A Study of Reverse Engineering on Rebuilding of Surfaces and Modification of Surface Features

博士 === 國立臺灣科技大學 === 機械工程系 === 87 === This thesis studies the following two topics: feature-based surface rebuilding and feature-based surface modification. Upon them, some algorithms are developed to achieve the targets. The first topic induces to the algorithms of data-preprocessing and feature-bas...

Full description

Bibliographic Details
Main Authors: Jiung-Ming Huang, 黃俊明
Other Authors: Wei-Cherng Lu
Format: Others
Language:zh-TW
Published: 1999
Online Access:http://ndltd.ncl.edu.tw/handle/48043652445192166780
Description
Summary:博士 === 國立臺灣科技大學 === 機械工程系 === 87 === This thesis studies the following two topics: feature-based surface rebuilding and feature-based surface modification. Upon them, some algorithms are developed to achieve the targets. The first topic induces to the algorithms of data-preprocessing and feature-based surface fitting. Data-preprocessing smoothes the position of data points and helps get simple sample points for feature-based surface fitting. In surface fitting, a surface is fitted through the sample points to rebuild the surface features with only a few control points. The second topic induces to the algorithms of features extraction, division of featured control points and the modification of surface features. In features extraction, the differentiation of a B-spline curve is discussed and an algorithm to find the boundaries of featured curve segments are developed. Featured control points are defined to be the control points within a closed feature boundary, and will be moved in the modification of the local feature. In the computer demonstration, feature points of curve segments are identified and concise feature models are built. Also, in the demonstration, the boundaries for surface features are clearly identified and it helps the definition of the modifying features. And in the pulling up, pressing down, moving sideward or even in the delete or recovery of a feature, it give good results to the algorithms proposed in this thesis.