卡尔曼粒子滤波的视频车辆跟踪算法研究

近年来，视频车辆跟踪作为城市智能交通系统（ITS）的一个关键技术受到关注。本文针对传统粒子滤波的非线性、非高斯性可能导致跟踪过程的不稳健性，提出一种基于卡尔曼粒子滤波的视频车辆跟踪算法，该算法利用基于重要区域的目标颜色直方图统计模型对视频车辆目标进行建模，并将其应用于卡尔曼滤波更新中，通过采用Mean Shift算法将卡尔曼滤波器引用到粒子滤波器当中，对车辆的运行轨迹进行校正，实现了局部线性滤波，实现了在保持跟踪系统整体上的非线性、非高斯性的同时，兼顾其局部的线性高斯特性。实验结果表明，本文所提出的方法与传统粒子滤波方法相比，能够更准确地对车辆进行跟踪，同时保证了在复杂环境下性能的稳健性。

Research on Video Vehicle Tracking Algorithm Based on Kalman and Particle Filter

Recently, video vehicle tracking as a key technology of intelligent transportation system(ITS) has got more attention. This paper introduces a video vehicle tracking algorithm based on Kalman and particle filter. The algorithm improves the traditional particle filter, whose non-linear and non-Gaussian may result in non-robustness of tracking process, the algorithm uses the targets color histogram statistical model based on the key regional to model video vehicle, and applies it to update Kalman filter. Then through the use of Mean Shift algorithm, the Kalman filter is added to the particle filter to calibrated the vehicle running tracking so that the experiment achieves a partial linear filtering, maintaining tracking system as a whole on the non-linear and non-Gaussian, and at the same time it takes into account the local characteristics of a linear Gaussian. Experimental results show that the proposed method in comparison with the traditional particle filtering can be more accurate on tracking of vehicles and ensure the robustness of performance in a complex environment.