基于稀疏和正交约束非负矩阵分解的高光谱解混

doi:10.11772/j.issn.1001-9081.2019010105

计算机应用 ›› 2019, Vol. 39 ›› Issue (8): 2276-2280.DOI: 10.11772/j.issn.1001-9081.2019010105

基于稀疏和正交约束非负矩阵分解的高光谱解混

陈善学, 储成泉

重庆邮电大学通信与信息工程学院, 重庆 400065

收稿日期:2019-01-15 修回日期:2019-03-15 发布日期:2019-08-14 出版日期:2019-08-10
通讯作者: 储成泉
作者简介:陈善学(1966-),男,安徽合肥人,教授,博士,主要研究方向:图像处理、数据压缩;储成泉(1993-),男,江苏南通人,硕士研究生,主要研究方向:高光谱图像解混。

Hyperspectral unmixing based on sparse and orthogonal constrained non-negative matrix factorization

CHEN Shanxue, CHU Chengquan

School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China

Received:2019-01-15 Revised:2019-03-15 Online:2019-08-14 Published:2019-08-10

摘要/Abstract

摘要： 针对基于非负矩阵分解（NMF）的高光谱解混存在的容易陷入局部极小值和受初始值影响较大的问题，提出一种稀疏和正交约束相结合的NMF的线性解混算法SONMF。首先，从传统的基于NMF的高光谱线性解混方法出发，分析高光谱数据本身的理化特性；然后，结合丰度的稀疏性和端元的独立性两个方面，将稀疏非负矩阵分解（SNMF）和正交非负矩阵分解（ONMF）两种方法结合应用到高光谱解混当中。模拟数据和真实数据实验表明，相比顶点成分分析法（VCA）、SNMF和ONMF这三种参考解混算法，所提算法提高了线性解混的性能；其中，评价指标光谱角距离（SAD）降低了0.012~0.145。SONMF能够结合两种约束条件的优势，弥补传统基于NMF线性解混方法对高光谱数据表达的不足，取得较好的效果。

关键词: 非负矩阵分解, 高光谱解混, 稀疏, 正交, 独立性

Abstract: Aiming at the problem that hyperspectral unmixing based on Non-negative Matrix Factorization (NMF) is easy to fall into local minimum and greatly affected by initial value, a linear unmixing algorithm based on Sparse and Orthogonal constrained Non-negative Matrix Factorization (SONMF) was proposed. Firstly, based on the traditional NMF hyperspectral linear unmixing method, the physical and chemical properties of the hyperspectral data was analyzed. Then the sparsity of the abundance and the independence of the endmember were combined together, two methods of Sparse Non-negative Matrix Factorization (SNMF) and Orthogonal Non-negative Matrix Factorization (ONMF) were combined and applied into hyperspectral unmixing. The experiments on simulation data and real data show that, compared with the three reference unmixing algorithms of Vertex Component Analysis (VCA), SNMF and ONMF, the proposed algorithm has improved the performance of linear unmixing, in which the Spectral Angle Distance (SAD) is reduced by 0.012 to 0.145. SONMF can combine the advantages of the two constraints to make up for the lack the expression of hyperspectral data by traditional NMF based linear unmixing methods, and achieve good results.

Key words: Non-negative Matrix Factorization (NMF), hyperspectral unmixing, sparsity, orthogonality, independence

中图分类号:

TN79

陈善学, 储成泉. 基于稀疏和正交约束非负矩阵分解的高光谱解混[J]. 计算机应用, 2019, 39(8): 2276-2280.

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.

参考文献

[1] 童庆禧,张兵,郑兰芬.高光谱遥感——原理、技术与应用[M].北京:高等教育出版社, 2006:15-30. (TONG Q X, ZAHNG B, ZHENG L F. Hyperspectral Remote Sensing——Principle, Technology and Application[M]. Beijing:Higher Education Press, 2006:15-30.)
[2] BIOUCAS-DIAS J M, PLAZA A, DOBIGEON N, et al. Hyperspectral unmixing overview:geometrical, statistical, and sparse regression-based approaches[C]//Proceedings of the 2011 IEEE International Geoscience and Remote Sensing Symposium. Piscataway, NJ:IEEE, 2011:1135-1138.
[3] 张兵,孙旭.高光谱图像混合像元分解[M].北京:科学出版社, 2015:6-22. (ZAHNG B, SUN X. Hyperspectral Image Unmixing[M]. Beijing:Science Press, 2015:6-22.)
[4] LEE D D, SEUNG S S. Learning the parts of objects by non-negative matrix factorization[J]. Nature, 1999, 401:788-791.
[5] 袁博.马尔可夫随机场的空间相关模型在非负矩阵分解线性解混中的应用[J].计算机应用,2017,37(12):3563-3568. (YUAN B. Application of MRF's spatial correlation model in NMF-based linear unmixing[J]. Journal of Computer Applications, 2017, 37(12):3563-3568.)
[6] WANG X, ZHONG Y, ZHANG L, et al. Spatial group sparsity regularized nonnegative matrix factorization for hyperspectral unmixing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2017, 55(11):6287-6304.
[7] RAJABI R, GHASSEMIAN H. Spectral unmixing of hyperspectral imagery using multilayer NMF[J]. IEEE Geoscience and Remote Sensing Letters, 2014, 12(1):38-42.
[8] ZHAO Y, ZHOU Z, WANG D, et al. Hyperspectral image unmixing algorithm based on endmember-constrained nonnegative matrix factorization[J]. Frontiers of Optoelectronics, 2016, 9(4):627-632.
[9] QIAN Y, JIA S, ZHOU J, et al. Hyperspectral unmixing via L_1/2 sparsity-constrained nonnegative matrix factorization[J]. IEEE Transactions on Geoscience & Remote Sensing, 2011, 49(11):4282-4297.
[10] 余肖玲.非负矩阵分解理论及其在高光谱解混中的应用[D]. 成都:成都理工大学, 2015:45-60. (YU X L. The theory of Nonnegative matrix factorization and its application in the hyperspectral unmixing[D]. Chengdu:Chengdu University of Technology, 2015:45-60.)
[11] NASCIMENTO J M P,BIOOUCASDIAS 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.
[12] 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, 2002, 39(3):529-545.
[13] POMPILI F, GILLIS N, ABSIL P, et al. Two algorithms for orthogonal nonnegative matrix factorization with application to clustering[J]. Neurocomputing, 2014, 141:15-25.
[14] 李孟杰,谢强,丁秋林.基于正交非负矩阵分解的K-means聚类算法研究[J].计算机科学,2016,43(5):204-208. (LI M J, XIE Q, DING Q L. Orthogonal Non-negative Matrix Factorization for K-means Clustering[J]. Computer Science, 2016, 43(5):204-208.)
[15] BIOUCASDIAS J M, NASCIMENTO J M P. Hyperspectral subspace identification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2008, 46(8):2435-2445. CHEN Shanxue, born in 1966, Ph.D., professor. His research interests including image processing, data compression.CHU Chengquan, born in 1993, M.S. candidate. His research interests including hyperspectral image unmixing.

基于稀疏和正交约束非负矩阵分解的高光谱解混

Hyperspectral unmixing based on sparse and orthogonal constrained non-negative matrix factorization

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	周烁, 仇润鹤, 唐旻俊. 基于禁忌搜索和Q-learning的CR-NOMA系统的功率分配算法[J]. 计算机应用, 2021, 41(7): 2026-2032.
[2]	裴春阳, 樊宽刚, 马政. 基于边缘保留分解和改进稀疏表示的医学图像融合[J]. 计算机应用, 2021, 41(7): 2092-2099.
[3]	杨蒙蒙, 张爱华. 基于灰度共生矩阵和同步正交匹配追踪的分形图像压缩[J]. 计算机应用, 2021, 41(5): 1445-1449.
[4]	夏玉杰, 时永鹏, 高雅, 孙鹏. 降低滤波器组多载波信号峰均比的边信息嵌入选择性映射方法[J]. 计算机应用, 2021, 41(5): 1425-1431.
[5]	刘紫燕, 马珊珊, 白鹤. 基于改进智能水滴的正交匹配追踪混合预编码算法[J]. 计算机应用, 2021, 41(5): 1419-1424.
[6]	林筠超, 万源. 基于图结构优化的自适应多度量非监督特征选择方法[J]. 计算机应用, 2021, 41(5): 1282-1289.
[7]	任晓奎, 刘鹏飞, 陶志勇, 刘影, 白立春. 基于单快拍信号到达角估计算法的室内入侵检测[J]. 计算机应用, 2021, 41(4): 1153-1159.
[8]	王锦凯, 贾旭. 基于迁移孪生非负矩阵分解的静脉识别算法[J]. 计算机应用, 2021, 41(3): 898-903.
[9]	时永鹏, 张俊杰, 夏玉杰, 高雅, 张尚伟. 基于NOMA的5G超密网计算迁移与资源分配策略[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3319-3324.
[10]	欧跃发, 杨鸣坤, 慕德俊, 柯捷, 马文涛. 基于广义最大Versoria准则的稀疏自适应滤波算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3325-3331.
[11]	葛晨宇, 董良, 许伊昆, 常毅, 张宏鸣. 基于总变分低秩组稀疏的全球雷达数据修复算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3353-3361.
[12]	裴仪瑶, 郭会明, 张丹普, 陈文博. 基于定位不确定性的鲁棒3D目标检测方法[J]. 计算机应用, 2021, 41(10): 2979-2984.
[13]	张艳娇, 乔志伟. 基于多残差UNet的CT图像高精度稀疏重建[J]. 计算机应用, 2021, 41(10): 2964-2969.
[14]	林腾涛, 查思明, 陈蕾, 龙显忠. 图趋势过滤诱导的噪声容错多标记学习模型[J]. 计算机应用, 2021, 41(1): 8-14.
[15]	王丽娟, 陈少敏, 尹明, 许跃颖, 郝志峰, 蔡瑞初, 温雯. 基于近邻图改进的块对角子空间聚类算法[J]. 计算机应用, 2021, 41(1): 36-42.