In biological spectral microscopy, the standard in the field is to use normalized endmembers in linear unmixing algorithms. In most cases, the endmembers are normalized to the maximum value. I have not found a reference that explains why endmembers are normalized prior to linear unmixing. By normalizing, are we not removing information about relative intensity of objects?
I am asking my question here because I am curious if endmember normalization is routine in remote sensing as well.