Central Difference Information Filter with Interacting Multiple Model for Robust Maneuvering Object Tracking

In this paper, we introduce a new framework to combine the central difference information filter (CDIF) with the interacting multiple model (IMM) method for maneuvering object tracking. The CDIF has been recently introduced for solving object tracking problem using multiple sensors. The CDIF uses Stirling's interpolation to generate a number of sigma points for approximating the distribution of Gaussian random variables and does not require the calculation of Jacobians. However, the general CDIF method has difficulties to handle maneuvering objects, due to the changing of system model. In the literature, the IMM method is a natural way to estimate the discontinuities of object motion, by running a bank of filters in parallel with multiple models. Here, our contribution is to use the CDIF in the IMM framework (IMM-CDIF), which has better capabilities to handle maneuvering object tracking problem. In the end, a bearing only tracking experiment is demonstrated, and shows that the new IMM-CDIF method has lower mean square error (MSE) comparing with the original CDIF method.