Automatic Segmentation of Satellite Image using Self Organizing Feature Map (SOFM)An Artificial Neural Network (ANN) Approach

Abstract

Satellite Image Processing is one of the key areasof computer science which is used for various important applications of remote sensing and Geographic Information System (GIS). Image segmentation is one of the key research areas in the field of image processing and analysis. Image segmentation is needed for various intermediate as well as final processes for various analyses. There are various algorithms available for image segmentation. Self Organizing Feature Map (SOFM), an Artificial Neural Network, is one of such algorithms. Based on the dataset, SOFM performs the training of the neural network and then segment the input satellite image into various homogenous clusters. Later the similar regions are merged and results are refined. This paper presents detailed working of SOFM together with the experiment conducted for satellite image segmentation.

Keywords:Artificial Neural Network, Self Organizing Feature Map (SOFM), Image Processing, Segmentation, Clustering