|
|
尺度自适应与抗遮挡的KCF方法研究
|
Abstract:
随着当前社会智能化的快速发展以及计算机视觉技术的兴起，目标跟踪技术在日常生活的诸多领域都得到了广泛应用。自相关滤波思想于2010年被引入目标跟踪领域后，基于此思想的跟踪器在性能方面的优良表现引起了国内外研究学者的关注。经典的核相关滤波目标跟踪算法(KCF)在较简单的场景下表现了较好的跟踪效果，但面对跟踪目标尺度变化及目标被遮挡的问题时，KCF算法无法进行稳定跟踪。本文提出了一种基于KCF的尺度自适应及抗遮挡的跟踪方法，利用尺度系数组对跟踪窗口进行尺度缩放，通过计算响应值确定跟踪窗口大小。同时利用图像响应值对目标进行遮挡判断，若目标未遮挡，则按原KCF算法执行；若目标被遮挡，则对图像提取ORB特征，在后续帧中重新确定目标位置，完成后续跟踪。OTB100数据集的实验表明本文方法对目标跟踪的可行性和有效性。
With the rapid development of social intelligence and the rise of computer vision technology, target tracking technology has been widely used in many areas of daily life. After the idea of autocorrelation filtering was introduced into the field of target tracking in 2010, the excellent performance of the tracker based on this idea has attracted the attention of researchers at home and abroad. The classic kernel correlation filtering target tracking algorithm (KCF) performs better tracking effect in relatively simple scenes, but the KCF algorithm cannot track stably in the face of the scale change of the tracking target and the occlusion of the target. In this paper, a KCF-based scale adaptive and anti-occlusion tracking method is proposed. The tracking window is scaled using the scale system array, and the response value of the specified region and template is calculated, and then the tracking window scale is determined. At the same time, the image response value is used to judge the occlusion of the target. If the target is not occluded, the original KCF algorithm is used; if the target is occluded, the ORB feature is extracted from the image, and the target position is redefined in the subsequent frame to complete the follow-up tracking. Experiments on OTB100 data set show that the proposed method is feasible and effective for target tracking.
| [1] | Henriques, J.F., Caseiro, R., Martins, P., et al. (2014) High-Speed Tracking with Kernelized Correlation Filters. IEEE Transactions on Pattern Analysis and Machine Intelligence, 37, 583-596.
https://doi.org/10.1109/TPAMI.2014.2345390 |
| [2] | Bolme, D.S., Beveridge, J.R., Draper, B.A. and Lui, Y.M. (2010) Visual Object Tracking Using Adaptive Correlation Filters. 2010 IEEE Computer Society Conference on Com-puter Vision and Pattern Recognition, San Francisco, 13-18 June 2010, 2544-2550. https://doi.org/10.1109/CVPR.2010.5539960 |
| [3] | Henriques, J.F., Caseiro, R., Martins, P., and Batista, J. (2012) Exploiting the Circulant Structure of Tracking-by- Detection with Kernels. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y. and Schmid, C., Eds., Computer Vision—ECCV 2012. Lecture Notes in Computer Science (Vol. 7575), Spring-er, Berlin, Heidelberg.
https://doi.org/10.1007/978-3-642-33765-9_50 |
| [4] | Liu, T., Wang, G. and Yang, Q.X. (2015) Real-Time Part-Based Visual Tracking via Adaptive Correlation Filters. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, 7-12 June 2015, 4902-4912.
https://doi.org/10.1109/CVPR.2015.7299124 |
| [5] | Akin, O., Erdem, E., Erdem, A. and Mikolajczyk, K. (2016) Deformable Part-Based Tracking by Coupled Global and Local Correlation Filters. Journal of Visual Communication and Image Representation, 38, 763-774.
https://doi.org/10.1016/j.jvcir.2016.04.018 |
| [6] | Ma, C., Yang, X.K., et al. (2015) Long-Term Correlation Tracking. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, 7-12 June 2015, 5388-5396. https://doi.org/10.1109/CVPR.2015.7299177 |
| [7] | Bibi, A. and Ghanem, B. (2015) Multi-Template Scale-Adaptive Kernelized Correlation Filters. 2015 IEEE International Conference on Computer Vision Workshop (ICCVW), Santiago, 7-13 December 2015, 613-620.
https://doi.org/10.1109/ICCVW.2015.83 |
| [8] | Danelljan, M., et al. (2015) Learning Spatially Regularized Correla-tion Filters for Visual Tracking. 2015 IEEE International Conference on Computer Vision (ICCV), Santiago, 7-13 De-cember 2015, 4310-4318.
https://doi.org/10.1109/ICCV.2015.490 |
| [9] | Danelljan, M., et al. (2016) Discriminative Scale Space Tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39, 1561-1575. https://doi.org/10.1109/TPAMI.2016.2609928 |
| [10] | Wang, M., Liu, Y. and Huang, Z. (2017) Large Margin Object Tracking with Circulant Feature Maps. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, 21-26 July 2017, 4800-4808.
https://doi.org/10.1109/CVPR.2017.510 |
| [11] | Wu, Y., Lim, J. and Yang, M.H. (2017) Online Object Tracking: A Benchmark. 2013 IEEE Conference on Computer Vision and Pattern Recognition, Portland, 23-28 June 2013, 2411-2418. https://doi.org/10.1109/CVPR.2013.312 |
| [12] | Wu, Y., Lim, J. and Yang, M.H. (2015) Object Tracking Benchmark. IEEE Transactions on Pattern Analysis & Machine Intelligence, 37, 1834-1848. https://doi.org/10.1109/TPAMI.2014.2388226 |
