Convolution van be set up to do all sorts of filters.
Consider the following 1x3 matrix: (-1 0 1). If you convolve this along a line (a video perhaps) you would have a basic edge detector, because solid blocks of intensity would cancel itself out, but when you hit a luma change, the output would have a magnitude greater or lower than 1 (giving you the gradient).
If you mean clock edge detection, then you're indeed correct. It has nothing to do with it. But when it comes to video edge detection, then it's a technique that is regularly used...
Just do a google search on Sobel and edge detection, you'll find some interesting stuff.
Convolution van be set up to do all sorts of filters.
Consider the following 1x3 matrix: (-1 0 1). If you convolve this along a line (a video perhaps) you would have a basic edge detector, because solid blocks of intensity would cancel itself out, but when you hit a luma change, the output would have a magnitude greater or lower than 1 (giving you the gradient).
Yeah thanks for the piece of info buddy. I thought quite a different when I heard the name edge detection (I thought purely in terms of signals only, not to an image\video frame)