基于核Fisher判别的群结构更新模型及群目标跟踪算法

doi:10.12305/j.issn.1001-506X.2022.10.03

Abstract

Abstract:

The traditional group structure model, such as the group evolution network model, can judge the group splitting and merging by comparing the Mahalanobis distance between two targets and by the size of the threshold set according to the priori knowledge. As the tracking effect depends on the threshold, it is difficult to deal with various complex situations in the group target tracking. In this paper, the problem of grouping is regarded as a binary classification problem, a group structure update model based on kernel Fisher discriminant analysis (KFDA) is proposed. The group structure update model is obtained by off-line training, which meets the characteristics of group splitting and group merging, and is directly used for group structure updating. The simulation experiments of group target tracking combined with the box particle probability hypothesis density (BP-PHD) algorithm show that, compared with the group evolution network model, the proposed group structure update model is more accurate in the estimation of group structure, more stable in the estimation of number, and has better performance in the group target tracking.

Key words: group evolution network model, kernel Fisher discriminant analysis (KFDA), group target tracking, box particle filtering

CLC Number:

TN953

Haonan LIU, Liping SONG. Group structure update model and group target tracking algorithm based on kernel Fisher discriminant[J]. Systems Engineering and Electronics, 2022, 44(10): 3012-3019.

Figures/Tables 14

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Table 1

Fig.6

Fig.7

Fig.8

Fig.9

Fig.10

Fig.11

Fig.12

Table 2

References 20

1	MAHLER R P S . Multitarget Bayes filtering via first-order multitarget moments[J]. IEEE Trans.on Aerospace and Electronic Systems, 2004, 39 (4): 1152- 1178.
2	VO B N , MA W K . The Gaussian mixture probability hypothesis density filter[J]. IEEE Trans.on Signal Processing, 2006, 54 (11): 4091- 4104. doi: 10.1109/TSP.2006.881190
3	SIDENBLADH H. Multi-target particle filtering for the probability hypothesis density[C]//Proc. of the International Conference on Information Fusion, 2003: 800-806.
4	SCHIKORA M , GNING A , MIHAYLOVA L , et al. Box-particle probability hypothesis density filtering[J]. IEEE Trans.on Aerospace and Electronic Systems, 2014, 50 (3): 1660- 1672. doi: 10.1109/TAES.2014.120238
5	朱书军. 基于随机有限集的可分辨群目标跟踪算法研究[D]. 杭州: 杭州电子科技大学, 2017: 29-30.
	ZHU S J. Multiple resolvable groups tracking by using random finite set[D]. Hangzhou: Hangzhou Dianzi University, 2017: 29-30.
6	权太范. 目标跟踪新理论与技术[M]. 北京: 国防工业出版社, 2009.
	QUAN T F . Target tracking: advanced theory and techniques[M]. Beijing: National Defence Industry Press, 2009.
7	耿文东, 王元钦, 董正宏. 群目标跟踪[M]. 北京: 国防工业出版社, 2014.
	GENG W D , WANG Y Q , DONG Z H . Group-targets tracking[M]. Beijing: National Defence Industry Press, 2014.
8	AFTAB W , HOSTETTLER R , DE FREITAS A , et al. Spatio-temporal Gaussian process models for extended and group object tracking with irregular shapes[J]. IEEE Trans.on Vehicular Technology, 2019, 68 (3): 2137- 2151. doi: 10.1109/TVT.2019.2891006
9	GNING A, MIHAYLOVA L, MASKELL S, et al. Evolving networks for group object motion estimation[C]//Proc. of the IET Seminar on Target Tracking and Data Fusion: Algorithms and Applications, 2008: 99-106.
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	程轩, 宋骊平, 姬红兵. 基于演化网络模型的箱粒子CPHD群目标跟踪[J]. 系统工程与电子技术, 2018, 40 (5): 961- 967.
	CHENG X , SONG L P , JI H B . Box particle CPHD filter for group targets tracking based on evolution network model[J]. Systems Engineering and Electronics, 2018, 40 (5): 961- 967.
12	宋骊平, 刘宇航, 程轩. 箱粒子PHD演化网络群目标跟踪算法[J]. 控制与决策, 2018, 33 (1): 74- 80.
	SONG L P , LIU Y H , CHENG X . Box-particle evolution network PHD filter for group targets tracking[J]. Control and Decision, 2018, 33 (1): 74- 80.
13	李振兴, 刘进忙, 李松, 等. 基于箱式粒子滤波的群目标跟踪算法[J]. 自动化学报, 2015, 41 (4): 785- 798. doi: 10.16383/j.aas.2015.c140222
	LI Z X , LIU J M , LI S , et al. Group targets tracking algorithm based on box particle filter[J]. Acta Automatica Sinica, 2015, 41 (4): 785- 798. doi: 10.16383/j.aas.2015.c140222
14	杜伟, 房立清, 齐子元. 一种邻域自适应半监督局部Fisher判别分析算法[J]. 计算机应用研究, 2019, 36 (1): 99- 102. 99-102, 118
	DU W , FANG L Q , QI Z Y . Neighborhood adaptive semi-supervised local Fisher discriminant analysis algorithm[J]. Application Research of Computers, 2019, 36 (1): 99- 102. 99-102, 118
15	YANG W Y , LIANG W , XIN L , et al. Subspace semi-supervised Fisher discriminant analysis[J]. Acta Automatica Sinica, 2009, 35 (12): 1513- 1519. doi: 10.1016/S1874-1029(08)60120-2
16	郑建炜, 王万良, 姚晓敏, 等. 张量局部Fisher判别分析的人脸识别[J]. 自动化学报, 2012, 38 (9): 1485- 1495.
	ZHENG J W , WANG W L , YAO X M , et al. Face recognition using tensor local Fisher discriminant analysis[J]. Acta Automatica Sinica, 2012, 38 (9): 1485- 1495.
17	MIKA S, RATSCH G, WESTON J, et al. Fisher discriminant analysis with kernel[C]//Proc. of the IEEE Signal Processing Society Workshop, 1999: 41-48.
18	SCHIKORA M , GNING A , MIHAYLOVA L , et al. Box-particle probability hypothesis density filtering[J]. IEEE Trans.on Aerospace and Electronic Systems, 2014, 50 (3): 1660- 1672.
19	GHOJOGH B, KARRAY F, CROWLEY M. Fisher and kernel Fisher discriminant analysis: tutorial[EB/OL]. [2021-01-20]. https://arXiv.org/abs/1906.09436.
20	LI X R , JILKOV V . A survey of maneuvering target tracking part I: dynamic models[J]. IEEE Trans.on Aerospace and Electronic Systems, 2003, 39 (4): 1333- 1364.

