Observatorio de I+D+i UPM

Memorias de investigación
Research Publications in journals:
Automated tone grading of granite
Year:2017
Research Areas
  • Civil engineering and architecture,
  • Housing,
  • Construction,
  • Architecture,
  • Construction of engineering,
  • Walls and closings
Information
Abstract
The production of a natural stone processing plant is subject to the intrinsic variability of the stone blocks that constitute its raw material, which may cause problems of lack of uniformity in the visual appearance of the produced material that often triggers complaints from customers. The best way to tackle this problem is to classify the product according to its visual features, which is traditionally done by hand: an operator observes each and every piece that comes out of the production line and assigns it to the closest match among a number of predefined classes, taking into account visual features of the material such as colour, texture, grain, veins, etc. However, this manual procedure presents significant consistency problems, due to the inherent subjectivity of the classification performed by each operator, and the errors caused by their progressive fatigue. Attempts to employ automated sorting systems like the ones used in the ceramic tile industry have not been successful, as natural stone presents much higher variability than ceramic tiles. Therefore, it has been necessary to develop classification systems specifically designed for the treatment of the visual parameters that distinguish the different types of natural stone. This paper describes the details of a computer vision system developed by AITEMIN for the automatic classification of granite pieces according to their tone, which provides an integral solution to tone grading problems in the granite processing and marketing industry. The system has been designed to be easily trained by the end user, through the learning of the samples established as tone patterns by the user. Keywords: computer vision, granite, natural stone, tone grading.
International
Si
JCR
No
Title
Boletín Geológico y Minero
ISBN
0366-0176
Impact factor JCR
Impact info
Volume
10.21701/bolgeomin.128.2.001
Journal number
128 (2)
From page
271
To page
286
Month
SIN MES
Ranking
Participants
  • Autor: Juan Carlos Catalina Hernandez (UPM)
Research Group, Departaments and Institutes related
  • Creador: Grupo de Investigación: Recursos Minerales
S2i 2020 Observatorio de investigación @ UPM con la colaboración del Consejo Social UPM
Cofinanciación del MINECO en el marco del Programa INNCIDE 2011 (OTR-2011-0236)
Cofinanciación del MINECO en el marco del Programa INNPACTO (IPT-020000-2010-22)