基于块运动的阈值判断自适应运动估计算法  

Threshold judgment in adaptive motion estimation algorithm based on block motion

作　　者：魏文振 昂志敏[1] 

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《电子设计工程》2015年第24期179-182,共4页Electronic Design Engineering

摘　　要：现有的运动估计算法在保证视频整体质量的同时,在很大程度上提高了运算效率,但是仍然有着可以更进一步的提升。基于UMHexagonS算法在准确性、高效性上仍有提升的空间,本文在算法中加入阈值判断和对视频的运动剧烈程度的划分,有效减少搜索点数,进一步提高运算效率。通过在JM平台上的实验,验证了改进后算法的可靠性,可以有效提高运算速率,而且在视频监控领域有着具有很高的应用前景。Existing motion estimation algorithms ensure the overall quality of the video at the same time, has greatly improved operational efficiency, but may still have further improved. Base on UMHexagonS algorithm accuracy, efficiency, added to the threshold determination and the movement intensity of the video division in the algorithm, effectively reduce the search points, to further improve the operation efficiency. Experiments on JM platform to verify the reliability of the improved algorithm can effectively improve the computing speed.The improved algorithm has a high prospect in the field of video surveillance.

关 键 词：运动估计算法 UMHEXAGONS算法 阈值 运动类型自适应性 混合搜索模板 

分 类 号：TN919[电子电信—通信与信息系统]