基于YOLOv8算法的电厂场景安全帽佩戴检测识别方法

Detection and Recognition Method of Safety Helmet Wearing in Power Plant Scene Based on YOLOv8 Algorithm

安全帽佩戴对于电厂施工人员的安全至关重要,但在复杂的电厂环境中施工人员难免会出现掉帽的情况。为了判断施工人员是否佩戴安全帽,文章提出了一种基于YOLOv8电厂场景的安全帽检测识别方法。针对开源安全帽数据集在电厂场景样本数量不足的问题,采集、清洗并标注电厂场景数据,重新构建安全帽数据集。基于ultralytics框架,采用YOLOv8 Nano神经网络模型对数据集进行训练,得到FPS为91.7,AP50为93.5%的网络模型。实验结果表明:该方法有效和快速检测施工人员是否佩戴安全帽,具备较好的应用效果。

Safety helmet wearing is crucial to the safety of power plant construction personnel,but the construction personnel inevitably fall off the helmet in the complex power plant environment.In order to judge whether the construction personnel wear safety helmets,this paper proposes a safety helmet detection and recognition method based on the YOLOv8 power plant scene.Aiming at the problem of insufficient number of samples in the power plant scene of the open source safety helmet data set,the power plant scene data is collected,cleaned and labeled,and the safety helmet data set is reconstructed.Based on the ultralytics framework,the YOLOv8 Nano neural network model is used to train the data set,and a network model with FPS of 91.7 and AP50 of 93.5%is obtained.The experimental results show that this method can effectively and quickly detect whether the construction personnel wear the safety helmet,and has a good application effect.