[1]马洲俊,陈锦铭,刘浩林,等.基于注意力机制与可变卷积神经网络的卫星视频运动目标检测[J].南京师大学报(自然科学版),2025,48(04):78-86.[doi:10.3969/j.issn.1001-4616.2025.04.008]
　Ma Zhoujun,Chen Jinming,Liu Haolin,et al.Satellite Video Moving Object Detection Based on Deformable Convolutional Neural Network and Shuffle Attention[J].Journal of Nanjing Normal University(Natural Science Edition),2025,48(04):78-86.[doi:10.3969/j.issn.1001-4616.2025.04.008]
点击复制

基于注意力机制与可变卷积神经网络的卫星视频运动目标检测()

《南京师大学报（自然科学版）》[ISSN:1001-4616/CN:32-1239/N]

卷:
48
期数:
2025年04期
页码:
78-86
栏目:
计算机科学与技术
出版日期:
2025-08-20

文章信息/Info

Title:
Satellite Video Moving Object Detection Based on Deformable Convolutional Neural Network and Shuffle Attention
文章编号:
1001-4616(2025)04-0078-09
作者:
马洲俊1; 陈锦铭2; 刘浩林3; 张 卡3
(1.国网江苏省电力有限公司,江苏 南京 210019)
(2.国网江苏省电力有限公司电力科学研究院,江苏 南京 211103)
(3.南京师范大学地理科学学院,江苏 南京 210023)
Author(s):
Ma Zhoujun1; Chen Jinming2; Liu Haolin3; Zhang Ka3
(1.State Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 210019,China)
(2.Electric Power Research Institute,State Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 211103,China)
(3.School of Geography,Nanjing Normal University,Nanjing 210023,China)
关键词:
卫星视频; YOLOv8; 轻量级注意力机制; 可变形卷积; 辅助边框回归
Keywords:
satellite video; YOLOv8; lightweight attention mechanism; deformable convolution; auxiliary border regression
分类号:
P237; TP751
DOI:
10.3969/j.issn.1001-4616.2025.04.008
文献标志码:
A
摘要:
视频卫星能获得高空间分辨率的视频信息,为运动目标的检测和分析提供有效数据支撑. 然而,由于卫星视频图像中目标像素比例低、纹理细节不清晰、背景复杂等缺点,从卫星视频中检测运动目标存在很大困难. 为此,本文以YOLOv8为骨干网络,提出了一种基于注意力机制与可变卷积神经网络的卫星视频运动目标检测算法. 首先,设计C2f-DCN模块替换原模型骨干网络中的C2f模块,以提高模型对不同尺度目标的特征提取能力. 其次,在检测头前添加Shuffle Attention轻量级注意力机制,在保证模型计算速度的前提下增强重要特征,加强通道间信息沟通提高模型特征融合能力. 最后,为了提高模型的学习能力和推理效率,采用Inner-CIoU损失函数,并引入辅助边界框概念来解决卫星视频图像中目标像素比例小的问题. 利用SAT-MTB卫星视频影像数据集进行对比实验,实验结果表明本文算法的精确度、召回率、mAP50:95和F1分数分别为75.3%、62.8%、34.9%和68.48,相较于原始YOLOv8n网络,上述指标分别提高了11.6%、4.2%、3.0%和7.44,验证了本文方法的有效性和优越性.
Abstract:
Video satellites can obtain high spatial resolution video information,providing effective data support for the detection and analysis of moving targets. However,due to the disadvantages of low target pixel proportion,unclear texture details,and complex background in satellite video images,there are significant difficulties in detecting moving targets from satellite videos. Thus,on the basis of the backbone network of YOLOv8,this paper proposes a new detection method of motion targets in satellite video based on deformable convolutional neural network and Shuffle Attention. Firstly,a C2f-DCN module is designed to replace the C2f module in the original model backbone network for improving the model's ability of extracting features from targets with different types and different scales. Secondly,a lightweight Shuffle Attention mechanism is added in front of the detection head to strengthen important features while ensuring the computational speed of the model,enhancing information communication between channels,and improving the model's feature fusion ability. Finally,to improve the learning ability and inference efficiency of the model,the Inner-CIoU loss function is adopted,and the concept of auxiliary bounding boxes is introduced for solving the problem of small proportion of target pixels in satellite video images. Comparative experiments are conducted using the SAT-MTB satellite video image dataset,and the experimental results show that the accuracy,recall,mAP50:95,and F1 scores of the algorithm are 75.3%,62.8%,34.9%,and 68.48,respectively. Compared with the original YOLOv8n network,above indexes are improved by 11.6%,4.2%,3.0%,and 7.44. Thus,the effectiveness and superiority of the proposed method is verified.

参考文献/References:

[1]李贝贝,韩冰,田甜,等. 吉林一号视频卫星应用现状与未来发展[J]. 卫星应用,2018(3):23-27.
[2]廖一鸣,万剑华,臧文乾,等. 视频卫星数据的运动车辆提取[J]. 测绘科学,2018,43(4):144-149.
[3]吴昱舟,姚力波,刘勇,等. 基于帧差和背景建模的卫星视频目标检测[J]. 海军航空工程学院学报,2018,33(5):441-446.
[4]SHU M,ZHONG Y,LV P. Small moving vehicle detection via local enhancement fusion for satellite video[J]. International journal of remote sensing,2021,42(19):7189-7214.
[5]吴佳奇,蒋永华,沈欣,等. 决策树弱分类支持的卫星视频运动检测[J]. 武汉大学学报(信息科学版),2019,44(8):1182-1190.
[6]LEI J,DONG Y,SUI H. Tiny moving vehicle detection in satellite video with constraints of multiple prior information[J]. International journal of remote sensing,2021,42(11),4110-4125.
[7]ZHANG J,JIA X,HU J,et al. Moving vehicle detection for remote sensing video surveillance with nonstationary satellite platform[J]. IEEE transactions on pattern analysis and machine intelligence,2022,44(9):5185-5198.
[8]魏业文,李梅,解园琳,等. 基于改进Faster-RCNN的输电线路巡检图像检测[J]. 电力工程技术,2022,41(2):171-178.
[9]刘江,关向雨,温跃泉,等. 基于改进YOLOv4的GIS红外特征识别与温度提取方法[J]. 电力工程技术,2023,42(1):162-168.
[10]周文青,刘刚. 基于深度学习和无人机图像的架空线路缺陷巡检综述[J]. 电力工程技术,2024,43(2):73-82.
[11]LEI L,GUO D. Multitarget detection and tracking method in remote sensing satellite video[J]. Computational intelligence and neuroscience,2021,2021(1):7381909.
[12]CHEN R,LI X,LI S. A lightweight CNN model for refining moving vehicle detection from satellite videos[J]. IEEE access,2020,8:221897-221917.
[13]FENG J,ZENG D,JIA X,et al. Cross-frame key point-based and spatial motion information-guided networks for moving vehicle detection and tracking in satellite videos[J]. ISPRS journal of photogrammetry and remote sensing,2021,177:116-130.
[14]YU C,FENG Z,WU Z,et al. HB-YOLO:An improved YOLOv7 algorithm for dim-object tracking in satellite remote sensing videos[J]. Remote sensing,2023,15(14):3551.
[15]LI S,ZHOU Z,ZHAO M. A multitask benchmark dataset for satellite video:object detection,tracking,and segmentation[J]. IEEE transactions on geoscience and remote sensing,2023,61:5611021.
[16]RAHMAN S,RONY J H,UDDIN J,et al. A. Real-time obstacle detection with YOLOv8 in a WSN using UAV aerial photography[J]. Journal of imaging,2023,9(10):216.
[17]ZHU X,HU H,LIN S,et al. Deformable convnets V2:More deformable better results[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Long Beach,USA:IEEE,2019:9300-9308.
[18]ZHANG Q L,YANG Y B. SA-Net:Shuffle attention for deep convolutional neural networks[C]//Proceedings of IEEE International Conference on Acoustics,Speech and Signal Processing. Toronto,Canada:IEEE,2021:2235-2239.
[19]ZHANG H,XU C,ZHANG S J. Inner-IoU:more effective intersection over union loss with auxiliary bounding box[J/OL]. arXiv,2023,https://doi.org/10.48550/arXiv.2311.02877.
[20]GEVORGYAN Z. SIoU Loss:more powerful learning for bounding box regression[J/OL]. arXiv,2022,https://doi.org/10.48550/arXiv.2205.12740.
[21]WANG C,HE W,NIE Y,et al. Gold-YOLO:efficient object detector via gather-and-distribute mechanism[J/OL]. arXiv,2023,https://doi.org/10.48550/arXiv.2309.11331.
[22]LI Y,HOU Q,ZHENG Z,et al. Large selective kernel network for remote sensing object detection[J/OL]. arXiv,2023,https://doi.org/10.48550/arXiv.2303.09030.
[23]LIU Z,LIN Y,CAO Y,et al. Swin transformer:hierarchical vision transformer using shifted windows[C]//Proceedings of the IEEE International Conference on Computer Vision. Montreal,Canada:IEEE,2021:9992-10002.

备注/Memo

备注/Memo:
收稿日期:2024-10-11.
基金项目:国网江苏省电力有限公司科技项目(J2023121).
通讯作者:张卡,博士,教授,研究方向:摄影测量与遥感. E-mail:zhangka81@126.com
更新日期/Last Update: 2025-08-20