Skip to content

[IEEE TCYB 2021] Unsupervised Change Detection in Multitemporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network

Notifications You must be signed in to change notification settings

ChenHongruixuan/KPCAMNet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Unsupervised Change Detection in Multitemporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network

This is an official implementation of unsupervised multi-class change detection framework KPCAMNet in our IEEE TCYB 2022 paper: Unsupervised Change Detection in Multitemporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network.

Abstract

With the development of Earth observation technology, very-high-resolution (VHR) image has become an important data source of change detection. Nowadays, deep learning methods have achieved conspicuous performance in the change detection of VHR images. Nonetheless, most of the existing change detection models based on deep learning require annotated training samples. In this paper, a novel unsupervised model called kernel principal component analysis (KPCA) convolution is proposed for extracting representative features from multi-temporal VHR images. Based on the KPCA convolution, an unsupervised deep siamese KPCA convolutional mapping network (KPCA-MNet) is designed for binary and multi-class change detection. In the KPCA-MNet, the high-level spatial-spectral feature maps are extracted by a deep siamese network consisting of weight-shared KPCA convolutional layers. Then, the change information in the feature difference map is mapped into a 2-D polar domain. Finally, the change detection results are generated by threshold segmentation and clustering algorithms. All procedures of KPCA-MNet do not require labeled data. The theoretical analysis and experimental results in two binary change detection datasets and one multi-class change detection datasets demonstrate the validity, robustness, and potential of the proposed method.

Network architecture

Get started

Requirements

python==3.6.12
scikit-learn==0.19.0
imageio=2.9.0
numpy==1.13.3
gdal==2.2.2

Datasets

Two binary change detection datasets, WH and QU, and one multi-class change detection dataset, HY, are used in our work. You can download them here for reproducibility or your own research. [GoogleDrive], [BaiduCloud]

Dataset Pre-event image Post-event image Reference Image
WH
QU
HY

Train & infer

python train_KPCAMNet.py

Detection results

  • Predictions on the Hanyang dataset:
Pre-event image Post-event image Reference Prediction
  • The results of quantitative assessments:

Citation

If this code or dataset contributes to your research, please kindly consider citing our paper. We appreciate your support!🙂

@article{Wu2022Unsupervised,
  author={Wu, Chen and Chen, Hongruixuan and Du, Bo and Zhang, Liangpei},
  journal={IEEE Transactions on Cybernetics}, 
  title={Unsupervised Change Detection in Multitemporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network}, 
  year={2022},
  volume={52},
  number={11},
  pages={12084-12098},
  doi={10.1109/TCYB.2021.3086884}
}

Q & A

For any questions, please contact us.

About

[IEEE TCYB 2021] Unsupervised Change Detection in Multitemporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages