first commit

class_xiantiao_pc/chose_ROI.py

@@ -0,0 +1,93 @@
+import cv2
+import numpy as np
+import os
+
+# 全局变量
+drawing = False  # 是否正在绘制
+ix, iy = -1, -1  # 起始点
+roi_list = []    # 存储多个 ROI 坐标 [(x, y, w, h), ...]
+image_path = "1.png"  # <<< 修改为你自己的图像路径
+save_dir = "./roi_1/1.txt"  # 保存坐标的目录
+
+# 创建保存目录
+os.makedirs(save_dir, exist_ok=True)
+
+def draw_rectangle(event, x, y, flags, param):
+    global ix, iy, drawing, img_copy, roi_list
+
+    if event == cv2.EVENT_LBUTTONDOWN:
+        drawing = True
+        ix, iy = x, y
+
+    elif event == cv2.EVENT_MOUSEMOVE:
+        if drawing:
+            # 每次移动都恢复原始图像，重新画矩形
+            img_copy = img.copy()
+            cv2.rectangle(img_copy, (ix, iy), (x, y), (0, 255, 0), 2)
+            cv2.imshow("Select ROI", img_copy)
+
+    elif event == cv2.EVENT_LBUTTONUP:
+        drawing = False
+        w = x - ix
+        h = y - iy
+        if w != 0 and h != 0:
+            # 确保宽高为正
+            x_start = min(ix, x)
+            y_start = min(iy, y)
+            w = abs(w)
+            h = abs(h)
+            cv2.rectangle(img_copy, (x_start, y_start), (x_start + w, y_start + h), (0, 255, 0), 2)
+            cv2.imshow("Select ROI", img_copy)
+            # 添加到列表
+            roi_list.append((x_start, y_start, w, h))
+            print(f"已选择 ROI: (x={x_start}, y={y_start}, w={w}, h={h})")
+
+# 保存坐标到 .txt 文件的函数
+def save_rois_to_txt(rois, filepath):
+    with open(filepath, 'w') as file:
+        for roi in rois:
+            # 将每个 ROI 转换为字符串并写入文件，每行一个 ROI
+            line = ','.join(map(str, roi)) + '\n'
+            file.write(line)
+    print(f"💾 ROI 坐标已保存至: {filepath}")
+
+def select_roi(image_path):
+    global img, img_copy
+
+    img = cv2.imread(image_path)
+    if img is None:
+        print(f"❌ 无法读取图像: {image_path}")
+        return
+
+    img_copy = img.copy()
+    cv2.namedWindow("Select ROI")
+    cv2.setMouseCallback("Select ROI", draw_rectangle)
+
+    print("📌 使用鼠标左键拖拽选择 ROI")
+    print("✅ 选择完成后按 's' 键保存坐标")
+    print("⏭️  按 'n' 键跳过/下一步（可自定义）")
+    print("🚪 按 'q' 键退出")
+
+    while True:
+        cv2.imshow("Select ROI", img_copy)
+        key = cv2.waitKey(1) & 0xFF
+
+        if key == ord('s'):
+            # 保存坐标
+            base_name = os.path.splitext(os.path.basename(image_path))[0]
+            save_path = os.path.join(save_dir, f"{base_name}_rois1.txt")  # 修改了扩展名为 .txt
+            save_rois_to_txt(roi_list, save_path)  # 使用新的保存函数
+
+        elif key == ord('n'):
+            print("⏭️  跳到下一张图片（此处可扩展）")
+            break
+
+        elif key == ord('q'):
+            print("👋 退出程序")
+            cv2.destroyAllWindows()
+            return
+
+    cv2.destroyAllWindows()
+
+if __name__ == "__main__":
+    select_roi(image_path)

