canny算法 python

Canny算法是一种经典的边缘检测算法，主要用于提取图像中的边缘信息。在Python中，可以使用OpenCV库中的cv2.Canny函数来实现Canny算法。

具体实现步骤如下：

1. 导入OpenCV库和numpy库

import cv2
import numpy as np

2. 读取图像文件并转换为灰度图像

img = cv2.imread('image.jpg')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

3. 使用高斯滤波器对图像进行平滑处理

blur = cv2.GaussianBlur(gray, (5,5), 0)

4. 根据Sobel算子计算图像的梯度值和方向

grad_x = cv2.Sobel(blur, cv2.CV_16S, 1, 0, ksize=3)
grad_y = cv2.Sobel(blur, cv2.CV_16S, 0, 1, ksize=3)
grad_abs_x = cv2.convertScaleAbs(grad_x)
grad_abs_y = cv2.convertScaleAbs(grad_y)
grad = cv2.addWeighted(grad_abs_x, 0.5, grad_abs_y, 0.5, 0)

5. 对梯度值进行非极大值抑制处理

rows, cols = grad.shape
for i in range(1, rows-1):
    for j in range(1, cols-1):
        if grad[i,j] == 0:
            grad[i,j] = 1e-6
        if grad[i,j] > grad[i-1,j] and grad[i,j] > grad[i+1,j]:
            if grad[i,j] > grad[i,j-1] and grad[i,j] > grad[i,j+1]:
                grad[i,j] = grad[i,j]
            else:
                grad[i,j] = 0
        else:
            grad[i,j] = 0

6. 进行双阈值处理，将梯度值分为强边缘、弱边缘和非边缘三类

thresh_low = 50
thresh_high = 150
strong_edges = np.zeros_like(grad)
strong_edges[grad > thresh_high] = 255
weak_edges = np.zeros_like(grad)
weak_edges[(grad >= thresh_low) & (grad <= thresh_high)] = 255

7. 对弱边缘进行连接处理，将其转化为强边缘

rows, cols = strong_edges.shape
for i in range(1, rows-1):
    for j in range(1, cols-1):
        if weak_edges[i,j] == 255:
            if strong_edges[i-1,j] == 255 or strong_edges[i+1,j] == 255 or strong_edges[i,j-1] == 255 or strong_edges[i,j+1] == 255:
                strong_edges[i,j] = 255
                weak_edges[i,j] = 0

最终得到的strong_edges即为Canny算法提取出来的图像边缘信息。