计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (6A): 240700097-5.doi: 10.11896/jsjkx.240700097

• 图像处理&多媒体技术 • 上一篇    下一篇

融合多尺度特征的无人机图像中小目标检测算法

黄红1, 苏菡1,2, 闵鹏1   

  1. 1 四川师范大学计算机科学学院 成都 610101
    2 可视化计算与虚拟现实四川省重点实验室 成都 610101
  • 出版日期:2025-06-16 发布日期:2025-06-12
  • 通讯作者: 苏菡(susuhan@163.com)
  • 作者简介:(2936836202@qq.com)
  • 基金资助:
    人社部留学回国人员择优项目;四川省自然科学基金(2023NSFSC1080,2023NSFSC0210,2023YFS0202);国家林草局国际合作资金项目(护动函(2023)94号);成都市科技计划(2022-YF09-00019-SN);成都大熊猫繁育研究基地(2021CPB-B06)

Small Target Detection Algorithm in UAV Images Integrating Multi-scale Features

HUANG Hong1, SU Han1,2, MIN Peng1   

  1. 1 College of Computer Science,Sichuan Normal University,Chengdu 610101,China
    2 Visual Computing and Virtual Reality Key Laboratory of Sichuan Province,Chengdu 610101,China
  • Online:2025-06-16 Published:2025-06-12
  • About author:SU Han,born in 1979,Ph.D,professor,is a member of CCF(No.A3506M).Her main research interests include computer vision,intelligent information processing and biometric recognition.
  • Supported by:
    Ministry of Human Resources and Social Security’s Returned Overseas Students Selection Program,Sichuan Natural Science Foundation(2023NSFSC1080,2023NSFSC0210,2023YFS0202),National Forestry and Grassland Administration International Cooperation Fund Project (Hudonghan (2023) No. 94),Chengdu Science and Technology Plan(2022-YF09-00019-SN) and Chengdu Research Base of Giant Panda Breeding(2021CPB-B06).

PDF (PC)

102

摘要： 针对无人机航拍图像小目标检测任务中小目标分布过于密集导致互相遮挡产生的漏检误检问题,提出了一种多尺度特征融合的轻量化目标检测方法。首先,提出了多尺度遮挡模块,通过该模块增强网络的多尺度信息提取能力,缩小不同尺度间的语义差异,提高对遮挡小目标的检测性能;其次,提出更加高效的共享检测头策略,该策略将不同尺度的特征信息通过共享卷积共享到不同的检测头,显著降低模型的参数量,实现对模型的轻量化;最后,引入软化非极大值抑制方法来解决传统贪心非极大值抑制在密集遮挡场景下的漏检误检问题,进一步提高了检测精度。在Visdrone-2019和RSOD数据集上评估了改进模型的有效性,相比基准模型,改进模型的平均精度均值分别提升了9.0%和6.0%,模型参数量降低了12.6%。实验结果表明,改进算法在保证轻量化的同时能够提升无人机航拍图像目标检测的精度,能够帮助无人机系统更准确地识别和追踪目标,提高了任务执行的可靠性和效率。

关键词: 无人机航拍, 小目标检测, 深度学习, 非极大值抑制, YOLOv8

Abstract: To address the issues of missed and false detections caused by overly dense distributions of small objects in the task of small object detection in drone aerial imagery,which can lead to mutual occlusion,a lightweight object detection method with multi-scale feature fusion is proposed.Firstly,a multi-scale occlusion module is introduced to enhance the network’s multi-scale information extraction capability,reduce semantic differences between different scales,and improve detection performance for occluded small objects.Secondly,a more efficient shared detection head strategy is proposed,which shares feature information of different scales through shared convolution across different detection heads,significantly reducing the model’s parameter count and achieving model lightweighting.Finally,a softened non-maximum suppression(NMS) method is introduced to solve the pro-blem of missed and false detections in dense occlusion scenarios caused by traditional greedy NMS,further improving detection accuracy.The effectiveness of the improved model is evaluated on the Visdrone-2019 and RSOD datasets,with the mean average precision(mAP) of the improved model increases by 9.0% and 6.0% respectively compared to the baseline model,and the model parameters reduce by 12.6%.Experimental results show that the improved algorithm can enhance the accuracy of object detection in drone aerial imagery while ensuring lightweighting,helping drone systems to identify and track targets more accurately,thereby improving the reliability and efficiency of task execution.

