无迹卡尔曼滤波Cam-Shift视频监控跟踪算法改进

Algorithm Improvement for Cam-Shift Video Surveillance and Tracking Based on Unscented Kalman Filter

视频监控系统作为城市安防建设的重点工程,为创建“平安城市”民生工程保驾护航。传统的视频监控系统控制系统纯粹是“有监无控”,实时问题的处理及设备操控均需要人工干预才能处理,人力物力成本投入大,系统的操作精准度也难以保证。针对均值漂移(Mean-Shift)及Cam-Shift跟踪算法缺点,提出了一种融合无迹卡尔曼滤波器的Cam-Shift算法改进,通过实验仿真,提出的跟踪算法性能和效率都比传统算法更好,在视频监控运动目标跟踪效果较好。

Video surveillance system,as a key project of urban security construction,will protect the people’s livelihood project—“safe city”.However,the traditional video surveillance system is purely“monitoring without control”.The real-time problem processing and equipment control need manual intervention,the cost of manpower and material resources is great and the operation accuracy of the system is difficult to be guaranteed.In order to deal with the above problems,a Cam-Shift algorithm improvement based on unscented Kalman filter is proposed.Through the experimental simulation,the performance and efficiency of the tracking algorithm proposed are better than the traditional algorithm,and the effect of the moving target tracking in video surveillance is better.