铸坯表面缺陷自动评级中图像预处理的研究
|
摘要
钢铁质量检测一直是钢铁生产的一个重要环节,对钢铁连铸坯的生产效率和生产质量都有着重大的意义。钢铁连铸坯表面缺陷检测作为一种重要的钢铁质量检测技术,一直是钢铁生产领域关注的热点。近些年钢铁连铸坯表面缺陷评级技术中应用最为广泛的技术冷酸蚀表面缺陷检测法在准确性上有很好的效果,但由于该技术仍采用人工来进行,在检测效率和精确性上已逐渐无法满足生产的需求。
本文针对连铸坯冷酸蚀表面缺陷在自动化评级上的应用需求,提出了基于数字图像处理技术的解决方案,实现连铸坯冷酸蚀图像的缺陷评级自动化。本文着重对该自动评级技术中的图像预处理过程进行了深入研究。
图像预处理旨在通过有针对性的处理,实现图像在各方面统一化,消除影响图像特征提取的不利因素,为后续特征提取的工作提供必要的支持。对于图像预处理过程中,重点对光照补偿、图像分割进行了研究。针对图片中偏光现象严重的情况,本文通过对直方图均衡化法、光照探测补偿法的算法及效果的研究比较,提出了基于以上两种方法的改进的光照补偿算法,有效克服了偏光现象对连铸坯图像的影响,保证了图像分割的有效性。针对图像预处理中的图像分割问题,本文对不同的缺陷分别提出了研究算法,根据中心偏析、中心疏松的离散性特点,提出了全局阈值、局部阈值结合基于邻域的杂散点分割法的算法;针对内裂、特裂、角裂缺陷的表现特征和评级方法,提出了全局阈值结合区域生长法、基于边界跟踪的图像分割法,并在效率上进行比较分析。
Steel quality inspection , which has always been an important process in the production of iron and steel , is of great significance to the efficiency and quality of Continuously Cast steel billet production. Steel billet surface defect detection , as an important techniques of Steel quality inspection,has been the focus in this field .In recent years the most widely used technique of Steel billet surface defect detection is Cold Acid Etching ,has proved to be efficient, however ,this technique , which is a manual job,can not meet the needs of the efficiency and accuracy.
According to the requirements of automatic rating of the Steel billet surface defect by Cold Acid Etching,this paper propose a solution based on the techniques of digital image processing , achieving the purpose of the accurate automatic rating . This paper focus on the process of the pre-processing in this technique and carry out an in-depth study of it.
Image Pre-processing process is designed , with efficient operations, to make the images normalized , to eliminate the adverse effects against the Feature Extraction, to provide the necessary support to the Follow-up Feature Extraction process . In the image pre-processing process , this article mainly focus on the illumination compensation and the image segmentation . Against the serious problems caused by Polarized light,this article provide two Algorithms ,Histogram equalization and the illumination detection and compensation , and take a study and comparison . Basing on these two Algorithms , this paper provides an improved , successfully overcome the impact to the continuous casting steel billets caused by the phenomenon of polarized light. In the image segmentation process , this paper provide different algorithms for each defects. According to the characteristics of center segregation and center porosity , the paper provide the overall threshold and Local threshold combined with algorithm of Neighborhood-based spurious points segmentation; According to the characteristics of outlooks and rating rules of internal crack. corner crack. Special crack , the paper provides overall threshold Algorithm combined with edge-detection based segmentation Algorithm and Algorithm based on seeded region growing theory ,and do some analysis and comparison between them in efficiency .
本文针对连铸坯冷酸蚀表面缺陷在自动化评级上的应用需求,提出了基于数字图像处理技术的解决方案,实现连铸坯冷酸蚀图像的缺陷评级自动化。本文着重对该自动评级技术中的图像预处理过程进行了深入研究。
图像预处理旨在通过有针对性的处理,实现图像在各方面统一化,消除影响图像特征提取的不利因素,为后续特征提取的工作提供必要的支持。对于图像预处理过程中,重点对光照补偿、图像分割进行了研究。针对图片中偏光现象严重的情况,本文通过对直方图均衡化法、光照探测补偿法的算法及效果的研究比较,提出了基于以上两种方法的改进的光照补偿算法,有效克服了偏光现象对连铸坯图像的影响,保证了图像分割的有效性。针对图像预处理中的图像分割问题,本文对不同的缺陷分别提出了研究算法,根据中心偏析、中心疏松的离散性特点,提出了全局阈值、局部阈值结合基于邻域的杂散点分割法的算法;针对内裂、特裂、角裂缺陷的表现特征和评级方法,提出了全局阈值结合区域生长法、基于边界跟踪的图像分割法,并在效率上进行比较分析。
Steel quality inspection , which has always been an important process in the production of iron and steel , is of great significance to the efficiency and quality of Continuously Cast steel billet production. Steel billet surface defect detection , as an important techniques of Steel quality inspection,has been the focus in this field .In recent years the most widely used technique of Steel billet surface defect detection is Cold Acid Etching ,has proved to be efficient, however ,this technique , which is a manual job,can not meet the needs of the efficiency and accuracy.
