基于双马尔科夫链的势概率假设密度滤波
|
摘要
针对已有的基于双马尔科夫链(PMC)模型的势概率假设密度(PMC-CPHD)滤波算法无法实现的问题,将PMC-CPHD算法改进为多项式形式以便于算法的实现,并给出了改进算法的高斯混合(GM)实现。实验结果表明给出的GM实现能够有效实现多目标跟踪,并且比基于PMC模型的概率假设密度(PMC-PHD)算法的GM实现提高了目标个数估计的稳定性。
In view of the problem that the Cardinalized Probability Hypothesis Density(CPHD) probability hypothesis density filtering algorithm based on the Pairwise Markov Chains(PMC) model(PMC-CPHD) is not suitable for implementation, the PMC-CPHD algorithm is modified into a polynomial form to facilitate implementation, and the Gauss Mixture(GM) implementation of the improved algorithm is given. The experimental results show that the given GM implementation realizes multitarget tracking effectively, and improves the stability of the target number estimation compared with the GM implementation of the probability hypothesis density filtering algorithm based on the PMC model(PMC-PHD).
In view of the problem that the Cardinalized Probability Hypothesis Density(CPHD) probability hypothesis density filtering algorithm based on the Pairwise Markov Chains(PMC) model(PMC-CPHD) is not suitable for implementation, the PMC-CPHD algorithm is modified into a polynomial form to facilitate implementation, and the Gauss Mixture(GM) implementation of the improved algorithm is given. The experimental results show that the given GM implementation realizes multitarget tracking effectively, and improves the stability of the target number estimation compared with the GM implementation of the probability hypothesis density filtering algorithm based on the PMC model(PMC-PHD).
引文
[1]GAO Yiyue,JIANG Defu,and LIU Ming.Particle-gating SMC-PHD filter[J].Signal Processing,2017,130:64-73.doi:10.1016/j.sigpro.2016.06.017.
[2]杨丹,姬红兵,张永权.未知杂波条件下样本校正的势估计概率假设密度滤波算法[J].电子与信息学报,2018,40(4):912-919.doi:10.11999/JEIT170666.YANG Dan,JI Hongbing,and ZHANG Yongquan.Acardinalized probability hypothesis density filter with unknown clutter estimation using corrected sample set[J].Journal of Electronics&Information Technology,2018,40(4):912-919.doi:10.11999/JEIT170666.
[3]袁常顺,王俊,孙进平,等.一种幅度信息辅助多伯努利滤波算法[J].电子与信息学报,2016,38(2):464-471.doi:10.11999/JEIT150683.YUAN Changshun,WANG Jun,SUN Jinping,et al.Amulti-bernoulli filtering algorithm using amplitude information[J].Journal of Electronics&Information Technology,2016,38(2):464-471.doi:10.11999/JEIT150683.
[4]陈辉,韩崇昭.Rao-Blackwellized粒子势均衡多目标多伯努利滤波[J].控制理论与应用,2016,33(2):146-153.doi:10.7641/CTA.2016.50588.CHEN Hui and HAN Chongzhao.Rao-Blackwellized particle cardinality balanced multi-target multi-bernoulli filter[J].Control Theory&Applications,2016,33(2):146-153.doi:10.7641/CTA.2016.50588.
[5]QIU Hao,HUANG Gaoming,and GAO Jun.Variational bayesian labeled multi-bernoulli filter with unknown sensor noise statistics[J].Chinese Journal of Aeronautics,2016,29(5):1378-1384.doi:10.1016/j.cja.2016.05.002.
[6]朱书军,刘伟峰,崔海龙.基于广义标签多伯努利滤波的可分辨群目标跟踪算法[J].自动化学报,2017,43(12):2178-2189.doi:10.16383/j.aas.2017.c160334.ZHU Shujun,LIU Weifeng,and CUI Hailong.Multiple resolvable groups tracking using the GLMB filter[J].Acta Automatica Sinica,2017,43(12):2178-2189.doi:10.16383/j.aas.2017.c160334.
[7]PETETIN Y and DESBOUVRIES F.Bayesian multi-object filtering for pairwise Markovchains[J].IEEE Transactionson Signal Processing,2013,61(18):4481-4490.doi:10 .1109/TSP.2013.2271751.
[8]PIECZYNSKI W.Pairwise Markov chains[J].IEEETransactions on Pattern Analysis and Machine Intelligence,2003,25(5):634-639.doi:10.1109/TPAMI.2003.1195998.
[9]MAHLER R.Tracking targets with pairwise-Markov dynamics[C].International Conference on Information Fusion,Washington,DC,USA,2015:280-286.
[10]MAHLER R.On multitarget pairwise-Markov models[J].Society of Photo-Optical Instrumentation Engineers(SPIE),2015,9474,94740D:1-12.doi:10.1117/12.2177192.
[11]MAHLER R.Multitarget bayes filtering via first-order multitarget moments[J].IEEE Transactions on Aerospace and Electronic Systems,2003,39(4):1152-1178.doi:10.1109/TAES.2003.1261119.
[12]ERDINC O,WILLETT P,and BARSHALOM Y.Aphysical-space approach for the probability hypothesis density and cardinalized probability hypothesis density filters[C].Proceedings of the SPIE Conference on Signal and Data Processing of Small Targets,Orlando,USA,2006,6236,623619:1-12.doi:10.1117/12.673194.
