Internal memory optimization scheme for spatial scalable inter-layer prediction

Jinha Choi, Jonghyun Bae, Jaeseok Kim

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

Abstract

This paper proposes an internal memory optimization method for the spatial scalable inter-layer prediction. The spatial scalable prediction of the H.264 Scalable Video Coding (SVC) has additional inter-layer predictions. The up-scale processing of the inter-layer prediction consists of the Wiener interpolation filter and the deblocking filter. They need internal memories for the pixel buffering. Thus, they need memory optimizing for an efficient hardware implementation. In addition, the SVC motion prediction requires a significant amount of motion vector from the previous layer prediction and an additional interpolation process. These motion vectors cause the internal memory size to increase which may result in increasing hardware cost and power consumption. To optimize internal memories, this paper proposes a memory optimized architecture for the deblocking filter and a motion vector bit compression scheme of the inter-layer motion prediction. The proposed architecture can control the internal memory easily and the proposed compression scheme reduces motion vector storing area by about 66.49% with less change in hardware size.

Original languageEnglish
Title of host publicationTENCON 2009 - 2009 IEEE Region 10 Conference
DOIs
Publication statusPublished - 2009 Dec 1
Event2009 IEEE Region 10 Conference, TENCON 2009 - Singapore, Singapore
Duration: 2009 Nov 232009 Nov 26

Other

Other2009 IEEE Region 10 Conference, TENCON 2009
CountrySingapore
CitySingapore
Period09/11/2309/11/26

Fingerprint

Data storage equipment
Scalable video coding
Computer hardware
Interpolation
Memory architecture
Electric power utilization
Pixels
Hardware
Processing
Costs

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Computer Science Applications

Cite this

Choi, Jinha ; Bae, Jonghyun ; Kim, Jaeseok. / Internal memory optimization scheme for spatial scalable inter-layer prediction. TENCON 2009 - 2009 IEEE Region 10 Conference. 2009.
@inproceedings{8758b635cf7b4e32b649d0cd48c73137,
title = "Internal memory optimization scheme for spatial scalable inter-layer prediction",
abstract = "This paper proposes an internal memory optimization method for the spatial scalable inter-layer prediction. The spatial scalable prediction of the H.264 Scalable Video Coding (SVC) has additional inter-layer predictions. The up-scale processing of the inter-layer prediction consists of the Wiener interpolation filter and the deblocking filter. They need internal memories for the pixel buffering. Thus, they need memory optimizing for an efficient hardware implementation. In addition, the SVC motion prediction requires a significant amount of motion vector from the previous layer prediction and an additional interpolation process. These motion vectors cause the internal memory size to increase which may result in increasing hardware cost and power consumption. To optimize internal memories, this paper proposes a memory optimized architecture for the deblocking filter and a motion vector bit compression scheme of the inter-layer motion prediction. The proposed architecture can control the internal memory easily and the proposed compression scheme reduces motion vector storing area by about 66.49{\%} with less change in hardware size.",
author = "Jinha Choi and Jonghyun Bae and Jaeseok Kim",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/TENCON.2009.5395870",
language = "English",
isbn = "9781424445479",
booktitle = "TENCON 2009 - 2009 IEEE Region 10 Conference",

}

Choi, J, Bae, J & Kim, J 2009, Internal memory optimization scheme for spatial scalable inter-layer prediction. in TENCON 2009 - 2009 IEEE Region 10 Conference., 5395870, 2009 IEEE Region 10 Conference, TENCON 2009, Singapore, Singapore, 09/11/23. https://doi.org/10.1109/TENCON.2009.5395870

Internal memory optimization scheme for spatial scalable inter-layer prediction. / Choi, Jinha; Bae, Jonghyun; Kim, Jaeseok.

TENCON 2009 - 2009 IEEE Region 10 Conference. 2009. 5395870.

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

TY - GEN

T1 - Internal memory optimization scheme for spatial scalable inter-layer prediction

AU - Choi, Jinha

AU - Bae, Jonghyun

AU - Kim, Jaeseok

PY - 2009/12/1

Y1 - 2009/12/1

N2 - This paper proposes an internal memory optimization method for the spatial scalable inter-layer prediction. The spatial scalable prediction of the H.264 Scalable Video Coding (SVC) has additional inter-layer predictions. The up-scale processing of the inter-layer prediction consists of the Wiener interpolation filter and the deblocking filter. They need internal memories for the pixel buffering. Thus, they need memory optimizing for an efficient hardware implementation. In addition, the SVC motion prediction requires a significant amount of motion vector from the previous layer prediction and an additional interpolation process. These motion vectors cause the internal memory size to increase which may result in increasing hardware cost and power consumption. To optimize internal memories, this paper proposes a memory optimized architecture for the deblocking filter and a motion vector bit compression scheme of the inter-layer motion prediction. The proposed architecture can control the internal memory easily and the proposed compression scheme reduces motion vector storing area by about 66.49% with less change in hardware size.

AB - This paper proposes an internal memory optimization method for the spatial scalable inter-layer prediction. The spatial scalable prediction of the H.264 Scalable Video Coding (SVC) has additional inter-layer predictions. The up-scale processing of the inter-layer prediction consists of the Wiener interpolation filter and the deblocking filter. They need internal memories for the pixel buffering. Thus, they need memory optimizing for an efficient hardware implementation. In addition, the SVC motion prediction requires a significant amount of motion vector from the previous layer prediction and an additional interpolation process. These motion vectors cause the internal memory size to increase which may result in increasing hardware cost and power consumption. To optimize internal memories, this paper proposes a memory optimized architecture for the deblocking filter and a motion vector bit compression scheme of the inter-layer motion prediction. The proposed architecture can control the internal memory easily and the proposed compression scheme reduces motion vector storing area by about 66.49% with less change in hardware size.

UR - http://www.scopus.com/inward/record.url?scp=77951104840&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951104840&partnerID=8YFLogxK

U2 - 10.1109/TENCON.2009.5395870

DO - 10.1109/TENCON.2009.5395870

M3 - Conference contribution

SN - 9781424445479

BT - TENCON 2009 - 2009 IEEE Region 10 Conference

ER -