Conversion of vehicle position data into a driveable set of path segments

This paper describes a method for creating a driveable set of vehicle path segments from data taken while driving a desired path. The developed method reduces the amount of data that must be stored to accurately describe the path, takes into account steering performance constraints, and produces path segments that are held to a user specified average deviation from the original data set. The method uses zero phase filtering to reduce high frequency noise in the data set and then calculates the curvature of the data set at each point. The method then uses a unique combination of information in the unfiltered data set and the calculated curvature measurement of the data in order to obtain the series of driveable path segments.