Share Email Print

Proceedings Paper

Online real AdaBoost with co-training for object tracking
Author(s): Lizuo Jin; Zhiguo Bian; Xiaobing Li; Hong Pan; Siyu Xia
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

One of the major challenges of object tracking is to tackle appearance variations, possibly caused by the change of object postures, size, and occlusions. In this paper an adaptive tracking system is presented, which integrates online semisupervised classification and particle filter efficiently. To identify object pixels from background accurately, classifiers are trained online using real Adaboost which performs much better than its discrete version. In the system, uncorrelated features, color and texture are adopt to train two classifiers separately; the classifiers fused by voting generate confidence score for each pixel measuring its belonging to object or background in candidate regions; accumulated scores in each region are feed to particle filter for estimating object states; pixels with high scores augment the training set mutually and further classifiers are updated by co-training. The system is applied to vehicle and pedestrian tracking in real world scenarios and the experimental results show its robustness to large appearance variations and severe occlusions.

Paper Details

Date Published: 30 October 2009
PDF: 8 pages
Proc. SPIE 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis, 749503 (30 October 2009); doi: 10.1117/12.833166
Show Author Affiliations
Lizuo Jin, Southeast Univ. (China)
Zhiguo Bian, Southeast Univ. (China)
Xiaobing Li, Southeast Univ. (China)
Hong Pan, Southeast Univ. (China)
Siyu Xia, Southeast Univ. (China)

Published in SPIE Proceedings Vol. 7495:
MIPPR 2009: Automatic Target Recognition and Image Analysis

© SPIE. Terms of Use
Back to Top