class_xiantiao_pc/class_xiantiao.py

@@ -0,0 +1,85 @@
+import os
+import cv2
+from ultralytics import YOLO
+
+
+class ROIClassifier:
+    """ 封装 YOLO 分类模型 + ROI txt 推理 """
+
+    def __init__(self, model_path):
+        if not os.path.exists(model_path):
+            raise FileNotFoundError(f"模型文件不存在: {model_path}")
+
+        self.model = YOLO(model_path)
+        print(f"[INFO] 成功加载 YOLO 分类模型: {model_path}")
+
+    def load_roi(self, roi_txt_path):
+        """读取 ROI txt，可自动支持逗号/空格分隔"""
+        if not os.path.exists(roi_txt_path):
+            raise FileNotFoundError(f"ROI 文件不存在: {roi_txt_path}")
+        with open(roi_txt_path, "r") as f:
+            text = f.read().strip()
+        # 把逗号替换成空格，再 split
+        for ch in [",", ";"]:
+            text = text.replace(ch, " ")
+        parts = text.split()
+        if len(parts) != 4:
+            raise ValueError(f"ROI txt 格式错误，应为4个数字，解析得到: {parts}\n文件: {roi_txt_path}")
+        x1, y1, x2, y2 = map(int, parts)
+        return x1, y1, x2, y2
+
+    def classify(self, img_np, roi_txt_path):
+        """对 ROI 区域做分类，返回 0/1"""
+        h, w = img_np.shape[:2]
+        x1, y1, x2, y2 = self.load_roi(roi_txt_path)
+
+        # -------- ROI 边界安全裁剪 --------
+        x1 = max(0, min(x1, w - 1))
+        x2 = max(0, min(x2, w - 1))
+        y1 = max(0, min(y1, h - 1))
+        y2 = max(0, min(y2, h - 1))
+
+        if x2 <= x1 or y2 <= y1:
+            raise ValueError(f"ROI坐标无效: {x1, y1, x2, y2}")
+
+        # -------- 1. 裁剪 ROI --------
+        roi_img = img_np[y1:y2, x1:x2]
+
+        # -------- 2. resize 到 640×640（强制送给模型）--------
+        roi_img = cv2.resize(roi_img, (640, 640))
+
+        # -------- 3. YOLO 分类推理 --------
+        results = self.model.predict(roi_img, verbose=False)
+
+        cls = int(results[0].probs.top1)  # 0 或 1
+        return cls
+
+
+def class_xiantiao():
+    # ================== 配置 ==================
+    model_path = "/home/hx/yolo/ultralytics_yolo11-main/runs/train/cls/exp_xiantiao_cls/weights/best.pt"
+    img_path = "1.png"
+    roi_txt_path = "/home/hx/开发/ML_xiantiao/class_xiantiao/roi_1/1/1_rois1.txt"
+
+    # ================== 1. 加载模型 ==================
+    classifier = ROIClassifier(model_path)
+
+    # ================== 2. 加载图像 ==================
+    if not os.path.exists(img_path):
+        raise FileNotFoundError(f"图片不存在: {img_path}")
+
+    img_np = cv2.imread(img_path)
+    if img_np is None:
+        raise ValueError("图像加载失败，可能路径错误或图像文件损坏。")
+
+    # ================== 3. 推理 ==================
+    result = classifier.classify(img_np, roi_txt_path)
+
+    # ================== 4. 输出 ==================
+    print(f"\n===== 推理结果 =====")
+    print(f"ROI 文件：{roi_txt_path}")
+    print(f"分类结果：{result}  (0=异常 / 1=正常)\n")
+
+
+if __name__ == "__main__":
+    class_xiantiao()

class_xiantiao_pc/cls_quexian.py

@@ -0,0 +1,63 @@
+import os
+import cv2
+from ultralytics import YOLO
+
+
+class FullImageClassifier:
+    """封装 YOLO 分类模型，对整张图像进行二分类（无缺陷/有缺陷）"""
+
+    def __init__(self, model_path):
+        if not os.path.exists(model_path):
+            raise FileNotFoundError(f"模型文件不存在: {model_path}")
+
+        self.model = YOLO(model_path)
+        print(f"[INFO] 成功加载 YOLO 分类模型: {model_path}")
+
+    def classify(self, img_np):
+        """
+        对整张图像进行分类，返回类别 ID（0 或 1）
+
+        Args:
+            img_np (np.ndarray): BGR 格式的 OpenCV 图像 (H, W, C)
+
+        Returns:
+            int: 分类结果，{0: "有缺陷", 1: "无缺陷"}
+        """
+        # 直接 resize 整图到模型输入尺寸（YOLO 默认为 224x224 或 640x640，由训练决定）
+        # Ultralytics YOLO 会自动处理 resize，但显式指定更可控
+        resized_img = cv2.resize(img_np, (640, 640))
+
+        # 推理（verbose=False 关闭进度条）
+        results = self.model.predict(resized_img, verbose=False)
+
+        cls = int(results[0].probs.top1)  # 获取 top-1 类别索引
+        return cls
+
+
+def cls_quexian():
+    # ================== 配置 ==================
+    model_path = "/home/hx/yolo/ultralytics_yolo11-main/runs/train/cls/exp_xiantiao_cls/weights/best.pt"
+    img_path = "1.png"
+
+    # ================== 1. 加载模型 ==================
+    classifier = FullImageClassifier(model_path)
+
+    # ================== 2. 加载图像 ==================
+    if not os.path.exists(img_path):
+        raise FileNotFoundError(f"图片不存在: {img_path}")
+
+    img_np = cv2.imread(img_path)
+    if img_np is None:
+        raise ValueError("图像加载失败，可能路径错误或图像文件损坏。")
+
+    # ================== 3. 推理（整图）==================
+    result = classifier.classify(img_np)
+
+    # ================== 4. 输出 ==================
+    label_map = {0: "有缺陷", 1: "无缺陷"}
+    print(f"\n===== 推理结果 =====")
+    print(f"分类结果：{result} → {label_map.get(result, '未知')}\n")
+
+
+if __name__ == "__main__":
+    cls_quexian()

