基于计算机视觉木材表面缺陷检测方法研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
板材是木材应用需求量最大的品种,板材表面质量是评定板材质量的重要指标之一。随着木材加工业向机械化、自动化的大规模生产方向发展,人们对板材的加工质量,尤其是表面缺陷给予了越来越多的重视,因而表面缺陷检测技术变得越来越重要。
     本文基于机器视觉理论对木材表面缺陷进行了深入研究,结合数字图像处理技术和人工神经网络模式识别技术,研究了木材表面缺陷图像预处理、特征提取、模式识别问题,以Visual C++程序设计环境,开发了用于检测板材表面缺陷的定位和识别等图像处理算法。
     图像预处理是检测的第一步,对图像缺陷特征的正确提取是非常关键的。论文通过三种方式对图像的灰度直方图进行分析统计:(1)对每一像素作256级灰度直方图分析;(2)对4×4像素块作256级灰度直方图分析;(3)4×4像素块作16级灰度直方图分析。
     特征提取直接影响木材缺陷检测系统的识别率。论文首先从灰度直方图中根据是否有颜色突变来判断图片是否存在缺陷,缺陷图片在直方图中表现出双峰特征,通常次波峰即为缺陷部位,但这不是绝对的。若直方图曲线只有一个波峰,则可能是正常木材图片。经过实验统计,当次波峰值比主波峰的值大于1/10时,次波峰即是代表缺陷颜色。实现了缺陷检测的第一步,即把图像分为有缺陷和无缺陷两类。
     基于人工神经网络的模式识别具有对数据类型和分布函数没有限制、容忍度更高等优点,相适应于木材表面缺陷的复杂性,有很好的应用前景。论文以缺陷灰度均值、缺陷灰度方差和缺陷形状作为缺陷类型识别的特征量为输入,缺陷类型为输出,构建了系统的BP网络系统模型。论文以4种缺陷类型为输出,选用LMS对BP神经网络进行训练,对设计的神经网络系统进行了检测,实验结果表明系统的平均识别率为97%,证实了所设计系统的可行性和有效性。
Saw-timber is most needed in wood application, and the surface of saw-timber is one of the important factors to assess wood quality. With large-scaled mechanization and automation of wood processing, people begin to attach more and more importance to manufacturing quality, especially surface defect. Therefore, surface defect detection is becoming more and more important.
     Based on the theory of computer vision, a research on defect detection of the wood surface is made in paper. Image preprocess, feature extraction and pattem recognition of wood surface defect are also studied by means of digital image processing techniques and neural networks principles of pattern recognition technology. An arithmetic of image disposal was compiled for disfigurement detection to orientate and recognize in the environment of VC++.
     Image preprocess is the first step for detection, which is vital to the correct extraction of the defection feature. The grey-historam is analyzed through three methods in paper: (1) 256 grade grey-historam statistics for every pixel. (2) 256 grades grey-historam statistics for every 4×4 block pixels. (3) 16 grades grey-historarn statistics for every 4×4 block pixels.
     Extraction of characteristic quantity will directly affect the distinguishing ratio of wood disfigurement detection system. The paper firstly judged whether there existed defects in the pictures according to the color break in grey-historam. Defect pictures have double peaks in grey-historarn. Generally speaking, the subordinate peak is just the defect part. But it is not absolute. If there is only one peak in the grey-historan, it means that the picture is normal. Experiments show when the value of subordinate peak is 1/10 greater than host peak value, subordinate peak indicates defect color. It would realize the first step for wood detection. and the images are divided into two types: defect images and normal images.
     Because there is superior tolerance and no limit to data type and distributing function in pattern recognition of neural network, it has a good prospect in application so as to adapt to the complexity in wood surface defect. Using three character values including gray average value, gray variance and defect shape as input value, ten defect types as output. This paper posed BP neural network system model. The designed network system is tested by chosing four type of defects, and LMS arithmetic to train BP neural network. The result shows the average ratio of recognition is 97 percent. The system is viable and available.
