复杂背景下视频运动目标跟踪算法研究  

Research of moving target tracking algorithms under complex background

作　　者：徐蕾 刘曼[1] 彭月平[1] 

机构地区：[1]武警工程大学,陕西西安710086

出　　处：《微型机与应用》2017年第18期15-17,共3页Microcomputer & Its Applications

摘　　要：随着视频监控技术的进步,运动目标跟踪作为计算机视觉领域的难点问题,具有广阔的发展前景,其采用的算法种类繁多,效果也各有千秋。结合文献资料介绍了目前普遍适用的四种目标跟踪算法,对比分析了其优势和局限性,可以综合采用两种以上算法进行跟踪研究来提高目标跟踪识别的精度和效率。最后归纳总结了目标跟踪算法亟待解决的问题和研究趋势,对未来跟踪算法的发展进行展望。With the continuous development of video surveillance technology,moving target tracking as a difficult problem in the field of computer vision,has a wide development prospect. It has many kinds of algorithms and they have different effects. In this paper,four kinds of tracking algorithms are introduced,and their advantages and limitations are compared and analyzed. We can combine two or more algorithms to take the tracking research,that can improve the accuracy and efficiency of target tracking. Finally,the problems that tracking algorithm needs to improve and research trends of target tracking algorithm are summarized,also the future development of tracking algorithms is prospected.

关 键 词：运动目标跟踪 跟踪算法 均值漂移 粒子滤波 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]