Transmission - 3DTV: Processing and Transmission of 3D Video Signals - page 100

2. Temporal direct motion vector generation - In this method all the motion
vectors and reference indices for the missing picture are generated in a
similar pattern to the temporal direct mode algorithm.
3. Motion and residual upsampling (BISkip) - The residual data and motion
information is generated by upsampling the corresponding information
of the base layer while the motion compensation is performed at the
enhancement layer. Therefore, it is obvious that this algorithm can be
used only to conceal errors at the enhancement layer.
4. Reconstructed Base Layer upsampling (RU) - As the name implies, this is an
error concealment algorithm for the enhancement layers. In this method,
the reconstructed base layer image is upsampled using 6-tap filters to
recreate the lost frame.
5. Flexible Macroblock Ordering (FMO) - In FMO, the MB data will be orga-
nized into different slice groups. Then these groups will be re-arranged
according to a predefined pattern which is known to both the encoder and
the decoder, and are transmitted in separate packets. At the receiving end,
the decoder re-arranges the scattered data before decoding. The missing
portions, i.e. the erroneous segments, are predicted from the correctly
decoded segments of the video sequence. By using this method, the sys-
tem can minimize losses and perform predictions more accurately than
the frame concealment option. In the H.264/AVC standard there are seven
predefined patterns and the pattern ''type1'' gives the best performance
of all these patterns.
4.2.3 ForwardErrorCorrection(FEC)
FEC is another technique used in error-prone environments to detect and
correct bit errors in bit streams. In this method, at the transmission end some
additional data, known as parity bits, are enclosed in the bit stream to facilitate
error recovery. The selection of the appropriate FEC codes for a particular
application depends on the characteristics of the algorithm such as resource
consumption, memory usage, flexibility and computational complexity.
Some of the FEC codes used extensively in literature are listed below with
their key features:
1. Reed Solomon codes:
— Simple systematic linear block error correction code.
— With the increase in file size, channel bit-rate and error rate, the
amount of overhead added increases rapidly, resulting in a massive
extra bandwidth requirement.
— Requires lot of memory for the processing. But this can be improved
by means of memory management algorithms. However, that in turn
results in an increase in decoding time.
— Is capable of handling variable packet sizes with low additional over-
head. Yet, this increases the computational complexity.