Implementation of Ga Based Fodpso for Efficient Segmentation of Hyper Spectral Satellite Images

Abstract

Hyperspectral satellite images have a high trough of Geospatial information. A great number of traditional segmentation algorithms have been implemented for segmentation of hyperspectral satellite images. But there exist the problems of under or over segmentation which affects the data retrieval process. The wireless hyper-spectral images can identify minerals better than multispectral images because of their high spectral resolution.  However, a pixel might include more than one mineral, as wireless hyper-spectral images have low spatial resolution.  In these situations, the number of minerals can be estimated in mixed pixels but their spatial position cannot be known.  This is one of the biggest obstacles that prevent effective use of wireless hyper-spectral images in mineral exploration.  Hence it necessitates the exploration of some hybrid methodology for the extraction the information’s from the hyperspectral images. In this research work we have Proposed a, GA based FODPSO for high-resolution image processing which leads to an efficient segmentation. The main aim of this work is to propose a computationally intelligent and efficient method, for partitioning remote sensing images into multiple regions. After the separation of images, intelligent data retrieval process can be implemented to get the required information from the remote sensing satellite images. However, a pixel might include more than one mineral, as hyper-spectral images have low spatial resolution. So we have to apply the principle to convert from low resolution image to high resolution using interpolation technique then do segmentation.