Match template using pyramid method

edit

Hi all!

I'm pretty newbie working with OpenCV . I'm trying to implement an algorithm to improve the performance of template matching operation .

I saw this article (http://opencv-code.com/tutorials/fast...) a method whose steps seem to make sense and I'm trying to take it to python . The problem is that I do not quite understand the procedure. I always get in the first level (smallest image) a black image. I never obtain any white point. Also I do not understand why this algorithm uses 2 masks as I don't understand the procedure.

I wonder if someone can guide me in the procedure to be performed for template matching in partial regions of the image . Or any other method to improve performance.

Let me know if any thing is not clear !

Thanks