基于机器视觉的酒精棉纱外包装质量检测系统的研究与开发
|
摘要
传统的产品包装质量检测依靠人工完成,检测效率较低,并容易出现漏检错检。因此将数字图像处理技术应用于包装自动检测中,研究基于机器视觉的自动产品质量检测系统对实现生产的自动化、智能化具有重要意义。
本文主要研究了酒精棉纱外包装表面印刷质量的在线检测系统。系统的硬件部分包括CCD相机、图像采集卡、LED光源、光电传感器及电平转换电路等,为了实现图像采集卡、光源、系统的传送设备能够协同工作,设计了外部触发同步电路。
针对酒精棉纱外包装存在的各种缺陷,首先研究了如何利用图像采集技术和图像处理技术来对缺陷特征进行表达,在此基础上分别研究了图文印刷错误、印刷瑕疵、包装破损、包装封口处的不重合度、边缘裂口褶皱等缺陷的检测算法。为了提高算法的执行速度,根据本项目直线检测的特点,对Hough变换直线检测算法进行改进,减小了直线检测的计算量。针对包装边缘图像中存在的规律特征,利用灰度共生矩阵的方法描述该部分的纹理特征,并设计了神经元感知器的方法来完成边缘裂口、褶皱的检测。
最后本文对实现的酒精棉纱外包装表面印刷质量检测系统进行了调试与验证,对不同缺陷类型的图像进行了实验,结果表明系统的硬件设计和软件设计能够协同工作,系统性能达到了预期的设计目标。
Manual print quality control is still adopted in many situations, so it is very necessary to develop automatic print quality detection systems based on machine vision technique, which will benefit in many aspects.
The thesis mainly focuses on an on-line print quality examination system for printed alcohol tissue package.The hardware system is consisted of CCD camera, image grabber board, led lamp light source, electro-optical sensor and electrical level transfer circuit. An trigger circuit is designed to make CCD camera, image grabber board, and light source working synchronically.
In order to detect all kinds of flaws that possibly happen to printed alcohol tissue bags, features to express such flaws are studied by applying image collection method and image processing technology. These flaws may be printing error, bag breakage, superposition error at sealed edges. To improve the efficiency of algorithm, according to the characteristic of the line in the image, an improved hough transform algorithm for line detection is proposed, which has highly reduced the computational complexity. Texture is found in the image sealed area, and a method has been developed with co-occurrence matrix features and single neuron classification.
At the end of the thesis the system evaluation and debugging are implemented , it is proved that both the hardware and the software can work well as it is expected.
本文主要研究了酒精棉纱外包装表面印刷质量的在线检测系统。系统的硬件部分包括CCD相机、图像采集卡、LED光源、光电传感器及电平转换电路等,为了实现图像采集卡、光源、系统的传送设备能够协同工作,设计了外部触发同步电路。
针对酒精棉纱外包装存在的各种缺陷,首先研究了如何利用图像采集技术和图像处理技术来对缺陷特征进行表达,在此基础上分别研究了图文印刷错误、印刷瑕疵、包装破损、包装封口处的不重合度、边缘裂口褶皱等缺陷的检测算法。为了提高算法的执行速度,根据本项目直线检测的特点,对Hough变换直线检测算法进行改进,减小了直线检测的计算量。针对包装边缘图像中存在的规律特征,利用灰度共生矩阵的方法描述该部分的纹理特征,并设计了神经元感知器的方法来完成边缘裂口、褶皱的检测。
最后本文对实现的酒精棉纱外包装表面印刷质量检测系统进行了调试与验证,对不同缺陷类型的图像进行了实验,结果表明系统的硬件设计和软件设计能够协同工作,系统性能达到了预期的设计目标。
Manual print quality control is still adopted in many situations, so it is very necessary to develop automatic print quality detection systems based on machine vision technique, which will benefit in many aspects.
The thesis mainly focuses on an on-line print quality examination system for printed alcohol tissue package.The hardware system is consisted of CCD camera, image grabber board, led lamp light source, electro-optical sensor and electrical level transfer circuit. An trigger circuit is designed to make CCD camera, image grabber board, and light source working synchronically.
In order to detect all kinds of flaws that possibly happen to printed alcohol tissue bags, features to express such flaws are studied by applying image collection method and image processing technology. These flaws may be printing error, bag breakage, superposition error at sealed edges. To improve the efficiency of algorithm, according to the characteristic of the line in the image, an improved hough transform algorithm for line detection is proposed, which has highly reduced the computational complexity. Texture is found in the image sealed area, and a method has been developed with co-occurrence matrix features and single neuron classification.
At the end of the thesis the system evaluation and debugging are implemented , it is proved that both the hardware and the software can work well as it is expected.
引文
[1]朱飞虎.机器视觉原理及应用.自动化博览,2005,22(2):81-83.
[2]胡兴军,蔡叶菁,王健.机器视觉技术及其在包装印刷质量检测中的应用.丝网印刷,2004,11:35-37.
