Image-Uni Q Papers

November - 2004Time : 3 hours Total Marks : 100

Information Technology

201. Justify/Contradict following statements .

(a) Enhancement process does not change the information content of image.

(b) For digital image having salt pepper noise, median filter is the best filter.

(c) For continuous image histogram can be perfectly equalized, but it may not be so for the digital image.

(d) The statement that "Run-length coding gives data compression" is not always true.

(e) Quality of picture depends on the number of pixels and grey level that represent the picture.

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Figures to the right indicate full marks.(4) Assume suitable data if necessary.

6

8

6

8

2. Let f (n), f (n), f (n), ... f (n) be a set of N orthogonal sequences each of length N.0 1 2 N-12(a) Show that N orthogonal patten of size N N can be generated by using these

sequences. Prove the orthogonality of these patterns.

(b) Write four Walsh sequence of length N = 4, using these sequences generate sixteen orthogonal Walsh pattern.

(c) For the following 4 4 image show that any four patterns from the above sixteen pattern are required to fully represent the image. Identify these pattern :

´

´

3. (a) What is the impulse response of each filter ?

(i) Transfer function is :-1 -1 -1 -1 -1H (z , z ) = 1 - a z - a z - a z z - a z , z .1 1 2 1 2 3 1 2 4 1 2

(ii) Frequency response is

H(w , w ) = 1 - 2 a cos w - 2 a cos w .1 2 1 2

4

4

10

10

4

4

10

10

8

8

12

12

8

8

12

12

UNIVERSITY QUESTION PAPERSUNIVERSITY QUESTION PAPERS

UQP - (Information Technology) - 1

4

6

7

4

4

7

4

(b) (i) Evaluate minimum number of average bits sequence per symbol.

(ii) What is coding efficiency for the code you have obtain in Q.4(a).

(iii) What is sum-length coding ? Explain clearly.

(b) (i) Discuss thining algorithms for finding the skeleton of a region. Apply it to a square and a triangle.

(ii) Illustrate opening and closing operations with suitable example.

(b) (i) Show that one dimensional FFT algorithm can be used to compute two dimensional DFT of an digital image.

(ii) Using 4-point FFT algorithms evaluate 20 DFT of following image.

0

1

2

1

1

0

1

2

2

1

0

1

1

2

1

0

10

6

5. (a) Explain in detail enhancement techniques in special domain used for images.

6. (a) Write an expression for a two dimensional DFT. What is its relationship with one dimensional DFT ?

207. Write short notes on following :

(a) Chain code.

(b) Fourier descriptor.

(c) Hit-or miss transform.

(d) Uniform sampling and non uniform sampling.

(e) Dilation and Erosion.**********

84. (a) Given below is the table of eight symbol and their frequency of occurrence. Give Huffman code for each eight symbols.

Symbol

Frequency

s1 s2 s3 s4 s5 s6 s7s0

0.15 0.06 0.08 0.21 0.14 0.07 0.040.25

2

3 2

3 2-1 1

1

x uy = 0 u = 0

N - 1 N - 1

31

2 2| f(x, y) | = | F(u, u) | S SS S

A = U =

8

6

(b) (i) For the 2 2 transform A and the image U :´

(ii) Prove two dimensional energy. Conservation relation

Calculate the transformed image u and the basis image.

where f(x, y) represent 2-D image signal and F(u, u) represent transformed domain signal.


1. (a) Write DDA Algorithm to draw a line, calculate all points using the same algorithm to draw line between points x(10, 10) and y(20, 15).

June - 2005Time : 3 hours Total Marks : 100

10

N.B. : (1) Attempt any five questions.(2) Figures to the right indicate full marks.(3) Assume any suitable data wherever required but justify the same.

(b) Show that the first order difference of a chain code normalize it to rotation. Compute the first difference of code 0101030303323232212111.

10

2. (a) Define : (i) Eucledean distance (ii) City block distance (iii) Chess board distance

Consider the following subimage; determine

(i) Eucledean distance

(ii) City block distance

(iii) Chess-board distance between p and q

(b) Explain with suitable example region splitting and merging technique for image segmentation.

3. (a) For the 2 ´ 2 transform A and the image U

Calculate the transformed image V and the basis image.

(b) Explain with suitable example morphological filter to eliminate background noise.

4. (a) Define the following spatial filters

(i) Median (ii) Min. (iii)Max.

Consider the digital image

Calculate the value at point g(2, 2) = 7 for above filters.

