基于改进YOLOv5算法的电力工人作业安全关键装备检测

Detection of Key Safety Equipment for Electric Workers Based on Improved YOLOv5 Algorithm

传统的电力工人作业安全关键装备检测算法精度低、鲁棒性差.为此提出了一种基于YOLOv5的改进型目标检测算法,用于对绝缘手套、安全帽、作业人员进行检测.首先在原版YOLOv5中加入坐标注意力模块,提升对小目标特征提取的能力;其次用加权双向特征金字塔网络结构代替原有的特征金字塔网络结构,进一步提升特征提取的能力.实验结果表明平均精确度提升了1.8%,其精度达96.4%,平均精确度均值提升了0.4%,其均值达93.3%.所提算法改善了原版网络对小目标容易漏检、误检的问题,具有较强的实用性和先进性,能满足实时检测的要求,对电力行业安全有一定的促进作用.

The traditional detection algorithm for key equipment of electric workers′work safety has low accuracy and poor robustness.Therefore,an improved target detection algorithm based on YOLOv5 is proposed to detect insulating gloves,safety helmets and operators.Firstly,the coordinate attention module is added to the original YOLOv5 to improve the ability of small target feature extraction;Secondly,the weighted bi-directional feature pyramid network structure is used to replace the original feature pyramid network structure to further enhance the ability of feature extraction.The experimental results show that the average accuracy is improved by 1.8%,up to 96.4%,and the average accuracy is improved by 0.4%,up to 93.3%.The proposed algorithm improves the problem that the original network is easy to miss detection and falsely detect small targets.With strong practicability and progressiveness,it can meet the requirements of real-time detection and promote the security of the power industry.