python - change detection on frames for different videos -

I am analyzing a video and I am recording some actionable (tablet) devices. And I record it for two different devices. But "content" is similar. I quote the content in a quote from one video to another because of the light, aspect ratio, boundaries. But the action that is running on both videos is the same.

If you are looking at four images, then I am trying to ask.

Image1 (Device1)

Image2 (Device1)

Image3 (device2)

Image4 (device2)

The first two images are from device 1 and the next two devices are from 2. Is there any way that I can detect a change in the frame (the second image to the second). Zooming in action and when it is zoomed to a point and changes occur (Image2). Similarly, I want to capture this change in other video too. So something like trying to catch image 4 from image 4.

I have used algorithms like Manhattan, FD, color histogram. None of them works properly

You can calculate the average word height and width of the texts of the reference frames (Fig. 3) and Comapare it with test images (Fig. 4). This zoom will be able to detect these actions. To do this I think that the binaryization should be sufficient after the associated component analysis. It should give you an estimate of magnification. There are many algorithms available in Google.

Then you can use image interpolation techniques to do zooming.

Good luck.