A Multi-Feature Particle Filter Vehicle Tracking Algorithm Based on Adaptive Interpolation Moth-Flame Optimization

doi:10.16183/j.cnki.jsjtu.2021.037

Abstract

Abstract:

In order to solve the problem of low accuracy and the poor global searching ability of the moth-flame optimization algorithm, an improved adaptive interpolation moth-flame optimization algorithm is proposed, which is embedded into multi-feature particle filter to optimize. Besides, a multi-feature particle filter vehicle tracking algorithm based on adaptive interpolation moth-flame optimization is constructed. First, adaptive weights are added to the moths’ position updating mechanism to improve the global searching ability of the proposed algorithm. Next, the adaptive interpolation moth-flame optimization algorithm is used to optimize the sampling process. Then, in combination with the multi-feature adaptive fusion particle filter vehicle tracking algorithm, the particle distribution according to the latest observation information is continuously adjusted, so that the particles in the low weight layer can move to the area with higher weight to enhance the particle quality and avoid sample degradation. The experimental results show that the proposed algorithm can effectively reduce the number of sample particles required for state prediction, improve the tracking performance of the algorithm, and track the target vehicle accurately and stably under the interferences of occlusion, illumination, attitude, and scale changes.

Key words: vehicle tracking, particle filter, multi-feature fusion, moth-flame

CLC Number:

TP391.41

HUANG He, WU Kun, LI Xinrui, WANG Jun, WANG Huifeng, RU Feng. A Multi-Feature Particle Filter Vehicle Tracking Algorithm Based on Adaptive Interpolation Moth-Flame Optimization[J]. Journal of Shanghai Jiao Tong University, 2022, 56(2): 143-155.

Figures/Tables 14

Fig.1

Fig.2

Fig.3

Tab.1

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Tab.2

Objective evaluation table of Seq1

算法	N	$ε - err$	$R - OS$ /%	λ/%	v_R/ (帧·s^-1)
MAFPF	20	16	56.98	63.33	66
	100	7	74.72	98	16
多特征AWPSO- PF车辆跟踪	20	6	78.57	99.67	11
AIMFO-PF	20	4	80.99	100	27

Tab.2

Fig.9

Tab.3

Objective evaluation table of Seq2

算法	N	$ε - err$	$R - OS$ /%	λ/%	v_R/ (帧·s^-1)
MAFPF	20	10	60.73	80	99
	100	6	72.02	100	24
多特征AWPSO- PF跟踪	20	5	73.83	100	18
AIMFO-PF	20	4	75.29	100	30

Tab.3

Fig.10

Tab.4

Objective evaluation table of Seq3

算法	N	$ε - err$	$R - OS$ /%	λ/%	v_R/ (帧·s^-1)
MAFPF	20	4	59.18	63.67	57
	100	3	61.14	64	27
多特征 AWPSO-PF跟踪	20	3	73.68	97	21
AIMFO-PF	20	2	76.59	97.67	29

