You are on page 1of 14

Outline

Kinds of Coding
Need for Compression
Basic Types
Taxonomy
Performance Metrics

Two kinds of Coding of Information

Source Coding: Coding for efficient representation of


information. (Also called Compression)
Channel Coding: Coding for error free transmission of
information through a noisy medium

Need for Compression

Multimedia data need


Huge storage space
Large Bandwidth for transmission

Storage, Bandwidth Requirements are usually


much greater than availability
Compression is a viable technique

Multimedia Storage/Bandwidth
Requirements

Media type
Audio
CD-DA quality
Speech quality
Image
Gray scale (8bit) Color (24bit)
Video
NTSC
HDTV

Duration/Siz
e

Bandwidth

Storage

60 mins

1.412 Mbits/lsec.
64Kbits/sec.

635MB.
28.8MB.

640 480

60 mins.

~307KB.
~922KB.
221 Mbits/sec.

99.45 GB.
~922 GB.

Images and Video are usually:

Generated in Analog form


Captured in Analog form
Processed in Digital form
Stored in Analog/Digital form
Transmitted in Analog/Digital form

Advantages of Digital Media

Ease of processing
(for noise removal, quality enhancement, etc.
Robustness to error
(in storage/transmission)
Ease of integration of discrete media

A/D Conversion

Two major steps:


Sampling
Quantization

Compression Techniques

Two kinds of compression Lossless and Lossy


Compression techniques take advantage of redundancy in
digital images
Types of redundancies
Spatial redundancy: due to the correlation between
neighboring pixel values
Spectral redundancy: due to the correlation between
different color planes or spectral bands
Lossy techniques, in addition, take advantage of HVS
(Human Visual System) properties

Lossless and Lossy Compression

Lossless Compression:
No loss of information
Original data exactly recovered from compressed data
Lower compression ratio
Lossy Compression:
Loss of information
Perceptual loss of information reduced (controlled)
Higher compression ratio

Taxonomy of Image Compression

Lossless Coding techniques


(Entropy Coding)

Repetitive
Sequence
Encoding
RLE

Statistical
Encoding

Lossless
Predictive
Coding

Huffman
Arithmetic
LZW

DPCM

Bitplane
Encoding

Taxonomy of Image Compression

Lossy Coding techniques


(Source Coding)

Block
Lossy
Truncation Predictive
Coding
Coding
DPAM
ADPCM
Delta
modulation

Transform
Coding

Subband
Coding

DFT
DCT
Haar
Hadamard

Subbands
Wavelets

Fractal
Coding

Vector
quantization

Image Compression Standards


Standards

Possible Application Areas

CCITT T.4

Facsimile, Document Imaging

CCITT T.6

Facsimile, Document Imaging

JPEG

Photographic Imaging

JBIG

Facsimile, Document Imaging

CCITT H.261

Teleconferencing

MPEG-1

Video, Digital Storage Media(DSM)

MPEG-2

Video, HDTV, DSM

MPEG-4

Audio-visual Communications, Multimedia,


Remote sensing

Performance Metrics for Lossy


Techniques
Compression Ratio (CR)
CR

s iz e _ o f _ o r ig in a l _ d a ta
s iz e _ o f _ c o m p r e s s e d _ d a ta

Peak Signal to Noise Ratio (PSNR)


P S N R 2 0 lo g

10

p e a k _ d a ta _ v a lu e
RM SE

RMSE is the Root Mean Square Error between the original


and reconstructed data
Speed (Complexity) of encoding and decoding
Digital Image is a rectangular array of pixels (picture
elements)

Summary

Source Coding = Compression


Compression takes advantage of redundancies
Basic Types: Lossless and Lossy
Compression Ratio and Quality (Fidelity) Tradeoffs