Research on safety helmet wearing detection algorithm based on improved YOLOv8

Safe production, as an important part of the national economy and social development, it's directly related to the safety of people's lives and the stability of society. Through statistical analysis, it is found that about 80% of safety accidents are related to safety helmets. The reason is that most of them are caused by the victims not wearing safety helmets correctly according to the safety management regulations. It can be seen that wearing safety helmets correctly is a very important protective measure in the process of safety production management, and it should be paid enough attention. This paper mainly studies the application of safety helmets in high-risk enterprises such as construction and chemical industry. Through practical investigation, the existing safety helmet wearing detection algorithm has a poor detection accuracy due to the interference of distance, illumination and occlusion, this paper proposes a new algorithm for safety helmet wearing detection based on improved YOLOv8. After the algorithm is improved, the results show that the new algorithm proposed in this paper has an accuracy of 95.6% and a recall rate of 94.8%, which fully meets the accuracy requirements of safety helmet wearing detection in the process of safety production.