Low−complexity Lossless Compression of Hyper− and Ultra−spectral Images Based on Slepian−wolf Coding
چکیده
Distributed source coding makes it possible to develop compression algorithms with a low−complexity encoder, while most of the signal modeling is moved to the decoder. This structure is an excellent match to the remote sensing scenario, in which the on−board processing units have limited computational capabilities. Moreover, remote sensing images do not exhibit motion, which greatly simplifies the design of a distributed coder. In this paper we propose a new algorithm for lossless compression of remote sensing images, based on distributed source coding. The objective of this algorithm is to achieve very low−complexity encoding, with performance as close as possible to a full−complexity coder. The complexity reduction is obtained by coding each spectral channel separately, whereas high coding efficiency is achieved through joint decoding. [continued on the next page]