Scribd
Upload
34 views16 pages

Chapter 5 Multi

The document discusses lossy compression algorithms. It introduces key concepts like distortion measures, rate-distortion theory, quantization, and transform coding. Quantization reduces the number of distinct values in a source to enable more efficient representation, and is central to lossy compression. Transform coding operates on blocks of samples rather than individual samples for improved efficiency.

Uploaded by

adu g
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
34 views16 pages

Chapter 5 Multi

The document discusses lossy compression algorithms. It introduces key concepts like distortion measures, rate-distortion theory, quantization, and transform coding. Quantization reduces the number of distinct values in a source to enable more efficient representation, and is central to lossy compression. Transform coding operates on blocks of samples rather than individual samples for improved efficiency.

Uploaded by

adu g
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

Chapter 5

Lossy Compression Algorithms

04/14/2024 1
Outline
1. Introduction
2. Distortion Measures
3. The Rate-Distortion Theory
4. Quantization
5. Transform Coding

04/14/2024 2
Introduction

• Hence, most multimedia compression algorithms are lossy.


• What is lossy compression ?
- It is the compression technique which will lose data in the original source
while trying to keep the visible quality at the almost same amount.
- It reduces non sensitive information to the human eyes and the compressed
media will not be the media that was available before compression.
- The compressed data is not the same as the original data, but a close
approximation of it.
- Yields a much higher compression ratio than that of lossless compression.

04/14/2024 3
Cont. …
 Advantages: Can reduce the file size more than in the Lossless
Compression
 Disadvantages: The original file cannot be taken after the decompression

Motivating Example

4 04/14/2024
Cont. …
The available bandwidth is limited!!
Only 16 bits available for the transmission of the data
Compression is necessary!!
Encode: drop the least significant bits
Encode data: 8*2-bits=16-bits

5 04/14/2024
Distortion Measures
 A distortion measure is a mathematical quantity that specifies how close an
approximation is to its original, using some distortion criteria.
 For example, if the reconstructed image is the same as original image except that it is
shifted to the right by one vertical scan line,
 an average human observer would have a hard time distinguishing it from the original
and would therefore conclude that the distortion is small.
 However, when the calculation is carried out numerically, we find a large distortion,
because of the large changes in individual pixels of the reconstructed image.
 The problem is that we need a measure of perceptual distortion, not a more naive
numerical approach.

04/14/2024 6
Cont. …

 The three most commonly used distortion measures in image compression are:
1. mean square error (MSE)
σ2, xn: input data sequence
yn : reconstructed data sequence
N : length of data sequence

04/14/2024 7
Cont. …

2. Signal to noise ratio (SNR), in decibel units (dB)


σ2
x : average square value of original data sequence
σ2
d : MSE

3. Peak signal to noise ratio (PSNR), in decibel units (dB) For 8 bit image
(video), xpeak = 255

04/14/2024 8
The Rate-Distortion Theory
 The rate distortion theory Lossy compression always involves a tradeoff between rate
and distortion.
 Rate is the average number of bits required to represent each source symbol.
 Within this framework, the tradeoff between rate and distortion is represented in the
form of a rate-distortion function R(D).
 Intuitively, for a given source and a given distortion measure, if D is a tolerable
amount of distortion, R(D) specifies the lowest rate at which the source data can be
encoded while keeping the distortion bounded above by D.

9 04/14/2024
Cont. …
 The rate-distortion function is meant to describe a fundamental limit for the
performance of a coding algorithm and so can be used to evaluate the performance of
different algorithms
 The below Figure shows a typical rate-distortion function.
 Notice that the minimum possible rate at D = 0, no loss, is the entropy of the source
data.
 The distortion corresponding to a rate R(D) ⇒ 0 is the maximum amount of distortion
incurred when “nothing” is coded

Typical Rate Distortion


Function.

10 04/14/2024
Quantization
• Quantization in some form is the heart of any lossy scheme.
• Without quantization, we would indeed be losing little information.
• The source we are interested in compressing may contain a large number of distinct
output values (or even infinite, if analog).
• To efficiently represent the source output, we have to reduce the number of distinct
values to a much smaller set, via quantization

04/14/2024 11
Cont. …
Types of Quantization
Uniform Scalar Quantization
• A uniform scalar quantizer partitions the domain of input values into equally spaced
intervals, except possibly at the two outer intervals.
• The endpoints of partition intervals are called the quantizer's decision boundaries.

- Midrise quantizers have even number of output levels.


- Midtread quantizers have odd number of output levels, including zero
as one of them
Nonuniform Scalar Quantization
• If the input source is not uniformly distributed, a uniform quantizer may be
inefficient.
• without having to increase the total number of decision levels, we can enlarge the
region in which the source is sparsely distributed.
• Such Nonuniform quantizer's thus have nonuniformly defined decision boundaries
04/14/2024 12
Cont. …
Vector Quantization
 According to Shannon's original work on information theory, any compression system
performs better if it operates on vectors or groups of samples rather than individual
symbols or samples.

 We can form vectors of input samples by concatenating a number of consecutive


samples into a single vector.
 For example, an input vector might be a segment of a speech sample, a group of
consecutive pixels in an image, or a chunk of data in any other format.

13 04/14/2024
Transform Coding
 From basic principles of information theory, we know that coding vectors is more
efficient than coding scalars
 To carry out such an intention, we need to group blocks of consecutive samples from
the source input into vectors
Discrete Cosine Transform (DCT)
 The Discrete Cosine Transform (DCT), a widely used transform coding technique, is
able to perform decorrelation of the input signal in a data-independent manner.
 Because of this, it has gained tremendous popularity.
 It helps separate the image into pars of different importance(with respect to the
image’s visual quality.

14 04/14/2024
Cont. …

Graphical Illustration of 8×8 2D DCT basis.

Let N=8
Each images in figures is 8*8 image hence there are 64 possible images, that represent 64 basic
functions of 8*8 images

04/14/2024 15 1
5
End of Chapter 5

Next Chapter 6
Image Compression Standards

16 04/14/2024

You might also like