Research/ PLANETARY ROBOTICS
Deep Learning Based Real-Time Lunar Terrain Detection for Autonomous Landing Approach
Kumar Sheshank Shekhar, Harsha Avinash Tanti, Abhirup Datta
Learning-based MethodsOpenAlex topic: Advanced Neural Network Applications · Perception & State EstimationOpenAlex keyword: real-time object detection · Design, Mechanisms & HardwareOpenAlex keyword: edge computing
Space exploration and research have led to many advancements in launch vehicles, landers, and rovers. Conducting in-situ observations requires identifying a safe landing location. To choose a safe landing site, this paper discusses computer vision technology for landers. This study utilizes the YOLOv5n model to identify the moon’s terrains. It is observed in this study that an accuracy of 92% can be achieved with near realtime detection using AI-edge devices.