Cryptography Reference
In-Depth Information
Group of pictures
Group of pictures
Macroblock
Macroblock
8
8
8
8
Slice
Slice
Picture
Picture
Block
Block
Fig. 2.15.
Video bitstream hierarchy.
Video sequence: This is the highest layer which defines the context valid for
the whole sequence. A video sequence consists of sequence header, the
number of group of pictures (GOP), and end-of-sequence code. The se-
quence header initializes the state of the decoder. This allows a decoder
to decode any sequence without being affected by past decoding history.
Group of pictures (GOP): A GOP consists of a series of one or more pictures
intended to allow random access into the sequence. A GOP always starts
with an I-picture.
Picture: The picture is the primary coding unit of the video sequence which
consists of 3 rectangular matrices representing one luminance (Y ) and two
chrominance (C
b
and C
r
) components.
Slice: A slice is the layer for Intra frame addressing and (re)synchronization.
It consists of one or more contiguous macroblocks. Slices should have the
similar amount of data.
Macroblock (MB): A typical MB contains four Y blocks, one or two C
b
and
C
r
blocks depending on whether it is 4:2:0 or 4:2:2 sub-sampled. MB is
the basic unit for motion prediction and compensation.
blocks: A block is an 88 pixel section of luminance or chrominance compo-
nents. It is the basic unit for DCT based Intra frame coding.
2.4.2 Picture Types
Most of video coding standards have three main picture types, and they are,
I-Picture: These are Intra coded picture, which means they are coded inde-
pendent of any other pictures. All the MBs in an I-picture are Intra coded
using DCT plus quantization plus zigzag scanning and run length coding
followed by VLC coding.
