AirFormer: Learning-Based Object Detection for Mars Helicopter

In future multiagent Mars detection schemes, the Mars helicopter can assist the scientific missions of Mars rovers by providing navigation information and scientific objects. However, Mars surface exhibits a complex topography with diverse objects and similar textures to the background, posing a gre...

Full description

Saved in:
Bibliographic Details
Main Authors: Yifan Qi, Xueming Xiao, Meibao Yao, Yonggang Xiong, Lei Zhang, Hutao Cui
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10746318/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850266393646202880
author Yifan Qi
Xueming Xiao
Meibao Yao
Yonggang Xiong
Lei Zhang
Hutao Cui
author_facet Yifan Qi
Xueming Xiao
Meibao Yao
Yonggang Xiong
Lei Zhang
Hutao Cui
author_sort Yifan Qi
collection DOAJ
description In future multiagent Mars detection schemes, the Mars helicopter can assist the scientific missions of Mars rovers by providing navigation information and scientific objects. However, Mars surface exhibits a complex topography with diverse objects and similar textures to the background, posing a great challenge for existing CNN-based object detection networks. In this article, we propose a novel deep learning-based object detection framework, AirFormer, for Mars helicopter. AirFormer embeds a new feature-fusion attention module, MAT, which injects various receptive field sizes into labels. This fusion module is capable of capturing the interrelations between objects with each other while simultaneously reducing computational complexity. In addition, we published a synthetic dataset from the viewpoint of the Mars helicopter: SynMars-Air, which refers to the data collected by the ZhuRong rover. Extensive experiments are conducted to validate the performance of AirFormer compared to SOTA methods. The results show that our method achieved the highest accuracy both on synthetic and real Mars landscapes.
format Article
id doaj-art-e8ee5b776b094b898d120d7e3e70f3dc
institution OA Journals
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-e8ee5b776b094b898d120d7e3e70f3dc2025-08-20T01:54:11ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-011810011110.1109/JSTARS.2024.349234610746318AirFormer: Learning-Based Object Detection for Mars HelicopterYifan Qi0Xueming Xiao1https://orcid.org/0000-0001-6511-0524Meibao Yao2https://orcid.org/0000-0002-7069-6782Yonggang Xiong3https://orcid.org/0000-0002-3058-4436Lei Zhang4Hutao Cui5CVIR lab, Changchun University of Science and Technology, Changchun, ChinaCVIR lab, Changchun University of Science and Technology, Changchun, ChinaSchool of Artificial Intelligence, Jilin University, Changchun, ChinaShanghai Research Institute for Intelligent Autonomous Systems, Tongji University, Shanghai, ChinaKey Lab of Opto-electronic Measurement and Optical Information Transmission Technology, Ministry of Education, Changchun, ChinaSchool of Astronautics, Harbin Institute of Technology, Harbin, ChinaIn future multiagent Mars detection schemes, the Mars helicopter can assist the scientific missions of Mars rovers by providing navigation information and scientific objects. However, Mars surface exhibits a complex topography with diverse objects and similar textures to the background, posing a great challenge for existing CNN-based object detection networks. In this article, we propose a novel deep learning-based object detection framework, AirFormer, for Mars helicopter. AirFormer embeds a new feature-fusion attention module, MAT, which injects various receptive field sizes into labels. This fusion module is capable of capturing the interrelations between objects with each other while simultaneously reducing computational complexity. In addition, we published a synthetic dataset from the viewpoint of the Mars helicopter: SynMars-Air, which refers to the data collected by the ZhuRong rover. Extensive experiments are conducted to validate the performance of AirFormer compared to SOTA methods. The results show that our method achieved the highest accuracy both on synthetic and real Mars landscapes.https://ieeexplore.ieee.org/document/10746318/Mars datasetMars explorationobject detectiontransformer model
spellingShingle Yifan Qi
Xueming Xiao
Meibao Yao
Yonggang Xiong
Lei Zhang
Hutao Cui
AirFormer: Learning-Based Object Detection for Mars Helicopter
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Mars dataset
Mars exploration
object detection
transformer model
title AirFormer: Learning-Based Object Detection for Mars Helicopter
title_full AirFormer: Learning-Based Object Detection for Mars Helicopter
title_fullStr AirFormer: Learning-Based Object Detection for Mars Helicopter
title_full_unstemmed AirFormer: Learning-Based Object Detection for Mars Helicopter
title_short AirFormer: Learning-Based Object Detection for Mars Helicopter
title_sort airformer learning based object detection for mars helicopter
topic Mars dataset
Mars exploration
object detection
transformer model
url https://ieeexplore.ieee.org/document/10746318/
work_keys_str_mv AT yifanqi airformerlearningbasedobjectdetectionformarshelicopter
AT xuemingxiao airformerlearningbasedobjectdetectionformarshelicopter
AT meibaoyao airformerlearningbasedobjectdetectionformarshelicopter
AT yonggangxiong airformerlearningbasedobjectdetectionformarshelicopter
AT leizhang airformerlearningbasedobjectdetectionformarshelicopter
AT hutaocui airformerlearningbasedobjectdetectionformarshelicopter