Revolutionizing Retail with Touchless Cart Technology Through Gesture-Controlled Shopping

The COVID-19 pandemic brought the demand for non-contact technologies to the next level in high-traffic areas such as supermarkets, where physical contact should be minimized for hygiene and safety. Grand View Research projects the global touchless sensing industry will reach a size of USD 15.3 billion by 2027. This project thus presents a fully non-contact solution for shopping based on both digit and alphabetic gesture recognition technology. Custom trained CNN model on the American Sign Language (ASL) dataset is used to make real-time hand gesture recognition able to select items, specify quantities, and place orders without touching carts or checkout terminals by making simple hand movements. It processes the hand gestures with great accuracy-93% accuracy on the custom dataset which recognizes both digits (0-9), as well as alphabets from A-Z. Real-time visual feedback and gesture validation with animated confirmation loops are the strengths of this system and lessen errors due to intuitive interaction with the user. The system is designed to be accessible for users with disabilities while also targeted to future voice command interfaces. Future improvements include broadened categories of products, wider gesture recognition, and integration with vending machines. This is an innovative hygienic shopping experience using practical applications of machine learning and computer vision through custom gesture recognition.