Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps

Yiqing Huang, Qin Liu, Takeshi Ikenaga

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

One Super Hi-Vision (SHV) 4k×4k@60fps fractional motion estimation (FME) engine is proposed in our paper. Firstly, two complexity reduction schemes are proposed in the algorithm level. By analyzing the integer motion cost, 48% clock cycle is saved based on our mode pre-filtering scheme. By further check the motion cost of neighboring search points, our directional one-pass scheme can achieve reduction of 50% clock cycle and 36% hardware cost. Secondly, in the hardware level, two parallel improved schemes namely 16-Pel interpolation and MB-parallel processing are given out. Thirdly, one unified pixel block loading scheme is proposed. About 28.67% to 80.68% pixels are reused and the related memory access is saved. Furthermore, one parity pixel organization scheme is proposed to solve memory access conflict of MB-parallel processing. By using TSMC 0.18μm technology in worst work conditions (1.62V, 125°C), our FME engine can achieve real-time processing for SHV 4k×4k@60fps with 976.5k gates hardware.

Original languageEnglish
Title of host publication2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09
DOIs
Publication statusPublished - 2009
Event2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09 - Rio De Janeiro
Duration: 2009 Oct 52009 Oct 7

Other

Other2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09
CityRio De Janeiro
Period09/10/509/10/7

Fingerprint

Motion estimation
Pixels
Engines
Hardware
Clocks
Processing
Data storage equipment
Costs
Interpolation

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Huang, Y., Liu, Q., & Ikenaga, T. (2009). Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps. In 2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09 [5293323] https://doi.org/10.1109/MMSP.2009.5293323

Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps. / Huang, Yiqing; Liu, Qin; Ikenaga, Takeshi.

2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09. 2009. 5293323.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Huang, Y, Liu, Q & Ikenaga, T 2009, Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps. in 2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09., 5293323, 2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09, Rio De Janeiro, 09/10/5. https://doi.org/10.1109/MMSP.2009.5293323
Huang Y, Liu Q, Ikenaga T. Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps. In 2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09. 2009. 5293323 https://doi.org/10.1109/MMSP.2009.5293323
Huang, Yiqing ; Liu, Qin ; Ikenaga, Takeshi. / Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps. 2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09. 2009.
@inproceedings{366166304856435fa06ae0ab05302378,
title = "Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps",
abstract = "One Super Hi-Vision (SHV) 4k×4k@60fps fractional motion estimation (FME) engine is proposed in our paper. Firstly, two complexity reduction schemes are proposed in the algorithm level. By analyzing the integer motion cost, 48{\%} clock cycle is saved based on our mode pre-filtering scheme. By further check the motion cost of neighboring search points, our directional one-pass scheme can achieve reduction of 50{\%} clock cycle and 36{\%} hardware cost. Secondly, in the hardware level, two parallel improved schemes namely 16-Pel interpolation and MB-parallel processing are given out. Thirdly, one unified pixel block loading scheme is proposed. About 28.67{\%} to 80.68{\%} pixels are reused and the related memory access is saved. Furthermore, one parity pixel organization scheme is proposed to solve memory access conflict of MB-parallel processing. By using TSMC 0.18μm technology in worst work conditions (1.62V, 125°C), our FME engine can achieve real-time processing for SHV 4k×4k@60fps with 976.5k gates hardware.",
author = "Yiqing Huang and Qin Liu and Takeshi Ikenaga",
year = "2009",
doi = "10.1109/MMSP.2009.5293323",
language = "English",
isbn = "9781424444649",
booktitle = "2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09",

}

TY - GEN

T1 - Highly parallel fractional motion estimation engine for Super Hi-Vision 4k×4k@60fps

AU - Huang, Yiqing

AU - Liu, Qin

AU - Ikenaga, Takeshi

PY - 2009

Y1 - 2009

N2 - One Super Hi-Vision (SHV) 4k×4k@60fps fractional motion estimation (FME) engine is proposed in our paper. Firstly, two complexity reduction schemes are proposed in the algorithm level. By analyzing the integer motion cost, 48% clock cycle is saved based on our mode pre-filtering scheme. By further check the motion cost of neighboring search points, our directional one-pass scheme can achieve reduction of 50% clock cycle and 36% hardware cost. Secondly, in the hardware level, two parallel improved schemes namely 16-Pel interpolation and MB-parallel processing are given out. Thirdly, one unified pixel block loading scheme is proposed. About 28.67% to 80.68% pixels are reused and the related memory access is saved. Furthermore, one parity pixel organization scheme is proposed to solve memory access conflict of MB-parallel processing. By using TSMC 0.18μm technology in worst work conditions (1.62V, 125°C), our FME engine can achieve real-time processing for SHV 4k×4k@60fps with 976.5k gates hardware.

AB - One Super Hi-Vision (SHV) 4k×4k@60fps fractional motion estimation (FME) engine is proposed in our paper. Firstly, two complexity reduction schemes are proposed in the algorithm level. By analyzing the integer motion cost, 48% clock cycle is saved based on our mode pre-filtering scheme. By further check the motion cost of neighboring search points, our directional one-pass scheme can achieve reduction of 50% clock cycle and 36% hardware cost. Secondly, in the hardware level, two parallel improved schemes namely 16-Pel interpolation and MB-parallel processing are given out. Thirdly, one unified pixel block loading scheme is proposed. About 28.67% to 80.68% pixels are reused and the related memory access is saved. Furthermore, one parity pixel organization scheme is proposed to solve memory access conflict of MB-parallel processing. By using TSMC 0.18μm technology in worst work conditions (1.62V, 125°C), our FME engine can achieve real-time processing for SHV 4k×4k@60fps with 976.5k gates hardware.

UR - http://www.scopus.com/inward/record.url?scp=74349120940&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=74349120940&partnerID=8YFLogxK

U2 - 10.1109/MMSP.2009.5293323

DO - 10.1109/MMSP.2009.5293323

M3 - Conference contribution

AN - SCOPUS:74349120940

SN - 9781424444649

BT - 2009 IEEE International Workshop on Multimedia Signal Processing, MMSP '09

ER -