Convolution kernel

A convolution kernel is a small mathematical tool that can be used to find patterns in data. It looks like a square or it can be a circle and it can have any number of sections in it - usually it has a 3x3 or 5x5 grid. To use the kernel, you move it around an image, one section at a time, checking to see if the color or shapes of the section you are checking match what the kernel is looking for. If it finds something that looks like the pattern it's looking for, the kernel marks that spot in the image and then moves on to the next section. By doing this over and over, it can detect patterns and features and this helps a computer to recognize the shapes, edges, or textures that are in the image.