For a few different images, Experiment with some low-level edgel detection operations, (e.g. thresholded gradient magnitude, thinned gradient magnitude, laplacian zero crossings, etc.) to get a feel for how they operate. Try combining with some smoothing operations to see what sort of a difference that makes. Turn in your original image and some of the more interesting processed versions, and a description of what you noticed about the various methods of labelling edgels.
Back to vision course main page