HT TP: //qpa Pe R.W But .Ac .In: 2010-11 Image Processing

Name : ……………………………………………………………
htt
Roll No. : ……………………………………………..…………..
Invigilator’s Signature : ………………………………………..
CS/B.Tech (CSE)/SEM-7/CS-704G/2010-11
p:/
2010-11
IMAGE PROCESSING
Time Allotted : 3 Hours Full Marks : 70
/q
The figures in the margin indicate full marks.
Candidates are required to give their answers in their own words
pap
as far as practicable.
GROUP – A
( Objective Type Questions )
er.
1. State True or False for any ten of the following : 10 × 1 = 10
i) Quantization is a measure of no. of points with same

gray level.
wb
ii) An Nth power transformation of the input gray level

makes the image brighter.
iii) Histogram equalization is done to enhance brightness.

ut .
iv) A 2nd order operator is more potent than the 1st order
derivative operator as a sharpening spatial filter.
a c.
v) In DFT of an image noise represents low frequency

components.
vi) A median filter is a linear filter.

in
7405 [ Turn over

vii) Huffman coding is a technique for error free image

compression.
htt
viii) Hough transform is used to find boundary of a region.
ix) A highpass filter is called smoothing frequency domain

p:/
filter.
x) In a coloured image every pixel will have 5 gray levels

associated with it.
/q
xi) Fourier transform is linear.
xii) The gray level of 8 bit image is ranging from 0 to 256.

pap
GROUP – B
( Short Answer Type Questions )
Answer any three of the following. 3 × 5 = 15
2. Explain how to shrink an image.

er.
3. In histogram equalization, explain properties of the transform

that produces new gray level. Show for continuous PDF how
the previous transformation can be achieved to equalize
wb
histogram.
4. Explain the terms ‘adjacency’, ‘connectivity’ in the aspect of

digital images.
ut .
5. Write down the 2-dimensional DFT and inverse DFT

expressions of some function. Show that DFT and its inverse
are linear processes.
a c.
6. Write down the Hadamard transformation H1. Show how H2

is obtained from H1. What will be the Hadamard transform of
T
[1 2 3 4] by H2 ? Write the Hadamard matrix H3 .
! !
7. Write briefly about colour image processing.
in
!
! ! !
7405 2
GROUP – C
htt
( Long Answer Type Questions )
Answer any three of the following. 3 × 15 = 45
8. a) What is meant by histogram matching ? Explain how it

p:/
could be done. We have rk = zk , k = 0, 1, 2, 3.
Pr (ri ) = 0 "25, i = 0, 1, 2, 3. Pz (z0 ) = 0 "5, Pz (z1 ) = 0 "25, Pz (z2 ) = 0, Pz (z3 ) = 0 "25.
Show how can the histogram (r, Pr ) could be matched to
histogram (z, Pz ) . ! 7
/q
!
b) How does ! single
derivative and double derivative
perform as sharpening spatial filter ? Compare their
pap
!
performances. For the one dimensional image shown
below compute single and double derivative. Comment
on both the image and its single and double derivative. 6
5 5 4 3 2 1 0 0 0 6 0 0 0 0 1 3 1 0 0 0 0 7 7 7
er.
c) What is image sampling and quantization ? 2
9. a) Briefly explain ideal highpass filter. What are their

shortcomings ? How these shortcomings are overcome
wb
in Butterworth or Gaussian highpass filter ? 6
b) What conditions does distance measure between pixel

satisfy ? How are Euclidean, City Block and Chess
ut .
Board distances defined ? For the latter two write

matrices to elucidate. 5
c) A & B shown below are two images, find A/B, A – B,

a c.
A ∨ B, A ∧ B.
"7 6 3% "4 2 1%
$ ' $ '
A = $2 1 4' & B = $5 3 6' 4
$ ' $ '
#5 7 1& #7 1 2&
in
7405 3 [ Turn over

!
10. a) If f (x, y) & h (x, y) are two images and f (x, y) ✳ h (x, y)
is their convolution, then show
htt
f ( x, y) ! h ( x, y) " F(u, v) H (u, v) constitute a DFT pair
where F ( u, v ) & H ( u, v ) are the DFT of f ( x, y ) and
h ( x, y ) respectively. 5
! b) Let H (u, v) be a filter function and suppose H (u, v) = A .
p:/
So the net effect of filtering will be to multiply the image
by a constant. Explain using convolution theorem why
in spatial domain as well the image is multiplied by the
! same constant. ! 4
/q
c) Consider a binary source alphabet A = { 0, 1 }. Consider
a binary symmetric channel. Show that when there is
no possibility of channel error the capacity is maximum,
whereas when there is possibility of error = 1/2, no
pap
information can be transferred. 6

11. a) Discuss briefly about Huffman code with suitable
example. 6
b) Explain what is Haar transform ? Write down the
properties of Haar transform. 5
er.
c) Explain Hough transform and its use. 4

12. a) Explain how segmentation is done by region spliting
and merging. 5
b) Explain how a gradient operator detects edges. Discuss
wb
different kinds of gradient operators for detecting edge.

Explain function of compass operators by showing
compass operators for N, NW, W, SW, S, SE, E and
NE. 8
c) What is histogram ? 2
ut .
13. Write short notes on any three of the following : 3×5

a) Lossy compression
b) Fidelity criteria
c) Discrete cosine transform
a c.
d) Sharpening spatial filters

e) Least mean square filters.
in
7405 4

HT TP: //qpa Pe R.W But .Ac .In: 2010-11 Image Processing

Uploaded by

Copyright:

Available Formats

HT TP: //qpa Pe R.W But .Ac .In: 2010-11 Image Processing

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

HT TP: //qpa Pe R.W But .Ac .In: 2010-11 Image Processing

Uploaded by

Copyright:

Available Formats

Name : ……………………………………………………………

1. State True or False for any ten of the following : 10 × 1 = 10

i) Quantization is a measure of no. of points with same

ii) An Nth power transformation of the input gray level

iii) Histogram equalization is done to enhance brightness.

v) In DFT of an image noise represents low frequency

vi) A median filter is a linear filter.

7405 [ Turn over

vii) Huffman coding is a technique for error free image

ix) A highpass filter is called smoothing frequency domain

x) In a coloured image every pixel will have 5 gray levels

xii) The gray level of 8 bit image is ranging from 0 to 256.

2. Explain how to shrink an image.

3. In histogram equalization, explain properties of the transform

4. Explain the terms ‘adjacency’, ‘connectivity’ in the aspect of

5. Write down the 2-dimensional DFT and inverse DFT

6. Write down the Hadamard transformation H1. Show how H2

8. a) What is meant by histogram matching ? Explain how it

c) What is image sampling and quantization ? 2

9. a) Briefly explain ideal highpass filter. What are their

in Butterworth or Gaussian highpass filter ? 6

b) What conditions does distance measure between pixel

Board distances defined ? For the latter two write

c) A & B shown below are two images, find A/B, A – B,

7405 3 [ Turn over

information can be transferred. 6

c) Explain Hough transform and its use. 4

different kinds of gradient operators for detecting edge.

13. Write short notes on any three of the following : 3×5

d) Sharpening spatial filters

You might also like