I was having a hard time visualizing the effects of this filter and found that this site helped: https://setosa.io/ev/image-kernels/.

If this convolution were applied to an image with uniform color, like mentioned in lecture, this would produce the same color as 5x(color) - 4x(color) = 1x(color). However, on edges where there is a sharp change in color, either the top and bottom or right and left sets of -1x will be skewed and make the dark colors darker and the light colors lighter.

Thanks for sharing! It's very helpful. I was confused about how convolution works in terms of matrix multiplication. Now I understand that it is element by element multiplication.