A Novel Object Tracking Algorithm Based on Point Pattern Matching

摘要

This paper proposes a novel point pattern matching algorithm. Compared with most published works, the proposed algorithm is characterized by its outstanding real time performance, making it the only applicable algorithm for object tracking among the existing ones. Evolved with the definition of point pattern matching, analyzing the object tracking task, it presents a specific description of the novel algorithm. The experimental results validate its real time performance and robustness.