Mask detection device based on YOLOv3 framework

In early 2020, novel coronavirus pneumonia broke out. In order to prevent the spread of the disease, governments around the world asked the masses to wear masks. However, there are still many people who do not wear masks in public places. To solve this problem, this paper proposes a mask detection device based on yolo3 framework. The device uses the yolov3 algorithm to extract the face prediction area, and uses the gray image to calculate the skin exposure rate of the mouth and nose of the face, so as to judge whether the recognized person is wearing a mask or not and whether the mask is wearing correctly. The model is deployed on the hardware to facilitate the staff to carry the detection. The experimental results show that the recognition rate is 86.6%.