Person Following Robot with Vision-based and Sensor Fusion Tracking Algorithm

Abstract

The person following robot ApriAttendaTM equipped with a stereo camera and Vision System and LRF is introduced. ApriAttendaTM has the Vision-Based Tracking system and the Vision-Based Motion Control system. ApriAttendaTM can do the person following motion using the tracking information. Moreover, ApriAttendaTM used LRF as another sensor for the tracking performance gain. The respective problems of the vision and LRF tracking systems are pointed out and an improvement method based on the idea of the Vision-LRF Sensor Fusion system is proposed. One feature of this new system is that the fusion rate changes depending on the congestion information of the environment. The experimental movement results of applying these systems to ApriAttendaTM are reported. The efficiency of the proposed method is confirmed by the experiment. As discussed here, efforts to achieve an advanced application using sensors independently are subject to an unavoidable limit. So, a system design integrating information from two or more types of sensor is required. Because the vision data containing abundant information plays a key role in the complex system, further development of the vision system is desirable.