摘要
目标跟踪是计算机视觉中的关键问题,有着广泛的应用,由于现实场景中仍存在着众多问题,目标跟踪仍然具有巨大的研究价值。针对传统的核相关滤波器(KCF)算法在跟踪过程中存在的问题,文章提出一种结合位置相关滤波器和尺度相关滤波器的目标跟踪算法,同时依据跟踪效果,采取了自适应的模型更新策略。文章选取了公开视频序列进行测试,相较于KCF算法跟踪准确率提高了10%,尤其在光照变化、目标尺度变化等情况下有较强的适应性。
Target tracking is a key problem in computer vision and has a wide range of applications. Since there are still many problems in the real scene, target tracking still has great research value. Aiming at the problems of the traditional kernel correlation filter(KCF) algorithm in the tracking process, this paper proposes a tracking algorithm which combines position correlation filter and scale correlation filter, an adaptive model updating strategy is adopted according to the tracking effect. This paper selects open video sequences for testing, compared with KCF algorithm, the tracking accuracy is improved by 10%, especially in the case of illumination change and target scale change.
引文
[1]Naiyan Wang,Jianping Shi,Dit-Yan Yeung,and Jiaya Jia Understanding and diagnosing visual tracking systems[C]ICCV,2015:3101-3109.
[2]Bolme D S,Beveridge J R,Draper B A,et al.Visual objec tracking using adaptive correlation filters[C].CVPR,20102544-2550.
[3]Jo o F.Henriques,Caseiro R,Martins P,et al.Exploiting the Circulant Structure of Tracking-by-Detection with Kernels[M]//Computer Vision-ECCV 2012.Springer Berlin Heidelberg,2012.
[4]Henriques J F,Caseiro R,Martins P,et al.High-Speed Tracking with Kernelized Correlation Filters[J].IEEETransactions on Pattern Analysis and Machine Intelligence2015,37(3):583-596.
[5]Sch lkopf,Bernhard.Learning with kernels:support vector machines,regularization,optimization,and beyond[M]MIT Press,2003.
[6]Xiang-Rong Z,Chun Y,Li-Cheng J.Semi-Supervised SAR Target Recognition Based on Laplacian Regularized Least Squares Classification[J].Journal of Software,201021(8):586-596.
[7]Danelljan M,H ger,Gustav,Khan F S,et al.Discriminative Scale Space Tracking[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2016,39(8):1561-1575.
[8]Cehovin L,Leonardis A,Kristan M.Visual object tracking performance measures revisited[J].IEEE Transactions on Image Processing,2016,25(3):1-1.
[9]The Visual Object Tracking VOT2013 Challenge Results[C]//2013 IEEE International Conference on Computer Vision Workshops(ICCVW).IEEE Computer Society,2013.
[10]Kristan M,Leonardis A,Matas J,et al.The Visual Objec Tracking VOT2016 Challenge Results[C]//IEEE International Conference on Computer Vision Workshops.IEEEComputer Society,2016.
[11]Kristan M,Matas J,Leonardis A,et al.A Novel Performance Evaluation Methodology for Single-Target Trackers[J]IEEE Transactions on Pattern Analysis&Machine Intelligence,2016,38(11):2137-2155.
[12]Li Y,Zhu J.A Scale Adaptive Kernel Correlation Filter Tracker with Feature Integration[C]//European Conference on Computer Vision.Springer,Cham,2014.
[13]Vojir T,Noskova J,Matas J.Robust scale-adaptive meanshift for tracking.[J].Pattern Recognition Letters,2014,49(3):250-258.