Share Email Print

Proceedings Paper

Chain code technique for the classification and orientation of simple objects in a machine vision task
Author(s): David Kerr; Jonathan T. Wakenshaw
Format Member Price Non-Member Price
PDF $17.00 $21.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Many routine pick-and-place tasks require a combinational software analysis approach in that a particular object must first be recognized before orienting a robot gripper or other tool to pick it up. The first step requires the segmentation of pattern feature from the image in order to make the classification. The second step concerns the determination of the position and orientation of the classified object. We present an approach to this two-stage problem that utilizes only the Freeman chain code of the object outline, rather than the image itself. We show that, given the chain code, it is possible to segment a number of specific geometrical pattern features that can be used to identify the object. From the same code, it is further demonstrated that the object location can be specified by computing its center of mass and minor axis of inertia. It is thus possible to identify and locate entities within an image given only their chain codes. The algorithms are demonstrated on a variety of simple shapes. The method is at present restricted to solid shapes, but could be extended to include objects of greater complexity.

Paper Details

Date Published: 27 August 1999
PDF: 11 pages
Proc. SPIE 3836, Machine Vision Systems for Inspection and Metrology VIII, (27 August 1999); doi: 10.1117/12.360281
Show Author Affiliations
David Kerr, Loughborough Univ. (United Kingdom)
Jonathan T. Wakenshaw, Loughborough Univ. (United Kingdom)

Published in SPIE Proceedings Vol. 3836:
Machine Vision Systems for Inspection and Metrology VIII
John W. V. Miller; Susan Snell Solomon; Bruce G. Batchelor, Editor(s)

© SPIE. Terms of Use
Back to Top