空间可靠性和相关滤波器联合学习的跟踪算法

doi:10.19665/j.issn1001-2400.2021.05.020

摘要/Abstract

摘要：

判别式相关滤波器采用循环移位产生负样本的方式不可避免带来了边界效应。基于背景感知的相关滤波跟踪算法试图利用裁剪矩阵获取更多真实的负样本,既有效缓解了边界效应的影响,又增强了对背景信息的学习。然而,裁剪矩阵的使用缺乏对空间不同位置可靠性的学习,可能会导致背景信息对滤波器的学习占据主导地位。为解决该问题,将空间可靠性的学习引入相关滤波算法中,通过交替方向法与滤波器进行联合迭代求解,加强了滤波器对空间可靠性区域的学习,增强了滤波器的对目标与背景的判别力。此外,为优化模型更新策略,提出了一种基于感知哈希算法的自适应模型更新方法,提升了滤波器学习的有效性。所提出的算法在标准视觉跟踪数据集上进行了全面评估,验证了该算法在性能上的有效性以及实时性。

Abstract:

The discriminant correlation filter (DCF) uses the cyclic shift to generate negative samples,which inevitably brings boundary effects.The background-aware correlation filter (BACF) attempts to use the clipping matrix to obtain more real negative samples.The method can not only effectively alleviate the influence of the boundary effect,but also enhance the learning of background information.However,the use of the clipping matrix lacks the learning of the spatial reliability of different positions,which may cause the background information to dominate the learning of the filter.In order to solve this problem,this paper introduces the learning of spatial reliability into the correlation filter.And the Alternate Direction Method is used to iteratively obtain the solution of spatial reliability and the filter.Our method can strengthen the learning of the spatial reliability region and enhance the filter's ability to discriminate targets and background.In addition,in order to optimize the model update strategy,an adaptive model update method based on the Perceptual Hash Algorithm is proposed,which improves the effectiveness of filter learning.The proposed algorithm has been comprehensively evaluated on standard visual tracking datasets.The results verify the effectiveness and real-time performance of the algorithm.

Key words: visual tracking, correlation filter, spatial reliability, joint learning, perceptual hashing algorithm, adaptive learning

中图分类号:

TN391.41

张飞,马时平,张立朝,何林远,仇祝令,韩永赛. 空间可靠性和相关滤波器联合学习的跟踪算法[J]. 西安电子科技大学学报, 2021, 48(5): 167-177.

ZHANG Fei,MA Shiping,ZHANG Lichao,HE Linyuan,QIU Zhuling,HAN Yongsai. Joint spatial reliability and correlation filter learning for visual tracking[J]. Journal of Xidian University, 2021, 48(5): 167-177.

