RESEARCH ARTICLE
Copyright © 2016 American Scientific Publishers
All rights reserved
Printed in the United States of America
Journal of
Computational and Theoretical Nanoscience
Vol. 13, 1–8, 2016
Research on Shape Feature Recognition of Boundary
Representation CAD Model
Jihua Wang
College of Information Science and Engineering, Shandong Normal University, Jinan, 250014, China
B-Rep (Boundary Representation) CAD model is being widely used in representation of industrial
product, so its feature recognition has acquired widespread research interests in computer vision
and 3D model retrieval fields. We present one approach of feature recognition based on the idea
of human visual mechanism. Surfaces as the visual shape features, and solids as well as shells as
the topological relations, were extracted from the neutral STEP (Standard for Exchange of Product
Model Data) files of B-Rep model. Towards three surface types of NURBS, analytical and poly
loop, the properties of surface boundary and region are established based on curvature and other
geometric index. So B-Rep CAD model is characterized as the hierarchical tree with solid layer,
shell layer and surface layer for object recognition and retrieval, and the corresponding experiments
verified the effectiveness of the method of shape feature recognition.
Keywords: Shape Features, B-Rep CAD Model, Solid, Shell, STEP, Object Recognition,
3D Retrieval, NURBS, Topology Scheme, Geometry Schema.
1. INTRODUCTION
B-Rep (boundary representation) CAD model to repre-
sent widespread industrial products is widely shared and
used in STEP format (Standard for Exchange of Product
Model Data), so the shape feature reco gnition of B-Rep
CAD model need arouse enough attention from researchers
in computer vision and 3D (three dimensional) model
retrieval fields. 3D object recognition in computer vision
and content-based 3D model retrieval in design field have
fascinated humans in the recent decades due to their the-
oretical and practical values. Computer vision is to match
the 2D (two dimensional) image features with the 3D
object features,
1
in fact, it is 2D–3D feature matching. 3D
model retrieval is mainly to compute the feature similar-
ity extracted from 3D object representation (3D model), it
is 3D-3D feature matching. In the both fields, 3D object
representation or 3D model format in computer is required
to solve in advance, then its features can be recognized
and matched. 3D multimedia and visualization models in
the general field are currently the dominant in triangular
mesh format such as .STL, .OBJ, and .VRML. Industrial
products widely exist in real life, they are represented by
B-Rep CAD model commonly accepted in STEP format,
but the research on shap e feature recognition o f B-Rep
CAD model is not enough. The current feature extrac-
tion algorithms for B-Rep CAD model retrieval mostly
adopt the mesh-based methods, using the appearance con-
tents such as outline, skeleton, material, colour, grain, etc.
The current feature detection algorithms for B-Rep CAD
model in computer vision are very few. So the shape fea-
ture recognition and feature modeling for B-Rep CAD
model in STEP format is of great significance, no matter
for computer vision or 3D model retrieval.
The paper proposes the shape feature recognition
approach for B-Rep CAD model which is more consis-
tent with perceptual intuition, since it is based on the
idea of human visual mechanism in psychology widely
cited in computer vision field. Surfaces, as the key visual
and functional features in ecological psychology and affor-
dance theory,
33
are first extracted from the neutral STEP
file of B-Rep CAD model, and the geometric properties of
surface boundary and region are computed as shape fea-
tures. Solids and shells in B-Rep CAD model which reflect
the clear topological relations among surfaces are also
extracted from STEP file. Therefore B-Rep CAD m odel
is characterized by the hierarchical tree with solid layer,
shell layer and surface layer, which is seen as its feature
model for 3D CAD model recognition and retrieval.
The claimed novelty of this research lays in the shape
feature recognition of B-Rep CAD model based o n sur-
faces in STEP file. (1) Surface boundary and region prop-
erties to reflect the object function or semantic are the most
important human visual perception contents. (2) Solid and
shell properties reflect the topological relations among sur-
faces, thus the h ierarchical tree with solid layer, shell layer
and surface layer as the feature model more contains the
J. Comput. Theor. Nanosci. 2016, Vol. 13, No. 12 1546-1955/2016/13/001/008 doi:10.1166/jctn.2016.6348 1