Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels
This article explores the challenges of integrating two deep learning neural networks, YOLOv5 and RT-DETR, to enhance the recognition of unmanned aerial vehicles (UAVs) within the optical-electronic channels of Sensor Fusion systems. The authors conducted an experimental study to test YOLOv5 and Fas...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-11-01
|
| Series: | Modelling |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-3951/5/4/92 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850037337527943168 |
|---|---|
| author | Vladislav Semenyuk Ildar Kurmashev Dmitriy Alyoshin Liliya Kurmasheva Vasiliy Serbin Alessandro Cantelli-Forti |
| author_facet | Vladislav Semenyuk Ildar Kurmashev Dmitriy Alyoshin Liliya Kurmasheva Vasiliy Serbin Alessandro Cantelli-Forti |
| author_sort | Vladislav Semenyuk |
| collection | DOAJ |
| description | This article explores the challenges of integrating two deep learning neural networks, YOLOv5 and RT-DETR, to enhance the recognition of unmanned aerial vehicles (UAVs) within the optical-electronic channels of Sensor Fusion systems. The authors conducted an experimental study to test YOLOv5 and Faster RT-DETR in order to identify the average accuracy of UAV recognition. A dataset in the form of images of two classes of objects, UAVs, and birds, was prepared in advance. The total number of images, including augmentation, amounted to 6337. The authors implemented training, verification, and testing of the neural networks exploiting PyCharm 2024 IDE. Inference testing was conducted using six videos with UAV flights. On all test videos, RT-DETR-R50 was more accurate by an average of 18.7% in terms of average classification accuracy (Pc). In terms of operating speed, YOLOv5 was 3.4 ms more efficient. It has been established that the use of RT-DETR as the only module for UAV classification in optical-electronic detection channels is not effective due to the large volumes of calculations, which is due to the relatively large number of parameters. Based on the obtained results, an algorithm for combining two neural networks is proposed, which allows for increasing the accuracy of UAV and bird classification without significant losses in speed. |
| format | Article |
| id | doaj-art-8879dd36f68c40a6aecd9c94d3325c9c |
| institution | DOAJ |
| issn | 2673-3951 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Modelling |
| spelling | doaj-art-8879dd36f68c40a6aecd9c94d3325c9c2025-08-20T02:56:52ZengMDPI AGModelling2673-39512024-11-01541773178810.3390/modelling5040092Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance ChannelsVladislav Semenyuk0Ildar Kurmashev1Dmitriy Alyoshin2Liliya Kurmasheva3Vasiliy Serbin4Alessandro Cantelli-Forti5Power Engineering and Radioelectronics Department, M. Kozybayev North Kazakhstan University, Petropavl 150000, KazakhstanInformation and Communication Technologies Department, M. Kozybayev North Kazakhstan University, Petropavl 150000, KazakhstanNatural Sciences, Head of Scientific Research Organization Department, M. Kozybayev North Kazakhstan University, Petropavl 150000, KazakhstanInformation and Communication Technologies Department, M. Kozybayev North Kazakhstan University, Petropavl 150000, KazakhstanCyber Security, Information Processing and Storage Department, K.I. Satpayev Kazakh National Research Technical University, Petropavl 150000, KazakhstanRadar and Surveillance Systems, National Laboratory, 56124 Pisa, ItalyThis article explores the challenges of integrating two deep learning neural networks, YOLOv5 and RT-DETR, to enhance the recognition of unmanned aerial vehicles (UAVs) within the optical-electronic channels of Sensor Fusion systems. The authors conducted an experimental study to test YOLOv5 and Faster RT-DETR in order to identify the average accuracy of UAV recognition. A dataset in the form of images of two classes of objects, UAVs, and birds, was prepared in advance. The total number of images, including augmentation, amounted to 6337. The authors implemented training, verification, and testing of the neural networks exploiting PyCharm 2024 IDE. Inference testing was conducted using six videos with UAV flights. On all test videos, RT-DETR-R50 was more accurate by an average of 18.7% in terms of average classification accuracy (Pc). In terms of operating speed, YOLOv5 was 3.4 ms more efficient. It has been established that the use of RT-DETR as the only module for UAV classification in optical-electronic detection channels is not effective due to the large volumes of calculations, which is due to the relatively large number of parameters. Based on the obtained results, an algorithm for combining two neural networks is proposed, which allows for increasing the accuracy of UAV and bird classification without significant losses in speed.https://www.mdpi.com/2673-3951/5/4/92dronessensor fusionneural networksvision transformersinferencetraining |
| spellingShingle | Vladislav Semenyuk Ildar Kurmashev Dmitriy Alyoshin Liliya Kurmasheva Vasiliy Serbin Alessandro Cantelli-Forti Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels Modelling drones sensor fusion neural networks vision transformers inference training |
| title | Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels |
| title_full | Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels |
| title_fullStr | Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels |
| title_full_unstemmed | Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels |
| title_short | Study of the Possibility to Combine Deep Learning Neural Networks for Recognition of Unmanned Aerial Vehicles in Optoelectronic Surveillance Channels |
| title_sort | study of the possibility to combine deep learning neural networks for recognition of unmanned aerial vehicles in optoelectronic surveillance channels |
| topic | drones sensor fusion neural networks vision transformers inference training |
| url | https://www.mdpi.com/2673-3951/5/4/92 |
| work_keys_str_mv | AT vladislavsemenyuk studyofthepossibilitytocombinedeeplearningneuralnetworksforrecognitionofunmannedaerialvehiclesinoptoelectronicsurveillancechannels AT ildarkurmashev studyofthepossibilitytocombinedeeplearningneuralnetworksforrecognitionofunmannedaerialvehiclesinoptoelectronicsurveillancechannels AT dmitriyalyoshin studyofthepossibilitytocombinedeeplearningneuralnetworksforrecognitionofunmannedaerialvehiclesinoptoelectronicsurveillancechannels AT liliyakurmasheva studyofthepossibilitytocombinedeeplearningneuralnetworksforrecognitionofunmannedaerialvehiclesinoptoelectronicsurveillancechannels AT vasiliyserbin studyofthepossibilitytocombinedeeplearningneuralnetworksforrecognitionofunmannedaerialvehiclesinoptoelectronicsurveillancechannels AT alessandrocantelliforti studyofthepossibilitytocombinedeeplearningneuralnetworksforrecognitionofunmannedaerialvehiclesinoptoelectronicsurveillancechannels |