Real time 3-D scene analysis by fast features detection

Author

Li, Xiao-Di

Author_Institution

Dept. of Electron. Eng., Chinese Univ. of Hong Kong, Shatin, Hong Kong

Volume

3

fYear

1995

fDate

5-9 Aug 1995

Firstpage

380

Abstract

Although intensity images can be obtained quickly, it is difficult to derive 3-D information. On the other hand, it is expensive or time consuming to get 3-D range data. In this paper, the range data obtained by a 2-D laser range finder is used to complement the intensity images for 3-D scene reconstruction. A sensor fusion method is described for fast computation of the 3-D scene. A polyhedral 3-D static model (PSM) for scene analysis is presented. The recovered features can match with the PSM to determine the location of the robot and reconstruct the scene

Keywords

distance measurement; feature extraction; image reconstruction; mobile robots; optical radar; robot vision; sensor fusion; 2-D laser range finder; 3-D information; 3-D range data; 3-D scene reconstruction; fast features detection; intensity images; polyhedral 3-D static model; real time 3-D scene analysis; sensor fusion method; Cameras; Computer vision; Image analysis; Image reconstruction; Laser fusion; Layout; Mobile robots; Robot vision systems; Sensor fusion; Sensor phenomena and characterization;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Robots and Systems 95. 'Human Robot Interaction and Cooperative Robots', Proceedings. 1995 IEEE/RSJ International Conference on

Conference_Location

Pittsburgh, PA

Print_ISBN

0-8186-7108-4

Type

conf

DOI

10.1109/IROS.1995.525913

Filename

525913