Parallel architecture for a pel-recursive motion estimation algorithm

Presents a parallel architecture for a pixel-recursive motion estimation algorithm. It is a linear array of processors, each consisting of an initialization part, a data-routing part, and an updating part. The initialization part performs a prediction of the motion vector. The routing part constitutes routing path along which previous frame data are routed from processors that store such data to processors that request it. A clocked version of the router is presented. The updating part calculates an update to the predicted motion vector. The architecture proposed is derived in a systematic way and is parameterized with respect to certain window sizes. It is thus completely different from existing pel-recursive motion estimation architectures. >