Wavelength matters

Let’s review: We’ve established that it’s preferable to use monochromatic rather than white light, and that you may want to match the wavelength of your lights to the sensitivity of your camera. Now let’s talk about how wavelength affects resolution.

Yes, the wavelength of the light you use affects the resolution of your vision system. It’s all to do with the ‘Airy disk diameter’ and the ‘Rayleigh resolution limit,’ neither of which I will explain because I’m an engineer and not a physicist. (Try Wikipedia, or Oldham Optical.)

But for now, understand that the straight rays of light we draw on lens diagrams should actually be thick lines with fuzzy edges because light passing through an aperture is always subject to diffraction. In cross-section, the ray would look like a disk rather than a point, hence the term “Airy disk,” (OK, so I don’t know who Airy was – does it matter?)

The diameter of the disk, D, is given by the formula:

D = 2.44 x λ x F

where λ is the wavelength of the light, and F the ‘f number’ of the lens. So if λ gets smaller the disk diameter will also reduce, and the spot of light will be less fuzzy.

And just what does this mean in the real machine vision world?

Simple: it means that shorter wavelength light will yield sharper images. As will a lens with a smaller f number. So you might like to consider using blue lights rather than red, even though the response of a CCD to blue is rather weak. As always, it’s a trade-off.