Autonomous crater detection on asteroids using a fully-convolutional neural network

TL;DR

This paper demonstrates an autonomous crater detection method on Ceres using a U-Net neural network trained on lunar data and fine-tuned with transfer learning, achieving high accuracy and potential for automated planetary surface analysis.

Contribution

It introduces a transfer learning approach with a U-Net model for crater detection on Ceres, adapting lunar-trained models to asteroid imagery with promising accuracy.

Findings

Achieved up to 97.19% accuracy on unseen Ceres images

Transfer learning effectively adapts lunar-trained models to Ceres data

Post-processing improves crater detection for cataloguing and navigation

Abstract

This paper shows the application of autonomous Crater Detection using the U-Net, a Fully-Convolutional Neural Network, on Ceres. The U-Net is trained on optical images of the Moon Global Morphology Mosaic based on data collected by the LRO and manual crater catalogues. The Moon-trained network will be tested on Dawn optical images of Ceres: this task is accomplished by means of a Transfer Learning (TL) approach. The trained model has been fine-tuned using 100, 500 and 1000 additional images of Ceres. The test performance was measured on 350 never before seen images, reaching a testing accuracy of 96.24%, 96.95% and 97.19%, respectively. This means that despite the intrinsic differences between the Moon and Ceres, TL works with encouraging results. The output of the U-Net contains predicted craters: it will be post-processed applying global thresholding for image binarization and a template matching algorithm to extract craters positions and radii in the pixel space. Post-processed craters will be counted and compared to the ground truth data in order to compute image segmentation metrics: precision, recall and F1 score. These indices will be computed, and their effect will be discussed for tasks such as automated crater cataloguing and optical navigation.