Image Kernel Convolutions (Filters/Masks) Visually Explained
In this video we cover image kernels, convolution matrices, or masks, that are uses for photo editing effects and feature detection in AI using a visual tool found at https://setosa.io/ev/image-kernels/
Kernel (Image Processing) Wikipedia https://en.wikipedia.org/wiki/Kernel_...)
How Blurs & Filters Work - Computerphile • How Blurs & Filters Work - Computerphile
Finding the Edges (Sobel Operator) - Computerphile • Finding the Edges (Sobel Operator) - Compu...
Sobel Operator Wikipedia https://en.wikipedia.org/wiki/Sobel_o...
Image Gradient Wikipedia https://en.wikipedia.org/wiki/Image_g...
Summary:
Iterate through every pixel
select values from making a grid of the same size of the image kernel center around the pixel of interest
multiply corresponding values with image kernel weight, add those values, the output value usually (0-255) or black to white in grayscale operations is now that value of the corresponding pixel (pixel of interest/center of grid)
for edges you can duplicate the image or ignore
http://minorityprogrammers.org/