Motion Estimation for H.264/AVC using Programmable Graphics Hardware

We present an efficient implementation of motion estimation (ME) for H.264/AVC using programmable graphics hardware. The cost function for ME in H.264/AVC depends on the motion vector (MV) predictor which is the median MV of three neighboring coded blocks. Previous implementations assume no dependen...

Full description

Saved in:
Bibliographic Details
Main Authors: Chi-Wang Ho, Au, O.C., Gary Chan, S.-H., Shu-Kei Yip, Hoi-Ming Wong
Format: Conference Proceeding
Language:English
Subjects:
Arithmetic
Automatic voltage control
Central Processing Unit
Cost function
Graphics
Hardware
Motion estimation
Pipelines
Rendering (computer graphics)
Streaming media
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present an efficient implementation of motion estimation (ME) for H.264/AVC using programmable graphics hardware. The cost function for ME in H.264/AVC depends on the motion vector (MV) predictor which is the median MV of three neighboring coded blocks. Previous implementations assume no dependency among adjacent blocks, which is not true for H.264/AVC, they also perform unsatisfactorily because of their low arithmetic intensity, which is defined as operation per word transferred. To overcome the dependency problem, we introduce a new implementation which performs ME on block-by-block basis. Moreover, we can adjust the arithmetic intensity easily to optimize the performance on different graphics cards. Experimental results show that our implementation is substantially faster (by 10 times) than our SIMD optimized CPU implementation
ISSN:1945-7871
1945-788X
DOI:10.1109/ICME.2006.262617