According to the requirements of automatic rating of the Steel billet surface defect by Cold Acid Etching,this paper propose a solution based on the techniques of digital image processing , achieving the purpose of the accurate automatic rating . This paper focus on the process of the pre-processing in this technique and carry out an in-depth study of it.
Image Pre-processing process is designed , with efficient operations, to make the images normalized , to eliminate the adverse effects against the Feature Extraction, to provide the necessary support to the Follow-up Feature Extraction process . In the image pre-processing process , this article mainly focus on the illumination compensation and the image segmentation . Against the serious problems caused by Polarized light,this article provide two Algorithms ,Histogram equalization and the illumination detection and compensation , and take a study and comparison . Basing on these two Algorithms , this paper provides an improved , successfully overcome the impact to the continuous casting steel billets caused by the phenomenon of polarized light. In the image segmentation process , this paper provide different algorithms for each defects. According to the characteristics of center segregation and center porosity , the paper provide the overall threshold and Local threshold combined with algorithm of Neighborhood-based spurious points segmentation; According to the characteristics of outlooks and rating rules of internal crack. corner crack. Special crack , the paper provides overall threshold Algorithm combined with edge-detection based segmentation Algorithm and Algorithm based on seeded region growing theory ,and do some analysis and comparison between them in efficiency .
引文
[1]张韶轩.当前钢铁检测技术发展及应用探析.学问.科教探索,1998,24(20): 18-18.
[2]谢艳峰.俄罗斯新利佩茨克钢铁公司热带钢自动表面检测技术.轧钢.2006,2:14-14
[3]出岛胜郎,胡风.钢铁厂在线质量检测技术的现状.国外计量,1993,3(2):10-13
[4]彭一江.钢铁分析检测与自动化.冶金自动化.2004,3(3) :10-14
[5]刘金升,白彦峰.钢铁分析检测与自动化.经济技术协作信息.2007,947(4):73~73
[6]刘胜国.连铸板坯快速硫印技术简介.重钢技术.1992,5(5): 37-40
[7]陈立新,周学东.硫印照片自动判级系统的设计与实现.Metallurgical Industry Automation .2004, 2(5):18-21
[8]吴华芳.连铸钢坯硫印检验及现场应用.检验经验,1999.15(4) :29-31
[9] Rafael C Gonzalez, Winz Paul. Digital Image Processing. Addison.Wesley Publishing Company.America. 1977:218-223
[10]崔屹.数字图像处理技术与应用.电子工业出版社.第一版. 1997年:234-237
[11]许庆太.连铸坯凝固组织和缺陷的冷酸蚀法宏观检验.炼钢.2003年,19(5):27-31
[12]史宸兴.连铸钢坯质量.北京冶金工业出版社.第一版1980年:45-51
[13] Wang Ying-hui, Ning Xiao-juan,Yang Chun-xia,and et al.A method of illumination compensation for human face image based on quotient image.In: Information Sciences,2008:2705–2721
[14] Javier Ruiz-del-Solar, Julio Quinteros. Illumination compensation and normalization in eigenspace-based face recognition: A comparative study of different pre-processing approaches. In:Pattern Recognition Letters, 2008 : 1966–1979
[15]于殿泓.图像检测与处理技术.西安电子科技大学出版社.2006,第一版:442-446
[16]皋军.图象的灰度直方图均衡化的实现.盐城工业学报.2002,14(4):63-65
[17]姜小磊.图像增强中直方图均衡化合理性的一种解释.电气电子教学学报.2008年,30(1):60-62
[18]刘宏,李锦涛,苗军.多方法融合来解决人脸检测中的光照补偿.系统仿真学报.2001年,13(5):486-490
[19]任艳斐.直方图均衡化在图像处理中的应用.计算机与信息技术.2007年,3(4):37-38
[20] Abdullah-Al-Wadud M.A Spatially Controlled Histogram Equalization for Image Enhancement. in 23rd International Symposium on Computer and Information Sciences,2008: 526-531
