A new motion vector composition algorithm for H.264 multiple reference frame motion estimation

Tsz Kwan Lee, Yui Lam Chan, Chang Hong Fu, Wan Chi Siu

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

Abstract

Multiple reference frame motion estimation (MRF-ME) is an important tool in H.264 to improve coding efficiency. However, it penalizes the encoder in computational complexity. When the number of reference frames increases, therequired computation expands proportionally. Therefore, various motion vector composition algorithms have been proposed to reduce the computational complexity of the encoder. However, it is found that they only perform well in a limited range of reference frames. In this paper, a new composition algorithm is proposed to compose a resultant motion vector from a set of candidate motion vectors. The proposed algorithm is especially suited for temporally remote reference frames in MRF-ME. Compared with existing algorithms, experimental results show that the new algorithm can deliver a remarkable improvement on the rate-distortion performance.
Original languageEnglish
Title of host publicationAPSIPA ASC 2009 - Asia-Pacific Signal and Information Processing Association 2009 Annual Summit and Conference
Pages473-476
Number of pages4
Publication statusPublished - 1 Dec 2009
EventAsia-Pacific Signal and Information Processing Association 2009 Annual Summit and Conference, APSIPA ASC 2009 - Sapporo, Japan
Duration: 4 Oct 20097 Oct 2009

Conference

ConferenceAsia-Pacific Signal and Information Processing Association 2009 Annual Summit and Conference, APSIPA ASC 2009
Country/TerritoryJapan
CitySapporo
Period4/10/097/10/09

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Electrical and Electronic Engineering
  • Communication

Fingerprint

Dive into the research topics of 'A new motion vector composition algorithm for H.264 multiple reference frame motion estimation'. Together they form a unique fingerprint.

Cite this