IMPROVED ALIGNMENT ESTIMATION FOR AUTONOMOUS DOCKING OF MOBILE ROBOTS

摘要

This paper presents the methodology for an improved visual tracking, intended for the autonomous control of mobile robots. We propose the cumulated use of the image processing method and positioning sensors for an improved visual tracking. The method primarily uses detection, tracking, and recognition techniques to locate the targets. The use of such methods will enable the robot to monitor the real-time continuous changes in the orientation and the alignment with respect to the target module. The proposed methodology is implemented on a subsisting genderless coupling mechanism which is integrated into a multi-directional hybrid locomotion module to test the alignment accuracy in an autonomous docking procedure. The long-term objective is to demonstrate the autonomous docking and self-reconfiguration of multiple mobile robots.