基于改进YOLOv7的交通路口目标识别算法  

Target Recognition Algorithm of Traffic Intersection Based on Improved YOLOv7

作　　者：江晟 张仲义 汪宗洋 于晴 JIANG Sheng;ZHANG Zhongyi;WANG Zongyang;YU Qing(School of Physics,Changchun University of Science and Technology,Changchun 130022,China;Institute of Deep Perception Technology,Wuxi 214000,Jiangsu Province,China)

机构地区：[1]长春理工大学物理学院,长春130022 [2]江苏集萃深度感知技术研究所,江苏无锡214000

出　　处：《吉林大学学报（理学版）》2024年第3期665-673,共9页Journal of Jilin University:Science Edition

基　　金：吉林省科技发展计划重点研发项目(批准号:20210203214SF).

摘　　要：针对交通路口车辆目标检测算法存在精确度低、少检、漏检等问题,提出一种基于改进YOLOv7的交通路口目标识别算法.该算法首先利用前馈式卷积注意力机制CBAM从通道注意力和空间注意力两者提升网络对关键特征的注意力,提高网络的运行速率,优化网络的特征提取能力;其次采取空间层到深度层连接全维动态卷积组成一个新的学习模块,以此结构改进YOLOv7特征学习方式,提升特征表达能力;最后在实际采集的交通路口数据集上进行实验.实验结果表明,该方法在对应数据集上平均精度达到96.1%,训练耗时降低至16.71 h,因此针对交通路口小目标检测有明显的识别优势.Aiming at the problems of low accuracy,under-detection,and missed detection in the vehicle target detection algorithm at traffic intersections,we proposed a target recognition algorithm of traffic intersection based on improved YOLOv7.Firstly,the algorithm used the feed-forward convolutional attention mechanism CBAM to enhance the network’s attention to key features from both channel attention and spatial attention,improve the network’s running speed,and optimize the network’s feature extraction capabilities.Secondly,a new learning module was formed by connecting the spatial layer to depth layers to form a full-dimensional dynamic convolution,which improved the YOLOv7 feature learning method and enhanced the feature expression ability.Finally,the experiments were conducted on the actual collected traffic intersection dataset.The experimental results show that the proposed method achieves an average accuracy of 96.1%on the corresponding dataset,and the training time is reduced to 16.71 h.Therefore,it has obvious recognition advantages for small target detection at traffic intersections.

关 键 词：深度学习 目标检测 卷积神经网络 注意力机制 全维动态卷积 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]