基于全卷积神经网络的单幅隧道图像裂纹提取算法

计算机工程与科学 ›› 2022, Vol. 44 ›› Issue (05): 845-854.

基于全卷积神经网络的单幅隧道图像裂纹提取算法

仇静博1,2,燕雪峰1,汪俊1,郭延文3，魏明强1,2

(1.南京航空航天大学计算机科学与技术学院，江苏南京 211106；
2.南京航空航天大学模式分析与机器智能工业和信息化部重点实验室，江苏南京 211106；
3.南京大学计算机软件新技术国家重点实验室，江苏南京 210023)

收稿日期:2021-11-17 修回日期:2022-01-04 接受日期:2022-05-25 出版日期:2022-05-25 发布日期:2022-05-24

Crack extraction from single tunnel image based on fully convolutional neural network

QIU Jing-bo1,2，YAN Xue-feng1，WANG Jun1，GUO Yan-wen3，WEI Ming-qiang1,2

（1.College of Computer Science & Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 211106;
2.MIIT Key Laboratory of Pattern Analysis and Machine Intelligence,
Nanjing University of Aeronautics and Astronautics,Nanjing 211106;
3.State Key Laboratory for Novel Software Technology,Nanjing University,Nanjing 210023,China）

Received:2021-11-17 Revised:2022-01-04 Accepted:2022-05-25 Online:2022-05-25 Published:2022-05-24

摘要/Abstract

摘要： 提出一种基于全卷积神经网络的单幅隧道图像裂纹提取算法，能够有效避免复杂背景下的伪裂纹噪声点干扰，实现对隧道裂纹的精确分割。首先，构建深度残差网络模型提取裂纹特征；其次，使用改进的全卷积神经网络中的反卷积操作恢复裂纹特征图的尺寸和裂纹细节；为了提升裂缝提取的精细程度，提出一个细节修复模块来保持裂缝的完整性与边缘细节；最后，公开一个裂纹数据集NUAACrack-2000，包含2 000幅隧道裂纹图像与精准标注标签。实验表明，提出的算法在避免噪声点干扰方面优于传统图像分割算法；在保留提取裂纹的整体性以及边缘细节处理方面优于基于机器学习的主流裂纹提取算法。

关键词: 裂纹提取, 图像分割, NUAACrack-2000, 全卷积神经网络

Abstract: A single tunnel image crack extraction algorithm based on fully convolutional neural network is proposed, which can effectively avoid the interference of pseudo-crack noise points in complex backgrounds and achieve accurate segmentation of tunnel cracks. First, a deep residual network model is constructed to extract crack features. Second, in order to recover the size of the crack feature map as well as the crack details, the feature map size is recovered using a deconvolution operation in an improved fully convolutional neural network. In order to improve the fineness of crack extraction, a detail repair module is proposed, which can maintain the integrity and edge details of cracks. Finally, a crack dataset NUAACrack-2000 is published, which contains 2000 tunnel crack images and accurately labeled labels. Extensive experiments show that the proposed algorithm outperforms traditional image segmentation algorithms in avoiding noise point interference. It is superior to the mainstream crack extraction algorithms based on machine learning in preserving the integrity of the extracted cracks and processing edge details.

Key words: crack extraction, image segmentation, NUAACrack-2000, fully convolutional neural network

仇静博, 燕雪峰, 汪俊, 郭延文, 魏明强, . 基于全卷积神经网络的单幅隧道图像裂纹提取算法[J]. 计算机工程与科学, 2022, 44(05): 845-854.

QIU Jing-bo, YAN Xue-feng, WANG Jun, GUO Yan-wen, WEI Ming-qiang, . Crack extraction from single tunnel image based on fully convolutional neural network[J]. Computer Engineering & Science, 2022, 44(05): 845-854.

