Performance Comparison of Rock Detection Algorithms for Autonomous Planetary Geology

Abstract

Detecting rocks in images is a valuable capability for autonomous planetary science. Rock detection facilitates selective data collection and return. It also assists with image analysis on Earth. This work reviews seven rock detection algorithms from the autonomous science literature. We evaluate each algorithm with respect to several autonomous geology applications. Tests show the algorithms' performance on Mars Exploration Rover imagery, terrestrial images from analog environments, and synthetic images from a Mars terrain simulator. This provides insight into the detectors' performance under different imaging conditions.