Identificación de variedades de aceitunas a partir del endocarpo utilizando visión por computador

  1. E. M. Ortega Vázquez 1
  2. D. Martínez Gila 1
  3. S. Satorres Martínez 1
  4. J. Gómez Ortega 1
  5. J. Gámez García 1
  1. 1 Universidad de Jaén
    info

    Universidad de Jaén

    Jaén, España

    ROR https://ror.org/0122p5f64

Libro:
XXXIX Jornadas de Automática: actas. Badajoz, 5-7 de septiembre de 2018
  1. Inés Tejado Balsera (coord.)
  2. Emiliano Pérez Hernández (coord.)
  3. Antonio José Calderón Godoy (coord.)
  4. Isaías González Pérez (coord.)
  5. Pilar Merchán García (coord.)
  6. Jesús Lozano Rogado (coord.)
  7. Santiago Salamanca Miño (coord.)
  8. Blas M. Vinagre Jara (coord.)

Editorial: Universidad de Extremadura

ISBN: 978-84-9749-756-5,  978-84-09-04460-3

Año de publicación: 2018

Páginas: 1014-1021

Congreso: Jornadas de Automática (39. 2018. Badajoz)

Tipo: Aportación congreso

DIALNET GOOGLE SCHOLAR lock_openRUC editor

Resumen

The identification of olives is of great importance for a multitude of factors, including harvesting, olive oil production process and trade exchanges. Precisely identifying varieties is a time-consuming task in addition to trained experts or specific and expensive equipment. When applying the traditional method, also known as morphological, a specialist assesses morphological features using the olive endocarp. In this paper a proposal to automate this identification methodology is presented. Endocarp images are acquired and analyzed to extract the endocarp features, processed by Wilk’s Lambda. Then, the varieties are identified by two classifiers: partial least squares discriminant analysis and different support vector machines. The proposal has been tested on a set of 250 samples from 5 varieties from the south of Spain.

Fuente de los datos: Dialnet