蚁群优化粒子滤波视觉跟踪算法

Ant Optimized Particle Filter For Visual Tracking

针对传统的基于再采样方法的粒子滤波算法存在的样本贫乏现象,提出采用蚁群优化的思想取代再采样步骤,通过有效模拟的蚂蚁觅食的生物特性,抑制样本贫乏现象,从而提高目标跟踪的准确性.实验结果表明,该算法能够有效处理目标快速运动、目标遮挡、交互等难题,表现出较好的鲁棒性.

The thought of ant optimization replacing resampling step is proposed to solve the sample deficiency problem caused by the traditional particle filter based on resampling method. By effectively simulating the foraging behavior of ants, the proposed algorithm can suppress the impoverishment problem of sample and consequently im- prove the target tracking accuracy. The experimental results show that the proposed algorithm is robust and efficient in dealing with the issues of rapid motion, object occlusion and interaction.