Original scientific paper
Intelligent Machine Vision System for Automated Quality Control in Ceramic Tiles Industry
Tomislav KESER
orcid.org/0000-0001-7300-0659
; Faculty of Electrical Engineering, J.J. Strossmayer University of Osijek, Osijek, Hrvatska
Željko HOCENSKI
; Faculty of Electrical Engineering, J.J. Strossmayer University of Osijek, Osijek, Hrvatska
Verica HOCENSKI
; KIO ceramic industry, Orahovica, Croatia
Abstract
Intelligent system for automated visual quality control of ceramic tiles based
on machine vision is presented in this paper. The ceramic tiles production
process is almost fully and well automated in almost all production stages
with exception of quality control stage at the end. The ceramic tiles quality
is checked by using visual quality control principles where main goal is to
successfully replace man as part of production chain with an automated
machine vision system to increase production yield and decrease the
production costs. The quality of ceramic tiles depends on dimensions and
surface features. Presented automated machine vision system analyzes
those geometric and surface features and decides about tile quality by
utilizing neural network classifier. Refined methods for geometric and
surface features extraction are presented also. The efficiency of processing
algorithms and the usage of neural networks classifier as a substitution for
human visual quality control are confirmed.
Keywords
Ceramic tiles; Machine vision; Neural networks; Visual inspection; Quality control
Hrčak ID:
56738
URI
Publication date:
30.4.2010.
Visits: 3.051 *