Entity

Time filter

Source Type

Laval, Canada

Chen X.,McMaster University | Tharmarasa R.,McMaster University | Pelletier M.,Radar FLIR | Kirubarajan T.,McMaster University
IEEE Transactions on Aerospace and Electronic Systems | Year: 2012

In this paper, based on Poisson point processes, two new methods for joint nonhomogeneous clutter background estimation and multitarget tracking are presented. In many scenarios, after the signal detection process, measurement points provided by the sensor (e.g., sonar, infrared sensor, radar) are not distributed uniformly in the surveillance region as assumed by most tracking algorithms. On the other hand, in order to obtain accurate results, the target tracking filter requires information about clutter's spatial intensity. Thus, nonhomogeneous clutter spatial intensity has to be estimated from the measurement set and the tracking filter's output. Also, in order to take advantage of existing tracking algorithms, it is desirable for the clutter estimation method to be integrated into the tracker itself. Nonhomogeneous Poisson point processes, whose intensity function are assumed to be a mixture of Gaussian functions, are used to model clutter points here. Based on this model, a recursive maximum likelihood (ML) method and an approximated Bayesian method are proposed to estimate the nonhomogeneous clutter spatial intensity. Both clutter estimation methods are integrated into the probability hypothesis density (PHD) filter, which itself also uses the Poisson point process assumption. The mean and the covariance of each Gaussian function are estimated and used to calculate the clutter density in the update equation of the PHD filter. Simulation results show that both methods are able to improve the performance of the PHD filter in the presence of slowly time-varying nonhomogeneous clutter background. © 1965-2011 IEEE.


Chen X.,McMaster University | Tharmarasa R.,McMaster University | Pelletier M.,Radar FLIR | Kirubarajan T.,McMaster University
IEEE Transactions on Aerospace and Electronic Systems | Year: 2013

Based on Poisson point processes, multitarget multi-Bernoulli processes, and set calculus, a Bayesian method is presented to estimate the nonhomogeneous clutter background while simultaneously tracking multiple targets. A major feature of the proposed approach is the seamless integration of clutter estimation with standard multitarget tracking algorithms like the multiple hypothesis tracker (MHT) and the joint integrated probabilistic data association (JIPDA) tracker by exploiting the association events and their probabilities constructed and calculated in standard multitarget tracking algorithms. © 1965-2011 IEEE.


He X.,McMaster University | Tharmarasa R.,McMaster University | Kirubarajan T.,Radar FLIR | Pelletier M.,Radar FLIR
IEEE Transactions on Aerospace and Electronic Systems | Year: 2013

In most multiple hypothesis tracking (MHT) implementations, the data association is solved using Murty's algorithm. However, since Murty's algorithm has no control over the diversity of measurement-to-track associations, often, the top associations vary only slightly. To overcome this problem and to provide more flexibility in the selection of hypotheses, a modified Murty's algorithm, which can achieve any user-defined (or adaptable) diversity of association of different types of tracks, is proposed. © 1965-2011 IEEE.

Discover hidden collaborations