(1) School of Electrical Engineering and Automation, Tianjin University, 300072 Tianjin, P.R. China
Abstract:
A static door number recognition system for mobile robots in office buildings is presented. The motivation behind the development of such a system is to update the localization results of dead reckoning or improve other distance-sensor based localization. Emphasis is placed on how to reduce the processing time of a template-matching based recognition method to enable real-time image-base localization for a mobile robot. A gap-based heuristic idea is put forward to increase the recognition accuracy. The experimental results show that door numbers can be rapidly and correctly identified under different environmental conditions and postures of the camera.