Spectral Reconstruction of Hyperion Data Based on FLAASH Model

A method of spectral reconstruction based on FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hyper cubes) is introduced in this paper. A scene of Hyper ion image is adopted with hyper spectral data in October 14th, 2008 in the Mountainous regions of Guizhou. The FLAASH model is adopted spectral reconstruction five typical objects. Atmospheric water vapor absorption band are removed (1346nm-1467nm and 1800nm-1971nm). The results show that, Vegetation reflectance in the visible band has a significant reflectance peak in green band, and have high reflection in the near infrared region due to multiple scattering effects. Water in the entire spectral range has a low reflection. Experiment and reasonable result could be achieved to improve performance of Hyperion.