Improved Joint Probabilistic Data Association for multiple target tracking in clutter based on Imitation learning

Document Type : Original Article

Authors

1 Master's degree, Malik Ashtar University of Technology, Tehran, Iran

2 Associate Professor, Malek Ashtar University of Technology, Tehran, Iran

Abstract

in multi-target tracking system In order to find the appropriate observation for each track, it is necessary to associate the sensor data. The presence of multiple targets along with the increase of noisy observations increases the order of the calculations and reduces the tracking accuracy. The Joint Probabilistic Data Association Algorithm is an effective algorithm for solving the problem of simultaneously associating multiple observations to multiple targets by generating all possible hypotheses. In general, hypothesis means assigning a maximum of one observation to each track in such a way that none of the observations inside the hypothesis are assigned to track more than once. Despite the optimal performance of the JPDA algorithm, the increase in noisy observations and the intersection of the tracks generates too many assignment hypotheses, decrease in accuracy, and slow down and even stop the tracking process. The algorithm proposed in this article by using the imitation learning algorithm with the real-time training process and removing the hypothesis generation step when there is a possibility of explosion of dimensions in the JPDA algorithm, along maintaining the approximate accuracy of estimating the states of each target, increase the speed of the assignment in the presence of noise and clutter observations. In this algorithm, the Imitation learning model has been mediated in order to extract the required information from the sensor observations and has eliminated the need for hypothesis generation in general. The simulation results show that the proposed algorithm, while maintaining the tracking accuracy in comparison to JPDA, has a lower calculation than the JPDA algorithm. In the situation in which the JPDA faces the problem of computational dimension explosion due to the overlap of target observations or the number of clutters, the proposed approach using the learned model shows a suitable performance and prevent the dimension explosion.

Keywords

Smiley face

References

[1] Blackman, S.S. and R. Popoli, Design and analysis of modern tracking systems. 1999: Artech House Publishers.
 
[2] Y. Bar-Shalom, W.D.B., Multitargte-Multisensor Tracking Application and Advances. Artech House, 2000. III.
 
[3] C. Qu, Y. Zhang, X. Zhang, and Y. Yang, “Reinforcement Learning-Based Data Association for Multiple Target Tracking in Clutter,” Sensors, vol. 20, no. 22, p. 6595, Nov. 2020, doi: 10.3390/s20226595.
 
[4] Turkmen, I. and K. Guney, Cheap Joint Probabilistic Data Association with Adaptive Neuro-Fuzzy Inference System State Filter for Tracking Multiple Targets in Cluttered Environment. AEU - International Journal of Electronics and Communications, 2004. 58(5): p. 349-357.
 
[5] Song, T.L. and D. Lee, A probabilistic nearest neighbor filter algorithm for m validated measurements. Signal Processing, IEEE Transactions on, 2006. 54: p. 2797-2802.
 
[6] Wang, J., et al., Deep Reinforcement Learning for Data Association in Cell Tracking. Frontiers in Bioengineering and Biotechnology, 2020. 8.
       [7] Gruyer, D., et al., Multi-Hypotheses Tracking using the Dempster-Shafer Theory Application to ambiguous road context. Information Fusion, 2015. 29.
 
[8] Chen, S.-l. and Y.-b. Xu, A new joint possibility data association algorithm avoiding track coalescence. International Journal of Intelligent Systems and Applications, 2011. 3(2): p. 45-51.
[9] Habtemariam, B., et al., A multiple-detection joint probabilistic data association filter. IEEE Journal of Selected Topics in Signal Processing, 2013. 7(3): p. 461-471.
 
[10] Saad, E., et al., Filtered Gate Structure Applied to Joint Probabilistic Data Association Algorithm for Multi-Target Tracking in Dense Clutter Environment. International Journal of Computer Science Issues (IJCSI), 2011. 8(2): p. 161.
 
      [11] Sinha, A., et al., Track quality based multitarget tracking approach for global nearest-neighbor association. IEEE Transactions on Aerospace and Electronic Systems, 2012. 48(2): p. 1179-1191.
 
       [12] He, S., H.-S. Shin, and A. Tsourdos, Multi-sensor multi-target tracking using domain knowledge and clustering. IEEE Sensors Journal, 2018. 18(19): p. 8074-8084.
 
        [13] Liang-Qun, L. and X. Wei-Xin, Intuitionistic fuzzy joint probabilistic data association filter and its application to multitarget tracking. Signal Processing, 2014. 96: p. 433-444
 
         [14] Sathyan, T., et al., A multiple hypothesis tracker for multitarget tracking with multiple simultaneous measurements. IEEE Journal of Selected Topics in Signal Processing, 2013. 7(3): p. 448-460.
 
          [15] Satapathi, G.S. and P. Srihari, Soft and evolutionary computation based data association approaches for tracking multiple targets in the presence of ECM. Expert Systems with Applications, 2017. 77: p. 83-104.
 
          [16] Zhang, L., Y. Li, and R. Nevatia. Global data association for multi-object tracking using network flows. in 2008 IEEE conference on computer vision and pattern recognition. 2008. IEEE.
 
          [17] Oh, S., S. Russell, and S. Sastry, Markov chain Monte Carlo data association for multi-target tracking. IEEE Transactions on Automatic Control, 2009. 54(3): p. 481-497.
 
         [18] Karlsson, R. and F. Gustafsson. Monte Carlo data association for multiple target tracking. in IEE Target Tracking: Algorithms and Applications (Ref. No. 2001/174). 2001. IET.
 
         [19] Wu, Z., et al., Coupling detection and data association for multiple object tracking. 2012. 1948-1955.
 
       [20] Sengupta, D. and R.A. Iltis, Neural solution to the multitarget tracking data association problem. IEEE Transactions on Aerospace and Electronic Systems, 1989. 25(1): p. 96-108.
Radar
Volume 10, Issue 2 - Serial Number 28
Number 28, Autumn and Winter Quarterly
January 2023

Files

History

  • Receive Date: 20 October 2022
  • Revise Date: 08 December 2022
  • Accept Date: 01 January 2023
  • Publish Date: 21 January 2023

Share

How to cite

Statistics