基于<strong>K-means++</strong>聚类算法和<strong>SSIM</strong>指标的金属板材腐蚀区域识别

doi:10.11902/1005.4537.2023.112

中国腐蚀与防护学报

2024, Vol. 44

Issue (2): 437-444 CSTR: 32134.14.1005.4537.2023.112 DOI: 10.11902/1005.4537.2023.112

研究报告

本期目录 | 过刊浏览

基于K-means++聚类算法和SSIM指标的金属板材腐蚀区域识别

龙梦翔, 付桂翠, 万博(

), 张钟庆

北京航空航天大学可靠性与系统工程学院北京 100191

Corrosion Area Identification of Sheet Metal Based on K-means++ Clustering Algorithm and SSIM Index

LONG Mengxiang, FU Guicui, WAN Bo(

), ZHANG Zhongqing

School of Reliability and Systems Engineering, Beihang University, Beijing 100191, China

引用本文:

龙梦翔, 付桂翠, 万博, 张钟庆. 基于K-means++聚类算法和SSIM指标的金属板材腐蚀区域识别[J]. 中国腐蚀与防护学报, 2024, 44(2): 437-444.
Mengxiang LONG, Guicui FU, Bo WAN, Zhongqing ZHANG. Corrosion Area Identification of Sheet Metal Based on K-means++ Clustering Algorithm and SSIM Index[J]. Journal of Chinese Society for Corrosion and protection, 2024, 44(2): 437-444.

全文: PDF(9730 KB) HTML

摘要:

材料腐蚀后会在表面产生锈斑、裂纹和鼓泡等多种腐蚀特征现象，通过观察腐蚀特征现象可判断材料腐蚀程度。目前主要通过人工目测的方式对材料的腐蚀情况进行判断，但其存在结果无法量化、效率低下等不足。本文采用K-means++聚类算法对金属板材图像像素RGB值进行聚类，分离腐蚀区域和未腐蚀区域；采用图像结构相似性指标(SSIM)判断聚类各区域是否发生腐蚀。结果表明：将K-means++聚类中心数量k设定为5，可有效根据图像颜色分布划分出各聚类区域；相比峰值信噪比PSNR和均方误差MSE，结构相似性指标SSIM与图像是否发生腐蚀具有较强相关性，将SSIM指标阈值设定为0.95，可根据SSIM指标有效判断各聚类区域是否发生腐蚀；本文所用方法相比人为根据像素颜色划分腐蚀区域，具有更高的识别效率，且准确率不低于90%。本研究可用于金属板材环境试验后防腐性能自动化评价。

关键词 ：腐蚀评估, 图像处理, K-means++算法, 聚类, 图像相似度, 智能诊断, 腐蚀面积

Abstract：

Various corrosion characteristics such as rust spots and cracks will appear on the surface of metal plates after corrosion. Their corrosion degree can be determined by corrosion characteristics. At present, the corrosion degree of metal plates is mainly judged by manual visual inspection. But it has many non-ignorable shortcomings such as low consistency and low efficiency etc. In this paper, the RGB values of image pixels of corroded metal sheets were collected and then clustered by means of K-means++ clustering algorithm, afterwards the relevant corroded- and uncorroded-regions were separated. Whether corrosion occurred or not was judged in each cluster area by means of image structural similarity index SSIM. The results show that setting the number of clustering centers ‘k’ to 5 can effectively delineate each clustering area based on the image color distribution. Compared to peak signal-to-noise ratio and mean square error, the structural similarity index SSIM is strongly correlated with the occurrence of erosion. Setting the SSIM index threshold at 0.95 can effectively judge whether erosion occurred in each cluster area. Compared to manually dividing corrosion areas based on pixel color, our method had a higher identification efficiency and an accuracy of not less than 90%. This research can be applied to automate the evaluation of the corrosion degree of metal sheets after environmental testing.

Key words： corrosion evaluation image processing K-means++ algorithm clustering image similarity intelligent diagnosis corrosion area

收稿日期: 2023-04-14 32134.14.1005.4537.2023.112

ZTFLH:

TP391.41

通讯作者: 万博，E-mail:wanbo@buaa.edu.cn，研究方向为元器件可靠性

Corresponding author: WAN Bo, E-mail: wanbo@buaa.edu.cn

作者简介: 龙梦翔，男，2000年生，硕士生

	服务

	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	龙梦翔
	付桂翠
	万博
	张钟庆
44.8K

链接本文:

https://www.jcscp.org/CN/10.11902/1005.4537.2023.112 或 https://www.jcscp.org/CN/Y2024/V44/I2/437

