| 基于多特征融合的均值偏移视频目标跟踪算法 |
| |
| 引用本文： | 王宾,潘建寿,王琳,崔宇巍. 基于多特征融合的均值偏移视频目标跟踪算法[J]. 小型微型计算机系统, 2006, 27(9): 1746-1749 |
| |
| 作者姓名： | 王宾 潘建寿 王琳 崔宇巍 |
| |
| 作者单位： | 西北大学,信息科学与技术学院,陕西,西安,710069 |
| |
| 摘 要： | 为了有效地实现复杂环境下视频目标跟踪，在分析了基于颜色分布的均值偏移跟踪算法的基础上，提出了一种联合空间信息的改进均值偏移算法，此算法是将目标空间位置预测与均值偏移算法定位结果通过加权得到最终目标位置，实验表明，改进算法将目标颜色分布与目标空间运动信息相结合，保证了跟踪效果的稳定性和鲁棒性．
|
| 关 键 词： | 均值偏移算法 空间特征 目标跟踪 颜色分布 |
| 文章编号： | 1000-1220（2006）09-1746-04 |
| 收稿时间： | 2005-06-08 |
| 修稿时间： | 2005-06-08 |
Video Target Tracking of Mean-Shift Algorithm Based on Multi-Feature Fusion |
| |
| WANG Bin,PAN Jian-shou,WANG Lin,CUI Yu-wei. Video Target Tracking of Mean-Shift Algorithm Based on Multi-Feature Fusion[J]. Mini-micro Systems, 2006, 27(9): 1746-1749 |
| |
| Authors: | WANG Bin PAN Jian-shou WANG Lin CUI Yu-wei |
| |
| Abstract: | Efficient video target tracking is challenging problem under complex environments.Mean-Shift Algorithm(MSA) based on color distribution is analyzed.A target tracking of MSA combining spatial feature is proposed in this paper,it is an improvement of MSA.The improved algorithm is weighting the spatial estimating location and the result of MSA to obtain the final location of target in this paper.In the experiments,the improved algorithm increased the reliability and robustness of tracking because of integrating the color distribution with spatial feature of target. |
| |
| Keywords: | mean-shift algorithm spatial feature target tracking color distribution |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
