最新推送

zjsh_code/muju_cls/train/3cls.yaml

@@ -0,0 +1,30 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+# YOLO11-cls image classification model. For Usage examples see https://docs.ultralytics.com/tasks/classify
+
+# Parameters
+nc: 3 # number of classes
+scales: # model compound scaling constants, i.e. 'model=yolo11n-cls.yaml' will call yolo11-cls.yaml with scale 'n'
+  # [depth, width, max_channels]
+  n: [0.50, 0.25, 1024] # summary: 151 layers, 1633584 parameters, 1633584 gradients, 3.3 GFLOPs
+  s: [0.50, 0.50, 1024] # summary: 151 layers, 5545488 parameters, 5545488 gradients, 12.2 GFLOPs
+  m: [0.50, 1.00, 512] # summary: 187 layers, 10455696 parameters, 10455696 gradients, 39.7 GFLOPs
+  l: [1.00, 1.00, 512] # summary: 309 layers, 12937104 parameters, 12937104 gradients, 49.9 GFLOPs
+  x: [1.00, 1.50, 512] # summary: 309 layers, 28458544 parameters, 28458544 gradients, 111.1 GFLOPs
+
+# YOLO11n backbone
+backbone:
+  # [from, repeats, module, args]
+  - [-1, 1, Conv, [64, 3, 2]] # 0-P1/2
+  - [-1, 1, Conv, [128, 3, 2]] # 1-P2/4
+  - [-1, 2, C3k2, [256, False, 0.25]]
+  - [-1, 1, Conv, [256, 3, 2]] # 3-P3/8
+  - [-1, 2, C3k2, [512, False, 0.25]]
+  - [-1, 1, Conv, [512, 3, 2]] # 5-P4/16
+  - [-1, 2, C3k2, [512, True]]
+  - [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32
+  - [-1, 2, C3k2, [1024, True]]
+  - [-1, 2, C2PSA, [1024]] # 9
+
+# YOLO11n head
+head:
+  - [-1, 1, Classify, [nc]] # Classify

zjsh_code/muju_cls/train/train_3cls_main.py

@@ -0,0 +1,18 @@
+from ultralytics import YOLO
+
+if __name__ == '__main__':
+    model = YOLO(r'3cls.yaml')
+    results = model.train(
+        data='/home/dy/dataset/muju',
+        epochs=300,
+        imgsz=640,
+        batch=4,
+        workers=5,
+        device='0',
+        project='runs/train/muju_cls',
+        name='exp_cls',
+        exist_ok=False,
+        optimizer='AdamW',
+        lr0=0.0005,
+        patience=0,
+    )

zjsh_code/muju_cls/val/main_pc.py

@@ -0,0 +1,73 @@
+import cv2
+from ultralytics import YOLO
+
+# ---------------------------
+# 配置路径（请按需修改）
+# ---------------------------
+MODEL_PATH = "muju.pt"
+IMAGE_PATH = "./test_img/class0.png"  # 替换为你要测试的单张图片路径
+
+# 类别映射：必须与训练时 data.yaml 的 names 顺序一致
+CLASS_NAMES = {
+    0: "模具车非f块",
+    1: "模具车f块",
+    2: "有遮挡"
+}
+
+
+# ---------------------------
+# 单张图片推理函数
+# ---------------------------
+def classify_single_image(model_path, image_path, class_names):
+    # 加载模型
+    model = YOLO(model_path)
+    print(f"模型加载成功: {model_path}")
+
+    # 读取图像
+    img = cv2.imread(image_path)
+    if img is None:
+        print(f"无法读取图像: {image_path}")
+        return
+
+    print(f"📷 正在推理: {image_path}")
+
+    # 推理
+    results = model(img)
+    probs = results[0].probs.data.cpu().numpy()
+    pred_class_id = int(probs.argmax())
+    pred_label = class_names[pred_class_id]
+    confidence = float(probs[pred_class_id])
+
+    # 输出结果
+    print("\n" + "=" * 40)
+    print(f"🔍 预测结果:")
+    print(f"   类别: {pred_label}")
+    print(f"   置信度: {confidence:.4f}")
+    print(f"   类别ID: {pred_class_id}")
+    print("=" * 40)
+
+    # （可选）在图像上显示结果并保存/显示
+    # 这里我们只打印，不保存。如需可视化，取消下面注释：
+    """
+    label_text = f"{pred_label} ({confidence:.2f})"
+    cv2.putText(img, label_text, (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1.0, (0, 255, 0), 2)
+    cv2.imshow("Result", img)
+    cv2.waitKey(0)
+    cv2.destroyAllWindows()
+
+    # 或保存带标签的图
+    # output_img_path = image_path.replace(".jpg", "_result.jpg")
+    # cv2.imwrite(output_img_path, img)
+    # print(f"带标签图像已保存: {output_img_path}")
+    """
+
+
+# ---------------------------
+# 运行入口
+# ---------------------------
+if __name__ == "__main__":
+    classify_single_image(
+        model_path=MODEL_PATH,
+        image_path=IMAGE_PATH,
+        class_names=CLASS_NAMES
+    )

