Information Technology Reference
In-Depth Information
7 Ongoing and Future Work Regarding BBC Dirac Video
Codec
Efforts are going on across the world in order to improve the different modules of
Dirac to make it suitable for real time video conferencing system. Different
research papers have been published to address the computational complexity of
the Dirac video codec. There is also working to embed new features in Dirac such
as Scalability, Audio etc. Some of the ongoing work and future directions on
Dirac video codec are now discussed.
7.1 Scalability
Scalability is an appealing concept. It can be seen as functionality where parts of a
total bit-stream of a video are extracted as different sub-streams. Each sub-stream
represents reduced source contents in temporal, spatial or SNR resolution as
compared to the original bit-stream. The sub-streams are divided into layers i.e.
base layer and enhancement layers. The base layer is a layer which is first encoded
into sub-stream from the source content with low frame rate, low temporal or
spatial resolution. The base layer is independently decodable and it produces an
acceptable baseline quality. The residual information between base layer and
source content is then encoded into one or more sub-streams and each sub-stream
represents an enhancement layer. The enhancement layers cannot be
independently decodable and can be added to the base layer to provide the user
with high perceptual quality. The base layer should be reached at the end user
correctly; otherwise the video will not be recovered properly. At decoder side,
video can be decoded at any supported spatial, temporal or SNR resolution
extracting a portion of the initial bit-stream. Scalability have been introduced in
many video codecs that uses Wavelet Transform. In [16], Fully-scalable wavelet
video coding using in-band motion compensated temporal filtering is described.
Scalability allows great advantage in broadcasting/multicasting to users of
heterogeneous capabilities.
There are three types of Scalability that is as follows:
a) SNR Scalability
b) Spatial Scalability
c) Temporal Scalability.
7.1.1 SNR Scalability
SNR scalability involves generating two or more layers of the same
spatiotemporal resolution but different video qualities from a single video source
Search WWH ::
Custom Search