Total least squares 3-D motion estimation

Author

Diamantaras, K.I. ; Papadimitriou, Th. ; Strintzis, M.G. ; Roumeliotis, M.

Author_Institution

Dept. of Inf., Technol. Educ. Inst., Thessaloniki, Greece

Volume

1

fYear

1998

fDate

4-7 Oct 1998

Firstpage

923

Abstract

A new method for estimating 3D motion parameters from point correspondences is presented in this paper. The problem formulation leads to the solution of an overdetermined linear system of equations. The total least squares (TLS) method is found to be the most suitable one for estimating the solution since our model includes noise both in the observation data and in the system matrix. The translation parameters are obtained immediately from the above solution whereas the rotation parameters are estimated from the solution of another TLS problem. Tests of our method on artificial data and on real images show its robustness against Gaussian additive noise and against digitalization noise introduced by finite pixel resolution

Keywords

Gaussian noise; image sequences; least squares approximations; motion estimation; parameter estimation; singular value decomposition; 3D motion parameters estimation; Gaussian additive noise; SVD; TLS problem; artificial data; digitalization noise; finite pixel resolution; motion estimation; observation data; optical flow; overdetermined linear system of equations; point correspondences; real images; rotation parameters; system matrix; total least squares method; translation parameters; Additive noise; Equations; Gaussian noise; Least squares approximation; Linear systems; Motion estimation; Noise robustness; Parameter estimation; Pixel; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing, 1998. ICIP 98. Proceedings. 1998 International Conference on

Conference_Location

Chicago, IL

Print_ISBN

0-8186-8821-1

Type

conf

DOI

10.1109/ICIP.1998.723670

Filename

723670