约束条件下的优选进化目标跟踪算法分析

1.西安建筑科技大学 信息与控制工程学院,陕西 西安 710055; 2.北京航空航天大学 软件学院,北京 100191

视频监控; 目标跟踪; 遗传算法; 粒子滤波

Analysis of target tracking algorithm of preferred evolutionary under constraints
WANG Yan-ni1,WANG Jian-ping2

(1.School of Information and Control Engineering,Xi'an University of Architecture and Technology,Xi'an 710055,China; 2.School of Software,Beihang University,Beijing 100191,China)

video surveillance; target tracking; genetic algorithm; particle filter

DOI: 10.13800/j.cnki.xakjdxxb.2017.0416

备注

针对传统粒子滤波算法的退化和容易受到背景干扰等问题,结合遗传算法思想,提出一种约束条件下的优选进化粒子滤波的视频目标跟踪算法。首先采用约束条件检测目标,在参考帧生成初始父代粒子种群。利用遗传算法的特点,设置优选进化策略,选择优秀粒子遗传到子代。其次通过制定合适的交叉及变异规则,对未选中的粒子进行相应的运算直到迭代次数。最后通过每个粒子的权值计算其状态,获得运动目标轨迹。仿真实验结果表明,该方法不受背景变化等影响,不仅实现简单,而且有较稳定的跟踪效果。

In order to solve the problem of the degradation of traditional particle filter algorithm and the interference of background,a new target tracking algorithm of preferred genetic evolutionary is proposed.Firstly,according to the constraints,the moving targets are detected and initial parent particle population are generated in the reference frame.The excellent particles are selected to be offspring based on optimal evolution strategies.Then,making appropriate crossover and mutation rules,the unselected particles are not selected to perform the operation until the number of iterations.Finally,the target trajectory is obtained by calculating object states of each particle.Simulation results show that the new algorithm is not affected by the background changes.It is not only simple,but also more stable for tracking effect than the existing algorithm.