S 3 Net: Superpixel-Guided Self-Supervised Learning Network for Multitemporal Image Change Detection
- 1 August 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 20 (1545598X) , 1-5
Abstract
Deep learning (DL) have recently achieved outstanding performance in change detection of multitemporal images. However, most existing DL-based change detection methods still suffer from the problem of insufficient labeled training samples. To overcome this limitation, an unsupervised superpixel-guided self-supervised learning network (S3Net) is proposed for detecting changes occurred on the land surface. By performing principal component analysis on two input images, a triple-channel pseudo-color image containing the main information of both images is first generated, which is used for superpixel segmentation to produce homogeneous image objects. Then, a siamese network composing of two identical subnetworks with shared weight based on transfer learning is trained for pretext task in a self-supervised learning way, aiming to obtain multiscale object-level spatial feature difference images. On this basis, a high-quality difference image is generated by incorporating the pixel-level and object-level difference information using a simple weighted fusion strategy, which can be analyzed by thresholding to produce the final binary change map. The experimental results on four real-world datasets from different sensors show that the proposed approach can obtain superior performance in comparison with several state-of-the-art change detection methods, which further demonstrates its effectiveness and practicability. We make our data and code publicly available (https://github.com/OMEGA-RS/S3Net_CD).Keywords
Funding Information
- National Natural Science Foundation of China (62103311)
- National Natural Science Foundation of China (62006188)
This publication has 19 references indexed in Scilit:
- Oil Spill Detection via Multitemporal Optical Remote Sensing Images: A Change Detection PerspectiveIEEE Geoscience and Remote Sensing Letters, 2017
- A Deep Convolutional Coupling Network for Change Detection Based on Heterogeneous Optical and Radar ImagesIEEE Transactions on Neural Networks and Learning Systems, 2016
- Automatic Change Detection in Synthetic Aperture Radar Images Based on PCANetIEEE Geoscience and Remote Sensing Letters, 2016
- Deep Learning for Remote Sensing Data: A Technical Tutorial on the State of the ArtIEEE Geoscience and Remote Sensing Magazine, 2016
- Deep Residual Learning for Image Recognition2016
- A Superresolution Land-Cover Change Detection Method Using Remotely Sensed Images With Different Spatial ResolutionsIEEE Transactions on Geoscience and Remote Sensing, 2016
- Change detection from remotely sensed images: From pixel-based to object-based approachesISPRS Journal of Photogrammetry and Remote Sensing, 2013
- SLIC Superpixels Compared to State-of-the-Art Superpixel MethodsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2012
- A Multilevel Parcel-Based Approach to Change Detection in Very High Resolution Multitemporal ImagesIEEE Geoscience and Remote Sensing Letters, 2008
- A Threshold Selection Method from Gray-Level HistogramsIEEE Transactions on Systems, Man, and Cybernetics, 1979