Hyperspectral Analysis Techniques for Remote Sensing

Imaging spectrometers (hyperspectral imaging sensors) and various airborne and spaceborne multispectral scanners have been used for the remote detection of mineral in the soil and for the study of vegetation. New applications are developing for hyperspectral sensors in the area of environmental monitoring. The combination of imaging with spectroscopy allows their data to be tied to maps and geographical information systems and, by use of global positioning system (GPS), report in real-time on the exact location and extent of events of interest. This capability is of particular importance in emergency response, toxic cloud tracking, oil spill detection, and more. OKSI is currently developing the first staring hyperspectral imaging sensor (TIRIS) that operates in passive mode in the long wave infrared portion of the spectrum for use in such applications. The development of chemometric techniques that are geared for operations with such a sensor will open the road for major advances in remote sensing and chemical detection. Under Phase I OKSI will develop novel algorithmic paradigms that are based on deterministic transform-domain analysis of spectral signatures, and utilize spatial information to obtain signal statistics. These techniques were originally developed for target and countermeasure discrimination and their application to chemical detection will be explored.