i’m using OpenCV for a school project using OpenCV for some object tracking. I know that there are multiple tracker in OpenCV, but can someone explain roughly how a tracker algorithm works in this case. And is it possible to code a primal version of that ? (not using blackbox like these trackers would be better for my project).

Thank you all (you can go a bit deeper in the explanation, and can add some mathematicals aspect)

I’m trying to code a simple object tracking in python but I want to do most of the job myself or at least comprehend what i’m using. In my code i’m using OpenCV but the tracker part make me a bit worried.