Multispectral images will be acquired in 18 different, non-uniformly distributed wavelengths ranging from UV (405 nm) to short wave NIR (970 nm). Briefly, the procedure for image acquisition and segmentation will be described, the system will be subjected to a light set up procedure known as “autolight” and will be calibrated radiometrically and geometrically. The samples will be transferred to Petri dishes, will be placed inside the Ulbricht sphere and multispectral images of the samples’ surface will be acquired. In order for redundant information (related to the background, fat, and not to the sample) to be removed image segmentation will be performed using the VideometerLab system software (version 2.12.39). The contrast between the sample material and the other irrelevant objects is maximized, in order to enable a threshold operation. Then, canonical discriminant analysis is employed, resulting in the production of a segmented image.