Model-based recognition of arbitrary surfaces from range data

Authors:
Jeffrey A. Bloom;Chang Y. Choo;William I. Kwak
Affiliations:
Worcester Polytechnic Institute, Department of Electrical Engineering, Worcester, Massachusetts;Worcester Polytechnic Institute, Department of Electrical Engineering, Worcester, Massachusetts;Digital Equipment Corporation, 500 Donald Lynch Blvd, Marlboro, Massachusetts
Venue:
IEA/AIE '90 Proceedings of the 3rd international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
Year:
1990

Citing 5
Cited 0

Three-dimensional object recognition

ACM Computing Surveys (CSUR) - Annals of discrete mathematics, 24
Computational geometry: an introduction

Computational geometry: an introduction
Principles of pictorial information systems design

Principles of pictorial information systems design
Computer Vision

Computer Vision
Design, analysis, and application of a three-dimensional vision system in computer integrated manufacturing

Design, analysis, and application of a three-dimensional vision system in computer integrated manufacturing

Quantified Score

Hi-index	0.01

Visualization

Abstract

Recognition of arbitrary surfaces is a difficult and largely unsolved problem in computer vision. In this paper, we present a technique to develop a piecewise planar, triangular patch model of an object surface from its range data, and a recognition technique to be used with this model. A large set of object surface data points are segmented into triangular patches using a small number of knot points. The recognition technique extracts and stores in the form of attributed connection graph the features of all known surfaces from their triangular patch models. A similar model can be built from an unknown surface and its features detected. These features and relationships between the features are compared with those of the known surfaces for recognition.