ENGLISH / MAGYAR
Kövess
minket

Global localization based on grid mapping with long range sensors

2016-2017/II.
Dr. Fiala Péter

Global localization refers to recognition of a specific, unique scene that can be located anywhere around the world.

Based on the premise, that a scene has been learned by mapping long range sensor data to a grid map,

the same scene shall be detected during free driving by comparison of the stored map with the map that is created online during the driving. The thesis work covers development and evaluation of an appropriate method for the above recognition process, with the following aspects also included:

 

-       Basic mechanism for matching stored and online scenes

-       Prevention of false-positive matches

-       Coarse-to-fine approaches

-       Sensitivity to entry point and orientation

-       Prototype implementation using a lidar sensor


1
1