W.M. Elmahboub and E. Yankey (USA)
Remote sensing of planetary surface, spectroscopy, and mathematical modeling
Investigation of Mars digital data through image processing and spectral analysis has been an area of growing concern in the scientific community. Classification relies on inherent spectral reflectance properties of Mars data. Image data are first clustered into natural spectral groupings present in the scene. Algorithm in C++ has been applied to Hubble Telescope digital data to degrade the effect of dust particles noise and to reduce the inaccuracy of image classification using trial and error procedure and incorporating ERDAS IMAGINE (commercial software). Spectral reflectance of Mars digital data are compared to the mineral laboratory spectroscopy. The degradation of noise shows improvement in the spectral analysis of Mars data. The spectral analysis shows presence of different minerals and abundance of iron oxide leading to the conclusion that the algorithm is reliable in minerals spectral identification.
Important Links:
Go Back
