Scalable Object-Based Indexing of HD Videos: A JPEG2000- Oriented solution

Claire Morand (1) , Jenny Benois-Pineau (1) , Jean-Philippe Domenger (1, 2)

1 LaBRI - Laboratoire Bordelais de Recherche en Informatique
2 GRAVITE - Graph Visualization and Interactive Exploration

Video indexing technique is crucial in multimedia applications. In the case of HD (High Definition) Video, the principle of scalability is of great importance. The wavelet decomposition used in the JPEG2000 standard provides this property. In this paper, we propose a scalable descriptor based on objects. First, a scalable moving object extraction method is constructed. Using the wavelet data, it relies on the combination of a robust global motion estimation with a morphological color segmentation at a low spatial resolution. It is then refined using the scalable order of data. Second, a descriptor is built only on the objects found at the previous step. This descriptor is based on multiscale histograms of wavelet coefficients of moving objects

Mots clés: segmentation par object des séquences vidéos; descripteur scalble; pyramide des ondelettes

Domaines: Multimédia [cs.MM]

Claire Morand, Jenny Benois-Pineau, Jean-Philippe Domenger. Scalable Object-Based Indexing of HD Videos: A JPEG2000- Oriented solution. 2008. ⟨hal-00349165⟩