基于FasterRCNN的口罩佩戴检测研究

Research on Mask Wearing Detection Based on Faster RCNN

在新冠病毒(COVID-19)全球肆虐的大背景下,为有效阻止新冠病毒在人群中传播,多地要求在公众场合佩戴口罩,针对这一问题,提出一种基于FasterRCNN算法的口罩佩戴检测方法,手动通过labelimg进行数据集标定,再对深度学习模型进行训练,实验结果表明,该深度学习算法的检测的准确率达到97.57%,可以有效检测人员的口罩佩戴情况。

In the context of the global raging COVID-19 virus(COVID-19),in order to effectively prevent the spread of the new coronavirus among the population,many places require masks to be worn in public.To address this problem,a mask wearing test based on the Faster RCNN algorithm is proposed.Methods:Manually calibrate the data set through labelimg,and then train the deep learning model.The experimental results show that the detection accuracy of the deep learning algorithm reaches 97.57%,which can effectively detect the wearing of masks.