A low complexity lossless data compressor IP-core for satellite images

Abstract

As the technology advances, space imaging systems use equipment of increasing resolutions. Hence, it is necessary to ensure that this great quantity of data arrives at its destination reliably. Among some variables involved, data compression plays an important role to accomplish this requirement. In this context, this paper proposes a digital hardware approach of a low complexity satellite image lossless compressor based on prediction and Golomb-Rice coding, which has achieved excellent results considering hardware and compression performance. In order to validate and analyze the compressor, a functional verification and FPGA prototyping methodology were followed. Given an image set from Brazilian's National Institute for Space Research (INPE, in Portuguese acronyms), its results on FPGA show that this compressor achieves compression ratio around 3.4, comparable value to related works in this area, and throughput of 28 MPixel/s (224 Mbit/s). Taking advantage of images nature, its compression can be parallelized through simultaneous multi-cores compressors. For example, using 5 cores, this work is able to compress those images in a rate of 142 MPixel/s (1.1 Gbit/s). All these features make it useful and effective as a current remote sensing imaging system.