Browse Prior Art Database

Inferred Three Dimensional Machine Vision Data

IP.com Disclosure Number: IPCOM000061424D
Original Publication Date: 1986-Jun-01
Included in the Prior Art Database: 2005-Mar-09
Document File: 1 page(s) / 12K

Publishing Venue

IBM

Related People

Colson, JC: AUTHOR

Abstract

There is described a technique permitting the inference of information for describing a 6 degree of freedom robotic manipulator goal from a two dimensional image of the object to be manipulated.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 76% of the total text.

Page 1 of 1

Inferred Three Dimensional Machine Vision Data

There is described a technique permitting the inference of information for describing a 6 degree of freedom robotic manipulator goal from a two dimensional image of the object to be manipulated.

It is desired in most general vision aided robotic assembly tasks to have three dimensional information about an object to be robotically manipulated. Conventional ways of obtaining this information are costly in both hardware and time, typically requiring two or more cameras and complex computations. The present system provides an improvement by using a single camera for determining three dimensional information.

Assuming an object at rest is at rest in one of plural stable states, recognition parameters for the object may be taught to the vision-aided robotic system. Then, the manipulator movement relative to the object is taught using the following sequence. A picture of the object at rest in one of its stable states is made. A coordinate system generated representing that position and orientation of the object relative to a predetermined coordinate system. The coordinate system describes the object in three degrees of freedom, and a third dimension is chosen perpendicular to the image plane. The system user moves the manipulator to a desired position and orientation in order to teach a transformation representing the desired position in orientation relative to the object frame.

Further, assuming that an object is in anoth...