Multiple target tracking using thermal imaging and radar sensors
Ioannis Kyriakides · 2016
Combining measurements from different types of sensors has the potential to improve tracking performance. This is attributed to the diversity of information provided by the sensors on the state of the targets. Moreover, compressively acquiring measurements has the potential to enable the collection of measurements from high resolution sensors at low sampling rates. This can further improve tracking performance while simplifying hardware design. However, the fusion of measurements of diverse origin for the purpose of sequential target state estimation is challenging due to the varying resolution and relationship of measurements with the target states. In this work the problem of combining measurements from two sensors having different type and resolution capabilities for the purpose of multiple target tracking is tackled using a sequential Monte Carlo method. The proposed algorithm is able to work with measurements that have a non-linear relationship with the target state and that are acquired either compressively or at the Nyquist rate. A scenario involving multiple targets observed by a thermal imager and a radar sensor is used to assess the tracking performance of the proposed method.