| 摘要 |
A method for hyperspectral image dimension reduction (300) including receiving (302) a hyperspectral image (200) having a plurality of pixels (202). A set of basis vectors (208) is established (304) at least in part with respect to the spectral vectors (204) of the initial hyperspectral image (200). For each pixel (202) of the hyperspectral image (200), the spectral vector (204) is read and decomposed, i.e. unmixed (532), with the basis vector set (208) to provide at least a reduced dimension vector for each pixel (202). Collectively the reduced dimension vectors for each pixel represent the dimensionally reduced image (306). A system (100) operable to perform the method (300) is also provided. |
