基于深度学习的口罩佩戴检测与人群聚集预警系统

Mask Wearing Detection and Crowd Gathering Warning System Based on Deep Learning

为了提高新冠疫情防控的效率和范围,提出一种基于深度学习的口罩佩戴检测与人群聚集预警系统,该系统包含口罩佩戴检测、人群聚集检测以及智能场景分类模块。智能场景分类模块使用深度学习分类算法自动识别摄像头机位种类,从而对低机位的摄像头中的画面进行准确且实时的口罩佩戴检测,人群聚集检测模块能够快速计算出高机位的广角摄像头画面中的人数,判断是否存在大规模人群聚集,从而有效提高疫情防控的效率与范围。实验证明,系统在各种不同的摄像头画面下均能准确判断摄像头机位并进行准确快速的口罩佩戴检测以及人群聚集检测。

In order to improve the efficiency and scope of COVID-19 prevention and control,this paper proposes a mask wearing detection and crowd gathering warning system based on deep learning,which includes mask wearing detection,a crowd gathering detection,and an intelligent scene classification module.The intelligent scene classification module allows the system to automatically identify the height of the current camera position and perform a real-time mask wearing detection algorithm for the low-position camera.The crowd gathering detection module can calculate the number of people in the high position camera instantly to determine whether there is a large-scale crowd gathering.The experiment proves that the system can accurately determine the camera position and perform accurate and real-time mask wearing detection and crowd gathering under different positions of the camera.