A hardware-friendly algorithm for compressing hyperspectral images

Abstract

The on-board compression of remote sensed hyperspectral images is an important task nowadays. One of the main difficulties is that the compression of these images must be performed in the satellite which carries the hyperspectral sensor, where the available power, time, and computational resources are limited. Moreover, it is important to achieve high compression ratios without compromising the quality of the decompressed image for the ulterior hyperspectral imaging applications. The HyperLCA compressor aims to fulfill these requirements, providing an efficient lossy compression process that allows achieving very high compression ratios while preserving the most relevant information for the subsequent hyperspectral applications. One extra advantage of the HyperLCA compressor is that it allows to fix the compression ratio to be achieved. In this work, the effect of the specified compression ratio in the computational burden of the compressor has been evaluated, also considering the rest of the input parameters and con figurations of the HyperLCA compressor. The obtained results verify that the computational cost of the HyperLCA compressor decreases for higher compression ratios, with independence of the specified con figuration. Additionally, the obtained results also suggest that this compressor could produce real-time compression results for on-board applications.