Motion estimation with integrated motion models

Authors:
Mahesh Nalasani;W. David Pan;Seong-Moo Yoo
Affiliations:
Univ. of Alabama in Huntsville, Huntsville, AL;Univ. of Alabama in Huntsville, Huntsville, AL;Univ. of Alabama in Huntsville, Huntsville, AL
Venue:
ACM-SE 42 Proceedings of the 42nd annual Southeast regional conference
Year:
2004

Citing 2
Cited 2

MPEG Video Compression Standard

MPEG Video Compression Standard
Motion estimation with integrated motion models

ACM-SE 42 Proceedings of the 42nd annual Southeast regional conference

Motion estimation with integrated motion models

ACM-SE 42 Proceedings of the 42nd annual Southeast regional conference
Complexity accuracy tradeoffs of Lie operators in motion estimation

Pattern Recognition Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

Conventional motion estimation algorithms rely on motion vectors characterizing translations and thus have limitations in capturing transformations of objects in video scenes such as scaling, rotations and deformations. In this paper, we introduce integrated motion models based on the Lie derivatives to improve the motion estimation accuracy. Our simulations on two standard video sequences show that an average improvement of 0.7dB in PSNR (peak signal-to-noise ratio) can be achieved in motion estimation.