class_xiantiao_pc/detect.py

@@ -0,0 +1,115 @@
+import os
+import cv2
+from ultralytics import YOLO
+
+
+class ObjectDetector:
+    """封装 YOLO 目标检测模型，检测图像中的缺陷"""
+
+    def __init__(self, model_path):
+        if not os.path.exists(model_path):
+            raise FileNotFoundError(f"模型文件不存在: {model_path}")
+        self.model = YOLO(model_path)
+        print(f"[INFO] 成功加载 YOLO 目标检测模型: {model_path}")
+
+    def detect(self, img_np, conf_threshold=0.5):
+        # 注意：这里先不设 conf 阈值，以便后续按类别筛选最高分
+        results = self.model.predict(img_np, conf=0.0, verbose=False)  # 获取所有预测
+        detections = []
+        for result in results:
+            boxes = result.boxes.cpu().numpy()
+            for box in boxes:
+                if box.conf.item() >= conf_threshold:  # 在 Python 层过滤
+                    detection_info = {
+                        'bbox': box.xyxy[0],
+                        'confidence': box.conf.item(),
+                        'class_id': int(box.cls.item())
+                    }
+                    detections.append(detection_info)
+        return detections
+
+
+def detect_quexian(img_path="1.png", model_path="/home/hx/yolo/ultralytics_yolo11-main/runs/train/cls/exp_xiantiao_cls/weights/best.pt",
+                   conf_threshold=0.5, debug=False):
+    """
+    检测木条图像中的孔洞/裂缝缺陷，并返回是否为良品。
+    每个类别仅保留置信度最高的一个框。
+
+    Args:
+        img_path (str): 输入图像路径
+        model_path (str): YOLO 检测模型路径（必须是 detect 任务训练的）
+        conf_threshold (float): 置信度阈值
+        debug (bool): 是否启用调试模式（打印详细信息）
+
+    Returns:
+        bool: True 表示无缺陷（良品），False 表示有缺陷（不良品）
+    """
+    # 1. 加载图像
+    if not os.path.exists(img_path):
+        raise FileNotFoundError(f"图片不存在: {img_path}")
+    img_np = cv2.imread(img_path)
+    if img_np is None:
+        raise ValueError("图像加载失败，可能路径错误或图像文件损坏。")
+
+    # 2. 加载模型并检测（获取所有 ≥ conf_threshold 的框）
+    detector = ObjectDetector(model_path)
+    all_detections = detector.detect(img_np, conf_threshold=conf_threshold)
+
+    # 3. 按类别分组，取每个类别中置信度最高的框
+    best_per_class = {}
+    for det in all_detections:
+        cls_id = det['class_id']
+        if cls_id not in best_per_class or det['confidence'] > best_per_class[cls_id]['confidence']:
+            best_per_class[cls_id] = det
+
+    # 转为列表（用于后续处理）
+    top_detections = list(best_per_class.values())
+
+    # 4. 判定是否有缺陷：只要有一个类别有框，就算有缺陷
+    has_defect = len(top_detections) > 0
+    is_good = not has_defect
+
+    # 5. 可视化：只绘制每个类别的最高置信度框
+    label_map = {0: "hole", 1: "crack"}
+    vis_img = img_np.copy()
+    for det in top_detections:
+        x1, y1, x2, y2 = map(int, det['bbox'])
+        conf = det['confidence']
+        cls_id = det['class_id']
+        label = f"{label_map.get(cls_id, '未知')} {conf:.2f}"
+        cv2.rectangle(vis_img, (x1, y1), (x2, y2), (0, 0, 255), 2)
+        cv2.putText(vis_img, label, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 0, 255), 2)
+
+    # 6. 调试信息（仅在 debug=True 时输出）
+    if debug:
+        print(f"\n===== 缺陷检测结果 (DEBUG 模式) =====")
+        print(f"置信度阈值: {conf_threshold}")
+        print(f"有效类别数量: {len(top_detections)}")
+        for i, det in enumerate(top_detections):
+            cls_name = label_map.get(det['class_id'], '未知')
+            bbox_int = det['bbox'].astype(int).tolist()
+            print(f"  - 类别 '{cls_name}' 最高置信度框: 置信度={det['confidence']:.3f}, bbox={bbox_int}")
+
+    # 7. 显示结果图像
+    cv2.imshow('Detection Results', vis_img)
+    cv2.waitKey(0)
+    cv2.destroyAllWindows()
+
+    return is_good
+
+
+if __name__ == "__main__":
+    # 示例：启用 debug 模式
+    is_good_product = detect_quexian(
+        #img_path="/home/hx/开发/ML_xiantiao/class_xiantiao_pc/test_image/val/1.jpg",
+        img_path="/home/hx/开发/ML_xiantiao/class_xiantiao_pc/test_image/train/微信图片_20251216095823_227.jpg",
+        model_path="/home/hx/yolo/ultralytics_yolo11-main/runs/train/exp_detect/weights/best.pt",
+        conf_threshold=0.5,
+        debug=True  # 改为 False 即静默模式
+    )
+
+    # 主程序最终输出（简洁版）
+    if is_good_product:
+        print("产品合格")
+    else:
+        print("产品存在缺陷")

