An initialization method of non-negative matrix factorization for hyperspectral data unmixing based on spectral shape and information dissimilarity
YUAN Deyou1, YUAN Lin2
1. School of Mathematics dissimilarity and Statistics, Nanyang Institute of Technology, Nanyang 473004, China; 2. School of Economics and Management, Nanyang Institute of Technology, Nanyang 473004, China
Abstract：When blind signal separation technique is applied to unmixing hyperspectral data, a good initialization is vital for improving separating precision. Aimed at the hyperspectral data with relatively high spatial resolution and simple surface features, the authors put forward a reasonable hypothesis that the data contain pure pixel or approximate pure pixel corresponding to the each type of end-members, and proposed a new initialization method of non-negative matrix factorization(NMF), which has great potential in pixel unmixing. By calculating parameters to quantify the spectral shape and information difference among candidate pixels, this method extracts pure pixels from mixed pixels, recognizes the information dissimilarity among different types of pure pixels and choose the existing pixels that are most suitable for representing each type of end-members as NMF’s initial values. The experimental results show that the method proposed in this paper can improve NMF’s decomposition accuracy of hyperspectral data significantly, and its performance is better than that of other NMF initialization methods.
袁德有, 袁林. 基于谱线形状与信息量差异的高光谱解混NMF初始化方法[J]. 国土资源遥感, 2017, 29(4): 114-119.
YUAN Deyou, YUAN Lin. An initialization method of non-negative matrix factorization for hyperspectral data unmixing based on spectral shape and information dissimilarity. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(4): 114-119.
[1] Berry M W,Browne M,Langville A N,et al.Algorithms and applications for approximate nonnegative matrix factorization[J].Computational Statistics and Data Analysis,2007,52(1):155-173. [2] 李二森,张保明,杨 娜,等.非负矩阵分解在高光谱图像解混中的应用探讨[J].测绘通报,2011(3):7-10. Li E S,Zhang B M,Yang N,et al.Discussion of the NMF’s application for hyperspectral imagery unmixing[J].Bulletin of Surveying and Mapping,2011(3):7-10. [3] Boutsidis C,Gallopoulos E.SVD based initialization:A head start for nonnegative matrix factorization[J].Pattern Recognition,2008,41(4):1350-1362. [4] Alshabrawy O S,Ghoneim M E,Awad W A,et al.Underdetermined blind source separation based on fuzzy C-means and semi-nonnegative matrix factorization[C]//Proceedings of 2012 Federated Conference on Computer Science and Information Systems.Wroclaw,Poland:IEEE,2012:695-700. [5] 翟亚利,吴 翊.NMF初始化研究及其在文本分类中的应用[J].计算机工程,2008,34(16):191-193,197. Zhai Y L,Wu Y.Study of non-negative matrix factorization initialization and its application to text classification[J].Computer Engineering,2008,34(16):191-193,197. [6] Lee D D,Seung H S.Algorithms for non-negative matrix factorization[C]//Proceedings of the 13th International Conference on Neural Information Processing Systems.Denver,CO:ACM,2000:535-541. [7] 余先川,安卫杰,吕中华,等.一种基于光谱角和光谱距离自动加权融合分类方法[J].地质学刊,2012,36(1):33-36. Yu X C,An W J,Lyu Z H,et al.Automatic weighting fusion classification method based on spectral angle and spectral distance[J].Journal of Geology,2012,36(1):33-36. [8] Wang R J,Zhan Y J,Zhou H F.RETRACTED:A method of underdetermined blind source separation with an unknown number of sources[J].Engineering Applications of Artificial Intelligence,2011:556-562.doi:10.1016/j.engappai.2011.06.003. [9] 毋文峰,陈小虎,苏勋家,等.基于峭度的ICA特征提取和齿轮泵故障诊断[J].机械科学与技术,2011,30(9):1583-1587. Wu W F,Chen X H,Su X J,et al.ICA feature extraction and fault diagnosis based on Kurtosis for a gear pump[J].Mechanical Science and Technology for Aerospace Engineering,2011,30(9):1583-1587.
2017, Vol. 29 
