Fast and Robust Self-Representation Method for Hyperspectral Band Selection
Authors:
Open Access
- 25 August 2017
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 10 (11) , 5087-5098
Abstract
In this paper, a fast and robust self-representation (FRSR) method is proposed to select a proper band subset from hyperspectral imagery (HSI). The FRSR assumes the separability structure of the HSI band set and transforms the problem of separable nonnegative matrix factorization into the robust self-representation (RSR) model. Then, the FRSR incorporates structured random projections into the RSR model to improve computational efficiency. The solution of FRSR is formulated into optimizing a convex problem and the augmented Lagrangian multipliers are adopted to estimate the proper factorization localizing matrix in the FRSR. The selected band subset is constituted with the bands corresponding to the r largest diagonal entries of the factorization localizing matrix. The experimental results show that FRSR outperforms state-of-the-art techniques in classification accuracy with lower computational cost.Keywords
Funding Information
- National Science Foundation of China (41671342, U1609203, 41401389)
- China Postdoctoral Science Foundation (2015M570668, 2016T90732)
- Public Projects of Zhejiang Province (2016C33021)
- K. C. Wong Magna Fund in Ningbo University
This publication has 46 references indexed in Scilit:
- Ensemble manifold regularized sparse low-rank approximation for multiview feature embeddingPattern Recognition, 2015
- A New Band Selection Method for Hyperspectral Image Based on Data QualityIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014
- Fast and Robust Recursive Algorithmsfor Separable Nonnegative Matrix FactorizationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
- Advances in Hyperspectral Image Classification: Earth Monitoring with Statistical Learning MethodsIEEE Signal Processing Magazine, 2013
- Robustness Analysis of Hottopixx, a Linear Programming Model for Factoring Nonnegative MatricesSIAM Journal on Matrix Analysis and Applications, 2013
- Clustering-based hyperspectral band selection using sparse nonnegative matrix factorizationJournal of Zhejiang University SCIENCE C, 2011
- Finding Structure with Randomness: Probabilistic Algorithms for Constructing Approximate Matrix DecompositionsSIAM Review, 2011
- K-nearest neighborScholarpedia, 2009
- Band Selection for Hyperspectral Image Classification Using Mutual InformationIEEE Geoscience and Remote Sensing Letters, 2006
- A joint band prioritization and band-decorrelation approach to band selection for hyperspectral image classificationIEEE Transactions on Geoscience and Remote Sensing, 1999