Manifold regularized sparse constraint nonnegative matrix factorization with superpixel algorithm for hyperspectral unmixing

doi:10.11772/j.issn.1001-9081.2019030534

Journal of Computer Applications ›› 2019, Vol. 39 ›› Issue (10): 3100-3106.DOI: 10.11772/j.issn.1001-9081.2019030534

• Frontier & interdisciplinary applications • Previous Articles

Manifold regularized sparse constraint nonnegative matrix factorization with superpixel algorithm for hyperspectral unmixing

LI Denggang¹, CHEN Xiangxiang², LI Huali³, WANG Zhongmei¹

1. College of Traffic Engineering, Hunan University of Technology, Zhuzhou Hunan 410082, China;
2. MediaTek(Shenzhen) Incorporated, Shenzhen Guangdong 518063, China;
3. College of Electrical and Information Engineering, Hunan University, Changsha Hunan 410002, China

Received:2019-04-02 Revised:2019-05-22 Online:2019-05-31 Published:2019-10-10
Supported by:
This work is partially supported by the Scientific Research Project of Hunan Provinical Education Department (18C0513).

基于超像素的流形正则化稀疏约束NMF混合像元分解算法

李登刚¹, 陈香香², 李华丽³, 王忠美¹

1. 湖南工业大学交通工程学院, 湖南株洲 412007;
2. 联发软件设计(深圳)有限公司, 广东深圳 518063;
3. 湖南大学电气与信息工程学院, 长沙 410082

通讯作者: 陈香香
作者简介:李登刚(1991-),男,湖南永州人,讲师,硕士,主要研究方向:高光谱图像处理;陈香香(1992-),女,湖南邵阳人,工程师,硕士,主要研究方向:数字图像处理、模式识别;李华丽(1984-),女,湖北孝感人,副教授,博士,主要研究方向:高光谱图像处理、模式识别;王忠美(1984-),男,湖北荆州人,讲师,博士,主要研究方向:高光谱图像处理、模式识别。
基金资助:
湖南省教育厅科学研究项目（18C0513）。

Abstract

Abstract: For the problems such as poor unmixing results and sensitivity to noise of traditional Nonnegative Matrix Factorization (NMF) applied to hyperspectral unmixing, a Manifold Regularized Sparse NMF with superpixel (MRS-NMF) algorithm for hyperspectral unmixing was proposed. Firstly, the manifold structure of hyperspectral image was constructed by superpixel segmentation based on entropy. The original image was divided into k-superpixel blocks, and the data points in each superpixel block with same property were labeled the same label. Weight matrices were defined between any two data points with the similar label in a superpixel block, and then the weight matrices were applied to the objective function of NMF to construct the manifold regularization constraint. Secondly, a quadratic parabola function was added to the objective function to complete the sparse constraint. Finally, the multiplicative iterative update rule was used to solve the objective function to obtain the solution formulas of endmember matrix and abundance matrix. At the same time, maximum iteration times and tolerate error threshold were set to get the final results by iterative operation. The proposed method makes full use of spectral and spatial information of hyperspectral images. Experimental results show that on synthetic data the unmixing accuracies of endmember and abundance based on proposed MRS-NMF are 0.016-0.063 and 0.01-0.05 respectively higher than those based on traditional methods like Graph-regularized L1/2-Nonnegative Matrix Factorization (GLNMF), L1/2NMF and Vertex Component Analysis-Fully Constrained Least Squares (VCA-FCLS); while on real hyperspectral images, the average unmixing accurary of endmember based on proposed MRS-NMF is 0.001-0.0437 higher than that of traditional GLNMF, Vertex Component Analysis (VCA) and Minimum Volume Constrained Nonnegative Matrix Factorization (MVCNMF). This proposed algorithm improves the accuracy of unmixing effectively with good robustness to noise.

Key words: hyperspectral unmixing, Nonnegative Matrix Factorization (NMF), superpixel segmentation, manifold regularization, sparseness

