Introduction - 3DTV: Processing and Transmission of 3D Video Signals - page 20

seen commercial usage so far, and is unlikely to do so. More recent standards
are capable of coding multiple views with much greater efficiency than this
MPEG-2 extension (see Section 3.2 for more details).
Significant 3DTV work has been carried out in Japan over the years. Sand's
(1992) paper describes research at the University of Tokyo, where an auto-
stereoscopic system with up to 24 channel multiple-views was displayed [33].
Nippon H os ¯oKyokai (NHK) developed a 3D HDTV relay system, which
was used in 1998 to transmit stereoscopic live images of the Nagano Winter
Games via satellite to a large-scale demonstration venue in Tokyo [36].
The recent arrival of affordable 3D television sets means that 3DTV
broadcasts are once again of interest to broadcasters. Many experimental
broadcasts are taking place across the world, particularly using the frame
compatible stereoscopic format (see Section 1.2.1). The frame compatible
format can already be deployed using existing digital broadcast technology,
such as DVB-S2. A commercial 3DTV channel was started by Sky Television
in the United Kingdom in October 2010.
1.2 3D Video Formats
This section provides an overview of the most common video formats
currently in use, or currently being considered by the research community.
The formats are summarized in Table 1.2, which also describes the respective
advantages and disadvantages. The frame compatible and service compatible
formats are currently the most commonly used for commercial applications,
and are supported by most consumer 3DTVs. The stereoscopic video formats
are relatively straightforward, as they consist of only two colour video views.
Therefore, most space in this section is given over to the description of the
depth-based formats, and multi-view formats.
1.2.1 FrameCompatibleandServiceCompatible
StereoscopicVideo
The first formats to be used in modern 3DTV systems will be stereoscopic,
as this type of video does not require excessive transmission bit-rates, and
is easier to capture than the other formats described in this section. All
stereoscopic formats suffer from a lack of flexibility in terms of rendering. It
is very difficult to change the users viewpoint, and to change the disparity
between the views presented to the user. This prevents users from adjusting
the presentation of the 3D video for comfortable viewing.
The frame compatible stereoscopic video format has been selected for use
in the first generation of 3DTV systems. Its principal advantage is that the
two views are packed together within a single video frame. This means
that the format is compatible with most existing digital television sys-
tems, as no change of resolution or frame rate is required to support 3D.