FPGA-based hyperspectral data compression using spectral unmixing and the pixel purity index algorithm

Abstract

Hyperspectral data compression is expected to play a crucial role in remote sensing applications. Most available approaches have largely overlooked the impact of mixed pixels and subpixel targets, which can be accurately modeled and uncovered by resorting to the wealth of spectral information provided by hyperspectral image data. In this paper, we develop an FPGA-based data compression technique based on the concept of spectral unmixing. It has been implemented on a Xilinx Virtex-II FPGA formed by several millions of gates, and with high computational power and compact size, which make this reconfigurable device very appealing for onboard, real-time data processing. © Springer-Verlag Berlin Heidelberg 2006.