Some steps towards a unified motion estimation procedure

This paper shows how prior knowledge on the video signal (e.g. spatial autocovariance, distribution of expected motion speed, noise spectrum, ...) can be integrated into a motion estimation procedure. Relations between different classes of motion algorithms (differential, tensor-based, block matching, ...) are discussed and perspectives for a unification and enhancement of such procedures are presented.

[1]  Hans Knutsson,et al.  Texture Analysis Using Two-Dimensional Quadrature Filters , 1983 .

[2]  Edward H. Adelson,et al.  The Design and Use of Steerable Filters , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  David J. Fleet,et al.  Likelihood functions and confidence bounds for total-least-squares problems , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[4]  Hagen Spies,et al.  Motion , 2000, Computer Vision and Applications.

[5]  Hans Knutsson,et al.  Signal processing for computer vision , 1994 .

[6]  Björn Johansson,et al.  A Theoretical Comparison of Different Orientation Tensors , 2002 .

[7]  Rudolf Mester,et al.  Subspace Methods and Equilibration in Computer Vision , 1999 .