基于双马尔可夫链的SMC-CBMeMBer滤波
|
摘要
大部分多目标跟踪滤波器都是假设目标及其量测符合隐式马尔可夫链(hidden Markov chain,HMC)模型,而HMC模型隐含的独立性假定在很多实际应用中是无效的,双马尔可夫链(pairwise Markov chain,PMC)模型相对于HMC模型更具有普适性。已有的基于PMC模型的势均衡多目标多伯努利(cardinality balanced multitarget multi-Bernoulli,CBMeMBer)滤波的高斯混合实现仅适用于线性高斯系统,针对基于PMC模型的非线性多目标跟踪系统,将每一条假设航迹的伯努利随机有限集用一组加权粒子来近似,提出了基于PMC模型的势均衡多目标多伯努利滤波的序贯蒙特卡罗(sequential Monte Carlo,SMC)方法实现(SMC-PMC-CBMeMBer)滤波。仿真实验结果验证了SMC-PMC-CBMeMBer算法的有效性,在基于PMC模型的非线性多目标跟踪系统中,SMC-PMC-CBMeMBer算法性能优于基于HMC模型的SMC-CBMeMBer滤波器和基于PMC模型的SMC-PHD滤波器。
Most multi-target tracking filters assume that one target and its observation follow a hidden Markov chain(HMC)model,but the implicit independence assumption of the HMC model is invalid in many practical applications,and a pairwise Markov chain(PMC)model is more universally suitable than the traditional HMC model.The existing Gauss mixture implementation of cardinality balanced multi-target multi-Bernoulli(CBMeMBer)filter based on the PMC model is only applicable to the linear Gauss system.Each hypothetical path of the Bernoulli random finite set is approximated by a set of weighted particles,and then the sequential Monte Carlo(SMC)implementation of the PMC-CBMeMBer filter is proposed for nonlinear systems.The experimental results show that SMC-PMC-CBMeMBer filter has better tracking performance than the SMC-HMCCBMeMBer filter and the SMC-PMC-PHD filter.
Most multi-target tracking filters assume that one target and its observation follow a hidden Markov chain(HMC)model,but the implicit independence assumption of the HMC model is invalid in many practical applications,and a pairwise Markov chain(PMC)model is more universally suitable than the traditional HMC model.The existing Gauss mixture implementation of cardinality balanced multi-target multi-Bernoulli(CBMeMBer)filter based on the PMC model is only applicable to the linear Gauss system.Each hypothetical path of the Bernoulli random finite set is approximated by a set of weighted particles,and then the sequential Monte Carlo(SMC)implementation of the PMC-CBMeMBer filter is proposed for nonlinear systems.The experimental results show that SMC-PMC-CBMeMBer filter has better tracking performance than the SMC-HMCCBMeMBer filter and the SMC-PMC-PHD filter.
引文
[1]MAHLER R.Advances in statistical multisource-multitarget information fusion[M].Norwood,MA,USA:ArtechHouse,2014.
[2]MAHLER R.Multitarget bayes filtering via first-order multitarget moments[J].IEEE Trans.on Aerospace and Electronic Systems,2003,39(4):1152-1178.
[3]DARYASAFAR N,SADEGHZADEH R A,NASER-MOGHA-DASI M.A technique for multitarget tracking in synthetic aperture radar spotlight imaging mode based on promoted PHD filtering approach[J].Radio Science,2017,52(2):248-258.
[4]MAHLER R.PHD filters of higher order in target number[J].IEEE Trans.on Aerospace and Electronic Systems,2007,43(4):1523-1543.
[5]DONG P,JING Z L,GONG D R,et al.Maneuvering multi-target tracking based on variable structure multiple model GMCPHDfilter[J].Signal Processing,2017,141(1):158-167.
[6]VO B T,VO B N,CANTONI A.The cardinality balanced multitarget multi-Bernoulli filter and its implementations[J].IEEE Trans.on Aerospace and Electronic Systems,2009,57(2):409-423.
[7]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,2017,65(8):1975-1987.
[8]WU W H,JIANG J,LIU W J,et al.Augmented state GM-PHDfilter with registration errors for multi-target tracking by Doppler radars[J].Signal Processing,2016,120(1):117-128.
