Simulating range cameras for complex terrain robot mobility

Author

Sheh, Raymond ; Sammut, Claude

Author_Institution

ARC Centre of Excellence for Autonomous Syst., Univ. of New South Wales, Sydney, NSW, Australia

fYear

2011

fDate

1-5 Nov. 2011

Firstpage

184

Lastpage

189

Abstract

We have developed three simple simulations for range cameras (also called 3D cameras or flash LIDARs) that may be integrated into existing 3D robot visualisation and simulation frameworks. They are able to generate data at better than video framerates with minimal additional processing by leveraging the OpenGL graphics engine. We have used these simulations in our work on applying machine learning to the task of autonomous complex terrain robot mobility and 3D SLAM. We found that with a simple noise model it is possible to obtain realistic results.

Keywords

SLAM (robots); cameras; control engineering computing; data visualisation; digital simulation; learning (artificial intelligence); mobile robots; optical radar; robot vision; 3D SLAM; 3D cameras; 3D robot simulation; 3D robot visualisation; OpenGL graphics engine; complex terrain robot mobility; flash LIDAR; machine learning; range cameras; Cameras; Engines; Noise; Robot vision systems; Three dimensional displays; 3D Vision; Complex Terrain Mobility; Computer Graphics; Range Camera; Robot Simulation; SLAM;

fLanguage

English

Publisher

ieee

Conference_Titel

Safety, Security, and Rescue Robotics (SSRR), 2011 IEEE International Symposium on

Conference_Location

Kyoto

Print_ISBN

978-1-61284-770-2

Type

conf

DOI

10.1109/SSRR.2011.6106763

Filename

6106763