汽车线束图纸的自动识别研究
|
摘要
汽车线束图纸是用来表示线束构件排布方式、线束连接关系的二维图纸,是进行线束工艺设计的主要数据来源。针对采用人工读图方法进行工艺设计产生的效率低问题,论文提出将图纸识别技术引入到线束图纸的自动识别和分析中,在分析电器件表示特点和图纸布局特点的基础上,提出基于规则的方法和基于图形相似匹配的方法。本文研究工作主要包括以下三个方面:
(1).提出基于规则的线束识别。首先,根据线束段的组成特点,建立线束段自动识别规则;然后,根据线束的连通无环性和唯一性,建立启发式规则,判断线束段的合法性和反馈识别缺失线束段。
(2).提出基于三角划分的多连通域图形精确匹配。首先,通过建立包围多边形、连通多边形概念,完成对多连通域图形的表示;然后,依据最近距离、无遮挡、包含原则添加辅助线,实现三角划分,解决包含连通的定位问题。在匹配过程中,首先通过匹配包围多边形实现多连通域图形的外轮廓定位;然后通过匹配连通多边形删除连通域,逐步收缩图形,实现精确匹配。
(3).提出基于自组织层次特征的电器件图形相似匹配。首先,通过对电器件所有连通域进行组合,建立自组织特征层次结构;其次,通过轮廓搜索和精确匹配实现自组织特征的定位;然后,通过建立自组织层次特征实现对图形相似程度的描述;最后,给出相似度计算公式,完成图形相似匹配的精确表示。
Automobile wire harness drawings represent component arrangements andwire harness connection relations and thus are the important data input in wireharness technology design. Due to the inefficiency problem in technology designresulting from manual ways, drawing recognition technology is introduced to wireharness drawing analysis. On the basis of analyzing toward the electrical part graphand drawing layout, this dissertation put forward rule-based method and graphmatching method. The main research work is listed as follows:
(1). Introduced rule-based method to recognition of wire harness. First, rules areestablished by analyzing the element of wire harness segment; then accordingto heuristic rules which are acquired from the connected acyclic feature anduniqueness, the wire harness segment is recognized and validated again.
(2). Put forward accurate matching of multi-connected region graph based ontriangulation. First, the multi-connected region graph is represented byencircled polygon and connected polygon; then, the positioning problem ofincluded connectivity is solved through triangulation according to principlesof shortest-distance, non-overlay and inclusion. In the matching process,firstly the contour is positioned by matching encircled polygon, and then theconnected region is removed successfully according to matched connectedpolygon, eventually accurate matching of rotated and scaled graph isaccomplished by shrinking the graphic region step by step.
(3). Put forward similarity matching of electrical parts based on self-organization hierarchical feature. First, self-organization feature hierarchical structure is established by combining all the connected regions of electrical part effectively; second, the self-organization feature is positioned by searching contour and accurate matching, and the similar part is identified by establishing self-organization hierarchical feature. And at last the similarity is denoted by using similarity formula.
(1).提出基于规则的线束识别。首先,根据线束段的组成特点,建立线束段自动识别规则;然后,根据线束的连通无环性和唯一性,建立启发式规则,判断线束段的合法性和反馈识别缺失线束段。
(2).提出基于三角划分的多连通域图形精确匹配。首先,通过建立包围多边形、连通多边形概念,完成对多连通域图形的表示;然后,依据最近距离、无遮挡、包含原则添加辅助线,实现三角划分,解决包含连通的定位问题。在匹配过程中,首先通过匹配包围多边形实现多连通域图形的外轮廓定位;然后通过匹配连通多边形删除连通域,逐步收缩图形,实现精确匹配。
(3).提出基于自组织层次特征的电器件图形相似匹配。首先,通过对电器件所有连通域进行组合,建立自组织特征层次结构;其次,通过轮廓搜索和精确匹配实现自组织特征的定位;然后,通过建立自组织层次特征实现对图形相似程度的描述;最后,给出相似度计算公式,完成图形相似匹配的精确表示。
Automobile wire harness drawings represent component arrangements andwire harness connection relations and thus are the important data input in wireharness technology design. Due to the inefficiency problem in technology designresulting from manual ways, drawing recognition technology is introduced to wireharness drawing analysis. On the basis of analyzing toward the electrical part graphand drawing layout, this dissertation put forward rule-based method and graphmatching method. The main research work is listed as follows:
(1). Introduced rule-based method to recognition of wire harness. First, rules areestablished by analyzing the element of wire harness segment; then accordingto heuristic rules which are acquired from the connected acyclic feature anduniqueness, the wire harness segment is recognized and validated again.
