With the rapid development of information technology,the safety production of coal mines has been sig-nificantly improved,and various downhole safety systems have played an important role in the safety work of coal mines.Based on computer vision technology,by existing monitoring equipment or through the deployment of new equipment at the downhole construction site,this paper adopts the relevant methods of machine learning to auto-matically identify the wearing of safety helmets for miners,which realizes the rapid monitoring of the wearing of safety helmets for operators,reduces supervision risks,enhances the prevention awareness of workers and improves the informatization level of supervision.