Attitude Correction Using Corner Landmarks in 3-D SLAM

This paper discusses the use of three-dimensional (3-D) corner features extracted from a 2-D laser scanner as landmarks in a simultaneous localization and mapping solution. The corner landmarks can be identified and tracked using their angles, and the observations of the corners are directly used to correct attitude errors. The proposed method can be used to navigate a small unmanned aerial vehicle in urban and indoor environments.