DocumentCode
1691003
Title
Errors-in-variables modelling in optical flow problems
Author
Ng, Lydia ; Solo, Victor
Author_Institution
Dept. of Electron., Macquarie Univ., Sydney, NSW, Australia
Volume
5
fYear
1998
Firstpage
2773
Abstract
Although still in practice, the use of total least squares (TLS) in optical flow estimation is unreliable. The TLS implicitly assumes that the error terms affecting the partial derivatives of the image intensities are independent. The usual methods for estimating the partial derivatives ensures that the errors are strongly correlated. Due to this correlation, an alternative method is required to treat the resulting errors-in-variables (EN) problem. We propose a new method for estimating optical flow based on Sprent´s (1966, 1969) procedure. This method incorporates a general EIV model and provides a far simpler computational procedure than found in previous solutions
Keywords
error statistics; image sequences; least squares approximations; maximum likelihood estimation; stochastic processes; Nagel´s maximum likelihood method; Sprent´s procedure; TLS; computational procedure; correlation; errors-in-variables modelling; gray scale image sequence; image intensities; optical flow estimation; optical flow problems; partial derivatives; stochastic model; total least squares; Brightness; Equations; Error analysis; Image motion analysis; Least squares approximation; Linear systems; Optical devices; Optical noise; Pixel; Uncertainty;
fLanguage
English
Publisher
ieee
Conference_Titel
Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on
Conference_Location
Seattle, WA
ISSN
1520-6149
Print_ISBN
0-7803-4428-6
Type
conf
DOI
10.1109/ICASSP.1998.678098
Filename
678098
Link To Document