1 2 3 4 5 6 7 8

1

2

3

4

5

6

7

8

9

10

q

p

10

10

10

2

1 1

1 4-1 3

1 21A = , U =

10

0

2

1

1

2

1

0

1

0

5

0

1

7

6

6

6

6

5

6

7

7

5

6

5

6

10



Compute the grey level histogram of the output image obtained by enhancing the input by histogram equalization technique.

7. Write short notes any four :

(i) Image sampling (iv) Fourier Descriptor

(ii) Walsh transform (v) Bit plane coding

(iii) Psychovisual redundancy (vi) Butterworth filter.

Frequency 0 26 39 16 84 20 00

20

**********

(b) Draw and explain block diagram of JPEG encoder and decoder.

5. (a) Show that a high pass filtered image can be obtained in spatial domain as

Highpass = Original - Lowpass

10

10

(b) Suppose that a digital image is subjected to histogram equalization. Show that a second pass of histogram equalization will produce exactly the same result as the first pass.

6. (a) Draw and explain the 2-D fast wavelet transform filterbank.

(b) Grey level histogram of an image is given below :

Grey level 1 2 3 4 5 6 70

10

10

10

1. (a) Write DDA Algorithm to draw a line. Also explain the basic transformations in computer Graphics.

(b) Chain codes can be made invariant to translation and rotation. Explain. Compute the first difference of code 011103022232212111.

2. (a) Develop an algorithm for converting one pixel thick 8-connected path to 4-connected path.

(b) Suppose, a 8 ´ 8 pixel image represented by 4 bits/pixel has the following grey level distribution :

Modify above histogram such that the desired distribution is as follower,

December - 2005Time : 3 hours Total Marks : 100

10

N.B. : (1) Question No. 1 is compulsory.(2) Answer any four out of remaining six questions.(3) Figures to the right indicate full marks.(4) Assume suitable data if necessary.

Grey level (nx)

No. of pixels P(nx)

0

0

1/7

1/7

2/7

2/7

3/7

3/7

4/7

4/7

5/7

5/7

6/7

6/7

1

1

4

2

16

6

17

8

11

10

8

12

4

11

4

9

0

6

Grey level (ng)

No. of pixels P(ny)

10

10

10



(b) Explain in detail enhancement techniques in spatial domain used for images.

5. (a) Explain with suitable example region splitting and merging technique for image segmentation.

(b) Write an expression for a two dimensional DCT. What is its relationship with one dimensional DCT ? Find DCT for the given image.

6. (a) Explain the filter in spatial domain :

(i) Low pass filter (iv) Butterworth filter

(ii) High pass filter (v) Gaussian filter.

(iii) Median filter

(b) Explain in detail, what is image pyramid and sub-band coding.

7. Write short notes on following :-

(a) Hit or Miss transform

(b) Polygonal approximation and signatures

(c) Huffman coding and Run length coding

(d) Hough Transform.

2

1

1

2

0

1

0

1

1

0

0

2

0

1

1

3

10

10

10

10

10

20

**********

3. (a) Obtain Hadamard matrix H(8) from a simple second order matrix H(2), and draw flow diagram for H(8).

(i) Define Zequency and find it for H(8).

(ii) Obtain the watch matrix w(8) from H(8).

(b) Explain the basic principle of transform coding for image compression and illustrate same with the help of DFT and DCT.

4. (a) Explain dilation, erosion, opening and closing morphological operations. Also

erode the following region of 1's. Using a 3 ´ 3 square operator.

10

10

10

1

1

1

1

1

1

1

1

1

1

1

1

1

1

0

1

1

0

1

1

1

1

1

1

1

1

1

June - 2006Time : 3 hours Total Marks : 100


N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Assume suitable data wherever necessary and state it.

1. (a) What do you mean by sampling ? What are the different types of sampling ?

(b) Explain the terms Brightness adaptation and discrimination with reference to human visual system.

6

8


4. Write short notes on the following :-

(a) Redundancies in digital images

(b) Fidelity criteria

(c) Homomorphic filter

(d) Texture analysis.

5. (a) Histogram of a digital image with 8 quantization levels is given below. Perform Histogram Equalization. Derive the transformation function and the new histogram and draw them.

(b) Explain with suitable example the LZW coding and decoding technique for image compression.

6. (a) What is segmentation ? Explain the method of segmentation by region splitting and merging.

(b) Poorly illuminated images are the difficult ones to be segmented. Explain.

(c) Can variable length coding procedures be used to compress a histogram equalized nimage with 2 grey levels. Explain.

7. Write short notes on :-

(a) Delta Modulation and differential pulse code modulation (DPCM).

(b) Smoothing and sharpening Filters

(c) Gamma Correction

(d) RGB color model.

(Grey level) r

(Number of pixels withGrey level r) nr

0 1 2 3 4 5 6 7

175 12 250 98 153 35 220 57

**********

20

10

10

8

4

8

20

2. Justify / Contradict the following statements :-

(a) For digital images having salt pepper noise, median filter is the best filter.

(b) The first difference of a chain code normalizes it to rotation.

(c) Minimum cost algorithm is meant for detection of edges.

(d) Laplacian is better than gradient for detection of edges.

3. (a) Describe the general compression system model.

(b) Explain briefly the Isopreference curves with respect to Digital Image Processing.

(c) Define Moments, Normalized moments and Central moments.

20

8

6

6

(c) State and prove symmetry and periodicity property of DFT. 6


(b) Consider the two image subset S and S shown in the following figure for V = {1} 1 2determine whether these two subset are :

(1) 4-adjacent (2) 8-adjacent (3) M-adjacent.

2. (a) Perform Histogram Equalization for following. Obtain a plot of original as well as Equalized Histogram.

S1

0 0 0 0 0

1 0 0 1 0

1 0 0 1 0

0 0 1 1 1

0 0 1 1 1

0 0 1 1 0

0 1 0 0 1

1 1 0 0 0

0 0 0 0 0

0 0 1 1 1

S2

(b) Write DDA algorithm to draw a line. Also explain the basic transformation in Computer Graphics.

70 100 40 60 0 80 10 40

Intensity

No. of Pixels

0 1 2 3 4 5 6 7

3. (a) Assume that the edge starts in the first row and ends in the last row for the following grey image :

Sketch all possible paths and determine the edge corresponding to minimum cost path.

5 6 1

6 7 0

7 1 3

Row Column

5

10

10

10

(b) Apply K-L transform on.TGiven X = (0, 0, 0)1T X = (1, 0, 0)2T X = (1, 1, 0)3T X = (1, 0, 1)4

10



15

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Assume suitable data if necessary.

1. (a) Explain following terms :

(i) Isopreference curve (iv) Uniform and Non-uniform sampling

(ii) Mach band (v) Fourier descriptor.

(iii) Simultaneous contrast


6. (a) Explain Hough transform technique for detection of shape from image edges.

(b) Explain Shape No. and find the Shape No. for code :

0 1 1 1 0 3 0 2 2 2 3 2 2 1 2 1 1 1

7. Write short notes :

(a) Wavelet transform (d) Polygonal approximation

(b) Psychovisual redundancy (e) Euler No.

(c) Split and merge

0 1 1 0

1 0 0 0

Let the structuring element B = 1 1

perform (1) Erosion, (2) Dialation, (3) Opening.

10

10

20

**********

5. (a) Find Huffman code for following six symbols. The symbols and their probabilities are given tabular form.

(b) Consider the image :

0.1

Symbol

Probability

a1

0.4

a2

0.06

a3

0.1

a4

0.04

a5

0.3

a6

Also find it average length and entropy.

0 1 0 0

0 1 0 0 A =

10

10

4. Explain in detail following image enhancement tech in spatial domain :-

(a) Image negative (d) High pass filter

(b) Bit plane slicing (e) Low pass filter

(c) Contrast stratching (f) Median filter.

20

1. Justify/contradict following statements (any five) :

(a) Chain codes can be made invariant to translation and rotation.

(b) Median filter is the best solution to remove salt and pepper noise.

(c) All image compression techniques are invertible.

(d) If Kernel of an image transform is separable and symmetric, the transform is expressed in matrix form.

(e) Walsh transform is nothing but sequency ordered Hadamard matrix.

(f) Image resulting from poor illumination cannot be segmented easily.

May - 2007Time : 3 hours Total Marks : 100


20



Give 3 ´ 3 mask for each of them and explain their operation.

7. (a) Draw and explain 2D-Fast Wavelet Transform Filterbanks.

(b) Write short notes on any two :-

(i) Fourier Descriptors (ii) Chain Code (iii) Image Compression Standard.

10

10

(b) For 2 ´ 2 transform A and the image U

Calculate transformed image V and the basis image.

