2. Name : Sahil Jain
Roll No. : 104
Branch : ECE
Year : 3rd
3. CONTENTS
Overview
• A Little Theory
• Achieving Compression
• Various Algorithms
• Spatial Compression
• Spectral Compression
• Steps in Image Compression
• Temporal Compression
• Video Coding and Frames
• Motion Compensated Prediction
• Motion Estimation
• Block Matching Summary
• Basic VC Architecture
• Video Encoder
• Video Decoder
• Using Video Compression Standards
• Current Video Compression Standards
• Video Coding Standardization Organizations
• Dynamics of Video Standardization Process
• Video Compression Standards
• Scope of Development
• References and Further Reading
4. OVERVIEW
To reduce quantity of data used to represent digital video images.
•Saves space and bandwidth
• Saves energy
• Increases portability
• Reduce cost
Eg.- 720 X 1280 pixels/frame, progressive scanning @60
frames/sec
(720X 1280 ppf)(60 fps)(3 colors/pixel)(8 bits/color) =
1.3Gb/sec
• 20 Mbps HD channel bandwidth
• Requires compression by a factor of 70(equivalent to .35
bits/ pixel)
6. A Little Theory
Video – It is a 3-D array of color pixels.
• Two of the dimensions serve as the spatial domain for moving
pictures
• One dimension serves as the time domain
Data Frame – Set of all pixels that corresponds to single time
moment.
Human eye
• Not responsive to every detail
• Quantization
• Smoothening
• More responsive to brightness than chrominance
7. Achieving Compression
Reduce redundancy and irrelevancy.
Sources of Redundancy
• Temporal: Adjacent frames highly correlated
• Spatial: Nearby pixels are often correlated
• Color space: RGB components are correlated among themselves
• Relatively straight to exploit
Sources of Irrelevancy
• Perceptually unimportant information
• Difficult to model and exploit
8. Various Algorithms
4 Major ways to compress videos
DCT(Discrete Cosine Transform)
• Samples images at regular intervals
• Analyze frequency components present in sample
• Discard unimportant frequencies from point of view of human
eye
Vector Quantization(VQ)
• It looks at an array of data rather than individual values and
than averaging what it perceives
• Compresses the found redundant data keeping the desired
object
contd…
9. Fractal compression(FC)
• Form of vector quantization.
• Finds self-similar section of particular image, than uses fractal
algorithm to create the sections
Discrete Wavelet Transform(DWT)
• Mathematically transform image into frequency components
• Process is performed on entire frame, the end result is very
effective hierarchical representation of an image
• Every layer represents a frequency band
10. Spatial Compression
Removing or reordering information about field of color pixels to
conserve space
Neighboring pixels will have nearly the same brightness and color
values
Instead of sending the same number for each and every
sample, one number could be sent representing a block of sample
points in an area where the information content remain same
11. Spectral Compression
Human eye is much better in distinguishing luminance than
chrominance.
This can be used as advantage in conveying color information as
there is less precision required – a higher level of precision would
thus be „ saved‟.
Fewer samples required to convey color information – fewer
samples-> lesser bandwidth required
12. Steps in Image
Compression
If the color is represented in RGB mode, translate it to YCrCb mode
Divide the file into 8X8 blocks(group of 8 pixels = 1 block).
Transform the pixel information from the spatial domain into the
frequency domain with the Discrete Cosine Transform.
Quantize the resulting values by dividing each coefficient by an integer
value and rounding off to the nearest integer.
Look at the resulting coefficients in a zigzag order. Do a run-length
encoding of the coefficients ordered in this manner. Follow by Huffman
coding.
When conversion is done from RGB to YCrCb, it makes 4:4:4
format, but 2 color components are discarded. Thus bandwidth is
reduced by 50%
14. Temporal Compression
Redundancy between successive frames is known as temporal
compression.
Only changes from one frame to the next are encoded as often as
large number of pixels will be same on series of frames.
This type of compression relies on keyframes.
Keyframes stores still images which are used for frame
differencing.
15. Temporal Processing
Usually high frame rate: Significant temporal redundancy.
Possible representations along temporal dimension.
• Transform/subband methods
• Good for textbook case of constant velocity uniform global
motion.
• Inefficient for non uniform motion, i.e. real-world motion.
• Requires large number of frame stores.
• Leads to delay (Memory cost may also be an issue).
• Predictive methods
• Good performance using only 2 frame stores.
• However, simple frame differencing in not enough…
16. Video Coding and Frames
Goal: Exploit the temporal redundancy
Predict current frame based on previously coded frames
Three types of coded frames:
• I-frame: Intra-coded frame, coded independently of all
other frames
• P-frame: Predictively coded frame, coded based on
previously coded frame
• B-frame: Bi-directionally predicted frame, coded based on
both previous and future coded frames
18. Motion Compensated Prediction
Simple frame differencing fails when there is motion
Must account for motion
• Motion-compensated (MC) prediction
MC-prediction generally provides significant improvements
Questions:
• How can we estimate motion?
• How can we form MC-prediction?
19. Motion Estimation
Ideal situation:
• Partition video into moving objects
• Describe object motion
• Generally very difficult
Practical approach: Block-Matching Motion Estimation
• Partition each frame into blocks, e.g. 16x16 pixels
• Describe motion of each block
• No object identification required
• Good, robust performance
21. Block Matching ME Summary
Issues:
• Block size?
• Search range?
• Motion vector accuracy?
• Motion typically estimated only from luminance
Advantages:
• Good, robust performance for compression
• Resulting motion vector field is easy to represent (one MV per block)
and useful for compression
• Simple, periodic structure, easy VLSI implementations
Disadvantages:
• Assumes translational motion model Breaks down for more complex
motion
• Often produces blocking artifacts (OK for coding with Block DCT)
22. Basic VC Architecture
Exploiting the redundancies:
• Temporal: MC-prediction (P and B frames)
• Spatial: Block DCT
• Spectral: Color space conversion
Scalar quantization of DCT coefficients
Zigzag scanning, run length and Huffman coding of the nonzero
quantized DCT coefficients
25. Using Standards in Video Compression
Motivation for Standards
• Ensuring interoperability: Enabling communication between
devices made by different manufacturers
• Promoting a technology or industry
• Reducing costs
What do the Standards imply?
• Just the bitstream syntax and the decoding process(e.g. use
IDCT, but not how to implement the IDCT)
• Enables improved encoding & decoding strategies to be
employed in a standard-compatible manner
26. Current Video Compression Standards
STANDARD APPLICATION BIT RATE
JPEG Continuous-tone still-image Variable
compression
H.261 Video telephony and teleconferencing p x 64 kb/s
over ISDN
MPEG-1 Video on digital storage media (CD- 1.5 Mb/s
ROM)
MPEG-2 Digital Television > 2 Mb/s
H.263 Video telephony over PSTN < 33.6 kb/s
MPEG-4 Object-based coding, synthetic Variable
content, interactivity
H.264 From Low bitrate coding to HD Variable
encoding, HD-DVD, Surveillance,
Video conferencing.
27. Video Coding Standardization Organizations
Two key Organizations:
ITU-T (Video Coding Experts group, VCEG)
• International Telecommunications Union – Telecommunications
Standardization Sector (ITU-T, a United Nations
Organization, formerly CCITT)
ISO/IEC Moving Picture Experts Group (MPEG)
• International Standardization Organization and International
Electrotechnical Commission
28. Dynamics of Video Compression
Standardization
VCEG is older and more focused on conventional (esp. low-delay)
video coding goals (e.g. good compression and packet-loss/error
resilience)
MPEG is larger and takes on more ambitious goals (e.g. “object
oriented video”, “synthetic-natural hybrid coding”, and digital
cinema)
Sometimes the major organizations team up (e.g. ISO, IEC and
ITU teamed up for both MPEG-2 and JPEG)
contd…
29. Relatively little industry consortium activity (DV and organizations
that tweak the video coding standards in minor ways, such as
DVD, 3GPP, 3GPP2, SMPTE, IETF, etc.)
Growing activity for internet streaming media outside of formal
standardization (e.g., Microsoft, Real Networks, Quicktime)
30. MPEG-I
MPEG-1, the first lossy compression scheme developed by the MPEG
committee.
Used for CD-ROM video compression and as part of early Windows
Media players.
Uses DCT algorithm.
For slow moving frames e.g. In talk shows greater compression is
achieved.
For fast moving frames e.g. In sports channels lesser compression is
achieved.
The current wildly popular MP3 (MPEG-1, Part 3) audio standard is
actually the audio compression portion of the MPEG-1 standard and
provides about 10:1compression of audio files at reasonable quality.
