Principal components analysis (PCA) of X-ray transmission tomography images
made at several different energies can produce images that allow individua
l materials to be distinguished more clearly than in single-energy images.
Computer simulations and tests on a five-band mineral data set demonstrate
the application of PCA.