Assuming they see greyscale*, I would just show them two objects with the same greyscale shade (i.e. which look identical to them), and explain that to me they look different - my eyes can see things their's can't. I'd explain that in low light my vision is like theirs, but my eyes have something that can see more when the light is brighter.
If they have advanced enough physics to know what wavelengths are, then I would just explain that I can perceive different wavelengths. If they have a sense of hearing and can hear differences in pitch, then I'd explain that brightness is analogous to loudness and color is analogous to pitch - I can see differences in wavelength of light in the same way they can hear differences in wavelength in sound.
If they required proof, I'd do something similar to Dubber Rucky's suggestion. I'd give them two piles of identically shaped pieces of paper which are indistinguishable in greyscale, but of two different colors. I'd tell them to test me; make a shape from the pieces of paper in one pile against a background of pieces of paper from the other pile (which they obviously would be unable to see after construction), and then show it to me and see if I could identify it. They could even have a control group of members of their own species, to prove that it really isn't greyscale-identifiable.
*All of this would work too if they saw pure black and white only and nothing in between