A STUDY ON HYPERSPECTRAL DATA IMAGING BASED ON SPATIO SPECTRAL SCANNING

Abstract

Hyperspectral data is one of the important breakthroughs in remote sensing. It has a capability of separating the different and describing the objects of same size in detail. It processes the image in the large- number, narrow and it can produce data with sufficient resolution. Hyperspectral images used to identify the spectral information and differentiate unique materials. It also provides the information by the means of very accurate and extraction of various information that relates to remote based information. Hyperspectral remote sensing purpose is to measure the various components of the Earth system that is acquired as images for scientific applications and research purpose. Hyperspectral remote sensing can be categorized the information into two ways: Feature space and Spectral space. The feature space method is not efficient because the data dimension is determined to describe the patterns. The spectral based is used for extracting anomalous pixel vector at endmember extraction. This will validate the both synthetic and real Hyperspectral data images. The scatter matrix that is preserved in the means of pixel by spatial domain and the projection of optimal discriminative is obtained with the help of scattering of spectro-spatial thus maximizing a modified scatter of information.