协同运动状态估计的多目标跟踪算法

摘要/Abstract

摘要： 多目标跟踪在视频分析场景中有着广泛的应用,如人机交互、虚拟现实、自动驾驶、视频监控和机器人导航等。多目标跟踪问题可以表示为在已有的检测数据上进行目标轨迹关联,检测算法的准确性对跟踪性能起着关键性的作用。在基于检测的目标跟踪框架中,提出了一种协同运动状态估计的跟踪算法,该算法主要关注相邻帧之间的数据关联,从目标检测、目标运动状态估计和数据关联这3个方面来直接解决多目标跟踪面临的挑战。首先,对于目标检测,采用Multi Scale Convolutional Neural Network(MS-CNN)算法作为检测器,这是因为深度学习在检测的效益上优于传统的机器学习方法；其次,为了更好地预测目标的运动状态和处理目标间的遮挡,针对不同状态的目标采取不同的运动估计方法: 采用核相关滤波来评估处于跟踪状态的目标的运动状态,当目标处于遮挡状态时,采用卡尔曼滤波做运动估计；最后,采用Kuhn-Munkres算法对检测目标和跟踪轨迹做数据关联。通过大量的实验证实了算法的有效性,且实验结果表明算法的准确性很高。

关键词: 多目标跟踪,卡尔曼滤波,核相关滤波,数据关联,目标检测

Abstract: Multiple object tracking (MOT) is widely applied in video analysis scenarios,such as human interaction,virtual reality,autonomous driving,visual surveillance and robot navigation etc.MOT can be formulated as a sort of tracklets association in existing detection results,in which the accuracy of detection algorithm is entitled an essential role in tracking performance.We proposed a multiple object tracking algorithm via collaborative motion status estimation.The algorithm is based on the tracking-by-detection framework.The algorithm predominantly focuses on data association of adjacent video frames,tackling challenges of MOT from three aspects:object detection,object motion status estimation and data association.Firstly,as for object detection,multi scale convolutional neural network(MS-CNN) is adopted as the detector,since the advantage of deep learning in detection outweighs that of classical machine learning method.Se-condly,to better predict object motion status and handle occlusion among targets,different motion estimation methods are utilized according to different motion statuses.In tracking status,kernelized correlation filter is employed,while in occlusion status,the use of kalman filter is prioritized.Lastly,Kuhn-Munkres algorithm is adopted to work out data association between detections and tracklets.A substantial amount of experiments were carried out to estimate the efficiency.The results are quite positive,demonstrating high accuracy.

Key words: Multiple object tracking,Kalman filter,Kernelized correlation filter,Data association,Object detection

袁大龙,纪庆革. 协同运动状态估计的多目标跟踪算法[J]. 计算机科学, 2017, 44(Z11): 154-159. https://doi.org/10.11896/j.issn.1002-137X.2017.11A.032

YUAN Da-long and JI Qing-ge. Multiple Object Tracking Algorithm via Collaborative Motion Status Estimation[J]. Computer Science, 2017, 44(Z11): 154-159. https://doi.org/10.11896/j.issn.1002-137X.2017.11A.032

参考文献

[1] LUO W,XING J,ZHANG X,et al.Multiple Object Tracking:A Literature Review[J/OL].https://arXiv.org/abs/1409.7618.
[2] REID D.An Algorithm for Tracking Multiple Targets[J].IEEE Transactions on Automatic Control,1979,24(6):843-854.
[3] FORTMANN T,BAR-SHALOM Y,SCHEFFE M.Sonar Trackingof Multiple Targets Using Joint Probabilistic Data Association[J].IEEE Journal of Oceanic Engineering,1983,8(3):173-184.
[4] HAMID REZATOFIGHI S,MILAN A,ZHANG Z,et al.Joint Probabilistic Data Association Revisited[C]∥Proceedings of the IEEE International Conference on Computer Vision.2015:3047-3055.
[5] KIM C,LI F,CIPTADI A,et al.Multiple Hypothesis Tracking Revisited[C]∥In Proceedings of the IEEE International Confe-rence on Computer Vision.2015:4696-4704.
[6] BAE S H,YOON K J.Robust Online Multi-Object Trackingbased on Tracklet Confidence and Online Discriminative Ap-pearance Learning[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2014:1218-1225.
[7] YANG M,JIA Y.Temporal dynamic appearance modeling for online multi-person tracking[J].Computer Vision and Image Understanding,2016,153:16-28.
[8] XIANG Y,ALAHI A,Savarese S.Learning to Track:Online Multi-Object Tracking by Decision Making[C]∥Proceedings of the IEEE International Conference on Computer Vision.2015:4705-4713.
[9] CHOI W.Near-Online Multi-target Tracking with Aggregated Local Flow Descriptor[C]∥Proceedings of the IEEE International Conference on Computer Vision.2015:3029-3037.
[10] CAI Z,FAN Q,FERIS R S,et al.A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection[C]∥European Conference on Computer Vision.Springer InternationalPublishing,2016:354-370.
[11] HENRIQUES J F,CASEIRO R,MARTINS P,et al.High-Speed Tracking with Kernelized Correlation Filters[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(3):583-596.
[12] KALMAN R E.A New Approach to Linear Filtering and Prediction Problems[J].Journal of Basic Engineering,1960,82(1):35-45.
[13] REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards Real-Time Object Detection with Region Proposal Networks[C]∥Advances in Neural Information Processing Systems.2015:91-99.
[14] YANG F,CHOI W,LIN Y.Exploit All the Layers:Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:2129-2137.
[15] RIFKIN R,YEO G,POGGIO T.Regularized Least-SquaresClassification[J].Nato Science Series Sub Series III Computer and Systems Sciences,2003,190:131-154.
[16] YU F,LI W,LI Q,et al.POI:Multiple Object Tracking with High Performance Detection and Appearance Feature[C]∥European Conference on Computer Vision.Springer International Publishing,2016:36-42.
[17] LEAL-TAIX L,MILAN A,REID I,et al.MOTChallenge2015:Towards a Benchmark for Multi-Target Tracking[J/OL].https://arXiv.org/abs.1504.01942.
[18] GEIGER A,LAUER M,WOJEK C,et al.3D Traffic Scene Understanding from Movable Platforms[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(5):1012-1025.
[19] PIRSIAVASH H,RAMANAN D,FLOWKES C C.Globally-Optimal Greedy Algorithms for Tracking a Variable Number of Objects[C]∥2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).IEEE,2011:1201-1208.
[20] DCILE C,CAMPS O I,SZNAIER M.The Way They Move:Tracking Multiple Targets with Similar Appearance[C]∥Proceedings of the IEEE International Conference on Computer Vision.2013:2304-2311.
[21] MILAN A,ROTH S,SCHINDLER K.Continuous Energy Minimization for Multitarget Tracking[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(1):58-72.
[22] YOON J H,YANG M H,LIM J,et al.Bayesian Multi-objectTracking using Motion Context from Multiple Objects[C]∥2015 IEEE Winter Conference on Applications of Computer Vision (WACV).IEEE,2015:33-40.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed