UNIT – III IMAGE RESTORTATION
SESSION – 1
1. Recap –Image enhancement – Pictorial Quiz. To conduct this activity, we can
divide the class into four groups. Different images can be shown to learners.
One minute of time can be given to the learners for internal discussion and
answer the question.
We can show the following images to the learners:
1. First image shown is the test image which has been corrupted as shown in
second image. Identify the noise.
2. Identify the noise which has the probability density function shown below
3. If the histogram of an image is as shown below, identify the noise that has
been added to the image
4. Identify the noise which has this skewed PDF
5. If the histogram of an image is as shown below, identify the noise that has
been added to the image.
1. Impulse Noise
2. Uniform Noise
3. Exponential Noise
4. Gamma Noise
5. Rayleigh Noise
2. Introduction to image restoration – Brainstorming –
List of ideas contributed by the students on Image Restoration is summarized
by the facilitator.
3. Comparison of image enhancement and image restoration – Power Point
Presentation
www.comp.dit.ie/bmacnamee/.../dip/.../ImageProcessing1-Introduction.p...
4. Introduction to degradation function – Power Point Presentation
www-ee.uta.edu/dip/Courses/EE5351/Image_Restoration.ppt
5. Conclusion –
Learner led presentation. Summary of the session can be asked to present by a
learner.
SESSION – 2
1. Recap – Question and Answer
Compare image enhancement and image restoration
What are the different degradation models?
What is degradation?
What is circulantmatrix?
2. Degradation model for continuous function – Chalk and Talk –
Expression for continuous degradation
3. Degradation model for discrete function – Chalk and Talk –
Expression for discrete function
4.Conclusion and Summary – Questions and Answers
What is circulant matrix?
Write the equation for continuous function
Write the equation for discrete function in two dimensions.
SESSION – 3
1. Recap –Brainstorming –
Brainstorming is carried out by posing different questions
What is degradation?
Write the degradation equation for discrete function
2. Algebraic approach of restoration – Presentation slides
users.rowan.edu/~shreek/fall09/dip/lectures/lecture6.ppt
web.eecs.utk.edu/~qi/ece472-572/lecture08_restoration_deblur.ppt
3. Unconstrained restoration:Chalk and Talk Presentation. The derivation
and characteristics of unconstrained restoration is explained.
4. Conclusion and Summary – Summarization by Learners – We can conclude
by summarizing the whole session – Algebraic approach is used to estimate the
image degradation,Seeking of approximate value of the original image.
SESSION – 4
1. Recap – Quiz –
The entire class can be divided into different groups. May be each now is
assumed as one group. Sample questions
What is unconstrained restoration?
List the algebraic approaches used for restoration?
Give the equation for continous and discrete degradation function?
2. Language multiplier – PPT Chalk and Talk
The derivation and characteristics of Language multiplieris explained.
www.tcc.edu/VML/Mth163/documents/LagrangeMultipliersNew2.ppt
www2.units.it/ramponi/teaching/DIP/materiale/dip05.pdf
3. Conclusion and Summary – Summarization by learners
We introduce a computational method, based on algorithm of Lagrange multi-
pliers, to restore an image that has been blurred by uniform linear motion. We
are motivated by the problem of restoring blurry images via well developed
mathematical methods and techniques based on the Lagrange multipliers in
order to obtain an approximation of the original image.
SESSION – 5
1. Recap – Questions and Answers
What is Lagrange multiplier?
Why Lagrange multiplier is used in image restoration?
2. Constrained restoration Chalk and Talk Presentation
The derivation and characteristics of Constrained restorationis explained.
csce.uark.edu/~jgauch/5683/notes/ch05c.pdf
math.ewha.ac.kr/~jylee/SciComp/dip-diml.yonsei/chap5-1.pdf
3. Conclusion and Summary – Board activity
Write the matrix notation for smoothing.
Write the procedure for constrain least square restoration.
SESSION – 6
1. Recap – Recall by key word
Unconstraint restoration, constraint restoration, smoothing matrix, procedure
for constraint least square filter
2. Inverse filtering formulation – Chalk and Talk Presentation
The derivation of Inverse filtering is explained and its importance is stated.
www-ee.uta.edu/dip/Courses/EE5351/Image_Restoration.ppt
3. Removal of blur caused by uniform linear motion – Chalk and Talk
Presentation.
The derivation of Removal of blur caused by uniform linear motionis explained
and its importance is stated.
faculty.petra.ac.id/resmana/private/pcd/presentation/lecture6.ppt
4. Conclusion and Summary – Summarization by learners
Recall the degradation model:
Given H(u,v), one may directly estimate the original image
At (u,v) where H(u,v) = 0, the noise N(u,v) term will be amplified
SESSION – 7
1. Recap – Questions and Answers
How will you remove the blur caused by uniform motion?
Write the expression for inverse filtering
What are the disadvantages of inverse filtering?
Define inverse filtering.
What is the advantage of inverse filtering?
2. Least mean square (WIENER) filter Chalk and Talk Presentation
www2.units.it/ramponi/teaching/DIP/materiale/dip05.pdf
www.postech.ac.kr/~seungjin/courses/ml/handouts/handout15_4pp.pdf
3. Conclusion and Summary – Summarization by learners
Iterative wiener filter is an effective method to estimate the power spectral
density of the original image.
The mean square error decreases with the number of iterations increasing
until it converges.
SESSION – 8
1. Recap – Review – Assumptions of Wiener filter
The original image and noise are stoically independent
The power spectral density of the original image and noise are known.
Both the original image and noise are zero mean.
2. Geometric Transformation – Chalk and Talk Presentation
www.postech.ac.kr/~seungjin/courses/ml/handouts/handout15_4pp.pdf
jc-schools.net/ppt/GeometricTransformations.ppt
3. Conclusion and Summary – Question and Answer
What is geometric transformation?
What are the types if geometric transformation?
SESSION – 9
1. Recap – See and Tell
a. Original image
b. Distorted image using bilinear transform
c. Difference between a and B
d. Geometrically restored image using bilinear transform for gray level
interpolation
2. Spatial Transformation Chalk and Talk Presentation
www.cs.bgu.ac.il/.../Intensity%20Transformation%20and%20Spatial
%20...www.csie.ntnu.edu.tw/~violet/IP93/Chapter03.ppt
3. Gray level Interpolation Chalk and Talk Presentation
elect.eng.ankara.edu.tr/courses/.../Lec-4%20Restoration-Deconvolution.p...
4. Conclusion and Summary – Cross word
Across
4. Rubber sheet transformation
7. Salt and pepper noise
1. The filter removes salt noise only
2. One of the type of geometric transformation
3. Removes both salt and pepper noise
Down
1. Reconstruction of an image
2. Minimum mean square error filter
3. Filter that smoothes local degradation of image’
4. Falls to handle noise
