Question about detector

[groschakal]

Hi, I am a student on image processing.

I try to use your TLD version for a personal project and I have a question about it.

When I track an object on a video and this object are move out the scene and he comes back from other side, it is not detect. And it will be detect when he comes back at the last position where it was detect before move out.

I think the learning is good because I have a lot of positive patch of my object.

I supposed the detection of object work only on a patch near the last position where the object was detect, but I do not find where this is on your code.

I would like to continue to detect the object in all parts of the frame when the detection was lost and not just in the last position.

Can you help me ?

Thank you in advance.