基于序列蒙特卡罗滤波的人脸跟踪算法

Face Tracking Algorithm Based on Sequential Monte Carlo Filter

结合图像的色彩分布和空间布局,提出了一种基于HSV色彩和空间信息的序列蒙特卡罗滤波人脸跟踪算法。通过比较采样值和期望值的特征距离来计算采样状态对应的权值。利用加权采样值来估计未知量后验概率,当采样数趋于无穷时,由大数定理保证采样值分布逼近于真实值分布。仿真实验给出了利用加权采样对人脸跟踪的结果。实验表明,基于序列蒙特卡罗的人脸跟踪算法计算简单有效,能够准确预测人脸的位置并且很好地跟踪其运动轨迹。

Incorporating color distribution and spatial layout, this paper proposes a sequential Monte Carlo filter tracking face algorithm using color and spatial information in HSV color space. By computing the distance between sample and target, different weights associated with every sample and the posterior state vector can be computed. The samples distribution trends to the state distribution, whose validity is guaranteed by the strong law of large numbers （SLLN）. Experimental results show the efficient face tracking performance using weighted sampling.