("fitness", -self.loss.detach().cpu().numpy())
+
+        use_pixel = self.world_size <= 1 and RANK in {-1, 0}
+        mean_px: float | None = None
+        if use_pixel:
+            tmp_path: str | None = None
+            try:
+                fd, tmp_path = tempfile.mkstemp(suffix=".pt", prefix="archery_pxfit_")
+                os.close(fd)
+                from ultralytics.utils.torch_utils import unwrap_model
+
+                core = unwrap_model(self.ema.ema if self.ema else self.model)
+                torch.save({"ema": deepcopy(core).half(), "train_args": vars(self.args)}, tmp_path)
+                probe = YOLO(tmp_path)
+                stats = eval_val_pixel_error(
+                    probe,
+                    data_yaml,
+                    device=self.device,
+                    imgsz=imgsz,
+                    conf=conf,
+                )
+                mean_px = stats.get("mean_px")
+                if mean_px is None:
+                    raise RuntimeError("无有效 mean_px（检查 val 标签与检测是否为空）")
+            except Exception as exc:
+                print(f"\n⚠️ [best-by-pixel] 像素探针失败，本 epoch 仍用 mAP fitness: {exc}\n")
+                mean_px = None
+            finally:
+                if tmp_path:
+                    try:
+                        os.unlink(tmp_path)
+                    except OSError:
+                        pass
+
+        if mean_px is not None:
+            fitness = -float(mean_px)
+            metrics["metrics/mean_px(val)"] = float(mean_px)
+        else:
+            fitness = float(orig_fitness)
+
+        if not self.best_fitness or self.best_fitness < fitness:
+            self.best_fitness = fitness
+        return metrics, fitness
+
+    BT.validate = validate
+    BT._archery_best_by_pixel_installed = True
+
+
+def _fmt_csv_metric(v: float | int | None) -> str:
+    if v is None:
+        return ""
+    if isinstance(v, float):
+        return f"{v:.6g}"
+    return str(v)
+
+
+# 写入 results.csv 的列名（与 --best-by-pixel 的 metrics/mean_px(val) 区分，避免被 last.pt 回调覆盖 EMA 行）
+_PIXEL_METRIC_COLUMNS: tuple[tuple[str, str], ...] = (
+    ("pixel_error/mean_px", "mean_px"),
+    ("pixel_error/median_px", "median_px"),
+    ("pixel_error/p95_px", "p95_px"),
+    ("pixel_error/max_px", "max_px"),
+    ("pixel_error/n_points", "n_points"),
+    ("pixel_error/n_images", "n_images"),
+    ("pixel_error/skip_no_det", "skip_no_det"),
+    ("pixel_error/skip_no_gt", "skip_no_gt"),
+    ("pixel_error/skip_kpt_mismatch", "skip_kpt_mismatch"),
+)
+
+
+def _merge_pixel_metrics_into_results_csv(save_dir: str | Path, epoch_1based: int, stats: dict) -> None:
+    """在 Ultralytics 写完本 epoch 行之后，把像素指标列合并进 results.csv（扩展表头、补空列）。"""
+    csv_path = Path(save_dir) / "results.csv"
+    if not csv_path.is_file():
+        return
+    try:
+        with open(csv_path, newline="", encoding="utf-8") as f:
+            rows = list(csv.reader(f))
+    except OSError:
+        return
+    if len(rows) < 2:
+        return
+    header = list(rows[0])
+    for col_name, _ in _PIXEL_METRIC_COLUMNS:
+        if col_name not in header:
+            header.append(col_name)
+            for ri in range(1, len(rows)):
+                rows[ri].append("")
+    col_ix = {name: i for i, name in enumerate(header)}
+    rows[0] = header
+    target_ri: int | None = None
+    for ri in range(1, len(rows)):
+        row = rows[ri]
+        while len(row) < len(header):
+            row.append("")
+        try:
+            if int(float(row[0].strip())) == int(epoch_1based):
+                target_ri = ri
+        except (ValueError, IndexError):
+            continue
+    if target_ri is None:
+        return
+    row = rows[target_ri]
+    while len(row) < len(header):
+        row.append("")
+    for col_name, sk in _PIXEL_METRIC_COLUMNS:
+        row[col_ix[col_name]] = _fmt_csv_metric(stats.get(sk))
+    try:
+        with open(csv_path, "w", newline="", encoding="utf-8") as f:
+            w = csv.writer(f)
+            w.writerows(rows)
+    except OSError:
+        pass
+
+
+def _make_pixel_metrics_callback(data_yaml: str, every: int, imgsz: int, conf: float = 0.25):
+    def on_fit_epoch_end(trainer):
+        from ultralytics.utils import RANK
+
+        if RANK not in {-1, 0}:
+            return
+        if every <= 0:
+            return
+        ep = int(getattr(trainer, "epoch", -1))
+        if (ep + 1) % every != 0:
+            return
+        w = Path(trainer.save_dir) / "weights" / "last.pt"
+        if not w.is_file():
+            return
+        m = YOLO(str(w))
+        stats = eval_val_pixel_error(
+            m,
+            data_yaml,
+            device=trainer.device,
+            imgsz=imgsz,
+            conf=conf,
+        )
+        mean_px = stats.get("mean_px")
+        p95_px = stats.get("p95_px")
+        mean_s = f"{mean_px:.3f}" if mean_px is not None else "n/a"
+        p95_s = f"{p95_px:.3f}" if p95_px is not None else "n/a"
+        print(
+            f"\n[pixel-metrics] epoch {ep + 1}: mean_px={mean_s} p95_px={p95_s} "
+            f"n_points={stats.get('n_points', 0)} "
+            f"skip(det/gt/k)={stats['skip_no_det']}/{stats['skip_no_gt']}/{stats['skip_kpt_mismatch']}\n"
+        )
+        _merge_pixel_metrics_into_results_csv(trainer.save_dir, ep + 1, stats)
+
+    return on_fit_epoch_end
