基于改进YOLOv5s的监控视频车流量检测  

Surveillance video vehicle flowrate detection based on improved YOLOv5s

作　　者：熊显名[1] 刘雨鑫 黎恒 XIONG Xianming;LIU Yuxin;LI Heng(School of Optoelectronic Engineering,Guilin University of Electronic Technology,Guilin 541004,China;Guangxi Beitou IT Innovation Technology Investment Group Co.,Ltd.,Nanning 530200,China)

机构地区：[1]桂林电子科技大学光电工程学院,广西桂林541004 [2]广西北投信创科技投资集团有限公司,南宁530200

出　　处：《桂林电子科技大学学报》2024年第4期416-426,共11页Journal of Guilin University of Electronic Technology

基　　金：国家自然科学基金(61965005);广西自然科学基金(2019GXNSFDA185010)。

摘　　要：针对监控视频中车流量统计准确率不高,多车型车辆检测跟踪精度低、鲁棒性差等问题,提出了一种基于改进YOLOv5s的目标检测算法与DeepSort跟踪算法相结合的车流量检测方法。该方法对YOLOv5s的特征提取网络进行重构,以强化对目标重要特征的提取,提高检测器的检测精度。首先在Backbone主干网络中引入Swin Transformer模块,替换原算法中部分C3模块,增强模型全局化建模的能力,更好地捕捉上下文特征信息,扩大模型的感受野。然后对比不同的注意力机制,选择在Neck网络中接入GAM注意力,增强信息在通道与空间维度之间的跨维交互作用,减少信息损失,以强化网络的性能。最后对DeepSort跟踪算法的特征提取网络部分进行优化,并在车辆重识别数据集上重新进行训练,使其更适合对车辆的跟踪。实验结果表明,改进后的YOLOv5s与原算法相比提高了2.04%,在结合DeepSort算法后,在白天、傍晚、夜间等不同光照条件下车流量统计准确率分别达到97.5%、95.7%、85.1%。Aiming at the problems of low detection and tracking accuracy,poor robustness and low statistical accuracy of traffic flow of multi-type vehicles in video,a vehicle flow detection method based on improved YOLOv5s object detection algorithm and Deep-Sort tracking algorithm is proposed.This method reconstructs the feature extraction network of YOLOv5s to strengthen the extrac-tion of important features of the target and improve the detection accuracy of the detector.Firstly,the Swin Transformer module is introduced in the Backbone network to replace some C3 modules in the original algorithm,so as to enhance the global modeling ability of the model,better capture contextual feature information,and expand the receptive field of the model.Then,by comparing different attention mechanisms,GAM attention is selected in Neck network to enhance the cross-dimensional interaction of informa-tion between channels and spatial dimensions,reduce information loss and enhance network performance.Finally,the feature extrac-tion network part of the DeepSort tracking algorithm is optimized and re-trained on the vehicle re-recognition dataset to make it more suitable for vehicle tracking.Experimental results show that the improved YOLOv5s improves by 2.04%points compared with the original algorithm.Combined with DeepSort algorithm,the statistical accuracy of vehicle traffic in different lighting conditions such as day,evening and night can reach 97.5%,95.7%and 85.1%,respectively.

关 键 词：YOLOv5s目标检测 DeepSort目标跟踪 Swin Transformer GAM 车流量检测 

分 类 号：TN911.74[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]