An Enhanced Target Detection Algorithm for Maritime Search and Rescue Based on Aerial Images

被引:19
作者
Zhang, Yijian [1 ]
Yin, Yong [1 ]
Shao, Zeyuan [1 ]
机构
[1] Dalian Maritime Univ, Nav Coll, Dalian 116026, Peoples R China
基金
国家重点研发计划;
关键词
object detection; maritime search and rescue; unmanned aerial vehicles (UAVs); computer vision;
D O I
10.3390/rs15194818
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Unmanned aerial vehicles (UAVs), renowned for their rapid deployment, extensive data collection, and high spatial resolution, are crucial in locating distressed individuals during search and rescue (SAR) operations. Challenges in maritime search and rescue include missed detections due to issues including sunlight reflection. In this study, we proposed an enhanced ABT-YOLOv7 algorithm for underwater person detection. This algorithm integrates an asymptotic feature pyramid network (AFPN) to preserve the target feature information. The BiFormer module enhances the model's perception of small-scale targets, whereas the task-specific context decoupling (TSCODE) mechanism effectively resolves conflicts between localization and classification. Using quantitative experiments on a curated dataset, our model outperformed methods such as YOLOv3, YOLOv4, YOLOv5, YOLOv8, Faster R-CNN, Cascade R-CNN, and FCOS. Compared with YOLOv7, our approach enhances the mean average precision (mAP) from 87.1% to 91.6%. Therefore, our approach reduces the sensitivity of the detection model to low-lighting conditions and sunlight reflection, thus demonstrating enhanced robustness. These innovations have driven advancements in UAV technology within the maritime search and rescue domains.
引用
收藏
页数:18
相关论文
共 48 条
[31]  
Selvaraju RR, 2020, INT J COMPUT VISION, V128, P336, DOI [10.1109/ICCV.2017.74, 10.1007/s11263-019-01228-7]
[32]   AIR-YOLOv3: Aerial Infrared Pedestrian Detection via an Improved YOLOv3 with Network Pruning [J].
Shao, Yanhua ;
Zhang, Xingping ;
Chu, Hongyu ;
Zhang, Xiaoqiang ;
Zhang, Duo ;
Rao, Yunbo .
APPLIED SCIENCES-BASEL, 2022, 12 (07)
[33]  
Simonyan K, 2015, Arxiv, DOI arXiv:1409.1556
[34]   Object Detection Method for Grasping Robot Based on Improved YOLOv5 [J].
Song, Qisong ;
Li, Shaobo ;
Bai, Qiang ;
Yang, Jing ;
Zhang, Xingxing ;
Li, Zhiang ;
Duan, Zhongjing .
MICROMACHINES, 2021, 12 (11)
[35]   Hybrid-YOLO for classification of insulators defects in transmission lines based on UAV [J].
Souza, Bruno Jose ;
Stefenon, Stefano Frizzo ;
Singh, Gurmail ;
Freire, Roberto Zanetti .
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS, 2023, 148
[36]   FCOS: Fully Convolutional One-Stage Object Detection [J].
Tian, Zhi ;
Shen, Chunhua ;
Chen, Hao ;
He, Tong .
2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019), 2019, :9626-9635
[37]   Detection of Bottle Marine Debris Using Unmanned Aerial Vehicles and Machine Learning Techniques [J].
Tran, Thi Linh Chi ;
Huang, Zhi-Cheng ;
Tseng, Kuo-Hsin ;
Chou, Ping-Hsien .
DRONES, 2022, 6 (12)
[38]   SeaDronesSee: A Maritime Benchmark for Detecting Humans in Open Water [J].
Varga, Leon Amadeus ;
Kiefer, Benjamin ;
Messmer, Martin ;
Zell, Andreas .
2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022), 2022, :3686-3696
[39]   Rapid object detection using a boosted cascade of simple features [J].
Viola, P ;
Jones, M .
2001 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, VOL 1, PROCEEDINGS, 2001, :511-518
[40]   YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors [J].
Wang, Chien-Yao ;
Bochkovskiy, Alexey ;
Liao, Hong-Yuan Mark .
2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR, 2023, :7464-7475