Browse Prior Art Database

STRUCTURED LIGHT SENSORS FOR 3-D ROBOT VISION

IP.com Disclosure Number: IPCOM000128222D
Original Publication Date: 1983-Dec-31
Included in the Prior Art Database: 2005-Sep-15
Document File: 4 page(s) / 20K

Publishing Venue

Software Patent Institute

Related People

J.T. Schwartz: AUTHOR [+3]

Abstract

A structured light or active illumination visual sensor is one which illuminates objects using non-uniform beams of light on which properties varying from point to point through the geometric extent of the beam have been impressed. Any observable property can be used to form such a beam; among properties which one might think of exploiting are intensity, polarisation, spectral distribution of optical energy, coherence, and (temporal) modulation. Structured light sensors can be used to achieve 3-D vision, that is, to build visual sensors that report the true position in 3-dimensional space of points on the surface of a body or bodies being observed. Besides giving information on object proximity, sensor output of this kind should be very useful for identifying the objects present in a scene and determining their positions and orientations.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 29% of the total text.

Page 1 of 4

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

STRUCTURED LIGHT SENSORS FOR 3-D ROBOT VISION

BY

J.T. Schwartz

Technical Report No. 65 Robotics Report No. 8

March, 1983

p~36' Work on this paper has been supported in part by Office of Naval Research Grant N00014-82-K-0381, and by grants from the Digital Equipment Corporation, the Sloan Foundation, the System Development Foundation, and the IBM Corpration. Structured Light Sensors for 3-D Robot Vision

J. T. Schwartz Courant Institute of Mathematical Sciences New York University

A structured light or active illumination visual sensor is one which illuminates objects using non- uniform beams of light on which properties varying from point to point through the geometric extent of the beam have been impressed. Any observable property can be used to form such a beam; among properties which one might think of exploiting are intensity, polarisation, spectral distribution of optical energy, coherence, and (temporal) modulation. Structured light sensors can be used to achieve 3-D vision, that is, to build visual sensors that report the true position in 3-dimensional space of points on the surface of a body or bodies being observed. Besides giving information on object proximity, sensor output of this kind should be very useful for identifying the objects present in a scene and determining their positions and orientations.

The simplest form of structured light approach is the striped light scheme originally introduced by P. Will and K. Pennington of IBM Research. In this sensor (see Fig. 1) a plane P of light formed by passing collimated light from a source S through a slit is used to illuminate a body B. The resulting illuminated stripe on the body B is observed through a camera C offset from the source S. Then the 3-D position of each illuminated point X can be calculated as the unique point of intersection of the plane P with the known line L from the camera through X. Note that L is known since it is determined geometrically by the point at which the image of X falls on the retina of C, (See Fig. 1, below.) We also suppose this wedge of light to be structured, in the sense that some measurable property Q is imposed upon its separate rays; this property is assumed to be constant in ,each vertical plane of the wedge W, but to vary monotonically from left to right across the wedge, which is to say monotonically with the geometric parameter h shown in Fig. 2, which defines the vertical plane in which a given point of W lies.

In the Will-Pennington scheme, the whole surface of the body B can be determined by panning the plane P over the whole visual field. An improved scheme, which allows this whole surface to be determined more rapidly, is as follows. Suppose that the body B is illuminated by a wedge W of light as in Fig. 2 below.

The internal property Q imposed upon the wedge W of light must have the following properties:

New York University Page 1 Dec 31, 1983

Page 2 of 4

ST...