본문 바로가기
  • Home

Implementation of Unmanned Payment System Based on Object Detection

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2026, 12(2), pp.137~145
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : December 14, 2025
  • Accepted : February 21, 2026
  • Published : April 30, 2026

Je-Hui Lee 1 Dong-Geon Kim 1 Hyo-won Kim 1 Chae-Yeon Baek 1 KIM, TAEKOOK 1

1국립부경대학교

Accredited

ABSTRACT

This study aims to design and implement an object detection–based unmanned payment system to alleviate the high infrastructure costs and user reliability issues associated with conventional automated checkout technologies. The proposed system employs a YOLO-based object detection model to recognize products placed in a shopping cart in real time and applies a frame-accumulation-based tracking algorithm to compensate for transient recognition errors, thereby enabling robust identification of payable items. In addition, BLE(Bluetooth Low Energy) beacons are utilized to automatically detect customer entry and exit, while a mobile application provides real-time visualization of a virtual shopping cart and electronic receipts. Experimental results demonstrate that the proposed approach achieves stable recognition performance across various real-world retail environments and effectively reduces payment errors caused by momentary detection failures. The proposed system presents a lightweight unmanned payment architecture that can be implemented without expensive sensor infrastructure, and is expected to contribute to the practical adoption of unmanned services in small- and medium-sized retail stores.

Citation status

* References for papers published after 2024 are currently being built.