The frequency of a beam of light doesn't change from the creation to annihilation of the photon. This is not true of the wavelength, though.
[itex]f \lambda = c/n[/itex], where n is the index of refraction of the medium. The local n near the retinas will always be the same, so frequency and wavelength can be related if that's known.
If you want to generalize a bit and say "what specifies an electromagnetic plane wave" the answer would be the frequency (and direction of propagation). The wavelength is then determined by the formula above. This is all my way of saying the frequency is more an intrinsic property of a wave, wavelength more extrinsic.
Side note: The charts you see giving the color of light based on wavelength are based on wavelength in a vacuum, with n = 1.
In electronics, filters work with frequency, not wavelength. I don't know a great deal about how CCDs work, but I'm guessing E = hf is an fairly important formula. One would typically design a receiver to be sensitive to a particular range of frequencies, because once the photon enters the detector its' wavelength is determined by the local n; it doesn't matter what it was in the ocean.
For optical cavities, one designs such that the optical path length is an integer multiple of the wavelength. Basically, the two are highly related, although not completely interchangeable. The most convenient choice varies with application.