The progression of the images getting more refined as we increase the operations done on them make sense to me, but how do we obtain the points in the initial "rough" picture for us to process later on?