图1 K-means聚类算法流程图

图2 K-means++聚类算法初始簇中心确定流程图

图3 金属板材腐蚀前后图像

图4 各聚类区域图像

图5 金属板材聚类区域1腐蚀前后图像对比

表1 聚类区域图片相似性指标计算结果

图6 金属板材腐蚀区域识别结果

图7 金属板材腐蚀区域图像识别结果与人工“数格子”识别结果对比

1	Cao X L. Investigation of atmospheric corrosivity with ACM technology[D]. Wuhan: Academy of Machinery Science & Technology, 2005
1	曹献龙. ACM技术研究大气环境腐蚀严酷性[D]. 武汉: 机械科学研究院, 2005
2	Xu D, Yang X J, Li Q, et al. Review on corrosion test methods and evaluation techniques for materials in atmospheric environment[J]. J. Chin. Soc. Corros. Prot., 2022, 42: 447
2	徐迪, 杨小佳, 李清等. 材料大气环境腐蚀试验方法与评价技术进展[J]. 中国腐蚀与防护学报, 2022, 42: 447
3	Uhlig H H. The cost of corrosion to the United States[J]. Corrosion, 1950, 6: 29 doi: 10.5006/0010-9312-6.1.29
4	Fei Y W, Peng B, Xiao Y P. Review on the development of digital processing technology for metal corrosion images[J]. Dig. Technol. Appl., 2010, (11): 45
4	费逸伟, 彭彬, 校云鹏. 金属腐蚀图像数字处理技术发展综述[J]. 数字技术与应用, 2010, (11): 45
5	Wang M H, Wang H, Liu R, et al. Research on image recognition for NiCrAlY Coating/N5 high-temperature alloy system based on deep learning method[J]. J. Chin. Soc. Corros. Prot., 2022, 42: 583
5	王明好, 王欢, 刘叡等. 基于深度学习方法的N5/NiCrAlY涂层图像识别的研究[J]. 中国腐蚀与防护学报, 2022, 42: 583 doi: 10.11902/1005.4537.2021.234
6	Tan C, Sun Y, Li G F, et al. Image segmentation technology based on genetic algorithm[A]. Proceedings of the 2019 3rd International Conference on Digital Signal Processing[C]. Jeju Island, Republic of Korea, 2019: 27
7	Lei F, Xiong J B, Zhang L, et al. Image enhancement algorithm in metal corrosion area[J]. CAAI Trans. Intell. Syst., 2019, 14: 385
7	雷芳, 熊建斌, 张磊等. 金属腐蚀区域图像增强算法研究[J]. 智能系统学报, 2019, 14: 385
8	Li Z H, Zhang L, Wang J M, et al. Research on image acquisition and preprocessing method for visual measurement of pipeline inner surface corrosion[A]. 2017 13th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)[C]. Yangzhou, China, 2017: 364
9	Zhu Z T, Mu Z T, Su W G, et al. Corrosion grade evaluation of aluminum alloy based on image processing technique[J]. J. Nanjing Univ. Aeronaut. Astronaut., 2010, 42: 383
9	朱做涛, 穆志韬, 苏维国等. 基于图像处理技术的铝合金腐蚀等级评定方法[J]. 南京航空航天大学学报, 2010, 42: 383
10	Xu J, Luan Y H. Application of digital image processing technology in corrosion science[J]. Total Corros. Control, 2015, 29(1): 43
10	徐健, 栾艳华. 数字图像处理技术在腐蚀科学中的应用研究[J]. 全面腐蚀控制, 2015, 29(1): 43
11	Nan C H. Research on corrosion mechanism and damage evaluation of ocean engineering materials based on image recognition[D]. Xi'an: Xi'an University of Technology, 2021
11	南翠红. 基于图像识别的海洋工程材料腐蚀机理及损伤评价研究[D]. 西安: 西安理工大学, 2021
12	Huang X B, Zhang X L, Zhang Y, et al. A method of identifying rust status of dampers based on image processing[J]. IEEE Trans. Instrum. Meas., 2020, 69: 5407 doi: 10.1109/TIM.19
13	Petricca L, Moss T, Figueroa G, et al. Corrosion detection using A. I: a comparison of standard computer vision techniques and deep learning model[A]. The Sixth International Conference on Computer Science, Engineering and Information Technology[C]. Viena, Austria, 2016
14	Zhou G Y, Zhang Z J, Yin W L, et al. Surface defect detection of CFRP materials based on infrared thermography and Attention U-Net algorithm[J]. Nondestruct. Test. Eval., doi: 10.1080/10589759.2023.2191954
15	Ibrahim A, Dalbah A, Abualsaud A, et al. Application of machine learning to evaluate insulator surface erosion[J]. IEEE Trans. Instrum. Meas., 2020, 69: 314 doi: 10.1109/TIM.19
16	Wang S S, Ma S J, Che K, et al. Application status of machine learning in field of natural environment corrosion assessment and prediction[J]. J. Chin. Soc. Corros. Prot., 2023, 43: 441
16	王莎莎, 马帅杰, 车琨等. 机器学习在自然环境腐蚀评估与预测领域的应用现状[J]. 中国腐蚀与防护学报, 2023, 43: 441 doi: 10.11902/1005.4537.2022.147
17	Rahman A, Wu Z Y, Kalfarisi R. Semantic deep learning integrated with RGB feature-based rule optimization for facility surface corrosion detection and evaluation[J]. J. Comput. Civil Eng., 2021, 35: 4021018 doi: 10.1061/(ASCE)CP.1943-5487.0000982
18	MacQueen J. Classification and analysis of multivariate observations[A]. 5th Berkeley Symposium on Mathematical Statistics and Probability[C]. Los Angeles, USA, 1967: 281
19	Liu Y. Laryngoscope image classification by an improved convolutional neural network based on entropy weighted K-means clustering[D]. Changchun: Changchun University of Technology, 2022
19	刘阳. 基于加权熵K-means聚类的改进卷积神经网络下喉镜图像分类研究[D]. 长春: 长春工业大学, 2022
20	Zhong J, Lin C, Sheng Z W, et al. Quantum K-means algorithm based on Hamming distance[J]. J. Comput. Appl., 2023, 43(8): 2493
20	钟静, 林晨, 盛志伟等. 基于汉明距离的量子K-Means算法[J]. 计算机应用, 2023, 43(8): 2493
21	Arthur D, Vassilvitskii S. K-means++: The advantages of careful seeding[A]. Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms[C]. New Orleans, USA, 2006
22	Thorndike R L. Who belongs in the family?[J]. Psychometrika, 1953, 18: 267 doi: 10.1007/BF02289263
23	Rousseeuw P J. Silhouettes: a graphical aid to the interpretation and validation of cluster analysis[J]. J. Comput. Appl. Math., 1987, 20: 53
24	Tibshirani R, Walther G, Hastie T. Estimating the number of clusters in a data set via the gap statistic[J]. J. R. Stat. Soc., 2001, 63B: 411
25	Sugar C A, James G M. Finding the number of clusters in a dataset: An information-theoretic approach[J]. J. Am. Stat. Assoc., 2003, 98: 750 doi: 10.1198/016214503000000666
26	Hirsch R. Exploring Colour Photography: A Complete Guide[M]. Berlin: Laurence King Publishing, 2005
27	Wang Z, Bovik A C. Mean squared error: Love it or leave it? A new look at signal fidelity measures[J]. IEEE Signal Process. Mag., 2009, 26: 98
28	Wang Z, Bovik A C, Sheikh H R, et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Trans. Image Process., 2004, 13: 600 doi: 10.1109/TIP.2003.819861

