let us suppose we have following two image :

and :

at first time, it seems two identical, but we might see some kind of difference between them, maybe land is deformed or some other work(let say agricultural activies) are done in second images, or eroded land is presented on second version,here is my code for reading both image :

import cv2
import numpy as np
image1 =cv2.imread("image1.png")
image1 =cv2.resize(image1,(500,500))
image2 =cv2.imread("image2.png")
imag2 =cv2.resize(image2,(500,500))
image =np.hstack([image1,imag2])
cv2.imshow("image",image)
cv2.waitKey(0)
cv2.destroyAllWindows()

my question is how can i calculate changes between two image?let me clarify idea behind my qestion : let us suppose we have following image :

if i have this image and its version before erosion, then question is : how can i calculate structural difference between two image? is pixel by pixel calculation enough?i have reviewed some methods here :
difference between two image

but are they correct ways of estimation information difference between two image?