移动机器人部分遮挡条件下多部件行人跟踪算法

Multi-part Pedestrian Tracking Algorithm in Part Occlusion for Mobile Robot

针对复杂街道环境中的交通标志与公共设施对行人造成不同程度的遮挡而致使移动机器人跟踪目标失败,提出一种多部件粒子滤波器行人跟踪方法,引入弹簧弹力负载决策,降低单一跟踪器的错误风险,增强部分遮挡条件下的视觉目标跟踪鲁棒性。仿真分析表明,该算法能够在行人部分遮挡时呈现较好跟踪效果,降低街道复杂背景环境对跟踪的影响,为后续移动机器人视觉目标避障等应用提供支撑。

The obstacle of the traffic sign and facilities would cause different degrees of occlusion to up-right person object tracking system in the mobile robot visual tracking of the complex street environment. A multi-part pedestrian tracking algorithm based on particle filter is proposed for solving the part-occlusion tracking failure problems resulting in a wrong bounding box for the whole body. A novel spring elastic-load decision was introduced to coordinate and calibrate the actual position of each pedestrian part,in order to decrease the error risk of using a single tracker and increase the visual object tracking robustness under the partial cover conditions. With the simulation results,the algorithm achieves better performance in part occlusion of pedestrian,and weakens the wrongly tracking effects in complex street fore-and back-ground,so that it can provide a support for mobile robot visual object obstacle avoidance applications.