Key words: Drone aerial photography, Small object detection, Deep learning, Non-maximum suppression, YOLOv8

中图分类号: 

  • TP391
[1]WANG S,JIANG H,LI Z,et al.PHSI-RTDETR:A Light-weight Infrared Small Target Detection Algorithm Based on UAV Aerial Photography[J].Drones,2024,8(6):240.
[2]BAI Z,PEI X,QIAO Z,et al.Improved YOLOv7 Target Detection Algorithm Based on UAV Aerial Photography[J].Drones,2024,8(3):104.
[3]ZHU X,LYU S,WANG X,et al.TPH-YOLOv5:ImprovedYOLOv5 based on transformer prediction headfor object detection on drone-captured scenarios[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Piscataway:IEEE,2021:2778-2788.
[4]ZHOU J,SU T,LI K,et al.Small Target-YOLOv5:Enhancing the Algorithm for Small Object Detection in Drone Aerial Imagery Based on YOLOv5[J].Sensors,2023,24(1):134.
[5]ZHANG M,ZHANG B,LIU M,et al.Robust object detection in aerial imagery based on multi-scale detector and soft densely connected[J].IEEE Access,2020,8:92791-92801.
[6]MA J,SHAO W,YE H,et al.Arbitrary-oriented scene text detection via rotation proposals[J].IEEE Transactions on Multimedia,2018,20(11):3111-3122.
[7]ZHAO L L,ZHU M L.MS-YOLOv7:YOLOv7 based on multi-scale for object detection on UAV aerial photography[J].Drones,2023,7(3):188.
[8]LIU W,LU H,FU H,et al.Learning to Upsample by Learning to Sample[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2023:6027-6037.
[9]KANG M,TING C M,TING F F,et al.ASF-YOLO:A Novel YOLO Model with Attentional Scale Sequence Fusion for Cell Instance Segmentation[J].arXiv:2312.06458,2023.
[10]TIAN Z,SHEN C,CHEN H,et al.FCOS:A simple and strong anchor-free object detector[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,44(4):1922-1933.
[11]NEUBECK A,VAN GOOL L.Efficient non-maximum suppression[C]//18th International Conference on Pattern Recognition(ICPR’06).IEEE,2006:850-855.
[12]BODLA N,SINGH B,CHELLAPPAR,et al.Soft-NMS－improving object detection with one line of code[C]//Proceedings of the IEEE International Conference on Computer Vision.2017:5561-5569.
[13]DU D,ZHANG Y,SHI B L,et al.VisDrone-SOT2019:The vision meets drone single object tracking challenge results[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision Workshop(ICCVW).Seoul,Republic of Korea,2019:27-28.
[14]UTTARKABAT S,SUNKARA N D,PATRA B K.RSOD:Efficient Technique for Outlier Detection using Reverse Nearest Neighbors Statistics[C]//2020 4th International Conference on Computational Intelligence and Networks(CINE).2020.
[15]WANG J,YU J,HE Z.ARFP:A novel adaptive recursive feature pyramid for object detection in aerial images[J].Applied Intelligence,2022,52(11):12844-12859.
[16]REDMON J,FARHADI A.Yolov3:An incremental improve-ment[J].arXiv:1804.02767,2018.
[17]CHEN H,LIU H,SUN T,et al.MC-YOLOv5:A Multi-ClassSmall Object Detection Algorithm[J].Biomimetics,2023,8(4):342.
[18]CHEN N,LI Y,YANG Z,et al.LODNU:lightweight object detection network in UAV vision[J].The Journal of Supercomputing,2023,79(9):10117-10138.
[19]CAI Z,HONG Z,YU W,et al.CNXResNet:A Light-weightBackbone based on PP-YOLOE for Drone-captured Scenarios[C]//2023 8th International Conference on Signal and Image Processing(ICSIP).IEEE,2023:460-464.
[20]BOCHKOVSKIY A,WANG C Y,LIAOH Y M.Yolov4:Optimal speed and accuracy of object detection[J].arXiv:2004.10934,2020.
[21]DING W,ZHANG L.Building detection in remote sensing image based on improved YOLOv5[C]//2021 17th International Conference on Computational Intelligence and Security(CIS).IEEE,2021:133-136.
[22]LANG L,XU K,ZHANG Q,et al.Fast and accurate object detection in remote sensing images based on lightweight deep neural network[J].Sensors,2021,21(16):5460.
[23]LIU W,ANGUELOV D,ERHAN D,et al.Ssd:Single shotmultibox detector[C]//Computer Vision－ECCV 2016:14th European Conference,Amsterdam,The Netherlands,October 11-14,2016,Proceedings,Part I 14.Springer International Publishing,2016:21-37.
[24]HUANG W,LI G,CHEN Q,et al.CF2PN:A cross-scale feature fusion pyramid network based remote sensing target detection[J].Remote Sensing,2021,13(5):847.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发