+
+
+def main():
+    ap = argparse.ArgumentParser(description="YOLO Pose 训练（XPU/CUDA/CPU）")
+    ap.add_argument("--data", default="datasets/dataset_pose.yaml", help="data.yaml")
+    ap.add_argument("--model", default="yolo11x-pose.pt", help="预训练权重")
+    ap.add_argument("--epochs", type=int, default=100)
+    ap.add_argument("--imgsz", type=int, default=960, help="训练输入边长（默认 960）")
+    ap.add_argument("--batch", type=int, default=4, help="批大小；OOM 时减小")
+    ap.add_argument(
+        "--device",
+        default="auto",
+        help="auto | xpu | 0 | cuda | cpu（auto：XPU 优先）",
+    )
+    ap.add_argument(
+        "--no-amp",
+        action="store_true",
+        help="关闭混合精度（默认：CUDA 开启，XPU/CPU 关闭）",
+    )
+    ap.add_argument("--project", default="runs/pose")
+    ap.add_argument("--name", default="target_pose_train")
+    ap.add_argument("--workers", type=int, default=4)
+    ap.add_argument(
+        "--pixel-metrics-every",
+        type=int,
+        default=0,
+        help="每 N 个 epoch 在 val 上打印像素误差并写入 results.csv 对应 epoch 行（0=关闭）；需 labels 与 data.yaml 布局一致",
+    )
+    ap.add_argument(
+        "--pixel-metrics-conf",
+        type=float,
+        default=0.25,
+        help="--pixel-metrics-every 时 predict 置信度阈值（默认 0.25）",
+    )
+    ap.add_argument(
+        "--best-by-pixel",
+        action="store_true",
+        help="best.pt 与 early stopping 按验证集 mean 像素误差（同 pose_pixel_metrics），fitness=-mean_px；单卡有效，DDP 自动退回 mAP",
+    )
+    ap.add_argument(
+        "--pixel-fitness-conf",
+        type=float,
+        default=0.25,
+        help="--best-by-pixel 时 predict 置信度阈值（默认与 pixel-metrics 一致）",
+    )
+    ap.add_argument(
+        "--export-onnx",
+        action="store_true",
+        help="训练结束后导出 ONNX（需再设 --onnx-imgsz）",
+    )
+    ap.add_argument(
+        "--onnx-imgsz",
+        type=int,
+        nargs=2,
+        metavar=("H", "W"),
+        default=[224, 320],
+        help="导出 ONNX 的 [高, 宽]，默认 224 320（Maix 常用）",
+    )
+    ap.add_argument(
+        "--clear-label-cache",
+        action="store_true",
+        help="启动训练前删除 data.yaml 中 path 下的 labels/*.cache（修正标注后仍报 corrupt 时用）",
+    )
+    args = ap.parse_args()
+
+    device = _pick_device(None if args.device == "auto" else args.device)
+    use_amp = False if args.no_amp else _default_amp(device)
+
+    if isinstance(device, torch.device) and device.type == "xpu":
+        print(f"✅ 使用 Intel XPU: {device}")
+    elif device == 0 or device == "0":
+        print(f"✅ 使用 CUDA: {torch.cuda.get_device_name(0)}")
+    else:
+        print("⚠️ 使用 CPU，训练会较慢")
+
+    if isinstance(device, torch.device) and device.type == "xpu":
+        _patch_ultralytics_for_xpu()
+
+    data_yaml = args.data
+    if not os.path.isabs(data_yaml):
+        data_yaml = os.path.join(os.path.dirname(os.path.abspath(__file__)), data_yaml)
+    if not os.path.exists(data_yaml):
+        print(f"❌ 数据集配置不存在: {data_yaml}")
+        return
+
+    if args.clear_label_cache:
+        n_rm = _clear_ultralytics_label_caches(data_yaml)
+        print(f"🗑️ 已删除标签目录缓存 {n_rm} 个（labels/*.cache），将强制重新扫描标注。")
+
+    print(f"📦 加载模型: {args.model}（固定 task=pose）")
+    model = YOLO(args.model, task="pose")
+
+    if args.best_by_pixel:
+        _install_best_by_pixel_validate(data_yaml, args.imgsz, args.pixel_fitness_conf)
+        print(
+            "📌 已启用 --best-by-pixel：best.pt / patience 按验证集 mean 像素误差（fitness=-mean_px）；"
+            "反向传播仍为 Ultralytics 默认 pose/box loss。"
+        )
+
+    if args.pixel_metrics_every > 0:
+        model.add_callback(
+            "on_fit_epoch_end",
+            _make_pixel_metrics_callback(
+                data_yaml, args.pixel_metrics_every, args.imgsz, conf=args.pixel_metrics_conf
+            ),
+        )
+
+    model.train(
+        task="pose",
+        data=data_yaml,
+        epochs=args.epochs,
+        imgsz=args.imgsz,
+        batch=args.batch,
+        name=args.name,
+        project=args.project,
+        exist_ok=True,
+        save=True,
+        save_period=5,
+        device=device,
+        workers=args.workers,
+        lr0=0.0001,
+        lrf=0.01,
+        optimizer="AdamW",
+        momentum=0.937,
+        weight_decay=0.001,
+        warmup_epochs=0,
+        warmup_momentum=0.8,
+        warmup_bias_lr=0.1,
+        hsv_h=0.015,
+        hsv_s=0.7,
+        hsv_v=0.4,
+        degrees=5.0,
+        translate=0.0,
+        scale=0.2,
+        shear=0.0,
+        perspective=0.0000,
+        flipud=0.0,
+        fliplr=0.5,
+        mosaic=0.0,
+        mixup=0.0,
+        copy_paste=0.0,
+        box=6,
+        cls=0.5,
+        dfl=1.5,
+        pose=18.0,
+        kobj=0.5,
+        freeze=0,
+        seed=42,
+        verbose=True,
+        amp=use_amp,
+        patience=100,
+        cos_lr=True,
+    )
+
+    print("\n✅ 训练完成！")
+    print(f"📁 best: {args.project}/{args.name}/weights/best.pt")
+    print(f"📁 last: {args.project}/{args.name}/weights/last.pt")
+    print("📊 仅看像素误差可运行: python pose_pixel_metrics.py --model <best.pt> --data <yaml> --imgsz", args.imgsz)
+
+    if args.export_onnx:
+        h, w = args.onnx_imgsz
+        print(f"📦 导出 ONNX imgsz=[{h}, {w}] ...")
+        model.export(format="onnx", imgsz=[h, w], simplify=True, opset=17, dynamic=False)
+        print("✅ ONNX 完成")
+
+
+if __name__ == "__main__":
+    main()