[9]ZHANG Y Q,JI H B.A robust and fast partitioning algorithm for extended target tracking using a Gaussian inverse Wishart PHD filter[J].Knowledge-Based Systems,2016,95(1):125-141.
[10]LI T CH,SUN SH D,BOLIC'M,et al.Algorithm design for parallel implementation of the SMC-PHD filter[J].Signal Processing,2016,119(1):115-127.
[11]GAO Y Y,JIANG D F,LIU M.Particle-gating SMC-PHD filter[J].Signal Processing,2017,130(1):64-73.
[12]PETETIN Y,DESBOUVRIES F.Bayesian multi-object filtering for pairwise Markov chains[J].IEEE Trans.on Signal Processing,2013,61(18):4481-4490.
[13]PIECZYNSKI W.Pairwise Markov chains and Bayesian unsupervised fusion[C]∥Proc.of the International Conference on Information Fusion,2000:24-31.
[14]PIECZYNSKI W.Pairwise Markov chains[J].IEEE Trans.on Pattern Analysis and Machine Intelligence,2003,25(5):634-639.
[15]LIU J Y,WANG CH P,WANG W.Particle probability hypothesis density filter based on pairwise Markov chains[EB/OL].[2018-12-28].arXivpreprint,arXiv:1811.12211,2018.
[16]MAHLER R.Tracking targets with pairwise-Markov dynamics[C]∥Proc.of the International Conference on Information Fusion,2015:280-286.
[17]张光华,韩崇昭,连峰,等.Pairwise马尔可夫模型下的势均衡多目标多伯努利滤波器[J].自动化学报,2017,43(12):2100-2108.ZHANG G H,HAN C Z,LIAN F,et al.Cardinality balanced multi-target multi-Bernoulli filter for pairwise Markov model[J].Acta Automatica Sinica,2017,43(12):2100-2108.
[18]MAHLER R.On multitarget pairwise-Markov models[J].Society of Photo-Optical Instrumentation Engineers,2015,9474,94740D:1-12.
[19]LIU C H,SUN J P,ZHANG X W.Multi-target tracking algorithm based on noise-adaptive cardinality-balanced multi-Bernoulli filter[C]∥Proc.of the 13th IEEE International Conference on Signal Processing(ICSP),2017:1471-1475.
[20]YANG F,ZHANG W Y,LIANG Y,et al.Cardinality balanced multi-target multi-Bernoulli filter for target tracking with amplitude information[C]∥Proc.of the 19th International Conference on Information Fusion,2016:958-964.
[21]CHEN H,HAN C Z.A new sequence Monte Carlo implementation of cardinality balanced multi-target multi-Bernoulli filter[J].Acta Automatica Sinaca,2016,42(1):26-36.
[22]曹倬,冯新喜,蒲磊.基于高斯混合概率假设密度滤波器的扩展目标跟踪算法[J].系统工程与电子技术,2017,39(3):494-499.CAO Z,FENG X X,PU L.Extended targets tracking algorithm based on Gaussian-mixture probability hypothesis density filter[J].Systems Engineering and Electronic,2017,39(3):494-499.
[23]LI D,HOU C,YI D.Multi-Bernoulli smoother for multi-target tracking[J].Aerospace Science and Technology,2016,48(1):234-245.
[24]YANG J L,LI P,YANG L,et al.An improved ET-GM-PHDfilter for multiple closely-spaced extended target tracking[J].International Journal of Control,Automation and Systems,2017,15(1):468-472.
[2]MAHLER R.Multitarget bayes filtering via first-order multitarget moments[J].IEEE Trans.on Aerospace and Electronic Systems,2003,39(4):1152-1178.
[3]DARYASAFAR N,SADEGHZADEH R A,NASER-MOGHA-DASI M.A technique for multitarget tracking in synthetic aperture radar spotlight imaging mode based on promoted PHD filtering approach[J].Radio Science,2017,52(2):248-258.
[4]MAHLER R.PHD filters of higher order in target number[J].IEEE Trans.on Aerospace and Electronic Systems,2007,43(4):1523-1543.
[5]DONG P,JING Z L,GONG D R,et al.Maneuvering multi-target tracking based on variable structure multiple model GMCPHDfilter[J].Signal Processing,2017,141(1):158-167.
[6]VO B T,VO B N,CANTONI A.The cardinality balanced multitarget multi-Bernoulli filter and its implementations[J].IEEE Trans.on Aerospace and Electronic Systems,2009,57(2):409-423.
[7]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,2017,65(8):1975-1987.
[8]WU W H,JIANG J,LIU W J,et al.Augmented state GM-PHDfilter with registration errors for multi-target tracking by Doppler radars[J].Signal Processing,2016,120(1):117-128.
[9]ZHANG Y Q,JI H B.A robust and fast partitioning algorithm for extended target tracking using a Gaussian inverse Wishart PHD filter[J].Knowledge-Based Systems,2016,95(1):125-141.
[10]LI T CH,SUN SH D,BOLIC'M,et al.Algorithm design for parallel implementation of the SMC-PHD filter[J].Signal Processing,2016,119(1):115-127.
[11]GAO Y Y,JIANG D F,LIU M.Particle-gating SMC-PHD filter[J].Signal Processing,2017,130(1):64-73.
[12]PETETIN Y,DESBOUVRIES F.Bayesian multi-object filtering for pairwise Markov chains[J].IEEE Trans.on Signal Processing,2013,61(18):4481-4490.
[13]PIECZYNSKI W.Pairwise Markov chains and Bayesian unsupervised fusion[C]∥Proc.of the International Conference on Information Fusion,2000:24-31.
[14]PIECZYNSKI W.Pairwise Markov chains[J].IEEE Trans.on Pattern Analysis and Machine Intelligence,2003,25(5):634-639.
[15]LIU J Y,WANG CH P,WANG W.Particle probability hypothesis density filter based on pairwise Markov chains[EB/OL].[2018-12-28].arXivpreprint,arXiv:1811.12211,2018.
[16]MAHLER R.Tracking targets with pairwise-Markov dynamics[C]∥Proc.of the International Conference on Information Fusion,2015:280-286.
[17]张光华,韩崇昭,连峰,等.Pairwise马尔可夫模型下的势均衡多目标多伯努利滤波器[J].自动化学报,2017,43(12):2100-2108.ZHANG G H,HAN C Z,LIAN F,et al.Cardinality balanced multi-target multi-Bernoulli filter for pairwise Markov model[J].Acta Automatica Sinica,2017,43(12):2100-2108.
[18]MAHLER R.On multitarget pairwise-Markov models[J].Society of Photo-Optical Instrumentation Engineers,2015,9474,94740D:1-12.
[19]LIU C H,SUN J P,ZHANG X W.Multi-target tracking algorithm based on noise-adaptive cardinality-balanced multi-Bernoulli filter[C]∥Proc.of the 13th IEEE International Conference on Signal Processing(ICSP),2017:1471-1475.
[20]YANG F,ZHANG W Y,LIANG Y,et al.Cardinality balanced multi-target multi-Bernoulli filter for target tracking with amplitude information[C]∥Proc.of the 19th International Conference on Information Fusion,2016:958-964.
[21]CHEN H,HAN C Z.A new sequence Monte Carlo implementation of cardinality balanced multi-target multi-Bernoulli filter[J].Acta Automatica Sinaca,2016,42(1):26-36.
[22]曹倬,冯新喜,蒲磊.基于高斯混合概率假设密度滤波器的扩展目标跟踪算法[J].系统工程与电子技术,2017,39(3):494-499.CAO Z,FENG X X,PU L.Extended targets tracking algorithm based on Gaussian-mixture probability hypothesis density filter[J].Systems Engineering and Electronic,2017,39(3):494-499.
[23]LI D,HOU C,YI D.Multi-Bernoulli smoother for multi-target tracking[J].Aerospace Science and Technology,2016,48(1):234-245.
[24]YANG J L,LI P,YANG L,et al.An improved ET-GM-PHDfilter for multiple closely-spaced extended target tracking[J].International Journal of Control,Automation and Systems,2017,15(1):468-472.