Lossless Compression of Hyperspectral Images Using Hybrid Based Clustering DPCM
G. Mahalakshmi1, A. Sivasankar2
1G. Mahalakshmi, PG Scholar, VLSI Design, Anna University Of Technology Madurai, Madurai, India.
2A. Sivasankar, Assistant Professor, ECE, Anna University Of Technology Madurai, Madurai, India.
Manuscript received on June 05, 2013. | Revised Manuscript received on June 11, 2013. | Manuscript published on June 15, 2013. | PP: 32-34 | Volume-1 Issue-7, June 2013. | Retrieval Number: G0339061713
Open Access | Ethics and Policies | Cite
© The Authors. Published By: 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: This project explores the use of hybrid clustering technique for lossless compression method for Hyperspectral images. It is based on the joint use of fuzzy c means and nearest neighbor algorithms. In this method, linear prediction is performed using coefficients optimized for each spectral cluster separately. The difference between the prediction and original values is entropy coded using an adaptive range coder for each cluster. The result shows that this method has lower bit-per-pixel value. It is an extension to the existing lossless compression algorithm. Better partitioning of data is achieved. The technique starts with the fuzzy c means algorithm, performed as the first stage for an adequately high number of centroids and continues with the nearest neighbour algorithm executed for the clusters obtained in the first stage, as the set of initial objects to be merged for relatively complex shapes.
Keywords: Hyperspectral images, image compression, lossless compression.