[1]文权锐,高玮军.基于改进YOLOv5的道路病害检测算法研究[J].计算机技术与发展,2025,(02):191-198.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0298]
　WEN Quan-rui,GAO Wei-jun.Research on Road Disease Detection Algorithm Based on Improved YOLOv5[J].,2025,(02):191-198.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0298]
点击复制

基于改进YOLOv5的道路病害检测算法研究()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2025年02期
页码:
191-198
栏目:
新型计算应用系统
出版日期:
2025-02-10

文章信息/Info

Title:
Research on Road Disease Detection Algorithm Based on Improved YOLOv5
文章编号:
1673-629X(2025)02-0191-08
作者:
文权锐; 高玮军
兰州理工大学 计算机与通信学院,甘肃 兰州 730050
Author(s):
WEN Quan-rui; GAO Wei-jun
School of Computer and Communication,Lanzhou University of Technology,Lanzhou 730050,China
关键词:
目标检测; YOLOv5; 轻量化; 注意力机制; 道路病害检测; 特征提取
Keywords:
target detection; YOLOv5; light weight; attention mechanism; road disease detection; feature extraction
分类号:
TP391
DOI:
10.20165/j.cnki.ISSN1673-629X.2024.0298
摘要:
受道路病害特征不明显、与背景对比度低等复杂因素的影响,基于传统数字图像处理技术的识别算法精度低。 而现有深度学习目标检测算法参数量大,不适合移动端实时检测。 为了更高效、准确地识别道路病害,提出一种基于 YOLOv5 的改进算法。 该算法致力于实现道路检测的实时性和精确度之间的平衡,通过引入轻量级 Ghost 模块用更少的参数生成更多的特征图,减少模型参数;将基于 CSP 结构中的 C3Ghost 模块重新设计成一种新型的 InceptionGhost 模块,实现从不同尺度进行特征提取,保障模型精度。 通过在骨干网络添加 CBAM 注意力机制进一步加强特征提取能力。 由于采集到的样本分布不均匀,使用 Focal-EIoU 损失函数改善样本数量不平衡的问题,提高模型的鲁棒性。 实验结果表明,改进后的算法较原有的 YOLOv5s 模型,mAP@ 0. 5 提高了 2. 5 百分点,同时参数量压缩了 36. 6% ,检测速度提高了 16. 9% ;并通过与目前先进算法和其他改进算法的对比表明,该算法在精度和参数量方面显示出明显的优越性。
Abstract:
The accuracy of recognition algorithm based on traditional digital image processing technology is too low due to the influence of road disease features and low background contrast. However,the existing deep learning object detection algorithm has a large number of parameters,which is not suitable for real-time detection on mobile terminals. In order to identify road diseases more efficiently and ac-curately,an improved algorithm based on YOLOv5 is proposed. The algorithm is committed to achieving the balance between real-time and accuracy of road detection. By introducing lightweight Ghost module to generate more feature maps with fewer parameters,the model parameters are reduced. The C3Ghost module based on CSP structure is redesigned into a new InceptionGhost module to realize feature extraction from different scales and ensure model accuracy. The feature extraction capability is further enhanced by adding CBAM attention mechanism to the backbone network. Due to the uneven distribution of samples collected,Focal-EIoU loss function was used to improve the problem of unbalanced sample quantity and improve the robustness of the model. Experimental results show that compared with the original YOLOv5s model,the improved algorithm has an increase of 2. 5 percentage points in mAP@ 0. 5,a 36. 6% reduction in parameter quantity,and a 16. 9% increase in detection speed. By comparing experiments with the current advanced algorithms and other improved algorithms,the proposed algorithm shows obvious superiority in terms of accuracy and parameter quantity.

相似文献/References:

[1]刘晓明 李毓蕙 高燕 郑华强.基于目标区域清晰显示的H.264编码策略[J].计算机技术与发展,2010,(06):29.
　LIU Xiao-ming,LI Yu-hui,GAO Yan,et al.A Coding Strategy of H.264 Based on High-definition Display of Target Region[J].,2010,(02):29.
[2]刘翔 吴谨 祝愿博 康晓晶.基于视频序列的目标检测与跟踪技术研究[J].计算机技术与发展,2009,(11):179.
　LIU Xiang,WU Jin,ZHU Yuan-bo,et al.A Study of Object Detecting and Tracking Based on Video Sequences[J].,2009,(02):179.
[3]曙光 张超 蔡则苏.基于改进的混合高斯模型的目标检测方法[J].计算机技术与发展,2012,(07):60.
　SHU Guang,ZHANG Chao,CAI Ze-su.Target Detection Method Based on Improved Gaussian Mixture Model[J].,2012,(02):60.
[4]刘洁,李目,周少武.一种混沌混合粒子群优化RBF神经网络算法[J].计算机技术与发展,2013,(08):181.
　LIU Jie[],LI Mu[],ZHOU Shao-wu[].An Algorithm of Chaotic Hybrid Particle Swarm Optimization Based on RBF Neural Network[J].,2013,(02):181.
[5]蒋翠清,孙富亮,吴艿芯. 基于相对欧氏距离的背景差值法视频目标检测[J].计算机技术与发展,2015,25(01):37.
　JIANG Cui-qing,SUN Fu-liang,WU Nai-xin. Video Object Detection of Background Subtraction Method Based on Relative Euclidean Distance[J].,2015,25(02):37.
[6]卢官明,衣美佳. 步态识别关键技术研究[J].计算机技术与发展,2015,25(07):100.
　LU Guan-ming,YI Mei-jia. Research on Critical Techniques in Gait Recognition[J].,2015,25(02):100.
[7]高翔,朱婷婷,刘洋. 多摄像头系统的目标检测与跟踪方法研究[J].计算机技术与发展,2015,25(07):221.
　GAO Xiang,ZHU Ting-ting,LIU Yang. Research of Target Detection and Tracking Method for Multi-camera System[J].,2015,25(02):221.
[8]章文洁[][],黄旻[],张桂峰[]. 滤光片多光谱成像中运动目标场景误配准修正[J].计算机技术与发展,2016,26(01):18.
　ZHANG Wen-jie[][],HUANG Min[],ZHANG Gui-feng[]. Misregistration Correction for Moving Object Scene in Filter-type Multispectral Imaging[J].,2016,26(02):18.
[9]施泽浩,赵启军.基于全卷积网络的目标检测算法[J].计算机技术与发展,2018,28(05):55.[doi:10.3969/j.issn.1673－629X.2018.05.013]
　SHI Ze-hao,ZHAO Qi-jun.Object Detection Algorithm Based on Fully Convolutional Neural Network[J].,2018,28(02):55.[doi:10.3969/j.issn.1673－629X.2018.05.013]
[10]张夏清,茅耀斌. 一种改进的ViBe背景提取算法[J].计算机技术与发展,2016,26(07):36.
　ZHANG Xia-qing,MAO Yao-bin. An Improved ViBe Background Generation Method[J].,2016,26(02):36.
[11]何维娟,江 涛,王林飞,等.舰船遥感图像数据集 DSTD 的构建研究[J].计算机技术与发展,2022,32(07):82.[doi:10. 3969 / j. issn. 1673-629X. 2022. 07. 014]
　HE Wei-juan,JIANG Tao,WANG Lin-fei,et al.Research on DSTD Construction of Ship Remote Sensing Image Dataset[J].,2022,32(02):82.[doi:10. 3969 / j. issn. 1673-629X. 2022. 07. 014]

更新日期/Last Update: 2025-02-10