基于核函数的Camshift算法的收敛性分析  

CONVERGENCE ANALYSIS OF KERNEL BASED CAMSHIFT ALGORITHM

作　　者：曲巨宝[1] 

机构地区：[1]武夷学院数学与计算机系,福建武夷山354300

出　　处：《数值计算与计算机应用》2011年第2期148-158,共11页Journal on Numerical Methods and Computer Applications

摘　　要：Camshift算法因其具有自适应更新核窗、克服目标形变和光照变化等性能,而被广泛应用于智能跟踪和图像处理领域.但针对其收敛性基础理论研究,目前还没有相关报道.介绍了Camshift算法,指出了其在目标建模中静噪能力差等问题,提出了基于核函数的Camshift算法,并对算法的收敛性进行了严格的数学证明.实验和仿真结果表明,改进后的算法具有更好的鲁棒性和收敛速度,能够适应更为复杂的跟踪环境.The Camshift algorithm, because it＇s auto-adapted to renew the kernel window, over- coming target deformation and illumination change and so on, is used applied in the field of the intelligent track and the image processing widely. But about its convergence basic study, at present, it has not had the related report. Introduces the Camshift algorithm and points out some questions the bad squelch ability in target modeling and so on, proposes the Camshift algorithm that based on the kernel function, and carries on the strict mathematics proof to the algorithm convergence. The experiment and the simulation result indicates that after the improvement, algorithm has a better robustness and the convergence rate that can adaot to the more complex track circumstances.

关 键 词：CAMSHIFT 核函数 收敛性 跟踪 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]