Research Article
Part-Based Visual Tracking via Online Weighted P-N Learning
Algorithm 1
Tracking based on proposed method.
Initialization: | (1) Segment object into blocks and obtain ; | (2) Select the set of LFBs ; | (3) Compute the set of offset in first frame ; | (4) Generate positive samples and negative | samples for the LFB; | (5) Train classifier for the LFB with data via | weighted P-N learning, where , | and ; | Object Tracking: | (6) for to the end of the sequence do | (7) for to do | (8) Estimate via detecting LFB with classifier | in LK framework; | (9) Compute via (15); | (10) Retrain classifier ; | (11) end for | (12) Estimate via (16); | (13) Adjust the set of offset via (14); | (14) for to do | (15) Check each LFB with corresponding classifier | (16) if do | (17) Update based on Section 4.4; | (18) Generate positive samples and | negative samples for the LFB; | (19) Train new classifier for the LFB | (20) break; | (21) end if | (22) end for | End |
|