A new approach to pattern recognition of remote sensing image using artificial neural network

The performance is analyzed of Multilayered Neural Networks (MLNs) and its training rule-back propagation (BP) algorithm. Some new methods are used to improve its training speed and to avoid local minima. Spatial information is introduced into pattern recognition of remote sensing image using MLNs. The authors use three-layer modified MLNs as classifiers in pattern recognition of remote sensing images (Landsat-4 TM Data). The experimental results verify the usefulness of the approach.<<ETX>>