Estimation three-dimensional motion of rigid objects from noisy observations

Author

Philip, Johan

Author_Institution

Dept. of Math., R. Inst. of Technol., Stockholm, Sweden

Volume

13

Issue

1

fYear

1991

fDate

1/1/1991 12:00:00 AM

Firstpage

61

Lastpage

66

Abstract

An estimate is made of the motion of a rigid body from two noisy 2-D perspective projections using the least-squares method and the algebra of R.Y. Tsai and T.S. Huang (1984). The accuracy of the estimated motion parameters is influenced by the position of the features of the object used in the calculation. Four test variables are derived that indicate how the accuracy is affected, and they are used for discarding inaccurate estimates. Monte Carlo tests demonstrate the obtained accuracy

Keywords

Monte Carlo methods; computer vision; computerised pattern recognition; computerised picture processing; least squares approximations; Monte Carlo tests; accuracy; estimated motion parameters; inaccurate estimates; least-squares method; noisy 2-D perspective projections; noisy observations; rigid objects; three-dimensional motion; Algebra; Computer vision; Least squares approximation; Least squares methods; Monte Carlo methods; Motion estimation; Nonlinear equations; Nonlinear systems; Parameter estimation; Testing;

fLanguage

English

Journal_Title

Pattern Analysis and Machine Intelligence, IEEE Transactions on

Publisher

ieee

ISSN

0162-8828

Type

jour

DOI

10.1109/34.67631

Filename

67631