triangle_target.py

@@ -22,6 +22,143 @@ def _log(msg):
         pass
 
 
+def _read_triangle_direction_cfg():
+    """读取 config 中三角形方向/中心距校验参数。"""
+    try:
+        import config as cfg
+        return {
+            "enable": bool(getattr(cfg, "TRIANGLE_DIRECTION_VALIDATE_ENABLE", True)),
+            "min_pass": int(getattr(cfg, "TRIANGLE_DIRECTION_MIN_PASS", 3)),
+            "dot_min": float(getattr(cfg, "TRIANGLE_DIRECTION_DOT_MIN", 0.0)),
+            "to_center_dot_min": float(
+                getattr(cfg, "TRIANGLE_DIRECTION_TO_CENTER_DOT_MIN", 0.35)
+            ),
+            "center_dist_enable": bool(
+                getattr(cfg, "TRIANGLE_CENTER_DISTANCE_VALIDATE_ENABLE", True)
+            ),
+            "center_dist_tol": float(
+                getattr(cfg, "TRIANGLE_CENTER_DISTANCE_RATIO_TOL", 0.45)
+            ),
+        }
+    except Exception:
+        return {
+            "enable": True,
+            "min_pass": 3,
+            "dot_min": 0.0,
+            "to_center_dot_min": 0.35,
+            "center_dist_enable": True,
+            "center_dist_tol": 0.45,
+        }
+
+
+def _quad_combo_orient_penalty(cands_4):
+    """
+    四点组合评分用的方向惩罚（原 _score_quad 内 orient_pen 逻辑）。
+    TRIANGLE_DIRECTION_VALIDATE_ENABLE=False 时调用方应跳过（不加罚）。
+    """
+    orient_pen = 0.0
+    orient_vote = []
+    for c in cands_4:
+        cen = np.array(c["center_px"], dtype=np.float32)
+        rpt = np.array(c["right_pt"], dtype=np.float32)
+        vx = float(cen[0] - rpt[0])
+        vy = float(cen[1] - rpt[1])
+        if abs(vx) < 1e-6 or abs(vy) < 1e-6:
+            orient_pen += 1.0
+            orient_vote.append(None)
+            continue
+        if abs(vx) < abs(vy) * 0.15 or abs(vy) < abs(vx) * 0.15:
+            orient_pen += 0.5
+        if vx > 0 and vy > 0:
+            orient_vote.append(0)
+        elif vx < 0 and vy > 0:
+            orient_vote.append(1)
+        elif vx > 0 and vy < 0:
+            orient_vote.append(2)
+        else:
+            orient_vote.append(3)
+    valid_votes = [v for v in orient_vote if v is not None]
+    if valid_votes:
+        from collections import Counter
+        vc = Counter(valid_votes)
+        orient_pen += max(0, max(vc.values()) - 1) * 0.8
+    return orient_pen
+
+
+def _marker_inward_unit(marker):
+    """从直角顶点指向三角内部的单位向量；marker['center'] 为直角顶点。"""
+    right = np.array(marker["center"], dtype=np.float64)
+    corners = marker.get("corners")
+    if not corners or len(corners) < 3:
+        return None
+    cen = np.mean(np.array(corners, dtype=np.float64), axis=0)
+    inv = cen - right
+    n = float(np.linalg.norm(inv))
+    if n < 1e-6:
+        return None
+    return inv / n
+
+
+def _validate_triangle_direction(marker_centers, tri_markers, cfg):
+    """
+    校验：四角到候选靶心距离近似一致；各真实黑三角朝向靶心。
+    仅统计 tri_markers 中真实检出的角（不含几何补全的虚拟点）。
+    Returns:
+        (ok: bool, reason: str)
+    """
+    if not cfg.get("enable", True):
+        return True, ""
+
+    pts = np.array(marker_centers, dtype=np.float64).reshape(-1, 2)
+    if len(pts) < 3:
+        return True, ""
+
+    quad_center = np.mean(pts, axis=0)
+
+    if cfg.get("center_dist_enable", True) and len(pts) >= 3:
+        dists = np.linalg.norm(pts - quad_center, axis=1)
+        mean_d = float(np.mean(dists))
+        if mean_d > 1e-6:
+            ratio = (float(np.max(dists)) - float(np.min(dists))) / mean_d
+            tol = float(cfg.get("center_dist_tol", 0.45))
+            if ratio > tol:
+                return False, f"center_dist_ratio={ratio:.2f}>{tol:.2f}"
+
+    dot_need = max(
+        float(cfg.get("dot_min", 0.0)),
+        float(cfg.get("to_center_dot_min", 0.35)),
+    )
+    pass_n = 0
+    check_n = 0
+    for m in tri_markers or []:
+        if m.get("center") is None:
+            continue
+        check_n += 1
+        right = np.array(m["center"], dtype=np.float64)
+        to_center = quad_center - right
+        nc = float(np.linalg.norm(to_center))
+        if nc < 1e-6:
+            continue
+        inward = _marker_inward_unit(m)
+        if inward is None:
+            continue
+        dot_tc = float(np.dot(inward, to_center / nc))
+        if dot_tc >= dot_need:
+            pass_n += 1
+
+    if check_n == 0:
+        return True, ""
+
+    min_pass = int(cfg.get("min_pass", 3))
+    min_pass = max(1, min(min_pass, check_n))
+    if pass_n < min_pass:
+        return False, (
+            f"direction_pass={pass_n}/{check_n} need>={min_pass} "
+            f"(dot>={dot_need:.2f})"
+        )
+    return True, ""
+
+
 def _gray_suppress_bright_by_v(img_rgb, v_above: int):
     """
     RGB 输入：在 HSV 的 V 上，将亮度 >= v_above 的像素灰度置为 255。
@@ -224,7 +361,7 @@ def detect_triangle_markers(
         blackhat_kernel_frac = 0.018
     try:
         import config as _tcfg
-        _timing_log = bool(getattr(_tcfg, "TRIANGLE_TIMING_LOG", True))
+        _timing_log = bool(getattr(_tcfg, "ARCHERY_TIMING_ENABLE", True)) and bool(getattr(_tcfg, "TRIANGLE_TIMING_LOG", True))
     except Exception:
         _timing_log = True
 