31. MPEG-II
Evolved to meet the needs of compressing higher quality video.
used in today‟s video DVDs and digital broadcasts.
It uses bit rates typically ranging from 5 to 8 Mbits/s.
Uses DCT transforms but it also provides support for interlaced
video.
MPEG-2 is also the current standard for (HDTV) transmission.
includes additional color sub sampling, improved
compression, error correction and multi-channel extensions for
surround sound.
32. MPEG- III
MPEG-3 is the compression standard that never was.
Originally evolved to support HD content.
It turned out that MPEG- III could be done with minor changes to
MPEG-2. So MPEG-3 never happened.
Now there are Profiles of MPEG-2 that support HDTV as well as
Standard Definition Television(SDTV).
33. MPEG- IV
Goal: solve two video transport problems
• sending video over low-bandwidth channels
• achieving better compression than MPEG-2 for broadcast
signals.
The MPEG committee designed MPEG-4 to be a single standard
covering the entire digital media workflow from capture, authoring
and editing to encoding, distribution, playback and archiving.
Based on Apple Computer‟s QuickTime technology.
Used in a wide range of bit rates, from 64 Kbits/s to 1,800
Mbits/s.
35. H.264/AVC
H.264/MPEG4-AVC is a jointly developed standard by the ITU-T
Video Coding Experts Group (VCEG) and the ISO/IEC Moving
Picture Experts Group (MPEG) and has been standardized by the
ITU under the H.264 name.
H.264 uses techniques fairly different from MPEG-2 and can match
the best MPEG-2 quality at up to half the data rate.
Delivers excellent video quality across the entire bandwidth
spectrum from 3G to HDTV and everything in-between (from 40
Kbits/s to upwards of 10 Mbits/s).
The H.264 design incorporates a Video Coding Layer (VCL), and
network Abstraction Layer(NAL)
36. JPEG
JPEG stands for Joint photographic Experts Group.
It exploits the fact that the human eye will not notice small color
changes in an image.
Not a very good compression technique for full-color or grayscale
images.
37. AVI
Stands for Audio Video Interleaved.
Sound And Motion Picture File that conforms to the standards set
by Microsoft Windows Resource Interchange File Format (RIFF).
The video quality is good at smaller resolutions.
Only major drawback is that the files tend to be large.
To play an .avi, you could use Windows Media
Player, RealPlayer, or the DivX player.
38. .MOV
.mov is an Apple QuickTime motion video file format. Developed
by Apple Computer for viewing moving images.
This file extension identifies an Apple QuickTime movie.
.mov is a method of storing sound, graphics and movie files.
39. MJPEG
Short for Motion JPEG.
Best suited for broadcast resolution interlaced video, such as
NTSC or PAL.
Each video field is separately compressed into a JPEG image.
Also used for short files such as the short movies that can be
made by a digital camera.
Not good for movies that are smaller than TV resolutions and ill
suited for progressive scan computer monitors.
40. DivX
DivX is a software application that uses MPEG-4 standard to
compress digital video.
DivX Networks and the open source community are developing
DivX jointly.
41. Scope of Development
The MPEG committee continues to add video and graphics
standards, such as MPEG-7 and MPEG-21, to their standards
efforts.
Digital video technology has become a necessity due to the
increasing demand to include video data for personal use as well
as in the entertainment industry, the corporate world, the
government and defense.
Compression rates and the quality of data will continue to
improve, providing more efficient use of bandwidth, storage and
computing resources.
42. References and Further Reading
www.wikipedia.com
http://www.videomaker.com/article/10842/
http://www.h264encoder.com/
www.scribd.com
http://ivythesis.typepad.com/term_paper_topics/2009/08/video-compression-
techniques.html
http://drogo.cselt.stet.it/mpeg
J.G. Apostolopoulos and S.J. Wee, ``Video Compression Standards'', Wiley Encyclopedia
of Electrical and Electronics Engineering, John Wiley & Sons, Inc., New York, 1999.
V. Bhaskaranand K. Konstantinides, Image and Video Compression Standards:
Algorithms and Architectures, Boston, Massachusetts: KluwerAcademic
Publishers, 1997.
J.L. Mitchell, W.B. Pennebaker, C.E. Fogg, and D.J. LeGall, MPEG Video Compression
Standard, New York: Chapman & Hall, 1997.