引文
[1] 李坚,陆文达,刘一星等.木材科学[M].哈尔滨:东北林业大学出版社,1994.
    [2] 李坚.中国木材研究[M].哈尔滨:东北林业大学出版社
    [3] 徐明有.木材学[M].北京:中国林业出版社.2006
    [4] 张壁光.木材科学与技术研究进展[M].北京:中国环境科学出版社,2004
    [5] 朱玉杰,张泱等,进口木材检验[M].哈尔滨:东北林业大学出版社,2005
    [6] 刘一星,进出口木材检测技术[M].北京:化学工业出版社,2005
    [7] 沈隽,赵钟声,木材加工技术[M].北京:化学工业出版社,2005
    [8] 杨春梅,胡万义.木材缺陷检测理论及方法的发展[J].林业机械与木工设备,2004(3):8-10
    [9] 陈莲梅,我国未来家具行业的竞争与展望[J].林业机械与木工设备,2004(3):11-13
    [10] 石岭,王克奇.板材表面缺陷检测技术研究[J].林业机械与木工设备.2005(3):40-42
    [11] 王立海,杨学春,徐凯宏.木材缺陷无损检测技术研究现状[J].林业机械与木工设备,2002,27(3):35-38
    [12] Ryszard Szymant, Kent A. Mclonald. Defect detec-tion in lumber: State of the art [J]. Forest Productslournal, 1981 (11)
    [13] 杨学春,王立海.应力波技术在木材性质检测中的研究进展[J].森林工程,2002,18(2):11-12
    [14] 马玉真,陶立英.计算机视觉技术的应用[J].试验技术与试验机,2006,(1):64-65
    [15] 戚大伟,牟洪波.人工神经网络在木材缺陷检测中的应用[J].森林工程,.2006,22(1):21-23
    [16] 李红俊,韩冀皖.数字图像处理技术及其应用[J].计算机测量与控制,2002,10(9):620-622
    [17] 石争浩,冯亚宁.用于铸件表面缺陷机器视觉检测的快速图像处理技术[J].计算机测量与控制,2005,26(4):325-328
    [18] Ross, R. J. Pellerin, R. F. 1991 Nondestructive testing of assessing wood members in structures.. USDA, Forest Products Laboratory, FPL-GTR-70
    [19] 石争浩,冯亚宁.基于机器视觉的铸件表面缺陷人工神经网络检测方法[J].重型机械,2004(2):51-54
    [20] RobertJ. ross. James, Ward and Tenoelde. Identifying Bacterially Infected Oak By Stress Wave Nondestructiv Evaluation. Res. Pap. FPL-RP-512. Forest Products Laboratory, 1992
    [21] 韩彦芳,施鹏飞.基于数字图像处理的表面缺损检测技术[J].测控技术,2005,24(9):15-18
    [22] 何滢,曲兴华等.非金属工件电镀表面缺陷现场检测技术的研究[J].仪器仪表学报,2004,25(2):271~273
    [23] 王克奇.木质材料表而缺陷计算机视觉测量的识别方法[J].林业科学,1996.32(1):92-96.
