CAD轻量化网格模型的特征识别  

Feature recognition for CAD lightweight mesh model

作　　者：罗年猛[1] 殷吉功 刘云华[1] 朱林[1] 

机构地区：[1]华中科技大学国家CAD支撑软件工程技术研究中心,湖北武汉430074

出　　处：《计算机工程与设计》2014年第8期2876-2880,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(51075162;61173115)

摘　　要：为解决轻量化网格模型特征识别不能生成严格边界、速度慢,且不能满足后续应用要求的问题,针对三维CAD系统输出的轻量化文件具有特征数据连续性的特点,提出一种快速特征识别算法。利用面片法向量和平均法向量快速获得平面特征;剩下区域利用高斯映射和最小二乘法逐步获得柱面、锥面、球面以及自由曲面特征。该算法充分利用特征数据的连续性,无需建立整体拓扑结构,在保证质量的情况下运算量少、速度快。To solve the problems that the feature recognition of lightweight mesh model works slowly and fails to generate strict boundaries and meet the requirements of the subsequent application, considering the continuity characteristic of feature data of lightweight file output by 3D CAD system, a fast feature recognition method was proposed. The facet normal and the average normal were used to obtain plane features; features of the cylinder, the cone, the sphere and the freeform were gained gradually by using Gaussian mapping and least squares for the remaining region. The algorithm makes full use of the continuity characteristic of feature data without establishing the whole topology structure Therefore on the premise of meeting quality requirements, it is characterized by its small calculating amount and fast speed.

关 键 词：特征识别 轻量化 CAD系统 网格模型 连续性 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]