Introduction To Video Compression - MULTIMEDIA

A video consists of a time - ordered sequence of frames — images. An obvious solution to video compression would be predictive coding based on previous frames. For example, suppose we simply created a predictor such that the prediction equals the previous frame. Then compression proceeds by subtracting images: instead of subtracting the image from itself (i.e., use a derivative), we subtract in time order and code the residual error.

And this works. Suppose most of the video is unchanging in time. Then we get a nice histogram peaked sharply at zero — a great reduction in terms of the entropy of the original video, just what we wish for.

However, it turns out that at acceptable cost, we can do even better by searching for just the right parts of the image to subtract from the previous frame. After all, our naive subtraction scheme will likely work well for a background of office furniture and sedentary university types, but wouldn't a football game have players zooming around the frame, producing large values when subtracted from the previously static green playing field?

So in the next section we examine how to do better. The idea of looking for the football player in the next frame is called motion estimation, and the concept of shifting pieces of the frame around so as to best subtract away the player is called motion compensation.


All rights reserved © 2018 Wisdom IT Services India Pvt. Ltd DMCA.com Protection Status

MULTIMEDIA Topics