基于压缩感知的粒子滤波跟踪算法

针对运动目标跟踪存在的目标遮挡和光照变化问题,提出一种基于压缩感知的粒子滤波跟踪算法。将改进的压缩感知跟踪算法提取的特征融合到粒子滤波跟踪框架中,并对压缩感知提取的特征和原始粒子滤波中的颜色特征进行可信度判定,能够较好地处理图像序列中由于目标遮挡和光照变化所带来的影响。此算法在公开数据库中进行测试,实验结果表明,提出的算法与已有改进压缩感知跟踪算法和粒子滤波跟踪算法相比,鲁棒性更好,能准确实时地对目标进行跟踪。

Particle filtering tracking based on compressive sensing

To deal with the target occlusion problem and illumination changes in moving target tracking, a particle filtering algorithm based on compressive sensing is proposed. The extracted features are added by compressive sense of the improved compressive tracking (CT) algorithm into the framework of particle filtering tracking. The credibility of extracted features including the color features of original particle filtering and compressive sensing features is judged, which deals with the target occlusion effects and illumination changes. The algorithm is tested in public database and experimental results show that the proposed algorithm brings about better robustness and tracks targets accurately in real time in comparison with the improved CT algorithm and particle filtering algorithm.