encoding focus on specific object in video or quadrant of video

I'm wondering about encoding that would focus it's data on a specific object in the scene, or a quadrant of the video.

Meaning after the encode, the bottom-left corner of the video capture is super accurate and everything els开发者_开发问答e is blocky. Or an object actor is in focus to sharp detail and the rest of the encode is blocky or blurrier.

Anyone ever come across such a project?

What you are interested in is probably one of the following topics:

• Motion detection
• Spacial Redundancy
• Macroblocks

Besides wikipedia's Video Compression entry, always a decent source, I found this presentation that might be more palatable for a quick overview on the different techniques that you are interested in.