A simple image convolution

Key Takeaways

what's going on is I've got this little 3x3 grid of values that's marching along our original image and if we zoom in each one of those values is 1 nth and what I'm doing at each iteration is multiplying each of those values by the corresponding pixel that it sits on top of and of course in computer science we think of colors as little vectors of three values representing the red green and blue components when I multiply all these little values by 1 nth and I add them together it gives us an average along each color Channel and the corresponding pixel for the image on the right is defined to be that sum the overall effect as we do this for every single Pixel on the image is that each one kind of bleeds into all of its neighbors which gives us a blurrier version than the original in the lingo we'd say that the image on the right is a convolution of our original image with a little grid of values or more technically maybe I should say that it's the convolution with a 180° rotated version of that little grid of values