Abstract: Aiming at the problem that the fast tracking algorithm using spatio-temporal context (STC) will
inevitably lead to drift and even lose the target in long-term tracking, a new algorithm based on spatio-temporal
context that integrates long-term tracking with detecting is proposed in this paper. We track the target by the
fast tracking algorithm, and the cascaded search strategy is introduced to the detecting part to relocate the target
if the fast tracking fails. To a large extent, the proposed algorithm effectively improves the accuracy and stability
of long-term tracking. Extensive experimental results on benchmark datasets show that the proposed algorithm
can accurately track and relocate the target though the target is partially or completely occluded or reappears
after being out of the scene.
LU Jiahui (陆佳辉), CHEN Yimin* (陈一民), ZOU Yibo (邹一波), ZOU Guozhi (邹国志)
. Long-Term Tracking Based on Spatio-Temporal Context[J]. Journal of Shanghai Jiaotong University(Science), 2017
, 22(4)
: 504
-512
.
DOI: 10.1007/s12204-017-1863-z
[1] YILMAZ A, JAVAD O, SHAH M. Object tracking:A survey [J]. ACM Computing Surveys, 2006, 38(4):1-45.
[2] KIM D Y, JEON M. Spatio-temporal auxiliary particlefiltering with l1-norm based appearance model learningfor robust visual tracking [J]. IEEE Transactions onImage Processing, 2013, 22(2): 511-522.
[3] KALAL Z, MIKOLAJCZYK K, MATAS J. Trackinglearning-detection [J]. IEEE Transactions on PatternAnalysis & Machine Intelligence, 2011, 34(7): 1409-1422.
[4] HUA Y, ALAHARI K, SCHMID C. Occlusion andmotion reasoning for long-term tracking [C]//ECCV2014: European Conference on Computer Vision.Zurich, Switzerland: Springer, 2014: 172-187.
[5] PERNICI F. Facehugger: The ALIEN tracker appliedto faces [C]//ECCV 2012: European Conference onComputer Vision. Florence, Italy: Springer, 2012: 597-601.
[6] SUPANCIC J S, RAMANAN D. Self-paced learningfor long-term tracking [C]//CVPR 2013: ComputerVision and Pattern Recognition. Portland, Oregon,USA: IEEE, 2013: 2379-2386.
[7] HARE S, SAFFARI A, TORR P H. Struck: Structuredoutput tracking with kernels [J]. International Conferenceon Computer Vision, 2011, 23(5): 263-270.
[8] WU Y, LIM J, YANG M H. Online object tracking: Abenchmark [C]//CVPR 2013: Computer Vision andPattern Recognition. Portland, Oregon, USA: IEEE,2013: 2411-2418.
[9] ZHANG K H, ZHANG L, LIU Q S, et al. Fast visualtracking via dense spatio-temporal context learning[C]// ECCV 2014: European Conference on ComputerVision. Zurich, Switzerland: Springer, 2014: 127-141.
[10] LEWIS J P. Fast normalized cross-correlation [J]. Circuits,Systems and Signal Processing, 2009, 28(6):819-843.
[11] ROSS D A, LIM J, LIN R S, et al. Incremental learningfor robust visual tracking [J]. International Journal ofComputer Vision, 2008, 77: 125-141.
[12] GRABNER H, GRABNER M, BISCHOF H. Realtimetracking via on-line boosting [C]//Proceedings ofthe British Machine Vision Conference 2006. Edinburgh,UK: BMVA, 2006: 47-56.
[13] ADAM A, RIVLIIN E, SHIMSHONI I. Robustfragments-based tracking using the integral histogram[C]//CVPR 2006: Computer Vision and PatternRecognition. New York, USA: IEEE, 2006: 798-805.
[14] ORON S, BAR-HILLEL A, LEVI D, et al. Locally orderlesstracking [J]. International Journal of ComputerVision, 2015, 111(2): 213-228.
[15] ZHANG T Z, GHANEM B, LIU S, et al. Robust visualtracking via multi-task sparse learning [C]//CVPR2012: Computer Vision and Pattern Recognition.Providence, USA: IEEE, 2012: 2042-2049.
[16] BABENKO B, YANG M H, BELONGIE S. Robustobject tracking with online multiple instance learning[J]. IEEE Transactions on Pattern Analysis & MachineIntelligence, 2011, 33(8): 1619-1632.
[17] ZHANG K H, ZHANG L, YANG M H. Real-time compressivetracking [C]//ECCV 2012: European Conferenceon Computer Vision. Florence, Italy: Springer,2012: 864-877.
[18] COLLINS R T. Mean-shift blob tracking through scalespace [C]//CVPR 2003: Computer Vision and PatternRecognition. Madison, USA: IEEE, 2003: 234-240.