[13]NIE Yongfang and ZHANG Tao.An improved merging algorithm for the gaussian mixture probability hypothesis density filter[C].Chinese Control and Decision Conference,Chongqing,China,2017:5687-5691.doi:10.1109/CCDC.2017.7978180.
[14]DONG Peng,JING Zhongliang,GONG Deren,et al.Maneuvering multi-target tracking based on variable structure multiple model GMCPHD filter[J].Signal Processing,2017,141:158-167.doi:10.1016/j.sigpro.2017.06.004.
[15]YANG Jinlong,LI Peng,YANG Le,et al.An improved ET-GM-PHD filter for multiple closely-spaced extended target tracking[J].International Journal of Control,Automation and Systems,2017,15(1):468-472.doi:10.1007/s12555-015-0193-x.
[2]杨丹,姬红兵,张永权.未知杂波条件下样本校正的势估计概率假设密度滤波算法[J].电子与信息学报,2018,40(4):912-919.doi:10.11999/JEIT170666.YANG Dan,JI Hongbing,and ZHANG Yongquan.Acardinalized probability hypothesis density filter with unknown clutter estimation using corrected sample set[J].Journal of Electronics&Information Technology,2018,40(4):912-919.doi:10.11999/JEIT170666.
[3]袁常顺,王俊,孙进平,等.一种幅度信息辅助多伯努利滤波算法[J].电子与信息学报,2016,38(2):464-471.doi:10.11999/JEIT150683.YUAN Changshun,WANG Jun,SUN Jinping,et al.Amulti-bernoulli filtering algorithm using amplitude information[J].Journal of Electronics&Information Technology,2016,38(2):464-471.doi:10.11999/JEIT150683.
[4]陈辉,韩崇昭.Rao-Blackwellized粒子势均衡多目标多伯努利滤波[J].控制理论与应用,2016,33(2):146-153.doi:10.7641/CTA.2016.50588.CHEN Hui and HAN Chongzhao.Rao-Blackwellized particle cardinality balanced multi-target multi-bernoulli filter[J].Control Theory&Applications,2016,33(2):146-153.doi:10.7641/CTA.2016.50588.
[5]QIU Hao,HUANG Gaoming,and GAO Jun.Variational bayesian labeled multi-bernoulli filter with unknown sensor noise statistics[J].Chinese Journal of Aeronautics,2016,29(5):1378-1384.doi:10.1016/j.cja.2016.05.002.
[6]朱书军,刘伟峰,崔海龙.基于广义标签多伯努利滤波的可分辨群目标跟踪算法[J].自动化学报,2017,43(12):2178-2189.doi:10.16383/j.aas.2017.c160334.ZHU Shujun,LIU Weifeng,and CUI Hailong.Multiple resolvable groups tracking using the GLMB filter[J].Acta Automatica Sinica,2017,43(12):2178-2189.doi:10.16383/j.aas.2017.c160334.
[7]PETETIN Y and DESBOUVRIES F.Bayesian multi-object filtering for pairwise Markovchains[J].IEEE Transactionson Signal Processing,2013,61(18):4481-4490.doi:10 .1109/TSP.2013.2271751.
[8]PIECZYNSKI W.Pairwise Markov chains[J].IEEETransactions on Pattern Analysis and Machine Intelligence,2003,25(5):634-639.doi:10.1109/TPAMI.2003.1195998.
[9]MAHLER R.Tracking targets with pairwise-Markov dynamics[C].International Conference on Information Fusion,Washington,DC,USA,2015:280-286.
[10]MAHLER R.On multitarget pairwise-Markov models[J].Society of Photo-Optical Instrumentation Engineers(SPIE),2015,9474,94740D:1-12.doi:10.1117/12.2177192.
[11]MAHLER R.Multitarget bayes filtering via first-order multitarget moments[J].IEEE Transactions on Aerospace and Electronic Systems,2003,39(4):1152-1178.doi:10.1109/TAES.2003.1261119.
[12]ERDINC O,WILLETT P,and BARSHALOM Y.Aphysical-space approach for the probability hypothesis density and cardinalized probability hypothesis density filters[C].Proceedings of the SPIE Conference on Signal and Data Processing of Small Targets,Orlando,USA,2006,6236,623619:1-12.doi:10.1117/12.673194.
[13]NIE Yongfang and ZHANG Tao.An improved merging algorithm for the gaussian mixture probability hypothesis density filter[C].Chinese Control and Decision Conference,Chongqing,China,2017:5687-5691.doi:10.1109/CCDC.2017.7978180.
[14]DONG Peng,JING Zhongliang,GONG Deren,et al.Maneuvering multi-target tracking based on variable structure multiple model GMCPHD filter[J].Signal Processing,2017,141:158-167.doi:10.1016/j.sigpro.2017.06.004.
[15]YANG Jinlong,LI Peng,YANG Le,et al.An improved ET-GM-PHD filter for multiple closely-spaced extended target tracking[J].International Journal of Control,Automation and Systems,2017,15(1):468-472.doi:10.1007/s12555-015-0193-x.