See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/310133767

Comparison between Fourier transform And Wavelet transform in signal

compression

Chapter · June 2010

CITATIONS READS

0 2,384

1 author:

Najmuldeen Hashim
Universiti Kebangsaan Malaysia
2 PUBLICATIONS   0 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Najmuldeen Hashim on 14 November 2016.

The user has requested enhancement of the downloaded file.

Chapter one Introduction

Chapter One
Introduction

1.1) General Consideration

Signals convey information. Systems transform signals. We gain this
understanding by dissecting their structure (their syntax) and by examining their
interpretation (their semantics). For systems, we look at the relationship between
the input and output signals (this relationship is a declarative description of the
system) and the procedure for converting an input signal into an output signal
[1].
A Sound is a signal, one can show how a sound can be usefully
decomposed into components that themselves have meaning. Musical chord, for
example, can be decomposed into a set of notes. An image is a signal. We do not
discuss the biophysics of visual perception, but instead show that an image can
be usefully decomposed. We can use such decomposition, for example, to
examine what it means for an image to be sharp or blurred, and thus to
determine how to blur or sharpen an image.
Signals can be more abstract (less physical) than sound or images. They
can be, for example, a sequence of commands or a list of names. We develop
models for such signals and the systems that operate on them, such as a system
that interprets a sequence of commands from a musician and produces a sound
[1].
Mathematically, we model both signals and systems as functions. A
signal is a function that maps a domain, often time or space, into a range, often a
physical measure such as air pressure or light intensity. A system is a function
that maps signals from its domain into the output signals [1].

~1~
Chapter one Introduction

One of the principal uses of compression is the control of level in vocals.

Many singers train for years to achieve the degree of breath control necessary
for an even tone and expressive performance. Other vocalists rely on an
instinctive voice production technique, which may need help in the studio to
maintain a consistent level, and result in a vocal track which sits correctly in the
mix. The unprocessed signal has a large dynamic range between the highest and
lowest levels. Applying compression reduces the highest levels, reducing the
dynamic range, make-up gain is addenda to restore the original high level. The
result is a much more controlled and usable sound [2].

The basic Wavelet Transform is similar to the well known Fourier

Transform. Like the Fourier Transform, the coefficients are calculated by an
inner-product of the input signal with a set of orthonormal basis functions (this
is a small subset of all available wavelet transforms). The difference comes in
the way these functions are constructed, and more importantly in the types of
analyses they allow [3].

The key difference is that the Wavelet Transform is a multi-resolution

transform, that is, it allows a form of time–frequency analysis (or translation–
scale in wavelet speak). When using the Fourier Transform the result is a very
precise analysis of the frequencies contained in the signal, but no information on
when those frequencies occurred. In the wavelet transform we get information
about when certain features occurred, and about the scale characteristics of the
signal. Scale is analogous to frequency, and is a measure The most basic
definition of a wavelet is simply a function with a well defined temporal [3].

~2~
Chapter one Introduction

1.2) project planning

Input Using Dimension Compressed

reduction Signal
signal N Transform N (compression) less
samples samples samples

compressing the signal

Inverse
Compressed Transform
Original signal
signal

Reconstruction
The Block diagram of the compression process

The work of each block:

1) The Input signal block :
any input signal (Biomedical signal , Sound , Temperature ,..)
2) The using transform block:
F.T(Fourier Transform) or W.T(Wavelet Transform)
3) The dimension Reduction block:
Since the information concentrated in the low frequency components, the low
frequency components will be taken and the high frequency components will
be negligted. The number of samples resulting from the transformation
process will be reduced.
4) The inverse tranform block:
Reconstraction of the original signal from the compressed signal.

~3~
Chapter one Introduction

1.3) Compression
The compression of the signal is reducing the number of samples during
certain period.

1.3.1) Compression Process

The goal of any compression scheme is to obtain a parsimonious
representation of signals. The physicians are accepted the new compression
scheme, the signal reconstructed from the parsimonious representation must be a
very exact replica of the original because physicians regard every wave in the
signal as potentially important indicators.
Therefore, the signal compression scheme must have a good compression ratio
and yet maintain the reconstruction quality. The compression ratio measures
ability of the scheme to derive a sparse representation whereas the
reconstruction quality indicates the fidelity of the reconstructed signal to the
original signal.[4]

1.3.2) Types of compression

Types of compression can be divided into two categories, as Lossless and

Lossy compression. In lossless compression, the reconstructed signal after
compression is numerically to the original signal. In lossy compression scheme,
the reconstructed signal contains degradation relative to the original.
lossless compression, the reconstructed image after compression is
numerically to the original image. In lossy compression scheme, the
reconstructed image contains degradation relative to the original.In the case of
video, compression causes some information to be lost; some information at a

~4~
Chapter one Introduction

detail level is considered not essential for a reasonable reproduction of the

scene. This type of compression is called lossy compression.
Audio compression on the other hand, is not lossy, it is called lossless
compression. An important design consideration in an algorithm that causes
permanent loss of information is the impact of this loss in the futureuse of the
stored data. Lossy technique causes signal quality degradation in each
compression/decompression step. Careful consideration of the human visual
perception ensures that the degradation is often unrecognizable, though this
depends on the selected compression ratio. In general, lossy techniques provide
far greater compression ratios than lossless techniques. The following are the
some of the lossless and lossy data compression techniques [4].

1.3.3) Lossless coding techniques

 Run length encoding
 Huffman encoding
 Arithmetic encoding
 Entropy coding
 Area coding

1.3.4) Lossy coding techniques

 Predictive coding
 Transform coding (FT/DCT/Wavelets)

~5~
Chapter one Introduction

1.4) Transformation techniques

First of all, why do we need a transform, or what is a transform anyway?
Mathematical transformations are applied to signals to obtain a further
information from that signal that is not readily available in the time domain
signal. There are number of transformations that can be applied, among which
the Fourier transforms are probably by far the most popular but not alwayes the
best.

Why do we need the frequency information?

Often times, the information that cannot be readily seen in the time-domain can
be seen in the frequency domain.
Let's give an example from biological signals. Suppose we are looking at an
ECG signal (ElectroCardioGraphy, graphical recording of heart's electrical
activity). The typical shape of a healthy ECG signal is well known to
cardiologists. Any significant deviation from that shape is usually considered to
be a symptom of a pathological condition.
This pathological condition, however, may not always be quite obvious in the
original time-domain signal. Cardiologists usually use the time-domain ECG
signals which are recorded on strip-charts to analyze ECG signals. Recently, the
new computerized ECG recorders/analyzers also utilize the frequency
information to decide whether a pathological condition exists. A pathological
condition can sometimes be diagnosed more easily when the frequency content
of the signal is analyzed.
1.4.1) Introduction to Fourier Transform
The Fourier transform move a data sample from the time domain to the
frequency domain. This allows a seemingly random sample to be broken into its
major frequencies. It is used in all kinds of signal processing.

~6~
Chapter one Introduction

As the Fourier Transform is widely used in analyzing and interpreting

signals and images, we will first have a survey on it prior to going further to the
Wavelet Transform. J. Fourier discovered in the early 18 century that it is
possible to compose a signal by superposing a series of sine and cosine
functions (Fourier Transform). These sine and cosine functions are known as
basic functions (see Figure 2-1) and are mutually orthogonal. The transform
decomposes the signal into the basic functions, which means that it determines
the contribution of each basis function in the structure of the original signal.
These individual contributions are called the (Fourier) coefficients.
Reconstruction of the original signal from its Fourier coefficients is
accomplished by multiplying each basic function with its corresponding
coefficient and adding them up together, i.e. a linear superposition of the basic
functions.
1.4.2) Introduction to Discrete Fourier Transform (DFT)
DFT is an estimation of the Fourier Transform, which uses a finite
number of sample points of the original signal to estimate the Fourier Transform
of it. The order of computation cost for the DFT is in order of n 2 where n is the
length of the signal. Fast Fourier Transform (FFT) is an efficient implementation
of the Discrete Fourier Transform, which can be applied to the signal if the
samples are uniformly spaced. FFT reduces the computation complexity by
taking advantage of self similarity properties of the DFT [4]. If the input is a
non-periodic signal, the superposition of the periodic basic functions does not
accurately represent the signal. One way to overcome this problem is to extend
the signal at both ends to make it periodic. Another solution is to use Windowed
Fourier Transform (WFT). In this method the signal is multiplied with a window
function (see Figure 2-2) prior to applying the Fourier transform. The window
function localizes the signal in time by putting the emphasis in the middle of the
window and attenuating the signal to zero towards both ends [5].

