CAJ | 학술논문

研究运动估计（ME）中的经典块匹配算法（BMA）并重点分析菱形（DS）算法，然后利用相邻块之间空间关系，设计一种新的块匹配策略，以相邻宏块的搜索结果进行预测匹配，以减少搜索点数。最后以平均每块搜索点数和平均峰值信噪比为衡量指标验证了该算法的有效性。
연구운동고계（ME）중적경전괴필배산법（BMA）병중점분석릉형（DS）산법，연후이용상린괴지간공간관계，설계일충신적괴필배책략，이상린굉괴적수색결과진행예측필배，이감소수색점수。최후이평균매괴수색점수화평균봉치신조비위형량지표험증료해산법적유효성。
The paper studies the classical fast block-matching algorithm (BMA)in motion estimation (ME)and pays attention to analy-zing diamond search (DS)algorithm.Next by taking advantage of the neighbor contact between neighboring blocks,it designs a new block matching strategy which takes the search result of neighboring macro blocks to predict the match so that the number of search points can be reduced.Finally by taking both average search points of each block and average PSNR as measurements,the algorithm is validated.