参考文献［7］

［1］	Ma Chang-xia,Zhao Chun-xia,Hu Yong,et al.Pavement cracks detection based on NSCT and morphology［J］.Journal of Computer Aided Design & Computer Graphics,2009,21(12):1761-1767.(in Chinese)
［2］	Li Gang,He Yu-yao.A novel image detection and classification for pavement crack under non-uniform illumination［J］.Acta Photonica Sinica,2010,39(8):1405-1408.(in Chinese)
［3］	Zhao Hui-jie, Ge Wen-qian,Li Xu-dong.Detection of crack defect based on minimum error and pulse coupled neural networks［J］.Chinese Journal of Scientific Instrument,2012,33(3):637-642.(in Chinese)
［4］	Xu Wei,Tang Zhen-min,Lü Jian-yong.Pavement crack detection based on image saliency［J］.Journal of Image and Graphics,2013,18(1):69-77.(in Chinese)
［5］	Xu Huan,Li Zhen-bi,Jiang Yuan-yuan,et al.Pavement crack detection based on OpenCV and improved Canny operator［J］.Computer Engineering and Design,2014,35(12):4254-4258.(in Chinese)
［6］	Otsu N.A threshold selection method from gray-level histograms［J］.IEEE Transactions on Systems,Man,and Cybernetics,1979,9(1):62-66.
［7］	Yang Li,Wu Gui-fang,Sun Xiu-ming.Algorithm of pavement crack fusion based on multi-scale curvelet transforms［J］.Computer Engineering and Applications,2016,52(6):182-185.(in Chinese)
［8］	Long J,Shelhamer E,Darrell T.Fully convolutional networks for semantic segmentation［C］∥Proc of the IEEE Conference on Computer Vision and Pattern Recognition,2015:3431-3440.
［9］	Badrinarayanan V,Kendall A,Cipolla R.SegNet:A deep convolutional encoder-decoder architecture for image segmentation［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(12):2481-2495.
［10］	He K,Gkioxari G,Dollár P,et al.Mask R-CNN［C］∥Proc of the IEEE International Conference on Computer Vision,2017:2961-2969.
［11］	Chen L C,Papandreou G,Kokkinos I,et al.DeepLab:Semantic image segmentation with deep convolutional nets,atrous convolution,and fully connected CRFs［J］.IEEE Transactions on Pattern Analysis & Machine Intelligence,2018,40(4):834-848.
［12］	Zhao H,Shi J,Qi X,et al.Pyramid scene parsing network［C］ ∥Proc of the IEEE Conference on Computer Vision and Pattern Recognition,2017:2881-2890.
［13］	Silva W R L,Lucena D S.Concrete cracks detection based on deep learning image classification［C］∥Proc of the 18th International Conference on Experimental Mechanics，2018:489.1-489.6.
［14］	Kim B,Cho S.Automated vision-based detection of cracks on concrete surfaces using a deep learning technique［J］.Sensors,2018,18(10):3452.
［15］	Zou Q,Zhang Z,Li Q,et al.DeepCrack:Learning hierarchical convolutional features for crack detection［J］.IEEE Transactions on Image Processing,2018,28(3):1498-1512.
［16］	Mei Q,Gül M,Azim M R.Densely connected deep neural network considering connectivity of pixels for automatic crack detection［J］.Automation in Construction,2020,110:103018.
［17］	Ren Y,Huang J,Hong Z,et al.Image-based concrete crack detection in tunnels using deep fully convolutional networks［J］.Construction and Building Materials,2020,234:117367.
［18］	Simonyan K, Zisserman A.Very deep convolutional networks for large-scale image recognition［J］.arXiv:1409.1556,2014.
［19］	Szegedy C,Liu W,Jia Y,et al.Going deeper with convolutions［C］∥Proc of the IEEE Conference on Computer Vision and Pattern Recognition,2015:1-9.
［20］	He K,Zhang X,Ren S,et al.Deep residual learning for image recognition［C］∥Proc of the IEEE Conference on Computer Vision and Pattern Recognition,2016:770-778.
［21］	Hu J,Shen L,Sun G.Squeeze-and-excitation networks［C］∥Proc of the IEEE Conference on Computer Vision and Pattern Recognition,2018:7132-7141.
［22］	Zhang Z, Sabuncu M. Generalized cross entropy loss for training deep neural networks with noisy labels［C］∥Proc of Advances in Neural Information Processing Systems,2018:8778-8788.
［23］	Qin X,Zhang Z,Huang C,et al.U2-Net:Going deeper with nested U-structure for salient object detection［J］.Pattern Recognition,2020,106:107404.
［24］	Zhou H,Xie X,Lai J H,et al.Interactive two-stream decoder for accurate and fast saliency detection［C］∥Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2020:9141-9150.
［25］	Wang L,Lee C Y,Tu Z,et al.Training deeper convolutional networks with deep supervision［J］.arXiv:1505.02496,2015.
	附中文参考文献：
［1］	马常霞,赵春霞,胡勇,等.结合NSCT和图像形态学的路面裂缝检测［J］.计算机辅助设计与图形学学报,2009,21(12):1761-1767.
［2］	李刚,贺昱曜.不均匀光照的路面裂缝检测和分类新方法［J］.光子学报,2010,39(8):1405-1408.
［3］	赵慧洁,葛文谦,李旭东.最小误差准则与脉冲耦合神经网络的裂缝检测［J］.仪器仪表学报,2012,33(3):637-642.
［4］	徐威,唐振民,吕建勇.基于图像显著性的路面裂缝检测［J］.中国图象图形学报,2013,18(1):69-77.
［5］	徐欢,李振璧,姜媛媛,等.基于OpenCV和改进Canny算子的路面裂缝检测［J］.计算机工程与设计,2014,35(12):4254-4258.
［7］	杨莉,吴贵芳,孙秀明.基于多尺度曲波变换的路面裂缝融合算法［J］.计算机工程与应用,2016,52(6):182-185.

