Edge-oriented interpolation for fractional motion estimation in hybrid video coding

Fractional motion estimation, using the interpolation process, improves the quality of the compressed video by about 2dB in terms of PSNR over integer motion estimation. Most existing interpolation techniques in recent video coding standards, such as the symmetric 6-tap filter of the H.264/AVC stand...

Full description

Saved in:
Bibliographic Details
Main Authors: Kokhazadeh, Ali, Aminlou, Alireza, Hashemi, Mahmoud Reza
Format: Conference Proceeding
Language:English
Subjects:
Equations
Fractional motion estimation
H.264/AVC
Hybrid video coding
Image edge detection
Interpolation
Mathematical model
Motion estimation
PSNR
Video object edge
Video sequences
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Fractional motion estimation, using the interpolation process, improves the quality of the compressed video by about 2dB in terms of PSNR over integer motion estimation. Most existing interpolation techniques in recent video coding standards, such as the symmetric 6-tap filter of the H.264/AVC standard, do not perform well around the object edges. This increases the value of residuals, which in turn results in higher bit rate. In this paper, a new interpolation method has been proposed that considers the edges of video objects. Simulation results indicate that the proposed method, when used in the H.264/AVC encoder, improves PSNR by up to 1.0 dB (0.4 dB in average) with respect to the standard interpolation technique. This is achieved at the expense of up to 6% (3% in average) increase in the computational complexity.
ISSN:1945-7871
1945-788X
DOI:10.1109/ICME.2011.6011999