首页 | 本学科首页   官方微博 | 高级检索  
     

一种基于背景加权直方图的Mean-Shift跟踪算法
引用本文:刘欢敏,龙永新,张长隆.一种基于背景加权直方图的Mean-Shift跟踪算法[J].湖南工业大学学报,2015,29(6):69-73.
作者姓名:刘欢敏  龙永新  张长隆
作者单位:School of Computer and Communication,Hunan University of Technology,Hunan Keyshare Information Technology Co.,Ltd.,School of Computer and Communication,Hunan University of Technology and College of Electrical and Information Engineering,Hunan University, Hunan Keyshare Information Technology Co.,Ltd.
基金项目:国家自然科学基金资助项目(61170102)
摘    要:经典的Mean-Shift算法以颜色直方图为特征进行跟踪,但在视频目标跟踪中没有有效地利用跟踪目标的空间信息,因而当目标快速移动时,目标极易丢失。针对这一问题,提出了一种基于距离目标形心加权直方图的Mean-Shift跟踪算法。距离目标形心加权直方图可有效利用目标在图像中的位置信息,从而能实现复杂背景下的目标跟踪。并分别用改进的基于距离目标形心加权直方图的Mean-Shift跟踪算法和经典的Mean-Shift跟踪算法,对快速运动的目标进行跟踪实验。结果表明,改进的基于距离目标形心加权直方图的Mean-Shift跟踪算法,具有较强的鲁棒性和实时性,能有效实现复杂场景下的目标实时跟踪。

关 键 词:目标跟踪  Mean-Shift算法  加权直方图  HSV颜色空间
收稿时间:2015/9/20 0:00:00

A Mean-Shift Tracking Algorithm Based on Background Weighted Histogram
Abstract:The classic Mean-Shift tracking algorithm, characterized by the color histogram, does not make full use of the position information of the target, therefore it is easy to lose the target as the target fast moving. Proposes a Mean-Shift tracking algorithm based on weighted histogram of the distance to target center.The weighted histogram effectively uses the target spatial information in the image to achieve the target tracking in complex background. Applies respectively the improved Mean-Shift tracking algorithm and classic Mean-Shift tracking algorithm for fast moving target tracking tests. The result shows that the improved algorithm has good robustness and real-time performance and effectively achieves the target tracking in complex scenes.
Keywords:
本文献已被 CNKI 等数据库收录!
点击此处可从《湖南工业大学学报》浏览原始摘要信息
点击此处可从《湖南工业大学学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号