A Novel UAV Visual Positioning Algorithm Based on A-YOLOX
The application of UAVs is becoming increasingly extensive. However, high-precision autonomous landing is still a major industry difficulty. The current algorithm is not well-adapted to light changes, scale transformations, complex backgrounds, etc. To address the above difficulties, a deep learning...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-11-01
|
Series: | Drones |
Subjects: | |
Online Access: | https://www.mdpi.com/2504-446X/6/11/362 |
_version_ | 1797465534590091264 |
---|---|
author | Ying Xu Dongsheng Zhong Jianhong Zhou Ziyi Jiang Yikui Zhai Zilu Ying |
author_facet | Ying Xu Dongsheng Zhong Jianhong Zhou Ziyi Jiang Yikui Zhai Zilu Ying |
author_sort | Ying Xu |
collection | DOAJ |
description | The application of UAVs is becoming increasingly extensive. However, high-precision autonomous landing is still a major industry difficulty. The current algorithm is not well-adapted to light changes, scale transformations, complex backgrounds, etc. To address the above difficulties, a deep learning method was here introduced into target detection and an attention mechanism was incorporated into YOLOX; thus, a UAV positioning algorithm called attention-based YOLOX (A-YOLOX) is proposed. Firstly, a novel visual positioning pattern was designed to facilitate the algorithm’s use for detection and localization; then, a UAV visual positioning database (UAV-VPD) was built through actual data collection and data augmentation and the A-YOLOX model detector developed; finally, corresponding high- and low-altitude visual positioning algorithms were designed for high- and low-altitude positioning logics. The experimental results in the actual environment showed that the AP50 of the proposed algorithm could reach 95.5%, the detection speed was 53.7 frames per second, and the actual landing error was within 5 cm, which meets the practical application requirements for automatic UAV landing. |
first_indexed | 2024-03-09T18:23:50Z |
format | Article |
id | doaj.art-cc8fa6d9e8c741c498b09c87b51215d0 |
institution | Directory Open Access Journal |
issn | 2504-446X |
language | English |
last_indexed | 2024-03-09T18:23:50Z |
publishDate | 2022-11-01 |
publisher | MDPI AG |
record_format | Article |
series | Drones |
spelling | doaj.art-cc8fa6d9e8c741c498b09c87b51215d02023-11-24T08:06:56ZengMDPI AGDrones2504-446X2022-11-0161136210.3390/drones6110362A Novel UAV Visual Positioning Algorithm Based on A-YOLOXYing Xu0Dongsheng Zhong1Jianhong Zhou2Ziyi Jiang3Yikui Zhai4Zilu Ying5Department of Intelligent Manufacturing, Wuyi University, Jiangmen 529020, ChinaDepartment of Intelligent Manufacturing, Wuyi University, Jiangmen 529020, ChinaDepartment of Intelligent Manufacturing, Wuyi University, Jiangmen 529020, ChinaDepartment of Intelligent Manufacturing, Wuyi University, Jiangmen 529020, ChinaDepartment of Intelligent Manufacturing, Wuyi University, Jiangmen 529020, ChinaDepartment of Intelligent Manufacturing, Wuyi University, Jiangmen 529020, ChinaThe application of UAVs is becoming increasingly extensive. However, high-precision autonomous landing is still a major industry difficulty. The current algorithm is not well-adapted to light changes, scale transformations, complex backgrounds, etc. To address the above difficulties, a deep learning method was here introduced into target detection and an attention mechanism was incorporated into YOLOX; thus, a UAV positioning algorithm called attention-based YOLOX (A-YOLOX) is proposed. Firstly, a novel visual positioning pattern was designed to facilitate the algorithm’s use for detection and localization; then, a UAV visual positioning database (UAV-VPD) was built through actual data collection and data augmentation and the A-YOLOX model detector developed; finally, corresponding high- and low-altitude visual positioning algorithms were designed for high- and low-altitude positioning logics. The experimental results in the actual environment showed that the AP50 of the proposed algorithm could reach 95.5%, the detection speed was 53.7 frames per second, and the actual landing error was within 5 cm, which meets the practical application requirements for automatic UAV landing.https://www.mdpi.com/2504-446X/6/11/362deep learningdata synthesisA-YOLOXvisual positioning |
spellingShingle | Ying Xu Dongsheng Zhong Jianhong Zhou Ziyi Jiang Yikui Zhai Zilu Ying A Novel UAV Visual Positioning Algorithm Based on A-YOLOX Drones deep learning data synthesis A-YOLOX visual positioning |
title | A Novel UAV Visual Positioning Algorithm Based on A-YOLOX |
title_full | A Novel UAV Visual Positioning Algorithm Based on A-YOLOX |
title_fullStr | A Novel UAV Visual Positioning Algorithm Based on A-YOLOX |
title_full_unstemmed | A Novel UAV Visual Positioning Algorithm Based on A-YOLOX |
title_short | A Novel UAV Visual Positioning Algorithm Based on A-YOLOX |
title_sort | novel uav visual positioning algorithm based on a yolox |
topic | deep learning data synthesis A-YOLOX visual positioning |
url | https://www.mdpi.com/2504-446X/6/11/362 |
work_keys_str_mv | AT yingxu anoveluavvisualpositioningalgorithmbasedonayolox AT dongshengzhong anoveluavvisualpositioningalgorithmbasedonayolox AT jianhongzhou anoveluavvisualpositioningalgorithmbasedonayolox AT ziyijiang anoveluavvisualpositioningalgorithmbasedonayolox AT yikuizhai anoveluavvisualpositioningalgorithmbasedonayolox AT ziluying anoveluavvisualpositioningalgorithmbasedonayolox AT yingxu noveluavvisualpositioningalgorithmbasedonayolox AT dongshengzhong noveluavvisualpositioningalgorithmbasedonayolox AT jianhongzhou noveluavvisualpositioningalgorithmbasedonayolox AT ziyijiang noveluavvisualpositioningalgorithmbasedonayolox AT yikuizhai noveluavvisualpositioningalgorithmbasedonayolox AT ziluying noveluavvisualpositioningalgorithmbasedonayolox |