5. (a) Explain the method of segmentation of images by Region Splitting and merging.

(b) Explain the following morphological operation in terms of dilation and erosion operations : Hit or miss, Open, Close, Boundary, Skeleton.

6. (a) Assuming that edge starts in the first row and ends in the last row. For following grey level image, sketch all possible paths and determine edge corresponding to minimum cost path.

(b) For the 3 bit 4 ´ 4 size image perform following operations :

(i) Negation

(ii) Thresholding with T = 4

(iii) Intensity level slicing with

background with r = 2 and r = 51 2

(iv) Bit plane slicing for MSB and LSB plane

(v) Clipping with r = 2 and r = 51 2

2

1 1

1 4-1 3

1 21A = , U =

10

10

10

10

10

2

2

0

2

7

1

7

5

5

1

2

5

3

2

4

2

2

3

6

4

6

0

7

3

1

3. (a) Given below is the table of 8 symbols and their frequency of occurance :

Give Huffman code for each eight symbols.

(b) Write DDA algorithm to draw a line. Calculate all points using the same algorithm to draw line between points x(10, 10) and y(20, 15).

4. (a) Show that 2D - DCT of an image can be computed by row and column passes with 1D - DCT algorithm.

Symbol

Frequency

s1 s2 s3 s4 s5 s6 s7s0

0.15 0.06 0.08 0.21 0.14 0.07 0.040.25

10

10

10

2. (a) A 64 ´ 64 image, represented by 8 bits/ pixel has following grey level distribution :

Perform Histogram Equalization and give new distribution of grey level. Show plots of original and equalized image.

Grey level

No. of Pixel 75 280 416 635 1058 820 684128

1 2 3 4 5 6 70

10

(b) Explain basic principles of detecting following in the images.

(i) Points (ii) Lines (iii) Edges.

10


**********


201. Justify/Contradict following statements .

(a) Enhancement process does not change the information content of image.

(b) The first difference of chain code normalize it to rotation.

(c) For continuous image histogram can be perfectly equalized but it may not be so for the digital image.

(d) Quality of picture depends on the number of pixels and grey level that represent the image ?

(e) Image resulting from poor illumination cannot be segmented easily.

Electronics & Telecommunication

N.B. : (1) Question No. 1 is compulsory.(2) Answer any four out of remaining six questions.(3) Figures to the right indicate full marks.(4) Assume any suitable data if required but justify the same.

2

3 2

3 2-1 1

1

x uy = 0 u = 0

N - 1 N - 1

31

2 2| f(x, y) | = | F(u, u) | S SS S

62. (a) What is histogram of digital image ? What information one can get by observing histogram ?

Grey level nx

ny

No. of pixels p(nx)

pd (ny)

0

0

1/7

1/7

2/7

2/7

3/7

3/7

4/7

4/7

5/7

5/7

6/7

6/7

1

1

4

2

A = U =

16

6

17

8

11

10

8

12

4

11

4

9

0

6

14

5

6

8

7

(b) Suppose that a 8 8 pixel image represented by 4 bits/pixel has following grey level distribution.

´

(b) Prove that the magnitude of determinant of a unitary transform is unity. Also show that all eigenvalues of unitary matrix have unity magnitude.

4. (a) Explain the significance of edge extraction.

3. (a) (i) For the 2 2 transform A and the image U :´

(ii) Prove two dimensional energy. Conversation relation

Calculate the transformed image u and the basis image.

where f(x, y) represent 2-D image signal and F(u, u) represent transform image.

Modify the above histogram such that the desired distribution is as follows :

UQP - (Electronics & Telecommunication) - 10

Symbol

Probability

a1 a2 a3 a4 a5 a6 a7a0

0.008

-2

0

2

0.022

-1

0

1

0.06 0.18 0.13 0.07 0.480.05

-1

0

1

8

20

(b) The 3 3 mask shown below is used to compute the derivative in the x direction at each points in an image.

´

Give an ALU procedure to implement this operator.

7. Write short notes on :

(a) Image compression standard. (d) Discrete cosine transform.

(b) Fourier descriptor. (e) Connectivity of pixels.

(c) Signature.

**********

4

4

6

8

4

(b) (i) Write the expression for Discrete Fourier transform of applied to 2-D signal.

(ii) State and explain the properties of 2-D Fourier transform.

(ii) Generate Huffman code for following table in which eight symbol with corresponding probabilities are given :

6. (a) (i) How many Unique Huffman Codes are there for three symbol source ? Construct these codes.

