I want to implement in my comic book reader application the functionality of another application: when you double-click on a comic book block, this block is approaching. the example was shown on a gif file - https://disk.yandex.ru/i/8sO9SpoCtQxnQA . The most difficult thing is how to determine the coordinates of the block? Is there a library for such machine vision?

Is there any way to find out what kind of library is used for this in another application?