[3]陈梅兰.机器视觉关键技术与应用实例分析.现代计算机,2006,1:21-25.
[4]盛晶.代替人眼的机器视觉-全自动印刷品质量检测技术.印刷杂志,2005,9:43-45.
[5]韩斌,刘庆安.基于图像处理的印刷缺陷计算机自动检测.自动化技术与应用,2002,21(3):37-38.
[6]段峰,王耀南,雷晓峰,吴立钊等.机器视觉技术及其应用综述.自动化博览,2002(3):59-62.
[7]关胜晓.机器视觉及其应用发展.自动化博览,2005,22(3):88-92.
[8]龚修端,刘昕.印刷品质量实时检测技术.包装工程,2003,24(6):45-46.
[9]费耀平,谷士文.工业视觉系统及应用.长沙铁道学院学报,1997,15(1):25-30.
[10]戚飞虎,周源华.模式识别与图像处理.上海:上海交通大学出版社,2000,8-21.
[11]阮秋琦.数字图像处理学.北京:电子工业出版社,2001,12-41.
[12]孙宇清,马军.在线检测实时处理系统的设计与实现.计算机工程与应用,2000,36(12):172-173.
[13]柯尊友,黄胜华等.数字图像处理技术在动态实时检测控制中的应用研究.电子技术应用,1999,25(2):17-18.
[14]张建民.数字图像处理技术在印章鉴别中的应用.太原重型机械学院学报,1995,16(3):206-208.
[15]J.Caron,L.Duvieubourg.Defect Detection by Recursive Filters in Packaging Industry.Proceedings of 1997 international Conference on Information,Communications and Signal Processing.1997(2):1189-1191.
[16]LiTao,Peter Wity.Machine Vision in The Inspection of Pattemed Textile Webs.IEE Colloquium on Industrial Inspection.1997:91-95.
[17]Jung Hun Kim,Suk Ahn,A High-speed High-resolution Vision System for the Inspection of TFT LCD.IEEE International Symposium On Industrial Electronics.2001(6):101-105.
[18]J.R.Coaton,A.M.Marsden.Lamps and Lighting.1992:20-27.
[19]任明灿.基于计算机视觉鸡蛋品质检测的研究.硕士学位论文.上海交通大学,2007.
[20]吴越豪.非接触式药片缺粒检测研究顾士学位论文.浙江大学.2006.
[21]JAI UK Ltd.,EN-50081-1,CV-Al-20 Operation Manual,England,JAI UK Ltd.,2004.
[22]Coreco Imaging Inc.,OC-PV2M-VUM00,PC2-Vision User's Manual,Canada,Coreco Inc.,2002.
[23]王庆有.CCD应用技术.天津:天津大学出版社.2000,38-39.
[24]李春成,蒋利波,余明友等.印刷品质量高速在线检测系统设计.空军雷达学院学报,2005,19(4):61-63.
[25]贾永红.数字图像处理.武汉:武汉大学出版社.2003,22-25.
[26]刘榴娣.实用数字图像处理.北京:北京理工大学出版社.1998,176-180.
[27]王新成.高级图像处理技术.北京:中国科学技术出版社.2000,234-235.
[28]熊凌.计算机视觉中的图像匹配综述.湖北工业大学学报,2006,21(3):171-173.
[29]胡歌华,周曼丽,柳健.Hough变换的一种改进及其在直线检测中的应用.Radio Engineering of China,1999,29(5):42-44.
[30]杨波,汪同庆,叶俊勇等.一种直线提取的新方法.计算机工程,2003.29(1):45-47.
[31]何斌,马天予.Visual C++数字图像处理.北京:人民邮电出版社.2001,426-434.
[32]郭林.特殊票据票面OCR字符分割及识别方法研究.硕士学位论文.山东大学.2003.
[33]吴晓兵.印刷体光学字符的研究.博士学位论文.中国科学院自动化研究所.2004.
[34]熊新炎.基于指纹和手背静脉的多模态生物特征识别系统研究.硕士学位论文.哈尔滨工程大学.2007.
[35]何斌,马天予等.Visual C++数字图像处理(第二版).北京:人民邮电出版社.2001,473-477
[36]王丰元,高松,孙壮志.数学形态学原理及其在图像处理中的应用.山东工程学院学报,1999,13(2):5-8
[37]贾永红.数字图像处理.武汉:武汉大学出版社.2003,182-186
[38]韩力群.人工神经网络理论、设计及应用.北京:化学工业出版社.2002,20-21
[39]王洪元.人工神经网络技术及其应用.北京:科技出版社.2002,4-5
[40]田景文,高美娟.人工神经网络算法及应用.北京:北京理工大学出版社.2006,25-28
[41]曲中周.基于图像处理的香烟小包装在线实时检测系统.硕士论文.南京航空航天大学.2004.
[2]胡兴军,蔡叶菁,王健.机器视觉技术及其在包装印刷质量检测中的应用.丝网印刷,2004,11:35-37.
[3]陈梅兰.机器视觉关键技术与应用实例分析.现代计算机,2006,1:21-25.
[4]盛晶.代替人眼的机器视觉-全自动印刷品质量检测技术.印刷杂志,2005,9:43-45.
[5]韩斌,刘庆安.基于图像处理的印刷缺陷计算机自动检测.自动化技术与应用,2002,21(3):37-38.
[6]段峰,王耀南,雷晓峰,吴立钊等.机器视觉技术及其应用综述.自动化博览,2002(3):59-62.
[7]关胜晓.机器视觉及其应用发展.自动化博览,2005,22(3):88-92.
[8]龚修端,刘昕.印刷品质量实时检测技术.包装工程,2003,24(6):45-46.
[9]费耀平,谷士文.工业视觉系统及应用.长沙铁道学院学报,1997,15(1):25-30.
[10]戚飞虎,周源华.模式识别与图像处理.上海:上海交通大学出版社,2000,8-21.
[11]阮秋琦.数字图像处理学.北京:电子工业出版社,2001,12-41.
[12]孙宇清,马军.在线检测实时处理系统的设计与实现.计算机工程与应用,2000,36(12):172-173.
[13]柯尊友,黄胜华等.数字图像处理技术在动态实时检测控制中的应用研究.电子技术应用,1999,25(2):17-18.
[14]张建民.数字图像处理技术在印章鉴别中的应用.太原重型机械学院学报,1995,16(3):206-208.
[15]J.Caron,L.Duvieubourg.Defect Detection by Recursive Filters in Packaging Industry.Proceedings of 1997 international Conference on Information,Communications and Signal Processing.1997(2):1189-1191.
[16]LiTao,Peter Wity.Machine Vision in The Inspection of Pattemed Textile Webs.IEE Colloquium on Industrial Inspection.1997:91-95.
[17]Jung Hun Kim,Suk Ahn,A High-speed High-resolution Vision System for the Inspection of TFT LCD.IEEE International Symposium On Industrial Electronics.2001(6):101-105.
[18]J.R.Coaton,A.M.Marsden.Lamps and Lighting.1992:20-27.
[19]任明灿.基于计算机视觉鸡蛋品质检测的研究.硕士学位论文.上海交通大学,2007.
[20]吴越豪.非接触式药片缺粒检测研究顾士学位论文.浙江大学.2006.
[21]JAI UK Ltd.,EN-50081-1,CV-Al-20 Operation Manual,England,JAI UK Ltd.,2004.
[22]Coreco Imaging Inc.,OC-PV2M-VUM00,PC2-Vision User's Manual,Canada,Coreco Inc.,2002.
[23]王庆有.CCD应用技术.天津:天津大学出版社.2000,38-39.
[24]李春成,蒋利波,余明友等.印刷品质量高速在线检测系统设计.空军雷达学院学报,2005,19(4):61-63.
[25]贾永红.数字图像处理.武汉:武汉大学出版社.2003,22-25.
[26]刘榴娣.实用数字图像处理.北京:北京理工大学出版社.1998,176-180.
[27]王新成.高级图像处理技术.北京:中国科学技术出版社.2000,234-235.
[28]熊凌.计算机视觉中的图像匹配综述.湖北工业大学学报,2006,21(3):171-173.
[29]胡歌华,周曼丽,柳健.Hough变换的一种改进及其在直线检测中的应用.Radio Engineering of China,1999,29(5):42-44.
[30]杨波,汪同庆,叶俊勇等.一种直线提取的新方法.计算机工程,2003.29(1):45-47.
[31]何斌,马天予.Visual C++数字图像处理.北京:人民邮电出版社.2001,426-434.
[32]郭林.特殊票据票面OCR字符分割及识别方法研究.硕士学位论文.山东大学.2003.
[33]吴晓兵.印刷体光学字符的研究.博士学位论文.中国科学院自动化研究所.2004.
[34]熊新炎.基于指纹和手背静脉的多模态生物特征识别系统研究.硕士学位论文.哈尔滨工程大学.2007.
[35]何斌,马天予等.Visual C++数字图像处理(第二版).北京:人民邮电出版社.2001,473-477
[36]王丰元,高松,孙壮志.数学形态学原理及其在图像处理中的应用.山东工程学院学报,1999,13(2):5-8
[37]贾永红.数字图像处理.武汉:武汉大学出版社.2003,182-186
[38]韩力群.人工神经网络理论、设计及应用.北京:化学工业出版社.2002,20-21
[39]王洪元.人工神经网络技术及其应用.北京:科技出版社.2002,4-5
[40]田景文,高美娟.人工神经网络算法及应用.北京:北京理工大学出版社.2006,25-28
[41]曲中周.基于图像处理的香烟小包装在线实时检测系统.硕士论文.南京航空航天大学.2004.