Title: On combining vision-based hybrid classifiers for weeds detection in precision agriculture

Authors: Alberto Tellaeche Iglesias, Maria Guijarro, Gonzalo Pajares Martinsanz, Xavier-P. Burgos Artizzu, Angela Ribeiro Seijas

Addresses: Department Informatica y Automatica, E.T.S. Informatica – UNED, 28040 Madrid, Spain. ' Centro Superior de Estudios Felipe II, Ingenieria Tecnica de Informatica de Sistemas, 28300 Aranjuez, Spain. ' Department Ingenieria del Software e Inteligencia Artificial, Facultad Informatica, Universidad Complutense, 28040 Madrid, Spain. ' Instituto de Automatica Industrial, CSIC, Arganda del Rey, Madrid, Spain. ' Instituto de Automatica Industrial, CSIC, Arganda del Rey, Madrid, Spain

Abstract: One objective in precision agriculture is to minimise the volume of herbicides that are applied to the fields through the use of site-specific weed management systems. This paper outlines an automatic computer vision system for the detection and differential spraying of Avena sterilis, a noxious weed growing in cereal crops. With such purpose we have designed a multiple hybrid decision making system based on four different simple classifiers: Bayes, fuzzy k-means (FkM), support vector machines (SVM) and Hebbian learning. The performance of this multiple classifier is compared against other approaches, including simple versions of hybrid classifiers.

Keywords: machine vision; simple classifiers; Bayes; fuzzy clustering; FkM; support vector machines; SVM; Hebbian learning; hybrid classifiers; precision agriculture; weed detection; weed management; cereal crops; crop spraying; multiple hybrid decision making; weeds; classification.

DOI: 10.1504/IJRIS.2010.034905

International Journal of Reasoning-based Intelligent Systems, 2010 Vol.2 No.2, pp.100 - 109

Published online: 30 Aug 2010 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article