@@ -622,6 +759,8 @@ def detect_triangle_markers(
         bot_pair = sorted(by_y[2:], key=lambda i: pts_4[i][0])
         return top_pair[0], bot_pair[0], bot_pair[1], top_pair[1]
 
+    _dir_cfg_combo = _read_triangle_direction_cfg()
+
     def _score_quad(cands_4):
         pts = [np.array(c["center_px"]) for c in cands_4]
         legs = [c["avg_leg"] for c in cands_4]
@@ -641,7 +780,13 @@ def detect_triangle_markers(
         med_l = float(np.median(legs))
         leg_dev = max(abs(l - med_l) / (med_l + 1e-6) for l in legs)
 
-        score = (diag_ratio - 1.0) * 3.0 + (h_ratio - 1.0) + (v_ratio - 1.0) + leg_dev * 2.0
+        orient_pen = (
+            _quad_combo_orient_penalty(cands_4)
+            if _dir_cfg_combo.get("enable", True)
+            else 0.0
+        )
+
+        score = (diag_ratio - 1.0) * 3.0 + (h_ratio - 1.0) + (v_ratio - 1.0) + leg_dev * 2.0 + orient_pen
         return score, (tl, bl, br, tr)
 
     assigned = None
@@ -932,6 +1077,8 @@ def _assign_marker_ids_from_filtered(filtered, verbose=True):
         bot_pair = sorted(by_y[2:], key=lambda i: pts_4[i][0])
         return top_pair[0], bot_pair[0], bot_pair[1], top_pair[1]
 
+    _dir_cfg_combo = _read_triangle_direction_cfg()
+
     def _score_quad(cands_4):
         pts = [np.array(c["center_px"]) for c in cands_4]
         legs = [c["avg_leg"] for c in cands_4]
@@ -947,7 +1094,12 @@ def _assign_marker_ids_from_filtered(filtered, verbose=True):
         v_ratio = max(s_left, s_right) / (min(s_left, s_right) + 1e-6)
         med_l = float(np.median(legs))
         leg_dev = max(abs(l - med_l) / (med_l + 1e-6) for l in legs)
-        score = (diag_ratio - 1.0) * 3.0 + (h_ratio - 1.0) + (v_ratio - 1.0) + leg_dev * 2.0
+        orient_pen = (
+            _quad_combo_orient_penalty(cands_4)
+            if _dir_cfg_combo.get("enable", True)
+            else 0.0
+        )
+        score = (diag_ratio - 1.0) * 3.0 + (h_ratio - 1.0) + (v_ratio - 1.0) + leg_dev * 2.0 + orient_pen
         return score, (tl, bl, br, tr)
 
