Satellite Image Classification using Ant Colony Optimization and Neural Network

Abstract

From the last three decades, remote sensing has come up with great applications in the field of science and technology. The concept of remote sensing is the observation of earth using data acquired instruments satellites and aircrafts from the outer space. Remote sensing helps to monitor the environment states, natural resources availability and terrain features. The unique capabilities of remote sensing concept regarding earth observations are that it helps to monitor, forecast, understand and manage the available resources of earth. Here, remote sensing data used for the observations of land cover terrain features with the help of image classification process. It helps to obtain the geo spatial from satellite data that can be used in several applications of computing, research, space intelligence, defense etc. In this research work, we are using this image classification for the identification of land cover terrain features from the satellite data of Alwar region, India. Concept of ant colony optimization and neural network has been used for the classification. Ant colony is swarm intelligence based global optimization concept. The output from the ACO is used for the further optimization with neural network approach. Results are evaluated in terms of Overall accuracy and kappa coefficient. Results obtained using proposed integrated approach are efficient to declare the validate classification of image.