摘要
视觉目标跟踪是计算机视觉领域的重要研究方向之一,目前已广泛应用于视频监控、自动驾驶和视频分析等多个领域。近年来目标跟踪算法虽取得巨大进步,但由于视频中目标和背景信息的随机变化,包括光照变化、尺度变化、快速移动及存在遮挡等,故目标跟踪仍是一项具有挑战性的任务。基于相关滤波器(CF)的跟踪算法由于其在基准数据集上的出色性能而备受关注。主要对相关滤波类跟踪算法做简单梳理总结,首先介绍目标跟踪的研究意义,然后分析总结经典的相关滤波跟踪算法,最后论述当前相关滤波算法存在的问题,针对未来研究方向给出作者的看法。
Visual object tracking is one of the important research directions in the field of computer vision.It has been widely used in video surveillance,autopilot and video analysis.In recent years,although the object tracking algorithm has made great progress,object tracking is still a challenging task due to random changes in target and background information in the video,including illumination changes,scale changes,fast motion and occlusion.Correlation filter(CF)based tracking algorithms have received wide attention due to their excellent performance on the benchmark data set.The correlation filter tracking algorithms are summarized.Firstly,the research significance of object tracking is introduced.Then the classical correlation filter tracking algorithms are analyzed and summarized.Finally,the problems existing in the current algorithm are discussed,and the future research directions are given.
作者
朱伟杰
朱洪军
伍祥
吴锦华
刘晴晴
ZHU Weijie;ZHU Hongjun;WU Xiang;WU Jinhua;LIU Qingqing(College of Computer and Software Engineering, Anhui Institute of Information Technology, Wuhu 241000, China)
出处
《信息工程大学学报》
2019年第6期684-688,共5页
Journal of Information Engineering University
基金
安徽省教育厅高校自然科学重点项目(KJ2019A1291,KJ2018A0634,KT2019A1292)。
关键词
计算机视觉
目标跟踪
相关滤波算法
computer vision
object tracking
correlation filter algorithm