|
Descripción
|
|
|---|---|
| 3D video technology has been on the rise in the last years, with abundant research advances accompanied by a widespread adoption by the film industry and an increasing importance in consumer electronics. On a related note, there is the concept of multiview video, which encompasses 3D video, and can be defined as a video stream composed of two or more views. Multiview video enables advanced video features such as stereoscopic video, free viewpoint video, improved eye contact through use of virtual views, or shared virtual environments. The purpose of this thesis is to address a major obstacle towards the use of multiview video in communication systems ? the lack of support for this technology from existing signalling protocols, which makes it impossible to configure a multiview video session through standard means. Thus, our main objective is to extend the Session Initiation Protocol (SIP) to support the negotiation of multimedia sessions with multiview video streams. Our work can be summarized in three major contributions. In the first place, we have defined a signalling extension for the configuration of SIP sessions with 3D video. This extension changes the Session Description Protocol (SDP) to introduce a new media-level attribute and a new type of decoding dependency, which help describe the 3D video formats that can be used in a session, and the relationship between video streams composing a 3D video stream. The second contribution consists in a SIP extension to handle the signalling of video conferences with multiview video streams. Two new SIP event packages are defined to describe the capabilities and topology of conferencing terminals, on the one hand, and the spatial configuration and stream mapping of a conference, on the other. A way to integrate the exchange of this information into the initiation of a SIP conference is also described. For the third and final contribution, we introduce the concept of conference virtual space, as a coordinate system where all relevant objects in a conference (such as capture devices, displays, and users) are included. We explain how the virtual space relates to conference features such as eye contact, video scale and spatial faithfulness, and provide guidance on how to determine the features of a conference through virtual space analysis, and on the generation of virtual spaces as part of conference configuration. | |
|
Internacional
|
No |
|
ISBN
|
|
|
Tipo de Tesis
|
Doctoral |
|
Calificación
|
Sobresaliente |
|
Fecha
|
13/07/2012 |