A new method for atmospheric correction and de-noising of CRISM hyperspectral data

We propose a new method to perform atmospheric correction and de-noising on hyperspectral image cubes acquired by the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) board NASA’s Orbiter (MRO). The CRISM imager has had an important role in advancing our understanding of many aspects Martian mineralogy. Many mineral detections from data have been facilitated significant efforts development processing pipeline retrieve surface reflectance. However, some residuals remain spectra after correction, causing difficulty interpretation processed reflectance spectra. In addition, images are occasionally corrupted with high noise levels exhibiting heterogeneous statistical properties. This paper identifies cause such spectral distortions describe technique that simultaneously performs both each cube individually. particular, focuses 1.0–2.6 ?m wavelength region is applicable non-icy surfaces. Experimental results show able significantly mitigate various sources like gaseous absorptions, detector temperature, water ice aerosols, compared official called volcano scan variety scenes. Careful validations include qualitative examination artifacts ratioed non-ratioed comparison using multiple overlapping strengthen confidence approach.