摘要
随着计算机视觉技术的发展和计算机性能的不断提高,由带钢图像在线检测表面质量已成为国内外学者研究的热点课题。论文设计了一套带钢表面自动检测装置,完成图像采集、预处理、系统噪声特性分析、滤波及缺陷检测、分割等技术的开发,研究了由图像阴影恢复三维形状的算法以及图像网络传输等几个关键技术问题。论文主要研究内容和成果如下:
(1)根据带钢表面自动检测系统的技术要求,提出系统的总体设计方案,及软、硬件设计方案;
(2)提出利用VFW进行视频采集卡二次开发的方案和具体实施方法,实现了带钢表面图像的实时采集;
(3)通过对带钢表面图像的研究,提出噪声图像的获取方法,找出系统噪声特性,并针对此特性设计了一种智能滤波器;
(4)根据对带钢表面缺陷特性的研究,提出一种基于小区域的缺陷检测算法,并在此基础上提出了缺陷分割算法;
(5)通过对从图像阴影恢复形状理论的研究,提出了一种基于人工神经网络的三维图像恢复算法;
(6)提出网络通讯的软、硬件方案,实现局域网中图像的实时传输。
经实验验证,以上技术和方法都取得了较好的效果。
With the development of the computer-vision and computer capacity, automatic strip surface inspection based on image has become favorable question for study. This peper proposes automatic strip inspection system, and finishes the development acquiring image, pre-processing image, analyzing and dispelling noisy, inspecting and cuting apart defect. The paper researches the technique of shape recover from shading and network. The main content and achivement of this paper are listed as follows:
1.According to inspection system's technical requirement, its overall design plan is put forward. It includes hardware and software design scheme.
2.Advancing the method for acquiring image with VFW.
3.Putting forward the method for the acquirement the noisy image, finding the noisy's character and advancing the intelligencal method for dispelling noisy.
4.With the research of the defect's character, putting forward the method of defect recognition and cut apart.
5.Researching the shape from shading and putting forward the shape from shading neural algorithm.
6.Putting forward the design scheme about local area network's hardware and software.
Test result shows the technologies are effective.
引文
[1] 吴平川,路同浚,王炎.带钢表面自动检测系统研究现状与展望.钢铁,2001,6:70-75
[2] 陈研.冷轧带钢表面缺陷智能检测技术的发展.鞍钢技术,1998,9:25-30
[3] M.P.Simonis.频闪法在带钢表面检查中的应用.世界钢铁,1994.4:65-68
[4] 让.路易斯.萨隆,吉米.于贝尔,安德烈.克兰著,卢文斌译.用安装在火焰切割设备前德涡流探测器检验热连铸板坯的表面质量.无损探伤,1994,10(6):1-7
[5] Yngve Strom. Automatic surface inspection of continuously cast billets. Iron and Steel Engineer. 1992, 69(5):29-33
[6] 吴平川,路同浚,王炎.钢板表面缺陷的无损检测技术与应用.无损检测,2000,7:312-315
[7] 童华强.冷轧带钢表面缺陷及粗糙度检测系统的应用.钢铁研究,1997.3:55-60
[8] 罗志勇,刘栋玉,江涛.新型冷轧带钢表面缺陷在线检测系统.华中理工大学学报,1996.1:75-78
[9] Marco Vascotto. High Speed Surface Defect Identification on Steel. MPT, 1996(4): 70-73
[10]John C Badger. Automated Surface Inspection System. Iron and Steel Engineen, 1996(3): 48-51
[11]Suresh BR, Fundakowski RA, Levitt TS. A real-time automated visual inspection system for hot steel slabs. IEEE Trans Pattern Anal Machine Intell, 1983, PAMI-5 (6):563-572
[12]Thomas F Porter, Robert A Sytvester, Theodore Wbouyoueas. Automatic strip surface defect detection system. Iron and Steel Engineer, 1988, 55(12): 17-20
[13]David G Park, Martir P Levoi, VAN Haneghem A I. Practical of On-line Hot Strip Inspection System at Hoogovens. Iron and Steel Engineer, 1995, 72(7):40-43
[14]王斌,罗志勇,刘栋玉,吕新民.带钢表面重皮缺陷在线图像识别算法研究.华中理工大学学报,1996,8:35-37
[15]吕新民,罗志勇,刘栋玉,王斌.用面阵CCD实时检测运动带钢表面孔洞.光电工
程,1997,2:42-45
[16]宣梅灿.宝钢1420mm冷轧生产线新技术应用介绍.轧钢,1997(2):19-23
