摘要
针对目前主流口罩佩戴检测算法均需要对样本进行标注,使用的网络模型对电脑硬件配置要求高,无法适用于便携设备或移动端的问题,提出了一种基于MobileNet V2的口罩佩戴识别方法。首先,对口罩佩戴数据集进行构建,并进行数据扩充;然后,通过搭建MobileNet V2模型实现对口罩佩戴的识别;最后,通过选择合适的评价标准进行结果分析。结果表明:该方法对是否佩戴口罩的检测准确率可达99.83%,对口罩佩戴是否标准的检测准确率可达98.97%。该方法在保证准确率和速度的基础上,减小网络体积,适用性更加广泛。
In view of the problem that the current mainstream mask wearing detection algorithms all need to mark the samples,and the network model used has high requirements for computer hardware configuration,which is not suitable for portable devices or mobile terminals,this paper proposes a mask wearing identification method based on MobileNet V2.Firstly,the mask wearing data set was constructed and the data was expanded.Then,the MobileNet V2 model was built to realize the mask wearing recognition.Finally,the results were analyzed by selecting appropriate evaluation criteria.The results show that the detection accuracy of this method can reach 99.83%for whether the mask is worn or not and 98.97%for whether the mask is worn correctly or not.On the basis of ensuring the accuracy and speed,this method reduces the network volume and has wider applicability.
作者
金映谷
张涛
杨亚宁
王月
楚艳丽
JIN Ying-gu;ZHANG Tao;YANG Ya-ning;WANG Yue;CHU Yan-li(School of Electromechanical Engineering,Dalian Minzu University,Dalian Liaoning 116605,China;School of Information and Communication Engineering,Dalian Minzu University, Dalian Liaoning 116605, China)
出处
《大连民族大学学报》
2021年第5期404-409,431,共7页
Journal of Dalian Minzu University
