zjsh_yolov11/angle_base_point/point_test.py

import cv2
import numpy as np
from ultralytics import YOLO
import os

# ====================== 用户配置 ======================
MODEL_PATH = '/home/hx/yolo/ultralytics_yolo11-main/runs/train/exp_pose/weights/best.pt'
IMAGE_PATH = '/output_masks/3.png'
OUTPUT_DIR = '/home/hx/yolo/output_images'  # 保存结果的文件夹

os.makedirs(OUTPUT_DIR, exist_ok=True)

# ====================== 可视化函数 ======================
def draw_keypoints_on_image(image, kpts_xy, kpts_conf, orig_shape):
    """
    在图像上绘制关键点
    :param image: OpenCV 图像
    :param kpts_xy: (N, K, 2) 坐标
    :param kpts_conf: (N, K) 置信度
    :param orig_shape: 原图尺寸 (H, W)
    """
    colors = [(0, 0, 255), (255, 0, 0), (0, 255, 0), (255, 255, 0)]  # 1红, 2蓝, 3绿, 4青

    for i in range(len(kpts_xy)):
        xy = kpts_xy[i]  # (4, 2)
        conf = kpts_conf[i] if kpts_conf.ndim == 2 else kpts_conf[i:i+1]  # (4,) 或标量

        for j in range(len(xy)):
            x, y = xy[j]
            c = conf[j] if hasattr(conf, '__len__') else conf

            x, y = int(x), int(y)

            # 检查坐标是否在图像范围内
            if x < 0 or y < 0 or x >= orig_shape[1] or y >= orig_shape[0]:
                continue

            # 只绘制置信度 > 0.5 的点
            if c < 0.5:
                continue

            # 绘制圆圈
            cv2.circle(image, (x, y), radius=15, color=colors[j], thickness=-1)
            # 标注编号
            cv2.putText(image, f'{j+1}', (x + 20, y - 20),
                        cv2.FONT_HERSHEY_SIMPLEX, 1.5, colors[j], 5)

    return image

# ====================== 主程序 ======================
if __name__ == "__main__":
    print("正在加载模型...")
    model = YOLO(MODEL_PATH)
    print(f"✅ 模型加载完成: {MODEL_PATH}")

    img = cv2.imread(IMAGE_PATH)
    if img is None:
        raise FileNotFoundError(f"无法加载图像: {IMAGE_PATH}")
    print(f"✅ 图像加载成功: {IMAGE_PATH} (shape: {img.shape})")

    print("\n开始推理...")
    results = model(img)

    for i, result in enumerate(results):
        if result.keypoints is not None:
            kpts = result.keypoints
            orig_shape = kpts.orig_shape  # (H, W)

            # ✅ 正确获取坐标和置信度
            kpts_xy = kpts.xy.cpu().numpy()    # (N, K, 2)
            kpts_conf = kpts.conf.cpu().numpy() if kpts.conf is not None else np.ones(kpts_xy.shape[:2])

            print(f"\n 结果 {i + 1}:")
            print(f"   检测到 {len(kpts_xy)} 个实例")
            print(f"   坐标形状: {kpts_xy.shape}")
            print(f"   置信度形状: {kpts_conf.shape}")

            # 绘图
            img_with_kpts = draw_keypoints_on_image(img.copy(), kpts_xy, kpts_conf, orig_shape)

            # 保存
            base_name = os.path.basename(IMAGE_PATH)
            save_path = os.path.join(OUTPUT_DIR, f"keypoints_{base_name}")
            cv2.imwrite(save_path, img_with_kpts)
            print(f"✅ 结果已保存: {save_path}")

            # 显示（可选）
            display_img = cv2.resize(img_with_kpts, (1280, 720))  # 缩小显示
            cv2.imshow("Keypoints", display_img)
            print("按任意键关闭...")
            cv2.waitKey(0)
            cv2.destroyAllWindows()
        else:
            print("   ❌ 未检测到关键点")