Phobos Image Enhancement Using Unpaired Multi-Frame Acquisitions From Indian Mars Color Camera

The Mars Color Camera (MCC) of Indian Space Research Organisation (ISRO) captures the elusive moon of Mars, Phobos. MCC instrument is the eye of Mars Orbiter Mission (MOM) launched by ISRO on 05th November 2013 from the spaceport of India, Sriharikota. MCC has already provided more than 1000 images of Martian surface features covering different terrains. Recently, MCC has captured Phobos image at successive time instances on 01st July 2020. Multiple unpaired acquisitions of planetary remote sensing images often contain rich information to super resolve the spatial details of these images. This paper describes the techniques developed to enhance the Phobos image from MCC multi-frame acquisitions using image rectification and topographic data. In this regard, the co-registration of MCC frames is a crucial step. To co-register the data at sub-pixel level, we present a method, namely Random Sample Consensus (RANSAC) pruning based Scale Invariant Feature Transform (R-SIFT). The image composite is based on Medoid approach. It is further enhanced by contrast limited adaptive histogram equalization (CLAHE). The location of Phobos is computed using Spacecraft Planet Instrument Camera Matrix Event (SPICE) toolkit. For topographic correction, we rely on the Digital Elevation Model (DEM) of Phobos using publicly available resources. Visual inspection and image quality parameters, such as BRISQUE, RMSE, PSNR, and SSIM are estimated for both enhanced and topographically corrected image for assimilation purpose. After incorporating these techniques, the final Phobos image appears more representative, spatially enhanced, and has normalized radiometry to study its surface features.