Automated Shopping Cart: Reducing Long Queues One Cart At A Time

Abstract

Everyone has experienced waiting in long queues to get their items billed at a shopping cart. With this product, the proposed research work intends to reduce the time spent by shoppers in queues. Our automated shopping cart can reduce the time and manual labor required for weighing and billing the items a user wishes to buy. This article has developed a shopping cart with weight sensors built into the cart to measure the load of the product, and an android application with deep learning models(MobileNetV2 and EfficientNetV2) to classify the required item. The models were trained on a custom-built dataset with 51 classes. The mobile phone camera is used to capture an image of the product. The models are used to classify this item. Once the item is classified, the weight of the item is obtained from the load cells. A database analysis is performed to obtain the price of the item. With these details, the bill is generated as soon as the user is done with shopping without waiting in long queues.