一种WSN下的Mean Shift跟踪算法*

为有效地利用无线传感器网络跟踪移动目标，提出了一种基于高斯混合模型的Mean Shift跟踪算法。该算法运用高斯混合模型描述网络区域内目标信号分布的统计特征，利用Mean Shift区分目标信号与环境噪声，并对目标进行定位与跟踪。仿真实验结果表明，该算法在网络存在较大噪声，特别是网络存在大量异常传感器节点读数的情况下，定位精度高、受异常传感器节点读数影响较小。较之以往无线传感器网络目标跟踪算法，该算法具有更好的鲁棒性。

Novel Mean Shift algorithm for target tracking using WSN

This paper presented a Gaussian mixture model based Mean Shift tracking algorithm so as to effectively track the mobile target using wireless sensor networks. Firstly used Gaussian mixture model to describe the statistical characteristics of the target signals distributed in wireless sensor networks. Then used Mean Shift algorithm to discriminate the target signals from the background noises and estimate the target position. The simulation results show that the target location can be estimated accurately with the algorithm even when lots of outlying sensor readings exist in sensor networks. Compared with other existing algorithms, more robustness can be achieved in the algorithm.