[CF]提交文件

runs/detect/train36/results.csv

@@ -88,3 +88,14 @@ epoch,time,train/box_loss,train/cls_loss,train/dfl_loss,metrics/precision(B),met
 87,602.206,0.44036,0.3477,0.80881,0.97052,0.70044,0.84557,0.62866,1.95032,1.14522,3.08449,0.0002972,0.0002972,0.0002972
 88,610.04,0.34605,0.29355,0.82656,0.97052,0.70044,0.84557,0.62866,1.95032,1.14522,3.08449,0.0002774,0.0002774,0.0002774
 89,618.104,0.32283,0.27712,0.78311,0.94286,0.70219,0.83861,0.6253,1.94975,1.17272,3.1016,0.0002576,0.0002576,0.0002576
+90,627.396,0.31937,0.2843,0.81442,0.94286,0.70219,0.83861,0.6253,1.94975,1.17272,3.1016,0.0002378,0.0002378,0.0002378
+91,636.531,0.40592,0.3456,0.76639,0.94286,0.70219,0.83861,0.6253,1.94975,1.17272,3.1016,0.000218,0.000218,0.000218
+92,645.424,0.32162,0.25546,0.70747,1,0.70074,0.8321,0.61851,1.94364,1.07547,3.15414,0.0001982,0.0001982,0.0001982
+93,653.162,0.32978,0.27483,0.7712,1,0.70074,0.8321,0.61851,1.94364,1.07547,3.15414,0.0001784,0.0001784,0.0001784
+94,660.746,0.33426,0.27878,0.85918,1,0.70074,0.8321,0.61851,1.94364,1.07547,3.15414,0.0001586,0.0001586,0.0001586
+95,668.458,0.28988,0.27119,0.77331,0.99076,0.68085,0.83074,0.61428,1.94021,0.97339,3.19668,0.0001388,0.0001388,0.0001388
+96,676.066,0.33953,0.28199,0.79338,0.99076,0.68085,0.83074,0.61428,1.94021,0.97339,3.19668,0.000119,0.000119,0.000119
+97,683.81,0.32111,0.27168,0.82616,0.8869,0.7234,0.82786,0.61017,1.92784,0.89589,3.21123,9.92e-05,9.92e-05,9.92e-05
+98,691.47,0.32248,0.28341,0.71945,0.8869,0.7234,0.82786,0.61017,1.92784,0.89589,3.21123,7.94e-05,7.94e-05,7.94e-05
+99,699.162,0.33571,0.27017,0.72974,0.8869,0.7234,0.82786,0.61017,1.92784,0.89589,3.21123,5.96e-05,5.96e-05,5.96e-05
+100,707.196,0.33638,0.25993,0.77413,0.88668,0.7234,0.82078,0.60787,1.92305,0.88185,3.22406,3.98e-05,3.98e-05,3.98e-05

src/pic.py

@@ -20,7 +20,7 @@ while cap.isOpened():
                 detected_object = frame[y1:y2, x1:x2]
                 save_path = os.path.join('detected_objects', f'detected_object_{target_class}_{cap.get(cv2.CAP_PROP_POS_FRAMES)}.jpg')
                 cv2.imwrite(save_path, detected_object)
-                print(f"截图保存至：{save_path}")
+                print(f"保存：{save_path}")
     annotated_frame = results[0].plot()
     cv2.imshow('YOLOv8 Detection', annotated_frame)
     if cv2.waitKey(1) & 0xFF == ord(' '):

src/show.py

@@ -0,0 +1,42 @@
+import cv2
+from PIL import Image
+import os
+
+folder_path = 'C:/workspace/le-yolo/data/images/train'
+for filename in os.listdir(folder_path):
+    if filename.lower().endswith(('.png', '.jpg', '.jpeg', '.bmp', '.gif')):
+        img_path = os.path.join(folder_path, filename)
+        img = cv2.imread(img_path)
+        height, width, _ = img.shape
+
+
+file_path = 'C:/workspace/le-yolo/data/labels/aut'
+txt_files = os.listdir(file_path)
+for txt_file in txt_files:
+    file_path = os.path.join(folder_path, txt_file)
+    with open(file_path, "r", ) as f:
+        lines = f.readlines()
+    for line in lines:
+        # 解析标注信息
+        parts = line.strip().split()
+        category_id = int(parts[0])
+        category_name = 'person'
+        x_center = float(parts[1])
+        y_center = float(parts[2])
+        w = float(parts[3])
+        h = float(parts[4])
+        # 转换为绝对坐标
+        x1 = int((x_center - w / 2) * width)
+        y1 = int((y_center - h / 2) * height)
+        x2 = int((x_center + w / 2) * width)
+        y2 = int((y_center + h / 2) * height)
+        # 在图片上绘制矩形框
+        cv2.rectangle(img, (x1, y1), (x2, y2), (255, 0, 0), 2)
+        # 在图片上绘制标注类别
+        cv2.putText(img, category_name, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (255, 0, 0), 2)
+
+cv2.imshow('Annotated Image', img)
+cv2.waitKey(0)
+cv2.destroyAllWindows()
+
+cv2.imwrite('annotated_image.jpg', img)

src/test.py

@@ -1,3 +1,3 @@
 from ultralytics import YOLO
-model = YOLO(r"C:\workspace\le-yolo\runs\detect\train35\weights\best.pt")
+model = YOLO(r"C:\workspace\le-yolo\runs\detect\train36\weights\best.pt")
 results = model.predict("../res/4.mp4", show=True, save=True)