基于联合GLMB滤波器的可分辨群目标跟踪

doi:10.12305/j.issn.1001-506X.2024.04.09

摘要/Abstract

摘要：

针对联合广义标签多伯努利(joint generalized labeled multi-Bernoulli, J-GLMB)滤波算法中群目标之间距离较近、容易关联错误的问题, 结合超图匹配(hypergraph matching, HGM)提出一种基于HGM-J-GLMB滤波器的可分辨群目标跟踪算法。首先, 采用J-GLMB滤波器估计群内各目标的状态、数目及轨迹信息, 并利用HGM结果提升量测与预测状态之间的关联性能。其次, 通过图理论计算邻接矩阵, 获取群结构信息和子群数目。随后, 利用群结构信息估计协作噪声, 进而校正目标的预测状态。最后, 通过平滑算法改善滤波效果, 并设置轨迹长度阈值, 使其在平滑状态达到消除短轨迹的目的。仿真实验表明, 所提算法在线性系统下能有效提升群目标跟踪性能。

关键词: 多目标跟踪, 联合广义标签多伯努利滤波, 可分辨群目标, 超图匹配

Abstract:

Aiming at the problem of association errors between group targets close to each other in the joint generalized labeled multi Bernoulli (J-GLMB) filtering algorithm, a resolvable group target tracking algorithm based on HGM-J-GLMB filter is proposed by combining hypergraph matching (HGM). Firstly, the J-GLMB filter is used to estimate the state, number, and trajectory information of each target in the group, and the HGM results are used to improve the correlation performance between measurement states and prediction states. Secondly, the adjacency matrix is calculated by graph theory to obtain group structure information and the number of subgroups. Subsequently, the collaborative noise is estimated by group structure information to correct the predicted state of the target. Finally, the filtering effect is improved through smoothing algorithms, and a trajectory length threshold is set to achieve the goal of eliminating short trajectories while maintaining a smooth state. Simulation experiment results show that the proposed algorithm can effectively improve the performance of group target tracking in linear systems.

Key words: multi-target tracking, joint generalized labeled multi-Bernoulli (J-GLMB) filter, resolvable group target, hypergraph matching (HGM)

中图分类号:

TN953

齐美彬, 庄硕, 胡晶晶, 杨艳芳, 胡元奎. 基于联合GLMB滤波器的可分辨群目标跟踪[J]. 系统工程与电子技术, 2024, 46(4): 1212-1219.

Meibin QI, Shuo ZHUANG, Jingjing HU, Yanfang YANG, Yuankui HU. Resolvable group target tracking based on joint GLMB filter[J]. Systems Engineering and Electronics, 2024, 46(4): 1212-1219.

图/表 11

图1

图2

表1

表2

图3

图4

图5

图6

图7

图8

表3

参考文献 31

1	SINGER R A, STEIN J J. An optimal tracking filter for processing sensor data of imprecisely determined origin in surveillance systems[C]//Proc. of the IEEE Conference on Decision and Control, 1971: 171-175.
2	FORTMANN T E, BAR SHALOM Y, SCHEFFE M. Multi-target tracking using joint probabilistic data association[C]//Proc. of the IEEE 19th Conference on Decision and Control including the Symposium on Adaptive Processes, 1980: 807-812.
3	REID D . An algorithm for tracking multiple targets[J]. IEEE Trans.on Automatics Control, 1979, 24 (6): 1202- 1211.
4	ZHAO S J , WANG Y , WANG P Y , et al. Adaptive non-linear joint probabilistic data association for vehicle target tracking[J]. IEEE Access, 2021, 9, 14138- 14147. doi: 10.1109/ACCESS.2021.3052555
5	MAHLER R P . Statistical multisource-multitarget information fusion[M]. London: Artech House, 2007: 685- 686.
6	DA K , LI T C , ZHU Y F , et al. Recent advances in multisensor multitarget tracking using random finite set[J]. Frontiers of Information Technology & Electronic Engineering, 2021, 22 (1): 5- 24.
7	MAHLER R . PHD filters of higher order in target number[J]. IEEE Trans.on Aerospace and Electronic Systems, 2007, 43 (4): 1523- 1543. doi: 10.1109/TAES.2007.4441756
8	ORGUNER U, LUNDQUIST C, GRANSTRÖM K. Extended target tracking with a cardinalized probability hypothesis density filter[C]//Proc. of the 14th International Conference on Information Fusion, 2011.
9	BEARD M, REUTER S, GRANSTRÖM K, et al. A generalised labelled multi-Bernoulli filter for extended multi-target tracking[C]//Proc. of the 18th International Conference on Information Fusion, 2015: 991-998.
10	GNING A , MIHAYLOVA L , MASKELL S , et al. Group object structure and state estimation with evolving networks and Monte Carlo methods[J]. IEEE Trans.on Signal Processing, 2011, 59 (4): 1383- 1396. doi: 10.1109/TSP.2010.2103062
11	ZHU S J, LIU W F, WENG C L, et al. Multiple group targets tracking using the generalized labeled multi-Bernoulli filter[C]//Proc. of the 35th Chinese Control Conference, 2016: 4871-4876.
12	LIU W F , ZHU S J , WEN C L , et al. Structure modeling and estimation of multiple resolvable group targets via graph theory and multi-Bernoulli filter[J]. Automatica, 2018, 89, 274- 289. doi: 10.1016/j.automatica.2017.12.004
13	LIU W , CHI Y . Resolvable group state estimation with maneuver based on labeled RFS and graph theory[J]. Sensors, 2019, 19 (6): 1307. doi: 10.3390/s19061307
14	VO B N , VO B T , HOANG H G . An efficient implementation of the generalized labeled multi-Bernoulli filter[J]. IEEE Trans.on Signal Processing, 2016, 65 (8): 1975- 1987.
15	HOANG H G, VO B T, VO B N. A fast implementation of the generalized labeled multi-Bernoulli filter with joint prediction and update[C]//Proc. of the 18th International Conference on Information Fusion, 2015: 999-1006.
16	陈辉, 杜金瑞, 韩崇昭. 基于星凸形随机超曲面模型多扩展目标多伯努利滤波器[J]. 自动化学报, 2020, 46 (5): 909- 922.
	CHEN H , DU J R , HAN C Z . A multiple extended target multi-Bernouli filter based on star-convex random hypersurface model[J]. Acta Automatica Sinica, 2020, 46 (5): 909- 922.
17	DO C T , NGUYEN T T D , VAN NGUYEN H . Robust multi-sensor generalized labeled multi-Bernoulli filter[J]. Signal Processing, 2022, 192, 108368. doi: 10.1016/j.sigpro.2021.108368
18	ZASS R, SHASHUA A. Probabilistic graph and hypergraph matching[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2008.
19	NAWAZ M , KHAN S , QURESHI R , et al. Clustering based one-to-one hypergraph matching with a large number of feature points[J]. Signal Processing: Image Communication, 2019, 74, 289- 298. doi: 10.1016/j.image.2019.01.001
20	WU S G, XIAO J J. Tracking group targets using hypergraph matching in data association[C]//Proc. of the Conference on Signal and Data Processing of Small Targets, 2011: 81370.
21	YU H Y , AN W , ZHU R , et al. A hypergraph matching labeled multi-Bernoulli filter for group targets tracking[J]. IEEE Trans.on Information and Systems, 2019, 102 (10): 2077- 2081.
22	迟玉东. 面向群目标运动建模与跟踪的随机有限集方法[D]. 杭州: 杭州电子科技大学, 2020.
	CHI Y D. Random finite set method for group target motion modeling and tracking[D]. Hangzhou: Hangzhou Dianzi University, 2020.
23	ZHANG Z G , SUN J P , ZHOU H Y , et al. Group target tracking based on MS-member filters[J]. Remote Sensing, 2021, 13 (10): 1920. doi: 10.3390/rs13101920
24	SCHUHMACHER D , VO B T , VO B N . A consistent metric for performance evaluation of multi object filters[J]. IEEE Trans.on Signal Processing, 2008, 56 (8): 3447- 3457. doi: 10.1109/TSP.2008.920469
25	GAO L , BATTISTELLI G , CHISCI L , et al. Fusion-based multidetection multitarget tracking with random finite sets[J]. IEEE Trans.on Aerospace and Electronic Systems, 2021, 57 (4): 2438- 2458. doi: 10.1109/TAES.2021.3059093
26	VO B T , VO B N . Labeled random finite sets and multi-object conjugate priors[J]. IEEE Trans.on Signal Processing, 2013, 61 (13): 3460- 3475. doi: 10.1109/TSP.2013.2259822
27	VO B N , VO B T . A multi-scan labeled random finite set model for multi-object state estimation[J]. IEEE Trans.on Signal Processing, 2019, 67 (19): 4948- 4963. doi: 10.1109/TSP.2019.2928953
28	KIM D Y , VO B N , VO B T , et al. A labeled random finite set online multi-object tracker for video data[J]. Pattern Recognition, 2019, 90, 377- 389. doi: 10.1016/j.patcog.2019.02.004
29	GOSTAR A K , RATHNAYAKE T , TENNAKOON R , et al. Centralized cooperative sensor fusion for dynamic sensor network with limited field-of-view via labeled multi-Bernoulli filter[J]. IEEE Trans.on Signal Processing, 2020, 69, 878- 891.
30	NGUYEN T T D , KIM D Y . GLMB tracker with partial smoothing[J]. Sensors, 2019, 19 (20): 4419. doi: 10.3390/s19204419
31	LIANG G L , LI Q R , QI B , et al. Multitarget tracking using one time step lagged delta-generalized labeled multi-Bernoulli smoothing[J]. IEEE Access, 2020, 8, 28242- 28256. doi: 10.1109/ACCESS.2020.2971624

