Safety Helmet Detection: Adding Attention Mechanism to Yolov8 to Improve Detection Accuracy

Economic growth is inseparable from the construction of infrastructural facilities. In daily infrastructure construction and production, it is crucial to ensure the safety of construction and operating personnel. Major enterprises have invested a lot of energy in ensuring production safety. Safety helmets can effectively reduce external impact to a certain extent, and wearing safety helmets can effectively ensure the safety of construction workers. With the development of science and technology, deep learning technology is widely used in safety helmet detection in construction sites. This paper designs a construction site safety helmet-wearing detection system based on Yolov8 to identify construction workers wearing safety helmets. It meets the requirements of general scenarios by improving the network structure of YOLOv8 and adding an attention mechanism to improve the accuracy of safety helmet recognition. This system not only has fast training speed but also does not require a large dataset. Thus, it can train a higher-precision model with a smaller dataset and has a greater accuracy advantage compared to other models. During the production process, it can meet the requirements for the accuracy of safety helmet detection. Related experimental results also verify the effectiveness and superiority of our proposed model.