中文摘要：

像元分解是遥感图像信息挖掘的重要手段，非负矩阵分解模型应用于高光谱遥感图像混合像元分解时，分解的效果与算法所获局部最优解密切相关。本文将带正交性约束的非负矩阵分解用于光谱解混，保证了分解矩阵列向量的线性无关性，进而使分解所得端元光谱具有较大的独立性。通过试验分析，利用正交非负矩阵分解，实现了对1997年机载可见光及红外成像光谱仪(AVIRIS)高光谱图像的混合像元分解，结果表明，增加约束条件后的正交非负矩阵分解，能成功分离出6种端元光谱，解混出的端元光谱与参考光谱的光谱角距离更小，与真实地物的丰度谱图吻合度增强。

英文摘要：

Hyperspectral unmixing is a powerful tool for the remote sensing image mining.Nonnegative matrix factorization(NMF)has been adopted to deal with this issue,while the precision of unmixing is closely related with the local minimizer of NMF.Orthogonal NMF which imposes orthogonality constraints on the factor matrices can improve the clustering performance,since it ensures the independent of the endmember spactra.In the experimental test,ONMF was used to unmix the Urban scene which was captured by airborne visible/infrared imaging spectrometer(AVIRIS)in1997,numerical results showed that ONMF could extract the endmember signature and accurately estimate abundance maps.