基于多尺度特征融合的车辆及行人目标检测算法  

Vehicle and Pedestrian Target Detection Algorithm Based on Multi-Scale Feature Fusion

作　　者：李相衡 方虹苏 杨娅琳 杨炜[2] Li Xiangheng;Fang Hongsu;Yang Yalin;Yang Wei(Henan Polytechnic University,Jiaozuo 454000;Chang’an University,Xi’an 710064)

机构地区：[1]河南理工大学,焦作454000 [2]长安大学,西安710061

出　　处：《汽车工程师》2024年第8期1-7,共7页Automotive Engineer

基　　金：中央高校基金项目(300102229112)。

摘　　要：针对道路交通环境复杂多样,车辆及行人检测易出现错检及漏检的问题,提出一种基于多尺度特征融合的车辆及行人目标检测算法YOLOv8-RC。首先,在基础网络YOLOv8的结构中引入RCS-OSA模块代替原有模块,对所提取的特征信息进行增强及融合,并引入轻量级上采样算子内容感知特征重组(CARAFE)代替原上采样算子,提高网络对全局多尺度信息的融合能力。其次,通过公开数据集及网络收集的方式构建了由6000张车辆及行人目标图片构成的检测数据集,并采用准确率、召回率、平均精度均值mAP50及mAP50-95对算法检测效果进行定量评价,相比于YOLOv8-N,YOLOv8-RC的精确率提升1.7百分点,召回率提升1.2百分点,mAP50提升0.9百分点,mAP50-95提升0.5百分点,证明了算法的有效性。In response to the complex and diverse nature of the road traffic environment,where vehicle and pedestrian detection is prone to false and missed detections,this paper proposes a vehicle and pedestrian target detection algorithm YOLOv8-RC based on multi-scale feature fusion.Initially,the RCS-OSA module is introduced within the structure of the base network YOLOv8 to replace the original module,thereby enhancing and integrating the extracted feature information.Additionally,a lightweight Context-Aware Adaptive Feature Reorganization(CARAFE)is employed to replace the original upsampling operator,enhancing the network’s capability for global multi-scale information fusion.Subsequently,a detection dataset consisting of 6000 images of vehicle and pedestrian targets is constructed through public datasets and network collection.The algorithm’s detection performance is quantitatively evaluated using accuracy,recall rate,mean Average Precision at a 50%intersection over union threshold(mAP50),and mAP50-95.Compared to YOLOv8-N,YOLOv8-RC demonstrates an improvement of 1.7 percentage in accuracy,1.2 percentage in recall rate,0.9 percentage in mAP50,and 0.5 percentage in mAP50-95,thus validating the algorithm’s effectiveness.

关 键 词：深度学习 目标检测 YOLOv8 行人检测 

分 类 号：U491[交通运输工程—交通运输规划与管理]