The Research on Lightweight Traffic Sign Recognition Algorithm Based on Improved YOLOv5 Model

Tiande Liu, Changlei Dongye
{"title":"The Research on Lightweight Traffic Sign Recognition Algorithm Based on Improved YOLOv5 Model","authors":"Tiande Liu, Changlei Dongye","doi":"10.1109/CCAI57533.2023.10201317","DOIUrl":null,"url":null,"abstract":"Traffic sign detection is an important research direction in object detection, which has been widely used in intelligent transportation system, driving assistance, automatic driving and other fields. In practical applications, traffic sign detection algorithms are required to complete detection and recognition tasks quickly and accurately, which requires the algorithm model to be lightweight to meet the deployment conditions. Aiming at the existing traffic sign detection problems, a lightweight traffic sign detection network based on YOLOv5s model was constructed, which improved the detection performance of the network model on the premise of guaranteeing the computing speed. In order to ensure lightweight, YOLOv5s model was selected. Firstly, Dense CSP Module (DCM) was designed to enhance the effect of feature fusion. At the same time, the feature pyramid is improved, and reduced the number of parameters in the model. Experimental results show that compared with the original algorithm, the detection efficiency of the proposed algorithm is improved by 5.28%, and the experimental results on multiple data sets show obvious improvement effect. This is a lightweight model that works well in the area of traffic sign detection.","PeriodicalId":285760,"journal":{"name":"2023 IEEE 3rd International Conference on Computer Communication and Artificial Intelligence (CCAI)","volume":"180 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 IEEE 3rd International Conference on Computer Communication and Artificial Intelligence (CCAI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CCAI57533.2023.10201317","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Traffic sign detection is an important research direction in object detection, which has been widely used in intelligent transportation system, driving assistance, automatic driving and other fields. In practical applications, traffic sign detection algorithms are required to complete detection and recognition tasks quickly and accurately, which requires the algorithm model to be lightweight to meet the deployment conditions. Aiming at the existing traffic sign detection problems, a lightweight traffic sign detection network based on YOLOv5s model was constructed, which improved the detection performance of the network model on the premise of guaranteeing the computing speed. In order to ensure lightweight, YOLOv5s model was selected. Firstly, Dense CSP Module (DCM) was designed to enhance the effect of feature fusion. At the same time, the feature pyramid is improved, and reduced the number of parameters in the model. Experimental results show that compared with the original algorithm, the detection efficiency of the proposed algorithm is improved by 5.28%, and the experimental results on multiple data sets show obvious improvement effect. This is a lightweight model that works well in the area of traffic sign detection.
基于改进YOLOv5模型的交通标志轻量化识别算法研究
交通标志检测是物体检测中的一个重要研究方向,已广泛应用于智能交通系统、驾驶辅助、自动驾驶等领域。在实际应用中,交通标志检测算法需要快速准确地完成检测和识别任务,这就要求算法模型轻量化以满足部署条件。针对现有的交通标志检测问题,构建了基于YOLOv5s模型的轻型交通标志检测网络,在保证计算速度的前提下,提高了网络模型的检测性能。为保证轻量化,选用YOLOv5s型号。首先,设计密集CSP模块(Dense CSP Module, DCM)增强特征融合效果;同时,对特征金字塔进行了改进,减少了模型中的参数个数。实验结果表明,与原算法相比,本文算法的检测效率提高了5.28%,在多数据集上的实验结果显示出明显的改进效果。这是一个轻量级的模型,在交通标志检测领域工作得很好。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信