A Kalman filter based registration approach for asynchronous sensors in multiple sensor fusion applications

A Kalman filter based registration approach is proposed for multiple asynchronous sensors. In the approach, a linear time-varying measurement model is formulated using a first order approximation and is shown to be uniformly completely observable. The sensor registration errors are estimated based on the application of a modified two-stage Kalman estimator. The proposed registration approach is computationally efficient and is capable of handling asynchronous sensor measurements. Simulation and real-life data are used to demonstrate the effectiveness of the proposed approach. Results are compared with the popular least squares (LS) method.