基于改进YOLOv5的铝型材表面缺陷检测方法

doi:10.16088/j.issn.1001-6600.2023050801

摘要/Abstract

摘要： 针对铝型材表面缺陷不同类别尺寸差别较大,导致检测效果较差的情况,本文提出一种基于改进YOLOv5的铝型材表面缺陷检测算法。首先,在网络中嵌入CA(coordinate attention)注意力机制模块,使网络更好地抑制图像中无效样本的干扰,更多聚焦于有用信息;其次,在原有检测层上增加一个小目标检测层,获取和传递更为丰富且更具判别性的小目标特征,以解决对小目标缺陷检测精度低的问题,提高整体检测精度;最后,引入SIoU损失函数,用边界框回归之间的向量角度来重新定义损失函数,在有效减少总自由度损失的同时提高推理精度。将改进算法应用到天池铝型材数据集中进行验证,实验结果表明:该模型能有效识别铝型材表面不同种类的缺陷,较原YOLOv5算法mAP提高11.4个百分点,检测速度达到66.4 frame/s,能够满足目前铝型材工厂生产现场缺陷检测要求。

关键词: 缺陷检测, YOLOv5, 注意力机制, SIoU, 多尺度融合

Abstract: In view of the large difference of different sizes of aluminum surface defects leading to poor detection effect, an aluminum surface defect detection algorithm based on improved YOLOv5 is proposed. Firstly, CA attention mechanism module is embedded in the network to make the network better suppress the interference of invalid samples in the image and focus more on useful information. Secondly, a small target detection layer is added to the original detection layer to acquire and transmit more abundant and discriminant small target features, so as to solve the problem of low detection accuracy of small target defects and improve the overall detection accuracy. Finally, the SIoU loss function is introduced and the vector angle between boundary box regression is introduced to redefine the loss function, which effectively reduces the total freedom of loss and improves the reasoning precision. The improved algorithm is applied to Tianchi aluminum data set for verification. The experimental results show that the model can effectively identify different types of defects on aluminum profiles, which is 11.4% higher than the original YOLOv5 algorithm mAP, and the detection speed is up to 66.4 frame/s. The proposed method can meet the requirements of on-site defect detection in aluminum profile factories.

Key words: defect detection, YOLOv5, attention mechanism, SIoU, multiscale fusion

中图分类号: TP391.41

席凌飞, 伊力哈木·亚尔买买提, 刘雅洁. 基于改进YOLOv5的铝型材表面缺陷检测方法[J]. 广西师范大学学报（自然科学版）, 2024, 42(1): 111-119.

XI Lingfei, Yilihamu Yaermaimaiti, LIU Yajie. Surface Defect Detection Method for Aluminum Profile Based on Improved YOLOv5[J]. Journal of Guangxi Normal University(Natural Science Edition), 2024, 42(1): 111-119.

参考文献

[1] 杨林. 冷态热轧钢板表面缺陷检测系统技术研究[D].重庆:重庆大学,2015.
[2] 陈妍.冷轧带钢表面缺陷智能检测技术的发展[J].鞍钢技术,1998(9):27-32.
[3] 刘娜, 强锡富, 高若云. 图像处理在非接触测量中的应用[J]. 电测与仪表, 1998(11): 44-45.
[4] 罗志勇,刘栋玉,江涛,等.新型冷轧带钢表面缺陷在线检测系统[J].华中理工大学学报,1996, 24(1):75-78.
[5] LIU B, WU S J, ZOU S F. Automatic detection technology of surface defects on plastic products based on machine vision[C]// 2010 International Conference on Mechanic Automation and Control Engineering.Piscataway,NJ:IEEE Press, 2010: 2213-2216. DOI: 10.1109/MACE.2010.5536470.
[6] DENG S E,CAI W W,XU Q Y,et al. Defect detection of bearing surfaces based on machine vision technique[C]// 2010 International Conference on Computer Application and System Modeling (ICCASM 2010): Volume 4. Piscataway,NJ: IEEE Press,2010: 548-554. DOI: 10.1109/ICCASM.2010.5620311.
[7] 贾方庆. 基于机器视觉的带钢表面缺陷检测系统研究[D]. 重庆: 重庆大学, 2007.DOI: 10.7666/d.y1141210.
[8] 谭绍华. 基于机器视觉的带钢表面缺陷检测系统研究[D]. 武汉: 华中科技大学, 2012.DOI: 10.7666/d.D228012.
[9] 刘孟轲,吴洋,王逊.基于卷积神经网络的轨道表面缺陷检测技术实现[J].现代计算机(专业版),2017(29):65-69, 77.DOI: 10.3969/j.issn.1007-1423.2017.29.016.
[10] 刘雄祥. 基于卷积神经网络的铁轨表面缺陷识别研究[D]. 绵阳: 西南科技大学, 2018.
[11] DALAL N, TRIGGS B. Histograms of oriented gradients for human detection[C]// 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition(CVPR’05). Los Alamitos, CA: IEEE Computer Society,2005:886-893.DOI: 10.1109/CVPR.2005.177.
[12] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004,60(2):91-110.DOI: 10.1023/B:VISI.0000029664.99615.94.
[13] CHERKASSKY V. The nature of statistical learning theory[J]. IEEE Transactions on Neural Networks, 1997, 8(6):1564.DOI: 10.1109/TNN.1997.641482.
[14] SUVDAA B, AHN J, KO J. Steel surface defects detection and classification using SIFT and voting strategy[J]. International Journal of Software Engineering and Its Applications, 2012, 6(2): 161-166.
[15] CORTES C, VAPNIK V. Support-vector networks[J]. Machine Learning, 1995,20(3):273-297. DOI: 10.1007/
BF00994018.
[16] 缪徐超,张旭,徐洋,等.基于支持向量机的金属表面缺陷检测[J].智能计算机与应用,2021,11(2):207-209, 封三.
[17] MEDINA R, GAYUBO F, GONZÁLEZ L M, et al. Surface defects detection on rolled steel strips by Gabor filters[C]// Proceedings of the Third International Conference on Computer Vision Theory and Applications (VISIGRAPP 2008): Volume 1. Setúbal: SciTePress, 2008: 479-485. DOI: 10.5220/0001088504790485.
[18] 汤勃, 孔建益, 王兴东, 等. 基于图像处理的钢板表面缺陷支持向量机识别[J]. 中国机械工程, 2011, 22(12): 1402-1405.
[19] 曹阳. 基于机器视觉的铝型材表面缺陷检测研究[D].西安:陕西科技大学,2021.DOI: 10.27290/d.cnki.gxbqc.2021.000217.
[20] 程婧怡,段先华,朱伟.改进 YOLOv3 的金属表面缺陷检测研究[J].计算机工程与应用,2021,57(19):252-258. DOI: 10.3778/j.issn.1002-8331.2104-0324.
[21] 吴越,杨延竹,苏雪龙,等.基于Faster R-CNN 的钢板表面缺陷检测方法[J].东华大学学报(自然科学版),2021,47(3):84-89. DOI: 10.19886/j.cnki.dhdz.2020.0159.
[22] 李庆党,李铁林.基于改进YOLOv3算法的钢板缺陷检测[J].电子测量技术,2021,44(2):104-108.DOI: 10.19651/j.cnki.emt.2005366.
[23] REDMON J,DIVVALA S, GIRSHICK R, et al.You only look once: unified, real-time object detection[C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society,2016: 779-788. DOI: 10.1109/CVPR.2016.91.
[24] REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]// 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Los Alamitos, CA: IEEE Computer Society, 2017: 6517-6525. DOI: 10.1109/CVPR.2017.690.
[25] REDMON J, FARHADI A.YOLOv3: an incremental improvement[EB/OL]. (2018-04-08)[2023-05-08].https://arxiv.org/abs/1804.02767. DOI: 10.48550/arXiv.1804.02767.
[26] BOCHKOVSKIY A, WANG C Y, LIAO H Y M.YOLOv4: optimal speed and accuracy of object detection[EB/OL].(2020-04-23)[2023-05-08].https://arxiv.org/abs/2004.10934. DOI: 10.48550/arXiv.2004.10934 .
[27] 高志国, 袁牧, 童旸, 等. 基于PCBAM-YOLOv5网络的变电站表计缺陷检测[J]. 电子设计工程, 2023, 31(4): 77-81. DOI: 10.14022/j.issn1674-6236.2023.04.016.
[28] HOU Q B, ZHOU D Q, FENG J S. Coordinate attention for efficient mobile network design[C]// 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2021: 13708-13717. DOI: 10.1109/CVPR46437.2021.01350.
[29] GEVORGYAN Z.SIoU loss:more powerful learning for bounding box regression[EB/OL].(2022-05-25)[2023-05-08].https://arxiv.org/abs/2205.12740. DOI: 10.48550/arXiv.2205.12740.

