一种基于粒子滤波的红外目标跟踪方法

Infrared object tracking based on particle filter

为解决红外运动目标跟踪中的遮挡、形变等问题,提出一种基于粒子滤波的跟踪方法。该方法首先利用目标区域的灰度分布,建立了一种基于统计直方图的系统观测概率模型。并将飞机目标的运动看作惯性受限的非平稳过程,采用微分线性拟合模型作为系统状态转移模型。序列图像的实验表明:该算法能够在目标高速运动或发生遮挡的情况下稳健跟踪目标,其总体性能优于Mean Shift算法。

To deal with the problem of occlusion and deformation,a novel infrared moving object tracking method based on particle filter is presented in this paper.Firstly,it utilizes the intensity distribution to represent infrared object,and constructs the observation probability model by statistical histogram.Then,it treats the motion of aircraft as non-stationary process of confined inertia,and introduces the linear fitting prediction as state transition model.Experimental results on sequential images show that our method can track steadily when the object moves fast or is occluded,the overall performance of the proposed method is better than Mean Shift algorithm.