如何通过OpenCV findContours获取检测轮廓/圆点的中心坐标

需要检测尺寸为(3000, 2000, 3)的图像中圆点的(x, y)坐标，输入图像如下：

用户使用以下代码处理图像：

img = cv2.imread(path_dots_data + "01000.exr", cv2.IMREAD_ANYCOLOR |cv2.IMREAD_ANYDEPTH | cv2.IMREAD_UNCHANGED)
img = img - dark_img
img = img / img.max()
img = np.clip(a = img, a_min = 0.0, a_max = 1.0)

# Multiply by 255 before converting to uint 8 dtype-
img = img * 255.0
        
# Convert to uint8 dtype for opencv2 algos to work-
img = img.astype('uint8')

# Apply blurring to suppress high-frequency signals-
blur = cv2.medianBlur(img, 5)

# Convert from color to gray-scale-
gray = cv2.cvtColor(blur, cv2.COLOR_BGR2GRAY)

# Apply thresholding to detect white dots on black background-
thresh = cv2.threshold(gray * 5, 200, 255, cv2.THRESH_BINARY)[1]

# Apply findCountours()-
cnts = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]

len(cnts)
# 41

# Add circles around detected contours-
min_area = 0.1
white_dots = []
for c in cnts:
   area = cv2.contourArea(c)
   if area > min_area:
      cv2.drawContours(img, [c], -1, (36, 255, 12), 2)
      white_dots.append(c)


# The final output looks like:
plt.figure(figsize = (12, 10))
plt.imshow(img, cmap = 'gray')
plt.show()

处理后的输出图像如下：

用户的问题：如何获取输出图像中绿色线条标注的轮廓/圆点的中心(x, y)坐标？

可以通过两种实用方法获取轮廓的中心坐标：

方法一：利用轮廓矩计算中心 #

OpenCV的cv2.moments()函数可计算轮廓的矩，通过矩的数值推导中心坐标。修改原有循环代码如下：

min_area = 0.1
white_dots = []
dot_centers = []  # 存储所有圆点中心坐标
for c in cnts:
    area = cv2.contourArea(c)
    if area > min_area:
        cv2.drawContours(img, [c], -1, (36, 255, 12), 2)
        white_dots.append(c)
        
        # 计算轮廓矩
        M = cv2.moments(c)
        # 避免除数为0的异常
        if M["m00"] != 0:
            cX = int(M["m10"] / M["m00"])
            cY = int(M["m01"] / M["m00"])
            dot_centers.append((cX, cY))
            # 可选：在图像上用红色圆点标注中心
            cv2.circle(img, (cX, cY), 5, (0, 0, 255), -1)

执行后，dot_centers列表将保存所有检测到的圆点中心(x, y)坐标。

方法二：通过最小外接圆获取圆心 #

如果目标圆点接近正圆形，使用cv2.minEnclosingCircle()能直接得到圆心坐标，精度更高：

min_area = 0.1
white_dots = []
dot_centers = []
for c in cnts:
    area = cv2.contourArea(c)
    if area > min_area:
        cv2.drawContours(img, [c], -1, (36, 255, 12), 2)
        white_dots.append(c)
        
        # 获取最小外接圆的圆心与半径
        (x, y), radius = cv2.minEnclosingCircle(c)
        center = (int(x), int(y))
        dot_centers.append(center)
        # 可选：在图像上标注圆心
        cv2.circle(img, center, 5, (0, 0, 255), -1)

两种方法均可满足需求，可根据圆点的形状特征选择使用。

内容的提问来源于stack exchange，提问作者Arun