ROI(Region of interest) - python

import numpy as np

import cv2



img = cv2.imread('messi5.jpg')



print(img.shape) #returns a tuple of number of rows, columns, and channels

print(img.size) # returns  total number of pixels is accessed

print(img.dtype) # returns image datatype is obtained

b,g,r = cv2.split(img)

img = cv2.merge((b,g,r))



ball = img[280:340, 330:390]

img[273:333, 100:160] = ball



cv2.imshow('image', img)

cv2.waitKey(0)

cv2.destroyAllWindows()