Presentation 1999/7/16
Multiple Planar Surface and Motion Parameter Estimation Based on Correspondence of Feature Points
Yutaka HORIE, Akihiro MINAGAWA, Norio TAGAWA, Tadashi MORIYA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In the conventional method to estimate the planar surface parameters and the motion parameters with the assumption of feature points correspondence between input images, these parameters can be only estimated in case of optimized one motion and one plane parameters estimation problem. Therefore, in this paper, we proposed the estimation method for multiple motions and multiple planes as expansion to the conventional. However, this expansion causes the clustering problem that feature point is owned by incorrect plane, and the reliability of each plane is reduced. We assume that the observed noise model for the feature points is a two-dimensional Gaussian, and define the likelihood function with respect to this observed noise. Then, by using Gaussian mixture model, both the parameter estimation problem and the clustering problem can be formulated. To solve these problems, we apply the iterative computation method based on EM algorithm. The effectiveness of the proposed method can be confirmed by the computational simulation results.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Planar Surface Parameter Estimation / Motion Parameter Estimation / ML estimation / Gaussian Mixture Density Model / EM algorithm
Paper # PRMU99-61
Date of Issue

Conference Information
Committee PRMU
Conference Date 1999/7/16(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Multiple Planar Surface and Motion Parameter Estimation Based on Correspondence of Feature Points
Sub Title (in English)
Keyword(1) Planar Surface Parameter Estimation
Keyword(2) Motion Parameter Estimation
Keyword(3) ML estimation
Keyword(4) Gaussian Mixture Density Model
Keyword(5) EM algorithm
1st Author's Name Yutaka HORIE
1st Author's Affiliation Graduate School of Engineering, Tokyo Metropolitan University()
2nd Author's Name Akihiro MINAGAWA
2nd Author's Affiliation Graduate School of Engineering, Tokyo Metropolitan University
3rd Author's Name Norio TAGAWA
3rd Author's Affiliation Graduate School of Engineering, Tokyo Metropolitan University
4th Author's Name Tadashi MORIYA
4th Author's Affiliation Graduate School of Engineering, Tokyo Metropolitan University
Date 1999/7/16
Paper # PRMU99-61
Volume (vol) vol.99
Number (no) 182
Page pp.pp.-
#Pages 7
Date of Issue