Morphology Feature Rapid Selection for Asteroid Landing

Given the complex and numerous morphological features during asteroid landing, which lead to challenges such as huge number of feature combinations, high computational load in online feature selection and low computational efficiency, in this paper an online multi-type feature rapid selection method was proposed. Using rapid selection criteria for multi-type features combined with search region, a method for selecting single-frame image features was established. Additionally, considering the short sampling interval, high inter-frame image overlap, and high feature inheritance probability during asteroid landing, a method for inter-frame image feature inheritance selection was proposed. Simulation experiments show that the design of single-frame image search regions and inter-frame image feature inheritance methods substantially reduces the number of feature combinations, thereby greatly enhancing the efficiency of computational and online feature selection.