一种新型点模式模糊松驰匹配成象跟踪算法

A NEW IMAGE TRACKING ALGORITHM OF POINT PATTERN FUZZY RELAXATION MATCHING

本文将一种新型的点模式模糊松驰匹配算法用于目标跟踪.该算法对整体几何畸变、一定程度的非仿射畸变、多点少点畸变不敏感.新型匹配矢量集以及基准矢量的提出大大地减少了运算时间.海上多目标红外图象序列模拟结果表明:该算法不仅适用于单目标,而且适用于多目标的识别、定位和跟踪.

A new algorithm, i. e., the point pattern fuzzy relaxation matching algorithm is applied to the tracking of targets. This algorithm is insensitive not only to the affine distortion in the whole, but also to the inaffine distortion in point addition or deletion to some extent. The establishment of matching vector set and the conoept of standard vector can reduce the calculation time greatly. The computer-simulated results show that the algorithm is adaptive not only to the recognition, location and tracking of a single target, but also to those of multi-targets.