opencv边缘检测运用sobel算子源代码方法
import cv2
import numpy as np
import random
import math
img0
= cv2.imread
(r
"C:\Users\lenovo\Desktop\python\python_vision\image.jpg",1
)
imgInfo
= img0.shape
height
= imgInfo
[0
]
weight
= imgInfo
[1
]
gray
= cv2.cvtColor
(img0,cv2.COLOR_RGB2GRAY
)
imgG
= cv2.GaussianBlur
(gray,
(3,3
),0
)
dst
= np.zeros
((height,weight,1
),np.uint8
)
for i
in range
(0,height-2
):
for j
in range
(0,weight-2
):
gy
= imgG
[i,j
]*1+imgG
[i,j+1
]*2+imgG
[i,j+2
]*1-imgG
[i+2,j
]*1-imgG
[i+2,j+1
]*2-imgG
[i+2,j+2
]*1
gx
= imgG
[i,j
]*1-imgG
[i+2,j
]*1+imgG
[i+1,j
]*2-imgG
[i+1,j+2
]*2+imgG
[i+2,j
]*1-imgG
[i+2,j+2
]*1
grad
= math.sqrt
(gx*gx+gy*gy
)
if grad
>20:
dst
[i,j
] = 0
else:
dst
[i,j
] = 255
cv2.imshow
("dst",dst
)
cv2.waitKey
()
效果图片： 代码解释：
for i
in range
(0,height-2
):
for j
in range
(0,weight-2
):
gy
= imgG
[i,j
]*1+imgG
[i,j+1
]*2+imgG
[i,j+2
]*1-imgG
[i+2,j
]*1-imgG
[i+2,j+1
]*2-imgG
[i+2,j+2
]*1
gx
= imgG
[i,j
]*1-imgG
[i+2,j
]*1+imgG
[i+1,j
]*2-imgG
[i+1,j+2
]*2+imgG
[i+2,j
]*1-imgG
[i+2,j+2
]*1
grad
= math.sqrt
(gx*gx+gy*gy
)
if grad
>20:
dst
[i,j
] = 0
else:
dst
[i,j
] = 255
分别让3X3的像素乘以sobel算子，然后求gx+gy平方和的根值，判断其大于或者小于某一个值。 边缘检测有：Laplace算子，sobel算子和canny算子等等