[17]宣梅灿.宝钢1550mm冷轧生产线新技术介绍.宝钢技术,1999(4):36-42
[18]徐科,徐金梧,鹿守理.冷轧带钢表面自动监测系统的研究.钢铁,2000,10:63-66
[19]Jeffrey Richter.Windows核心编程.北京:机械工业出版社,2000
[20]李明柱,王西平,赵晓红.Windows NT中“Intranet组建和管理技术.北京:北京航空航天大学出版社,1999:25
[21]任观就,张永林.实时视频图像捕获的实现方法.计算机工程,2002.8:268-270
[22]王博,王毅.VideoCap视频捕捉软件的设计与开发.计算机工程与应用,2002.22:255-256
[23]Visual C++6.0多媒体开发实例.北京:机械工业出版社,2000
[24]吕凤军.数字图像处理编程入门.北京:清华大学出版社,1999
[25]何斌,马天予,王运坚.Visual C++数字图像处理.北京:人民邮电出版社,2001
[26]杨雪荣,康戈文,任文伟.数字视频图像噪声特性分析.实用测试技术,2002,2:29-30
[27]王爱民.图像分割研究综述.测试技术,2000,5:1-6
[28]周长发.精通Visual C++图像编程.北京:电子工业出版社,2000
[29]贾云得。机器视觉.北京:科学出版社,2000
[30]张静,王宏剐,王涌天.一种边缘提取得图像分割方法.光学技术,2001,9:424-426
[31]Yen J C, Chang F J, Chang S. A New Criterion for Automatic Multilevel Thresholding. IEEE Trans on Image Processing, 1995, 4(3) :370-377
[32]Papamarkos N, Gatos B. A New Approach for Multilevel Threshold Selection. CVGIP:Graphic Models and Image Processing, 1994, 56:357-370
[33]Huang L K, Wang M J J. Image Thresholding by Minimizing the Measure of Fuzziness. Pattern Recognition, 1995, 28 (1) :41-51
[34]Corneloup G, Moysan J,Maynin I E. BSCAN Image Segmentation by Thresholding Using Cooccurence Matrix Analysis. Pattern Recognition, 1996, 29(2):281-296
[35]Li L, Gong J, Chen W. Grey-Level Image Thresholding Based on Fisher Linear Projection of Two-Dimensional Histogram. Pattern Recognition, 1997, 30 (5) : 743-750
[36]盛锦华.利用单目图像得阴影恢复物体形状.合肥工业大学学报,1991,3:11—20
[37]Horn B.K.P, Books M J. The variational approach to skape from shading. CVGIP, 1986, 33:174-208
[38]章疏晋.图象理解与计算机视觉.北京:清华大学出版社,2000
[39]高文,陈熙霖.计算机视觉.北京:清华大学出版社,1999
[40]Cho S Y, Chow T W S. Shape Recover from Shading by a New Neural-Based Reflectance Model. IEEE Trans. Neural Networks, 1999,10(6):1536-1541
[41]Cho S Y, Chow T W S. Shape and Surface Measurement Technology by an Improved Shape from Shading Neural Algorithm. IEEE TRANSACTIONS ON INDUSTRONICS, 2000,47(1): 225-230
[42]Cho S Y, Chow T W S. A Neural-Learning-Based Reflectance Model for 3-D Shape Reconstruction. IEEE TRANSACTIONS ON INDUSTRONICS, 2000,47(6)1346-1350
[43]G.Q. Wei and G. Hirzinger, " Learning shape from shading by a multilayer network," IEEE Trans.Neural Networks, 1996,17(7):985-995
[44]Cho S Y, Chow T W S. Neural Computation Approach for Developing a 3-D Shape Reconstruction Model. IEEE Trans. Neural Networks, 2001,12(5):1204-1214
[45]阎平凡,张长水.人工神经网络与模拟进化计算.清华大学出版社.2000
[46]张国宝,陈维南,高明顺。照明方向及反射特性参数的鲁棒估计.计算机研究与发展,1998,3:274-278
[47]廖熠,赵荣椿.从明暗恢复形状(SFS)的几类典型算法分析与评价.中国图象图象学报, 2001,10:953-961
[48]Qinfen Zheng and Rama Chellappa. Estimation of Illuminant Direction, Albedo, and Shap from Shading. IEEE Trans. Pattern analysis machine intelligence 1991,13(7):680-702
[49]Cho S Y, Chow T W S. Training multilayer neural networks using fast global learning algorithm-least-squares and penalized optimization methods. Neurocomputing, 1999,25:115-131
[50]莫宁,俞宁.局域网视频数据传输的应用研究.工业控制计算机,2000,3:14-16
[51]刘建萍.谈网络OSI参考模型与TCP/1P参考模型.福建电脑,2002.6:25-26
[52]罗惠琼,杨亚玲,杨国渝.通信与网络技术及编程.成都:电子科技大学,2000
[53]汪翔,袁辉.Visual C++实践与提高.北京:中国铁道出版社,2001