摘要
本文的主要目的是实现印刷质量检测的自动化、精确化、高效率及低损耗,为印刷质量检测的规范化和数据化管理提供理论与实践依据。其重要性在于该研究有利于推动印刷科学技术的进步与发展,为人民的物质和文化生活提供高质量的印刷包装产品。本文的研究符合当前印刷质量检测快速、高效、精确的发展方向,研究成果运用于实际生产有助于提高印刷检测速度,降低生产成本,提高企业经济效益。
本文的主要研究方法:运用数字图像处理原理和印刷基本原理,针对印刷质量自动检测的关键环节,根据印刷实际条件,利用模板定位待检测图像;根据牛顿迭代优化法和抽样法确定实验,利用平均值比较法和方差分析法处理数据,对误差产生的原因进行分析。
本文主要工作:1)根据连通区域分析法确定了模板噪声去除的方法;2)实验得到了图像旋转过程中倾斜角度的最大允许误差;3)使用点定位和区域定位相结合,缩短了定位的时间,并提高了定位的精度;4)使用不同阈值制作模板,消除了使用单张印刷制作模板带来的误差,使模板的制作更加精确。
主要结论:1)用高斯模板去噪,可以最小限度地减少图像信息的损失;2)实验表明,图像旋转过程中允许的最大倾斜角度为0.002弧度;3)采用点定位和区域定位相结合的方法,缩短了图像定位的时间,提高了精度;4)阈值选取范围在最大灰度差的80-100%内,所制作的模板效果最好。总之,本文提出的印刷质量检测系统与传统的质量检测方法相比,更安全、可靠,具有较高的效率。
The main task of this thesis is to set up a digital and integration system to check and measure printing quality, which holds the four characters of automation, precision, efficiency and less cost, and thereby, finds basic principles and provides practice for normalization of printing-quality-measure system. The research follows, or even, guides the trend of printing technology, and undoubtedly, would help to accelerate the printing technology development and bring more benefits to the enterprises and people.
Based on traditional technology, this new system combines image processing theories and printing principles, and also provides new theory for the for the automatic printing quantity checking system. Image match-orientation and Newton Iteration method are two main methods used in the system. In the data analysis compare and equation data processing methods are used to find the reason of variation.
The main work:1)Selected the method of image noise reduction via connected-area-analysis. 2)Found the allowable max variation of image tilt based on the confirmatory experiment. 3)Took advantage to a less time cost algorithm and advance orientation precision by combination utilizing of orientation of dot and orientation of area in orientation of image. 4)Made template in different threshold in to reduce and eliminate printing quality
algorithm and advance orientation precision by combination utilizing of orientation of dot and orientation of area in orientation of image. 4)Made template in different threshold in to reduce and eliminate printing quality variation.
Experiment results have been shown that :1) Reducing noise of the image by Gaussian Template reaches the minimal lost of the image;2)The maximum tilt angle of the image is 0. 002 radian;3)The method of images orientation combined the dot orientation;4)When making template, using of 80% to 100% threshold value is efficient;5)The measurement method of the new system is more exact and efficient than traditional checking method.
引文
[1] 中国印刷通史,第十八章当代社会印刷发展概况第一节.http://www.cgan.net/book/books/print/g-history/gb9/181.htm
[2] 海德堡印刷专刊,短版彩印市场的新选择.2001(1):14-15.
[3] Gardner, Steven Joseph. A wireless printing application using Web services technology. UNIVERSITY OF LOUISVILLE:2002.
[4] 张林桂.以纸为承载的印刷100年也不会消失,广东印刷,2003(6):23-25.
[5] 全球印刷简要分析,今日印刷.2003.2:83.
[6] Liao J H, Yin B Z, Lao W W and Xie H. Photoelectric detection system in the recognition and sorting equipment of bills. SPIE 3558,646~651,1998.
[7] Thomas Friese, Tiziana Margaria, Alfred Hofmann. Integrating printed and online information Technical innovation. Int J STTT 1998(2): 202.
[8] 北京长城九千质量认证咨询有限公司软件事业部供稿,印刷企业电子化的ISO9000质量管理体系与企业计算机管理综合解决方案,印刷质量与标准化,2002.5:2-5.
[9] Bela Lombay ,Laszlo Szabo, Alex Yeboah .How to get it on paper and in print. N. Thorne Griscom, Department of Radiology, Children's Hospital, Boston,MA 02115 USA:877.
[10] 刘昕,陈路.基于画面的印品质量检测控制系统之研究.印刷技术,2001.11(2):45-46.
[11] Datenschnittstelle. Von ClP3 zu CIP4. Druck & Medien-Magazin, 2001.6: 51.
[12] John Macphee, Presses-past, present, and future, GATA, 1998.
[13] 李震宇,译.海德堡的新一代印刷控制系统CP2000.印刷世界,
2000.1:37.
[14] Doelle, Klaus. Application of fiber loading technology to improve paper strength and optical properties of lightwieght, high opacity printing and copy paper. THE UNIVERSITY OF WISCONSIN - MADISON:2002.