     assigned = None
@@ -1113,7 +1265,7 @@ def try_triangle_scoring(
 
     try:
         import config as _cfg_tl
-        _try_timing_log = bool(getattr(_cfg_tl, "TRIANGLE_TIMING_LOG", True))
+        _try_timing_log = bool(getattr(_cfg_tl, "ARCHERY_TIMING_ENABLE", True)) and bool(getattr(_cfg_tl, "TRIANGLE_TIMING_LOG", True))
         _crop_min_side = int(getattr(_cfg_tl, "TRIANGLE_CROP_ROI_MIN_SIDE_PX", 64))
     except Exception:
         _try_timing_log = True
@@ -1733,6 +1885,21 @@ def try_triangle_scoring(
             "is_virtual": bool(_is_virtual),
         })
 
+    # ---------- 方向 / 中心距校验（config.TRIANGLE_DIRECTION_*） ----------
+    _dir_cfg = _read_triangle_direction_cfg()
+    _dir_ok, _dir_reason = _validate_triangle_direction(
+        marker_centers, tri_markers, _dir_cfg
+    )
+    if not _dir_ok:
+        _log(f"[TRI] 方向校验失败: {_dir_reason}")
+        if _try_timing_log:
+            _log(
+                f"[TRI] timing_ms(try_triangle): {_tri_yolo_part} "
+                f"geometry={(time.perf_counter() - _t_seg) * 1000:.1f} "
+                f"total_try={(time.perf_counter() - _t_try0) * 1000:.1f} (方向校验失败)"
+            )
+        return out
+
     # ---------- 结果有效性校验（防 nan/inf 与退化角点） ----------
     try:
         import config as _cfg

version.py

@@ -4,7 +4,8 @@
 应用版本号
 每次 OTA 更新时，只需要更新这个文件中的版本号
 """
-VERSION = '1.2.12'
+VERSION = '1.2.15.1'
+
 
 # 1.2.0 开始使用C++编译成.so，替换部分代码
 # 1.2.1 ota使用加密包
@@ -19,10 +20,10 @@ VERSION = '1.2.12'
 # 1.2.10 config formal
 # 1.2.11 增加三角形的单应性算法，适配对应的靶纸
 # 1.2.110 关掉了黑色三角形算法，只用于测试
-
-
-
-
+# 1.2.13 修改wifi连接
+# 1.2.14 修改了icc登录部分
+# 1.2.15.1 增加了标靶判断 20 40 
+# 1.2.16.1 增加激光校准，三角形方向判断，时间开关
 
 
 

vision.py

@@ -10,6 +10,7 @@ import os
 import math
 import threading
 import queue
+import time
 from maix import image
 import config
 from logger_manager import logger_manager
@@ -531,6 +532,9 @@ def detect_circle_v3(frame, laser_point=None, img_cv=None):
     if img_cv is None:
         img_cv = image.image2cv(frame, False, False)
     logger = logger_manager.logger
+    _timing_on = bool(getattr(config, "VISION_TIMING_ENABLE", True))
+    _t0 = time.perf_counter() if _timing_on else None
+    _t1 = _t2 = _t3 = _t4 = _t5 = None
     from datetime import datetime
     logger.debug(f"[detect_circle_v3] begin {datetime.now()}")
     # -- 1. 缩图加速（与三角形路径保持一致）
@@ -554,6 +558,8 @@ def detect_circle_v3(frame, laser_point=None, img_cv=None):
     ellipse_params = None
     
     logger.debug(f"[detect_circle_v3] step 1 fin {datetime.now()}")
+    if _timing_on:
+        _t1 = time.perf_counter()
 
     # -- 2. HSV + 黄色掩码
     hsv = cv2.cvtColor(img_det, cv2.COLOR_RGB2HSV)
@@ -567,6 +573,9 @@ def detect_circle_v3(frame, laser_point=None, img_cv=None):
     mask_yellow = cv2.morphologyEx(mask_yellow, cv2.MORPH_CLOSE, kernel)
 
     logger.debug(f"[detect_circle_v3] step 2 fin {datetime.now()}")
+    if _timing_on:
+        _t2 = time.perf_counter()
+        _t3 = time.perf_counter()
 
     # -- 3. 红色掩码：在循环外只算一次
     mask_red = cv2.bitwise_or(
@@ -593,6 +602,9 @@ def detect_circle_v3(frame, laser_point=None, img_cv=None):
             red_candidates.append({"center": (int(xr), int(yr)), "radius": int(rr)})
 
     logger.debug(f"[detect_circle_v3] step 3 fin {datetime.now()}")
+    if _timing_on:
+        _t3 = time.perf_counter()
+        _t4 = time.perf_counter()
 
     # -- 4. 黄色轮廓循环（复用上面的红色候选列表）
     contours_yellow, _ = cv2.findContours(mask_yellow, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
@@ -642,6 +654,9 @@ def detect_circle_v3(frame, laser_point=None, img_cv=None):
             logger.debug("Debug -> 未找到匹配的红色圆圈，可能是误识别")
 
     logger.debug(f"[detect_circle_v3] step 4 fin {datetime.now()}")
+    if _timing_on:
+        _t4 = time.perf_counter()
+        _t5 = time.perf_counter()
 
