Free clustering optimal particle probability hypothesis density (PHD) filter

Abstract

As to the fact that it is difficult to obtain analytical form of optimal sampling density and tracking performance of standard particle probability hypothesis density (P-PHD) filter would decline when clustering algorithm is used to extract target states, a free clustering optimal P-PHD (FCO-P-PHD) filter is proposed. This method can lead to obtainment of analytical form of optimal sampling density of P-PHD filter and realization of optimal P-PHD filter without use of clustering algorithms in extraction target states. Besides, as sate extraction method in FCO-P-PHD filter is coupled with the process of obtaining analytical form for optimal sampling density, through decoupling process, a new single-sensor free clustering state extraction method is proposed. By combining this method with standard P-PHD filter, FC-P-PHD filter can be obtained, which significantly improves the tracking performance of P-PHD filter. In the end, the effectiveness of proposed algorithms and their advantages over other algorithms are validated through several simulation experiments.

Keywords

multiple target tracking / probability hypothesis density filter / optimal sampling density / particle filter / random finite set / clustering algorithm / state extraction

References

[1] VoB N, SinghS, DoucetA. Sequential Monte Carlo methods for multi-target filtering with random finite sets [J]. IEEE Transactions on Aerospace Electronic Systems, 2005, 41(4): 1224-1245

[2] SimandM, StrakaO. Sampling densities of particle filter: A survey and comparison [C]. Proc American Control Conference, 2007, New York, USA, IEEE: 4437-4442

[3] HolJ D, SchonT B, GustafssonF. On resampling algorithms for particle filters [C]. Proc IEEE Nonlinear Statistical Signal Processing Workshop, 2006, Cambridge, UK, IEEE: 79-82

[4] TangX, WeiPing. Improved peak extraction algorithm in SMC implementation of PHD filter [C]. Proc IEEE International Symposium on Intelligent Signal Processing and Communication Systems, 20101-4

[5] HigoT, TakadamaK. Hierarchical importance sampling instead of annealing [C]. Proc IEEE Congress Evolutionary Computation, 2007, Singapore, IEEE: 134-141

[6] BrasnettP, MihaylovaL, BullD, CanagarajahN. Improved proposal distribution with gradient measures for tracking [C]. Proc IEEE Workshop Machine Learning for Signal Processing, 2005, Mystic, CT, IEEE: 105-110

[7] StrakaO, SimandlM, DunikJ. Gaussian mixtures proposal density in particle filter for track-before-detect [C]. Proc IEEE International Conference on Information Fusion, 2009, Seattle, USA, IEEE: 270-277

[8] HuangY, DjuricP M. A hybrid importance function for particle filtering [J]. IEEE Signal Processing Letters, 2004, 11(3): 404-406

[9] MaccormickJ, IsardM. Partitioned sampling, articulated objects, and interface-quality hand tracking [C]. Proc 6th European Conference on Computer Vision, 2000, Dublin, Ireland, DBLP: 3-19

[10] WhiteleyN, SinghS, GodsillS. Auxiliary particle implementation of probability hypothesis density filter [J]. IEEE Transactions on Aerospace Electronic Systems, 2010, 46(3): 1437-454

[11] ArulampalamM S, MaskellS, GordonN, ClappT. A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking [J]. IEEE Transactions on Signal Processing, 2002, 50(2): 174-188

[12] ClarkD E, BellJ. Multi-target state estimation and track continuity for the particle PHD filter [J]. IEEE Transactions on Aerospace Electronic Systems, 2007, 43(4): 1441-1453

[13] LiuW-f, HanC-z, LianF, ZhuH-yan. Multitarget state extraction for the phd filter using MCMC approach [J]. IEEE Transactions on Aerospace Electronic Systems, 2010, 46(2): 883-864

[14] TangX, WeiPing. Multi-target state extraction for the particle probability hypothesis density filter [J]. IET Radar Sonar & Navigation, 2011, 5(9): 877-883

[15] LiuS, ChenR. Sequential Monte Carlo methods for dynamic systems [J]. Journal of American Statistical Association, 1995, 90(2): 567-576

[16] MorelandeM. A Sequential Monte Carlo method for PHD approximation with conditionally linear/Gaussian models [C]. Proc IEEE International Conference on Information Fusion, 2010, Edinburgh, UK, IEEE: 1-7

[17] CappeO, GodsillS J, MoulinesE. An overview of existing methods and recent advances in sequential monte carlo [J]. IEEE Journal & Magazine, 2007, 95(5): 899-924

[18] MahlerRStatistical multisource multitarget information fusion [M], 2007, London, Artech House: 311-313

[19] MahlerR. A theoretical foundation for the Stein-Winter probability hypothesis density (PHD) multi-target tracking approach [C]. Proc 2002 MSS National Symposium on Sensor and Data Fusion, 2000, San Antonio, USA, Sandia National Laboratories: 99-117

[20] MahlerR. Multi-target bayes filtering via first order multi-target moments [J]. IEEE Transactions on Aerospace Electronic Systems, 2003, 39(4): 1152-1178

[21] DoucetA, GodsillS, AndrieuC. On sequential Monte Carlo sampling methods for bayesian filtering [J]. Statistics and Computing, 2000, 10(3): 197-208

[22] WenQ, GaoJ, KosakaA. A particle filter framework using optimal importance function for protein molecules tracking [C]. Proc IEEE International Conference on Image Processing, 2005, Genova, Italian, IEEE: I-1161-4

[23] SchuhmacherD, VoB T, VoB N. A consistent metric for performance evaluation of multi-object filters [J]. IEEE Transactions on Signal Processing, 2008, 56(8): 3447-3457