相关文章 15

[1]	王天雨, 袁嘉伟, 齐芮, 李洋. 多类型知识增强的微博立场检测模型[J]. 广西师范大学学报（自然科学版）, 2024, 42(1): 79-90.
[2]	肖宇庭, 吕晓琪, 谷宇, 刘传强. 基于拆分残差网络的糖尿病视网膜病变分类[J]. 广西师范大学学报（自然科学版）, 2024, 42(1): 91-101.
[3]	赵伟, 田帅, 张强, 王耀申, 王思博, 宋江. 基于改进YOLOv5的平贝母检测模型[J]. 广西师范大学学报（自然科学版）, 2023, 41(6): 22-32.
[4]	宋冠武, 陈知明, 李建军. 基于ResNet-50的级联注意力遥感图像分类[J]. 广西师范大学学报（自然科学版）, 2023, 41(6): 80-91.
[5]	郭嘉梁, 靳婷. 基于语义增强的多模态情感分析[J]. 广西师范大学学报（自然科学版）, 2023, 41(5): 14-25.
[6]	吴正清, 曹晖, 刘宝锴. 基于注意力卷积神经网络的中文虚假评论检测[J]. 广西师范大学学报（自然科学版）, 2023, 41(5): 26-36.
[7]	唐侯清, 辛斌斌, 朱虹谕, 乙加伟, 张冬冬, 武新章, 双丰. 基于多尺度注意力倒残差网络的轴承故障诊断[J]. 广西师范大学学报（自然科学版）, 2023, 41(4): 109-122.
[8]	黄叶祺, 王明伟, 闫瑞, 雷涛. 基于改进的YOLOv5金刚石线表面质量检测[J]. 广西师范大学学报（自然科学版）, 2023, 41(4): 123-134.
[9]	邓希桢, 蒋明, 岑明灿, 罗玉玲. 基于熵图像静态分析技术的勒索软件分类研究[J]. 广西师范大学学报（自然科学版）, 2023, 41(3): 91-104.
[10]	魏明军, 周太宇, 纪占林, 张鑫楠. 基于YOLOv3的公共场所口罩佩戴检测方法[J]. 广西师范大学学报（自然科学版）, 2023, 41(1): 76-86.
[11]	王利娥, 王艺汇, 李先贤. POI推荐中的多源数据融合和隐私保护方法[J]. 广西师范大学学报（自然科学版）, 2023, 41(1): 87-101.
[12]	王宇航, 张灿龙, 李志欣, 王智文. 体现用户意图和风格的图像描述生成[J]. 广西师范大学学报（自然科学版）, 2022, 40(4): 91-103.
[13]	李正光, 陈恒, 林鸿飞. 基于双向语言模型的社交媒体药物不良反应识别[J]. 广西师范大学学报（自然科学版）, 2022, 40(3): 40-48.
[14]	万黎明, 张小乾, 刘知贵, 宋林, 周莹, 李理. 基于高效通道注意力的UNet肺结节CT图像分割[J]. 广西师范大学学报（自然科学版）, 2022, 40(3): 66-75.
[15]	张萍, 徐巧枝. 基于多感受野与分组混合注意力机制的肺结节分割研究[J]. 广西师范大学学报（自然科学版）, 2022, 40(3): 76-87.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed