基于OCR技术的飞机定检数据收集系统设计
|
摘要
飞机定检数据对分析飞机的适航性和监控飞行系统可靠性是非常重要的数据,但是现阶段该类数据的记录依然是以传统的纸质工卡为主。这样造成大量的定检数据不能用计算机来处理,不能对飞机性能进行趋势分析和故障预测。本研究正是基于以上背景,研究和探讨了基于OCR技术的飞机定检数据收集系统中的关键技术,并对飞机定检数据收集程序系统进行了详细设计和实现。本文的主要
工作在以下几个方面:
(1)对飞机定检数据收集的现状进行调研,主要说明了这一现状的不足。为解决这一问题,引入了OCR技术,并分析这一技术的可行性。
(2)进行系统需求分析,阐述了系统目标、主要功能、运行环境和用户分析。特别对我国维修电子化需求进行详细描述。
(3)完成了基于OCR技术的飞机定检工卡数据收集系统的总体设计。系统分四个功能模块:预处理、图像分割、特征提取及分类器设计和后处理。
(4)针对后处理关键技术,本研究提出并详细阐释了基于上下文语义词库的相似字型优选法的后处理方法,是提高识别精度和速度的保障。
(5)描述了一种基于中文分词技术的机务专业词库的建立方法,并利用Access2007软件建立专业词汇数据库。
(6)对基于OCR技术的飞机定检数据收集系统进行了程序实现,并用VB编制了一个友好的界面。通过对某一典型的飞机定检工卡的数据收集识别进行实例验证,证明用该程序系统进行飞机定检数据收集识别是切实有效的。
Aircraft scheduled maintenance data is very import for analyzing a aircraft airworthiness and monitoring reliability of systems of a aircraft .but now, these datum are written on paper job cards. So that, a lot of scheduled maintenance datum cannot been processed and used in trend analysis and prediction of failure for performance of a aircraft. The objectives of the paper are designing a system of collecting data from aircraft scheduled maintenance and researching on critical technology. This paper work is following:
(1)Research of current situation of collecting aircraft scheduled maintenance data; mainly disadvantages from current situation are explained. In order to solve this problem, OCR technology is applied and feasibility is analyzed.
(2)System requirements are studied, including system purpose, main function, operating environment and user analysis. Domestic e-maintenance requirements are described comprehensively.
(3) System design conception of collecting aircraft scheduled maintenance data based on OCR technology is presented. The system includes four functional modules, including preprocess image, segmentation, feature extraction and sorter design, and after-treatment.
(4)After-treatment is the critical technology for this system. Method of similar font identification based on context meaning is described.
(5) The concept of maintenance specialize vocabularies database is presented in this module. Database is established using Access2007.
(6) Program is coded with friendly interface. Through example test, this program system is proved effective to identify aircraft scheduled maintenance data.
工作在以下几个方面:
(1)对飞机定检数据收集的现状进行调研,主要说明了这一现状的不足。为解决这一问题,引入了OCR技术,并分析这一技术的可行性。
(2)进行系统需求分析,阐述了系统目标、主要功能、运行环境和用户分析。特别对我国维修电子化需求进行详细描述。
(3)完成了基于OCR技术的飞机定检工卡数据收集系统的总体设计。系统分四个功能模块:预处理、图像分割、特征提取及分类器设计和后处理。
(4)针对后处理关键技术,本研究提出并详细阐释了基于上下文语义词库的相似字型优选法的后处理方法,是提高识别精度和速度的保障。
(5)描述了一种基于中文分词技术的机务专业词库的建立方法,并利用Access2007软件建立专业词汇数据库。
(6)对基于OCR技术的飞机定检数据收集系统进行了程序实现,并用VB编制了一个友好的界面。通过对某一典型的飞机定检工卡的数据收集识别进行实例验证,证明用该程序系统进行飞机定检数据收集识别是切实有效的。
Aircraft scheduled maintenance data is very import for analyzing a aircraft airworthiness and monitoring reliability of systems of a aircraft .but now, these datum are written on paper job cards. So that, a lot of scheduled maintenance datum cannot been processed and used in trend analysis and prediction of failure for performance of a aircraft. The objectives of the paper are designing a system of collecting data from aircraft scheduled maintenance and researching on critical technology. This paper work is following:
(1)Research of current situation of collecting aircraft scheduled maintenance data; mainly disadvantages from current situation are explained. In order to solve this problem, OCR technology is applied and feasibility is analyzed.
(2)System requirements are studied, including system purpose, main function, operating environment and user analysis. Domestic e-maintenance requirements are described comprehensively.
(3) System design conception of collecting aircraft scheduled maintenance data based on OCR technology is presented. The system includes four functional modules, including preprocess image, segmentation, feature extraction and sorter design, and after-treatment.
(4)After-treatment is the critical technology for this system. Method of similar font identification based on context meaning is described.
(5) The concept of maintenance specialize vocabularies database is presented in this module. Database is established using Access2007.
(6) Program is coded with friendly interface. Through example test, this program system is proved effective to identify aircraft scheduled maintenance data.
引文
[1]荆涛,王仲.光学字符识别技术与展望[J].计算机工程,2003,29(2):1-2.
[2]张文国.汉王OCR证照识别系统及其应用[J].中国信息导报, 2003,(12) .
[3]荆涛,王仲.光学字符识别技术与展望[J].计算机工程, 2003,(02).
[4]孙家广,杨长贵.计算机图形学[M].北京:清华大学出版社,1995:566—567.
[5]杨智勇.复杂背景下的文本提取技术[D].江西师范大学硕士学位论文, 2004.
