feat: 2.18.9 靶规格YOLO分类重构

- 换用 model_325964（class_id: 0=40cm, 1=靶心, 2=20cm） - class_id=1 解析阶段直接过滤，不参与上报选择 - 上报仅在有效靶规格(20/40)中取置信度最高的一条 - 0 结果时按 RETRY_CONF_TH 自动重检一次 - 增加检测调试打印（raw/detections/上报/忽略计数）
2026-10-22 18:13:47 +08:00
parent 2459bb3b67
commit 496085cb93
6 changed files with 131 additions and 57 deletions
@@ -1,6 +1,6 @@
 id: t11
 name: t11
-version: 2.18.6
+version: 2.18.9
 author: t11
 icon: ''
 desc: t11
@@ -18,8 +18,8 @@ files:
  - laser_manager.py
  - logger_manager.py
  - main.py
-  - model_317828.cvimodel
-  - model_317828.mud
+  - model_325964.cvimodel
+  - model_325964.mud
  - network.py
  - ota_curl.sh
  - ota_manager.py
@@ -262,14 +262,18 @@ TRIANGLE_SAMPLE_PATCH_HALF_PX = 2
 # 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 = False

-# YOLO target size classification: class 0=20cm, class 1=40cm.
+# YOLO target size classification: class_id → label，下标即 class_id：0=40cm, 1=0(靶心), 2=20cm.
 TARGET_CLASS_YOLO_ENABLE = True
-TARGET_CLASS_YOLO_MODEL_PATH = APP_DIR + "/model_317828.mud"
-TARGET_CLASS_YOLO_LABELS = (20, 40)
-TARGET_CLASS_YOLO_CONF_TH = 0.66
-TARGET_CLASS_YOLO_IOU_TH = 0.45
-TARGET_CLASS_YOLO_RETRY_ON_EMPTY = False
+TARGET_CLASS_YOLO_MODEL_PATH = APP_DIR + "/model_325964.mud"
+TARGET_CLASS_YOLO_LABELS = (40, 0, 20)
+# 只有映射结果落在该集合内才用于上报（如 class 1→label 0 直接忽略）
+TARGET_CLASS_YOLO_VALID_LABELS = (20, 40)
+# 解析阶段直接丢弃的 class_id（1=靶心），不参与任何选择/上报
+TARGET_CLASS_YOLO_IGNORE_CLASS_IDS = (1,)
+TARGET_CLASS_YOLO_CONF_TH = 0.5
+# 首次检测 0 结果时用该阈值重新检测一次（≥CONF_TH 时按 CONF_TH 同参重跑）
 TARGET_CLASS_YOLO_RETRY_CONF_TH = 0.25
+TARGET_CLASS_YOLO_IOU_TH = 0.45
 TARGET_CLASS_YOLO_PRELOAD_ON_BOOT = True

 # ── 第二段 YOLO：仅在 Stage1 裁切出的靶环图上推理（与合成 stage2 训练数据一致）→ 子框内传统算法取直角点 ──
@@ -1,7 +1,7 @@

 [basic]
 type = cvimodel
-model = model_317828.cvimodel
+model = model_325964.cvimodel

 [extra]
 model_type = yolov5
@@ -9,5 +9,5 @@ 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 = 20, 40
+labels = 40, 00, 20

@@ -310,7 +310,9 @@ def _det_obj_score(o):


 def try_get_target_class_from_yolo(maix_frame, logger=None):
-    """Classify the current target as 20cm or 40cm; return None if unknown."""
+    """调用 YOLO 检测：class_id∈IGNORE_CLASS_IDS(默认 1) 的框在解析时直接过滤；
+    detections 输出其余全部框（不按置信度过滤）；
+    上报字段 label/class_id/confidence 仅在有效靶规格(20/40)中取置信度最高的一条。"""
    try:
        import config as cfg
    except Exception:
@@ -333,64 +335,132 @@ def try_get_target_class_from_yolo(maix_frame, logger=None):

    conf_th = float(getattr(cfg, "TARGET_CLASS_YOLO_CONF_TH", 0.5))
    iou_th = float(getattr(cfg, "TARGET_CLASS_YOLO_IOU_TH", 0.45))
-    labels = getattr(cfg, "TARGET_CLASS_YOLO_LABELS", (20, 40))
+
+    # class_id → 靶规格标签，下标即 class_id：0→40, 1→0, 2→20
+    labels = getattr(cfg, "TARGET_CLASS_YOLO_LABELS", (40, 0, 20))
    if isinstance(labels, str):
        labels = tuple(x.strip() for x in labels.split(",") if x.strip())
    labels = tuple(labels)

-    def _detect(threshold):
+    # 只有映射到这些标签的检测才算有效靶规格（class 1→label 0 直接忽略）
+    valid_labels = getattr(cfg, "TARGET_CLASS_YOLO_VALID_LABELS", (20, 40))
+    if isinstance(valid_labels, str):
+        valid_labels = tuple(
+            int(float(x.strip())) for x in valid_labels.split(",") if x.strip()
+        )
+    valid_labels = frozenset(int(float(v)) for v in valid_labels)
+
+    # 解析阶段就丢弃这些 class_id（1=靶心），不进入 detections，也不参与上报选择
+    ignore_class_ids = getattr(cfg, "TARGET_CLASS_YOLO_IGNORE_CLASS_IDS", (1,))
+    if isinstance(ignore_class_ids, int):
+        ignore_class_ids = (ignore_class_ids,)
+    if isinstance(ignore_class_ids, str):
+        ignore_class_ids = tuple(
+            int(float(x.strip())) for x in ignore_class_ids.split(",") if x.strip()
+        )
+    ignore_class_ids = frozenset(int(float(v)) for v in ignore_class_ids)
+    n_ignored = [0]
+
+    def _run(threshold):
        try:
-            raw = detector.detect(maix_frame, conf_th=threshold, iou_th=iou_th)
+            return detector.detect(maix_frame, conf_th=threshold, iou_th=iou_th)
        except Exception as exc:
+            print(f"[YOLO-TARGET] detect 异常: {exc}")
            if logger:
                logger.warning(f"[YOLO-TARGET] detect 异常: {exc}")
-            return []
-        return _normalize_objs(raw if raw is not None else [])
+            return None

