Scribd
Upload
41 views

Video Image Compression

Uploaded by

jason beryl
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
41 views

Video Image Compression

Uploaded by

jason beryl
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 16

VIDEO IMAGE

COMPRESSION

- Sagar I MSc 21872


What does mean video compression ?

> Video compression is the process of encoding a video


file in such a way that it consumes less space than the
original file.
VIDEO IMAGE COMPRESSION

The development of digital video technology has made it


possible to use digital video compression for a variety of
telecommunications applications.
Standardization of compression algorithms for video was
first initiated by CCITT (Consultative Committee on
International Telephony & Telegraphy) for
teleconferencing and video telephony.
Multimedia Standards For Video

P*64 JPEG MPEG


(CCITT) (ISO) (ISO)
Video Conferencing Still Images Stored Video

Video Image Video

H.261 Motion Video MPEG Encoding/


JPEG Baseline
(CIF,QCIF) Decoding

Audio

G.711 MPEG Audio


G.722 (Layers 1,2,3)
G.728
Requirements for full-motion Video
Compression

 Applications using MPEG standards can be symmetric or asymmetric.


 Symmetric applications that require essentially equal use of
compression & decompression.
 Asymmetric applications require frequent decompression.
 Symmetric applications require online input devices video cameras,
scanners & microphones for digitized sound.
 In addition to video & audio compression, this standards activity is
concerned with playback of video clips & sound clips.
 The MPEG standard has identified a number of such issues that have
been addressed by the standards activity. Let us review these Issues …
 Random Access
The expectations generated for multimedia systems are ability to play sound or
video clip from any frame with that clip, irrespective of on what kind of media
the information is stored.

 VCR paradigm
The VCR paradigm consists of the control functions typically found on VCR
such as play, fast forward, rewind, search forward & rewind search.

 Multiplexing Multiple Compressed Audio & Video Bit Stream


It is a special requirement of multimedia messaging system & other multimedia
applications is the need to multiplex multiple audio & video bit streams retrieved
from different storage centres on a network. It may have to be achieved in a
smooth manner to avoid the appearance of a jumpy screen.

 Edit ability
 Playback device Flexibility
CCITT H.261 Video Coding
Algorithms (P * 64)
 The linear quantizer uses a step algorithm that can be adjusted based on
picture quality & coding efficiency.
 The H.261 is a standard that uses a hybrid of DCT & DPCM (Differential
Pulse Code Modulation) schemes with motion estimation.
 It also define the data format. Each MB contains the DCT coefficients
(DCTCOEFF) of a block followed by an EOB (a fixed length end – of –
block maker).
 Each MB consists of block data & an MB header. A GOB (Group of
Blocks) consists of a GOB header.
 The picture layer consists of a picture header.
 The H.261 is designed for dynamic use & provides a fully contained
organization & a high level of interactive control.
Moving Picture Experts Group
(MPEG) Compression
 The MPEG standards consists of number of different standards.
 The MPEG 2 suite of standards consist of standard for MPEG 2 video ,
MPEG 2 audio & MPEG 2 systems. It is also defined at different levels ,
called Profiles.
 The main profile is designed to cover the largest number of applications.
 It supports digital video compression in the range of 2 to 15M bits/sec.
 It also provides a generic solution for television worldwide, including
cable, direct broadcast satellite, fibre optic media & optical storage media
(including digital VCRs).
MPEG Coding Methodology

 The above said requirements can be achieved only by incremental


coding of successive frames. It is known as interframe coding.
 If we access information randomly by frame requires coding confined
to a specific frame, then it is known as intraframe coding.
 The MPEG standard addresses these 2 requirements by providing a
balance between interframe & intraframe coding.
 The MPEG standard also provides for recursive & non recursive
temporal redundancy reduction.
 The MPEG video compression standard provides 2 basic schemes: discrete
– transform – based compression for the reduction of spatial redundancy &
block – based motion compensation for the reduction of temporal ( motion )
redundancy.
 During the initial stages of DCT compression, both the full motion MPEG
& still image JPEG algorithms are essentially identical.
 First an image is converted to the YUV colour space (a luminance /
chrominance colour space similar to that used for television ).
 The pixel data is then fed into a discrete cosine transform, which creates a
scalar quantization ( 2 Dimensional array representing various frequency
ranges represented in the image) of the pixel data.
 Following quantization, a number of compression algorithms are applied,
including run length and Huffman encoding.
 For full motion video ( MPEG 1 & 2 ), several more levels of block based
motion – compensated techniques are applied to reduce temporal
redundancy with both causal & noncausal coding to further reduce spatial
redundancy.
Moving Picture Types

 Moving picture consist of sequences of video pictures that are


played back a fixed number of frame per second. To achieve
the requirement of random access, a set of pictures can be
defined to form a group of pictures (GOP) consisting of one
or more of the following 3 types of pictures.

1. Intra pictures (I)


2. Unidirectionally predicted pictures(U)
3. Bidirectionally predicted pictures(B)
A Gap consists of consecutive pictures that begin with an intra
pictures.
The intra pictures is coded without any references to any other
pictures in the group.
Predicted pictures are coded with a reference to a past pictures,
either an intra picture or a unidirectionally predicted picture.
Bidirectionally predicted picture is never used as references
Motion Compensation for Coding MPEG.
MPEG Encoder :
It contains DCT quantizer, Huffman coder & Motion compensation. These represent the key
modules in the encoder.
Architecture of MPEG Encoder :

+
Macro Huffman
block DCT Q
Coder
input
-1
Q
MUX
IDCT

Motion
Com
pensation
Past Frame

Future
+ Huffman
Coder

Frame
Motion Compensation
 Motion compensation is the basis for most compression
algorithms for visual telephony and full – motion video.
 Motion Compensation assumes that the current picture is
some translation of a previous pictures.
 This creates the opportunity for using prediction &
interpolation.
 Prediction requires only the current frame & the reference
frame.
The Sequence of events for MPEG

 First an image is converted to the YUV colour space.


 The pixel data is then fed into a DCT, which creates a scalar
quantization of the pixel data.
 Following quantiztion, a number of compression algorithms
are applied, including run – length and Huffman encoding.
 For full – motion video, several more levels of motion
compensation compression & coding are applied.
THANK YOU 

You might also like