Autonomous construction of three-dimensional models from range data

摘要

This paper presents a method for autonomously constructing a complete general purpose surface description (model) of an object. The models are constructed based on a large number of noisy range images from an essentially arbitrary set of viewpoints around the object. The resulting models consist of a set of parametric surface patches. Results of testing the model building algorithm on two synthetic test objects are also presented. (C) 1997 Pattern Recognition Society. Published by Elsevier Science Ltd. [References: 13]