图/表 8

图1

图2

表1

图3

图4

图5

图6

图7

参考文献 24

[1]	宋建锋, 苗启广, 申猛, 等. 多特征融合的相关滤波红外单目标跟踪算法[J]. 西安电子科技大学学报, 2019, 46(05):142-147.
	SONG Jianfeng, MIAO Qiguang, SHEN Meng, et al. Algorithm for Tracking an Infrared Single Target Based on Correlation Filtering with Multi-feature Fusion[J]. Journal of Xidian University, 2019, 46(05):142-147.
[2]	王欣远, 肖嵩, 李磊, 等. 融合ELM和相关滤波的鲁棒性目标跟踪算法[J]. 西安电子科技大学学报, 2019, 46(01):57-63.
	WANG Xinyuan, XIAO Song, LI Lei, et al. Robust Target Tracking Algorithm Combining ELM and Correlation Filtering[J]. Journal of Xidian University, 2019, 46(01):57-63.
[3]	HENRIQUES J F, CASEIRO R, MARTINS P, et al. High speed Tracking with Kernelized Correlation Flters[J]. Proceedings of the IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(3):583-596.
[4]	DANELLJAN M, HAGER G, KHAN F S, et al. Learning Spatially Regularized Correlation Flters for Visual Tracking[C]// Proceedings of the IEEE International Conference on Computer Vision.Piscataway:IEEE, 2015:4310-4318.
[5]	GALOOGAHI K H, FAGG A, LUCEY S. Learning Background-aware Correlation Flters for Visual Tracking[C]// Proceedings of the IEEE International Conference on Computer Vision.Piscataway:IEEE, 2017:1135-1143.
[6]	HUANG Z Y, FU C H, LI Y M, et al. Learning Aberrance Repressed Correlation Flters for Real-time Uav Tracking[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision.Piscataway:IEEE, 2019:2891-2900.
[7]	DAI K N, WANG D, LU H C, et al. Visual Tracking Via Adaptive Spatially-regularized Correlation Flters[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2019:4670-4679.
[8]	FU C H, HUANG Z Y, LI Y M, et al. Boundary Effect-aware Visual Tracking for Uav Withonline Enhanced Background Learning and Multi-frame Consensus Verifcation[C].// 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems.Piscataway:IEEE, 2019:4415-4422.
[9]	KOZAT S S, VENKATESAN R, MIHC K M. Robust Perceptual Image Hashing Via Matrix Invariants[C]// 2004 International Conference on Image Processing,Piscataway:IEEE. 2004:3443-3446.
[10]	WU Y, LIM J, YANG M H. Online object tracking:A benchmark[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2013:2411-2418.
[11]	WU Y, LIM J, YANG M H. Object Tracking Benchmark[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 37(9):1834-1848.
[12]	MUELLER M, SMITH N, GHANEM B. A Benchmark and Simulator for Uav Tracking[C]// European Conference on Computer Vision.Berlin:Springer, 2016:445-461.
[13]	BOYD S, PARIKH N, CHU E, et al. Distributed Optimization and Statistical Learning Via the Alternating Direction Method of Multipliers[J]. Foundations and Trends in Machine Learning, 2010, 3(1):1-122. doi: 10.1561/2200000016
[14]	SHERMAN J and MORRISON W J. Adjustment of an Inverse Matrix Corresponding to a Change in One Element of a Given Matrix[J]. The Annals of Mathematical Statistics, 1950, 21(1):124-127. doi: 10.1214/aoms/1177729893
[15]	DALAL N, TRIGGS B. Histograms of Oriented Gradients for Human Detection[C]// Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2015:886-893.
[16]	DANELLJAN M, KHAN F S, FELSBERG M, et al. Adaptive color attributes for real-time visual tracking[C]// 2014 IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2014:1090-1097.
[17]	LI F, TIAN C, ZUO W M, et al. Learning Spatial-temporal Regularized Correlation Flters for Visual Tracking[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2018:4904-4913.
[18]	DANELLJAN M, BHAT G, KHAN F S, et al. Eco:Effcient Convolution Operators for Tracking[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2017:6938-6946.
[19]	LUKEZIC A, VOJIR T, ZAJC L C, et al. Discriminative Correlation Flter with Channel and Spatial Reliability[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2017:6309-6318.
[20]	BERTINETTO L, VALMADRE J, GOLODETZ S, et al. Staple:Complementary Learners for Real-Time Tracking[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2016:1401-1409.
[21]	BERTINETTO L, VALMADRE J, HENRIQUES J F, et al. Fully-convolutional Siamese Networks for Object Tracking[C]// European Conference on Computer Vision.Berlin:Springer, 2016:850-865.
[22]	LI B, YAN J J, WU W, et al. High Performance Visual Tracking with Siamese Region Proposal Network[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2018:8971-8980.
[23]	LI Y, ZHU J K, HOI S C H, et al. Robust Estimation of Similarity Transformation for Visual Object Tracking[C]// Proceedings of the AAAI Conference on Artificial Intelligence,Hawaii:AAAI, 2019, 33(01):8666-8673.
[24]	ZHANG J, MA S, SCLAROFF S. MEEM:Robust Tracking Via Multiple Experts Using Entropy Minimization[C]// European Conference on Computer Vision.Berlin:Springer, 2014:188-203.

参数	取值
η	0.012 5
γ	0.01
β_min	0.5
β_max	1.5
τ1	0.45
τ2	0.55