视频监控中的目标检测及自动跟踪技术  

Object Detection and Tracking in Video Surveillance

作　　者：董云云[1] 鲍海燕[1] 王溢琴[1] DONG Yun-yun BAO Hai-yan WANG Yi-qin(School of Information Technology and Engineering, Jinzhong University, Jinzhong 030619, Shanxi, China)

机构地区：[1]晋中学院信息技术与工程学院

出　　处：《山西师范大学学报（自然科学版）》2017年第1期65-69,共5页Journal of Shanxi Normal University(Natural Science Edition)

基　　金：山西省教改项目(J2015108);晋中学院教改项目(zl2016jg07;zl2014jg04)

摘　　要：目标跟踪是计算机视觉和图像处理领域的重要研究课题.本文提出了一种目标跟踪方法,包括检测和跟踪两个模块.检测模块完成识别目标,生成具有最佳学习的、可靠性的目标信息;跟踪模块使用特征点提取的目标信息跟踪感兴趣区域,该模块使用Camshift和光流估计来测算目标位置.实验结果表明,所提出的方法可以用来自动跟踪视频中移动的目标.Object detection and tracking is an important research topic in the field of computer vision and image processing. In the paper an object tracking system including detection module and tracking modul is proposed. The detection module identifies the object and generates standard object information with the best reliability of learning. The tracking module tracks the region of interest using the object information extracted by the feature points. Additionally,this module estimates the object position by Camshift and optical flow estimation. Experimental results show that the proposed method can be used to track the moving object in the video surveillance.

关 键 词：目标跟踪 SURF CAMSHIFT 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP393.03[自动化与计算机技术—控制科学与工程]