作　　者： ; ; ;

机构地区： 西安电子科技大学电子工程学院

出　　处： 《电子科技》 2007年第12期24-27,31,共5页

摘　　要： 针对复杂背景下运动目标跟踪问题,提出了一种基于Hausdorff距离的目标跟踪改进算法。作为模板与待匹配图像相似性的度量,传统的Hausdorff距离容易造成误匹配;现有的Hausdorff距离改进算法虽然可以从一定程度上克服上述缺点,但是计算复杂,难以满足实时性。为了能够很好的解决上述这些问题,达到稳定实时的跟踪,结合Hausdorff距离提出一种新的图像匹配策略。实验表明,该算法匹配精度高,适用性强,同时大大加快运算速度。 In order to solve the problem of moving object tracking in the clutter background, an improved object tracking algorithm based on Hausdorff distance is proposed. As a similar metric of the template and matching image, the traditional Hausdorff distance easily mismatches. Although improved algorithms available can overcome these shortcomings to some extent, they are so complicated in calculation to guarantee real -time operation. For a steady and real -time tracking, a novel image matching is proposed. Experimental results show that this improved algorithm has the advantages of simple calculation, high matching accuracy and robust applicability.

关 键 词： 目标跟踪 距离 图像匹配

领　　域： [自动化与计算机技术] [自动化与计算机技术]