Abstract:
Motion estimation, in video processing, is a technique for describing a frame in terms of
translated blocks of another reference frame. This technique increases the ratio of video
compression by the efficient use of redundancy information between frames. The Block
Matching based motion estimation methods, based on dividing frames into blocks and
calculating a motion vector for each block, are accepted as motion estimation standards in
video encoding systems by international enterprises, such as MPEG, ATSC, DVB and ITU. In this thesis study, a hardware implementation of 3-D Recursive Search Block Matching Algorithm for the motion estimation levels, global and local motion estimations, is presented.
