Study of real-time lossless data compression for hyperspectral imagery

This paper describes a study of real-time lossless data compression of hyperspectral imagery using prediction and entropy encoding. The main effort in developing a compression system, is to study predictors that can yield the best reduction of entropy and can be easily implemented in real-time. The Consultative Committee for Space Data System (CCSDS) recommended lossless algorithm is selected as the entropy encoder. Four predictor schemes have been selected for study. Three typical hyperspectral data sets acquired by the Airborne Visible/Infrared imaging Spectrometer (AVIRIS) and three acquired by the Compact Airborne Spectrographic Imager (casi) were used as test data. A lossless compression system with different predictors has been simulated and tested with the test data.