     # -- 5. 选最佳目标，坐标还原到原始分辨率
     if valid_targets:
@@ -669,7 +684,20 @@ def detect_circle_v3(frame, laser_point=None, img_cv=None):
         ellipse_params = be
         best_radius1 = best_radius * 5
     result_img = image.cv2image(img_cv, False, False)
-    logger.debug(f"[detect_circle_v3] step 5 fin {datetime.now()}")
+    if _timing_on:
+        _t5 = time.perf_counter()
+        _t_all = (_t5 - _t0) * 1000
+        _ms1 = (_t1 - _t0) * 1000
+        _ms2 = (_t2 - _t1) * 1000
+        _ms3 = (_t3 - _t2) * 1000
+        _ms4 = (_t4 - _t3) * 1000
+        _ms5 = (_t5 - _t4) * 1000
+        logger.info(
+            f"[VISION timing] total={_t_all:.1f}ms  "
+            f"resize={_ms1:.1f}  hsv_yellow={_ms2:.1f}  "
+            f"red_mask={_ms3:.1f}  yellow_loop={_ms4:.1f}  "
+            f"select_cv2img={_ms5:.1f}"
+        )
     return result_img, best_center, best_radius, method, best_radius1, ellipse_params
 
 def estimate_distance(pixel_radius):
@@ -921,6 +949,51 @@ def start_save_shot_worker():
         logger.info("[VISION] 存图 worker 线程已启动")
 
 
+def enqueue_save_raw_shot(frame, shot_id=None, photo_dir=None):
+    """
+    异步保存射箭原图（无算法标注）。需 SAVE_IMAGE_ENABLED 且 SAVE_RAW_SHOT_IMAGE_ENABLED。
+    文件名：{photo_dir}/shot_{shot_id}_raw.jpg
+    """
+    if not getattr(config, "SAVE_RAW_SHOT_IMAGE_ENABLED", False):
+        return
+    if not getattr(config, "SAVE_IMAGE_ENABLED", True):
+        return
+    if not shot_id:
+        return
+    if photo_dir is None:
+        photo_dir = config.PHOTO_DIR
+
+    try:
+        img_cv = image.image2cv(frame, False, False)
+        img_copy = np.copy(img_cv)
+    except Exception as e:
+        logger = logger_manager.logger
+        if logger:
+            logger.error(f"[VISION] enqueue_save_raw_shot 复制图像失败: {e}")
+        return
+
+    def _job():
+        try:
+            try:
+                if photo_dir not in os.listdir("/root"):
+                    os.mkdir(photo_dir)
+            except Exception:
+                pass
+            filename = f"{photo_dir}/shot_{shot_id}_raw.jpg"
+            out = image.cv2image(img_copy, False, False)
+            out.save(filename)
+            logger = logger_manager.logger
+            if logger:
+                logger.info(f"[VISION] 已保存射箭原图: {filename}")
+            prune_old_images_in_dir(photo_dir, config.MAX_IMAGES, logger, "[VISION]")
+        except Exception as e:
+            logger = logger_manager.logger
+            if logger:
+                logger.error(f"[VISION] 保存射箭原图失败: {e}")
+
+    threading.Thread(target=_job, daemon=True).start()
+
+
 def enqueue_save_shot(result_img, center, radius, method, ellipse_params,
                       laser_point, distance_m, shot_id=None, photo_dir=None,
                       yolo_roi_xyxy=None):
@@ -1010,3 +1083,63 @@ def detect_target(frame, laser_point=None):
         logger.debug("[VISION] 使用传统黄色靶心检测")
     return detect_circle_v3(frame, laser_point)
 
