PREDICTION OF MOTION WITH ADAPTIVE STEP SIZE BASED ON STATISTICAL SUM OF ABSOLUTE VALUES OF DIFFERENCES

摘要

PROBLEM TO BE SOLVED: To obtain a fixed step motion prediction method and a method for adaptively adjusting the size of a search window by changing the size of the search window with a history of motion between frames. SOLUTION: A digital video system with a compression format of a video sequence represented by a series of frames is provided (step 10). A 1st block of a current frame is selected (step 12). A matrix of a size of a block of luminance pixels of a prescribed frame is selected as an initial object matrix (step 14). A shot-term means comparison value of luminance pixel data between frames is provided (step 16). A retrieval pattern is specified by the use of the given short-term means value (step 18). An averaging statistic variable is updated (step 20). The product obtained is a result of predicting the change in a macro block position, and a search pattern size or a search repetition number is changed along with the history of motion between frames.