Suppose we have a video of two white points moving toward each other at a uniform rate against a black background. We encode it via MPEG. In one I frame the two points are 100 pixels apart; in the next I frame they have merged. The final point of merger happens to lie at the center of a 16 × 16 macro block.
(a) Describe how you might optimally encode the Y component of the intervening B (or P) frames. (b) Now suppose the points are in color, and that the color changes slowly as the points move. Describe what the encoding of the U and V values might look like.