+
+def sample_target_rgb_at_physical_radius(frame, target_center, target_radius_px, radius_cm=None, angles_deg=None, patch_half_px=None, black_thresh=None, timing=False):
+    """
+    在物方半径位置采样 RGB，判断黑/白靶。
+    返回: dict {ok, is_black, mean_rgb, samples, black_ratio, elapsed_ms}
+    """
+    logger = logger_manager.logger
+    if target_center is None or target_radius_px is None:
+        return {"ok": False, "reason": "no_target", "is_black": None, "elapsed_ms": 0.0}
+
+    radius_cm = float(radius_cm if radius_cm is not None else getattr(config, "TRIANGLE_SAMPLE_RADIUS_CM", 15.0))
+    angles_deg = tuple(angles_deg if angles_deg is not None else getattr(config, "TRIANGLE_SAMPLE_ANGLES_DEG", (0, 90, 180, 270)))
+    patch_half_px = int(patch_half_px if patch_half_px is not None else getattr(config, "TRIANGLE_SAMPLE_PATCH_HALF_PX", 2))
+    black_thresh = float(black_thresh if black_thresh is not None else getattr(config, "TRIANGLE_SAMPLE_BLACK_THRESH", 30.0))
+    timing_on = bool(timing) and bool(getattr(config, "TRIANGLE_SAMPLE_TIMING_ENABLE", True))
+    t0 = time.perf_counter() if timing_on else None
+
+    try:
+        img_cv = image.image2cv(frame, False, False)
+        h, w = img_cv.shape[:2]
+        cx, cy = float(target_center[0]), float(target_center[1])
+        scale = float(target_radius_px) / max(radius_cm, 1e-6)
+        samples = []
+        black_count = 0
+        for ang in angles_deg:
+            rad = math.radians(float(ang))
+            sx = int(round(cx + math.cos(rad) * radius_cm * scale))
+            sy = int(round(cy + math.sin(rad) * radius_cm * scale))
+            x0 = max(0, sx - patch_half_px)
+            y0 = max(0, sy - patch_half_px)
+            x1 = min(w, sx + patch_half_px + 1)
+            y1 = min(h, sy + patch_half_px + 1)
+            if x1 <= x0 or y1 <= y0:
+                continue
+            patch = img_cv[y0:y1, x0:x1]
+            mean_rgb = patch.reshape(-1, 3).mean(axis=0)
+            is_black = bool(np.all(mean_rgb < black_thresh))
+            black_count += 1 if is_black else 0
+            samples.append({"angle": float(ang), "xy": (sx, sy), "mean_rgb": tuple(float(v) for v in mean_rgb), "is_black": is_black})
+        black_ratio = float(black_count) / float(len(samples) or 1)
+        out = {
+            "ok": len(samples) > 0,
+            "is_black": black_ratio >= 0.5,
+            "mean_rgb": tuple(float(v) for v in (np.mean([s["mean_rgb"] for s in samples], axis=0) if samples else (0, 0, 0))),
+            "samples": samples,
+            "black_ratio": black_ratio,
+            "elapsed_ms": (time.perf_counter() - t0) * 1000.0 if timing_on else 0.0,
+        }
+        if logger:
+            logger.info(
+                f"[TRI-SAMPLE] radius_cm={radius_cm:.1f} black_thresh={black_thresh:.1f} "
+                f"black_ratio={black_ratio:.2f} is_black={out['is_black']} "
+                f"elapsed_ms={out['elapsed_ms']:.1f} samples={len(samples)}"
+            )
+        return out
+    except Exception as e:
+        if logger:
+            logger.error(f"[TRI-SAMPLE] 采样失败: {e}")
+        return {"ok": False, "reason": str(e), "is_black": None, "elapsed_ms": 0.0}
+

