Date of Publication :7th June 2016

Abstract: As the dimensionality of remotely sensed Hyper spectral images are increasing, compression is required to transmit and archive Hyper spectral data. In this paper, an adept method for Hyper spectral image compression is presented to effectively reduce the volume of Hyper spectral data. The objective of proposed method is to apply classical compression method where low frequency information is reserved and high frequency information is discarded, based on Discrete Wavelet Transform. The core idea of our proposed method is to apply tucker decomposition on the wavelet coefficient, exploit both the spectral and the spectral information in the images. Moreover, it also evaluates the compression ratios for biorthogonal wavelet family The obtained result shows better performance on Bior1.1 and Bior1.3 wavelets.

Reference :