Motion-Compensated Television Coding: Part 1

Television signals are generated by scanning a scene several times a second even though there may not be any change in the scene from one frame to the next. This results in a considerable frame-to-frame redundancy in the signal. Existence of this redundancy has long been recognized, and several measurements have been made to quantify it. However, the first real demonstration of a frame-to-frame coder which used redundancy between successive frames was made in 1969 by Mounts.1 Since then, several improvements have been made to the basic frame-to-frame encoder resulting in prototypes or real implementations of coder-decoder pairs.2"5 These are the subjects of two excellent surveys,6,7 the first one covering material up to 1972 and the second one up to 1978. As is evident from these works, most frame-toframe coders are based on the following: (i) Segmenting each television frame into two parts, one part which is the same as the previous frame and the other part (called the moving area) which has changed from the previous frame. (ii) Transmitting two types of moving area information: (a) addresses specifying the location of the picture elements in the moving 631