[15] 李琴,王强.印刷图像质量评价方法的研究.印刷世界,2001.5:23-25.
[16] 王晓红,徐昌权,周世生,等.文字印刷质量的抽样统计规律及其在模糊评判中的应用.西北纺织工学院学报,1999.6:145-149.
[17] 付忠良,罗中兵.阮波.一种改进的模板匹配算法,计算机应用,2002.3:28-30.
[18] 王岩松,阮秋琦.一种基于互相关的图像定位匹配算法研究及应用.北方交通大学学报,2002.4:20-24.
[19] 韩斌,刘以安,吴小俊,等.印刷缺陷检测中的图像对准.华东船舶工业学院学报(自然科学版),2002.4:37-40.
[20] 陈路.基于画面的印刷质量检测控制系统研究.西安理工大学硕士论文.2002.3.
[21] 韩斌,刘以安,王士同.基于图像处理的印刷缺陷计算机自动检测.计算机应用,2002.3(21):37-39.
[22] 章毓晋,黄翔宇,李睿.自动检测精细印刷品缺陷的初步方案.中国体视学与图像分析,2001.6:109-113.
[23] 鲁镇恶,谢勇.印刷品外观缺陷机器视觉的检测与识别.包装工程,2002.5:10-11.
[24] 冯波.彩色图像印刷质量评价相关问题浅析.印刷世界,2002.4:26.
[25] 中国印刷专利十七则.印刷杂志,专利之窗,2002.4.
[26] 陈德荣,编译.印刷品质量的检测与控制.北京:印刷工业出版社,1998:1.
[27] 刘世昌.印刷图像质量测控技术.北京:印刷工业出版社,1998:50-51:165-166.
[28] 胡成发.印刷色彩与色度学.北京:印刷工业出版社,1993:298-301.
[29] Hinderliter, H. &HUTTON, P.N. Thermal Plate Study, Graphic Arts Technical Foundation, 1998.
[30] Ulicheney R. Digital Halftoing. Cambridge, MA: MIT press, 1987.
[31] 冯波.彩色图像印刷质量评价相关问题浅析.印刷世界,2002.4:26-27.
[32] 汤顺青.色度学.北京:科学技术出版社,1990:112-145.
[33] 刘昕,郭锦.平版胶印工艺技术.北京:中国轻工业出版社,2002:19-21.
[34] 肖道钧.胶印工艺技术读本.北京:印刷工业出版社,1995:34-37.
[35] 胡成发.印刷色彩与色度学.北京:印刷工业出版社,1993:83-85;310-318.
[36] Gershon Elber, George Wolberg. Rendering traditional mosaics. The Visual Computer, 2003(19):67-78.
[37] 徐建华.图像处理与分析.北京:科学出版社,1992:140-145.
[38] David McDowell, Standards Consultant Eastman Kodak Company, Color Management What' s needed for Printing & Publishing ?.Standards articles published by the IPA Standards Committe, 2000:1-7.
[39] Soren Winslow, COLOR MANAGEMENT WORKFLOW ANALYSIS, Printed in Demark 2000. The Graphic Arts Institute of Denmark, 2000:8-10.
[40] Klaman, Marianne. Aspects on colour rendering, colour predic tion and colour control in printed media .KUNGLIGA TEKNISKA HOGSKOLAN (SWEDEN):2002:1-24.
[41] 龚修端,刘昕.印刷品质量实时监测系统,包装工程,2003(6):45—47.
[42] 周长发.精通Visual C++ 图像编程,北京:电子工业出版社,2000:343-348.
[43] Kenneth. R. Castleman. Digital Image Processing. Prentice Hall, New York, 1998:112-120.
[44] 周长发.精通Visual C++.net图像编程.北京:电子工业出版社.2000:348-356.
[45] 王润生.图像理解.长沙:国防科技大学出版社,1995:145-151.
[46] 何斌,马天予,王运坚,等.Visum C++ 数字图像处理.北京:人民邮电出版社,2001:281-282.
[47] R.L. Kashyap and R. chellapa, Decision rules for choice of neighbors in random field models of images comp. graph and Image Processing 15, 1991:301-318.
[48] G.R. Cross and A.K. Jain, Markov random field texture, IEEE T tans and Machine Intelligence, 5(1),1998:25-39.
[49] 吕风军.数字图像处理编程入门.北京:清华大学出版社,1999:28-33.
[50] 谢天保.印刷图像处理.西安理工大学教材科,1997:92-95.
[51] 章毓晋.图像理解与计算机视觉.北京:清华大学出版社,1998:163-164.
[52] A. Verikas, K. Malmqvist, and L. Bergman. Neural Networks Ba sed Colour Measuring for Process Monitoring and Control in Multicoloured Newspaper Printing. Neural Comput & Applic, 2000(9):227-242.
[53] 章毓晋,黄翔宇,李睿.自动检测精细印刷品缺陷的初步方案.中国体视学与图像分析,2001.6:109-113.
[54] 王万中,茆诗松.实验的设计与分析.上海:华东师范大学出版社,1997.5:10-18.