L2,1-Norm Regularized Double Non-Negative Matrix Factorization for Hyperspectral Change Detection
Open Access
Abstract
Hyperspectral image (HSI) change detection (CD) is an important technology for identifying surface changes using multi-temporal HSIs. Nevertheless, the high dimensionality of HSIs presents significant challenges for CD tasks, including issues such as lack of robustness and high computational costs in existing methods. To address those issues, this paper proposes an unsupervised simple and effective HSI CD model termed -norm regularized double non-negative matrix factorization (-DNMF). Specifically, the proposed model employs a symmetric double non-negative matrix factorization (NMF) framework to jointly analyze multitemporal HSIs, capturing their common and invariant structural information to construct a shared feature basis. Meanwhile, two non-negative feature weight matrices are learned to generate a differential image matrix that directly reflects the change regions. To enhance robustness against noise, an -norm constraint is imposed on the difference image matrix, ensuring that unchanged areas exhibit near-zero values while changed areas present nonzero values. Finally, comprehensive experiments performed on three benchmark hyperspectral datasets validated the efficacy of the proposed method, which is superior to some state-of-the-art ones regarding detection performance and computational cost.
Keywords
Funding Information
- National Natural Science Foundation of China (62401203, 2023NK2011, 2023NK2002, 2023JJ40333)
- Hunan Provincial Key Research and Development Program (62401203, 2023NK2011, 2023NK2002, 2023JJ40333)
- Hunan Provincial Natural Science Foundation of China (62401203, 2023NK2011, 2023NK2002, 2023JJ40333)
This publication has 42 references indexed in Scilit:
- Refined change detection in heterogeneous low-resolution remote sensing images for disaster emergency responseISPRS Journal of Photogrammetry and Remote Sensing, 2024
- Large kernel convolution application for land cover change detection of remote sensing imagesInternational Journal of Applied Earth Observation and Geoinformation, 2024
- CDasXORNet: Change detection of buildings from bi-temporal remote sensing images as an XOR problemInternational Journal of Applied Earth Observation and Geoinformation, 2024
- Progressive pseudo-label framework for unsupervised hyperspectral change detectionInternational Journal of Applied Earth Observation and Geoinformation, 2024
- Self-supervised learning-based oil spill detection of hyperspectral imagesScience China Technological Sciences, 2022
- Multi-sensor change detection for within-year capture and labelling of forest disturbanceRemote Sensing of Environment, 2022
- Building damage assessment for rapid disaster response with a deep object-based semantic change detection framework: From natural disasters to man-made disastersRemote Sensing of Environment, 2021
- Hyperspectral anomalous change detection based on joint sparse representationISPRS Journal of Photogrammetry and Remote Sensing, 2018
- A post-classification change detection method based on iterative slow feature analysis and Bayesian soft fusionRemote Sensing of Environment, 2017
- SVD based initialization: A head start for nonnegative matrix factorizationPattern Recognition, 2008