基于YOLOv5的矿工安全帽佩戴及自救器携带检测研究
Research on Miners’ Helmet Wearing and Self-Rescuer Carrying Detection Based on YOLOv5

为增强矿山工作人员佩戴安全帽和携带自救器的意识，有效预防生产安全事故发生，基于Pytorch框架和YOLOv5目标检测算法，实现了对安全帽和自救器的自动检测。结果显示：模型对安全帽和自救器有较高的检测精度，mAP达到84.1%。使用矿山企业现场的监控录像对模型进行测试，能够准确检测到视频中的安全帽和自救器。
To enhance the awareness of mine workers to wear safety helmets and carry self-rescuers, and effectively prevent production safety accidents, automatic detection of safety helmets and self- rescuers was realized based on the Pytorch framework and YOLOv5 target detection algorithm. The results show that the model has high detection accuracy for safety helmets and self-rescuers, and the mAP reaches 84.1%. The model was tested using surveillance video from the site of a mining enterprise and was able to accurately detect helmets and self-rescuers in the video.