DocumentCode :
1249441
Title :
Gaussian Mixture Model on Tensor Field for Visual Tracking
Author :
Zhan, Xueliang ; Ma, Bo
Author_Institution :
Beijing Lab. of Intell. Inf. Technol., Beijing Inst. of Technol., Beijing, China
Volume :
19
Issue :
11
fYear :
2012
Firstpage :
733
Lastpage :
736
Abstract :
Visual tracking remains a challenging problem because of both intrinsic appearance variability of object and extrinsic disturbance. To deal with this problem, we present a novel approach for tracking based on the tensor features. We convert the image into tensor field to yield more discriminating features and encode the target appearance probabilistically with gaussian mixture model (GMM). The model parameters are obtained by a modified EM algorithm using all tensor samples extracted from the target area. An incremental learning procedure is employed to update the model parameters for adapting to the appearance changes over time. Experimental results compared with three state-of-the-art methods demonstrate the good performance of the proposed algorithm under challenging conditions.
Keywords :
Gaussian processes; computer vision; feature extraction; image coding; learning (artificial intelligence); object tracking; tensors; GMM; Gaussian mixture model; extrinsic disturbance; image conversion; incremental learning; intrinsic appearance object variability; model parameters; modified EM algorithm; target appearance encoding; tensor field; tensor samples extraction; visual tracking; Adaptation models; Covariance matrix; Robustness; Signal processing algorithms; Target tracking; Tensile stress; Visualization; GMM; incremental learning; tensor field; visual tracking;
fLanguage :
English
Journal_Title :
Signal Processing Letters, IEEE
Publisher :
ieee
ISSN :
1070-9908
Type :
jour
DOI :
10.1109/LSP.2012.2209874
Filename :
6247466
Link To Document :
بازگشت