Site Detection for Autonomous Soft-Landing on Asteroids Using Deep Learning

Autor: Khilan Ravani, Radhakant Padhi, S. Mathavaraj
Rok vydání: 2021
Předmět:
Zdroj: Transactions of the Indian National Academy of Engineering.
ISSN: 2662-5423
2662-5415
DOI: 10.1007/s41403-021-00207-0
Popis: In this paper, to achieve the crucial task of detection of potential landing sites for soft landing, a deep learning-based novel approach is proposed. Furthermore, the applicability of object detection algorithms for detection of landing sites has been investigated in this paper. In particular, the usage of Mask-Region Convolutional Neural Networks for detecting the landing sites on the surface of an asteroid has been demonstrated. Next, an implementation procedure of the proposed design for site detection on an autonomous soft-landing on an asteroid has been recommended. However, deep learning-based techniques require a labeled data set, and unfortunately, no such labeled data is available for open access in the public domain. To address this issue, a possible procedure for creating such a labeled data set has also been discussed. The proposed design has been successfully demonstrated on asteroid Vesta.
Databáze: OpenAIRE
Externí odkaz: