基于外观相似性更新的相关滤波跟踪算法

doi:10.12305/j.issn.1001-506X.2022.01.16

摘要/Abstract

摘要：

针对核相关滤波(kernel correlation filter, KCF)算法在目标旋转、形变等复杂环境中容易产生模型漂移的问题, 提出了一种基于KCF自适应更新的目标跟踪算法(adaptive updating target tracking algorithm based on KCF, AUKCF)。该方法首先对响应进行多峰判断, 然后针对多峰现象使用显著性检测进行目标的重新定位, 减少模型漂移。为了保证显著性检测的准确性, 使用重检测手段进行显著性检测结果的校准。最后, 使用斯皮尔曼相关性判断目标是否存在遮挡、严重形变等问题, 并根据斯皮尔曼相关性结果决定是否进行模型的更新, 减少模型退化, 提高更新效率。在目标跟踪数据集OTB2015上进行测试, 实验结果表明, AUKCF相比KCF算法的精度和成功率分别提高14%和11.8%, 并且AUKCF算法比目前流行的深度学习算法更加简洁, 对设备性能要求更低, 算法实时性可以达到93.84 fps。

关键词: 目标跟踪, 显著性检测, 斯皮尔曼相关性, 模型退化, 模型更新

Abstract:

To solve the problem of kernel correlation filter (KCF) algorithm, which is prone to model drift in complicated environments, an algorithm using adaptive updating target tracking algorithm based on KCF(AUKCF) is proposed. This algorithm firstly makes multi-peak judgment on the response, then uses the saliency detection to relocate the target for the multi-peak phenomenon to reduce the model drift. To ensure the accuracy of the saliency detection, the redetection method is applied to calibrate the saliency detection result. Finally, Spearman correlation, which reduces model degradation and improves update efficiency is used to determine whether the target has problems such as occlusion and serious deformation, and it is determined whether to update the model according to the results of Spearman's correlation. Tested on the OTB2015 benchmark, the experimental results show that the accuracy and success rate of the AUKCF and KCF algorithms are increased by 14% and 11.8%, respectively. Compared with the current deep learning algorithm, the AUKCF algorithm has a simpler model and lower requirements for equipment, the real-time performance can reach 93.84 fps.

Key words: target tracking, saliency detection, Spearman correlation, model degradation, model update

中图分类号:

TP139

方澄, 路稳, 姬菁颖, 宋玉蒙, 梁斐菲, 罗志伟. 基于外观相似性更新的相关滤波跟踪算法[J]. 系统工程与电子技术, 2022, 44(1): 117-126.

Cheng FANG, Wen LU, Jingying JI, Yumeng SONG, Feifei LIANG, Zhiwei LUO. Correlation filter-tracking algorithm based on appearance similarity update[J]. Systems Engineering and Electronics, 2022, 44(1): 117-126.

图/表 6

图1

图2

表1

图3

图4

表2

参考文献 50

1	MAGGIO E , CAVALLARO A . Video tracking: theory and practice[M]. Oxford: Wiley, 2011.
2	ZHOU Y , WANG T , HU R , et al. Multiple kernelized correlation filters (MKCF) for extended object tracking using X-band marine radar data[J]. IEEE Trans.on Signal Processing, 2019, 67 (14): 3676- 3688. doi: 10.1109/TSP.2019.2917812
3	BOLME D S, BEVERIDGE J R, DRAPER B A, et al. Visual object tracking using adaptive correlation filters[C]//Proc. of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2010: 2544-2550.
4	HENRIQUES J F, CASEIRO R, MARTINS P, et al. Exploiting the circulant structure of tracking-by-detection with kernels[C]// Proc. of the European Conference on Computer Vision, 2012: 702-715.
5	ZHANG K H, ZHANG L, LIU Q S, et al. Fast visual tracking via dense spatio-temporal context learning[C]//Proc. of the European Conference on Computer Vision, 2014: 127-141.
6	DANELLJAN M, SHAHBAZ K F, FELSBERG M, et al. Adaptive color attributes for real-time visual tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 1090-1097.
7	POSSEGGER H, MAUTHNER T, BISCHOF H. In defense of color-based model-free tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2015: 2113-2120.
8	DALAL N, TRIGGS B. Histograms of oriented gradients for human detection[C]//Proc. of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005: 886-893.
9	HENRIQUES J F , CASEIRO R , MARTINS P , et al. High-speed tracking with kernelized correlation filters[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2014, 37 (3): 583- 596.
10	WANG N, LI S, GUPTA A, et al. Transferring rich feature hierarchies for robust visual tracking[J]. arXiv preprint arXiv: 1501.04587, 2015.
11	WANG M M, LIU Y, HUANG Z Y. Large margin object tracking with circulant feature maps[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 4021-4029.
12	WANG N, SHI J, YEUNG D Y, et al. Understanding and diagnosing visual tracking systems[C]//Proce. of the IEEE International Conference on Computer Vision, 2015: 3101-3109.
13	CHOI J, JIN C H, JEONG J, et al. Visual tracking using attention-modulated disintegration and integration[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 4321-4330.
14	LI Y, ZHU J, HOI S C H, et al. Robust estimation of similarity transformation for visual object tracking[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2019: 8666-8673.
15	BERTINETTO L, VALMADRE J, GOLODETZ S, et al. Staple: complementary learners for real-time tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 1401-1409.
16	MA C , HUANG J B , YANG X , et al. Robust visual tracking via hierarchical convolutional features[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2018, 41 (11): 2709- 2723.
17	VALMADRE J, BERTINETTO L, HENRIQUES J, et al. End-to-end representation learning for correlation filter based tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 2805-2813.
18	MA C, HUANG J B, YANG X, et al. Hierarchical convolutional features for visual tracking[C]//Proc. of the IEEE International Conference on Computer Vision, 2015: 3074-3082.
19	LUKEZIC A, MATAS J, KRISTAN M. D3S-A discriminative single shot segmentation tracker[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 7133-7142.
20	YANG T Y, XU P F, HU R B, et al. ROAM: recurrently optimizing tracking model[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 6718-6727.
21	DANELLJAN M, GOOL L V, TIMOFTE R. Probabilistic regression for visual tracking[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 7183-7192.
22	BERTINETTO L, VALMADRE J, HENRIQUES J F, et al. Fully-convolutional siamese networks for object tracking[C]//Proc. of the European Conference on Computer Vision, 2016: 850-865.
23	WANG Q, ZHANG L, BERTINETTO L, et al. Fast online object tracking and segmentation: a unifying approach[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2019: 1328-1338.
24	VOIGTLAENDER P, LUITEN J, TORR P H S, et al. Siam R-CNN: visual tracking by re-detection[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 6578-6588.
25	XU Y D, WANG Z Y, LI Z X, et al. SiamFC++: towards robust and accurate visual tracking with target estimation guidelines[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2020: 12549-12556.
26	XU T, FENG Z H, WU X J, et al. Joint group feature selection and discriminative filter learning for robust visual object tracking[C]//Proc. of the IEEE International Conference on Computer Vision, 2019: 7950-7960.
27	ITTI L , KOCH C , NIEBUR E . A model of saliency-based visual attention for rapid scene analysis[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 1998, 20 (11): 1254- 1259. doi: 10.1109/34.730558
28	HAREL J, KOCH C, PERONA P. Graph-based visual saliency[C]//Proc. of the Advances in Neural Information Processing Systems, 2007: 545-552.
29	余汉蓉, 林彬, 俞增林. 结合SAMF和视觉显著性的目标跟踪算法[J]. 电光与控制, 2021, 28 (1): 1- 5. doi: 10.3969/j.issn.1671-637X.2021.01.001
	YU H R , LIN B , YU Z L . Object tracking algorithm combining SAMF and visual saliency[J]. Electronics Optics & Control, 2021, 28 (1): 1- 5. doi: 10.3969/j.issn.1671-637X.2021.01.001
30	谢瑜, 陈莹. 融合稀疏重构图像显著性的相关滤波跟踪[J]. 计算机工程与应用, 2019, 55 (22): 106- 113, 224. doi: 10.3778/j.issn.1002-8331.1807-0234
	XIE Y , CHEN Y . Correlation filter tracking algorithm fusing image saliency via sparse reconstruction[J]. Computer Engineering and Applications, 2019, 55 (22): 106- 113, 224. doi: 10.3778/j.issn.1002-8331.1807-0234
31	DANELLJAN M, HAGER G, SHAHBAZ K F, et al. Learning spatially regularized correlation filters for visual tracking[C]//Proc. of the IEEE International Conference on Computer Vision, 2015: 4310-4318.
32	LI F, TIAN C, ZUO W M, et al. Learning spatial-temporal regularized correlation filters for visual tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 4904-4913.
33	HUANG Z Y, FU C H, LI Y M, et al. Learning aberrance repressed correlation filters for real-time UAV tracking[C]//Proc. of the IEEE International Conference on Computer Vision, 2019: 2891-2900.
34	LI Y, FU C, DING F, et al. Autotrack: towards high-performance visual tracking for UAV with automatic spatio-temporal regularization[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 11923-11932.
35	WU Y , LIM J , YANG M H . Object tracking benchmark[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2015, 37 (9): 1834- 1848. doi: 10.1109/TPAMI.2014.2388226
36	KRISTAN M, LEONARDIS A, MATAS J, et al. The sixth visual object tracking VOT2018 challenge results[C]//Proc. of the European Conference on Computer Vision Workshops, 2018.
37	DANELLJAN M, HÄGER G, KHAN F, et al. Accurate scale estimation for robust visual tracking[C]//Proc. of the British Machine Vision Conference, 2014.
38	LI Y, ZHU J K. A scale adaptive kernel correlation filter tracker with feature integration[C]//Proc. of the European Conference on Computer Vision, 2014: 254-265.
39	CHOI J, JIN C H, YUN S, et al. Attentional correlation filter network for adaptive visual tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 4807-4816.
40	WANG N, SONG Y, MA C, et al. Unsupervised deep tracking[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2019: 1308-1317.
41	LI P X, CHEN B Y, OUYANG W L, et al. Gradnet: gradient-guided network for visual object tracking[C]//Proc. of the IEEE International Conference on Computer Vision, 2019: 6162-6171.
42	ABDELPAKEY M H, SHEHATA M S, MOHAMED M M. Denssiam: end-to-end densely-siamese network with self-attention model for object tracking[C]//Proc. of the International Symposium on Visual Computing, 2018: 463-473.
43	DE A G, EVERSON R M. Part-based tracking by sampling[EB-OL]. [2020-11-10]https://arxiv-org/pdt/1805.08511.
44	BABENKO B , YANG M H , BELONGIE S . Robust object tracking with online multiple instance learning[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2010, 33 (8): 1619- 1632.
45	BATTISTONE F , PETROSINO A , SANTOPIETRO V . Watch out: embedded video tracking with BST for unmanned aerial vehicles[J]. Journal of Signal Processing Systems, 2018, 90 (6): 891- 900. doi: 10.1007/s11265-017-1279-x
46	HARE S , GOLODETZ S , SAFFARI A , et al. Struck: structured output tracking with kernels[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2015, 38 (10): 2096- 2109.
47	MARESCA M E, PETROSINO A. Clustering local motion estimates for robust and efficient object tracking[C]//Proc. of the European Conference on Computer Vision, 2014: 244-253.
48	ROSS D A , LIM J , LIN R S , et al. Incremental learning for robust visual tracking[J]. International Journal of Computer Vision, 2008, 77 (1/3): 125- 141.
49	ADAM A, RIVLIN E, SHIMSHONI I. Robust fragments-based tracking using the integral histogram[C]//Proc. of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2006: 798-805.
50	MARESCA M E, PETROSINO A. Matrioska: a multi-level approach to fast tracking by learning[C]//Proc. of the International Conference on Image Analysis and Processing, 2013: 419-428.

算法名称	显著性检测调节	自适应更新
KCF	-	-
KCFup	-	√
KCFsd	√	-
AUKCF	√	√

算法	相机抖动(准确率)	超出视野(准确率)	光照变化(准确率)	快速运动(准确率)	遮挡(准确率)	尺度变化(准确率)	准确率均值	准确率加权	准确率合并方差	EAO
AUKCF	0.522 8	0.464 6	0.476 6	0.461 0	0.425 7	0.363 5	0.452 4	0.465 4	0.469 0	0.119 9
DensSiam^[42]	0.468 7	0.472 2	0.486 9	0.441 5	0.420 2	0.348 5	0.439 7	0.445 7	0.456 5	0.167 8
PBTS^[43]	0.389 0	0.386 8	0.361 4	0.347 3	0.312 7	0.335 8	0.355 5	0.367 8	0.375 3	0.147 2
MIL^[44]	0.377 0	0.410 7	0.380 3	0.360 0	0.358 0	0.291 4	0.362 9	0.371 4	0.385 2	0.113 8
BST^[45]	0.242 0	0.308 2	0.270 6	0.236 8	0.260 8	0.236 8	0.259 2	0.262 5	0.272 6	0.113 7
struck2011^[46]	0.437 3	0.409 0	0.420 2	0.408 3	0.378 0	0.312 4	0.394 2	0.402 3	0.410 4	0.093 0
BDF^[47]	0.379 0	0.361 5	0.400 5	0.355 7	0.378 8	0.309 4	0.364 1	0.361 2	0.365 2	0.091 5
Matflow	0.402 9	0.416 9	0.405 5	0.354 9	0.360 8	0.314 1	0.375 8	0.384 3	0.398 2	0.090 0
MRSNCC	0.318 1	0.357 1	0.359 9	0.277 0	0.285 1	0.295 1	0.315 4	0.318 5	0.328 2	0.079 7
DSST^[37]	0.407 5	0.418 2	0.474 8	0.354 6	0.349 7	0.301 4	0.384 4	0.384 9	0.396 7	0.078 0
IVT^[48]	0.377 2	0.431 4	0.421 9	0.325 8	0.361 6	0.330 1	0.374 7	0.378 8	0.396 4	0.075 0
CPOINT	0.295 9	0.326 3	0.281 1	0.261 7	0.283 9	0.310 6	0.293 2	0.300 4	0.308 8	0.069 4
L1APG^[49]	0.420 3	0.467 1	0.391 0	0.378 5	0.382 1	0.299 7	0.389 8	0.407 8	0.428 3	0.068 6
FragTrack	0.392 2	0.403 7	0.373 5	0.348 0	0.325 0	0.328 4	0.361 8	0.374 3	0.390 1	0.068 2
Matrioska^[50]	0.431 3	0.437 5	0.428 9	0.383 5	0.364 8	0.295 8	0.390 3	0.402 0	0.416 0	0.064 2

[1]	仇祝令, 查宇飞, 李振宇, 李禹铭, 张鹏, 朱川. 基于多模型蒸馏的时间正则化相关滤波跟踪算法[J]. 系统工程与电子技术, 2022, 44(8): 2448-2456.
[2]	侯子林, 程婷, 彭瀚. 基于量测转换序贯滤波的GMPHD机动目标跟踪[J]. 系统工程与电子技术, 2022, 44(8): 2474-2482.
[3]	史浩然, 卢发兴, 祁江鑫, 杨光. 基于辅助信标的无人机协同目标跟踪[J]. 系统工程与电子技术, 2022, 44(7): 2302-2310.
[4]	金国栋, 薛远亮, 谭力宁, 许剑锟. 基于孪生神经网络的目标跟踪算法进展研究[J]. 系统工程与电子技术, 2022, 44(6): 1805-1822.
[5]	翟光, 王妍欣, 孙一勇. 基于低轨星网的多目标协同跟踪滤波技术[J]. 系统工程与电子技术, 2022, 44(6): 1957-1967.
[6]	王帅, 向建军, 彭芳, 唐书娟. 基于新最速下降法的目标跟踪算法[J]. 系统工程与电子技术, 2022, 44(5): 1512-1519.
[7]	辛怀声, 曹晨. 基于交互多模型的分组δ-广义标签多伯努利算法[J]. 系统工程与电子技术, 2022, 44(4): 1128-1138.
[8]	谢家豪, 黄树彩, 韦道知, 张曌宇, 王文豪. 基于P_EV准则的不确定混合多传感器联盟求解[J]. 系统工程与电子技术, 2022, 44(3): 819-826.
[9]	宋子壮, 杨嘉伟, 张东方, 王诗强, 张硕. 基于无锚框的红外多类别多目标实时跟踪网络[J]. 系统工程与电子技术, 2022, 44(2): 401-409.
[10]	姚云翔, 陈莹. 注意力机制下双模态交互融合的目标跟踪网络[J]. 系统工程与电子技术, 2022, 44(2): 410-419.
[11]	辛怀声, 宋鹏汉, 曹晨. 多模型广义标签多伯努利滤波器[J]. 系统工程与电子技术, 2022, 44(12): 3603-3613.
[12]	刘浩楠, 宋骊平. 基于核Fisher判别的群结构更新模型及群目标跟踪算法[J]. 系统工程与电子技术, 2022, 44(10): 3012-3019.
[13]	黄月平, 李小锋, 杨小冈, 齐乃新, 卢瑞涛, 张胜修. 基于相关滤波的视觉目标跟踪算法新进展[J]. 系统工程与电子技术, 2021, 43(8): 2051-2065.
[14]	赵跃新, 齐望东, 刘鹏, 袁恩, 徐兵. 三维AoA目标跟踪的二次约束卡尔曼滤波算法[J]. 系统工程与电子技术, 2021, 43(8): 2263-2272.
[15]	孙明杰, 周林, 顾金玲, 李培庚. 基于多域网络的红外目标跟踪算法[J]. 系统工程与电子技术, 2021, 43(5): 1176-1183.