Automatic Detection and Boundary Extraction of Lunar Craters Based on LOLA DEM Data

Impact-induced circular structures, known as craters, are the most obvious geographic and geomorphic features on the Moon. The studies of lunar carters’ patterns and spatial distributions play an important role in understanding geologic processes of the Moon. In this paper, we proposed a method based on digital elevation model (DEM) data from lunar orbiter laser altimeter to detect the lunar craters automatically. Firstly, the DEM data of study areas are converted to a series of spatial fields having different scales, in which all overlapping depressions are detected in order (larger depressions first, then the smaller ones). Then, every depression’s true boundary is calculated by Fourier expansion and shape parameters are computed. Finally, we recognize the craters from training sets manually and build a binary decision tree to automatically classify the identified depressions into craters and non-craters. In addition, our crater-detection method can provides a fast and reliable evaluation of ages of lunar geologic units, which is of great significance in lunar stratigraphy studies as well as global geologic mapping.