摘要
针对相关匹配跟踪算法,提出一种新的基于三步搜索法的特征相关跟踪算法。该算法利用Moravec算子提取目标灰度特征,并根据特征点选取相应的子模板,然后利用三步搜索法进行多子模板的相关匹配。该方法可以在大幅度缩短相关匹配的搜索过程的基础之上适应运动目标的旋转、缩放等情况。实验证明该算法实际应用的可行性。
In this paper, a new feature correlation target tracking algorithm based on Three-Step Search (TSS) is presented. In the algorithm, Moravec operator is used for abstracting the gray scale feature of the target, and templates are selected accordingly, then TSS is used for multi-template correlation matching. This algorithm can cut down the course of search for correlation matching, and is adaptable to target's rotating and scaling because of the feature matching. Experiments demonstrate the feasibility of this algorithm.
出处
《电光与控制》
2004年第4期38-40,共3页
Electronics Optics & Control
