Boundary Detection of Dispersal Impact Craters Based on Morphological Characteristics Using Lunar Digital Elevation Model

Lunar impact craters are the most typical geomorphic unit found on the moon and are of great significance for the study of terrain features. Currently, mostmainstream detection algorithms for lunar impact craters represent craters as circles, but these methods counter difficulty in expressing the real shapes in an accurate manner. It is difficult to conduct an advanced analysis, such as characteristic and spatial heterogeneity analyses of the lunar surface. The lunar terrain characteristics were analyzed using dispersal impact craters as the object, and a dispersal crater detection algorithm (CDA) based on a digital elevation model (DEMbased CDA) was proposed. By analyzing the surface catchment, the impact crater can be treated as a closed basin structure. The raw impact crater region is detected using watershed algorithm. This algorithm can detect the real boundaries of an impact crater based on the actual terrain. Several morphological factors, such as posture ratio and rectangle factor, are used to filter the raw impact crater regions. A suborbicular raw impact crater region is treated as the desired result after the elimination of false impact craters. For the remaining raw impact crater regions, terrain profile analysis is applied to accurately identify the boundaries of impact craters. The DEM data produced from the Chang'E-2 imagery and LRO laser altimeter data (20 and 60 m) are used in verification. The results show that the impact crater boundaries with real terrain characteristics can be effectively expressed and used in further exploration.