A1 Journal article (refereed), original research

Robot Vision System for Handling Sheets in a Manufacturing Cell


Publication Details

Authors: Ikonen Leena, Oinonen Olli, Kälviäinen Heikki

Publication year: 2001

Language: English

Related journal or series: Распознавание образов и анализ изображе... / Pattern Recognition and Image Analysis: Advances in Mathematical Theory and Applications

Volume number: 11

Issue number: 3

JUFO level of this publication:

Open Access: Not an Open Access publication


Abstract

In this paper, we present an industrial application for robotized handling of sheet metal parts in a workshop. Intelligent and reliable control and high-speed image processing are needed in this application. The parts are used for manufacturing castors and wheels. The problem is to find a pickable sheet from a box where the sheets lie disordered. First, the sheet to be picked is located roughly in the box. Then, it is picked by a robot, posed accurately, and finally moved to a next working step. A flexible machine vision system has been built by combining several image processing techniques such as edge detection, distance transforms, and blob analysis. The approach must satisfy industrial requirements for computing time and for reliable detection. The robot cell with vision consists of a robot, two cameras with three light sources, and a PC with a Matrox image processing board. Currently, the system is successfully used in real-world manufacturing in a workshop.


Last updated on 2017-22-03 at 15:46