基于线性光谱混合模型的光谱解混改进模型

An improved spectral unmixing modeling based on linear spectral mixing modeling

传统的基于线性光谱混合模型(LSMM)的解混方法采用迭代求解方式,复杂度较高,为此提出一种基于几何方式的模型求解方法。另一方面,LSMM采用固定谱形固定数量的光谱端元进行解混,影响了光谱解混精度,为此提出端元谱形的区域修正方法和端元子集的局域确定方法,从而建立基于柔性端元的新解混方式。实验表明了所提出的几何求解方法及柔性光谱端元方式的有效性。

Spectral unmixing is one of the important techniques of hyperspectral imagery processing.Traditional spectral unmixing method based on linear spectral mixing modeling（LSMM） is solved in terms of iteration manner,which suffers a heavy computational burden.In this case,a geometric solving method is proposed for LSMM instead of the iteration manner.Additionally,fixed number and fixed shape of endmembers are used in LSMM,leading to a discount unmixing accuracy.Aiming at this problem and constructing a new spectral unmixing way based on flexibly determined endmembers,the method of spectral shape modifying regionally and the method of endmember subset determining locally are proposed.Experiments show the validation of the geometric solving method of LSMM and the determination method of flexible endmembers.