Reweighted sparse nonnegative matrix decomposition for hyperspectral unmixing

被引：0

作者：

Jia Q. ^{[1
]}

Liao S. ^{[1
]}

Zhang Z. ^{[1
]}

Yang X. ^{[1
]}

机构：

[1] Rocket Force Engineering University, Xi'an

来源：

Hongwai yu Jiguang Gongcheng/Infrared and Laser Engineering | 2020年 / 49卷

关键词：

Alternating Direction Method of Multipliers (ADMM); Hyperspectral image unmixing; Nonnegative Matrix Factorization (NMF); Reweighted; Sparse constraint;

D O I：

10.3788/IRLA20200152

中图分类号：

学科分类号：

摘要：

In recent years, Nonnegative Matrix Factorization (NMF) methods for hyperspectral image unmixing have attracted widespread attention. However, due to the non -convexity of NMF problem, it cannot guarantee the uniqueness of the solution, and it is easy to fall into local minima. In order to reduce the solution space of NMF problem and improve the unmixing accuracy, a new method of reweighted sparse NMF (ARSNMF) was proposed. Firstly, considering the sparsity of abundance matrix, the sparse constraint was added to the NMF model. Then, considering that the calculation of the problem was complex and not easy to be optimized, it was converted into a form of reweighted sparse constraint, which not only achieved the sparse effect, but also solved the problem that was difficult to solve. In order to improve the convergence speed of the algorithm, the Alternating Direction Method of Multipliers (ADMM) was used to optimize the model, and the objective function was divided into several sub-problems for independent solution. Experiments based on simulation data and real data verify the effectiveness of the proposed algorithm. © 2020, Editorial Board of Journal of Infrared and Laser Engineering. All right reserved.

引用

共 18 条

[1] Zhang Bing, Gao Lianru, Hyperspectral Image Classification and Target Detection, pp. 1-102, (2011)
[2] He Fang, Wang Rong, Yu Qiang, Et al., Extraction of hyperspectral image features by weighted space spectral locally retained Projection, Optical and Precision Engineering, 25, 1, pp. 263-273, (2017)
[3] Yuan Jing, Zhang Yujin, Gao Fangping, Review of linear hyperspectral miscible models, Journal of Infrared and Millimeter Waves, 37, 5, pp. 43-61, (2018)
[4] Gan Shizhong, Xiao Zhitao, Chen Lei, Et al., Multi-target hyperspectral image unmixing algorithm based on high-order nonlinear model, Infrared and Laser Engineering, 48, 10, (2019)
[5] Tang Yi, Yan YongJian, He Mi, Et al., Spectral decom-missioning of successive hyperspectral images with L<sub>1/2</sub> regularization, Infrared and Laser Engineering, 48, 7, (2019)
[6] Lee D D, Seung H S., Learning the parts of objects by nonnegative matrix factorization, Nature, 401, 6755, pp. 788-791, (1999)
[7] Qian Y, Jia S, Zhou J, Et al., Hyperspectral unmixing via L<sub>1/2</sub> sparsity-constrained nonnegative matrix factorization, IEEE Transactions on Geoence and Remote Sensing, 49, 11, pp. 4282-4297, (2011)
[8] Lu X, Wu H, Yuan Y, Et al., Manifold regularized sparse NMF for hyperspectral unmixing, IEEE Transactions on Geoscience & Remote Sensing, 51, 5, pp. 2815-2826, (2013)
[9] Zuoyu Z, Shouyi L, Hexin Z, Et al., Bilateral filter regularized l2 sparse nonnegative matrix factorization for hyperspectral unmixing, Remote Sensing, 10, 6, (2018)
[10] Zhang S, Li J, Li H C, Et al., Spectral-spatial weighted sparse regression for hyperspectral image unmixing, IEEE Transactions on Geoscience & Remote Sensing, 56, 6, pp. 3265-3276, (2018)

← 1 2 →