基于双目视觉的智能跟踪行李车的设计

Design of Intelligent Object Tracking Baggage Vehicle Based on Binocular Vision

利用空间同一点在双摄像机画面上的视差,设计了一种基于双目视觉的智能跟踪行李车系统。采用CMOS摄像头搭建双目视觉系统,并用张正友法对主、从摄像头进行标定。通过对主、从视场内同一运动目标进行检测与匹配,并根据主、从视场中目标上同一点的视差计算出三维空间坐标,从而实现对目标的跟踪。采用模块化设计了系统硬件部分,并详细描述了目标识别算法和行李车运动控制策略。仿真结果表明,系统在单一背景下能较好地跟踪运动目标。

An intelligent object tracking baggage vehicle design based on binocular vision is presented according to the target＇ s paral- lax on the cameras＇ screen. The binocular visual system is composed of CMOS cameras calibrated by the Zhangzhengyou＇ s method. The detection and matching of moving object in different view fields calculate the target coordinates to achieve object tracking. System hardware design is modular. Target recognition algorithms and motion control strategy of baggage vehicle are described in details. Simu- lation results show that the system can track moving targets under a simple background well.