基于unscented粒子滤波的红外弱小目标跟踪

Tracking of infrared small target based on unscented particle filtering

为有效解决非线性环境中的红外弱小目标跟踪问题,提出基于unscented粒子滤波的目标跟踪算法。状态转移先验概率中未考虑当前测量对状态估计的作用,为克服传统粒子滤波算法采用状态转移先验概率作为粒子滤波建议分布的缺点,采用UKF生成粒子滤波的建议分布(UPF),并从中抽样粒子。由于考虑到当前观测值在状态后验估计中产生的影响,改善了目标状态估计的性能,且实验所需粒子数目大大少于传统粒子滤波算法所需粒子数目。用实际红外图像对所提算法做了仿真实验,结果表明,用该方法得到的状态估计结果优于用传统粒子滤波算法和用扩展卡尔曼滤波作为建议分布的粒子滤波算法获得的结果。

A new sequential Monte Carlo method for is proposed tracking infrared point target. Owing to the transition prior does not take into account the current observation, so the effect of lots of particles could be negligible. Instead of using transition prior as proposal distribution, UKF （unscented Kalman filter） is used to generate the proposal distributions （UPF）, which takes the current measurement into account so as to improve the tracking performance greatly with less particles. To evaluate the efficiency of the UPF in the field discussed, It is applied to the real infrared point target tracking and the obtained results are compared with generic particle filtering and the particle filtering with EKF generating proposal distribution. Experimental results show that UPF has advantages in the field of state estimation problem.