class_xiantiao_pc/divid_val.py

@@ -0,0 +1,108 @@
+import os
+import shutil
+import random
+from pathlib import Path
+
+
+def split_train_to_val(train_dir, val_dir, ratio=0.1, seed=42):
+    """
+    从 train_dir 随机抽取 ratio 比例的数据到 val_dir。
+    自动判断是分类结构（有子文件夹）还是平铺结构（无子文件夹）。
+
+    Args:
+        train_dir (str): 训练集路径
+        val_dir (str): 验证集路径（会自动创建）
+        ratio (float): 抽取比例，如 0.1 表示 10%
+        seed (int): 随机种子，保证可复现
+    """
+    train_path = Path(train_dir)
+    val_path = Path(val_dir)
+
+    if not train_path.exists():
+        raise FileNotFoundError(f"训练目录不存在: {train_path}")
+
+    # 设置随机种子
+    random.seed(seed)
+
+    # 获取所有一级子项
+    items = [p for p in train_path.iterdir()]
+
+    # 判断是否为分类结构：所有子项都是目录
+    is_classification = all(p.is_dir() for p in items) and len(items) > 0
+
+    if is_classification:
+        print("📁 检测到分类结构（含类别子文件夹）")
+        for class_dir in items:
+            class_name = class_dir.name
+            src_class_dir = train_path / class_name
+            dst_class_dir = val_path / class_name
+            dst_class_dir.mkdir(parents=True, exist_ok=True)
+
+            # 获取该类下所有文件（只取图像，但会连带移动同名标签）
+            files = [f for f in src_class_dir.iterdir() if f.is_file()]
+            if not files:
+                print(f"  ⚠️ 类别 '{class_name}' 为空，跳过")
+                continue
+
+            # 随机抽取
+            num_val = max(1, int(len(files) * ratio))  # 至少抽1个
+            val_files = random.sample(files, num_val)
+
+            # 移动文件（包括可能的同名标签）
+            for f in val_files:
+                # 移动主文件
+                shutil.move(str(f), str(dst_class_dir / f.name))
+                # 尝试移动同名不同扩展名的标签（如 .txt）
+                for ext in ['.txt', '.xml', '.json']:
+                    label_file = f.with_suffix(ext)
+                    if label_file.exists():
+                        shutil.move(str(label_file), str(dst_class_dir / label_file.name))
+
+            print(f"  ✅ 类别 '{class_name}': {len(val_files)} / {len(files)} 已移至 val")
+
+    else:
+        print("📄 检测到平铺结构（无类别子文件夹）")
+        val_path.mkdir(parents=True, exist_ok=True)
+
+        # 获取所有文件，按“主文件”分组（如 img.jpg 和 img.txt 视为一组）
+        all_files = [f for f in train_path.iterdir() if f.is_file()]
+        # 提取所有不带扩展名的 stem（去重）
+        stems = set(f.stem for f in all_files)
+        file_groups = []
+        for stem in stems:
+            group = [f for f in all_files if f.stem == stem]
+            file_groups.append(group)
+
+        if not file_groups:
+            print("⚠️ 训练目录为空")
+            return
+
+        # 随机抽取组
+        num_val = max(1, int(len(file_groups) * ratio))
+        val_groups = random.sample(file_groups, num_val)
+
+        # 移动每组所有文件
+        for group in val_groups:
+            for f in group:
+                shutil.move(str(f), str(val_path / f.name))
+
+        print(f"✅ 平铺结构: {len(val_groups)} 组 / {len(file_groups)} 组 已移至 val")
+
+    print(f"\n🎉 分割完成！验证集已保存至: {val_path}")
+
+
+# ======================
+# 使用示例
+# ======================
+if __name__ == "__main__":
+    # 修改为你自己的路径
+    #TRAIN_DIR = "/media/hx/04e879fa-d697-4b02-ac7e-a4148876ebb0/dataset/cls-new/19cc/train"
+    #VAL_DIR = "/media/hx/04e879fa-d697-4b02-ac7e-a4148876ebb0/dataset/cls-new/19cc/val"
+    TRAIN_DIR = "/home/hx/开发/ML_xiantiao/image/datasetr1/train"
+    VAL_DIR = "/home/hx/开发/ML_xiantiao/image/datasetr1/val"
+    split_train_to_val(
+        train_dir=TRAIN_DIR,
+        val_dir=VAL_DIR,
+        ratio=0.1,  # 抽取 10%
+        seed=25  # 随机种子
+    )

