How to implement a pre built template for use with TLD?
I have successfully implemented TLD in openCV. Is there anyway I can train the detector once, save what it learned on to a file and then use the file to initialise the tracker at some other point of time for further tracking?