Distributed Lossless Coding Techniques for Hyperspectral Images

In this paper, we present a novel distributed coding scheme for lossless, progressive and low complexity compression of hyperspectral images. Hyperspectral images have several unique requirements that are vastly different from consumer images. Among them, lossless compression, progressive transmission, and low complexity onboard processing are three most prominent ones. To satisfy these requirements, we design a distributed coding scheme that shifts the complexity of data decorrelation to the decoder side to achieve lightweight onboard processing after image acquisition. At the encoder, the images are subsampled in order to facilitate successive encoding and progressive transmission. At the decoder, we generate the side information with adaptive region-based predictor by taking full advantage of the decoded subsampled images and previously decoded neighboring bands based on the assumptions that the objects appearing in different bands are highly correlated. The proposed progressive transmission via subsampling enables the spectral correlation to be refined successively, resulting in gradually improved decoding performance of higher-resolution layers as more sub-images are decoded. Experimental results on the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data demonstrate that the proposed scheme is able to achieve competitive compression performance comparing with the-state-of-the-art 3D schemes, including existing distributed source coding (DSC) schemes. The proposed scheme has even lower encoding complexity than that of the conventional 2D schemes.