(2). Put forward accurate matching of multi-connected region graph based ontriangulation. First, the multi-connected region graph is represented byencircled polygon and connected polygon; then, the positioning problem ofincluded connectivity is solved through triangulation according to principlesof shortest-distance, non-overlay and inclusion. In the matching process,firstly the contour is positioned by matching encircled polygon, and then theconnected region is removed successfully according to matched connectedpolygon, eventually accurate matching of rotated and scaled graph isaccomplished by shrinking the graphic region step by step.
(3). Put forward similarity matching of electrical parts based on self-organization hierarchical feature. First, self-organization feature hierarchical structure is established by combining all the connected regions of electrical part effectively; second, the self-organization feature is positioned by searching contour and accurate matching, and the similar part is identified by establishing self-organization hierarchical feature. And at last the similarity is denoted by using similarity formula.
引文
[1]李光耀,李慎国.我国汽车用电线束现状及发展趋势[J].汽车与配件,2006,40:28-30.
[2]2005年版中国线束产品市场竞争研究.http://www.hdcmr.com/article/yjbg/05/04/1007.html[EB/OL],2006-11.
[3]混合动力汽车兴起—车用线束日渐重要.http://tech.digitimes.com.tw/Show News.aspx?zCatId=335&zNotesDocId=790D7731F13B7678482570F40044B3BD[EB/OL],2006-01-16.
[4]王玉,邢渊,阮雪榆.机械产品设计重用策略研究[J].机械工程学报,2002,38(5):145-148.
[5]王玉.工程设计重用理论及其关键技术研究[D].上海交通大学,2001:2-5.
[6]董玉德,赵韩,王平,张国平等.工程图纸识别与理解的研究现状分析[J].合肥工业大学学报(自然科学版),2005,28(1):29-33.
[7]Xu X W,Bai Y B.Computerising scanned engineering documents[J].Computers Industry,2000,42(1):59-71.
[8]Li B,Suen C V.A knowledge—based thinning algorithm[J].Pattern Recognition,1991,24(12):1211-1221.
[9]朱建新,周济.工程图信息系统—EDIS[J].高技术通讯,1994,(10):17-19.
[10]Nagasamy V,Langrana N A.Engineering drawing processing and vectorization system[J].CVGIP,1990,49:379-397.
[11]S.D.Connell,A.K.Jain.Template-Based On-Line Character Recognition[J].Pattern Recognition,2000,43(1):1-14.
[12]李伟青,谭建荣,彭群生.基于图段结构的整体识别方法的研究[J].计算机学报,1998,21(8):753-758.
[13]LaViola Jr.,Joseph J.;Zeleznik,Robert C.A Practical Approach for Writer-Dependent Symbol Recognition Using a Writer-Independent Symbol Recognizer[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2007,29(11):1917-1926.
[14]Nezis K,Vosniakos G.Recognizing 2D shape features using a neural network and heuristics[J].Computer Aided Design,1997,29(7):523-539.
[15]Nishida H.Shape recognition by integrating structural descriptions and geometrical/statistical transforms[J].Computer Vision and Image Understanding,1996,64(2):248-262.
[16]王若瑜,蔡蛛华,曹阳,杨士杰,李恒.基于规则的建筑结构图钢筋用量自动识别系统[J].软件学报,2002,3(4):574-579.
[17]沈怡,蔡士杰,高晓.建筑工程图符号的特征匹配识别方法[J].计算机辅助设计与图形学学报,2003,15(9):1065-1069.
[18]Yu Y,Samal A,Seth S C.A system for recognizing a large class of engineering drawings[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,1997,19(8):868-890.
[19]R.J Qureshi,J.Y Ramel,U.Ali,H.Cardot.Graph Based Recognition of Isolated Graphic Symbols[A]IEEE International Conference on Emerging Technologies(ICET'05)[C],17-18 September 2005,Islamabad- Pakistan,pp.109-114.
[20]Di Stefano P,Bianconi F,Di Angelo L.An approach for feature semantics recognition in geometric models[J].Computer Aided Design,2004,36(10):993-1009.
[21]Legg L,Hsu C,Bruderlin B D,et al.Inferring3D CAD models from free hand sketches and constraints[J].Computer Aided Design,1995,27(4):243-254.
[22]高纬,彭群生.基于二维视图特征的三维重建[J].计算机学报,1999,22(5):481-485.
