Why CCD sensors will become obsolete

It seems the massively-talented Andy Wilson and I are thinking along the same lines when it comes to the future of camera technology. While I was scribbling “Noise and CMOS sensors” and “Exploring the CCD vs. CMOS issue” he was hard at work on the extended-length “Machine Vision's Future Centers on CMOS and Consumer Advances” (Vision Systems Design, December 1^st, 2012.)

Andy’s article is interesting because he explores the drivers behind the growth of CMOS sensor technology, and because he talks about where it’s going. And where it’s going is described, to an extent anyway, in the complimentary VSD article, “Low-cost embedded devices boost end-user applications.” (Also Dec 1^st ’12.)

One big takeaway from these is that automotive applications are rapidly becoming the driver. Vehicle manufacturers want to put ever-more intelligence into our cars, (probably because we’ve demonstrated that we’re too dumb to handle them safely.) What’s more, they will offer huge volumes to those who can drive the cost down.

What does this mean for the developers and users of industrial machine vision? If want to take advantage of these economies of scale I suggest we look for ways to subvert blind spot warning, collision detection, and other such technologies, to our service. Otherwise we’ll be left playing with the scraps from the big boys table.