Given image :

14(b) Prepare edge map at the given image using the following operations. Assume suitable threshold values.

(i) Sobel (ii) Prewitt (iii) Laplacian (iv) Robert's.

65 62 22 36 42

47 46 17 50 46

11 16 24 30 23

47 51 11 42 46

(ii) Under what circumstances cosine transform approximates K-L transform.

53 54 16 58 55

5. (a) (i) What is K-L transform ? Why it is called as 'Method of principle components' 6


June - 2005

Time : 3 hours Total Marks : 100

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Figures to the right indicate full marks.(4) Assume suitable data if necessary.


2. (a) What is the impulse response of each filter

(i) The transfer function is-1 -1 -1 -1 +1 -1H(z , z ) = l - a z - a z - a z z - a z z1 2 1 1 2 2 3 1 2 4 1 2

(ii) Frequency Response is

H (w , w ) = 1 - 2 cos w - 2 cos w1 2 1 2µ µ

6

(b) Let

(c) Three column vector are given below

Prove that they are orthogonal. Using these vectors generate 3 ´ 3 nine orthogonal patterns. P (x, y), i, j, x, y = 0, 1, 2.ij

3. (a) An image of size 5 ´ 5 quantized to 16 grey level is given below. Operate on this image by any two edge extraction

operator of size 3 ´ 3 use appropriate threshold value. Give outputs before and after thersholding. Give your comments.

(b) Assuming that edge starts on the first column and ends in last column. For following grey level image.


1 1

1 4-1 3

1 2A = , B =

Find Kronecker products A B and B A. Comments on you result.* *

f =0

0w0w0w

f =1

0w1w2w

f =2

0w2w4w

-j2p/3where w = e

10

12

15

13

1

12

10

14

4

12

15

15

2

15

14

15

3

10

12

1

2

8

10

2

1

j

k 0 1 2

0

1

2

2

1

6

1

1

8

0

7

2

6

8

10

10

1. (a) Define : (i) Spatial Resolution (ii) Intensity Resolution

(iii) Temporal Resolution (iv) PSF (Point Spread Function)

(v) Image Enhancement.

(b) Reason the following contentions.m(i) For N ´ N Image with 2 grey level, points lying on the isopreference curve

in N-M plane corresponds to image of equal subjective quality.

(ii) Moments are used as a similarity measure in comparisions of two regions of an image with the other regions.

(iii) For continuous image histogram can be perfectly equalised, but it may not be so for digital image.

(iv) Brightness discrimination is poor at low levels of illuminations.

10

2.5 4´


(b) Explain following techniques : (i) Arithmetic Coding. (ii) Bit Plane Coding.

7. Write short note on following (any four) :

(i) Fourier Descriptors. (iv) Wavelets.

(ii) Moments. (v) RGB colour model.

(iii) Uniform sampling and non uniform sampling.

10

20

**********

4. (a) What are different types of redundancies in digital images.

(b) Explain following method of data compression :

(i) Huffman Coding (ii) Runlength Coding (iii) Transform Coding.

5. (a) Explain in detail enhancement technique in spatial domain used for images.

6

12

8

(b) (i) Discuss thining algorithm for finding skeleton of a region. Apply it to a square and triangle.

(ii) Illustrate opening and closing operations with suitable examples.

6. (a) Explain with examples the following :

(i) Thinning. (ii) Global processing via graph theoretical techniques.

6

6

10


1. (a) Explain Isopreference curves with respect to digital image-processing.

(b) Reason the contention : Storing of images in frame buffers has many advantages over storing them in computer memory.

(c) State and prove Parsevels energy relation with respect to 2-D image signal.

(d) Explain Hit and Miss transform.

(e) State the difference between convolution and correlation.

2. (a) The 3 ´ 3 mask shown below is frequently used to compute the derivative in x direction at each point in an image.

Give an ALU procedure to implement this operation.

(b) Explain in brief frequency domain enhancement technique. State and explain difference between, restoration and enhancement.

3. (a) A digital image with eight Quantization level is given below.

f(x, y) = | i - j | for i, j = 0, 1, 2, 3, 4, 5, 6, 7.

Perform histogram equalisation. Derive the transformation function and draw new histogram.

December - 2005Time : 3 hours Total Marks : 100

20

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four out of remaining six questions.(3) Make suitable assumptions wherever necessary and state them clearly.

