LOSSLESS COMPRESSION OF AVIRIS IMAGES

Adaptive DPCM methods using linear prediction are described for the lo ssless compression of hyperspectral (224-band) images recorded by the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). The methods h ave two stages-predictive decorrelation (which produces residuals) and residual encoding. Good predictors are described, whose performance c losely approaches limits imposed by sensor noise. It is imperative tha t these predictors make use of the high spectral correlations between bands. The residuals are encoded using variable-length coding (VLC) me thods, and compression is improved by using eight codebooks whose desi gn depends on the sensor's noise characteristics. Rice coding has also been evaluated; it loses 0.02-0.05 b/pixel compression compared with better VLC methods but is much simpler acid faster. Results for compre ssing ten AVIRIS images are reported.