-    def _candidates(objs):
-        found = []
-        for obj in objs:
-            class_id = _det_obj_class_id(obj)
-            if class_id is None or class_id < 0 or class_id >= len(labels):
-                continue
-            try:
-                label = int(float(labels[class_id]))
-            except (TypeError, ValueError):
-                continue
-            if label in (20, 40):
-                found.append((label, class_id, _det_obj_score(obj)))
-        return found
-
-    objects = _detect(conf_th)
-    candidates = _candidates(objects)
-    if logger and objects:
-        logger.info(
-            "[YOLO-TARGET] 原始框=%d, 解析类别=%s"
-            % (
-                len(objects),
-                [(_det_obj_class_id(o), _det_obj_score(o)) for o in objects[:8]],
+    def _obj_box(o):
+        try:
+            return (
+                round(float(o.x), 1),
+                round(float(o.y), 1),
+                round(float(o.w), 1),
+                round(float(o.h), 1),
            )
-        )
-    if not candidates and bool(
-        getattr(cfg, "TARGET_CLASS_YOLO_RETRY_ON_EMPTY", False)
-    ):
-        retry_th = float(getattr(cfg, "TARGET_CLASS_YOLO_RETRY_CONF_TH", conf_th))
-        if 0 < retry_th < conf_th:
-            candidates = _candidates(_detect(retry_th))
+        except (TypeError, ValueError, AttributeError):
+            return None

-    if not candidates:
+    def _parse(raw_result):
+        out = []
+        n_ignored[0] = 0
+        for obj in _normalize_objs(raw_result if raw_result is not None else []):
+            class_id = _det_obj_class_id(obj)
+            if class_id is None or class_id in ignore_class_ids:
+                n_ignored[0] += 1
+                continue
+            label = None
+            if 0 <= class_id < len(labels):
+                try:
+                    label = int(float(labels[class_id]))
+                except (TypeError, ValueError):
+                    label = None
+            out.append(
+                {
+                    "class_id": class_id,
+                    "label": label,
+                    "confidence": round(_det_obj_score(obj), 4),
+                    "box_xywh": _obj_box(obj),
+                }
+            )
+        return out
+
+    # 第 1 次检测
+    raw = _run(conf_th)
+    detections = _parse(raw)
+    print(f"[YOLO-TARGET] 第1次 conf_th={conf_th} iou_th={iou_th} raw={raw!r}")
+    print(
+        f"[YOLO-TARGET] 检测结果={detections}"
+        f"（已忽略 class_id∈{tuple(sorted(ignore_class_ids))} 共{n_ignored[0]} 条）"
+    )
+
+    # 0 结果时重新检测一次（阈值取更小值，仍为 0 则同参数重跑）
+    if not detections:
+        retry_th = float(getattr(cfg, "TARGET_CLASS_YOLO_RETRY_CONF_TH", 0.25) or 0)
+        if not (0 < retry_th < conf_th):
+            retry_th = conf_th
+        print(f"[YOLO-TARGET] 0 结果，重新检测 第2次 conf_th={retry_th}")
+        raw_retry = _run(retry_th)
+        detections_retry = _parse(raw_retry)
+        print(f"[YOLO-TARGET] 第2次 raw={raw_retry!r}")
+        print(
+            f"[YOLO-TARGET] 检测结果={detections_retry}"
+            f"（已忽略 class_id∈{tuple(sorted(ignore_class_ids))} 共{n_ignored[0]} 条）"
+        )
+        if detections_retry:
+            raw, detections = raw_retry, detections_retry
+
+    if logger:
+        logger.info(f"[YOLO-TARGET] 检测结果={detections}")
+
+    if not detections:
        if logger:
-            logger.warning("[YOLO-TARGET] 当前帧未识别到 20/40，按未知处理")
+            logger.warning("[YOLO-TARGET] 未检测到目标（已重检 1 次）")
        return None

-    label, class_id, confidence = max(candidates, key=lambda item: item[2])
-    result = {"label": label, "class_id": class_id, "confidence": confidence}
-    if logger:
-        logger.info(
-            f"[YOLO-TARGET] 当前帧分类={label}, class_id={class_id}, "
-            f"conf={confidence:.3f}"
-        )
-    return result
+    # 上报口径：只在有效靶规格(20/40)里选置信度最高的一条，忽略 label=0 等无效类
+    valid = [d for d in detections if d["label"] in valid_labels]
+    if valid:
+        picked = max(valid, key=lambda d: d["confidence"])
+    else:
+        picked = {
+            "class_id": None,
+            "label": None,
+            "confidence": None,
+            "box_xywh": None,
+        }
+
+    print(
+        f"[YOLO-TARGET] 上报={picked['label']}, class_id={picked['class_id']}, "
+        f"conf={picked['confidence']}, 有效候选={[(d['label'], d['confidence']) for d in valid]}"
+    )
+
+    return {
+        "label": picked["label"],
+        "class_id": picked["class_id"],
+        "confidence": picked["confidence"],
+        "detections": detections,
+    }


 def _det_to_src_xyxy(o, coord_mode: str, src_w: int, src_h: int, net_w: int, net_h: int):
@@ -4,6 +4,6 @@
 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 = '2.18.6'
+VERSION = '2.18.9'