Analysis of Multispectral Microscope Images Using Neural Networks

Abstract

In this study, neural networks have been used to classify the clusters in a scatter diagram formed from images in a multispectral set. The use of multispectral images in surface and materials science has been applied to a number of techniques in recent years. The use of neural networks to analyse multispectral Landsat images of the Earth has already been reported, but has not yet been applied in other fields employing multispectral images. The work reported here used two different neural networks : a three-layer neural net with back propagation and a self-organizing map. The results suggest that neural nets can provide a useful addition to the current multispectral analysis procedures in materials and surface science.