Limited field-of-view multimodal sensor adaptation for data association

We have investigated the utility of field-of-view adaptation for multimodal sensing in cluttered multi-target environments. Measurement data from multiple integrated sensors are collected at a fusion center, which employs a soft association procedure to integrate them into the estimation procedure. A variance penalty model for the limited fields-of-view property is incorporated into the state estimation procedure. This model also forms the basis of an optimization problem that determines the best next-step sensing parameters for the changing target environment. Numerical simulations demonstrate the benefit of the proposed method for both tracking and association metrics compared to a non-adaptive tracker.