10-1

0

1

-2

0

2

-1

0

1

10

10



(b) Explain following methods of data compression indicating clearly whether they are lossy or lossless. Also state the type of redundancy they can minimize (any four) -

(i) Huffman Coding (iv) Differential Pulse Code Modulation (DPCM)

(ii) Transform Coding (v) Transform Coding.

(iii) Adaptive Delta Modulation

4. (a) Three column vectors are given below :

Prove that they are orthogonal.

(b) Using these vectors generate 3 ´ 3 nine orthogonal patterns

P (x, y) i, j, x, y = 0 1 2.ij

(c) Evaluate the transform coefficients a so that image I(x, y) is given by ij

where I(x, y) = i = 0

2

SI(x, y) = a p (x, y)ij ijj = 0

2

S3

1

2

2

3

1

1

2

3

f =0

0w0w0w

f =1

0w1w2w

f =2

0w2w4w

-j2p/3where w = e

10

5

7

8

6. (a) State and explain 1-D and 2-D forward and inverse transformation Kernel. Write the expression for 1-D and 2-D Discrete cosine transform. Express 1-D and 2-D DCT in term of above transformation Kernels.

(b) Prove that the magnitude of determinant of unitary transform is unity. Also show that all eigenvalues of unitary matrix have unity magnitude.

(c) Describe edge detection via graph theoretic techniques.

7. Write short notes on following (any five) :-

(a) Fourier Descriptor (d) Dilation and Erosion

(b) Signature (e) Moments

(c) Wavelets (f) Uniform and nonuniform sampling.

8

6

6

20

**********

5. (a) Find the DFT of 2-D image signal shown below -

Use DITFFT or DIFFFT algorithms.

(b) What is threshold ? Explain how to obtain the optimum threshold for image segmentation.

f(x, y) =

4

3

4

1

2

2

2

2

1

4

3

3

2

2

2

4

8

5

(c) Describe the region growing and spliting and merging method of image segmentation.

7


201. Answer the following questions (any four) :

(a) Explain why Brightness Discriminations is poor at Low Level of Illumination.

(b) Explain convolution in the spatial and the frequency domain. Derive the relationship between the two domain.

(c) Compare between Contract Stretching and Histogram Equalization.

(d) Explain the role and nature of image fidility criteria in image compression.

(e) A 2-dimensional DFT can be obtained using a 1-dimensional DFT algorithm twice. Explain.

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Assume suitable additional data if required, state and justify the assumptions made.

82. (a) Define "Image" as a two-dimensional array and state what is meant by Grey Level Distribution.

(b) Can variable-length coding procedure be used to compress histogram equalised nimage with 2 Grey Levels ? Explain.

(c) Low pass filter is a smoothing filter. Justify.

8

4

123. (a) Suppose that 8 ´ 8 Pixel P image represented by 4 bits/pixel has following grey level distribution.

Grey Levels

No. of Pixels

0 1/7 2/7 3/7 4/7 5/7 6/7

4 16 17 11 8 4 4 0

7/7


(c) Write Kernel matrix of size 4 (four) and application of Slant and Haar Transform, for the given image f(x, y), find the Slant and Haar Transform matrix.

5 4

f(x, y) =

4 3 2

4 6 6 3

2 6 6 3

2 3 5 2

(b) Why ? Fourier transform and the frequency domain tools are so useful for image enhancement. With the help of neat block diagram explain the basic of filtering in the frequency domain. Give the reasons of shifting the origin.

4. (a) What is the difference between Hadamard and Walsh Hadamard Transform ? Is the Hadamard matrix orthogonal ?

8

5

(b) If the Kernel of an image transform is separable and symmetric, the transform can be expressed in matrix form. Explain.

5

Draw Original specified and modified histogram.

2 6 8 10 12 11 9 6

Modify the above histogram such that the desired distribution is as follows :

Grey Levels

No. of Pixels

0 1/7 2/7 3/7 4/7 5/7 6/7 7/7


8

6

207. Write short notes on any four of the following :

(a) Median filtering.

(b) Connectivity of pixels.

(c) Zooming of a digital image.

(d) Uniform and non uniform sampling.

(e) Predictive coding.

(f) Fast Hadamard transform.

**********

5. (a) Detect the boundary and segment of the given image using graph theoretical approach. Also draw the graph for the given image.

(b) Explain the basic principles of detecting the following in image-

(i) Point (ii) Lines (iii) Edges (iv) Region Soliting

(c) Describe how Hough transform is used for boundary shape detection.

