Share Email Print
cover

Proceedings Paper

A 3D terrain reconstruction method of stereo vision based quadruped robot navigation system
Author(s): Zhuo Ge; Ying Zhu; Guanhao Liang
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

To provide 3D environment information for the quadruped robot autonomous navigation system during walking through rough terrain, based on the stereo vision, a novel 3D terrain reconstruction method is presented. In order to solve the problem that images collected by stereo sensors have large regions with similar grayscale and the problem that image matching is poor at real-time performance, watershed algorithm and fuzzy c-means clustering algorithm are combined for contour extraction. Aiming at the problem of error matching, duel constraint with region matching and pixel matching is established for matching optimization. Using the stereo matching edge pixel pairs, the 3D coordinate algorithm is estimated according to the binocular stereo vision imaging model. Experimental results show that the proposed method can yield high stereo matching ratio and reconstruct 3D scene quickly and efficiently.

Paper Details

Date Published: 23 January 2017
PDF: 7 pages
Proc. SPIE 10322, Seventh International Conference on Electronics and Information Engineering, 1032242 (23 January 2017); doi: 10.1117/12.2265369
Show Author Affiliations
Zhuo Ge, Beijing Institute of Technology (China)
Ying Zhu, Beijing Institute of Technology (China)
Guanhao Liang, Beijing Institute of Technology (China)


Published in SPIE Proceedings Vol. 10322:
Seventh International Conference on Electronics and Information Engineering
Xiyuan Chen, Editor(s)

© SPIE. Terms of Use
Back to Top