Acknowledgements
We are deeply indebted to our supervisor [Link], [Link]., Associate
Professor, Department of Electronics and Communication Engineering for valuable guidance
constant encouragement constructive criticism and keen interest evinced throughout the
course of my mini project work. I am really fortunate to associate myself with such an
advising and helping guide in every possible way, at all stages, for the successful completion
of this mini project work.
We express our deep sense of gratitude to [Link], [Link].,Ph.D. Head of
the department of Electronics and Communication Engineering, for their valuable guidance
and constant encouragement given to me during this mini project work and the course
We express our gratitude to principal [Link], M.E., Ph.D.
and respectable director [Link], M.E., Ph.D. for supporting me in the
completion of mini project work successfully by providing the facilities. I am pleased to
express my heart full thanks to my FACULTY in the department of ECE of Sree
Vidyanikethan Engineering College for their moral support and good wishes.
Finally, we have a notion to express my sincere thanks to my friends and all those
who guided, inspired and helped me in the completion of mini project.
Project Associates
[Link]
[Link] KUMAR
[Link] KUMAR
[Link]
[Link]
[Link] KUMAR
i
Abstract
A new and efficient algorithm for high-density salt and pepper noise removal in images and
videos is proposed. The existing non-linear filter like Standard Median Filter (SMF),
Adaptive Median Filter (AMF), Decision Based Algorithm (DBA) and Robust Estimation
Algorithm (REA) shows better results at low and medium noise densities. At high noise
densities, their performance is poor. In this project, a novel decision-based filter, called the
multiple thresholds switching (MTS) filter using DBUTM Algorithm, is proposed to restore
images corrupted by salt-pepper impulse noise. The filter is based on a detection-estimation
strategy. The impulse detection algorithm is used before the filtering process, and therefore
only the noise-corrupted pixels are replaced with the estimated central noise-free ordered
mean value in the current filter window. The new impulse detector, which uses multiple
thresholds with multiple neighbourhood information of the signal in the filter window, is very
precise, while avoiding an undue increase in computational complexity. For impulse noise
suppression without smearing fine details and edges in the image, extensive experimental
results demonstrate that our scheme performs significantly better than many existing, well-
accepted decision-based methods.
ii
CONTENTS
ACKNOWLEDGEMENTS i
ABSTRACT ii
LIST OF FIGURES vii
Chapter 1 INTRODUCTION 01
1.1 Overview 01
1.2 Software Requirements 02
1.3 Outline of the project 03
Chapter 2 FUNDAMETALS OF DIGITAL IMAGE 04
PROCESSING
2.1 Fundamentals of Digital Image 04
2.2 Image 04
2.3 A Simple Image Model 04
2.4 Fundamentals of Image Processing 05
2.5 Image Compression 05
2.6 Image Compression model 07
2.7 Image Compression Type 08
2.7.1 Lossy Image Compression 08
2.7.2 Lossless Image Compression 08
Chapter 3 BASICS OF MATLAB 09
3.1 Introduction 09
3.2 The MATLAB System 10
3.2.1 Development Environment 10
3.2.2 The MATLAB Mathematical Function 10
Library
3.2.3 The MATLAB Language 10
3.2.4 Handle Graphics 10
3.2.5 The MATLAB Application program Interface 10
3.3 Develop Environment 11
3.3.1 Introduction 11
3.4 Quitting MATLAB 11
3.4.1 Starting MATLAB 11
3.4.2 Quitting MATLAB 11
3.5 MATLAB Desktop 12
3.6 Desktop Tools 12
3.6.1 Common Window 12
3.6.2 Common History 13
3.6.3 Help Browser 13
3.6.4 Current Directort Browser 13
3.6.5 Search Path 13
3.6.6 Workspace Browser 13
3.6.7 Array Editor 14
3.6.8 Editor/Debugger 14
3.7 Manipulating Matrices 15
3.7.1 Entering Matrices 15
3.7.2 Expressions 16
iii
3.7.3 Variables 16
3.7.4 Numbers 17
3.7.5 Operator 17
3.7.6 Functions 17
3.8 Graphic Use Interface 18
3.8.1 Creating GUI With Guide 19
3.8.2 GUI Development Environment 19
3.8.3 The Implementation Of GUI 19
3.8.4 Features Of The Guide-Generated 20
Application M-file
3.8.5 Beginning The Implementation Process 20
3.8.6 User Interface Controls
Chapter 4 NOISE 25
4.1 Introduction 25
4.2 Noise Generation 25
4.2.1 Brief Description 25
4.2.2 How it Works 26
4.2.3 Detector Noise 27
4.2.4 Salt and pepper Noise 27
4.2.5 Gaussion Noise 28
4.3 Median Filtering 29
4.3.1 Brief Description 30
4.3.2 How it Works 31
4.4 Mean Filter 31
4.4.1 Brief Description 31
4.4.2 How it Works 32
4.5 Crimmins Speckle Removal 32
4.5.1 Brief Description 32
4.5.2 How it Works 32
Chapter 5 IMAGE RESTORATION 35
5.1 Introduction 35
5.2 What is image restoring 36
5.2.1 Literature Survey 36
5.2.2 Requirement for Restoration 37
5.2.3 Blur Parameter 38
5.3.4 Types of Noise 39
Chapter 6 DBUTM FILTERING 41
6.1 Introduction 41
6.2 Shear Sort Algorithm 41
6.2.1 Correctness 43
6.2.2 Analysis 43
6.3 DBUTM Algorithm 44
6.3.1 Algorithm 45
6.4 Noise Filtering 49
Mat Lab Coding 49
Chapter 7 RESULTS 55
7.1 Noise Added Image 56
7.2 Apply Filtering 57
7.3 Restored Output 57
iv
Chapter 8 CONCLUSION 58
REFERENCES 59
List Of Figures
Figure No. Title Page No.
Figure 2.1 Block Diagram of Fundamental Sequence inDIP 07
System
2.2 Block Diagram of Image Compression 07
2.3 Block Diagram of Image Decompression 08
4.1 1D Gaussion Distribution With Mean 0 and Standard
deviation 1
4.2 Calculating the Median Value of Pixel
Neighbourhood
4.3 3x3 Averageing kernel often used in mean filtering
4.4 Crimmins Speckle Removal Algorithm
5.1 Blur Parameters
6.1 Original matrix before sorting
6.2 Step 1-Row sorting
6.3 Step 3- Column sorting
6.4 Step 4- Row sorting
v
6.5 Rows sorted in alternate direction
6.6 Block Diagram of DBUTM Filtering
6.7 Illustration for DBUTM
7.1 Results for 70% Noise corrupted Lena Image
7.2 Noise added image
7.3 Filtered Image
vi
LIST OF ABBREVIATIONS
DBUTM : Decisin Based Unsymmetric Trimmed Median Filter.
GUI : Graphic User Interface.
MSE : Mean Squared Error.
SMF : Standard Median Filter.
DBA : Decision Based Algorithm.
PSNR : Peak Signal to Noise Ratio.
MRI : Magnetic Resonance Imaging.
REA : Robust Estimation algorithm
MTS : Multiple Threshold switching
vii