| 一种针对视频对象快速移动和遮挡的改进mean-shift跟踪算法 |
| |
| 引用本文： | 唐勇,孙磊,孙宾.一种针对视频对象快速移动和遮挡的改进mean-shift跟踪算法[J].燕山大学学报,2010,34(1):52-55,66. |
| |
| 作者姓名： | 唐勇 孙磊 孙宾 |
| |
| 作者单位： | 燕山大学,信息科学与工程学院,河北,秦皇岛,066004 |
| |
| 基金项目： | 国家自然科学基金资助项目(50675189) |
| |
| 摘 要： | 针对经典mean-shift算法在目标发生快速移动或受到局部遮挡时容易跟踪失败的问题,提出一种改进的mean-shift算法。首先以Bhattacharyya系数作为目标运动状态判断因子:当目标发生快速移动时,采用SSD(Sumof Square Difference)算法进行全局搜索,以此来指定当前帧mean-shift算法的初始跟踪位置;当目标受到局部遮挡时,采用背景相似度算法,对目标模板和候选模型区域中像素点进行加权,降低背景像素的干扰,然后由SSD算法进行局部匹配,对mean-shift算法的跟踪结果进行校正。经实验表明,改进算法对发生快速移动或受到局部遮挡的目标,能进行有效的跟踪。
|
| 关 键 词： | 目标跟踪 mean-shift 快速移动 遮挡 |
A robust mean-shift tracking method under fast-moving and occlusion |
| |
| TANG Yong,SUN Lei,SUN Bin.A robust mean-shift tracking method under fast-moving and occlusion[J].Journal of Yanshan University,2010,34(1):52-55,66. |
| |
| Authors: | TANG Yong SUN Lei SUN Bin |
| |
| Institution: | College of Information Science and Engineering;Yanshan University;Qinhuangdao;Hebei 066004;China |
| |
| Abstract: | An improved mean-shift-based tracking algorithm was proposed to solve the poor tracking ability under fast-moving and partial occlusion.Bhattacharyya coefficient analyzing was used to determine the tracking status:when the object moves fast,SSD algorithm is applied to choose the start point for the mean-shift tracking;when the object undergoes partial occlusion,firstly the background pdf is used to separate the object region from the background;secondly,SSD algorithm is employed to modify the result from th... |
| |
| Keywords: | object tracking mean-shift fast move occlusion |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
