计算机科学

计算机科学 ›› 2026, Vol. 53 ›› Issue (6A): 250600228-8.doi: 10.11896/jsjkx.250600228

• 图像处理&多媒体技术 • 上一篇    下一篇

面向景象匹配导航系统基准图的语义感知主动学习方法

单程程1, 梅春1, 李未亭1, 郭原源2, 钱伟行2, 熊智3   

  1. 1 国家电投集团江苏海上风力发电有限公司 江苏 盐城 224000
    2 南京师范大学电气与自动化工程学院 南京 210023
    3 南京航空航天大学自动化学院 南京 211106
  • 出版日期:2026-06-16 发布日期:2026-06-12
  • 通讯作者: 单程程(374468731@qq.com)
  • 基金资助:
    国家自然科学基金(62373194)

Semantic Perception Active Learning Method for the Datum Map of Scene Matching Navigation System

SHAN Chengcheng1, MEI Chun1, LI Weiting1, GUO Yuanyuan2, QIAN Weixing2, XIONG Zhi3   

  1. 1 SPIC Jiangsu Offshore Wind Power Co.,Ltd.,Yancheng,Jiangsu 224000,China
    2 School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210023,China
    3 College of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China
  • Published:2026-06-16 Online:2026-06-12
  • About author:SHAN Chengcheng,born in 1996,bachelor,engineer,registered safety engineer.His main research interests include offshore wind power industry engineering construction,production and operation maintenance,and transportation vessel management.
  • Supported by:
    National Natural Science Foundation of China(62373194).

PDF (PC)

28

摘要： 航空遥感目标检测技术所获取的高精度基准图语义信息,可有效提升景象匹配导航系统的感知维度。由于遥感图像尺度大、目标密度高且标注成本昂贵,限制了高性能检测模型的训练与应用。针对景象匹配场景的语义目标检测中面临的源域与目标域数据分布差异及目标域标注数据不足的问题,提出了一种高效的主动学习方法来优化目标域样本的选择。该方法利用源域已标注数据和目标域未标注数据,通过主动学习策略从目标域中选取具有高信息量的样本进行人工标注,从而弥补目标域数据标注不足带来的影响。文中设计了3种主动学习评分函数——一致性评分、判别器评分和余弦差异评分,分别从检测框预测不一致性、领域归属概率以及特征差异角度等方面来评估目标域样本的标注价值。同时,还构建了一套针对目标检测任务的评估框架,考虑每张图像的整体标注效果并量化每个检测框的标注成本。实验证明,该最佳方法能够在目标域上将主动学习的目标检测性能提高约3.29%,同时将目标域所需的标注边界框数量减少约17.6%。所提方法在标注资源有限的条件下,有效提升了目标检测性能与领域迁移能力,能够应对源域与目标域之间因分布差异导致的模型性能下降问题,为遥感场景中的跨域目标检测提供了新的解决思路,为构建高精度的语义基准图提供了可靠的数据支持,进而提升了景象匹配导航的精度和可靠性。

关键词: 景象匹配, 目标检测, 主动学习, 领域自适应, 判别器

Abstract: The high-precision datum semantic information obtained by aerial remote sensing target detection technology can effectively improve the perception dimension of scene matching navigation system.Due to the large scale,high target density and high annotation cost of remote sensing images,the training and application of high-performance detection models are limited.In order to solve the problems of difference in data distribution between source domain,target domain and insufficient annotation data in target domain in semantic object detection in scene matching scenes,an efficient active learning method is proposed to optimize the selection of target domain samples.This method uses the labeled data in the source domain and the unlabeled data in the target domain to select high-information samples from the target domain for manual labeling through the active learning strategy,so as to make up for the impact of insufficient data labeling in the target domain.This paper proposes three active learning scoring functions,namely consistency score,discriminator score and cosine difference score,which are designed to evaluate the labeling value of target domain samples from the perspectives of detection frame prediction inconsistency,domain belonging probability and feature difference,respectively.At the same time,a set of evaluation framework for object detection tasks is constructed,which considers the overall annotation effect of each image and quantifies the annotation cost of each detection frame.Experiments show that the proposed method can improve the object detection performance of active learning by about 3.29% on the target domain,and reduce the number of labeled bounding boxes required by the target domain by about 17.6%.Under the condition of limited annotation resources,this method can effectively improve the performance of object detection and domain migration,and can cope with the problem of model performance degradation caused by distribution differences between source and target domains,provide a new solution for cross-domain object detection in remote sensing scenes,and provide reliable data support for the construction of high-precision semantic datum maps,thereby improving the accuracy and reliability of scene matching navigation.

Key words: Scene matching, Object detection, Active learning, Domain adaptation, Discriminators

中图分类号: 

  • TN911.73
[1] KHODABANDEH M,SMITH J,BROWN T,et al.A resilient learning method for domain-adaptive object detection [C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2019:480-490.
[2] LI Y,ZHANG P,GAO R,et al.Cross-domain teacher adaptation for object detection [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2022:7581-7590.
[3] ZHUANG C,HAN P,WANG M,et al.IFAN:Instance-level alignment networks for adaptive object detection [C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020:13122-13129.
[4] ZHU J,MA R,LI D,et al.Unpaired image translation using cycle-consistent generative adversarial networks [C]//Proceedings of the IEEE International Conference on Computer Vision.2017.
[5] MENKE M,JOHNSON L,REED B.AWADA:Foreground-aware adversarial learning for cross-domain object detection [EB/OL].SSRN 4272713,2023.
[6] WANG R,LIU J,ZHANG D.Extending version-space learning to multi-label active learning with class imbalance [J].Pattern Recognition,2023,142:109690.
[7] SCHNEEGANS J,LEE A,HUGHES P.An empirical study of active learning methods for object detection [C]//Proceedings of the Workshop on Interactive Adaptive Learning.2022:49-67.
[8] LI Y,DU J,ZHAO S,et al.Deep active learning strategies for object detection [J].Information Sciences,2021,579:418-433.
[9] HAUSSMANN E,MARTINEZ R,COLEMAN F,et al.Scalable strategies for active learning in object detection [C]//Proceedings of the IEEE Intelligent Vehicles Symposium(IV).2020:1430-1435.
[10] LIANG S,SUN C,ZHENG K.Improving reliability of neuralnetworks in out-of-distribution image detection [EB/OL].arXiv:1706.02690,2017.
[11] CHEN S,LI E,WANG Y.Semi-supervised active learning forenhanced object detection [J].Electronics,2023,12(2):375.
[12] SU J,XU M,QIAN D,et al.Active adversarial adaptation across domains [C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.2020:739-748.
[13] PENG J,SHI P,XUE S,et al.TPM:Multi-object tracking via tracklet-plane association [J].Pattern Recognition,2020,107480.
[14] XIE B,XIONG L,LIANG N,et al.Active learning for domain adaptation using energy-based modeling [C]//Proceedings of the AAAI Conference on Artificial Intelligence.2022:8708-8716.
[15] SOHN K,WILSON T,COOPER H,et al.CReST:Class-rebalancing self-training framework for semi-supervised learning with imbalance [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:10857-10866.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市两江新区洪湖西路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