10 7

Image =

2 2

5 7 2

5 1 0

8

6

66. (a) Explain the origin of data redundancy in digital images. How these redundancies are reduced ?

(b) Explain the basic principles of transform coding for image compression. Illustrate the same with the help of DFT and DCT.

(c) Construct the IGS code of the given grey level set :

Grey level - 75, 105, 98, 175, 200, 130, 210, 220

Decode the received IGS code 0 1 1 1.

1. Justify/contradict following statements :-

(a) If Kernel of an image transform is separable and symmetric, the transform can be expressed in matrix form.

(b) All image compression techniques are invertible.

(c) Chain codes can be made invariant to translation and rotation.

(d) Median filter is best solution to remove salt and pepper noise.

(e) Walsh transform is nothing but sequency ordered Hadamard matrix.


20




(b) Calculate Entropy of the source.

(c) Find its average length.

(d) Calculate efficiency of the code obtained in 3(a).

4. (a) Show that 2D-DCT of an image can be computed by row and column passes with 1D-DCT algorithm.

(b) For 2 ´ 2 transform A and the image U

Calculate transformed image V and the basis image.

5. (a) Explain the method of segmentation of images by region splitting and merging.

(b) Explain the following morphological operation in terms of dilation and erosion operations Hit or Miss, Open, Close, Boundary, Skeleton.

2

1 1

1 4-1 3

1 21A = , U =

4

3

3

10

10

10

10

Perform Histogram equalization and give new distribution of grey level. Show plots of original and equalised image.

(b) Explain basic principles of detecting following in the images :-

(i) Points (ii) Lines (ii) Edges.

Give 3 ´ 3 mask for each of them and explain their operation.

3. (a) Given below is the table of 8 symbols and their frequency of occurrence :-

Give Huffman Code for each eight symbols.

Symbol

Frequency

S1 S2 S3 S4 S5 S6 S7S0

0.15 0.06 0.08 0.21 0.14 0.07 0.040.25

10

10

**********

6. (a) Explain different types of color models and where they are used. How do convert colors from RGB to HSI model ?

(b) Assuming that edge starts in the first row and ends in the last row, for following grey-level image, sketch all possible paths an determine edge corresponding to minimum cost path.

2

2

0

2

7

1

7

5

5

10

10

7. Write short notes on (any four) :-

(a) Fourier descriptors (d) Image Compression Standard

(b) DPCM / DCT Encoder (e) Pseudocolor Image Processing.

(c) 2D-wavelet transform filterbank

20

2. (a) A 64 ´ 64 image, represented by 8 bit/pixel has following Grey level distribution :-

Grey level

No. of Pixel 75 280 416 635 1058 820 684128

1 2 3 4 5 6 70

10


5

12

(b) What do you understand by Gamma correction - explain.

5. (a) Illustrate the following techniques :

(i) Arithmetic coding (ii) Bit plane coding.

152. (a) (i) Explain zooming of an image. Does it increase the information content of an image.


20

20


(i) Image compression standards.

(ii) Wavelet transform.

(iii) Connectivity

(iv) Sampling and quantization.

3. Write notes on :

(i) Fidelity criteria (iii) Boundary descriptors

(ii) Basic operations in morphology (iv) Thresholding.

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) All questions carry equal marks.

1 2 5 2 1

2 3 4 1 2

4 1 3 1 1

4 2 5 1 1

1 1 7 1 3

(ii) Perform zooming on the following image by replication and linear interpolation. Is the result same in both the cases.

84. (a) Explain the different approaches used to describe texture of digital images.

12(b) Explain with examples the following :

(i) Thinning.

(ii) Global processing via graph Theoretic Techniques.

Computer Engineering

UQP - (Computer Engineering) - 18

12

**********

8(b) Plot the histogram for the following image. Perform histogram equalization and then plot the equalized histogram and histogram equalized image.

10

8

10

6. (a) Explain filtering in spatial domain.

(b) Write the 8 8 matrix for Walsh transform. Write the characteristics of Walsh transform. Also show the Walsh basis functions for N = 4.

´

7. (a) Explain image averaging and image subtraction.

(b) Explain the following boundary descriptors :

(i) Fourier descriptors, (ii) Moments, (iii) Shape numbers.

1 1 5 3 3

1 2 6 6 7

1 4 0 6 2

4 4 2 5 2

7 4 0 2 2


1. Reason the following contentions :-

(a) The first difference of chain code normalise to Rotation.

