Localization and Mapping

Planar avoidance localiziation and mapping is an extension of existing simultaneous localization and mapping (SLAM) from 2 dimensions to 3 dimensions. SLAM was developed for 2D robots moving at a relatively slow pace. Mapping a large number of points was not a problem in this regime. However, when applying SLAM to aerospace endeavors speed and precision is critical. It is at this point that an algorithm such as PALM is needed to develop 3D maps of the environment. These maps are used by the vehicles both for navigation in the environment and vehicle localization. This is a first step towards polygon mesh type mapping which is currently limited by real-time computational constraints.

A basic algorithm for converting point clouds obtained from a camera or laser scanner into planes is shown below. Note how the points are grouped according to which plane they belong.

Movie