融合Kalman滤波的自适应带宽Mean Shift算法

来源期刊:中南大学学报(自然科学版)2011年第z1期

论文作者:傅荟璇 王宇超 孙枫

文章页码:784 - 788

关键词:目标跟踪;Mean Shift算法;Kalman滤波

Key words:object tracking; Mean Shift algorithm; Kalman filter

摘    要:针对Mean Shift算法在复杂背景中存在颜色干扰问题和核函数带宽固定的缺陷,提出一种融合Kalman滤波的自适应带宽Mean Shift跟踪算法。将Mean Shift算法得到的目标位置作为Kalman滤波的测量值,利用Kalman滤波预测出下一帧目标的位置,以预测值作为Mean Shift算法迭代运算的初始值,并以仿射变换来描述目标尺寸变化,利用连续两帧中匹配窗口的最大相关系数确定最优匹配窗口搜索目标。跟踪试验结果表明:算法在背景与目标颜色相近和目标尺寸变化等复杂情况下都能对目标进行准确跟踪,具有较强的抗干扰能力。

Abstract: According to the defects that Mean Shift algorithm exists color interference in a complex background and kernel function bandwidth of Mean Shift was not changeable, integrated Kalman filter to adaptive bandwidth Mean Shift Tracking algorithm was proposed. Taking object position obtained from Mean Shift as Kalman filter measure value, Kalman filter is used to get the predicted starting position of Mean Shift in every frame. The object size change by the affine transformation is described, and the window is matched using continuous two frames in the maximum correlation coefficient to determine the optimal matching window search object. The tracking experiment shows that the algorithm in the complex situation, such as background and the object having similar color and the object size changing, can carry on the accurate tracking to the object, showing a good anti-interference ability.

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号