[23]王耘,刘衍聪,应道宁.工程特征的识别与理解—问题与思考[J].中国机械工程,1997,8(1):80-83.
[24]Fujisaki Atsushi,Samitsu Norikazu.Development of flexible flat cable harnesses for automobile roof modules.Furukawa Review,2002,21:87-91.
[25]Osada Yasufumi,Nakano Hiroyuki.Development of wire harness modules.Furukawa Review,1999,18:7-11.
[26]黄海云.MRP系统在汽车线束生产中的应用[J].华中理工大学学报,2000,28(2):96-97.
[27]刘华,毛宁.基于Web的汽车线束行业工艺信息系统研究[J].机电产品开发与创新,2005,18(5):78-79.
[28]何士双,徐本柱,吴黄,刘晓平.基于AutoCAD的线束电器件库设计与实现[A].计算机技术与应用进展2007(下)[C]:937-940.
[29]吴黄,石慧,刘晓平.基于虚框与虚点的线束工艺图建模研究[A].计算机技术与应用进展2007(下)[C]:1026-1030.
[30]程光春,徐本柱,刘晓平.基于参数化的多连通图的表示[A].计算机技术与应用进展2007(下)[C]:998-1002.
[31]周美立.相似工程学[M].北京:机械工业出版社,1998.
[32]刘雪梅,李爱平,朱文博.零部件建库方法研究[J].工程图学学报,2006,27(1):30-34.
[33]杨勇.数据库系统中BLOB对象的管理[J].微电子学与计算机,2006,23(7): 147-149.
[34]王恪典,徐海波,基于AutoCAD的网络化CAD实现[J].工程图学学报,2004,25(2):8-14.
[35]马建,滕弘飞,孙治国,杨宏宇.图形匹配问题[J].计算机科学,2001,28(4):61-64.
[36]Liu H C,Srinsth M D.Partial shape classification using contour matching in distance transformation.IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,12(11):1072-1079.
[37]S.Smithies,K.Novins,J.Arvo.A Handwriting-Based Equation Editor[A].Proc.Graphics Interface Conf.[C],1999,pp.84-91.
[38]Davis L.Shape matching using relaxation techniques.IEEE Transactions on Pattern Analysis and Machine Intelligence.1979,PAM-1(1):66-72.
[39]Liang Lin,Song-Chun Zhu,Yongtian Wang.Layered Graph Match with Graph Editing[A].IEEE Conf.on Computer Vision and Pattern Recognition(CVPR)[C],2007,pp.1-8.
[40]王雪飞,胡青泥.基于分层加权的多边形图形匹配[J].工程图学学报,2002,23(2):89-96.
[41]谭建荣,岳小莉,陆国栋.图形相似的基本原理、方法及其在结构模式识别中的应用[J].计算机学报,2002,25(9):959-967.
[2]2005年版中国线束产品市场竞争研究.http://www.hdcmr.com/article/yjbg/05/04/1007.html[EB/OL],2006-11.
[3]混合动力汽车兴起—车用线束日渐重要.http://tech.digitimes.com.tw/Show News.aspx?zCatId=335&zNotesDocId=790D7731F13B7678482570F40044B3BD[EB/OL],2006-01-16.
[4]王玉,邢渊,阮雪榆.机械产品设计重用策略研究[J].机械工程学报,2002,38(5):145-148.
[5]王玉.工程设计重用理论及其关键技术研究[D].上海交通大学,2001:2-5.
[6]董玉德,赵韩,王平,张国平等.工程图纸识别与理解的研究现状分析[J].合肥工业大学学报(自然科学版),2005,28(1):29-33.
[7]Xu X W,Bai Y B.Computerising scanned engineering documents[J].Computers Industry,2000,42(1):59-71.
[8]Li B,Suen C V.A knowledge—based thinning algorithm[J].Pattern Recognition,1991,24(12):1211-1221.
[9]朱建新,周济.工程图信息系统—EDIS[J].高技术通讯,1994,(10):17-19.
[10]Nagasamy V,Langrana N A.Engineering drawing processing and vectorization system[J].CVGIP,1990,49:379-397.
[11]S.D.Connell,A.K.Jain.Template-Based On-Line Character Recognition[J].Pattern Recognition,2000,43(1):1-14.
[12]李伟青,谭建荣,彭群生.基于图段结构的整体识别方法的研究[J].计算机学报,1998,21(8):753-758.