目标位置	不同时刻
目标位置	2~12 s	13~29 s	30~40 s
1		[400, -20, 400, -23]
2		[400, -7, 470, -23]
3		[-400, 20, 410, -23]
4		[405, -7, 470, -23]
5		[-400, 22-300, 15]
6		[-405, 22, -300, 15]
7	[-400, 22, -300, 15]	[-400, 22, -300, 15]	[~, 18, ~, -15]
8	[-400, 20, 0, 14]	[~, 20, ~, -23]	[~, 20, ~, -23]

索引	群结构划分
1	g₁={1, 3}, g₂={2, 4}, g₃={5, 6, 7}, g₄={8}
2	g₁={1, 3, 8}, g₂={2, 4}g₃={5}, g₄={6, 7}
3	g₁={1, 3, 8}, g₂={2, 4}, g₃={5, 6, 7}
4	g₁={1, 8}, g₂={3}, g₃={2, 4}, g₄={5, 6, 7}
5	g₁={1, 3, 8}, g₂={2, 4}, g₃={5, 6}, g₄={7}
6	g₁={1, 3, 8}, g₂={2}, g₃={4}, g₄={5, 6, 7}
7	g₁={1, 3, 8}, g₂={2, 4}, g₃={5, 6}, g₄={7}
8	g₁={1}, g₂={2, 4}, g₃={3, 8}, g₄={5, 6, 7}
9	g₁={1, 3, 8}, g₂={2, 4}, g₃={5}, g₄={6, 7}
10	g₁={1, 3, 8}, g₂={2, 4}, g₃={5, 7}, g₄={6}
11	g₁={1}, g₂={3}, g₃={2, 4}, g₄={5, 6, 7}, g₅={8}
12	g₁={1, 3}, g₂={2}, g₃={4}, g₄={5, 6, 7}, g₅={8}
13	g₁={1, 3}, g₂={2, 4}, g₃={5}, g₄={6, 7}, g₅={8}
14	g₁={1, 3}, g₂={2, 4}, g₃={6}, g₄={5, 7}, g₅={8}
15	g₁={1, 3}, g₂={2, 4}, g₃={5, 6}, g₄={7}, g₅={8}
16	g₁={1}, g₂={2, 4}, g₃={3}, g₄={5, 7}, g₅={6}, g₆={8}
17	g₁={1}, g₂={2, 4}, g₃={3}, g₄={5, 6}, g₅={7}, g₆={8}
18	g₁={1, 3}, g₂={2, 4}, g₃={5}, g₄={6}, g₅={7}, g₆={8}

[1]	CHENG Xuan, SONG Liping, JI Hongbing, ZOU Zhibin. Group target tracking algorithm based on labeled box particle probability hypothesis density [J]. Systems Engineering and Electronics, 2019, 41(8): 1677-1685.
[2]	CHENG Xuan, SONG Liping, JI Hongbing. Box particle CPHD filter for group targets tracking based on evolution network model [J]. Systems Engineering and Electronics, 2018, 40(5): 961-967.
[3]	CHANG Zhi-peng，CHENG Long-sheng. Automatic optimization algorithm of multiple parameters for kernel Fisher discriminant analysis [J]. Journal of Systems Engineering and Electronics, 2013, 35(1): 212-217.

Group structure update model and group target tracking algorithm based on kernel Fisher discriminant

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 14

References 20

Related Articles 3

Recommended Articles

Metrics

Comments