首页 | 本学科首页   官方微博 | 高级检索  
     

基于SIFT特征一致性的目标跟踪算法
引用本文:黎云汉,楼京京.基于SIFT特征一致性的目标跟踪算法[J].控制工程,2011,18(6):966-969.
作者姓名:黎云汉  楼京京
作者单位:义乌工商职业技术学院机电信息分院,浙江义乌,322000
基金项目:义乌市科技计划项目(09-1-31)
摘    要:针对光照条件突然变化情况下混合目标模型Mean Shift算法无法准确跟踪目标的缺点,提出了一种基于SIFT特征一致性的目标跟踪算法.算法用SIFT特征来匹配帧间的感兴趣区域,同时使用包含初始帧信息和前一帧信息的混合目标模型Mean Shift算法计算帧间感兴趣区域的直方图,以直方图分布距离最小为原则计算Mean Sh...

关 键 词:mean  shift  SIFT  目标跟踪

Object Tracking Algorithm Based on SIFT Features Correspondence
LI Yun-han , LOU Jing-jing.Object Tracking Algorithm Based on SIFT Features Correspondence[J].Control Engineering of China,2011,18(6):966-969.
Authors:LI Yun-han  LOU Jing-jing
Affiliation:LI Yun-han,LOU Jing-jing(School of Electro-mechanical and Information Technology,Yiwu Industrial & Commercial College,Yiwu 322000,China)
Abstract:To the problem that the mean shift algorithms with hybird target model can not track the object accurately when sudden changes in illumination,a tracking algorithm based on SIFT features correspondence is proposed.The algorithm uses SIFT features to match the region of interests across the frames.The mean shift algorithm of hybrid target model using the information of the first frame and the previous frame is applied to conduct the similarity search via color histograms simultaneously.The mean shift vector ...
Keywords:mean shift  SIFT  object tracking  
本文献已被 CNKI 万方数据 等数据库收录!
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