G	G′
$\boldsymbol{V}=\left\{z_1, z_2, \cdots, z_m\right\}$	$\boldsymbol{V}^{\prime}=\left(x_1, x_2, \cdots, x_n\right)$
$E \subseteq V^d$	$E^{\prime} \subseteq V^{\prime d}$
$e \in E$	$e^{\prime} \in E^{\prime}$
$e_{i, j}=z_i-z_j$	$e_{i, j}^{\prime}=x_{k \mid k-1, i}-x_{k \mid k-1, j}$

滤波器	J-GLMB	HGM-J-GLMB	HGM-J-GLMB+平滑
OSPA距离	11.461 3	10.737 2	8.766 5
OSPA位置估计	8.441 9	8.400 2	6.460 6
OSPA势估计	3.019 4	2.337 0	2.305 9
群OSPA势估计	3.658 4	2.718 3	2.714 4

[1]	曾舒雅, 饶彬. 基于动力学守恒定律的弹道目标关联方法[J]. 系统工程与电子技术, 2024, 46(2): 684-691.
[2]	毕文豪, 周杰, 张安, 刘力. 杂波环境下基于最大熵模糊聚类的JPDA算法[J]. 系统工程与电子技术, 2023, 45(7): 1920-1927.
[3]	刘政玮, 陈映, 鲁耀兵. 适用于多目标轨迹小角度交叉的PHD滤波器[J]. 系统工程与电子技术, 2023, 45(4): 982-990.
[4]	安雷, 李召瑞, 吉兵. 杂波环境下可移动主被动传感器长时调度方法[J]. 系统工程与电子技术, 2023, 45(1): 165-174.
[5]	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.
[6]	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.
[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]	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.
[9]	尹立凡, 张奕群, 王硕, 孙承钢. 直方图概率多假设跟踪方法技术综述[J]. 系统工程与电子技术, 2021, 43(11): 3118-3125.
[10]	李正杰, 谢军伟, 张浩为, 张昭建. 基于集中式MIMO雷达的功率带宽联合分配算法[J]. 系统工程与电子技术, 2020, 42(5): 1041-1049.
[11]	卢晓东, 崔涛, 王伟, 程承. 考虑配准时空误差的多目标标签多伯努利滤波[J]. 系统工程与电子技术, 2020, 42(4): 904-911.
[12]	彭华甫, 黄高明, 田威, 邱昊. 基于幅度信息的标签多伯努利滤波算法[J]. 系统工程与电子技术, 2018, 40(12): 2636-2641.
[13]	刘建业, 王华, 周晚萌. 基于GA-SA的低轨星座传感器资源调度算法[J]. 系统工程与电子技术, 2018, 40(11): 2476-.
[14]	宋喜玉, 郑娜娥, 高留洋. 多目标跟踪的分布式MIMO雷达有限阵元选取算法[J]. 系统工程与电子技术, 2017, 39(9): 1992-1997.
[15]	谷晓琳, 周石琳, 雷琳. 基于m-best数据关联和小轨迹关联多目标跟踪算法[J]. 系统工程与电子技术, 2017, 39(7): 1640-1646.