python - OpenCV:Canny 边缘检测器获得 minEnclosingCircle

Question

我正在使用 Canny 边缘检测器检测白色背景上的对象，并想在其周围绘制一个矩形和一个圆圈。我可以获得边界矩形的坐标，但不能获取 OpenCV 函数 minAreaRect 和 minEnclosingCircle 的坐标。

import cv2
import numpy as np

img = cv2.imread(image.path, 0)
edges = cv2.Canny(img, 100, 200)

#Bounding Rectangle works
x, y, w, h = cv2.boundingRect(edges)

#This does not work
(x,y),radius = cv2.minEnclosingCircle(edges)

#This also does not work
rect = cv2.minAreaRect(edges)

错误:

Traceback (most recent call last):
  File "/home/hschneider/workspace/onspiration/website/venv/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 3296, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "<ipython-input-28-f9e34ac01335>", line 1, in <module>
    cv2.minEnclosingCircle(edges)
cv2.error: OpenCV(4.1.0) /io/opencv/modules/imgproc/src/shapedescr.cpp:160: error: (-215:Assertion failed) count >= 0 && (depth == CV_32F || depth == CV_32S) in function 'minEnclosingCircle'

我想，这是因为 Canny 边缘检测器的结果格式错误，但我找不到如何转换它以使其有效。

Answer 1

这些函数之间的区别在于，boundingRect 作用于图像，而 minEnclosingCircle 和 minAreaRect 作用于二维点集。要从 Canny 的输出中获取点集，您可以按照 this tutorial 中的建议使用 findCountours :

# im2, contours, hierarchy = cv.findContours(thresh, 1, 2) # OpenCV 3.x
contours, hierarchy = cv.findContours(thresh, 1, 2)        # OpenCV 4.x
cnt = contours[0]

rect = cv.minAreaRect(cnt)

(x,y),radius = cv.minEnclosingCircle(cnt)