如何遍历每个连接的组件标签像素

pip install opencv-python

import cv2

# 加载图像
image = cv2.imread('path_to_your_image.jpg')

# 转换为灰度图像
gray_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

# 应用二值化处理
_, binary_image = cv2.threshold(gray_image, 127, 255, cv2.THRESH_BINARY)

# 查找轮廓
contours, hierarchy = cv2.findContours(binary_image, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

for contour in contours:
    # 获取轮廓的边界框
    x, y, w, h = cv2.boundingRect(contour)
    
    # 提取轮廓区域
    roi = image[y:y+h, x:x+w]
    
    # 在这里处理roi中的像素
    # 例如，你可以计算平均颜色、统计像素值等
    average_color_per_row = numpy.average(roi, axis=0)
    average_color = numpy.average(average_color_per_row, axis=0)
    print(f"Average color of contour: {average_color}")