广东工业大学学报

• •    

半监督遥感图像建筑物变化检测算法

柴文光, 罗崇熙   

  1. 广东工业大学 计算机学院, 广东 广州 510006
  • 收稿日期:2024-03-26 出版日期:2024-09-27 发布日期:2024-09-27
  • 作者简介:柴文光(1969–)，男，副教授，博士，主要研究方向为人工智能与应用技术、计算机系统结构及高性能计算，E-mail：chaiwg@gdut.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(61772143)；广东省重点领域研发计划项目(2021B0101220006)

Semisupervised Remote Sensing Image Building Change Detection Algorithm

Chai Wen-guang, Luo Chong-xi   

  1. School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2024-03-26 Online:2024-09-27 Published:2024-09-27

HTML

PDF (PC)

45

摘要： 建筑物的变化检测在遥感图像处理和模式识别领域中具有重要意义，但在深度学习算法的应用中，数据标注一直是一个显著的挑战，特别是在变化检测的场景下。因此，针对基于深度学习的变化检测算法中数据标注的难题，本文提出了一种半监督学习方法。该方法采用了融合双时相特征的孪生网络来进行特征提取，并构建了一个教师-学生网络框架以实施模型的半监督训练。为了进一步提升半监督变化检测的准确度，本文在深度特征上引入了随机扰动，以此来实现一致性正则化。此外，在图像深度特征的层面上，本文还提出通过捕获双时相图像特征的差异来形成决策边界以区分双时相图变化的方法。该方法在Levir-CD和WHU_Building两个公开数据集上分别实现了83.04%和85.57%的交并比(Intersection over Union, IoU)。实验结果表明，本文提出的方法在使用少量标注数据的前提下，能够达到与全监督训练相近的性能。

关键词: 遥感图像, 变化检测, 一致性正则化, 半监督学习

Abstract: Change detection in buildings holds significant importance in the fields of remote sensing image processing and pattern recognition. However, data annotation has always been a prominent challenge in the application of deep learning algorithms, especially in change detection scenarios. To address the data annotation challenges in deep learning-based change detection algorithms, this paper proposes an innovative semi-supervised learning method. This method employs a Siamese network that fuses bi-temporal features for feature extraction and constructs a teacher-student network framework for semi-supervised model training. To further enhance the accuracy of semi-supervised change detection, this paper introduces random perturbations in deep features to achieve consistency regularization. Additionally, on the level of image deep features, the paper proposes a method for forming decision boundaries by capturing differences in bi-temporal image features to distinguish changes in bi-temporal images. This method achieved Intersection over Union (IoU) scores of 83.04% and 85.57% on the Levir-CD and WHU Building datasets, respectively. Experimental results show that the proposed method can achieve performance levels close to fully supervised training with a limited amount of labeled data.

Key words: remote sensing images, change detection, consistency regularization, semi-supervised learning

中图分类号: 

  • TP391.4
[1] ZHANG C, YUE P, TAPETE D, et al. A deeply supervised image fusion network for change detection in high resolution bi-temporal remote sensing images [J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 166: 183-200.
[2] RODRIGO C D, BERTRAND L S, ALEXANDREl B. Fully convolutional siamese networks for change detection[C]//2018 25th IEEE International Conference on Image Processing (ICIP). Athens: IEEE, 2018: 4063-4067.
[3] CHEN P, ZHANG B, HONG D, et al. FCCDN: Feature constraint network for VHR image change detection [J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 187: 101-119.
[4] 张枫幸, 黄健, 李浩. 轻量级双侧输入D-WNet航空影像建筑物变化检测[J]. 激光与光电子学进展, 2024, 61(8): 3.
ZHANG F X, HUANG J, LI H. Light-weight bilateral input D-WNet for building change detection in aerial imagery [J]. Laser & Optoelectronics Progress, 2024, 61(8): 3.
[5] TARVAINEN A, VALPOLA H. Mean teachers are better role models: weight-averaged consistency targets improve semi-supervised deep learning results [J]. Advances InNeural Information Processing Systems, 2017, 30: 1195-1204.
[6] SOHN K, BERTHELOT D, CARLINI N, et al. Fixmatch: Simplifying semi-supervised learning with consistency and confidence [J]. Advances in Neural Information Processing Systems, 2020, 33: 596-608.
[7] YANG L, QI L, FENG L, et al. Revisiting weak-to-strong consistency in semi-supervised semantic segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 7236-7246.
[8] BANDARA W G C, PATEL V M. Revisiting consistency regularization for semi-supervised change detection in remote sensing images[EB/OL]. arXiv: 2204.08454 (2022-4-18) [2024-4-16]. http://arxiv.org/abs/2204.08454.
[9] XIE E, WANG W, YU Z, et al. SegFormer: simple and efficient design for semantic segmentation with transformers [J]. Advances in Neural Information Processing Systems, 2021, 34: 12077-12090.
[10] LIN T Y, DOLLÁR P, GIRSHICK. et al. Feature pyramid networks for object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Hawaii: IEEE, 2017: 2117-2125.
[11] LOSHCHILOV I, HUTTER F. Decoupled weight decay regularization[EB/OL]. arXiv: 1711.05101 (2017-11-14) [2024-4-16]. https://arxiv.org/abs/1711.05101.
[12] FANG S, LI K, LI Z. Changer: feature interaction is what you need for change detection[J/OL]. arXiv: 2209.08290v1(2022-09-17)[2024-01-20]. https://github.com/likyoo/open-cd.
[13] MITTAL S, TATARCHENKO M, BROX T. Semi-supervised semantic segmentation with high-and low-level consistency [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 43(4): 1369-1379.
[14] PENG D, LORENZO B, ZHANG Y J, et al. SemiCDNet: a semisupervised convolutional neural network for change detection in high resolution remote-sensing images [J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(7): 5891-5906.
[15] FANG S, LI K, SHAO J. et al. SNUNet-CD: a densely connected Siamese network for change detection of VHR images [J]. IEEE Geoscience and Remote Sensing Letters, 2021, 19: 1-5.
[16] CODEGONI A, LOMBARDI G, FERRARI A. TINYCD: a (not so) deep learning model for change detection [J]. Neural Computing and Applications, 2023, 35(11): 8471-8486.
[17] BANDARA W G C, PATEL V M. A transformer-based siamese network for change detection[C]//IGARSS 2022-2022 IEEE International Geoscience and Remote Sensing Symposium. Malaysia: IEEE, 2022: 207-210.
[18] KOLESNIKOV A, BEYER L, ZHAI X, et al. Big transfer (bit): general visual representation learning[C]//Computer Vision-ECCV. Glasgow: ECCV, 2020: 491-507.
[19] SUN C, WU J, CHEN H, et al. SemiSANet: a semi-supervised high-resolution remote sensing image change detection model using Siamese networks with graph attention [J]. Remote Sensing, 2022, 14(12): 2801.
[20] SUN C, CHEN H, DU C, et al. SemiBuildingChange: a semi-supervised high-resolution remote sensing image building change detection method with a pseudo bitemporal data generator [J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 1-19.
[1] 徐锦华, 李斯. 结合GAN损失与预训练模型的半监督SPECT重建方法[J]. 广东工业大学学报, 2024, 41(0): 0-.
[2] 杨运龙, 梁路, 滕少华. 一种双路网络语义分割模型[J]. 广东工业大学学报, 2022, 39(01): 63-70.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发