As part of NASA's Game Changing Development LuNaMaps Project, our group is involved in producing shape models of the lunar surface to support NASA science and exploration activities at the south pole and beyond. This includes the need for medium resolution terrain models to support optical navigation along the approach trajectory towards the destination near the Lunar South Pole. We used a technique known as Shape-from-Shading (SfS), implemented within the available Ames Stereo Pipeline (ASP) software package (Alexandrov & Beyer 2018). We processed images of the lunar surface taken by the LROC camera to refine existing altimeter-based topographic maps into more accurate terrain maps for a potential approach path. Because it is a computationally demanding task, SfS has thus far mostly been used to retrieve high-resolution (0.5-2 meters/pixel) terrain models of sites with a limited area [Alexandrov 2018]. Our application is in this sense very unusual: we target medium-resolution (~30 m/pix) terrain models over a 18,000 km2 strip between 60-80S. We developed a sophisticated semi-automated pipeline embedding both ASP and ISIS image processing tools, as well as relevant python libraries.