A Comparative Study Between Neural Network and Maximum Likelihood in the Satellite Image Classification

Antonio Gabriel Rodrigues, Rossana Baptista Queiroz, Arthur Tórgo Gómez

Abstract

In this paper it's showed a comparative study between two techniques of satellite image classification. The studied techniques are the Maximum Likelihood statistical method and an Artificial Intelligence technique based in Neural Networks. The analyzed images were scanned by CBERS 1 satellite and supplied by Brazilian National Institute for Space Research (INPE). These images refer to Province of Rondonia area and were obtained by CBERS 1 IR-MSS sensor.

References

  1. Bittencourt, H. R., 2001. Reconhecimento estatístico de padrões: o caso da discriminação logística aplicada a classificação de imagens Digitais obtidas por Sensores remotos. In Congresso Brasileiro de Computação - CBComp 2001.
  2. Braga, A. de P, Ludemir, T.B., Carvalho, A.C.P.de L.F., 2000. Redes Neurais Artificiais Teoria e Aplicações. Livros Técnicos e Científicos Editora.
  3. Erbert, M., 2001. Introdução ao Sensoriamento Remoto. Master Tesis, Universidade Federal do Rio Grande do Sul.
  4. Fahlman, S. E., Lebiere, C., 1990. The Cascade-Correlation Learning Architecture. School of Computer Science Carnegie Mellon University, Pittsburgh.
  5. Haykin, S., 1999. Redes Neurais - Princípios e Prática. Bookman, 2nd edition.
  6. INPE, 2002.Tutorial SPRING. INPE, São José dos Campos.
  7. Osório, F. S, Amy, B, 1999. A hibrid System for Constructive machine learning. Neurocomputing, v. 28.
  8. Rennó, C. D., 1998. Avaliação das Incertezas nas Classificações de Máxima Verossimilhança e Contextual de Modas Condicionais Iterativas em Imagens Jers na Região de Tapajós, Estado do Pará. INPE, São José dos Campos.
  9. Rodrigues, A. G., Queiroz, Rossana B., Gómez, A. T., 2003. Estudo Comparativo entre os Métodos de Classificação de Imagens de Satélite Máxima Verossimilhança Gaussiana e Redes Neurais. In Congresso Regional de Iniciação Científica e Tecnológica em Engenharia - CRICTE 2003.
  10. Todt, V. D., Formaggio, A. R., Shimabukuro, Y., 2003. Identificação de áreas Desflorestadas na Amazônia através de uma Rede Neural Artificial Utilizanto Imagens Fração Derivadas dos Dados do IR-MSS/CBERS. In XI Simpósio Brasileiro de SensoriamentoRemoto.
Download


Paper Citation


in Harvard Style

Gabriel Rodrigues A., Baptista Queiroz R. and Tórgo Gómez A. (2004). A Comparative Study Between Neural Network and Maximum Likelihood in the Satellite Image Classification . In Proceedings of the First International Workshop on Artificial Neural Networks: Data Preparation Techniques and Application Development - Volume 1: ANNs, (ICINCO 2004) ISBN 972-8865-14-7, pages 1-8. DOI: 10.5220/0001130100010008


in Bibtex Style

@conference{anns04,
author={Antonio Gabriel Rodrigues and Rossana Baptista Queiroz and Arthur Tórgo Gómez},
title={A Comparative Study Between Neural Network and Maximum Likelihood in the Satellite Image Classification},
booktitle={Proceedings of the First International Workshop on Artificial Neural Networks: Data Preparation Techniques and Application Development - Volume 1: ANNs, (ICINCO 2004)},
year={2004},
pages={1-8},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001130100010008},
isbn={972-8865-14-7},
}


in EndNote Style

TY - CONF
JO - Proceedings of the First International Workshop on Artificial Neural Networks: Data Preparation Techniques and Application Development - Volume 1: ANNs, (ICINCO 2004)
TI - A Comparative Study Between Neural Network and Maximum Likelihood in the Satellite Image Classification
SN - 972-8865-14-7
AU - Gabriel Rodrigues A.
AU - Baptista Queiroz R.
AU - Tórgo Gómez A.
PY - 2004
SP - 1
EP - 8
DO - 10.5220/0001130100010008