Pattern Recognition in Multispectral Satellite Images Using Concurrent Self-Organizing Modular Neural Networks

Abstract

We investigate multispectral space image classification using the new artificial computational intelligence model called Concurrent Self-Organizing Maps (CSOM), representing a winner-takes-all collection of concurrent small modular self-organizing artificial neural networks. For comparison, we evaluate the performances of Bayes classifier. The implemented neural/statistical classifiers are evaluated using a LANDSAT TM image with 7 bands (multi-sensor data fusion) composed by a set of 7-dimensional pixels, out of which a subset contains labelled pixels, corresponding to seven thematic categories of Earth images taken from space. The best experimental result leads to the recognition rate of 95.29 %. The model has defence applications for Earth surveillance from space.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2006
Accession Number
ADA479800

Entities

People

  • Armand-dragos Ropot
  • Victor-emil Neagoe

Tags

DTIC Thesaurus Topics

  • Abstracts
  • Artificial Satellites
  • Classification
  • Detectors
  • Image Classification
  • Information Operations
  • Information Systems
  • Instructions
  • Machine Learning
  • Military Operations
  • Multispectral
  • Neural Networks
  • Pattern Recognition
  • Recognition
  • Satellite Imaging
  • Standards
  • Training

Fields of Study

  • Computer science

Readers

  • Atmospheric Remote Sensing.
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks
  • Space
  • Space - Space Objects