Slip Decision Algorithm and Power Consumption Analysis of Lunar Exploration CubeRover Using Star Motion Blurring on Cortex-Based Embedded PC; [Cortex 기반 임베디드 PC에서의 별 모션 블러링 현상을 활용한 달 탐사 큐브로버의 미끄러짐 판단 알고리즘 및 전력소비 분석]

Abstract

This study proposes a camera-vision-based method for slip perception and decision through image preprocessing utilizing star motion blurring phenomena to enable safe navigation for exploration rovers on hazardous terrains, such as soft sand, which may induce slippage. Aiming for real-time performance in low-power embedded environments, we developed a Cortex-based printed circuit board. Through evaluations on validation datasets, the slip detection algorithm showed that the contour tracing technique achieved 31% higher accuracy compared to the Hough transformation. In addition, frames per second performance and power consumption comparisons confirmed that the slip detection algorithm using the contour tracing technique is more suitable for autonomous driving environments. This research presents an optimal algorithm and hardware configuration for real-time hazard detection on low-power embedded systems, particularly for CubeRover applications. We expect that these findings will contribute to the development and validation of CubeRover-based exploration rovers in autonomous navigation scenarios. © 2024 Korean Society of Remote Sensing.