Tab.4

References 15

[1]	GUCHHAIT A, B M, D K. A hybrid V2V system for collision-free high-speed Internet access in intelligent transportation system[J]. Transactions on Emerging Telecommunications Technologies, 2018, 29(3):1-9.
[2]	ISARD M, BLAKE A. Condensation—Conditional density propagation for visual tracking[J]. International Journal of Computer Vision, 1998, 29(1):5-28. doi: 10.1023/A:1008078328650 URL
[3]	赵康, 王正勇, 何小海, 等. 基于改进粒子滤波的视觉目标跟踪[J]. 四川大学学报(自然科学版), 2019, 56(5):875-882.
	ZHAO Kang, WANG Zhengyong, HE Xiaohai, et al. Visual object tracking based on improved particle filter[J]. Journal of Sichuan University (Natural Science Edition), 2019, 56(5):875-882.
[4]	金志刚, 卫津津, 罗咏梅, 等. 基于改进的颜色和SURF特征的粒子滤波目标跟踪[J]. 计算机工程与应用, 2015, 51(22):193-198.
	JIN Zhigang, WEI Jinjin, LUO Yongmei, et al. Targets tracking based on improved color histogram and SURF features using particle filter[J]. Computer Engineering and Applications, 2015, 51(22):193-198.
[5]	LIU C G, LIU P, LIU J F, et al. 2D articulated pose tracking using particle filter with partitioned sampling and model constraints[J]. Journal of Intelligent and Robotic Systems, 2010, 58(2):109-124. doi: 10.1007/s10846-009-9346-6 URL
[6]	张琪, 胡昌华, 乔玉坤. 基于权值选择的粒子滤波算法研究[J]. 控制与决策, 2008, 23(1):117-120.
	ZHANG Qi, HU Changhua, QIAO Yukun. Particle filter algorithm based on weight selected[J]. Control and Decision, 2008, 23(1):117-120.
[7]	田梦楚, 薄煜明, 吴盘龙, 等. 基于萤火虫优化粒子滤波的新型机动目标跟踪算法[J]. 控制与决策, 2017, 32(10):1758-1766.
	TIAN Mengchu, BO Yuming, WU Panlong, et al. Novel target tracking method based on firefly algorithm optimized particle filter[J]. Control and Decision, 2017, 32(10):1758-1766.
[8]	NARAYANA M, NENAVATH H, CHAVAN S, et al. Intelligent visual object tracking with particle filter based on modified grey wolf optimizer[J]. Optik, 2019, 193:162913. doi: 10.1016/j.ijleo.2019.06.013 URL
[9]	胡振涛, 潘泉, 梁彦, 等. 基于进化采样的粒子滤波算法[J]. 控制理论与应用, 2009, 26(3):269-273.
	HU Zhentao, PAN Quan, LIANG Yan, et al. The particle filter algorithm based on evolution sampling[J]. Control Theory & Applications, 2009, 26(3):269-273.
[10]	EBERHART R, KENNEDY J. A new optimizer using particle swarm theory[C]//MHS’95. Proceedings of the Sixth International Symposium on Micro Machine and Human Science. Piscataway, NJ, USA: IEEE, 1995: 39-43.
[11]	MIRJALILI S. Moth-flame optimization algorithm: A novel nature-inspired heuristic paradigm[J]. Knowledge-Based Systems, 2015, 89:228-249. doi: 10.1016/j.knosys.2015.07.006 URL
[12]	TARIQ R, IQBAL Z, AADIL F. IMOC: Optimization technique for drone-assisted VANET (DAV) based on moth flame optimization[J]. Wireless Communications and Mobile Computing, 2020, 2020:1-29.
[13]	SAYED F, KAMEL S, TAHER M A, et al. Enhancing power system loadability and optimal load shedding based on TCSC allocation using improved moth flame optimization algorithm[J]. Electrical Engineering, 2021, 103(1):205-225. doi: 10.1007/s00202-020-01072-w URL
[14]	SHI J Y, ZHANG D Y, XUE F, et al. Moth-flame optimization-based maximum power point tracking for photovoltaic systems under partial shading conditions[J]. Journal of Power Electronics, 2019, 19(5):1248-1258.
[15]	刘峰, 韩艳丽, 王铎. 自适应权重粒子群优化的粒子滤波算法[J]. 计算机仿真, 2013, 30(11):330-333.
	LIU Feng, HAN Yanli, WANG Duo. Particle swarm optimization particle filtering algorithm based on adaptive weight[J]. Computer Simulation, 2013, 30(11):330-333.

测试函数	算法	最优值	均值	标准差	测试函数	算法	最优值	均值	标准差
Sphere	AWPSO	0.79	2.89	2.01	Ackley	AWPSO	1.58	3.41	2.11
	MFO	1.28×10^-14	3.18×10^-14	4.97×10^-14		MFO	1.84×10^-8	3.58×10^-7	3.75×10^-7
	IMFO	2.61×10^-19	1.44×10^-15	3.62×10^-15		IMFO	2.49×10^-10	2.82×10^-8	4.61×10^-8
	AMFO	8.16×10^-44	4.44×10^-39	3.94×10^-39		AMFO	8.88×10^-16	2.01×10^-14	5.61×10^-14
	AIMFO	1.23×10^-48	5.46×10^-44	7.12×10^-44		AIMFO	8.88×10^-16	8.88×10^-16	8.88×10^-16
Schwefel’s 2.22	AWPSO	0.56	2.01	0.69	Penalized1	AWPSO	1.03	3.98	2.61
	MFO	8.66×10^-10	5.21×10^-9	3.82×10^-9		MFO	6.45×10^-16	2.59×10^-12	5.15×10^-11
	IMFO	1.31×10^-12	1.82×10^-10	2.99×10^-10		IMFO	5.35×10^-16	5.52×10^-15	7.45×10^-15
	AMFO	5.16×10^-24	8.41×10^-23	9.51×10^-23		AMFO	4.83×10^-16	2.53×10^-15	1.32×10^-15
	AIMFO	2.13×10^-28	5.05×10^-27	6.48×10^-27		AIMFO	2.12×10^-17	8.11×10^-16	1.21×10^-16