首页 | 本学科首页   官方微博 | 高级检索  
     

一种工程图检索方法在匹配优化问题中的应用
引用本文:赵晔,翟世梅. 一种工程图检索方法在匹配优化问题中的应用[J]. 安徽大学学报(自然科学版), 2012, 36(2): 67-73
作者姓名:赵晔  翟世梅
作者单位:1. 西安工业大学理学院,陕西西安,710032
2. 西安高科技研究所,陕西西安,710025
基金项目:国家自然科学基金资助项目(2009JM1017);陕西省自然科学基金资助项目(2009JM1017)
摘    要:以二维工程图检索系统的设计与开发为背景,提出一种基于非精确图匹配的二维工程图检索方法.该方法首先将二维工程图转化为一种基于图元的属性化邻接图;然后计算目标模型与被检索模型的属性化邻接图之间顶点相容程度矩阵与边相容程度矩阵,并由此建立顶点匹配矩阵M的目标优化函数;最后运用Sinkhorn行列交替规范化方法求解匹配优化问题.在匹配过程中,充分利用属性邻接图的顶点与边属性信息动态裁剪搜索空间,实现快速匹配.实验结果表明,该方法能够检索到不同相似程度的二维工程图,并且检索效率也能满足实际要求.

关 键 词:工程图检索  非精确图匹配  属性化邻接图  Sinkhorn  二次优化  连续法

The application in matching optimization problem of engineering drawing retrieval
ZHAO Ye , ZHAI Shi-mei. The application in matching optimization problem of engineering drawing retrieval[J]. Journal of Anhui University(Natural Sciences), 2012, 36(2): 67-73
Authors:ZHAO Ye    ZHAI Shi-mei
Affiliation:1.College of Sciences,Xi’an Technological University,Xi’an 710032,China; 2.Xi’an High-tech Research Institute,Xi’an 710025,China)
Abstract:Based on the engineering drawing system design and implementation,an engineering drawing retrieval method based on inexact graph matching was presented.First,a representation of primitive attributed relational graph(ARG) for each engineering drawing was extracted;and then the vertex compatibility matrix and edge compatibility matrix between the ARGs of the target and searched model were calculated,and the measure of the similarity between the two models was created.Finally,the Sinkhorn’s alterative normalization method was adopted to solve the matching optimal problem.In the matching process,the ARG’s node and edge attributes were used to prune the search space.Experimental results showed that this method was able to support the inexact model retrieval and its efficiency meets the requirements of practical applications.
Keywords:engineering drawings retrieval  inexact graph matching  attributed relational graph  Sinkhorn  quadratic optimal method  continuation method
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号