Hello everyone,
I’ll start by stating that the pattern I want to recognize is very close from a Voronoi grid, so let’s take this as reference.
I have a bunch of differents Voronoi patterns like so:
and I need my program to be able to recognize which pattern is present in the image I’m testing. The pattern could be 3D oriented in the image I’ll try to compare, so it has to be able to handle perspective deformation, etc. like so:
I’m not an advanced programmer but I have some good notions, more specially in c#.
I’ve now been googling stuff for few days to learn as much as possible about recognitions algorythms.
I’m here to ask you if you know any better methods that the one I’m gonna list, or if you can tell me which one may produce a better result before I start. I’m going to list them from the most efficient to the less efficient (from what I think I understood so far)
Option 1) Using SimpleBlobDetector to collects datas from my original patterns, and making a comparison with the image I’ll try to compare.
Option 2) Using SURF algorythm to do the same (Even tho I don’t know how to extract datas from it at the moment)
Option 3) Proceed a HAAR training for every patterns in different angles, and compare it afterward.
Let me know if you have any idea or good link to share, any help is apreciated
Thank you for reading