AISRP workshop '97 presentation,    E. Merényi, U of Arizona, LPL

1. The Challenge

Spectral imagers in space missions produce data with

• unprecedented volume
• unprecedented richness of scientific information note 1.1

We need to sift through data collected by orbiting satellites in (near-)real time, AND exploit the data properly through sufficiently sensitive discrimination of many types of intricate spectral signatures

• to trigger action on discovery (by classification of spectral signatures)
• to use limited downlink capacity to transmit extracted relevant scientific knowledge (e.g. thematic maps of surface covers) rather than only a small percentage of raw data

Even on Earth, with no hard time limits, hyperspectral image exploitation poses a great challenge because of the high data dimensionality of the spectra, for mathematical and technical reasons. note 1.2