Bayesian Probability for Kernel-based Target Tracking

摘要：

In this paper,a Bayesian Probability tracking algorithm framework for kernel-based(BKBT) is proposed,it can performs well in dynamic background scene. The algorithm is based on HSV color feature model,modeling the tracking target by bayesian theory,generating the probability density distribution image,finally,iterations localization the target using the mean shift,experiments show that the BKBT algorithm can successfully and robustly track the target when the background changes

关键词： bayesian probability BKBT truncationfactor hsv

作者: Wang Hao Geng Wei Fang Baofu Hu Xiaorong Meng Fanhui

作者单位: Information School of Computer and Information Hefei University of Technology Hefei city,China

会议类型: 国际会议

会议名称: 2010 International Forum on Computer Science-Technology and Applications(2010 国际计算机科学技术应用论坛 IFCSTA 2010)

会议地点: 南宁

会议语种:英文

页码: 188-191

在线出版日期: 2010-12-10（万方平台首次上网日期，不代表论文的发表时间）