如何获取图像中像素的x，y（坐标）数组

Question

I need all coordinates of certain colored pixels in image by opencv - python

I guess return value of cv2.inRange may contains coordinates of pixels, but can`t understands how it is constructed

Is there any way to get this?

Answer 1

import cv2
import numpy as np


img = cv2.imread("sample.jpg")
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)

# Create an array with HSV between 110,50,50 and 130,255,255
array = cv2.inRange(hsv, np.array([110,50,50]), np.array([130,255,255]))

np.nonzero(array)
# (array([ 1,  1,  1,  2, 12, 13, 15, 15, 15, 18, 19]),
#  array([13, 14, 18, 22, 21,  1, 20, 22, 23,  1, 30]))

That means that pixels (1, 13), (1, 14), (1, 18), (2, 22), etc. are between our given values. If you want them already paired:

np.transpose(np.nonzero(array))
# array([[ 1, 13],
#        [ 1, 14],
#        ...
#        [18,  1],
#        [19, 30]])