|
摘要：
|
|
在冲裁模智能设计CBR系统的研究与开发过程中,零件的识别是难点之一,零件的识别也就是零件相似性匹配。从几何匹配的角度来看,零件外轮廓的相似性匹配是最重要的。本文针对冲裁件外轮廓的相似性匹配提出了一种两级相似性度量方法,该方法已经应用到冲裁模智能设计CBR系统中的图形匹配模块,实践证明,此法具有较好的识别质量和计算机处理效率。
|
|
Recognition of parts is one of the difficulties in the process of research and development of CBR system of blanking die intelligence design. Part recognition means the parts' similarity matching. From the viewpoint of geometric matching, parts' outline similarity matching is the most important in the process. Aiming to resolve this problem, the author proposes a two-level measuring method. The method has been applied to graphics matching module in the CBR system and is proved a good one in the aspects of recognition as well as of computer processing efficiency.
|
|
基金项目：
|
|
广州市科技攻关项目(2004Z3D0101)
|
|
作者简介：
|
|
|
|
参考文献：
|
[1]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,(2):8996.
[2]ParkJongH,UmBongS.Anewapproachtosimilarityre trievalof2Dgraphicobjectsbasedondominantshapes[J].PatternRecognition,1999,20(6):591616.
[3]AmarM,AggarwalJK.Contourregistrationbyshape specif icpointsforshapematching.CVGIP1983,22:396408.
[4]HuttenlocherDanielP,KlandermanGregoryA.ComparingImagesUsingtheHausdorffDistance[J].IEEETransac tionsonPatternAnalysisandMachineIntelligence,1993,15:850863.
|
|
服务与反馈：
|
|
【文章下载】【加入收藏】
|
|
|
|
