基于粒子滤波模型目标跟踪算法研究

本文基于视频图像中目标的基本特征——灰度特征,提出了一种基于粒子滤波模型实现目标跟踪的方法。文中介绍了粒子滤波模型的原理,并且在该原理的基础上,提出了目标跟踪的算法框架。详细闸述了实现该算法具体的流程,给出了视频序列图像的实时仿真结果。基于粒子滤波模型的目标跟踪算法,不但在非线性非高斯状态下,能够对目标进行有效跟踪,还可以消除由光照变化以及部分遮挡造成的影响,具有较强的鲁棒性。

Study on Object Tracking Algorithm based on Particle Filter Model

This paper propose a algorithm of object tracking based on particle filter model using gray-level in the video sequence. The principium of particle filter model is introduced, based on which the frame of object tracking is put forward. It also illustrates the process of how to implement the arithmetic. Our result indicates that the algorithm can not only track object effectively in Non-linear/Non-Gaussian state, but also eliminate the infection of illumination variety and partly shelter. The arithmetic is robust highly.