3D Wavelet Block Tree Coding for Hyperspectral Images
Shrish Bajpai1, Naimur Rahman Kidwai2, Harsh Vikram Singh3
1Shrish Bajpai, Doctoral Student, Department of Electronic Engineering, A. P. J. Abdul Kalam Technical University, Lucknow, India.
2Naimur Rahman Kidwai, Professor, Department of Electronics and Communications Engineering, Ahangirabad Educational Trust Group Of Institutions, Barabanki, India.
3Harsh Vikram Singh, Associate Professor, Department of Electronic Engineering, Kamla Nehru Institute of Technology, Sultanpur, India.
Manuscript received on 03 April 2019 | Revised Manuscript received on 10 April 2019 | Manuscript Published on 13 April 2019 | PP: 64-68 | Volume-8 Issue-6C April 2019 | Retrieval Number: F12200486C19/19©BEIESP
Open Access | Editorial and Publishing 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: A novel hyperspectral image compression scheme based on set partitioned compression scheme is proposed. This compression scheme uses the 3D wavelet transform to exploits the both, inter sub-band & intra sub-band correlation, among the wavelet coefficients of transformed hyperspectral images. The compression scheme is based on the spatial oriented trees (SOT) which is the basic unit in block. Block is in cube shape having the coefficients m*m*m in contrast to a single coefficient in three dimension set partitioning in hierarchical trees compression scheme. Each SOT has a root node in LLL band with the child and descendent blocks in high frequency sub-band. So, proposed wavelet based compression scheme uses the features of both zeroblock & zerotree base compression schemes.
Keywords: Compression Schemes, Hyperspectral Imaging, Performance Comparison, Set Partition Compression, Wavelet Transform.
Scope of the Article: Communications