[1]	朱烨森, 蔡锟, 胡葆文, 夏云秋, 胡涛勇, 黄一. 海底管道CO₂ 腐蚀特性及预测模型研究进展[J]. 中国腐蚀与防护学报, 2023, 43(6): 1225-1236.
[2]	张新新,高志明,胡文彬,伍志鹏,韩连恒,卢丽花,修妍,夏大海. Q235钢在薄液膜下腐蚀行为与图像信息的相关性研究[J]. 中国腐蚀与防护学报, 2017, 37(5): 444-450.
[3]	刘涛, 艾军, 张丽芳, 张鹏飞, 杨朝辉, 徐春林. 基于图像处理技术的钢箱梁防腐涂层寿命预测实验研究[J]. 中国腐蚀与防护学报, 2013, 33(5): 407-412.
[4]	黄家怿邱于兵郭兴蓬. 采用聚类分析研究X70钢在库尔勒土壤中初期电化学噪声特征[J]. 中国腐蚀与防护学报, 2009, 29(6): 453-458.
[5]	唐其环 . 从灰色聚类结果看化学成分对钢大气腐蚀的影响[J]. 中国腐蚀与防护学报, 2003, 23(4): 239-242 .
[6]	张伦武; 汪学华; 封先河 . 大气腐蚀环境及环境因素模糊聚类分析[J]. 中国腐蚀与防护学报, 2003, 23(2): 112-115 .
[7]	钟庆东; 谢巧敏; 周国定 . 模糊聚类分析评价防锈油膜电化学不均匀性[J]. 中国腐蚀与防护学报, 2000, 20(6): 368-373 .
[8]	李长荣;屈祖玉;丁乙;李英义;杨德钧;朱日彰;顾建宁;李春光. 土壤腐蚀性因素的聚类分析[J]. 中国腐蚀与防护学报, 1994, 14(2): 138-142.

Viewed

Full text

Abstract

Cited

Shared

Discussed