CAD Data Composition Method from 3D Sensed Data Based on Feature Extraction

Abstract

This paper proposes a new method for constructing CAD data from 3D sensed data for the purpose of realizing reverse engineering. In CAD data of many industrial products, geometrical features such as flat planes and cylinders are defined on the boundary surfaces. In the proposed method flat planes, cylinders and cones can be extracted robustly from the complicated geometry or high noisy sensed and CAD data that contains the features can be constructed. The robust feature extraction is done through the following approaches: (1) Estimating geometric parameters of the features using the Gaussian image: and (2) Extracting the features by using “growing surface method”. The method has been applied to some 3D sensed data and the validity of the method was confirmed.