Abstract
To identify mask wearing quickly and automatically in public places is particularly important for epidemic prevention and control. In this paper, we present a real-time mask wearing detection algorithm based on improved YOLOv5s, which speeds up the reasoning speed by 5~10% and achieves a detection accuracy of more than 96%. The proposed algorithm can be easily deployed in Raspberry PI. We also design a Web-based mask wearing detection system consisting of two parts: the cloud subsystem and the edge subsystem. The cloud part mainly realizes data storage, model training, equipment monitoring, big data visualization and other functions. The edge part uses Raspberry Pie as the core deployment equipment to complete data collection, model reasoning, information early warning and other functions. Our system features the advantages of high real-time, low cost and low network traffic. It can be widely deployed in the supermarket, parks, intelligent light poles and other open scenes, resulting in greater practical application value.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have