[13]LaViola Jr.,Joseph J.;Zeleznik,Robert C.A Practical Approach for Writer-Dependent Symbol Recognition Using a Writer-Independent Symbol Recognizer[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2007,29(11):1917-1926.
[14]Nezis K,Vosniakos G.Recognizing 2D shape features using a neural network and heuristics[J].Computer Aided Design,1997,29(7):523-539.
[15]Nishida H.Shape recognition by integrating structural descriptions and geometrical/statistical transforms[J].Computer Vision and Image Understanding,1996,64(2):248-262.
[16]王若瑜,蔡蛛华,曹阳,杨士杰,李恒.基于规则的建筑结构图钢筋用量自动识别系统[J].软件学报,2002,3(4):574-579.
[17]沈怡,蔡士杰,高晓.建筑工程图符号的特征匹配识别方法[J].计算机辅助设计与图形学学报,2003,15(9):1065-1069.
[18]Yu Y,Samal A,Seth S C.A system for recognizing a large class of engineering drawings[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,1997,19(8):868-890.
[19]R.J Qureshi,J.Y Ramel,U.Ali,H.Cardot.Graph Based Recognition of Isolated Graphic Symbols[A]IEEE International Conference on Emerging Technologies(ICET'05)[C],17-18 September 2005,Islamabad- Pakistan,pp.109-114.
[20]Di Stefano P,Bianconi F,Di Angelo L.An approach for feature semantics recognition in geometric models[J].Computer Aided Design,2004,36(10):993-1009.
[21]Legg L,Hsu C,Bruderlin B D,et al.Inferring3D CAD models from free hand sketches and constraints[J].Computer Aided Design,1995,27(4):243-254.
[22]高纬,彭群生.基于二维视图特征的三维重建[J].计算机学报,1999,22(5):481-485.
[23]王耘,刘衍聪,应道宁.工程特征的识别与理解—问题与思考[J].中国机械工程,1997,8(1):80-83.
[24]Fujisaki Atsushi,Samitsu Norikazu.Development of flexible flat cable harnesses for automobile roof modules.Furukawa Review,2002,21:87-91.
[25]Osada Yasufumi,Nakano Hiroyuki.Development of wire harness modules.Furukawa Review,1999,18:7-11.
[26]黄海云.MRP系统在汽车线束生产中的应用[J].华中理工大学学报,2000,28(2):96-97.
[27]刘华,毛宁.基于Web的汽车线束行业工艺信息系统研究[J].机电产品开发与创新,2005,18(5):78-79.
[28]何士双,徐本柱,吴黄,刘晓平.基于AutoCAD的线束电器件库设计与实现[A].计算机技术与应用进展2007(下)[C]:937-940.
[29]吴黄,石慧,刘晓平.基于虚框与虚点的线束工艺图建模研究[A].计算机技术与应用进展2007(下)[C]:1026-1030.
[30]程光春,徐本柱,刘晓平.基于参数化的多连通图的表示[A].计算机技术与应用进展2007(下)[C]:998-1002.
[31]周美立.相似工程学[M].北京:机械工业出版社,1998.
[32]刘雪梅,李爱平,朱文博.零部件建库方法研究[J].工程图学学报,2006,27(1):30-34.
[33]杨勇.数据库系统中BLOB对象的管理[J].微电子学与计算机,2006,23(7): 147-149.
[34]王恪典,徐海波,基于AutoCAD的网络化CAD实现[J].工程图学学报,2004,25(2):8-14.
[35]马建,滕弘飞,孙治国,杨宏宇.图形匹配问题[J].计算机科学,2001,28(4):61-64.
[36]Liu H C,Srinsth M D.Partial shape classification using contour matching in distance transformation.IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,12(11):1072-1079.
[37]S.Smithies,K.Novins,J.Arvo.A Handwriting-Based Equation Editor[A].Proc.Graphics Interface Conf.[C],1999,pp.84-91.
[38]Davis L.Shape matching using relaxation techniques.IEEE Transactions on Pattern Analysis and Machine Intelligence.1979,PAM-1(1):66-72.
[39]Liang Lin,Song-Chun Zhu,Yongtian Wang.Layered Graph Match with Graph Editing[A].IEEE Conf.on Computer Vision and Pattern Recognition(CVPR)[C],2007,pp.1-8.
[40]王雪飞,胡青泥.基于分层加权的多边形图形匹配[J].工程图学学报,2002,23(2):89-96.
[41]谭建荣,岳小莉,陆国栋.图形相似的基本原理、方法及其在结构模式识别中的应用[J].计算机学报,2002,25(9):959-967.