원문정보
초록
영어
To improve and enhance the estimation speed and accuracy of motion estimation, this paper proposes a fast diamond search algorithm with improved search strategy; researches motion estimation algorithm by using pattern block matching; analyzes and set up fast searching algorithm model based on the fast convergence characteristics of three step search, the central distribution characteristics of new three step search and the approximate global optimal characteristics of diamond search; use the coarse positioning and fortified halfway stop technologies to establish improved diamond search algorithm; gives the method of reducing the number of search bits. Through experiment and analysis, the results show that the improved diamond search algorithm can effectively reduce the matching points
목차
1. Introduction
2．Motion Estimation Algorithm Based on Pattern Block Matching
2.1. Block Matching Criterion
2.2. Searching Mode Based on Pattern Block Matching
3. Study of the Traditional Fast Search Algorithm
3.1. Three Step Search (TSS)
3.2. New Three-Step Search (NTSS)
3.3. The Diamond Search Algorithm (DS)
4. Improved Search Algorithm
5. Experimental Results and Analysis
5.1. Comparison of Average Matching Points with Different Algorithms
5.2. Comparison of Saved Points
5.3. Comparison of Peak Signal To Noise Ratio (PSNR) with Different Algorithms
5.4. Compensating Frame Results of Different Motion Estimation Fast Algorithm
6. Conclusions
References