[6]马立权,李维,蔡韩辉.手写数字识别中的预处理技术研究[J].仪器仪表学报,2001,22(3): 263-265.
[7]汪益民,梅林,张义超.基于OCR技术的书写文字识别系统设计[J].甘肃科技,2003,23 (8):17-19.
[8]郭捷,施鹏飞.基于颜色和纹理分析的车牌定位方法[J].中国图象图形学报, 2002.5 (A), 472-476.
[9] F. Leclerc and R. Plamondon, Automatic signature verification: the state of the art, 1989-1993[J]. International Journal of Pattern Recognition and Artificial Intelligence, 1994.3(1):643-660.
[10] A.A. Kholmatov, Biometric Identity Verification Using On-Line & Off-line Sig- nature Verification [D], Master’s thesis, Sabanci University,2003.
[11] E.A.Fadhel and P.Bhattacharyya, Ability of neural network for offline signature verification[C]. the Int. Conf. Know. Based Computer System.(KBCS 2000), Mumbai, India,Dec.Reco-gnition,2003,Calcutta,2003.
[12] C.Gonzalez, P.Wintz, Digital Image Processing. [M],2nd edition, Addison-Wesley, MA.1987
[13] V.E. Ramesh, M.Narasimha Murty, Off-line signature verification using geneti- cally optimized weighted features. [J]. Pattern Recognition, 1999, 32:217-233.
[14]张纯,张涛,黄笑.中文商务名片识别系统的实现[J].中文信息学报,2000,14(2):21-25.
[15] Symmetrix Remote Data Facility Product Description Guide[M].EMC Corporation,2000.
[16]陈国平,张明新,付跃文等.高性能的多体印刷英文识别系统的实现[J] .计算机工程与应用, 2006 (12) :183-186.
[17]陈艳,孙羽菲,张玉志.灰度图像中字符切分方法的研究.[J]中文信息学报,2004,(04).
[18] David A.Patterson. John L.Hennessy.Computer Architecture:A Quantitative Approach.[M].北京:机械工业出版社,1999.
[19]周源,白顺科.基于上下文语义词库的OCR识别方法[J].中国制造业信息化,2008,37(23): 62-64.
[20]张旭.一个基于词典与统计的中文分词算法[D]:[硕士学位论文].西安:电子科技大学,2006.
[21]祁正华.基于无词库的中文分词方法的研究[D]:[硕士学位论文].南京:南京邮电学院软件学院,2005.
[22]高荣芳,赵安科,刘茜.数据库原理与应用(第二版),西安电子科技大学出版社,2009
[2]张文国.汉王OCR证照识别系统及其应用[J].中国信息导报, 2003,(12) .
[3]荆涛,王仲.光学字符识别技术与展望[J].计算机工程, 2003,(02).
[4]孙家广,杨长贵.计算机图形学[M].北京:清华大学出版社,1995:566—567.
[5]杨智勇.复杂背景下的文本提取技术[D].江西师范大学硕士学位论文, 2004.
[6]马立权,李维,蔡韩辉.手写数字识别中的预处理技术研究[J].仪器仪表学报,2001,22(3): 263-265.
[7]汪益民,梅林,张义超.基于OCR技术的书写文字识别系统设计[J].甘肃科技,2003,23 (8):17-19.
[8]郭捷,施鹏飞.基于颜色和纹理分析的车牌定位方法[J].中国图象图形学报, 2002.5 (A), 472-476.
[9] F. Leclerc and R. Plamondon, Automatic signature verification: the state of the art, 1989-1993[J]. International Journal of Pattern Recognition and Artificial Intelligence, 1994.3(1):643-660.
[10] A.A. Kholmatov, Biometric Identity Verification Using On-Line & Off-line Sig- nature Verification [D], Master’s thesis, Sabanci University,2003.
[11] E.A.Fadhel and P.Bhattacharyya, Ability of neural network for offline signature verification[C]. the Int. Conf. Know. Based Computer System.(KBCS 2000), Mumbai, India,Dec.Reco-gnition,2003,Calcutta,2003.
[12] C.Gonzalez, P.Wintz, Digital Image Processing. [M],2nd edition, Addison-Wesley, MA.1987
[13] V.E. Ramesh, M.Narasimha Murty, Off-line signature verification using geneti- cally optimized weighted features. [J]. Pattern Recognition, 1999, 32:217-233.
[14]张纯,张涛,黄笑.中文商务名片识别系统的实现[J].中文信息学报,2000,14(2):21-25.
[15] Symmetrix Remote Data Facility Product Description Guide[M].EMC Corporation,2000.
[16]陈国平,张明新,付跃文等.高性能的多体印刷英文识别系统的实现[J] .计算机工程与应用, 2006 (12) :183-186.
[17]陈艳,孙羽菲,张玉志.灰度图像中字符切分方法的研究.[J]中文信息学报,2004,(04).
[18] David A.Patterson. John L.Hennessy.Computer Architecture:A Quantitative Approach.[M].北京:机械工业出版社,1999.
[19]周源,白顺科.基于上下文语义词库的OCR识别方法[J].中国制造业信息化,2008,37(23): 62-64.
[20]张旭.一个基于词典与统计的中文分词算法[D]:[硕士学位论文].西安:电子科技大学,2006.
[21]祁正华.基于无词库的中文分词方法的研究[D]:[硕士学位论文].南京:南京邮电学院软件学院,2005.
[22]高荣芳,赵安科,刘茜.数据库原理与应用(第二版),西安电子科技大学出版社,2009