CounterVision: Self-Checkout System

  • Unique Paper ID: 170235
  • Volume: 11
  • Issue: 6
  • PageNo: 3235-3242
  • Abstract:
  • Tremendous change has happened in the retail industry, especially in higher demand for efficient, touchless, and friendly systems of checkout [4]. This work proposes a camera-based self-checkout system known as CounterVision. This must replace common methods of barcodes by using advanced computer vision and deep learning techniques [8]. CounterVision employs the object detection model called YOLOv7 to instantly identify and process several products in challenging retail scenarios that depend on the changing lighting conditions and filled shelves [1]. In this work, the system is used with a mounted single camera with background subtraction and bounding box tracking to minimize the computational costs with high discovery accuracy [8]. Extensive testing resulted in the system achieving a mean Average Precision of 1.0 and recall of 0.97 at 0.5 IoU and 0.4 confidence threshold, underlining its robustness and reliability [1][8]. Although it does away with bar codes, CounterVision is one of the automated retailing solutions for cost-effectiveness by eliminating waiting time and improving the speed of operations; this means human error is eliminated [8]. This paper explores the actualities of scalability and diversity in datasets and more potential upgrades and use cases that can include edge computing and real-time connectivity for inventory management systems [5][8]. Reconfiguring checkout enables CounterVision to make retail places smarter, faster, and more reliable [4].

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 6
  • PageNo: 3235-3242

CounterVision: Self-Checkout System

Related Articles