[1]	吴从中, 董浩, 方静. 基于注意力机制的自适应滤波遥感图像分割网络[J]. 计算机工程与科学, 2022, 44(11): 2010-2018.
[2]	曾增峰, 环宇翔, 邹卓, 郑立荣. 多尺度深度监督的高度近视萎缩病变分割方法[J]. 计算机工程与科学, 2021, 43(07): 1264-1272.
[3]	蒋芸, 高静, 王发林. SAG-Net：用于联合视盘和视杯分割的新型跳过注意力指导网络[J]. 计算机工程与科学, 2021, 43(07): 1273-1282.
[4]	张俊鹏, 刘辉, 李清荣. 基于FCN-LSTM的工业烟尘图像分割[J]. 计算机工程与科学, 2021, 43(05): 906-916.
[5]	陈爱莲, 丁正龙, 詹曙. 基于对抗学习和多尺度特征融合的前列腺MR图像分割[J]. 计算机工程与科学, 2021, 43(04): 697-703.
[6]	朱倩倩, 车文刚, 苗晗. 数字化藏文古籍中多样性字体的实现方法研究[J]. 计算机工程与科学, 2020, 42(11): 2073-2079.
[7]	赵仁和，王军锋. 自适应尺度的局部强度聚类图像分割模型[J]. 计算机工程与科学, 2020, 42(06): 1043-1048.
[8]	徐秋平. 基于人机交互式图割的目标快速提取[J]. 计算机工程与科学, 2020, 42(02): 299-306.
[9]	王卫民1,符首夫1,顾榕蓉1,王东升1,何林容2,关文斌3. 基于卷积神经网络的虫情图像分割和计数方法[J]. 计算机工程与科学, 2020, 42(01): 110-116.
[10]	任佳1,2，张胜男1，董超2，赵敏钧1. 基于改进模糊C均值的海面目标图像分割算法[J]. 计算机工程与科学, 2019, 41(05): 858-864.
[11]	吴尚智1，佘志用2，张霞1，赵慧琴3. 利用变精度粗糙熵的图像分割算法[J]. 计算机工程与科学, 2018, 40(10): 1837-1843.
[12]	易三莉，张桂芳，贺建峰，李思洁. 基于最大类间方差的最大熵图像分割[J]. 计算机工程与科学, 2018, 40(10): 1874-1881.
[13]	刘汉强，张青. 区间模糊谱聚类图像分割方法[J]. 计算机工程与科学, 2018, 40(09): 1611-1616.
[14]	谭志国1，2,欧建平1,张军1,沈先耿2. 多特征组合的深度图像分割算法[J]. 计算机工程与科学, 2018, 40(08): 1429-1434.
[15]	郭伟，李红达，邢宇哲. 基于SLIC的自适应多主体图像分割算法[J]. 计算机工程与科学, 2018, 40(08): 1459-1467.