Exploration of Unmixing and Classification of Hyperspectral Imagery
Nagarajan Munusamy1, Rashmi. P. Karchi2

1Nagarajan Munusamy, Department of Multimedia and Web Technology, KSG College of Arts and Science, Coimbatore (Tamilnadu), India.
2Rashmi. P.Karchi, Department of Computer Science, Bharathiar University, Coimbatore (Tamilnadu), India.
Manuscript received on 05 May 2019 | Revised Manuscript received on 12 May 2019 | Manuscript published on 30 May 2019 | PP: 723-733 | Volume-8 Issue-7, May 2019 | Retrieval Number: G5605058719/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Hyperspectral imaging is the vital method and an effective tool to quantify as well identify dissimilar objects from remotely recognized spectral information. Using OMEGA instrument, the Mars Region is imaged using an unprecedented spatial and spectral combination of resolution spectrometer. Hyperspectral images provide high resolution, and its spectral range gives the ability to identify chemical mixture in the atmosphere of Mars more precisely than before. Due to the inadequate spatial resolution of Hyperspectral sensors mixed pixel arises. Such mixed pixels contain more than one distinct material, which is called endmembers. These hyperspectral images provide good resolution, and the range of spectra will give the ability to identify the chemical species present in the atmosphere of Mars more accurately than before. The proposed methodology is evaluated on the real hyperspectral datasets. The integration of Unmixing algorithm termed “Non-Linear Hybrid Approach for Regularized Simultaneous Forward-Backward Greedy Algorithm (NonLHA-RSFBGA)” with the Singular Spectrum Analysis approach, resulting in a better level of classification using the ART classifier for the identification/classification of the Mineral endmember. The results of the proposed method for the classification of the endmembers in hyperspectral imagery is promising.
Keyword: Endmember, Hyperspectral image classification, Mixed pixel, Unmixing. 
Scope of the Article: Classification