Deformable Object Tracking Based on Lie Algebra
-
摘要: 动态几何变形是图像跟踪技术面临的突出难题之一. 本文提出基于李代数的变形目标跟踪方法, 用Gabor特征表征目标, 以仿射李群建立目标几何变形, 利用李代数和李群之间的指数映射将参数的最优化求解从欧氏空间转至光滑流形, 实现了对变形目标的稳定跟踪.从物理层面分析了目标跟踪过程中的参数几何变换的实质, 从理论上对在光滑流形上进行迭代求解的优点进行了详细分析, 并对其收敛性做出了证明.图像序列跟踪测试的对比实验表明, 本文方法较现有基于欧氏空间的算法在收敛速度、跟踪稳定性和精确性方面有显著提高.Abstract: Dynamic deformation of object is a distinct problem in image-based tracking. A novel deformable object tracking method based on Lie algebra is presented. This method uses Gabor feature as target token, models deformation using affine Lie group, and optimizes parameters directly on manifold, which can be solved by exponential mapping between Lie group and its Lie algebra. We physically illustrate the essential of the geometric transformation in the tracking, then analyze the advantage of our method and give a direct proof of local quadratic convergence of the algorithm. The experimental results demonstrate that the Lie algebra based method makes significant improvements in convergence speed, tracking stability and accuracy in comparison to Euclidean based algorithms.
-
Key words:
- Lie group /
- Lie algebra /
- object tracking /
- geometric deformation /
- exponential mapping
-
[1] Montera D A, Rogers S K, Ruck D W, Oxley M E. Object tracking through adaptive correlation. Optical Engineering, 1994, 33(1): 294-302[2] Harry H S, Marshall A D, Markham K C. Tracking targets in FLIR images by region template correlation. In: Proceedings of the Acquisition, Tracking, and Pointing. Orlando, USA: SPIE, 1997. 221-232[3] Shi J, Tomasi C. Good features to track. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Seattle, USA: IEEE, 1994. 593-600[4] Chien S, Sun S. Adaptive window method with sizing vectors for reliable correlation-based target tracking. Pattern Recognition, 2000, 33(2): 237-249[5] Hager G D, Belhumeur P N. Efficient region tracking with parametric models of geometry and illumination. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(10): 1025-1039[6] Jurie F, Dhome M. Hyperplane approximation for template matching. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(7): 996-1000[7] Baker S, Matthews I. Lucas-Kanade 20 years on: a unifying framework. International Journal on Computer Vision, 2004, 56(3): 221-255[8] Buenaposada J M, Baumela L. Real-time tracking and estimation of plane pose. In: Proceedings of the 16th International Conference on Pattern Recognition. Quebec, Canada: IEEE, 2002. 697-700[9] Zhu G P, Zeng Q S, Wang C H. Efficient edge-based object tracking. Pattern Recognition, 2006, 39(11): 2223-2226[10] Schreiber D. Robust template tracking with drift correction. Pattern Recognition Letters, 2007, 28(12): 1483-1491[11] Li Hong-You, Wang Tong-Qing, Ye Jun-Yong. Tracking moving target using drift correction algorithm. Acta Automatica Sinica, 2009, 35(3): 310-314(李宏友, 汪同庆, 叶俊勇. 基于主动漂移矫正的运动目标跟踪算法. 自动化学报, 2009, 35(3): 310-314)[12] Tsao T, Wen Z Q. Image-based target tracking through rapid sensor orientation change. Optical Engineering, 2002, 41(3): 697-703[13] Tsao T R, Wen J Z. Rapid target tracking algorithm using Gabor representation of target signature and Lie derivatives. In: Proceedings of the Acquisition, Tracking, and Pointing. Bellingham, USA: SPIE, 2001. 93-101[14] Mann S, Picard R W. Video orbits of the projective group: a simple approach to featureless estimation of parameters. IEEE Transactions on Image Processing, 1997, 6(9): 1281-1295[15] Drummond T, Cipolla R. Application of Lie algebras to visual servoing. International Journal on Computer Vision, 2000, 37(1): 21-41[16] Benhimane S, Malis E. Homography-based 2D visual tracking and servoing. International Journal of Robotics Research, 2007, 26(7): 661-676[17] Taylor C J, Kriegman D J. Minimization on the Lie Group SO(3) and Related Manifolds, Technical Report No.9405, Yale University,USA, 1994[18] Bayro-Corrochano E, Ortegon-Aguilar J. Lie algebra template tracking. In: Proceedings of the 17th International Conference on Pattern Recognition. Cambridge, UK: IEEE, 2004. 56-59[19] Bayro-Corrochano E, Ortegon-Aguilar J. Lie algebra approach for tracking and 3D motion estimation using monocular vision. Image and Vision Computing, 2007, 25(6): 907-921[20] Tuzel O, Porikli F, Meer P. Learning on Lie groups for invariant detection and tracking. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA: IEEE, 2008. 1-8[21] Mégret R, Authesserre J, Berthoumieu Y. Bidirectional composition on Lie groups for gradient-based image alignment. IEEE Transactions on Image Processing, 2010, 19(9): 2369-2381[22] Ying Shi-Hui, Peng Ji-Gen, Zheng Kai-Jie, Qiao-Hong. Lie group method for data set registration problem with anisotropic scale deformation. Acta Automatica Sinica, 2009, 35(7): 867-874(应时辉, 彭济根, 郑开杰, 乔红. 含各向异性尺度形变数据集匹配问题的Lie群方法. 自动化学报, 2009, 35(7): 867-874)[23] Liu Yun-Peng, Li Guang-Wei, Shi Ze-Lin. Projective registration algorithm based on Riemannian manifold. Acta Automatica Sinica, 2009, 35(11): 1378-1386(刘云鹏, 李广伟, 史泽林. 基于黎曼流形的图像投影配准算法. 自动化学报, 2009, 35(11): 1378-1386)[24] Hall B C. Lie Groups, Lie Algebras, and Representations: an Elementary Introduction. New York: Springer, 2003[25] Porat M, Zeevi Y Y. The generalized Gabor scheme of image representation in biological and machine vision. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1988, 10(4): 452-468[26] Wu X, Bhanu B. Gabor wavelet representation for 3-D objects recognition. IEEE Transactions on Image Processing, 1997, 6(1): 47-64[27] Li Qing-Yang, Wang Neng-Chao, Yi Da-Yi. Numerical Analysis. Beijing: Tsinghua University Press, 2001. 64-81(李庆扬, 王能超, 易大义. 数值分析. 北京: 清华大学出版社, 2001. 64-81)
点击查看大图
计量
- 文章访问数: 2250
- HTML全文浏览量: 80
- PDF下载量: 1421
- 被引次数: 0