Knowledge-based robot vision system for automated part handling

• Published in: South African Journal of Industrial Engineering Vol. 19; no. 1; pp. 119 - 135
• Main Authors: Van Niekerk, T.I., Hua, T., Wang, J., Hattingh, D.G.
• Format: Journal Article
• Language: English
• Published: Bedfordview South African Institute of Industrial Engineers (SAIIE) 01.05.2008; South African Institute for Industrial Engineering; The Southern African Institute for Industrial Engineering
• Subjects: Artificial intelligence; Data collection; Engineering, Industrial; Feature extraction; Knowledge management; Neural networks; Pattern recognition; Robots; Sensors; Vision systems
• ISSN: 1012-277X; 2224-7890; 2224-7890

Hierdie artikel beskryf 'n kennis-gebaseerde visiesisteemalgoritme wat in 'n industriële robotsisteem ingesluit word om sodoende intelligente komponenthantering te bewerkstellig. 'n Kontinue wasige beheerder is gebruik om allerlei objekinligting deur middel van 'n effektiewe berekeningsmetode te bepaal. Die ontwikkelde algoritme vir aan-lyn komponentherkenning maak gebruik van wasige logika en word bewys as 'n effektiewe metode om geometriese inligting van objekte te bepaal. Die voorgestelde grensvektormetode verskaf voldoende inligting en maak geometriese rekonstruksie van die objek moontlik om greepbeplanning te kan doen. Voorts is 'n komponenthanteringsmodel ontwikkel deur die grypkenmerke af te lei uit die geometriese eienskappe. This paper discusses an algorithm incorporating a knowledge-based vision system into an industrial robot system for handling parts intelligently. A continuous fuzzy controller was employed to extract boundary information in a computationally efficient way. The developed algorithm for on-line part recognition using fuzzy logic is shown to be an effective solution to extract the geometric features of objects. The proposed edge vector representation method provides enough geometric information and facilitates the object geometric reconstruction for gripping planning. Furthermore, a part-handling model was created by extracting the grasp features from the geometric features.