DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR

For drone-based detection tasks, accurately identifying small-scale targets like people, bicycles, and pedestrians remains a key challenge. In this paper, we propose DV-DETR, an improved detection model based on the Real-Time Detection Transformer (RT-DETR), specifically optimized for small target d...

Full description

Saved in:
Bibliographic Details
Main Authors: Xiaolong Wei, Ling Yin, Liangliang Zhang, Fei Wu
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/24/22/7376
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846152388378886144
author Xiaolong Wei
Ling Yin
Liangliang Zhang
Fei Wu
author_facet Xiaolong Wei
Ling Yin
Liangliang Zhang
Fei Wu
author_sort Xiaolong Wei
collection DOAJ
description For drone-based detection tasks, accurately identifying small-scale targets like people, bicycles, and pedestrians remains a key challenge. In this paper, we propose DV-DETR, an improved detection model based on the Real-Time Detection Transformer (RT-DETR), specifically optimized for small target detection in high-density scenes. To achieve this, we introduce three main enhancements: (1) ResNet18 as the backbone network to improve feature extraction and reduce model complexity; (2) the integration of recalibration attention units and deformable attention mechanisms in the neck network to enhance multi-scale feature fusion and improve localization accuracy; and (3) the use of the Focaler-IoU loss function to better handle the imbalanced distribution of target scales and focus on challenging samples. Experimental results on the VisDrone2019 dataset show that DV-DETR achieves an mAP@0.5 of 50.1%, a 1.7% improvement over the baseline model, while increasing detection speed from 75 FPS to 90 FPS, meeting real-time processing requirements. These improvements not only enhance the model’s accuracy and efficiency but also provide practical significance in complex, high-density urban environments, supporting real-world applications in UAV-based surveillance and monitoring tasks.
format Article
id doaj-art-09cf52d87e404578a7d9cb44a447a412
institution Kabale University
issn 1424-8220
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-09cf52d87e404578a7d9cb44a447a4122024-11-26T18:21:46ZengMDPI AGSensors1424-82202024-11-012422737610.3390/s24227376DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETRXiaolong Wei0Ling Yin1Liangliang Zhang2Fei Wu3School of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai 201620, ChinaSchool of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai 201620, ChinaSchool of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai 201620, ChinaSchool of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Shanghai 201620, ChinaFor drone-based detection tasks, accurately identifying small-scale targets like people, bicycles, and pedestrians remains a key challenge. In this paper, we propose DV-DETR, an improved detection model based on the Real-Time Detection Transformer (RT-DETR), specifically optimized for small target detection in high-density scenes. To achieve this, we introduce three main enhancements: (1) ResNet18 as the backbone network to improve feature extraction and reduce model complexity; (2) the integration of recalibration attention units and deformable attention mechanisms in the neck network to enhance multi-scale feature fusion and improve localization accuracy; and (3) the use of the Focaler-IoU loss function to better handle the imbalanced distribution of target scales and focus on challenging samples. Experimental results on the VisDrone2019 dataset show that DV-DETR achieves an mAP@0.5 of 50.1%, a 1.7% improvement over the baseline model, while increasing detection speed from 75 FPS to 90 FPS, meeting real-time processing requirements. These improvements not only enhance the model’s accuracy and efficiency but also provide practical significance in complex, high-density urban environments, supporting real-world applications in UAV-based surveillance and monitoring tasks.https://www.mdpi.com/1424-8220/24/22/7376transformersmall target detectionreal-time taskRT-DETR algorithm
spellingShingle Xiaolong Wei
Ling Yin
Liangliang Zhang
Fei Wu
DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR
Sensors
transformer
small target detection
real-time task
RT-DETR algorithm
title DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR
title_full DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR
title_fullStr DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR
title_full_unstemmed DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR
title_short DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR
title_sort dv detr improved uav aerial small target detection algorithm based on rt detr
topic transformer
small target detection
real-time task
RT-DETR algorithm
url https://www.mdpi.com/1424-8220/24/22/7376
work_keys_str_mv AT xiaolongwei dvdetrimproveduavaerialsmalltargetdetectionalgorithmbasedonrtdetr
AT lingyin dvdetrimproveduavaerialsmalltargetdetectionalgorithmbasedonrtdetr
AT liangliangzhang dvdetrimproveduavaerialsmalltargetdetectionalgorithmbasedonrtdetr
AT feiwu dvdetrimproveduavaerialsmalltargetdetectionalgorithmbasedonrtdetr