class_xiantiao_pc/image_test.py

@@ -0,0 +1,71 @@
+import cv2
+import numpy as np
+import os
+
+
+def detect_cracks_in_wood(image_path, output_dir="output"):
+    """
+    检测木条图像中的裂缝缺陷（线状或不规则形状）
+
+    Args:
+        image_path (str): 输入图像路径
+        output_dir (str): 结果保存目录
+    """
+    os.makedirs(output_dir, exist_ok=True)
+
+    # 1. 读取图像
+    img = cv2.imread(image_path)
+    if img is None:
+        raise FileNotFoundError(f"无法读取图像: {image_path}")
+
+    original = img.copy()
+    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
+
+    # 2. 使用高斯模糊去除噪声，并增强裂缝对比度
+    blurred = cv2.GaussianBlur(gray, (5, 5), 0)
+
+    # 3. 使用Canny边缘检测算法查找边缘
+    edges = cv2.Canny(blurred, 50, 150)
+
+    # 4. 形态学操作：使用闭运算填充裂缝间的空隙
+    kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3, 3))
+    closed_edges = cv2.morphologyEx(edges, cv2.MORPH_CLOSE, kernel, iterations=2)
+
+    # 5. 查找轮廓
+    contours, _ = cv2.findContours(closed_edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+
+    # 6. 筛选疑似裂缝的轮廓
+    min_length = 200  # 最小长度（像素），根据实际情况调整
+    max_gap = 10  # 连接断开点的最大距离（像素）
+
+    crack_contours = []
+    for cnt in contours:
+        length = cv2.arcLength(cnt, False)  # 计算曲线长度
+        if length > min_length:
+            crack_contours.append(cnt)
+
+    # 7. 在原图上绘制检测到的裂缝
+    result_img = original.copy()
+    cv2.drawContours(result_img, crack_contours, -1, (0, 0, 255), 2)  # 红色框
+
+    # 8. 保存结果
+    base_name = os.path.splitext(os.path.basename(image_path))[0]
+    cv2.imwrite(os.path.join(output_dir, f"{base_name}_edges.png"), edges)
+    cv2.imwrite(os.path.join(output_dir, f"{base_name}_closed_edges.png"), closed_edges)
+    cv2.imwrite(os.path.join(output_dir, f"{base_name}_result.png"), result_img)
+
+    print(f"✅ 检测完成！共找到 {len(crack_contours)} 个疑似裂缝")
+    print(f"📁 结果已保存至: {output_dir}")
+
+    # （可选）显示
+    cv2.imshow("Original", cv2.resize(original, (640, 480)))
+    cv2.imshow("Edges", cv2.resize(edges, (640, 480)))
+    cv2.imshow("Closed Edges", cv2.resize(closed_edges, (640, 480)))
+    cv2.imshow("Result", cv2.resize(result_img, (640, 480)))
+    cv2.waitKey(0)
+    cv2.destroyAllWindows()
+
+
+if __name__ == "__main__":
+    image_path = "1.jpg"  # ← 替换为你的木条图像路径
+    detect_cracks_in_wood(image_path)