    [24] 程万胜,冷显国等.一种带钢表面缺陷实时检测系统的研制及应用[J].冶金自动化2006(3):47-52
    [25] Sun Jose chang, James R Olson, Paul C. Wang. NMR imaging of internal feature in wood[J]. Forest Products Journal, 1989(6)
    [26] 刘松平,刘菲菲.先进的可视化成像无损检测技术及其应用[J].航空制造技术,2006,(01):69-73
    [27] 沈建中,无损检测的几个热点问题和技术[J].无损检测.2005,27(1):24-26
    [28] 胡英成,顾继友.木材及人造板物理力学性能无损检测技术研究的发展与展望[J].2002,15(4):39-45
    [29] Niemz P, etal. Vergleichende untersuchungen zur bestimmung des ddynamischen E-moduls mittels schall-laufzeitund resonanzfrequenzmessung. Holzforschung und Holzverwertung. 1997, 49(5): 91-93
    [30] 张建合,郭广平.国内外飞速发展的热像无损检测技术[J].无损探伤.2005,29(1):1-4
    [31] Toshimitsu ISH Ⅱ. Yoshizo OKAMOTO. Investination of Qualification, Certification and Reference Testing Methods on Nondestructive Inspection by Means of Infrared Thermongraphv[J].非破坏检查, 1999, 48(10): 642-652
    [32] 薛书文,洪伟铭.脉冲加热红外热成像无损检测技术回顾[J].湛江师范学院学报,2004,25(6):22-25
    [33] Mime J M. A transient thermal method measuring the depths of subsurface flaws in metals[M]. British Ioumal of NDT, 1988, 30(5): 77-80
    [34] Costs J, Larrieq P. Surface Characaerizalion with 3-Demen-sional Roughness Parameter[C]. Proceedings of the iSth International wood Machining Seminar, 2001, 473-476.
    [35] Faust T D. Real Time Measurement of Veneer Surface Rough-ness by Image Analysis[J]. Forest Prod. J., 1987, 37(6): 34-40
    [36] Vav ilov V, Grinza to E Bison P G etal Surface tmnsient tempemture inversion for hidden corrosion characterization theory and application[J]. Int J Heat Mass Transfe, 1996, 39(2): 335-371
    [37] 邓志刚,邓荣斌.板材缺陷的微波检测[J].林业机械与木工设备,1998,26(6):30-31
    [38] 薛书文.利用红外热成像技术检测金属内部近表缺陷[J].红外与毫米波学报,1998,17(6):424-428.
    [39] 林文树,杨慧敏.超声波与应力波在木材内部缺陷检测中的对比研究[J].林业科技,2005,30(2):39-41
    [40] 吕立仁,陈兆斌.超户声波检测木材质量的可行性探讨[J].铁道技术监督.1999(6):33-34
    [41] 韩玉杰,朱国玺,田中千秋..木材表面缺陷的激光在线检测技术[J].木材工业.2002,16(3):30-32
    [42] Robert J. Ross, Roy F. Pellerin, Norbert Volny, etc. In-spection of Timber Brides Using Stress Wave Timing Nonde-structive Evaluation Tools. Pap. FPL G TR-114. Forest Products Laboratory.. 1999
    [43] 赵元松,李大鹏。基于有限元方法的钢板焊缝缺陷红外无损检测[J]。机电工程。2005,22(12):39-42
    [44] 郑世才。我国射线检测技术近年的发展[J]。无损检测。2004,26(4):163-167
    [45] 李巍,李庆丰,丁淑杰。无损检测自动化中的数字图像自动分析技术[J]。黑龙江电子技术。1999(1):40-42
    [46] 许文台.X射线检测木材实验报告[J].森林工业通讯,1979(2):34-36
    [47] 青郁,刘自强,崔广兴.木材内腐X光电视检测研究[J].东北林业大学学报,1983(4)
    [48] 林文树.应力波与超声波在木材内部缺陷检测中的对比研究[D].东北林业大学,2005
    [49] 杨学春.基于应力波原木内部腐朽检测理论及试验的研究[D].东北林业大学,2004
    [50] D. Ouis. Detection of decay in logs through measuring the dampening of bending vibrations bymeans of a room acoustic technique[J]. Wood Science and Technology. 2000, 34: 221-236.
    [51] 王业琴,王克奇,白雪冰,石岭.计算机视觉木材表面色差检测的研究[J].林业科技, 2005, 30 (2): 36-38
    [52] 赵茂程,郑加强,林小静等.基于BP神经网络的树形识别系统研究[J].林业科学,2004,40(1):99-102.
    [53] 谢永华.基于分形理论木材表面缺陷识别的研究[D].东北林业大学,2006
    [54] 林文树.MATLAB在木材内部缺陷检测图像处理中的应用[J].林业科技,2005,(6):49-51.
