parse
1. Read in pictures
2.Channel separation
3. If the pixel value is between [100, 200], assign a value of 128. Values greater than 200 are assigned a value of 255, and values less than 100 are assigned a value of 0.
Source code
import cv2
img_raw_path="past/unet-test_result0-0-1-0.png"
img_raw=cv2.imread(img_raw_path)
(r,g,b)=cv2.split(img_raw)
r_copy=r.copy()
r_copy[ r_copy.any in range(100,200)]=128
r_copy[r_copy>200]=255
r_copy[r_copy<100]=0
cv2.imwrite("New/unet-test_result0-0-1-0.png",r_copy)image display
The original picture is as follows.
After processing the picture, the gap becomes clear.
