基于混合粒子滤波的多目标跟踪

Journal of Systems Engineering and Electronics ›› 2009, Vol. 31 ›› Issue (8): 1795-1800.

基于混合粒子滤波的多目标跟踪

李少军, 朱振福

目标与环境光学特征国防科技重点实验室, 北京, 100854

收稿日期:2008-09-27 修回日期:2008-12-05 出版日期:2009-08-20 发布日期:2010-01-03
作者简介:李少军(1977- ),男,博士后,主要研究方向为模式识别、图像末制导、嵌入式计算.E-mail:robbieshaojun@yahoo.com.cn

Multi-target tracking based on mixtures of particle filtering

LI Shao-jun, ZHU Zhen-fu

Key Lab. of Optical Features of Targets and Environment, Beijing 100854, China

Received:2008-09-27 Revised:2008-12-05 Online:2009-08-20 Published:2010-01-03

摘要/Abstract

摘要： 针对可变数量的多个红外弱小目标的检测与跟踪问题,提出了基于混合概率密度模型的多目标先跟踪后检测方法,开发了一种t分布混合粒子滤波器.在混合粒子滤波器中,利用每个分量粒子滤波器的输出信息,根据序列似然比假设检验,检测每个被跟踪目标的存在性.通过估计目标在离散占据网格上的出现概率,检测新目标的出现.混合粒子滤波器使用单独的粒子滤波器独立估计每个被跟踪目标的状态,避免算法的计算量随着目标数量增加呈指数增长的问题.仿真实验证明混合粒子滤波器能够跟踪目标数量可变的弱小目标,能够同时检测目标的消失和出现.

Abstract: For the problem of detecting and tracking a variable number of dim small targets in IR image sequences,a multi-target track-before-detect approach based on the mixture model of probability densities is proposed,and a mixture of t distribution particle filters(MTPF) is developed for the implementation of the proposed approach.In the mixture of particle filters,the existence of each tracked target is detected by using the sequential likelihood ratio test estimated by the output of the component particle filter.New targets are detected by the appearance probabilities in the discrete occupancy grid in the image frame.The proposed approach overcomes the curse of multi-target dimensionality by independently estimating each target state with a separate particle filter,and avoids the exponential increase in the estimation complexity.Importance resampling is carried out in each component particle filter individually.Simulation experiments illustrate that the MTPF algorithm can detect and track the variable number of dim small targets in the IR images.It can also detect the disappearance and appearance of targets simultaneously.

中图分类号:

TN27

李少军, 朱振福. 基于混合粒子滤波的多目标跟踪[J]. Journal of Systems Engineering and Electronics, 2009, 31(8): 1795-1800.

LI Shao-jun, ZHU Zhen-fu. Multi-target tracking based on mixtures of particle filtering[J]. Journal of Systems Engineering and Electronics, 2009, 31(8): 1795-1800.

参考文献

[1] Kirubarajan T,Bar-Shalom Y.Probabilistic data association techniques for target tracking in clutter[C]//Proc.of the IEEE,2004,92(3):536-557.
[2] Liebens M,Sakiyama T.Miura J.Visual tracking of multiple persons in a heavy occluded space using person model and joint probabilistic data association[C]//IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems,2006:547-552.
[3] Davey S J.Tracking possibly unresolved targets with PMHT[C]//IEEE Information,Decision and Control,2007:47-52.
[4] Streit R L.Studies in probabilistic multi-hypothesis tracking and related topics[R].Naval Undersea Warfare Center,Publication SES-98-101,Newport,RI,1998.
[5] Reed I,Gagliardi R,Stotts L B.Application of three-dimensional filtering to moving target detection[J].IEEE Trans,on Aerospace and Electronic Systems,1988,24(4).327-335.
[6] Singer P F.Performance analysis of a velocity filter bank[C]//SPIE Proc.of Signal and Data Processing of Small Targets,1997,3163:96-107.
[7] Artes-Rodriguez A,Perez-Cruz F,Fernandez-Lorenzana R,et al.On the uncertainty in sequential hypothesis testing[C]//5th IEEE International Symposium on Biomedical Imaging:From Nano to Macro,2008:1223-1226.
[8] Tonissen S M,Evans R J.Performance of dynamic programming track before detect algorithm[J].IEEE Trans,on Aerospace and Electronic Systems,1996,32(4),1440-1451.
[9] Kreucher C M,Kastella K,Hero A O.Tracking multiple targets using a particle filter representation of the joint multitarget probability density[J].Signal and Data Processing of Small Targets,Proc.of SPIE,2003,5204:258-269.
[10] Kreucher C M,Kastella K,Hero A O.Multi-target tracking using the joint multitarget probability density[J].IEEE Trans,on Aerospace and Electronic Systems,2005:1396-1414.
[11] Li J,Godsill Ng W,Vermaak J S.Online multi-target detection and tracking using sequential Monte Carlo methods[C]//Proc.of 8th International Conference on Information Fusion,2005:25-28.
[12] Vermaak J,Doucet A,Perez P.Maintaining multi-modality through mixture tracking[C]//Proc.of 9th IEEE Int.Conference on Computer Vision,2003,2:1110-1116.
[13] King O,Forsyth D A.How does condensation behave with a finite number of samples[C]//Proc.of 6th European Conference Computer Vision,Dublin,Ireland,2000,1:695-709.
[14] Stone L D,Barlow C A,Corwin T L.Bayesian multiple target tracking[M].Boston:Artech House Inc.,1999.
[15] Lawrence D Stone.A Bayesian approach to multiple-target tracking[M]//Hall David L,Llinas James,ed.Handbook of multisensor data fusion.London:CRC Press,2001.
[16] Peel D,McLachlan G J.Robust mixture modeling using the t distribution[J].Statistics and Computing,2000,10:339-348.
[17] Liu Chuanhai,Rubin D B.ML estimation of the t distribution using EM and its extensions,ECM and ECME[J].Statistica Sinica,1995,5:19-39.
[18] Arulampalam M S,Maskell S,Gordon T C.A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking[J].IEEE Trans,on Signal Processing,2002,50(2):174-188.
[19] Boers Y,Driessen H.Particle filter based track before detect algorithms[C]//Proc.of SPIE,Signal and Data Processing of Small Targets,2003,5204:20-30.
[20] Boers Y,Dries H.A particle-filter-based detection scheme[J].IEEE Signal Processing Letters,2003,10(10):300-302.
[21] Samond D J,Birch H.A particle filter for track-before-detect[C]//Proc.of the American Control Conference,2001:3755-3760.