An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images
Unmanned aerial vehicle (UAV) image detection algorithms are critical in performing military countermeasures and disaster search and rescue. The state-of-the-art object detection algorithm known as you only look once (YOLO) is widely used for detecting UAV images. However, it faces challenges such a...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-05-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/12/10/2274 |
_version_ | 1797600311592878080 |
---|---|
author | Lijia Cao Pinde Song Yongchao Wang Yang Yang Baoyu Peng |
author_facet | Lijia Cao Pinde Song Yongchao Wang Yang Yang Baoyu Peng |
author_sort | Lijia Cao |
collection | DOAJ |
description | Unmanned aerial vehicle (UAV) image detection algorithms are critical in performing military countermeasures and disaster search and rescue. The state-of-the-art object detection algorithm known as you only look once (YOLO) is widely used for detecting UAV images. However, it faces challenges such as high floating-point operations (FLOPs), redundant parameters, slow inference speed, and poor performance in detecting small objects. To address the above issues, an improved, lightweight, real-time detection algorithm was proposed based on the edge computing platform for UAV images. In the presented method, MobileNetV3 was used as the YOLOv5 backbone network to reduce the numbers of parameters and FLOPs. To enhance the feature extraction ability of MobileNetV3, the efficient channel attention (ECA) attention mechanism was introduced into MobileNetV3. Furthermore, in order to improve the detection ability for small objects, an extra prediction head was introduced into the neck structure, and two kinds of neck structures with different parameter scales were designed to meet the requirements of different embedded devices. Finally, the FocalEIoU loss function was introduced into YOLOv5 to accelerate bounding box regression and improve the localization accuracy of the algorithm. To validate the performance of the proposed improved algorithm, we compared our algorithm with other algorithms in the VisDrone-Det2021 dataset. The results showed that compared with YOLOv5s, MELF-YOLOv5-S achieved a 51.4% reduction in the number of parameters and a 38.6% decrease in the number of FLOPs. MELF-YOLOv5-L had 87.4% and 47.4% fewer parameters and FLOPs, respectively, and achieved higher detection accuracy than YOLOv5l. |
first_indexed | 2024-03-11T03:46:42Z |
format | Article |
id | doaj.art-683cc39636f840ae909027c4819b73a5 |
institution | Directory Open Access Journal |
issn | 2079-9292 |
language | English |
last_indexed | 2024-03-11T03:46:42Z |
publishDate | 2023-05-01 |
publisher | MDPI AG |
record_format | Article |
series | Electronics |
spelling | doaj.art-683cc39636f840ae909027c4819b73a52023-11-18T01:10:08ZengMDPI AGElectronics2079-92922023-05-011210227410.3390/electronics12102274An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV ImagesLijia Cao0Pinde Song1Yongchao Wang2Yang Yang3Baoyu Peng4School of Automation & Information Engineering, Sichuan University of Science & Engineering, Yibin 644000, ChinaSchool of Automation & Information Engineering, Sichuan University of Science & Engineering, Yibin 644000, ChinaChair of Automatic Control Engineering, Technical University of Munich, 80333 Munich, GermanySchool of Automation & Information Engineering, Sichuan University of Science & Engineering, Yibin 644000, ChinaSchool of Computing Science and Engineering, Sichuan University of Science & Engineering, Yibin 644000, ChinaUnmanned aerial vehicle (UAV) image detection algorithms are critical in performing military countermeasures and disaster search and rescue. The state-of-the-art object detection algorithm known as you only look once (YOLO) is widely used for detecting UAV images. However, it faces challenges such as high floating-point operations (FLOPs), redundant parameters, slow inference speed, and poor performance in detecting small objects. To address the above issues, an improved, lightweight, real-time detection algorithm was proposed based on the edge computing platform for UAV images. In the presented method, MobileNetV3 was used as the YOLOv5 backbone network to reduce the numbers of parameters and FLOPs. To enhance the feature extraction ability of MobileNetV3, the efficient channel attention (ECA) attention mechanism was introduced into MobileNetV3. Furthermore, in order to improve the detection ability for small objects, an extra prediction head was introduced into the neck structure, and two kinds of neck structures with different parameter scales were designed to meet the requirements of different embedded devices. Finally, the FocalEIoU loss function was introduced into YOLOv5 to accelerate bounding box regression and improve the localization accuracy of the algorithm. To validate the performance of the proposed improved algorithm, we compared our algorithm with other algorithms in the VisDrone-Det2021 dataset. The results showed that compared with YOLOv5s, MELF-YOLOv5-S achieved a 51.4% reduction in the number of parameters and a 38.6% decrease in the number of FLOPs. MELF-YOLOv5-L had 87.4% and 47.4% fewer parameters and FLOPs, respectively, and achieved higher detection accuracy than YOLOv5l.https://www.mdpi.com/2079-9292/12/10/2274lightweightFocalEIoUUAV imageattention mechanismembedded device |
spellingShingle | Lijia Cao Pinde Song Yongchao Wang Yang Yang Baoyu Peng An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images Electronics lightweight FocalEIoU UAV image attention mechanism embedded device |
title | An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images |
title_full | An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images |
title_fullStr | An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images |
title_full_unstemmed | An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images |
title_short | An Improved Lightweight Real-Time Detection Algorithm Based on the Edge Computing Platform for UAV Images |
title_sort | improved lightweight real time detection algorithm based on the edge computing platform for uav images |
topic | lightweight FocalEIoU UAV image attention mechanism embedded device |
url | https://www.mdpi.com/2079-9292/12/10/2274 |
work_keys_str_mv | AT lijiacao animprovedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT pindesong animprovedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT yongchaowang animprovedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT yangyang animprovedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT baoyupeng animprovedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT lijiacao improvedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT pindesong improvedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT yongchaowang improvedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT yangyang improvedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages AT baoyupeng improvedlightweightrealtimedetectionalgorithmbasedontheedgecomputingplatformforuavimages |