What is exciting is the growth of both the processing capacity and the sensor resolutions. Given that vision requires huge processing resources and the proliferation of quad / 8 core processors there is enough horsepower availible to time wise effectively process live images. Couple that with the advent of cheap, high resolution cameras you have a nexus of opportunity to provide "real time" live vision processing for the masses. Very exciting times indeed.
The application of vision is indeed becoming more prevalent in embedded systems. It is fascinating to see how vision is being applied to problems in the industrial and consumer spaces that have traditionally been handled by other, more cumbersome solutions. The replacement of gaming controllers with stereo vision based solution is one example among many. An example of one approach to adding custom embedded vision to a product is outlined in the following TI white paper: http://www.ti.com/lit/wp/spry232/spry232.pdf
Blog Doing Math in FPGAs Tom Burke 2 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...