TY - JOUR
T1 - Improving UAV Aerial Imagery Detection Method via Superresolution Synergy
AU - Wang, Dianwei
AU - Gao, Zehao
AU - Fang, Jie
AU - Li, Yuanqing
AU - Xu, Zhijie
N1 - Publisher Copyright:
© 2025 The Authors.
PY - 2025
Y1 - 2025
N2 - Unmanned aerial vehicles (UAVs) have emerged as versatile tools across various industries, providing valuable insights through aerial image analysis. However, the efficacy of UAV-deployed image detection systems is often limited by the resolution of captured images and the altitudinal constraints of UAV operations. This article introduces a novel integration of the detection system with superresolution networks and image reconstruction techniques, inspired by the exceptional visual capabilities of eagles, to enhance image detail and detection recall from aerial perspectives. The superresolution component utilizes advanced algorithms to upscale the resolution of images captured by UAVs, thereby improving the granularity and clarity of the visual data. Concurrently, image reconstruction techniques are applied to enhance the quality of original images further. In addition, we propose an innovative adaptive feature fusion technique, which not only surpasses traditional concatenation methods in integrating multiscale features effectively but also demonstrates remarkable improvement in feature utilization and further refinement of the fusion process. Extensive experiments conducted on VisDrone2019 and DOTA datasets demonstrate that our integrated system significantly outperforms existing methods in terms of detection precision and recall. Compared to YOLOv5s, recall and mAP50 have increased by 8.89% and 11.11%, respectively, with only a slight increase in the number of parameters and complexity.
AB - Unmanned aerial vehicles (UAVs) have emerged as versatile tools across various industries, providing valuable insights through aerial image analysis. However, the efficacy of UAV-deployed image detection systems is often limited by the resolution of captured images and the altitudinal constraints of UAV operations. This article introduces a novel integration of the detection system with superresolution networks and image reconstruction techniques, inspired by the exceptional visual capabilities of eagles, to enhance image detail and detection recall from aerial perspectives. The superresolution component utilizes advanced algorithms to upscale the resolution of images captured by UAVs, thereby improving the granularity and clarity of the visual data. Concurrently, image reconstruction techniques are applied to enhance the quality of original images further. In addition, we propose an innovative adaptive feature fusion technique, which not only surpasses traditional concatenation methods in integrating multiscale features effectively but also demonstrates remarkable improvement in feature utilization and further refinement of the fusion process. Extensive experiments conducted on VisDrone2019 and DOTA datasets demonstrate that our integrated system significantly outperforms existing methods in terms of detection precision and recall. Compared to YOLOv5s, recall and mAP50 have increased by 8.89% and 11.11%, respectively, with only a slight increase in the number of parameters and complexity.
KW - Eagle-eye vision system
KW - object detection
KW - unmanned aerial vehicle (UAV) imagery
KW - YOLOv5
UR - http://www.scopus.com/inward/record.url?scp=85215368636&partnerID=8YFLogxK
U2 - 10.1109/JSTARS.2024.3525148
DO - 10.1109/JSTARS.2024.3525148
M3 - Article
AN - SCOPUS:85215368636
SN - 1939-1404
VL - 18
SP - 3959
EP - 3972
JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
ER -