Information Technology Reference
In-Depth Information
Table 2
Generation of 3D surface
Layer
Mode
# of depth pixels
# of triangles
Shape of surface
Regular mesh
-
4
2
Up-left
14
20
Down-left
14
20
Boundary
Up-right
14
20
Down-right
14
20
Full modeling
25
44
1
st
layer
5
4
2
nd
layer
6
6
Feature
Points
3
rd
layer
7
8
4
th
layer
8
10
color image using texture mapping. As a result, we can realize a fast rendering system
to support 3D video contents based on video-plus-depth in real time.
4 MPEG-4-Based 3D Video Contents Exploitation
In order to deliver 3D video contents, a multimedia framework is needed. Tradi-
tional multimedia frameworks, such as MPEG-1 and MPEG-2, merely deal with
efficient coding issues and synchronization problems between video and audio. In
addition, they do not provide any framework to support interactive functionalities
to users. Hence, we direct attention to the MPEG-4 multimedia framework [20]
that supports streaming functionality for a variety of media objects and provides
flexible interactivity.
In this chapter, we design a new node for a depth image sequence in the
MPEG-4 system to provide a practical solution to stream 3D video contents while
supporting a variety of user-friendly interactions. Figure 8 illustrates the overall
system architecture to generate the 3D video contents based on video-plus-depth
in the MPEG-4 multimedia framework.
At the sender side, we generate high-resolution video-plus-depth using the hy-
brid camera system as introduced in Section 2. Video-plus-depth is then spatio-
temporally combined with other multimedia, such as audio and computer graphics
models, using the MPEG-4 Binary Format for Scene (BIFS). The MPEG-4 BIFS
is a scene descriptor that contains the spatio-temporal relationship between each
multimedia object and some interactivity information. The video-plus-depth data
are compressed by two video coders; one for color image sequence and the other
