I'm trying to code something that would run two videos and turn any redundant pixels into a solid color.

-the first video would be a camera moving on a fixed track.
-the second video would be the camera moving on the same track but with a person moving in the foreground.
-I would then code the background video to act as a marker to delete anything in the second video with the person that is redundant.

Why two videos? If I'm just deleting the redundant pixels, object tracking one video would (I guess) be all I would need. But once I've got the two-video differential working, I could instead set all redundant pixels to desaturate completely, or blur slightly, etc. I guess, unless that's still easier with just one video.

If you could give me some starting materials or information to get me along this project I would be most appreciative.

I've been playing around with Processing, but accomplishing this would give me a great and in-depth knowledge of the code.