class_xiantiao_pc/resize_dataset_image.py

@@ -0,0 +1,57 @@
+import os
+import cv2
+
+# ----------------------------
+# 配置
+# ----------------------------
+SOURCE_ROOT_DIR = "/home/hx/开发/ML_xiantiao/image/dataset1"       # 原始图片根目录
+TARGET_ROOT_DIR = "/home/hx/开发/ML_xiantiao/image/datasetr1"    # 输出根目录
+CLASSES = ["class0", "class1"]  # 类别列表
+TARGET_SIZE = 640           # resize 尺寸
+SUBSETS = ["train", "val", "test"]
+
+# ----------------------------
+# 全局 ROI (x, y, w, h)
+# ----------------------------
+GLOBAL_ROI = [3,0,694,182]
+# ----------------------------
+# 主处理函数
+# ----------------------------
+def process_images():
+    x, y, w, h = GLOBAL_ROI
+    for subset in SUBSETS:
+        for class_dir in CLASSES:
+            src_dir = os.path.join(SOURCE_ROOT_DIR, subset, class_dir)
+            tgt_dir = os.path.join(TARGET_ROOT_DIR, subset, class_dir)
+            os.makedirs(tgt_dir, exist_ok=True)
+
+            if not os.path.exists(src_dir):
+                print(f"警告: 源目录 {src_dir} 不存在，跳过")
+                continue
+
+            for file in os.listdir(src_dir):
+                if not (file.endswith(".jpg") or file.endswith(".png")):
+                    continue
+
+                img_path = os.path.join(src_dir, file)
+                img = cv2.imread(img_path)
+                if img is None:
+                    print(f"❌ 无法读取图片: {img_path}")
+                    continue
+
+                h_img, w_img = img.shape[:2]
+                x1, y1 = max(0, x), max(0, y)
+                x2, y2 = min(w_img, x + w), min(h_img, y + h)
+
+                cropped = img[y1:y2, x1:x2]
+                if cropped.size == 0:
+                    print(f"❌ 裁剪结果为空: {file}")
+                    continue
+
+                resized = cv2.resize(cropped, (TARGET_SIZE, TARGET_SIZE))
+                tgt_path = os.path.join(tgt_dir, file)
+                cv2.imwrite(tgt_path, resized)
+                print(f"✅ 图片处理完成: {subset}/{class_dir}/{file}")
+
+if __name__ == "__main__":
+    process_images()

class_xiantiao_pc/roi_1/1/1_rois1.txt

@@ -0,0 +1 @@
+3,0,694,182

class_xiantiao_pc/test_image/train/1.txt

@@ -0,0 +1 @@
+1 0.614113 0.455275 0.664973 0.245135

class_xiantiao_pc/test_image/train/1（另一个复件）.txt

@@ -0,0 +1 @@
+1 0.614113 0.455275 0.664973 0.245135

class_xiantiao_pc/test_image/train/1（复件）.txt

@@ -0,0 +1 @@
+1 0.614113 0.455275 0.664973 0.245135

class_xiantiao_pc/test_image/train/1（第 3 个复件）.txt

@@ -0,0 +1 @@
+1 0.614113 0.455275 0.664973 0.245135

class_xiantiao_pc/test_image/train/1（第 4 个复件）.txt

@@ -0,0 +1 @@
+1 0.614113 0.455275 0.664973 0.245135

class_xiantiao_pc/test_image/train/1（第 5 个复件）.txt

@@ -0,0 +1 @@
+1 0.614113 0.455275 0.664973 0.245135

class_xiantiao_pc/test_image/train/2.txt

@@ -0,0 +1,2 @@
+0 0.325184 0.190563 0.102830 0.112333
+0 0.532322 0.519080 0.408382 0.188687