[21] Tang LR, Zhang J, Qi B.An Improved Fuzzy Image Enhancement Algorithm. In: Fifth International Conference On Fuzzy Systems And Knowledge Discovery, 2008:186-189
[22] Chen DW, Qian Y.An Improved Algorithm of Adaptive Fuzzy Image Enhancement. In:International Symposium On Computer Science And Computational Technologu,2008:594-598
[23]郑庆,闵帆,陈雷霆基于复合变换的人脸光照补偿方案.计算机应用研究2008,25(2):507-508
[24]梁晓辉,游志胜.自适应的彩色图像光照补偿新方法.光电工程.2006,33(2):94-97
[25]程培英、沈会良.一种光照不均匀图像的恢复增强方法.计算机应用与软件。2008,25(10):52-55
[26]罗海霞,龙永红,谭雪利.同态滤波在光照补偿中的应用.湖南工业大学学报.2008,22(5):23-27
[27]谢仪.常用数字图像分割算法研究.电脑开发与应用. 2005,2l(6) :41-43
[28]罗晓辉.阈值化图像分割算法研究.科技信息. 2008,5(7):185-185
[29]石荣刚,李志远,江涛.图像分割的常用方法及其应用.现代电子技术. 2007, 12(251) : 111-114
[30]徐正光,奠东来.张剁欣.基于递归的二值图像连通域像素标记算法.计算机工程. 2006,32(24) :186-189
[31]侯立华.图像分割方法综述.科技创新导报.2008,3(22) :249-249
[32]曾磊.基于Otsu法自适应阈值的图像分割研究.信息技术.2008,3(10) :88-90
[33] Chao Wang.An Efficient Illumination Compensation based on Plane-fit for Face Recognition.In:2008 10th Intl. Conf. on Control, Automation, Robotics and Vision:7-10
[34]罗胜.一种边界-区域相结合的图像分割算法.光电工程. 2008,35(l2) :101-106
[35]张静,王宏刚,王涌天.一种边缘提取的图像分割方法.光学技术.2001,27(5):424-429
[36]卢晓菁,陈锻生.一种人脸图像光照补偿的新方法.小型微型计算机系统.2008,10(2):1864-1867
[37]王茜蓓,彭中,刘莉.一种基于自适应阈值的图像分割算法.北京理工大学学报.第23卷第4期2003年8月:521-524
[38] P.A. Arbelaez and L.D. Cohen,Path variation and image segmentation.In:Proc. EMMCVP'03, 2003:246–260.
[39]杨威,张田文,师海峰.一种用于二值图象分割的快速聚类算法.计算机研究与发展. 1998, 35(8):719-723
[40] Shi Jianbo, J Malik. Normalized cuts and image segmentation.Pattern Analysis and Machine Intelligence,2000:888-905
[41] Potocnik B,Zazula D.Automated analysis of a sequence of ovarian ultrasound images. Image vision and computing,2007,33(2):217–225
[42] Kaiping Wei, Tao Zhang, Xianjun Shen.An Improved Threshold Selection Algorithm Based on Particle Swarm Optimization for Image Segmentation.In Third International Conference on Natural Computation,2007:83-85
[43] Kostas Haris, Serafim N.Hybrid image segmentation using watersheds and fast region merging. IEEE Transactions on Image Processing,1998:1684-1699
[44] Sharon E, Brandt A, Basri R.Fast multiscale image segmentation. Computer Vision and Pattern Recognition, 2000:70-77
[45] G. Koepfler, C Lopez, and J M Morel.A multiscale algorithm for image segmentation variational method.SIAM Journal on Numerical Analysis, 2007:282–299
[46] Tabb M,Ahuja N.Multiscale image segmentation by integrated edge and region detection. IEEE Transactions on Image Processing,1997 :642-645
[2]谢艳峰.俄罗斯新利佩茨克钢铁公司热带钢自动表面检测技术.轧钢.2006,2:14-14
[3]出岛胜郎,胡风.钢铁厂在线质量检测技术的现状.国外计量,1993,3(2):10-13
[4]彭一江.钢铁分析检测与自动化.冶金自动化.2004,3(3) :10-14
[5]刘金升,白彦峰.钢铁分析检测与自动化.经济技术协作信息.2007,947(4):73~73
[6]刘胜国.连铸板坯快速硫印技术简介.重钢技术.1992,5(5): 37-40
[7]陈立新,周学东.硫印照片自动判级系统的设计与实现.Metallurgical Industry Automation .2004, 2(5):18-21
[8]吴华芳.连铸钢坯硫印检验及现场应用.检验经验,1999.15(4) :29-31
[9] Rafael C Gonzalez, Winz Paul. Digital Image Processing. Addison.Wesley Publishing Company.America. 1977:218-223
[10]崔屹.数字图像处理技术与应用.电子工业出版社.第一版. 1997年:234-237
[11]许庆太.连铸坯凝固组织和缺陷的冷酸蚀法宏观检验.炼钢.2003年,19(5):27-31
[12]史宸兴.连铸钢坯质量.北京冶金工业出版社.第一版1980年:45-51
[13] Wang Ying-hui, Ning Xiao-juan,Yang Chun-xia,and et al.A method of illumination compensation for human face image based on quotient image.In: Information Sciences,2008:2705–2721
