Contour-based automatic crater recognition using digital elevation models from Chang'E missions

In order to provide fundamental information for exploration and related scientific research on the Moon and other planets, we propose a new automatic method to recognize craters on the lunar surface based on contour data extracted from a digital elevation model (DEM). Through DEM and image processing, this method can be used to reconstruct contour surfaces, extract and combine contour lines, set the characteristic parameters of crater morphology, and establish a crater pattern recognition program. The method has been tested and verified with DEM data from Chang'E-1 (CE-1) and Chang'E-2 (CE-2), showing a strong crater recognition ability with high detection rate, high robustness, and good adaptation to recognize various craters with different diameter and morphology. The method has been used to identify craters with high precision and accuracy on the Moon. The results meet requirements for supporting exploration and related scientific research for the Moon and planets. We have developed a new crater detection method base on contour lines extracted directly from digital elevation model.Our method employs a set of geometry and topographic feature parameters to character crater's morphology.A discriminant formula consisting of the these parameters with respective weights is trained by Fisher's discriminant analysis.Our method has a detection rate up to 87% on test data.