Removal of salt-and-pepper noise in THEMIS infrared radiance and emissivity spectral data of the martian surface

The quality of false color images of THEMIS multispectral infrared data is significantly reduced by the presence of white noise. Presented here is the description of the algorithm developed for the removal of white noise from THEMIS data to produce higher quality false color imagery in radiance and emissivity without loss of spatial resolution. The algorithm utilizes inter-band correlations of surface features and principal component analysis techniques to separate noise from the radiation originating from the planet surface. This algorithm is a standard process available to users of THEMIS data for the production of false color imagery.