摘要： 针对传统非负矩阵分解（NMF）法用于高光谱图像混合像元分解时产生的分解结果精度不高、对噪声敏感等问题，提出一种基于超像素的流形正则化稀疏约束NMF混合像元分解算法——MRS-NMF。首先，通过基于熵率的超像素分割来构造高光谱图像的流形结构，把原图像分割为k个超像素块并把每个超像素块中具有相似性质的数据点标上相同的标签，定义像素块内有相同标签的任意两个数据点之间的权重矩阵，然后将权重矩阵应用于NMF的目标函数中以构造出流形正则化约束项；第二，在目标函数中添加二次抛物线函数以完成稀疏约束；最后，采用乘法迭代更新法则求解目标函数以得到端元矩阵和丰度矩阵的求解公式，同时设置最大迭代次数和容忍误差阈值，迭代运算得到最终结果。该方法有效利用了高光谱图像的光谱和空间信息。实验结果表明，在模拟的高光谱数据中，与传统的流形稀疏约束的非负矩阵分解（GLNMF）、L1/2-NMF和顶点成分分析-全约束最小二乘法（VCA-FCLS）等方法相比，MRS-NMF可以提高0.016~0.063的端元分解精度和0.01~0.05的丰度分解精度；而在真实的高光谱图像中，MRS-NMF较传统的GLNMF、顶点成分分析法（VCA）、最小体积约束的非负矩阵分解（MVCNMF）等方法可以平均提高0.001~0.0437的端元分解精度。所提MRS-NMF算法有效地提高了混合像元分解的精度，同时具有较好的抗噪性能。

关键词: 混合像元分解, 非负矩阵分解, 超像素分割, 流形正则化, 稀疏性

CLC Number:

TP751.1

LI Denggang, CHEN Xiangxiang, LI Huali, WANG Zhongmei. Manifold regularized sparse constraint nonnegative matrix factorization with superpixel algorithm for hyperspectral unmixing[J]. Journal of Computer Applications, 2019, 39(10): 3100-3106.

李登刚, 陈香香, 李华丽, 王忠美. 基于超像素的流形正则化稀疏约束NMF混合像元分解算法[J]. 计算机应用, 2019, 39(10): 3100-3106.

References

[1] 徐晨光, 邓承志, 朱华生. 近似稀疏约束的多层非负矩阵分解高光谱解混[J]. 红外与激光工程, 2018, 47(11):257-265. (XU C G, DENG C Z, ZHU H S. Approximate sparse regularized multilayer NMF for hyperspectral unmixing[J]. Infrared and Laser Engineering, 2018, 47(11):257-265.)
[2] 王瀛, 何欣, 左方. 基于最大整体包容度约束非负矩阵分解的高光谱遥感图像混合像元分析算法[J]. 光子学报, 2018, 47(3):136-144. (WANG Y, HE X, ZUO F. Mixed data analysis algorithm based on maximum overall coverage constraint nonnegative matrix factorization for hyperspectral image[J]. Acta Photonica Sinica, 2018, 47(3):136-144.)
[3] NASCIMENTO J M P, DIAS J M B. Vertex component analysis:a fast algorithm to unmix hyperspectral data[J]. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(4):898-910.
[4] BOARDMAN J W, KRUSE F A, GREEN R O. Mapping target signatures via partial unmixing of AVIRIS data[EB/OL].[2019-01-10]. http://www.hgimaging.com/PDF/boardman95.PDF.
[5] WINTER M E. N-FINDR:an algorithm for fast autonomous spectral end-member determination in hyperspectral data[J]. Proceedings of the SPIE, 1999, 3753:266-275.
[6] PLAZA A, MARTINEZ P, PEREZ R, et al. Spatial/spectral endmember extraction by multidimensional morphological operations[J]. IEEE Transactions on Geoscience and Remote Sensing, 2002, 40(9):2025-2041.
[7] HEINZ D C, CHANG C. Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2001, 39(3):529-545.
[8] WANG J, CHANG C. Applications of independent component analysis in endmember extraction and abundance quantification for hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(9):2601-2616.
[9] MIAO L, QI H. Endmember extraction from highly mixed data using minimum volume constrained nonnegative matrix factorization[J]. IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(3):765-777.
[10] WANG N, DU B, ZHANG L. An endmember dissimilarity constrained non-negative matrix factorization method for hyperspectral unmixing[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2013, 6(2):554-569.
[11] QIAN Y, JIA S, ZHOU J, et al. Hyperspectral unmixing via L_1/2 sparsity constrained nonnegative matrix factorization[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(11):4282-4297.
[12] WU C, SHEN C. Spectral unmixing using sparse and smooth nonnegative matrix factorization[C]//Proceedings of 21st International Conference on Geoinformatics. Piscataway:IEEE, 2013:1-5.
[13] LU X, WU H. Manifold regularized sparse NMF for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 51(5):2815-2826.
[14] RAJABI R, GHASSEMIAN H. Sparsity constrained graph regularized NMF for spectral unmixing of hyperspectral data[J]. Journal of the Indian Society of Remote Sensing, 2015, 43(2):269-278.
[15] CAI D, HE X, HAN J, et al. Graph regularized nonnegative matrix factorization for data representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(8):1548-1560.
[16] BELKIN M, NIYOGI P. Laplacian eigenmaps and spectral techniques for embedding and clustering[C]//Proceedings of the 14th International Conference on Neural Information Processing Systems. Cambridge, MA:MIT Press, 2001:585-591.
[17] LEE D D, SEUNG S H. Learning the parts of objects by non-negative matrix factorization[J]. Nature, 1999, 401(6755):788-791.
[18] LIU M Y, TUZEL O, RAMALINGAM S, et al. Entropy rate superpixel segmentation[C]//Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE, 2011:2097-2104.
[19] LI D G, LI S T, LI H L. Hyperspectral image unmixing based on sparse and minimum volume constrained nonnegative matrix factorization[C]//Proceedings of the 6th Chinese Conference on Pattern Recognition. Berlin:Springer, 2014:44-52.
[20] CHANG C, DU Q. Estimation of number of spectrally distinct signal sources in hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2004, 42(3):608-619.

