Lunar Surface Reconstruction from Apollo MC Images

Abstract

The last three Apollo lunar missions (15, 16, and 17) carried an integrated photogrammetric mapping system of a metric camera (MC), a high-resolution panoramic camera, a star camera, and a laser altimeter. Recently images taken by the MC were scanned by Arizona State University (ASU); these images contain valuable information for scientific exploration, engineering analysis, and visualization of the Moon’s surface. In this article, we took advantage of the large overlaps, the multi viewing, and the high ground resolution of the images taken by the Apollo MC in generating an accurate and reliable surface of the Moon. We started by computing the relative positions and orientations of the exposure stations through a rigorous photogrammetric bundle adjustment process. We then generated a surface model using a hierarchical correlation-based matching algorithm. The matching algorithm was implemented in a multi-photo scheme and permits the exclusion of obscured pixels. The generated surface model was registered with LOLA topographic data and the comparison between the two surfaces yielded an average absolute difference of 36 m. These results look very promising and demonstrate the effectiveness of the proposed algorithm in accounting for depth discontinuities, occlusions, and image-signal noise.