### Skapa referens, olika format (klipp och klistra)

**Harvard**

Bergroth, F. (2011) *Planar ray shooting for robot localization using laser range finders*. Göteborg : Chalmers University of Technology

** BibTeX **

@mastersthesis{

Bergroth2011,

author={Bergroth, Fredrik},

title={Planar ray shooting for robot localization using laser range finders},

abstract={The purpose of this study has been improve a localization technique used in a robot developed at the Adaptive systems group at Chalmers University of Technology. Currently, the robot uses a simple algorithm to compute laser ray shooting queries among line segments representing a map, rather than preprocessing the set of line segments into a data structure that answers queries more quickly. Two different algorithms are presented that answer such queries, as well as the data structures upon which they are built. Persistent search trees and some core concepts of computational geometry are used in the first algorithm, while the second algorithm stores its information in a binary space partitioning tree. The
latter technique was implemented and shown to outperform the simple algorithm for the particular case when the maps consisted of axis{parallel line segments. In maps when the
number of segments was on the order of 200, the algorithm was shown to be twice as fast, and nearly three times as fast when the number of segments reached 500. Finally, some
suggestions are presented regarding improvements of the algorithm.},

publisher={Institutionen för teknisk fysik, Chalmers tekniska högskola},

place={Göteborg},

year={2011},

note={32},

}

** RefWorks **

RT Generic

SR Print

ID 176456

A1 Bergroth, Fredrik

T1 Planar ray shooting for robot localization using laser range finders

YR 2011

AB The purpose of this study has been improve a localization technique used in a robot developed at the Adaptive systems group at Chalmers University of Technology. Currently, the robot uses a simple algorithm to compute laser ray shooting queries among line segments representing a map, rather than preprocessing the set of line segments into a data structure that answers queries more quickly. Two different algorithms are presented that answer such queries, as well as the data structures upon which they are built. Persistent search trees and some core concepts of computational geometry are used in the first algorithm, while the second algorithm stores its information in a binary space partitioning tree. The
latter technique was implemented and shown to outperform the simple algorithm for the particular case when the maps consisted of axis{parallel line segments. In maps when the
number of segments was on the order of 200, the algorithm was shown to be twice as fast, and nearly three times as fast when the number of segments reached 500. Finally, some
suggestions are presented regarding improvements of the algorithm.

PB Institutionen för teknisk fysik, Chalmers tekniska högskola,

LA eng

OL 30