METHOD OF CODING MOTION VECTOR BASED ON 2-DIMENSIONAL MINIMUM BIT PREDICTION

摘要

PURPOSE: A method of coding a motion vector based on 2-dimensional minimum bit prediction is provided to search a motion vector where bits of a 2-dimensional predictive error become minimum between a presently-coded motion vector and peripheral motion vectors, and to code 2-dimensional mode information presenting which motion vector is used to find predictive error information, thereby increasing a motion picture compression coding efficiency. CONSTITUTION: A motion vector where predictive error bits with a presently-coded motion vector become minimum is searched, among peripheral motion vectors and the presently-coded motion vector. 2-dimensional predictive error information of the searched motion vector and the presently-coded motion vector is coded. The motion vector is coded by a method of transmitting 2-dimensional mode information presenting which motion vector is used for a predictive value, among the peripheral motion vectors. The motion vector is decoded based on the 2-dimensional predictive error coding information by using the mode information.