Abstract
Joint detection and tracking weak target is a challenging problem whose complexity is intensified when there are multiple targets present at the same time. Some Probability Hypothesis Density (PHD) based track-before-detect (TBD) particle filters (PHD-TBD) are proposed to solve this issue; however, the performance is unsatisfactory especially when the number of targets is large because some assumptions in PHD are violated. We propose to modify the general PHD-TBD filter in two aspects to make the PHD processing available for TBD scenarios. First, the distribution of false alarms is approximated as the Poisson distribution through a threshold method, and then a clustering technique is proposed to solve the overestimation of the target number. A typical TBD scenario is used to test the effectiveness of the proposed method. Simulation results indicate that the proposed method outperforms the general method in terms of estimation accuracy and computational complexity.
Highlights
Joint detection and tracking a low signal-to-noise ratio (SNR) target, referred to as weak, dim or stealthy target, is a thorny problem based on the traditional threshold detection method
THE GENERAL Probability Hypothesis Density (PHD) FILTER FOR TBD In Section 2, we have modelled the multitarget TBD observations and collection of states as random finite set (RFS), and a PHD and TBD approach (PHD-TBD) approach can be formulated
A threshold method is used to adapt the distribution of false alarms to Poisson distribution, and a clustering technique is used to solve the problem of overestimation of the target number
Summary
Joint detection and tracking a low signal-to-noise ratio (SNR) target, referred to as weak, dim or stealthy target, is a thorny problem based on the traditional threshold detection method. The random finite set (RFS) theory [14], [15] has drawn wide attention and has been applied to many fields [16], [17] It provides a systematic and rigorous procedure to solve the multitarget tracking problem. The PHD filter operates on the single-target state space and avoids the combinatorial problems that arise from data association. The proposed method is shown to be a computationally efficient solution to the multitarget tracking problems with the varying number of targets, it assumes the false alarms to be a constant number, which is violated with the second assumption above, and the results need to be checked in more detail. The surviving target motion is denoted as a RFS Sk|k−1 (xk−1), the RFS of target birth at time k is denoted as k , and the RFS of targets spawning from the target with xk−1 is represented by Bk|k−1 (xk−1), the multitarget state Xk is modelled by the union of RFSs as
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
