Conference on Intelligent Data Understanding 2010 (CIDU)

LUNAR TERRAIN AND ALBEDO RECONSTRUCTION FROM APOLLO
IMAGERY

ARA V NEFIAN, TAEMIN KIM, MICHAEL BROXTON, AND ZACH MORATTO*

Abstract. Generating accurate three dimensional planetary models and albedo maps is becoming
increasingly more important as NASA plans more robotics missions to the Moon in the coming
years. This paper describes a novel approach for separation of topography and albedo maps
from orbital Lunar images. Our method uses an optimal Bayesian correlator to refine the stereo
disparity map and generate a set of accurate digital elevation models (DEM). The albedo maps
are obtained using a multi-image formation model that relies on the derived DEMs and the Lunar-
Lambert reflectance model. The method is demonstrated on a set of high resolution scanned
images from the Apollo era missions.