Share Email Print

Proceedings Paper

On Describing A Robotic Scene
Author(s): Vincent Hayward; Stephane Aubry
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The focus of this article is the presentation of the issues involved in the design of a system able to automatically or semi-automatically construct a global, world-centered model of an industrial scene. In a robotic context, the uses of such a model could be the detection and prevention of collisions, and the interactive off-line programming of manipulators. More generally, such a model can be useful for a variety of automation tasks, including graphic simulation and machining operations sequencing. We first review range acquisition and solid modelling techniques in light of their respective appropriateness to the above problem. Then we explain how the process of constructing a scene differs from both object recognition and image processing, since we are a priori not interested in isolating distinguishing features of objects but rather in faithfully describing entire scenes. We also discuss how the merging of the several views of the scene is a central element of the scene acquisition process and how great care must be exercised in order to maintain connectivity consistency. Finally, we show that data acquisition, solid modelling as well the actual process of constructing the scene are closely interrelated and that these three issues, all of which are necessary components of the scene description process, should not be considered independently of each other.

Paper Details

Date Published: 19 February 1988
PDF: 8 pages
Proc. SPIE 0848, Intelligent Robots and Computer Vision VI, (19 February 1988); doi: 10.1117/12.942799
Show Author Affiliations
Vincent Hayward, Computer Vision and Robotics Laboratory (Canada)
Stephane Aubry, Computer Vision and Robotics Laboratory (Canada)

Published in SPIE Proceedings Vol. 0848:
Intelligent Robots and Computer Vision VI
David P. Casasent; Ernest L. Hall, Editor(s)

© SPIE. Terms of Use
Back to Top