DocumentCode :
2239088
Title :
Multi-scale structure from multi-views by Δ2G filtered 3D voting
Author :
Mori, Takeaki ; Suzuki, Satoshi ; Horikoshi, Tsutomu ; Yasuno, Takayuki
Author_Institution :
NTT Corp., Yokosuka, Kanagawa, Japan
fYear :
1993
fDate :
15-17 Jun 1993
Firstpage :
662
Lastpage :
663
Abstract :
A multiscale 3-D structure recovery method is proposed. It divides the scene space into voxels and votes for them using the sight lines of the camera from coarse to fine. The mechanism behind the influence of the image quantization error and the position error of the camera´s optical center is analyzed. Both errors can be diminished by Δ2G filtering the voted voxel space. A multiscale 3-D structure of objects is successfully generated from a real image sequence of six frames using this method
Keywords :
error correction; filtering and prediction theory; image recognition; measurement errors; roundoff errors; Δ2G filtered 3D voting; camera sight lines; image quantization error; multiple views; multiscale 3-D structure recovery; multiviews; optical centre position error; voxels; Cameras; Filtering; Humans; Image analysis; Image sequences; Laboratories; Layout; Optical filters; Optical noise; Optical sensors; Quantization; Telegraphy; Telephony; Voting;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Vision and Pattern Recognition, 1993. Proceedings CVPR '93., 1993 IEEE Computer Society Conference on
Conference_Location :
New York, NY
ISSN :
1063-6919
Print_ISBN :
0-8186-3880-X
Type :
conf
DOI :
10.1109/CVPR.1993.341039
Filename :
341039
Link To Document :
بازگشت