One of the most common image filtering methods is performed using the act of convolution. More specifically, a convergence takes place between the original image and a table of factors called a mask or kernel. The act of convergence can visually reminiscent of array multiplication, but differs significantly. In general In this case, the paintings are not the same size, but the mask is small e.g. 3x3 while the image is much larger. In these cases the image is split in subdivisions of the same size as the mask and in each subdivision is executed the convergence.