![]() I see little reason for anyone to buy this from you. Unfortunately for you shape matching is so prevalent, simple, but also very limited in it's use. What are the actual advantages of your methods? Are they fast? Accurate? Do they match well under occlusions and deformations? Is it able to detect manufacturing defects? Some numbers would be appreciated to back any claims you have, especially if you are trying to sell. While this would be impressive for a beginner (nice work), the solutions themselves are not impressive as far as we can see here. A nice demo, a little confusing at first. From this I assume you're methods work on binary input pairs. I assume that's the shape matching score an green is a positive match? It also clearly written which method you are showcasing, and when the relative rotation of the shape is displayed. ![]() On it some numbers are display which are sometimes red, sometimes green. ![]() Top right is a binary template, so this is what we are looking for? Then top left is the segmented input image.
0 Comments
Leave a Reply. |