InSAR Simulation and Speeded-Up Robust Features Algorithm for Terrain Relative Navigation in PSRs on the Moon

Landing in the permanently shadowed regions (PSRs) on the Moon requires high-resolution topographic information and accurate navigation. Owing to low Sun elevation angles, there is no direct solar illumination in PSR, making it difficult to acquire high-resolution optical images for terrain relative navigation (TRN). Synthetic aperture radar (SAR) onboard lunar orbiter can acquire the high-resolution digital elevation model (DEM) of PSR with the interference phases from repeat-passes, or alternatively, from multiantenna observations in a single orbit pass. In this article, SAR images from dual-antenna observations obtained in single orbit passes are simulated with two-scale model and Range-Doppler algorithm for the interference phases based on the DEM data from the lunar orbiter laser altimeter (LOLA). Hence, we generate DEMs of PSR in two prominent lunar south polar craters, Shoemaker and Shackleton. After geometric correction, the influence of radar parallax in DEM data are removed. The generated DEM data are used to illustrate the possibility of TRN in PSR with the image-matching algorithm. The slope angle image of the PSR from the generated DEM is taken as the reference image for navigation, while high-resolution slope angle image from LOLA DEM data is taken as the real-time image from the flyer. The speeded-up robust features algorithm matches the feature points in the reference image and real-time image. The location of the matched points determines the position and motion vector of the flyer. The simulation proves the DEM data from InSAR can provide detailed topographic information and can be used for navigation in regions of permanent shadows.