CAJ | 학술논문

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。
위해결경양화망격모형특정식별불능생성엄격변계、속도만，차불능만족후속응용요구적문제，침대삼유CAD계통수출적경양화문건구유특정수거련속성적특점，제출일충쾌속특정식별산법。이용면편법향량화평균법향량쾌속획득평면특정；잉하구역이용고사영사화최소이승법축보획득주면、추면、구면이급자유곡면특정。해산법충분이용특정수거적련속성，무수건립정체탁복결구，재보증질량적정황하운산량소、속도쾌。
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 characte-ristic 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 .