2nd International ICST Conference on Mobile Multimedia Communications

Research Article

Motion driven adaptive transform based on wavelet transform for enhanced video coding

  • @INPROCEEDINGS{10.1145/1374296.1374307,
        author={Nikola Sprljan and Marta Mrak and Ebroul Izquierdo},
        title={Motion driven adaptive transform based on wavelet transform for enhanced video coding},
        proceedings={2nd International ICST Conference on Mobile Multimedia Communications},
        publisher={ACM},
        proceedings_a={MOBIMEDIA},
        year={2006},
        month={9},
        keywords={Adaptive wavelet transform scalable video coding intra coding.},
        doi={10.1145/1374296.1374307}
    }
    
  • Nikola Sprljan
    Marta Mrak
    Ebroul Izquierdo
    Year: 2006
    Motion driven adaptive transform based on wavelet transform for enhanced video coding
    MOBIMEDIA
    ACM
    DOI: 10.1145/1374296.1374307
Nikola Sprljan1,*, Marta Mrak1,*, Ebroul Izquierdo1,*
  • 1: University of London, London, UK
*Contact email: nikola.sprljan@elec.qmul.ac.uk, marta.mrak@elec.qmul.ac.uk, ebroul.izquierdo@elec.qmul.ac.uk

Abstract

Spatial wavelet transform in video coding has traditionally been applied in a non-adaptive fashion. However, since motion compensation introduces specific structure into video frames, adaptive spatial transform can introduce additional coding gain. As the structure introduced by motion compensation depends on applied motion parameters, the same motion information can be used as an adaptation parameter in spatial decomposition. Specifically, in this paper motion driven adaptation on wavelet transform in its lifting implementation is proposed. The introduced computational complexity of the proposed method is low as only a few additional multiplications are introduced per lifting step. Since in this case the same motion information is available at both encoder and decoder, transmission of additional side information is not needed. Besides better energy compaction, the proposed scheme introduces considerable gain when applied in scalable video coding.