MODEL DESCRIBING VIDEO STORED IN MEMORY AND CONVERTED IN CODE

摘要

PROBLEM TO BE SOLVED: To provide a rate-distortion model (R-D) considering an interframe dependency for an optimum bit allocation in a video code conversion for an error resistance.

SOLUTION: The accurate rate-distortion (R-D) model is provided for converting a video in code. One model describes rate-distortion characteristics for re-quantizing the video by considering the interframe dependency. Other models estimate the relationship of a distortion to an error propagation in a motion compensative video, and characterize the insertion rates of an intra-block and a resynchronizing marker. These models are used for an optimum bit-allocation system for converting the video in code.

COPYRIGHT: (C)2006,JPO&NCIPI