基于上下文线索的红外目标跟踪  

Infrared Target Tracking Algorithm Based on Contextual Cue

作　　者：彭耿 周伟[2] 孙艳丽[3] PENG Geng;ZHOU Wei;SUN Yanli(Naval Academy,Beijing 100161;Research Institute of Information Fusion,Naval Aeronautical and Astronautical University,Yantai 264001;Department of Basic Experiment,Naval Aeronautical and Astronautical University,Yantai 264001)

机构地区：[1]海军研究院,北京100161 [2]海军航空工程学院信息融合研究所,烟台264001 [3]海军航空工程学院基础实验部,烟台264001

出　　处：《计算机与数字工程》2019年第6期1353-1356,共4页Computer & Digital Engineering

基　　金：海军重点基金项目“光学图像舰船目标识别定位技术”(编号:NAA-201502-JN1-ZH008);国家高技术研究发展计划(863计划)(编号:2007AA12Z149);国防科技重点实验室基金(编号:9140C910402150C91387)资助

摘　　要：传统的跟踪算法通常忽略目标与其背景的时空相关性,不能有效的区分红外目标和背景边缘,产生偏移现象。针对这一问题,论文通过分析目标和相邻背景之间的统计关联关系,在贝叶斯理论框架下提出了一种基于目标上下文线索的快速鲁棒跟踪算法。实验结果表明跟踪算法计算简单、速度快,有较高的跟踪精度和鲁棒性。The traditional tracking algorithm usually ignores the relevance of time and space of the target and its background,so it cannot distinguish between infrared target and background edge effectively,and produces deviation. Aiming at the problem,in this paper,the fast and robust tracking algorithm in the Bayesian theory framework based on the contextual cue of target is proposed by analyzing the statistical correlation between target and adjacent background. The experiment results show that the tracking algorithm is simple and fast,and has higher tracking accuracy and robustness.

关 键 词：统计关联 上下文线索 目标跟踪 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]