| Since December 2019,the COVID-19 has been at a high state.Wearing a mask is the easiest and most effective way to cut off the transmission of the virus and inhibit the rebound of the epidemic.Therefore,wearing masks in public places has become a requirement under the normalization of epidemic prevention.Individuals often do not wear masks because of their weak awareness of epidemic prevention,which brings great challenges to our epidemic prevention work.At present,the detection of mask wearing in public places is mainly carried out by manual detection,which not only consumes a lot of human resources,but also is prone to false detection and missed detection.It is impossible to accurately monitor the wearing of masks by pedestrians in various areas of public places in real time.To solve this problem,this paper develops a mask wearing detection system based on deep learning by applying object detection technology.The system supports the detection of pedestrians wearing masks in real-time video,and provides statistical analysis functions.It mainly includes six modules,which are real-time video detection,key point profile,offline video detection,mask wearing situation statistics,picture detection,and system configuration management.The real-time video detection module mainly realizes the real-time detection of the video captured by the camera;the key point profile mainly shows the mask wearing detection and potential risk level of different detection areas;the mask wearing statistics mainly provides statistical analysis of the detection results;the offline video detection and picture detection mainly realize the detection of the video and pictures uploaded by users,and the system configuration management mainly realizes the user management and device management functions.In terms of algorithms,Mask wearing dataset was created.Modify and train the YOLOv5 algorithm model,and integrate it with the attention mechanism to realize the mask wearing detection algorithm based on YOLOv5-Attention,it improves the precision of mask wearing detection In terms of system technology architecture,this paper uses Spring Boot as the development framework of the back-end system,Mybatis as the data persistence framework,My SQL as the database,and Bootstrap framework to realize the development of the front-end page.In terms of system implementation,the system requirements analysis,outline design and detailed design are completed in accordance with the software development life cycle.Finally,based on the design and implementation,the system is tested and the work is summarized and prospected. |
