Acknow Ledgeme NT
Acknow Ledgeme NT
Acknow Ledgeme NT
ACKNOW LEDGEME NT
I would like to thank God for not letting me down at the time of
crisis and showing methe silver lining in the dark clouds.I wish to express my deep gratitude to Mr. M.D.Singh, Sr.Lecturer,
Electrical &Instrumentation Engineering Department for providing his uncanny guidance and supportthroughou t the preparation
of the thesis report.I am also thankful to Dr. Smarajit Ghosh, Professor & Head, Electrical & InstrumentationE ngineering
Department, for the motivation and inspiration that triggered me for thethesis work.I would also like to mention a special thank for Dr. R.K.Sharma,
Dean of AcademicAffairs and Mr. Sunil Kumar Singla, Sr. Lecturer, Electrical & InstrumentationE ngineering Department for
providing their untimely support.I would also like to thank all the staff members and my co-students who were alwaysthere at the need of the
hour and provided with all the help and facilities, which I requiredfor the completion of the thesis
ABSTRAC T
Historically, the field of image processing grew from electrical engineering as anextension of
the signal processing branch. The massive amount of data required forimages is a primary reason for the development of
many sub areas within the field of computer imaging such as image segmentation and compression.Wha tever may be the way of
transmission, the data tends to get noisy and thereby thefurther processing does not lead to good results. Hence, it is very essential to keep thedata
close to originality.The prime focus of this thesis is related to the pre processing of an image. The preprocessing being worked
upon is the de noising of images. In order to achieve this interms of the concerned work, wavelet transforms have been applied:
Discrete wavelettransform and Un decimate d Discrete wavelet transform .In this thesis, a new thresholding technique has been presented
alongwith the standardthreshold ing techniques like soft and hard thresholding. And a comparative analysis of different combinations of
the suggested threshold values and thresholding techniques hasbeen carried out very efficiently. A new constraint, of either
thresholding the low passcomponents or keeping them as such before applying the inverse DWT and UDWT, hasalso been added. This
has been done in order to find more possible combinations that canlead to the best denoising technique.MATL AB codes have been developed
TABLE OF CONTENT S
ContentsPage No.
CERTIFICATE i ACKNOWLED GEMENT ii ABSTRACT iii
LITERATURE S URVEY 4CHAP TER 3. WAVEL ET TRANSFOR MS 123.1 Introdu ction to Wavelet 123.2 Mathemati cal Representatio n of Wavelet 133.
3 Translation and Scale in WT 143. 4 MultiResolution Analy sis in WT 153.5 Properties of Wa velet 163.6 Types of Wavelet Trans forms 163.6.1 De
composition Proc ess 173.6.2 Comp osition Process 1 83.6.3 Pyramidal Decomposition 1 93.6.4 Wavelet P acket Decomposit ion 203.7 Undeci mated Wavelet T
ransform 213.8 Wavelet Families 223.9 Wavelet D omain Advantage s 23CHAPTER 4 . IMAGE DENO ISING 244.1 Intr oduction 24
v4.2 Wavelet De noising 244.3 Im age Types 264.4 I mage File Format s 274.5 Wavelet I mage Denoising Using Custom Th resholding 28CH APTER 5. THE
EXPLICIT ALG ORITHM 29CH APTER 6. EXPE RIMENTAL RE SULTS 40CHAP TER 7. DISCUS SION OF RESU LTS 92CHAPTE R 8. CONCLUSI
ON AND FUTU RE SCOPE 1008 .1 Conclusion 10 08.2 Future Scop e 102 APPENDIX A 103 APPENDIX B 106
DWT Discrete W avelet Transform UDWT Undecim ated Discrete Wa velet TransformP SNR Peak Signal to Noise RatioM AD Median Abso lute DeviationST
D Standard Devia tionUNIV Univer sal ThresholdMS MAD SoftSS ST D SoftUS UNIV SoftMH MAD H ardSH STD Hard UH UNIV Hard MC MAD Custo