融合光流检测与模板匹配的目标跟踪算法  

Object tracking algorithm fused with optical flow detection and template matching

作　　者：魏国剑[1] 侯志强[1] 李武[1] 余旺盛[1] 

机构地区：[1]空军工程大学信息与导航学院,西安710077

出　　处：《计算机应用研究》2014年第11期3498-3501,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(61175029);陕西省自然科学基金资助项目(2011JM8015)

摘　　要：针对传统的目标跟踪算法需要人工选择目标且不能较好地处理目标的尺度变化问题,提出融合光流检测与模板匹配的目标跟踪算法。首先通过结合光流信息与图像分割结果从视频中自动地检测和提取运动目标,实现基于检测的跟踪;当检测跟踪结果不可靠时,再利用模板匹配定位目标位置,实现基于匹配的跟踪;最后,通过自动更新模板,使得跟踪框能够自适应目标的尺度变化。实验结果表明该算法能够在自适应目标尺度变化的同时获得较为稳定的跟踪结果。与其他三种算法相比,所提方法在目标的自动检测提取与尺度自适应方面具有优势。Object must be selected manually and scale variation can’t be handled with properly when using traditional algo-rithms.To overcome this problem,this paper proposed a new object tracking algorithm fused with optical flow detection and template matching.Firstly,it automatically detected and extracted object by combining the optical flow and the image segmenta-tion results,accomplishing detection based tracking.It used the template matching technique for target locating when the result of the detection based tracking was unreliable,accomplishing matching based tracking.Finally,with the help of the template’s auto update,adapted the target’s changing in scale by the tracking window.The experimental results demonstrate that the pro-posed method can not only adapt object’s scale variation,but acquires stable tracking results.Compared with three other algo-rithms,the proposed method has advantages in automatic detection and scale self-adaptation.

关 键 词：目标跟踪 目标检测 光流法 模板匹配 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]