The joint probabilistic data-association filter (JPDAF)[1] is a statistical approach to the problem of plot association (target-measurement assignment) in a target tracking algorithm. Like the probabilistic data association filter (PDAF), rather than choosing the most likely assignment of measurements to a target (or declaring the target not detected or a measurement to be a false alarm), the PDAF takes an expected value, which is the minimum mean square error (MMSE) estimate for the state of each target. At each time, it maintains its estimate of the target state as the mean and covariance matrix of a multivariate normal distribution. However, unlike the PDAF, which is only meant for tracking a single target in the presence of false alarms and missed detections, the JPDAF can handle multiple target tracking scenarios. A derivation of the JPDAF is given in.[2]
The JPDAF is one of several techniques for radar target tracking and for target tracking in the field of computer vision.
The coalescence problem
editA common problem observed with the JPDAF is that estimates of closely spaced targets tend to coalesce over time.[3][4] This is because the MMSE estimate is typically undesirable when target identity is uncertain.[5]
Variants of the JPDAF algorithm have been made that try to avoid track coalescence. For example, the Set JPDAF[6] uses an approximate minimum mean optimal sub pattern assignment (MMOSPA) instead of an approximate MMSE estimator. The JPDAF*,[7] modifies how the target-measurement association probabilities are computed, and variants of the global nearest-neighbor JPDAF (GNN-JPDAF) (a best-hypothesis tracker)[8] use the global nearest neighbor (GNN) estimate in place of the mean but compute the covariance matrix as in the normal JPDAF: as a mean-squared error matrix.
Implementations
edit- Matlab: The PDAF, JPDAF, Set JPDAF, JPDAF*, GNN-JPDAF and multiple other exact and approximate variants of the JPDAF are implemented in the
singleScanUpdatefunction that is part of the United States Naval Research Laboratory's free Tracker Component Library.[9] The sample code indemo2DDataAssociationdemonstrates how the algorithms can be used in a simple scenario.
References
edit- ^ Bar-Shalom, Yaakov; Daum, Fred; Huang, Jim (December 2009). "The probabilistic data association filter". IEEE Control Systems Magazine. 29 (6): 82–100. doi:10.1109/MCS.2009.934469. S2CID 6875122.
- ^ Bar-Shalom, Yaakov; Li, Xiao-Rong (1995). Multitarget-multisensor tracking : principles and techniques, 1995. Yaakov Bar-Shalom. ISBN 978-0964831209.
- ^ Fitzgerald, Robert (November 1985). "Track Biases and Coalescence with Probabilistic Data Association". IEEE Transactions on Aerospace and Electronic Systems. AES-21 (6): 822–825. Bibcode:1985ITAES..21..822F. doi:10.1109/TAES.1985.310670. S2CID 6544485.
- ^ Bar-Shalom, Yaakov (1986). "Comments on "Track Biases and Coalescence with Probabilistic Data Association"". IEEE Transactions on Aerospace and Electronic Systems. AES-22 (5): 661–662. Bibcode:1986ITAES..22..661.. doi:10.1109/TAES.1986.310734.
- ^ Crouse, David (23 May 2013). Advances in displaying uncertain estimates of multiple targets. Proceeding of SPIE 8745, Signal Processing, Sensor Fusion, and Target Recognition XXII. Baltimore.
- ^ Svensson, Lennart; Svensson, Daniel; Willett, Peter (July 2009). Set JPDA algorithm for tracking unordered sets of targets. Proceedings of the 12th International Conference of Information Fusion. Seattle. pp. 1187–1194.
- ^ Blom, H.A.P.; Bloem, E.A. (2000). "Probabilistic data association avoiding track coalescence". IEEE Transactions on Automatic Control. 45 (2): 247–259. doi:10.1109/9.839947.
- ^ Drummond, Oliver (October 1999). Best hypothesis target tracking and sensor fusion. Proceedings of SPIE: Signal and Data Processing of Small Targets Conference. Denver. pp. 586–600.
- ^ "Tracker Component Library". Matlab Repository. Retrieved January 5, 2019.
- ^ "Stone Soup Github Repo". GitHub.
- ^ "Stone Soup JPDA Tutorial Docs".
- ^ "Stone Soup JPDA Tutorial Code". GitHub.