Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks

Yanjie Guo; Yimin Deng; Haibin Duan

doi:10.1007/978-3-030-04212-7_38

Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks

Yanjie Guo
, Yimin Deng^*
, Haibin Duan

^*Corresponding author for this work

School of Automation Science and Electrical Engineering

Beihang University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Scopus citations

Abstract

Autonomous aerial refueling (AAR) is an important capability for the future development of unmanned aerial vehicles (UAVs). A robust and accurate algorithm of detecting the drogue is crucial to such a capability. In this paper, we present an innovative algorithm based on the adaptive boosting algorithm and convolutional neural networks (CNN) classifier with improved focal loss (IFL). The IFL function addresses the sample imbalance during the training stage of the CNN classifier. The pytorch deep learning framework with the graphics processing units (GPUs) is used to implement the system. Real scenario images that contain drogue carried by UAVs are for training and testing. The results show that the algorithm not only accelerates the speed but also improves the accuracy.

Original language	English
Title of host publication	Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings
Editors	Seiichi Ozawa, Andrew Chi Sing Leung, Long Cheng
Publisher	Springer Verlag
Pages	437-443
Number of pages	7
ISBN (Print)	9783030042110
DOIs	https://doi.org/10.1007/978-3-030-04212-7_38
State	Published - 2018
Event	25th International Conference on Neural Information Processing, ICONIP 2018 - Siem Reap, Cambodia Duration: 13 Dec 2018 → 16 Dec 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11304 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	25th International Conference on Neural Information Processing, ICONIP 2018
Country/Territory	Cambodia
City	Siem Reap
Period	13/12/18 → 16/12/18

Keywords

Adaboost
Autonomous aerial refueling
CNN
Sample imbalance

Access to Document

10.1007/978-3-030-04212-7_38

Cite this

Guo, Y., Deng, Y., & Duan, H. (2018). Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks. In S. Ozawa, A. C. S. Leung, & L. Cheng (Eds.), Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings (pp. 437-443). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11304 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-04212-7_38

Guo, Yanjie ; Deng, Yimin ; Duan, Haibin. / Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks. Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. editor / Seiichi Ozawa ; Andrew Chi Sing Leung ; Long Cheng. Springer Verlag, 2018. pp. 437-443 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{edbd1aa1e50e48669f90c3e3a7fa8c29,

title = "Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks",

abstract = "Autonomous aerial refueling (AAR) is an important capability for the future development of unmanned aerial vehicles (UAVs). A robust and accurate algorithm of detecting the drogue is crucial to such a capability. In this paper, we present an innovative algorithm based on the adaptive boosting algorithm and convolutional neural networks (CNN) classifier with improved focal loss (IFL). The IFL function addresses the sample imbalance during the training stage of the CNN classifier. The pytorch deep learning framework with the graphics processing units (GPUs) is used to implement the system. Real scenario images that contain drogue carried by UAVs are for training and testing. The results show that the algorithm not only accelerates the speed but also improves the accuracy.",

keywords = "Adaboost, Autonomous aerial refueling, CNN, Sample imbalance",

author = "Yanjie Guo and Yimin Deng and Haibin Duan",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2018.; 25th International Conference on Neural Information Processing, ICONIP 2018 ; Conference date: 13-12-2018 Through 16-12-2018",

year = "2018",

doi = "10.1007/978-3-030-04212-7\_38",

language = "英语",

isbn = "9783030042110",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "437--443",

editor = "Seiichi Ozawa and Leung, \{Andrew Chi Sing\} and Long Cheng",

booktitle = "Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings",

address = "德国",

}

Guo, Y, Deng, Y & Duan, H 2018, Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks. in S Ozawa, ACS Leung & L Cheng (eds), Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11304 LNCS, Springer Verlag, pp. 437-443, 25th International Conference on Neural Information Processing, ICONIP 2018, Siem Reap, Cambodia, 13/12/18. https://doi.org/10.1007/978-3-030-04212-7_38

Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks. / Guo, Yanjie; Deng, Yimin ; Duan, Haibin.
Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. ed. / Seiichi Ozawa; Andrew Chi Sing Leung; Long Cheng. Springer Verlag, 2018. p. 437-443 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11304 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks

AU - Guo, Yanjie

AU - Deng, Yimin

AU - Duan, Haibin

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2018.

PY - 2018

Y1 - 2018

N2 - Autonomous aerial refueling (AAR) is an important capability for the future development of unmanned aerial vehicles (UAVs). A robust and accurate algorithm of detecting the drogue is crucial to such a capability. In this paper, we present an innovative algorithm based on the adaptive boosting algorithm and convolutional neural networks (CNN) classifier with improved focal loss (IFL). The IFL function addresses the sample imbalance during the training stage of the CNN classifier. The pytorch deep learning framework with the graphics processing units (GPUs) is used to implement the system. Real scenario images that contain drogue carried by UAVs are for training and testing. The results show that the algorithm not only accelerates the speed but also improves the accuracy.

AB - Autonomous aerial refueling (AAR) is an important capability for the future development of unmanned aerial vehicles (UAVs). A robust and accurate algorithm of detecting the drogue is crucial to such a capability. In this paper, we present an innovative algorithm based on the adaptive boosting algorithm and convolutional neural networks (CNN) classifier with improved focal loss (IFL). The IFL function addresses the sample imbalance during the training stage of the CNN classifier. The pytorch deep learning framework with the graphics processing units (GPUs) is used to implement the system. Real scenario images that contain drogue carried by UAVs are for training and testing. The results show that the algorithm not only accelerates the speed but also improves the accuracy.

KW - Adaboost

KW - Autonomous aerial refueling

KW - CNN

KW - Sample imbalance

UR - https://www.scopus.com/pages/publications/85059012129

U2 - 10.1007/978-3-030-04212-7_38

DO - 10.1007/978-3-030-04212-7_38

M3 - 会议稿件

AN - SCOPUS:85059012129

SN - 9783030042110

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 437

EP - 443

BT - Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings

A2 - Ozawa, Seiichi

A2 - Leung, Andrew Chi Sing

A2 - Cheng, Long

PB - Springer Verlag

T2 - 25th International Conference on Neural Information Processing, ICONIP 2018

Y2 - 13 December 2018 through 16 December 2018

ER -

Guo Y, Deng Y , Duan H. Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks. In Ozawa S, Leung ACS, Cheng L, editors, Neural Information Processing - 25th International Conference, ICONIP 2018, Proceedings. Springer Verlag. 2018. p. 437-443. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-04212-7_38

Drogue detection for autonomous aerial refueling based on adaboost and convolutional neural networks

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this