Road-Map-Assisted ground moving target tracking using gausian misture PHD filter
Yang Bin, Jun Wang, Yuan Chang-shun · 2016
In ground tracking applications, the majority of targets are vehicles, which are moving on roads. Since road maps can be achieved easily, the track performance and track continuity can be enhanced by exploiting road map information. In this paper, we adopt a one-dimensional road coordinates motion model and use linear road segments to approximate the curved ground coordinates road. Incorporating this modeled road map information, a probability hypothesis density (PHD) multiple target filter is applied to estimate these road constrained targets. We use Gaussian Mixture (GM) variant to implement the closed-form PHD filter in road coordinates linear Gaussian systems. The simulation results show that the approach of using GM-PHD filter with road map information can yield an effective ground moving target tracker.