Browse Prior Art Database

Research Directions in Inclustrial Machine Vision: A Workshop Summary

IP.com Disclosure Number: IPCOM000131564D
Original Publication Date: 1982-Dec-01
Included in the Prior Art Database: 2005-Nov-11
Document File: 11 page(s) / 44K

Publishing Venue

Software Patent Institute

Related People

John F. Jarvis: AUTHOR [+3]

Abstract

Industrial machine vision has many potential applications. By identifying the major problems in several specialized areas, researchers are hoping to effectively direct the course of future investigations. Within the general discipline of computer vision are a number of application areas sufficiently distinct to warrant identification as separate fields. One of these is the application of computer vision to problems found in the industrial and manufacturing domains. Industrial machine vision is simply the use of computer processing of images, usually visual, that are part of the manufacturing process. Robot vision and automated visual inspection are complementary areas of this same field. Analysis of nonvisual images based on range data, tactile sensors, or other array presentations of physical parameters is also part of industrial machine vision. Unique characteristics of this field include the ability - - necessity, in fact -- to control the visual environment, well-defined performance and success criteria, and an extreme sensitivity to system cost and performance measures. In general, there is good control of the variability that can occur in images, although such images may not be particularly simple.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 9% of the total text.

Page 1 of 11

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

This record contains textual material that is copyright ©; 1982 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Contact the IEEE Computer Society http://www.computer.org/ (714-821-8380) for copies of the complete work that was the source of this textual material and for all use beyond that as a record from the SPI Database.

Research Directions in Inclustrial Machine Vision: A Workshop Summary

John F. Jarvis,

Bell Laboratories

Industrial machine vision has many potential applications. By identifying the major problems in several specialized areas, researchers are hoping to effectively direct the course of future investigations.

Within the general discipline of computer vision are a number of application areas sufficiently distinct to warrant identification as separate fields. One of these is the application of computer vision to problems found in the industrial and manufacturing domains. Industrial machine vision is simply the use of computer processing of images, usually visual, that are part of the manufacturing process. Robot vision and automated visual inspection are complementary areas of this same field. Analysis of nonvisual images based on range data, tactile sensors, or other array presentations of physical parameters is also part of industrial machine vision.

Unique characteristics of this field include the ability - - necessity, in fact -- to control the visual environment, well-defined performance and success criteria, and an extreme sensitivity to system cost and performance measures. In general, there is good control of the variability that can occur in images, although such images may not be particularly simple.

Three dimensional vision

What is 3-D vision from the viewpoint of industrial applications? A vision system is considered 3- D if, for example, an object can be recognized under any possible set of rotations. Similarly, if the internal representations and manipulations of the vision program are intrinsically 3-D, then the system is a 3-D vision system. Location of an object in space for grasping by a robot requires that both the position and orientation of the object be determined.

Current methods for obtaining 3-D information about the environment include disparity stereo, range images, and structured light. If the object's surface reflectivities

are well characterized, 3-D information can be extracted from shadows and shading (spatial intensity variations). Stereo imaging, the reconstruction of a 3-D scene from multiple images taken from different viewpoints, is a triangulation scheme. The same object feature points from two images are located, allowing range information to be derived from the camera separation vector and the angles to the object points. Increasing separation between the cameras increases the accuracy of the triangulation at the cost of greater difficulty in locating and matching feature poin...