yolo_te.py

@@ -0,0 +1,267 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""Standalone live camera + single YOLO runner.
+
+不复用项目内的 `camera_manager` / `target_roi_yolo` / `config` / `logger_manager`。
+
+功能：
+- 独立初始化摄像头
+- 实时读取帧
+- 独立加载单个 YOLO 模型并推理
+- 画出检测框、ROI、FPS
+
+适用场景：
+- 单独验证一个模型是否能跑
+- 验证实时帧率
+- 验证 ROI 是否裁对
+- 不进入主业务射箭流程
+"""
+
+from __future__ import annotations
+
+import os
+import time
+from dataclasses import dataclass
+
+
+@dataclass
+class RunnerConfig:
+    camera_width: int = 640
+    camera_height: int = 480
+    model_path: str = "/root/model_278702.mud"
+    conf_th: float = 0.7
+    retry_conf_th: float = 0.5
+    class_ids: tuple = (0,)
+    merge_mode: str = "union"
+    coord_mode: str = "native"
+    roi_margin_frac: float = 0.11
+    min_box_side_px: int = 8
+
+
+def log(msg: str):
+    print(msg)
+
+
+class DummyLogger:
+    def info(self, msg):
+        log(msg)
+
+    def warning(self, msg):
+        log(msg)
+
+    def error(self, msg):
+        log(msg)
+
+
+class StandaloneYOLORunner:
+    def __init__(self, cfg: RunnerConfig):
+        self.cfg = cfg
+        self.logger = DummyLogger()
+        self._last_fps_t = time.perf_counter()
+        self._frames = 0
+        self._fps = 0.0
+        self._camera = None
+        self._det = None
+
+    def _import_maix(self):
+        try:
+            from maix import camera, image, nn
+            return camera, image, nn
+        except Exception as e:
+            raise RuntimeError(f"maix import failed: {e}")
+
+    def _init_camera(self):
+        camera, _, _ = self._import_maix()
+        if self._camera is not None:
+            return self._camera
+        try:
+            self._camera = camera.Camera(
+                width=self.cfg.camera_width,
+                height=self.cfg.camera_height,
+                format=camera.RGB888,
+            )
+        except Exception:
+            self._camera = camera.Camera(width=self.cfg.camera_width, height=self.cfg.camera_height)
+        return self._camera
+
+    def _load_detector(self, model_path: str):
+        _, _, nn = self._import_maix()
+        if not model_path or not os.path.isfile(model_path):
+            return None
+        return nn.YOLOv5(model=model_path, dual_buff=False)
+
+    @staticmethod
+    def _get_class_id(obj):
+        for key in ("class_id", "cls", "label", "category", "cat_id", "id"):
+            if hasattr(obj, key):
+                v = getattr(obj, key)
+                if v is None:
+                    continue
+                try:
+                    return int(float(v))
+                except Exception:
+                    pass
+        return None
+
+    @staticmethod
+    def _normalize_boxes(raw):
+        out = []
+        for o in raw or []:
+            if isinstance(o, (list, tuple)) and len(o) >= 6:
+                class Box:
+                    pass
+                b = Box()
+                b.x, b.y, b.w, b.h, b.score, b.class_id = map(float, o[:6])
+                out.append(b)
+            else:
+                out.append(o)
+        return out
+
+    def _det_to_xyxy(self, det, obj):
+        x = float(getattr(obj, "x", 0.0))
+        y = float(getattr(obj, "y", 0.0))
+        w = float(getattr(obj, "w", 0.0))
+        h = float(getattr(obj, "h", 0.0))
+        return x, y, x + w, y + h
+
+    def _run_detector(self, det, img, conf_th, class_ids):
+        if det is None:
+            return []
+        raw = det.detect(img, conf_th=conf_th)
+        objs = self._normalize_boxes(raw if raw is not None else [])
+        out = []
+        for o in objs:
+            cid = self._get_class_id(o)
+            if cid is not None and cid not in class_ids:
+                continue
+            out.append(o)
+        return out
+
+    def _calc_fps(self):
+        self._frames += 1
+        now = time.perf_counter()
+        dt = now - self._last_fps_t
+        if dt >= 1.0:
+            self._fps = self._frames / dt
+            self._frames = 0
+            self._last_fps_t = now
+        return self._fps
+
+    def _draw_text(self, img, lines):
+        try:
+            import cv2
+            y = 24
+            for line in lines:
+                cv2.putText(img, line, (10, y), cv2.FONT_HERSHEY_SIMPLEX, 0.55, (0, 255, 0), 1, cv2.LINE_AA)
+                y += 20
+        except Exception:
+            pass
+
+    def _clip_roi(self, x0, y0, x1, y1, w, h):
+        x0 = max(0, min(int(x0), w - 1))
+        y0 = max(0, min(int(y0), h - 1))
+        x1 = max(x0 + 1, min(int(x1), w))
+        y1 = max(y0 + 1, min(int(y1), h))
+        return x0, y0, x1, y1
+
+    def _merge_boxes(self, boxes):
+        if not boxes:
+            return None
+        x0 = min(b[0] for b in boxes)
+        y0 = min(b[1] for b in boxes)
+        x1 = max(b[2] for b in boxes)
+        y1 = max(b[3] for b in boxes)
+        return x0, y0, x1, y1
+
+    def _run_single_yolo(self, frame, img_cv):
+        h, w = int(img_cv.shape[0]), int(img_cv.shape[1])
+        if self._det is None:
+            self._det = self._load_detector(self.cfg.model_path)
+        det = self._det
+        if det is None:
+            return []
+
+        boxes = self._run_detector(det, frame, self.cfg.conf_th, self.cfg.class_ids)
+        if not boxes and self.cfg.retry_conf_th < self.cfg.conf_th:
+            boxes = self._run_detector(det, frame, self.cfg.retry_conf_th, self.cfg.class_ids)
+
+        xyxy = []
+        for obj in boxes:
+            x0, y0, x1, y1 = self._det_to_xyxy(det, obj)
+            if (x1 - x0) < self.cfg.min_box_side_px or (y1 - y0) < self.cfg.min_box_side_px:
+                continue
+            if self.cfg.coord_mode == "native":
+                x0, y0, x1, y1 = self._clip_roi(x0, y0, x1, y1, w, h)
+            xyxy.append((x0, y0, x1, y1))
+        return xyxy
+
+    def run(self):
+        _, image, _ = self._import_maix()
+        cam = self._init_camera()
+        log("[YOLOTE] standalone runner started")
+
+        while True:
+            try:
+                frame = cam.read()
+            except Exception as e:
+                log(f"[YOLOTE] camera read failed: {e}")
+                time.sleep(0.02)
+                continue
+
+            if frame is None:
+                time.sleep(0.01)
+                continue
+
+            try:
+                img_cv = image.image2cv(frame, False, False)
+            except Exception as e:
+                log(f"[YOLOTE] image2cv failed: {e}")
+                time.sleep(0.01)
+                continue
+
+            import cv2
+            t0 = time.perf_counter()
+            boxes = self._run_single_yolo(frame, img_cv)
+            t1 = time.perf_counter()
+
+            for i, (bx0, by0, bx1, by1) in enumerate(boxes):
+                cv2.rectangle(img_cv, (int(bx0), int(by0)), (int(bx1) - 1, int(by1) - 1), (0, 255, 0), 2)
+                cv2.putText(img_cv, f"B{i}", (int(bx0), max(0, int(by0) - 4)), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 1, cv2.LINE_AA)
+
+            fps = self._calc_fps()
+            self._draw_text(
+                img_cv,
+                [
+                    f"FPS: {fps:.1f}",
+                    f"YOLO: {(t1 - t0)*1000.0:.1f} ms",
+                    f"Boxes: {len(boxes)}",
+                    "Ctrl+C to exit",
+                ],
+            )
+
+            try:
+                frame_out = image.cv2image(img_cv, False, False)
+                if hasattr(cam, "show"):
+                    cam.show(frame_out)
+                else:
+                    try:
+                        frame_out.show()
+                    except Exception:
+                        pass
+            except Exception as e:
+                log(f"[YOLOTE] show failed: {e}")
+
+            time.sleep(0.001)
+
+
+def main():
+    cfg = RunnerConfig()
+    runner = StandaloneYOLORunner(cfg)
+    try:
+        runner.run()
+    except KeyboardInterrupt:
+        log("[YOLOTE] interrupted")
+
+
+if __name__ == "__main__":
+    main()