Transcript MPEG-1Standard - Florida International University SCIS

MPEG-1Standard
By
Alejandro Mendoza
Introduction
• The major goal of video compression is
to represent a video source with as few
bits as possible while preserving the
level of quality required for the given
video application.
MPEG Consists of two stream layers:
• System layer contains timing and other information
needed to demultiplex the audio and video during
playback.
• The compression layer which includes the audio and
video streams.
Video Stream Data Hierarchy
•
•
•
•
Group of picture: a header and a series of pictures
Picture: primary coding unit of a video sequence
Slice: one or more continuous macroblock
Macroblock: a 16-pixel by 16-line contains four Y blocks for
brightness, one Cb block and one Cr block for color
• Block: an 8-pixel by 8-line set of values of brightness or color
component
Audio Stream Data Hierarchy
• Contains packet and frames
• Audio frame header: synchronization, ID, bitrate, and
sampling frequency information
• Error-checking code
• Audio data: information used to reconstruct the
sampled audio data
• Ancillary data: user-defined data
Inter-Picture Coding
• Intra Pictures: coded using only information present in
the picture itself, no motion compensation
• Predicted picture: coded with respect to the nearest
previous I- or P- pictures, called forward prediction
Inter-Picture Coding
• Bidirectional Pictures: pictures that use both a past
and future picture as reference. They provide the
most compression and don’t propagate errors since
they are never used as a reference
Video Stream Composition
MPEG-1 audio
Standardizes three different coding
schemes for digitized sound waves
Layer I: Lowest Complexity, uses Psychoacoustic
model
Layer II: One encoder serves many decoders
Layer III: A.K.A. MP3, Enhanced frequency
resolution, use Huffman Coding
How Does it Work
Sound is pressure differences in air. When pick up
by a microphone and fed through an amplifier
becomes voltage level. The voltage is sample by
the computer a number of time per second. For
CD quality the sample rate is 44000 times per
second and each sample has a resolution of 16 bits
giving a total of 1.4Mbits per second.
Compression / Masking Effect
•
•
•
•
Can compress bitstream of 32kbit/s to 384kbit/s
Removes irrelevant and redundant parts of the signal
Sound we do not hear are thrown away
Uses psychoacoustic principle to deal with the way the
human brain perceives sound.
• Frequency range of 20Hz to 20kHz divided into 32 subbands
• Use mid frequencies since the ear is less sensitive in the
high and low frequencies
• High signal-to-noise ratio at 90dB
MPEG-1 Audio Encoding