I am using the Sobel filter for an RGB image. I have found two different ways to do that and the results look a little different. What is the difference between these two methods?
Is there an advantage over one of the two methods? I am trying to find a structure of the image that is parallel to the x axis.