zjsh_code/obb_main/train/obb_data.yaml

@@ -0,0 +1,18 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+# DOTA8 dataset 8 images from split DOTAv1 dataset by Ultralytics
+# Documentation: https://docs.ultralytics.com/datasets/obb/dota8/
+# Example usage: yolo train model=yolov8n-obb.pt data=dota8.yaml
+# parent
+# ├── ultralytics
+# └── datasets
+#     └── dota8  ← downloads here (1MB)
+
+# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
+path: ./dataset
+train: train  # 数据集路径下的train.txt
+val: val  # 数据集路径下的val.txt
+test: test  # 数据集路径下的test.txt
+
+nc: 1
+names: ['clamp']
+

zjsh_code/obb_main/train/train_obb_zengqiang.py

@@ -0,0 +1,42 @@
+from ultralytics import YOLO
+
+if __name__ == '__main__':
+    # ✅ 推荐：加载官方预训练模型 或 自己的 best.pt
+    model = YOLO(r'yolo11-obb.yaml')  #
+
+    results = model.train(
+        data='obb_data.yaml',
+        epochs=500,                    # 减少 epochs，配合早停
+        patience=0,                   # 50 轮无提升则停止
+        imgsz=640,
+        batch=4,
+        workers=10,
+        device='0',
+        project='runs/train',
+        name='exp_obb5',               # 建议递增实验编号
+        exist_ok=False,
+
+        # 优化器
+        optimizer='AdamW',
+        lr0=0.0005,                    # 更稳定的学习率
+        weight_decay=0.01,
+        momentum=0.937,
+
+        # 数据增强（OBB 关键）
+        degrees=5.0,                  # 随机旋转 ±10°
+        translate=0.1,                 # 平移
+        scale=0.5,                     # 缩放比例
+        shear=1.0,                     # 剪切
+        flipud=0.0,                    # 不推荐上下翻转（角度易错）
+        fliplr=0.5,                    # ✅ 水平翻转，OBB 支持良好
+        hsv_h=0.015,                   # 色调扰动
+        hsv_s=0.7,                     # 饱和度
+        hsv_v=0.4,                     # 明度
+
+        # 其他
+        close_mosaic=50,               # 最后10轮关闭 Mosaic 增强，提升稳定性
+        val=True,                      # 每轮验证
+
+        #
+        amp=False,
+    )

zjsh_code/obb_main/train/yolo11-obb.yaml

@@ -0,0 +1,47 @@
+# Ultralytics YOLO 🚀, AGPL-3.0 license
+# YOLO11 Oriented Bounding Boxes (OBB) model with P3-P5 outputs. For Usage examples see https://docs.ultralytics.com/tasks/obb
+
+# Parameters
+nc: 1 # number of classes
+scales: # model compound scaling constants, i.e. 'model=yolo11n-obb.yaml' will call yolo11-obb.yaml with scale 'n'
+  # [depth, width, max_channels]
+  n: [0.50, 0.25, 1024] # summary: 344 layers, 2695747 parameters, 2695731 gradients, 6.9 GFLOPs
+  s: [0.50, 0.50, 1024] # summary: 344 layers, 9744931 parameters, 9744915 gradients, 22.7 GFLOPs
+  m: [0.50, 1.00, 512] # summary: 434 layers, 20963523 parameters, 20963507 gradients, 72.2 GFLOPs
+  l: [1.00, 1.00, 512] # summary: 656 layers, 26220995 parameters, 26220979 gradients, 91.3 GFLOPs
+  x: [1.00, 1.50, 512] # summary: 656 layers, 58875331 parameters, 58875315 gradients, 204.3 GFLOPs
+
+# YOLO11n backbone
+backbone:
+  # [from, repeats, module, args]
+  - [-1, 1, Conv, [64, 3, 2]] # 0-P1/2
+  - [-1, 1, Conv, [128, 3, 2]] # 1-P2/4
+  - [-1, 2, C3k2, [256, False, 0.25]]
+  - [-1, 1, Conv, [256, 3, 2]] # 3-P3/8
+  - [-1, 2, C3k2, [512, False, 0.25]]
+  - [-1, 1, Conv, [512, 3, 2]] # 5-P4/16
+  - [-1, 2, C3k2, [512, True]]
+  - [-1, 1, Conv, [1024, 3, 2]] # 7-P5/32
+  - [-1, 2, C3k2, [1024, True]]
+  - [-1, 1, SPPF, [1024, 5]] # 9
+  - [-1, 2, C2PSA, [1024]] # 10
+
+# YOLO11n head
+head:
+  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
+  - [[-1, 6], 1, Concat, [1]] # cat backbone P4
+  - [-1, 2, C3k2, [512, False]] # 13
+
+  - [-1, 1, nn.Upsample, [None, 2, "nearest"]]
+  - [[-1, 4], 1, Concat, [1]] # cat backbone P3
+  - [-1, 2, C3k2, [256, False]] # 16 (P3/8-small)
+
+  - [-1, 1, Conv, [256, 3, 2]]
+  - [[-1, 13], 1, Concat, [1]] # cat head P4
+  - [-1, 2, C3k2, [512, False]] # 19 (P4/16-medium)
+
+  - [-1, 1, Conv, [512, 3, 2]]
+  - [[-1, 10], 1, Concat, [1]] # cat head P5
+  - [-1, 2, C3k2, [1024, True]] # 22 (P5/32-large)
+
+  - [[16, 19, 22], 1, OBB, [nc, 1]] # Detect(P3, P4, P5)