~7~
Chapter one Introduction

1.4.3) Introduction to Wavelet Transform

The wavelet transform is a method of converting a function(or signal)in to
anther from which makes certain features of the original signal more amenable.
Wavelet transforms allow a signal to be decomposed such that frequency
characteristics and the location of particular features in a time series may be
highlighted simultaneously [5].
Plotting the wavelet transform allows a picture to be built up of the correlation
between the wavelet at various scales or locations and the signal. In subsequent
chapters we will cover the wavelet transform in more mathematical detail.

1.4.4) Introduction to Discrete Wavelet Transform

The DWT of a signal x is calculated by passing it through a series of
filters. First the samples are passed through a low pass filter with impulse
response g resulting in a convolution of the two, as in equation bellow [6]:

y[n]  ( x  g )   x[k ]g[n  k ]
k  
............ (1.1)

The signal is also decomposed simultaneously using a high-pass filter h.

The outputs give the detail coefficients (from the high-pass filter) and
approximation coefficients (from the low-pass). It is important that the two
filters are related to each other and they are known as a quadrature mirror filter.

~8~
Chapter one Introduction

However, since half the frequencies of the signal have now been removed, half
the samples can be discarding according to Nyquist’s rule.
For many signals, the low-frequency content is the most important part,
which gives the signal identity. The high-frequency content, on the other hand,
imparts flavor or nuance. Consider the human voice, if you remove the high-
frequency components, the voice sounds different, but you can still tell what's
being said. However, if you remove enough of the low-frequency components,
you hear gibberish.

1.5) Literature Survey

 A New Algorithm for Voice Signal Compression (VSC) , [2009]
Analysis the signal to be suitable for Limited Storage Devices
Using MatLab (International Journal of Computer and Electrical
Engineering, Vol. 1, No. 5 December).
 Yasir A.Ahmed AL.Obaidi [2009].design Wavelet Based Image
Compression Using VHDL.
 Ali H.A Al.Tememy [2005].design Heart Disease Diagnosis Using
Discrete Wavelet Transform and Neural Network. This thesis shows an
approach for ECG signal processing based on Artificial Neural Networks
(ANN) and transform domains (Discrete Wavelet Transform (DWT) and
Fourier Transform (FT)). In part of this work, the signal compression is
done to make it suitable to the neural network.
 Quality-Controlled [2005] Compression Method using Wavelet
Transform for Electrocardiogram Signals (International Journal of
Biological and Life Sciences).
 Z. Lu, D.Y. Kim, W.A. Pearlman, Jan 2000, Wavelet Compression
of Signals by Set Partitioning in Hierarchical Trees (SPIHT), IEEE
Transactions on Biomedical Engineering.

~9~
 Chapter one Introduction

1.6) The aim of work

 Design a compression scheme for any signal using wavelet transforms.

 Quantify and compare the performance of the designed scheme with the
compression technique using Fourier Transform, This comparison must
be made to choose the best transformation method.
 Also comparison must be made between different types of wavelet
transform.
 Implement the designed scheme as a matlab program.

1.7) project layout

This project consists of five chapters:

 Chapter one: describes an introduction to compression applications,
Fourier transform, wavelets, literature survey and aim of the work.
 Chapter two: contains the Fourier transform and Wavelet transform
theory.
 Chapter three: describes in details the algorithms and software design of
compression process.
 Chapter four: presents the results of FFT and DWT application on
different types of signals to perform the compression process.
 Chapter five: summarizes the main conclusions of the work presented in
this project, and gives suggestions for further work.

~ 10 ~

View publication stats