Fusion of THEMIS and TES for Accurate Mars Surface Characterization

In a recent NASA ROSES solicitation, NASA has expressed strong interest in improving surface characterization of Mars using orbital imagers. Thermal Emission Imaging System (THEMIS) and Thermal Emission Spectrometer (TES) are orbital multispectral imagers of Mars. THEMIS has 10 spectral bands in the 6-13 micrometers region and a spatial resolution of 100 m. TES has 143 spectral bands in the 5-50 micrometers range, but with low spatial resolution of 3x3 km. Although both have been used to map out the surface composition of Mars, there are some limitations. First, THEMIS has low spectral resolution that may not provide accurate surface characterization. Second, TES has low spatial resolution that cannot provide fine spatial details of surface characteristics. Roughly speaking, each TES pixel contains about 900 THEMIS pixels. It is therefore very challenging to fuse the two data sets. We propose a novel and accurate framework that can deal with the above challenge. The framework is based on the latest development in image registration, image fusion, anomaly detection, pixel classification using hyperspectral images, and concentration estimation. First, a two-step image registration algorithm is proposed to align the THEMIS and TES bands. Subpixel accuracy can be achieved. Second, a novel image fusion algorithm is proposed to fuse THEMIS and TES bands to generate an image cube with 143 bands of 100-m resolution images. Our algorithm has been proven to be better than state-of-the-art fusion algorithms. Third, we propose a novel anomaly detection algorithm for detecting interesting regions in the fused image cube. This algorithm is known as cluster kernel Reed Xiaoli (CKRX) and has high performance in anomaly detection. Fourth, a sparsity based approach is proposed to perform accurate rock classification. Finally, a deep learning based algorithm is proposed to estimate the chemical composition of the rocks for better surface characterization.