Robust Visual Tracking Based on Online Discrimination Dictionary Learning
Xue Mo-gen① Zhu Hong① Yuan Guang-lin②
①(Anhui Province Key Laboratory of Polarization Imaging Detection Technology, Army Officer Academy of PLA, Hefei 230031, China) ②(Eleventh Department, Army Officer Academy of PLA, Hefei 230031, China)
The existing subspace tracking methods have well solved appearance changes and occlusions. However, they are weakly robust to complex background. To deal with this problem, firstly, this paper proposes an online discrimination dictionary learning model based on the Fisher criterion. The online discrimination dictionary learning algorithm for template updating in visual tracking is designed by using the block coordinate descent and replacing operations. Secondly, the distance between the target candidate coding coefficient and the mean of target samples coding coefficients is defined as the coefficient error. The robust visual tracking is achieved by taking the combination of the reconstruction error and the coefficient error as observation likelihood in particle filter framework. The experimental results show that the proposed method has better robustness and accuracy than the state-of-the-art trackers.
Li Xi, Hu Wei-ming, Shen Chun-hua, et al.. A survey of appearance models in visual object tracking[J]. ACM Transactions on Intelligent Systems and Technology, 2013, 4(4): 1-48.
[2]
Ross D, Lim J, Lin R S, et al.. Incremental learning for robust visual tracking[J]. International Journal of Computer Vision, 2008, 77(1-3): 125-141.
[3]
Mei Xue and Ling Hai-bin. Robust visual tracking using L1 minimization[C]. IEEE International Conference on Computer Vision, Kyoto, 2009: 1436-1443.
[4]
Black M J and Jepson A D. Eigentracking: Robust matching and tracking of articulated objects using a view-based representation[C]. European Conference on Computer Vision, London, 1996: 329-342.
[5]
Bao Cheng-long, Wu Yi, Ling Hai-bin, et al.. Real time robust L1 tracker using accelerated proximal gradient approach[C]. IEEE Conference on Computer Vision and Pattern Recognition, Providence, 2012: 1830-1837.
[6]
Xing Jun-liang, Gao Jin, Li Bing, et al.. Robust object tracking with online multi-lifespan dictionary learning[C]. IEEE International Conference on Computer Vision, Sydney, 2013: 665-672.
[7]
Mairal J, Bach F, Ponce J, et al.. Online dictionary learning for sparse coding[C]. The 26th International Conference on Machine Learning, Montreal, 2009: 539-547.
[8]
Wang Nai-yan, Wang Jing-dong, and Yeung D. Online robust non-negative dictionary learning for visual tracking[C]. IEEE International Conference on Computer Vision. Sydney, 2013: 657-664.
[9]
Yang Meng, Zhang Lei, Feng Xiang-chu, et al.. Sparse representation based Fisher discrimination dictionary learning for image classification[C]. IEEE International Conference on Computer Vision, Barcelina, 2011: 543-550.
[10]
Richtarik P and Takac M. Iteration complexity of randomized block-coordinate decent methods for minimizing a composite function[J]. Mathematical Programming, 2014, 144(1): 1-38.
[11]
Rosasco L, Verri A, Santoro M, et al.. Iterative projection methods for structured sparsity regularization[R]. MIT, 2009.
[12]
Wu Yi, Lim J, and Yang M. Online object tracking: a benchmark[C]. IEEE Conference on Computer Vision and Pattern Recognition, Portland, 2013: 2411-2418.
[13]
Zhang Kai-hua, Zhang Lei, and Yang M H. Fast compressive tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(10): 2002-2015.
[14]
Wang Dong, Lu Hu-chuan, and Yang M H. Online object tracking with sparse prototypes[J]. IEEE Transactions on Image Processing, 2013, 22(1): 314-325.
[15]
Zhang Tian-zhu, Ghanem B, Liu Si, et al.. Robust visual tracking via structured multi-task sparse learning[J]. International Journal of Computer Vision, 2013, 101(2): 367-383.
