Multi-sensor joint configuration and tracking algorithm—Road constrained mobile sensor systems

This paper proposes an algorithm of mobile multi-sensor configuration and multi-object tracking considering the with the road-constrained condition. The multi-sensor configuration is modeled in the linear programming framework and multi-object states are estimated using linear estimation theory. By minimizing the power and cost (PaC) of all mobile sensors, we use convex optimization method to solve the multi-sensor configuration and further choice three controlling modalities for the mobile multi-sensor, i.e., forward, static or backward. We can approximately get the optimal sensors and their controlling modalities. Finally, we make measurement association through the proposed sequence multi-sensor MHT algorithm for the sensors and then estimate the object states through Kalman filter. We try to jointly deal with the interaction of sensor configuration, modal choice and multi-object tracking. The finally experimental results evaluate the tracking performance and verify the given algorithm.