智能监控系统中自适应抗遮挡目标跟踪  

Adaptive anti-occlusion moving target tracking of intelligent video surveillance system

作　　者：张颖[1] 李孟歆[1] 英宇 ZHANG Ying;LI Mengxin;YING Yu(College of Information and Control Engineering,Shenyang Jianzhu University,Shenyang 110168,China)

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168

出　　处：《现代电子技术》2018年第17期59-61,共3页Modern Electronics Technique

基　　金：住宅与城乡建设部项目(2015-K8-011);辽宁省教育厅项目(L2013225)~~

摘　　要：智能视频监控系统中的目标跟踪问题存在许多干扰因素,其中运动目标运动时的动态干扰及大面积遮挡问题是影响跟踪失败的最主要的原因。此外,光照和颜色近似干扰问题也是影响跟踪准确度的主要因素。为了达到运动目标的抗干扰及准确跟踪,在Meanshift算法完成目标准确跟踪的基础上,提出一种基于双重判别的自适应跟踪算法,将Meanshift算法与Kalman滤波算法相结合,在出现长时间大面积遮挡及动态背景干扰、光照,颜色近似干扰的情况下,完成目标的准确跟踪。实验结果表明,提出的方法有较好的稳定性和准确性。The target tracking of intelligent video surveillance system is influenced by multiple interference factors,in which the dynamic interference and large area occlusion act as the main reasons influencing the tracking failure while the target is moving.In addition,the approximate interference of illumination and color is also the main factor affecting the tracking accuracy.An adaptive tracking algorithm based on double-discrimination is proposed to achieve the purposes of anti?interference and accurate tracking of moving target,in which the Meanshift algorithm and Kalman filtering algorithm are combined to accomplish the accurate target tracking under the conditions of longtime large area occlusion,dynamic background interference,and approximate interference of illumination and color.The experimental results show that the proposed method has high stability and accuracy.

关 键 词：目标跟踪 MEANSHIFT算法 Bhattacharrya系数 KALMAN滤波 双重判别 抗遮挡 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]