基于深度学习的口罩佩戴状态目标检测系统设计  

Design of Mask Wearing Status Target Detection System Based on Deep Learning

作　　者：焦双健[1] 孙萌雪 Jiao Shuangjian;Sun Mengxue(College of Engineering,Ocean University of China,Qingdao 266100,China)

机构地区：[1]中国海洋大学工程学院,青岛266100

出　　处：《单片机与嵌入式系统应用》2021年第11期5-9,共5页Microcontrollers & Embedded Systems

摘　　要：提出一种基于深度目标检测方法的口罩佩戴识别检测系统,以有效监督人们佩戴口罩情况。该系统首先通过对大量不同场景下正确佩戴口罩、错误佩戴口罩、未佩戴口罩三类图片数据进行网络训练,然后利用训练好的离线网络模型即可对监控视频中的人员口罩佩戴情况进行实时自动监督,当检测出错误佩戴口罩以及未佩戴口罩人员时,自动发出报警,提醒相关区域人员正确佩戴口罩。In the paper,a mask wearing identification and detection system based on deep target detection method is proposed to effectively supervise people wearing masks.The system first trains the network by a large number of different scenarios wear a face mask,errors,wearing masks,not wearing masks three kinds of image data,then trains offline network model for monitoring video real-time automatic supervision of the personnel wore.When it is detected that the person wearing the mask incorrectly or not,it will automatically send an alarm to remind the personnel in the relevant area to wear the mask correctly.

关 键 词：口罩佩戴状态 深度学习 卷积神经网络 YOLO 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]