[14] Javier Ruiz-del-Solar, Julio Quinteros. Illumination compensation and normalization in eigenspace-based face recognition: A comparative study of different pre-processing approaches. In:Pattern Recognition Letters, 2008 : 1966–1979
[15]于殿泓.图像检测与处理技术.西安电子科技大学出版社.2006,第一版:442-446
[16]皋军.图象的灰度直方图均衡化的实现.盐城工业学报.2002,14(4):63-65
[17]姜小磊.图像增强中直方图均衡化合理性的一种解释.电气电子教学学报.2008年,30(1):60-62
[18]刘宏,李锦涛,苗军.多方法融合来解决人脸检测中的光照补偿.系统仿真学报.2001年,13(5):486-490
[19]任艳斐.直方图均衡化在图像处理中的应用.计算机与信息技术.2007年,3(4):37-38
[20] Abdullah-Al-Wadud M.A Spatially Controlled Histogram Equalization for Image Enhancement. in 23rd International Symposium on Computer and Information Sciences,2008: 526-531
[21] Tang LR, Zhang J, Qi B.An Improved Fuzzy Image Enhancement Algorithm. In: Fifth International Conference On Fuzzy Systems And Knowledge Discovery, 2008:186-189
[22] Chen DW, Qian Y.An Improved Algorithm of Adaptive Fuzzy Image Enhancement. In:International Symposium On Computer Science And Computational Technologu,2008:594-598
[23]郑庆,闵帆,陈雷霆基于复合变换的人脸光照补偿方案.计算机应用研究2008,25(2):507-508
[24]梁晓辉,游志胜.自适应的彩色图像光照补偿新方法.光电工程.2006,33(2):94-97
[25]程培英、沈会良.一种光照不均匀图像的恢复增强方法.计算机应用与软件。2008,25(10):52-55
[26]罗海霞,龙永红,谭雪利.同态滤波在光照补偿中的应用.湖南工业大学学报.2008,22(5):23-27
[27]谢仪.常用数字图像分割算法研究.电脑开发与应用. 2005,2l(6) :41-43
[28]罗晓辉.阈值化图像分割算法研究.科技信息. 2008,5(7):185-185
[29]石荣刚,李志远,江涛.图像分割的常用方法及其应用.现代电子技术. 2007, 12(251) : 111-114
[30]徐正光,奠东来.张剁欣.基于递归的二值图像连通域像素标记算法.计算机工程. 2006,32(24) :186-189
[31]侯立华.图像分割方法综述.科技创新导报.2008,3(22) :249-249
[32]曾磊.基于Otsu法自适应阈值的图像分割研究.信息技术.2008,3(10) :88-90
[33] Chao Wang.An Efficient Illumination Compensation based on Plane-fit for Face Recognition.In:2008 10th Intl. Conf. on Control, Automation, Robotics and Vision:7-10
[34]罗胜.一种边界-区域相结合的图像分割算法.光电工程. 2008,35(l2) :101-106
[35]张静,王宏刚,王涌天.一种边缘提取的图像分割方法.光学技术.2001,27(5):424-429
[36]卢晓菁,陈锻生.一种人脸图像光照补偿的新方法.小型微型计算机系统.2008,10(2):1864-1867
[37]王茜蓓,彭中,刘莉.一种基于自适应阈值的图像分割算法.北京理工大学学报.第23卷第4期2003年8月:521-524
[38] P.A. Arbelaez and L.D. Cohen,Path variation and image segmentation.In:Proc. EMMCVP'03, 2003:246–260.
[39]杨威,张田文,师海峰.一种用于二值图象分割的快速聚类算法.计算机研究与发展. 1998, 35(8):719-723
[40] Shi Jianbo, J Malik. Normalized cuts and image segmentation.Pattern Analysis and Machine Intelligence,2000:888-905
[41] Potocnik B,Zazula D.Automated analysis of a sequence of ovarian ultrasound images. Image vision and computing,2007,33(2):217–225
[42] Kaiping Wei, Tao Zhang, Xianjun Shen.An Improved Threshold Selection Algorithm Based on Particle Swarm Optimization for Image Segmentation.In Third International Conference on Natural Computation,2007:83-85
[43] Kostas Haris, Serafim N.Hybrid image segmentation using watersheds and fast region merging. IEEE Transactions on Image Processing,1998:1684-1699
[44] Sharon E, Brandt A, Basri R.Fast multiscale image segmentation. Computer Vision and Pattern Recognition, 2000:70-77
[45] G. Koepfler, C Lopez, and J M Morel.A multiscale algorithm for image segmentation variational method.SIAM Journal on Numerical Analysis, 2007:282–299
[46] Tabb M,Ahuja N.Multiscale image segmentation by integrated edge and region detection. IEEE Transactions on Image Processing,1997 :642-645