基于改进YOLOv5s的车载人员安全带行为检测

Seatbelt Behavior Detection of Vehicle Occupants Based on Improved YOLOv5s

车载人员佩戴安全带行为的检测对于人的生命安全保障具有重要作用;针对目前车内复杂环境下车载人员佩戴安全带检测精度不高的问题,提出一种基于改进的YOLOv5s车载人员佩戴安全带的检测方法;该检测方法将YOLOv5s作为基础网络,在此基础上进行改进;为改善深度模型对特征信息的提取能力,采用RFB模块增大网络的感受野,并利用RFB模块多分支结构获得混合的感受野;加入ECA注意力通道模块,使得整个网络更加专注特征信息的提取;将原YOLOv5s的损失函数替换为EIOU,进一步提高网络对安全带的检测精度;经过实验结果表明,改进后网络与原YOLOv5s网络相比,其平均精度均值mAP提高了2.2%,查准率提升了5.1%;改进后的网络具有良好的提升效果,表明了该方法的有效性。

The detection of seatbelt wearing behavior of vehicle-borne personnel plays an important role in ensuring human life safety.Aiming at the low detection accuracy of seatbelt worn by vehicle occupants in complex environments,an improved detection method based on YOLOv5s is proposed.The detection method takes YOLOv5s as the basic network and improves on it.In order to improve the ability of the depth model to extract feature information,the receptive field of the network is expanded by using the receptive field RFB module,and the hybrid receptive field is obtained by using the multi-branch structure of the RFB module.Adding the efficient channel attention(ECA)modules to make the entire network more focused on extracting the feature information.The loss function of the original YOLOv5s is replaced by the EIOU to further improve the detection accuracy of the safety belt.The experimental results show that compared with the original YOLOv5s network,the mean average precision(mAP)of the improved network is increased by 2.2%,and the precision by 5.1%.The improved network has a good enhancement effect,which shows the effectiveness of the method.