Multi-Sensor Data Compression

ABSTRACT: The compression of hyperspectral data is made more difficult by the requirement that the compressed data adequately serve a variety of diverse applications. Lossy compression methods will be evaluated with the requirements for tactical HSI analysis in mind. We propose to investigate two compression approaches, one based on sampling and reconstruction and the other on a fractal description. To determine their potential applicability to HSI, hyperspectral target detection and terrain classification algorithms will be used. BENEFIT: The development of a compression standard for tactical hyperspectral systems, will greatly increase the utility of the data by the military community. In addition, the development of compression techniques will have broad applicability for space-based and airborne commercial instruments.