Crater detection using Bayesian classifiers and LASSO

Surveying a large amount of small sub-kilometer craters in planetary images is a challenging task due to their non-distinguishable features. In this paper, we integrate the LASSO (Least Absolute Shrinkage and Selection Operator) method with the Bayesian network classifier and propose an L1 Regularized Bayesian Network Classifier (L1-BNC) algorithm for this task. The L1-BNC algorithm uses the LASSO method not only to deal with high-dimensional crater features, but also to give a crater feature order for constructing a Bayesian network classifier. Our framework is evaluated on a large Martian image of 37,500 × 56,250m². Experimental results demonstrate that this proposed method gets higher prediction accuracy than the existing crater detection algorithms.