Find the higher degree of similarity between images and report the points in common and the differences

Question

I am trying to analyse due different images and find their similarities. In order to do this, I need to consider the coordinates of the pixels in both images, then select the points in common, i.e. those points with small distance or higher degree of similarity. However, I have some problems to select only those points in common for each of the two images using Mathematica.

EDIT:

img1 = Image[
   Import@FileNameJoin[{NotebookDirectory[], "simpson1.jpg"}]];
pixels1 = ImageData[img1]

img2 = Image[
   Import@FileNameJoin[{NotebookDirectory[], "simpson2.jpg"}]];
pixels2 = ImageData[img2]

grouping = GroupBy[pixels1, pixels1 - Nearest[pixels2]] ; it takes a bit to run, so I think it is not the right way to group the pixels and measure the similarity

Images used:

What I would expect is a list of coordinates for each image, which are pretty similar in both. Is there any way to do this using Mathematica software? This problem could be also interpreted as "spot the differences", if you prefer.

Answer 1

Perhaps this can help.

img1 = Import["https://i.stack.imgur.com/agPud.jpg"];
img2 = Import["https://i.stack.imgur.com/dCNCz.jpg"];
aligned = RemoveAlphaChannel@ImageAlign[img2, img1];
img2 - aligned // ImageAdjust // RemoveBackground

All the white pixels are the pixels that are the same in both images.

Usually, for these types of images, just subtraction should be enough, but since the sizes of the images are different, alignment is also necessary.