class_xiantiao_pc/test_image/train/2（另一个复件）.txt

@@ -0,0 +1,2 @@
+0 0.325184 0.190563 0.102830 0.112333
+0 0.532322 0.519080 0.408382 0.188687

class_xiantiao_pc/test_image/train/2（复件）.txt

@@ -0,0 +1,2 @@
+0 0.325184 0.190563 0.102830 0.112333
+0 0.532322 0.519080 0.408382 0.188687

class_xiantiao_pc/test_image/train/2（第 3 个复件）.txt

@@ -0,0 +1,2 @@
+0 0.325184 0.190563 0.102830 0.112333
+0 0.532322 0.519080 0.408382 0.188687

class_xiantiao_pc/test_image/train/2（第 4 个复件）.txt

@@ -0,0 +1,2 @@
+0 0.325184 0.190563 0.102830 0.112333
+0 0.532322 0.519080 0.408382 0.188687

class_xiantiao_pc/test_image/train/2（第 5 个复件）.txt

@@ -0,0 +1,2 @@
+0 0.325184 0.190563 0.102830 0.112333
+0 0.532322 0.519080 0.408382 0.188687

class_xiantiao_pc/test_image/train/3.txt

@@ -0,0 +1 @@
+0 0.670369 0.354613 0.214830 0.096716

class_xiantiao_pc/test_image/train/3（另一个复件）.txt

@@ -0,0 +1 @@
+0 0.670369 0.354613 0.214830 0.096716

class_xiantiao_pc/test_image/train/3（复件）.txt

@@ -0,0 +1 @@
+0 0.670369 0.354613 0.214830 0.096716

class_xiantiao_pc/test_image/train/3（第 3 个复件）.txt

@@ -0,0 +1 @@
+0 0.670369 0.354613 0.214830 0.096716

class_xiantiao_pc/test_image/train/3（第 4 个复件）.txt

@@ -0,0 +1 @@
+0 0.670369 0.354613 0.214830 0.096716

class_xiantiao_pc/test_image/train/3（第 5 个复件）.txt

@@ -0,0 +1 @@
+0 0.670369 0.354613 0.214830 0.096716

class_xiantiao_pc/test_image/train/4.txt

@@ -0,0 +1 @@
+0 0.535815 0.542963 0.570926 0.593823

class_xiantiao_pc/test_image/train/4（另一个复件）.txt

@@ -0,0 +1 @@
+0 0.535815 0.542963 0.570926 0.593823

class_xiantiao_pc/test_image/train/4（复件）.txt

@@ -0,0 +1 @@
+0 0.535815 0.542963 0.570926 0.593823

class_xiantiao_pc/test_image/train/4（第 3 个复件）.txt

@@ -0,0 +1 @@
+0 0.535815 0.542963 0.570926 0.593823

class_xiantiao_pc/test_image/train/微信图片_20251216095823_214.txt

@@ -0,0 +1 @@
+0 0.493643 0.417761 0.436794 0.658617

class_xiantiao_pc/test_image/train/微信图片_20251216095823_214（另一个复件）.txt

@@ -0,0 +1 @@
+0 0.493643 0.417761 0.436794 0.658617

class_xiantiao_pc/test_image/train/微信图片_20251216095823_214（复件）.txt

@@ -0,0 +1 @@
+0 0.493643 0.417761 0.436794 0.658617

class_xiantiao_pc/test_image/train/微信图片_20251216095823_219.txt

@@ -0,0 +1 @@
+1 0.682482 0.234558 0.050639 0.296622

class_xiantiao_pc/test_image/train/微信图片_20251216095823_219（另一个复件）.txt

@@ -0,0 +1 @@
+1 0.682482 0.234558 0.050639 0.296622

class_xiantiao_pc/test_image/train/微信图片_20251216095823_219（复件）.txt

@@ -0,0 +1 @@
+1 0.682482 0.234558 0.050639 0.296622

class_xiantiao_pc/test_image/train/微信图片_20251216095823_219（第 3 个复件）.txt

@@ -0,0 +1 @@
+1 0.682482 0.234558 0.050639 0.296622