Conference Paper

3DSEAM: a model for annotating 3D scenes using MPEG-7

LSRIMAG, Saint Martin d'Heres, France;
DOI: 10.1109/ISM.2005.2 Conference: Multimedia, Seventh IEEE International Symposium on
Source: DBLP

ABSTRACT The progress and the continuous evolution of computer capacities, as well as the emergence of the X3D standard have recently boosted the 3D domain. Associating some semantics with 3D contents becomes a major issue specially for reusing such contents or pieces of content after having extracted them from existing 3D scenes. In this paper, we address this issue by proposing a generic semantic annotation model for 3D, called 3DSEAM (3D semantics annotation model). 3DSEAM aims at indexing 3D contents considering visual, geometric and semantic aspects. 3DSEAM is instantiated using MPEG 7 that we have extended with 3D specific locators. These locators link some visual, geometric and semantic features to the corresponding X3D fragments. These features can then be used for indexing and querying.

  • [Show abstract] [Hide abstract]
    ABSTRACT: Nowadays, semantic information accompanying 3D graphical data, which is called metadata, has been realized to be essential for several purposes such that available 3D graphical data can be indexed and retrieved by means of its metadata and then other applications can make use of the 3D graphical data efficiently. Even though a number of works have focused on this topic so far, no temporal aspect of 3D graphical data has been taken into account. Especially no complete automatic metadata extraction process has been proposed. A manual process is possible with a small amount of 3D graphical data while impossible with a larger amount. Therefore, the manual process gets heavy and the cost in time and effort is so much required. In such a context, we introduce for 3D graphical data a temporal meta database system where an automatic metadata extraction method takes into consideration the temporal aspect of 3D graphical data. 3D graphical data is arranged in a scene formatted in the standard language VRML. Semantic information of the scene, i.e. metadata, will be automatically achieved. The structure of scene metadata is handled by using the temporal NIAM object oriented conceptual schema model at the abstraction level so that the semantic scene information can get more understandable to users, graphics designers, and application programmers as well. As a result, scene metadata is able to be easily stored and queried by means of some standardized database language such as SQL or OQL. With the aid of the extracted scene metadata, the respective 3D scenes are now exactly and conveniently accessed by other applications such as a virtual guide in transportation, architecture, tourism, etc.
    Computer and Information Technology, 2007. CIT 2007. 7th IEEE International Conference on; 11/2007
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: According to recent advances in computer science and engineering, three dimensional (3D) graphical data is of growing interest in many application areas. In this paper, a data management technique is proposed to handle 3D graphical data with the time dimension from a database perspective. Different from the existing works, our technique deals with 3D graphical data in all of the three various aspects: the visual aspect for the 3D graphical data part, the semantic aspect for the metadata part, and especially the temporal aspect for changes over the time. As a result, 3D graphical data can be managed properly with semantics and time. Therefore, many front-end applications can conveniently have access to them at both graphical and semantic data levels for sharability and reusability.
  • Source


Available from

Ioan Marius Bilasco