Open xinsuinizhuan opened 4 years ago
We actually wanted to evaluate this algorithm for our machine vision programs. Since it is not rotational and scale invariant, we have not pursued its development further. I have not looked into the performance of the algorithm. Kindly share performance figures that you are getting. I shall look into the speed up issue.
Do you have some rotational and scale invariant methods to template match?
Right now I do not have rotational and scale invariant methods to template match in OpenCV. However, other libraries like Halcon and LabView have robust implementation of said methods. I shall try to port those methods in OpenCV later on.
Do you have some methods of template match of the multi object、object have rotational and scale invariant?
It is so slowly, how to speed up it?