Store merchandise recognition system using image processing with RFID tag information for autonomous stocking and disposal tasks

In recent years, since the varieties of services provided in convenience stores have been increasing, the employees are required to conduct a lot of works simultaneously. The merchandises are sometimes not re-filled nor displayed properly and therefore the store losses the chance for selling the items. To solve the problem, we propose the autonomous merchandise stocking and disposal robotic shelf. In order to manipulate merchandises by robotic hand, the recognition of the kind of merchandise with its position and orientation are inevitable. As the first step to realize the autonomous shelf, in this paper, we propose the store merchandise recognition system using image processing with RFID tag information. We implemented prototype of the autonomous shelf equipped with RFID tag reader, CCD camera, and distance measuring sensor. We also conducted the experiments with the box type merchandise and confirmed the efficiency of the proposed system.