OpenCV霍夫变换检测圆形，Python

小咪咪 2023-02-12 10:21 44阅读 0赞

假设现在把图中的足球圆形轮廓检测出来，并用红线标记：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3poYW5ncGhpbA_size_16_color_FFFFFF_t_70][]

python代码：

import cv2
    import numpy as np
    
    if __name__ == "__main__":
        original_img = cv2.imread("c.jpg")
        gray = cv2.cvtColor(original_img, cv2.COLOR_BGR2GRAY)
        blur_img = cv2.GaussianBlur(gray, (3, 3), 0)
    
        circles = cv2.HoughCircles(image=blur_img, method=cv2.HOUGH_GRADIENT,
                                   dp=1,
                                   minDist=100,  # 两个圆之间圆心的最小距离.如果太小的，多个相邻的圆可能被错误地检测成了一个重合的圆。反之，这参数设置太大，某些圆就不能被检测出来。
                                   param1=100,
                                   param2=120,
                                   minRadius=80,  # 圆半径的最小值
                                   maxRadius=200)  # 圆半径的最大值
        circles = np.uint16(np.around(circles))
    
        for i in circles[0, :]:
            #画圆
            cv2.circle(original_img, (i[0], i[1]), i[2], (0, 0, 255), 2)
            #画圆心
            cv2.circle(original_img, (i[0], i[1]), 2, (0, 0, 255), 3)
    
        cv2.imshow('circle', original_img)
        cv2.waitKey(0)
        cv2.destroyAllWindows()

结果如下:

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3poYW5ncGhpbA_size_16_color_FFFFFF_t_70 1][]

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3poYW5ncGhpbA_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20200525180856554.jpg?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3poYW5ncGhpbA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3poYW5ncGhpbA_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20200525180929982.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3poYW5ncGhpbA==,size_16,color_FFFFFF,t_70