Home // INTELLI 2021, The Tenth International Conference on Intelligent Systems and Applications // View article

Vision-based Inspection System for OrnamentalStone Using a Weighted Hybrid Ensemble Classifier

Authors:
Liliana Antão
Rui Pinto
Manuel João Ferreira
Tiago Pinto
Gil Gonçalves

Keywords: Ornamental Stone, Ensemble Classifier, Convolutional Neural Network, Machine Vision.

Abstract:
With the increasing demands of customers in the ornamental stone industry, both in terms of the individual specifications of each product and in the delivery times, it is necessary to constantly adapt the manufacturing processes and their inherent complexity and, consequently, the automated systems that are essential to them. There is a strong movement of research in areas capable of generating non-destructive testing techniques applied to production systems in this sector. Currently, one of the main problems occurs during the ornamental stone slab polishing phase, where there is the need to monitor the polishing quality and diagnose possible defects in the surface of the slab. This can be used as feedback for self-correction and optimization of variables and process parameters in the polishing equipment. In this paper is proposed a monitoring system, based on machine vision techniques, used to detect defects in the surface of polished ornamental stone slabs. This approach is based on a weighted hybrid ensemble classifier, relying on image processing techniques and a Convolutional Neural Network. Results show that the ensemble classifier outperforms related classifiers.

Pages: 43 to 49

Copyright: Copyright (c) IARIA, 2021

Publication date: July 18, 2021

Published in: conference

ISSN: 2308-4065

ISBN: 978-1-61208-882-2

Location: Nice, France

Dates: from July 18, 2021 to July 22, 2021