简化车辆检测后的边界框过滤

src/distributed_hierarchical_attentive/main/tester/multi_lane_test.py

@@ -179,15 +179,10 @@ def detect_hog(self, image):
         gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
         rects, weights = self.hog.detectMultiScale(gray, winStride=(4, 4),
                                                    padding=(8, 8), scale=1.05)
-
-        bboxes = []
-        scores = []
-        for (x, y, w, h) in rects:
-            # 过滤掉太小的检测
-            if w > 50 and h > 50:
-                bboxes.append([x, y, x+w, y+h])
-                scores.append(1.0)  # HOG不返回置信度分数
-
+
+        # 使用列表推导式简化过滤和转换
+        bboxes = [[x, y, x + w, y + h] for (x, y, w, h) in rects if w > 50 and h > 50]
+        scores = [1.0] * len(bboxes)  # HOG不返回置信度分数
         return np.array(bboxes), np.array(scores)
 
     def detect_yolo(self, image):