基于联合模板稀疏表示的目标跟踪方法  

Object tracking method based on sparse representation of joint template

作　　者：张旭东[1] 陈仲海[1] 胡良梅[1] 杨慧[1] 董文菁 

机构地区：[1]合肥工业大学计算机与信息学院,合肥230009

出　　处：《控制与决策》2015年第9期1696-1700,共5页Control and Decision

基　　金：国家自然科学基金项目(61273237;61271121;61403116);中国博士后基金项目(2014M560507);中央高校基本科研业务费专项资金项目(2013HGBH0045)

摘　　要：针对传统基于稀疏表示的目标跟踪方法中,当场景中含有与目标相似的背景时容易出现跟踪漂移的问题,提出一种新的目标跟踪方法.该方法基于目标的局部二元模式特征,将目标外观模型同时用原始目标模板与当前帧部分粒子构成的联合模板稀疏表示,构建一个联合目标函数,将跟踪问题通过迭代转化为求解最优化问题.实验结果表明,所提出跟踪方法在解决遮挡、光照等问题的同时,对场景中含有与目标相似背景的序列具有较好的跟踪效果.When a scene contains a similar background to the object,the traditional tracking method based on sparse representation may produce the drifting problems,a new object tracking method is proposed.Based on the local binary pattern features（LBP） of the object,the object appearance model can sparse representation simultaneously with the original object template and the template built by some particles in the current frame,and build a joint objective function which can solve the tracking issue through an iterative optimization problem.Experimental results show that the track method can well track with occlusion and illumination issues,as well as the scene sequence with a similar background to the object.

关 键 词：目标跟踪 联合模板 联合目标函数 稀疏表示 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]