I have downloaded videos and ground truth data from here and I am trying to use it to test out the results of my background subtraction algorithm. I can't for the life of me how they have labelled the ground truth data so which image lines up with which frame of the image/timestamp.
As for how to use I am unsure, but I was thinking of perhaps subtracting the frame from the image somehow. perhaps cv.subtract( in openCV. Then for true positives and negatives and false positives and negatives? Am I on the right track?
Aucun commentaire:
Enregistrer un commentaire