一种基于边缘特征和改进Hausdorff距离的火箭目标跟踪算法  

A Rocket Object Tracking Algorithm Based on Edge Features and Improved Hausdorff Distance

作　　者：阳小燕[1] 李琳琳[1] 

机构地区：[1]成都理工大学核技术与自动化学院,四川成都

出　　处：《国际航空航天科学》2016年第2期34-40,共7页Journal of Aerospace Science and Technology

摘　　要：火箭发射时背景十分复杂,产生的大量烟雾也会影响跟踪效果,本文针对这一问题提出了一种基于边缘特征和改进Hausdorff距离的跟踪算法。首先将火箭的边缘特征与山体背景分离出来,再用改进的Hausdorff距离进行模板匹配。在目标跟踪过程当中,为了加快模板匹配速度,使用了3步搜索法。实验结果证明,该算法适时跟踪效果好,运算速度快。The background of the rocket launch is very complex, and a lot of smoke will affect the tracking results, so this paper proposes a tracking algorithm based on edge feature and improves Haus-dorff distance. Firstly, we separate the edge feature of the rocket from the mountain background, and then use the improved Hausdorff distance to carry out template matching. In the target tracking process, in order to speed up the template matching speed, the 3-step search method is applied. Experimental results show that this algorithm has good tracking performance and fast com- putation speed.

关 键 词：边缘特征 HAUSDORFF距离 目标跟踪 图像匹配 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]