Lines Extraction from Laser Scans Integrating Hough Transform, Total Least Squares and Successive Edges Following
Julie Stephany Berrío, Samith Aron Orozco Ordoñes, Eduardo Francisco Caicedo Bravo · 2012
This paper presents a lines segmentation method, based mainly on a robust line extraction algorithm called Hough transform to obtain the preliminary parameters of straight lines composed of a set of points provided by a 2D laser scanner such as those used in robot navigation systems. The lines parameters obtained by the Hough transform algorithm are fitted by the Total Least Squares (improving the precision), and the points associated with these parameters are again processed by a Successive Edge Following algorithm in order to obtain the Cartesian coordinates of the extreme points of the lines which describe the environment.