Manifold regularized sparse constraint nonnegative matrix factorization with superpixel algorithm for hyperspectral unmixing

基于超像素的流形正则化稀疏约束NMF混合像元分解算法

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics

[1]	WANG Jinkai, JIA Xu. Vein recognition algorithm based on Siamese nonnegative matrix factorization with transferability [J]. Journal of Computer Applications, 2021, 41(3): 898-903.
[2]	WANG Jinkai, JIA Xu. Vehicle face recognition algorithm based on NMF with weighted and orthogonal constraints [J]. Journal of Computer Applications, 2020, 40(4): 1050-1055.
[3]	LIU Ying, LIANG Nannan, LI Daxiang, YANG Fanchao. Hyperspectral image unmixing algorithm based on spectral distance clustering [J]. Journal of Computer Applications, 2019, 39(9): 2541-2546.
[4]	CHEN Shanxue, CHU Chengquan. Hyperspectral unmixing based on sparse and orthogonal constrained non-negative matrix factorization [J]. Journal of Computer Applications, 2019, 39(8): 2276-2280.
[5]	YANG Liangdong, YANG Zhixia. Incremental robust non-negative matrix factorization with sparseness constraints and its application [J]. Journal of Computer Applications, 2019, 39(5): 1275-1281.
[6]	XIANG Wen, ZHANG Ling, CHEN Yunhua, JI Qiumin. Single image super resolution algorithm based on structural self-similarity and deformation block feature [J]. Journal of Computer Applications, 2019, 39(1): 275-280.
[7]	CAO Dawei, HE Chaobo, CHEN Qimai, LIU Hai. Short text clustering algorithm based on weighted kernel nonnegative matrix factorization [J]. Journal of Computer Applications, 2018, 38(8): 2180-2184.
[8]	XIAO Wen, HU Juan. Performance analysis of frequent itemset mining algorithms based on sparseness of dataset [J]. Journal of Computer Applications, 2018, 38(4): 995-1000.
[9]	YUAN Zhengwu, CHEN Ran. Collaborative filtering recommendation algorithm based on multi-level hybrid similarity [J]. Journal of Computer Applications, 2018, 38(3): 633-638.
[10]	JIA Xu, SUN Fuming, LI Haojie, CAO Yudong. Image feature extraction method based on improved nonnegative matrix factorization with universality [J]. Journal of Computer Applications, 2018, 38(1): 233-237.
[11]	YU Keqin, WU Yingbo, LI Shun, JIANG Jiacheng, XIANG De, WANG Tianhui. Tensor factorization recommendation algorithm based on context similarity of mobile user [J]. Journal of Computer Applications, 2017, 37(9): 2531-2535.
[12]	LI Xuejun, ZHANG Kaihua, SONG Huihui. Unsupervised video segmentation by fusing multiple spatio-temporal feature representations [J]. Journal of Computer Applications, 2017, 37(11): 3134-3138.
[13]	JIN Liang, YU Jiong, YANG Xingyao, LU Liang, WANG Yuefei, GUO Binglei, Liao Bin. Video recommendation algorithm based on clustering and hierarchical model [J]. Journal of Computer Applications, 2017, 37(10): 2828-2833.
[14]	SUN Jing, CAI Xibiao, SUN Fuming. Multi-constraint nonnegative matrix factorization algorithm based on feature fusion [J]. Journal of Computer Applications, 2017, 37(10): 2834-2840.
[15]	XIAO Xiaoli, QIAN Yali, LI Danjiang, TAN Liubin. Clustering recommendation algorithm based on user interest and social trust [J]. Journal of Computer Applications, 2016, 36(5): 1273-1278.