Utilizing Two-dimensional LADAR Images for Three-dimensional Navigation: Conceptual Exploration

Andrey Soloviev and Maarten Uijt de Haag

Abstract:	This paper explores methodologies that use twodimensional (2D) LAser raDAR (LADAR) data to derive a three-dimensional navigation solution. Parameters of 3D planar surfaces (planes) are first extracted from the 2D LADAR data. The extracted plane parameters are then applied in the derivation of the 3D navigation solution. The main challenge of this problem is to derive parametric expressions for the 3D planar surfaces from the 2D LADAR data. Two methods for the plane parameter estimation are studied in the paper: 1) Intentional rotations of LADAR. The LADAR sensor is rotated intentionally (e.g. using a servo motor based configuration that is described in the paper) and obtains 2D images for three different elevation angles of the LADAR scanning plane. As a result, three intersect lines are obtained for a plane visible to the LADAR. The intersect lines are first extracted from the three LADAR images and the plane parameters are then estimated by fitting a plane through the three intersect lines. 2) Integration of LADAR measurements with inertial data. A 3D displacement vector is derived from both LADAR and inertial data. A vertical plane assumption (i.e. a zero plane tilt angle is assumed) is applied to estimate the displacement when using only LADAR measurements. However, non-zero tilt angles of the planes used for navigation will result in errors in the position change estimate derived from the LADAR images. Position changes derived from inertial data contain inertial drift errors but are free of any tilt angle induced errors. It is therefore suggested to combine inertial and LADAR data for the plane tilt estimation. An extended Kalman filter mechanization is developed to estimate plane tilt angles. This method is viewed as complementary to intentional LADAR rotations but does not require intentional LADAR rotations. It will be shown that non-zero pitch and roll angles of the LADAR scanning plane will generally improve the observability of the extended Kalman filter states. The paper uses simulation results and live LADAR data tests to illustrate the concepts and algorithms developed.
Published in:	Proceedings of the 63rd Annual Meeting of The Institute of Navigation (2007) April 23 - 25, 2007 Royal Sonesta Hotel Cambridge, MA
Pages:	602 - 613
Cite this article:	Soloviev, Andrey, de Haag, Maarten Uijt, "Utilizing Two-dimensional LADAR Images for Three-dimensional Navigation: Conceptual Exploration," Proceedings of the 63rd Annual Meeting of The Institute of Navigation (2007), Cambridge, MA, April 2007, pp. 602-613.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In