作者: José M. Bioucas-Dias , José M. P. Nascimento
DOI: 10.1117/12.620061
关键词:
摘要: Dimensionality reduction plays a crucial role in many hyperspectral data processing and analysis algorithms. This paper proposes new mean squared error based approach to determine the signal subspace imagery. The method first estimates noise correlations matrices, then it selects subset of eigenvalues that best represents least square sense. effectiveness proposed is illustrated using simulated real images.