Computer vision versus machine vision

After my little tirade about the growing reach of surveillance technology (“The ethics of computer vision”) I received an email asking if I could explain the difference between the two terms in the title of this posting. (Just goes to show you should never make assumptions about the background and experience of your audience.)

Wikipedia probably does this better than I can, but I’m going to use fewer words:

Computer vision is the science of analyzing and processing digital images. It has many possible applications, from supporting cellular biology research to surveillance, guidance of autonomous vehicles and industrial inspection.

Machine vision is the application of computer vision to industrial tasks. Inspection in all its various flavors is probably the main one, but process control and robot guidance figure in there too. This means machine vision is more engineering than science. Interestingly though, when talking about vision guided robotics we get into stereo vision and 3D reconstruction, which starts to move into what I consider the computer vision world.

I guess the boundaries are fuzzy but I think it’s fair to say that machine vision is a subset of computer vision. Hope that clears things up.