Biogeography Based Band Selection for Hyperspectral Image Classification

Hyperspectral image classification is a widely researched topic in the field of remote sensing. In this context, dimensionality reduction plays an important role as it can affect the efficiency of the learning algorithm and also reduce the redundancy. In this article, a different look on the problem has been explored where application of biogeography based optimization is used for efficient dimensionality reduction in hyperspectral images. After that SVM classification technique is used to categorize the image pixels. Establishing the betterment of the methods, qualitative and quantitative analysis of the methods have been made, which shows a promising result than other existing methods.