Information Technology Reference
In-Depth Information
2.2 Perceptually Optimized Coding
The perceptual importance of scene elements can be utilized to code and transmit
video content more efficiently. For example, the foreground objects of a video
frame can be coded with finer details than the background information. Perceptu-
ally optimized coding methods (e.g. ROI) are a common topic in 2D video coding
research, and can provide improved perceptual quality at reduced bitrates. For ex-
ample, the dynamic bit-allocation methods proposed in [14] and [15] utilize the
response of the human visual system to allocate bits for block-based and object-
based coding strategies respectively. Mixed-resolution coding of stereoscopic vid-
eo is also based on the response of the human visual system (e.g. the binocular
suppression theorem [16]). Reduced temporal-spatial coding of the right image
sequence has achieved overall bit-rate reduction, with no effect on the perceived
quality of stereoscopic video [17]. Therefore, the response of the human visual
system can be utilized to encode 2D/3D video more efficiently. Consequently,
compression schemes have been designed that use coarsely compressed depth im-
ages [18] and reduced resolution depth images [19] in order to optimize bit-rate.
The transmission of 3D video in particular can be prioritized based on the per-
ceptual importance of each component. For example, in the case of color plus depth
3D video, the color image sequence which is directly viewed by the human viewers
can be sent over a more protected channel than the depth sequence. However the
selection of priority levels should be made after careful observation of their effect
on the reconstructed quality of 3D video. For example, human perception
is
used to
decide the transmission strategy for 3D objects in [20]. In the presented framework,
3D video content is divided into segments, which are then coded and transmitted
based on their individual perceptual importance to achieve enhanced visual quality
at the receiving end. Perceptual importance is derived for these segments (back-
ground objects, foreground objects and depth information) by their expected trans-
mission distortion and corresponding distance from the viewing point.
2.3 Quality of Service for Wireless Local Area Networks
In order to enable wireless transmission of 3D videos in home environments, a
new standard IEEE 802.11e [21] is used. It has been selected because it provides
the Quality-of-Service (QoS) support for demanding multimedia applications with
stringent requirements. IEEE 802.11e supports two medium access mechanisms,
namely, controlled channel access and contention-based channel access, referred
to as Enhanced Distributed Channel Access (EDCA). EDCA provides the MAC
layer with per-class service differentiation. QoS support is realized with the intro-
duction of Traffic Categories (TCs) or Access Classes (ACs). With this approach,
frames are delivered through multiple back off instances within one station. The
implementation of legacy Distributed Coordination Function (DCF) and EDCA
with different traffic classes and independent transmission queues are shown in
Fig. 3. The priority levels for each TC can be differentiated based on the parame-
ters selected for Contention Window (CW) size (e.g. CW
min
, CW
max
), Arbitrary
