1: \begin{abstract}

2: This paper presents a robust  6-DOF relative navigation by

3: combining the iterative closet point  (ICP)  registration

4: algorithm and a noise-adaptive Kalman filter (AKF) in a

5: closed-loop configuration together with  measurements from a

6: laser scanner and an inertial measurement unit (IMU). In this

7: approach, the fine-alignment phase of the registration is

8: integrated with the filter innovation step  for estimation

9: correction while the filter estimate propagation provides the

10: coarse alignment needed to find the corresponding points at the

11: beginning of ICP iteration cycle.  The convergence of the ICP

12: point matching is monitored by a fault-detection logic and the

13: covariance associated with  the ICP alignment error is estimated

14: by a recursive algorithm.  This ICP enhancement has proven to

15: improve  robustness and accuracy of the pose tracking performance

16: and to automatically recover correct alignment whenever the

17: tracking is lost. The Kalman filter estimator is designed so as to identify the

18: required parameters such as IMU biases and location of the

19: spacecraft center-of-mass (CoM).  

20: \end{abstract}

21: