Cooperative multiple dynamic object tracking on moving vehicles based on Sequential Monte Carlo Probability Hypothesis Density filter

Abstract

Abstract-This paper proposes a generalized method for tracking of multiple objects from moving, cooperative vehicles -bringing together an Unscented Kalman Filter for vehicle localization and extending a Sequential Monte Carlo Probability Hypothesis Density filter with a novel cooperative fusion algorithm for tracking. The latter ensures that the fusion of information from cooperating vehicles is not limited to a fully overlapping Field Of View (FOV), as usually assumed in popular distributed fusion literature, but also allows for a perceptual extension corresponding to the union of the vehicles' FOV. Our method hence allows for an overall extended perception range for all cooperative vehicles involved, while preserving same or improving the accuracy in the overlapping FOV. This method also successfully mitigates noisy sensor measurement and clutter, as well as localization inaccuracies of tracking vehicles using Global Navigation Satellite Systems (GNSS). Finally, we extensively evaluate our method using a high-fidelity simulator for vehicles of varying speed and trajectories.

Recent advances in multisensor multitarget tracking using random finite set

“…One solution is to extend the prior FoVs of all sensors to an identical and larger FoV before fusion (Battistelli et al, 2017;Li SQ et al, 2018), but this may underestimate the existence probabilities of targets in non-public areas. Another widely used method is to split the information according to the FoVs' intersection (Gan et al, 2016;Vasic et al, 2016;Li TC et al, 2019c;Da et al, 2020b;Li GC et al, 2020). Fusion will be executed in each splitting area separately, followed by a reunion operation for all areas.…”

Distributed multi-sensor multi-view fusion based on generalized covariance intersection

“…. , Q, the set X p contains the tracks represented by the matched DIFs v1 p (x) and v2 p (x) in the sense that vl p (x) ≈ 0 (31) for l = 1, 2 and for any x ∈ X \ X p ;…”

“…However, both methods are prone to estimation errors and can underestimate the target number when nearby targets appear. A distributed fusion algorithm based on the SMC-PHD filter, abandoning the limitation of fully overlapping FoV, is proposed in [31], which classifies the received particles into common particle set and external particle set. As for the GM-PHD filter, [32] proposes a solution to handle different FoVs in the context of simultaneous localization and mapping (SLAM).…”

Statistical Information Fusion for Multiple-View Sensor Data in Multi-Object Tracking

“…In this scenario, the multi-sensor fusion solution should be capable of combining complementary information provided by various sensors. Examples of complementary fusion solutions include the usage of linear-like complementary filters for attitude estimation [24], using an extended Kalman filter for complementary fusion of multi-sensor data in mobile robotics [25], and using an unscented Kalman filter for complementary fusion of multiple Poisson densities from local PHD filters in robotic applications [26]. This paper presents a novel strategy for combining random set posteriors from sensor nodes with different and limited fields of view.…”