    [55] Wang Keqi et al., A Machine Vision System for Inspecting Wood Surface Defects by Using Neural Network, Journal of Northeast Forestry University, Vol. 7, No. 2, 1996
    [56] 邹定海.用于在线测的视觉检测系统[J].仪器仪表学报.1993,16(4):337-340
    [57] Tian B, Shaikh M A, Azii-Sadjadi M R, et al. A Study of Cloud Classification with Neural Networks Using Spectral and Textural Features. IEEETrans. NeuralNetworks, 1999, 10(1): 138-151
    [58] 段峰,王耀南,雷晓峰.机器视觉技术及其应用综述[J].自动化博览,2002(3):59-62.
    [59] Bradski G. Computer Vision Face For Use in a Percetual User Interface. Proceeding of Fourth IEEE Workshop on Appliaction s of Computer Vision, 1998: 214-219
    [60] Carfaqni, Monica, furferi. A real-time machine-vision system for monitoring the textile raising process[J]. Computers in Industry, 2005, 56: 831-842
    [61] 马颂德,张正友.计算机视觉一计算理论与算法基础[M].北京:科学出版社,1998
    [62] 中国视觉网.康耐视机器视觉系统在汽车制造业的应用.http://www.chinan.net/UploadFile/html/news/20061152033382720.htm
    [63] 深圳视觉龙.http://www.visiondragon.com
    [64] 木材缺陷图谱,中华人民共和国国家标准GB/T18000,1999
    [65] 谢佩军.基于机器视觉的晶振外壳缺陷在线抽检系统的研究[D],浙江工业大学,2006
    [66] Wren C, Azarbayejani A, Darrell T. &Pentland A. Pfinder. Real-TimeTracking of the Human Body. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, 19(7): 780~785
    [67] http://www.chuandong.cn/publisb/product/2006/12/product_2_13_38910.html
    [68] 李炜,黄心汉,王敏.基于机器视觉的带钢表面缺陷检测系统,华中科技大学学报(自然科学版)2003,31(2:72-74
    [69] H. Kong. The Design and Implementation of a real-time machine vision system for medical imaging applications[J]. ASME, 2001, 56(02): 198-206
    [70] 王慧琴.数字图像处理[M].北京:北京邮电大学出版社,2004
    [71] 孙恰,孙洪雨,白鹏.X射线焊缝图像中缺陷的实时检测方法[J].焊接学报,2004,25(2):115-118
    [72] 黎明,马聪等,机械加工零件表面纹理缺陷检测[J].中国图象图形学报A,2004,9(3):318-322
    [73] 余成波.数字图像处理及MATLAB实现[M].重庆:重庆大学出版社.2003
    [74] 周长发,精通Visual C++图像编程[M],北京:电子工业出版社。2000
    [75] 张晓波。基于图像处理技术的表面缺陷自动检测系统研究[D]。天津大学2004.12
    [76] 沈庭芝,方予文。数字图像处理及模式识别[M],北京:北京理工大学出版社,1998
    [77] 刘丹。计算机图像处理的数学和算法基础[M]。北京:国防工业出版社。2005
    [78] 冈萨雷斯.阮秋琦等译.数字图像处理[M]。北京:电子工业出版社
    [79] 何滢,曲兴华等,非金属工件电镀表面缺陷现场检测技术的研究,仪器仪表学报,2004,25(20):271~273
    [80] 刘文耀.光电图像处理[M].电子工业出版社.2002
    [81] [日]谷口庆治.数字图象处理—应用篇[M].科学出版社,2002:54-55
    [82] Amoroso C, Chella A., Morreale V &Stonnolo Soccer Robot Based on Neural Networks. The on Robocup Stockholm Sweden, 1999: P. A Segmentation System for Third International Workshop: 5-10]
    [83] Karayamnis N. B Venestsanoporlos A. N Artifical Neural Networks [M]. Kluwer Academic Publishs. 1993
    [84] Hahn-Ming Lee. Learning efficiency improvement of back-propagation algorithm by error saturation prevention method. Neurocomputing. 2001, (41): 125-143.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700