Implement Canny Edge Detection from Scratch with Pytorch

The Canny filter is certainly the most known and used filter for edge detection. I will explain step by step the canny filter for contour detection . Step by step because the canny filter is a multi-stage filter . The Canny filter is rarely integrated in a Deep Learning model. So I will describe the different parts and at the same time implement it with Pytorch . It can be customized almost without limit, I have allowed myself some deviations.

Let me introduce what a convolution matrix (or kernel ) is. A convolution matrix describes a filter that we are going to pass over an input image. To make it simple, the kernel will move over the whole image, from left to right, from top to bottom by applying a convolution product . The output of this operation is called a filtered image .