(b) Spectral approach using Fourier spectrum is useful in texture matching.

(c) Brightness discrimination is poor at low levels of illumination.

(d) Quality of picture depends on the number of pixels and number of grey level that represent the picture.

2. (a) A contrast stretching is performed as per the following characterize :

v = 7 3 £ m £ 5

= u otherwise.

where u and v are the intensities for the input and output image respectively. Determine the o/p image when, contrasts stretching g is applied on following:

(b) State and explain the features of median filtering. Complete the o/p of median filtering in the following case.

(i) y(m) = {2 4 8 3 2} and w = {-1 0 1 2}

(ii) y(m) = {8 2 4 3 4} and w = {-1 0 1}


20

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Figures to the right indicate full marks.

12

0

2

2

1

2

1

3

4

3

1

0

5

4

5

2

2

5

3

3

6

1

1

2

0

6

6

6

1

1

2

8

UQP - (Computer ) - 19 Engineering

(c) Evaluate the transform coefficient a so that given image I(x, y) is given by -ij

where I(x, y) =

8

i = 0

2

SI(x, y) = a P (x, y)ij ijj = 0

2

S3

1

2

2

3

1

1

2

3

3. (a) Three column vectors are given below :

f =0

0w0w0w

f =1

0w1w2w

f =2

0w2w4w

6

-j2p/3where w = eProve that they are orthogonal.

(b) Using these vector generate 3 ´ 3 nine orthogonal patterns p (x, y), i, j, x, y = 0,1,2.ij 6


6. (a) What are different types of redundancies in digital images ?

(b) Explain following method of data compression indicating clearly whether they are lossy or lossless. Also state the type of redundancy they can minimize (any four) :

(i) Huffman coding. (iv) Differential pulse code Modulation (DPCM).

(ii) Transform coding. (v) Runlength Coding on bit planes.

(iii) Adaptive delta modulation (ADM).

7. Write short notes on (any five) :

(a) JPEG. (e) Dilation and Erosion.

(b) Fourier Descriptor. (f) Moments.

(c) Discrete Cosine transform. (g) Shape number.

(d) Hit and Miss transform.

8

12

20

**********

10

12

15

13

1

12

10

14

4

12

15

15

2

15

14

15

3

10

12

1

2

8

10

2

1

5. (a) Give following masks of size 3 ´ 3 and explain their usefulness in image processing :

(i) Sobel operator (iv) High pass filter

(ii) Roberts operators (v) Median filter

(iii) Prewitt operator (vi) Frei-chen operator.

12

(b) Histogram of a digital image with eight quantization level is shown below perform histogram equalisation. Derive the transformation function and new histogram.

12

Grey level

No. of grey level 140 50 60 70 170 130 160220

1 2 3 4 5 6 70

hr

r

(b) An image of size 5 ´ 5 quantized to 16 grey levels is shown below. Operate on this image by

any two edge extraction operators of size 3 ´ 3. Use appropriate thereshold should value. Give output before and after thresholding. Give your comments.

8

4. (a) What is histogram of a digital image ? What information one can get by observing the histogram of images ?

8

2. (a) Write a note on Wavelet transform.

(b) Classify Image Compression Techniques.

(c) Name and explain any four zero memory operations.

3. (a) What do you mean by Image Averaging.

(b) Texture-an Image descriptor. Explain.

(c) Can Graph theoretic techniques be used in Image Segmentation ? If so, how?

4. (a) What do you understand by Variable Length Coding techniques ? Explain at least two techniques.

(b) What is constant area coding technique ?

(c) What is Chain code and what are its applications.

5. (a) Define and explain -Moments, Normalized moments, and central moments.

(b) What are Morphological operations ? For a region explain Boundary extraction operation and then region filling operation on the extracted boundary.

6. (a) How do you get a Skeleton of a region using Medial Axis Transformation ?

(b) Histogram of a digital image with 8 quantization levels is given below. Perform Histogram Equalization. Derive the transformation function and the new histogram

7. (a) Write a note on Discrete Cosine Transform and its application in Image Compression.

(b) Write 8 ´ 8 Haar transform matrix and its Signal Flow Graph. Using this Butterfly diagram (Signal Flow Graph) compute Haar transform for x(n) = {1,2,3,4,1,2,3,4}.

(Grey level) r

(Number of pixels withgrey level r) nr

0 1 2 3 4 5 6 7

250 550 450 250 100 150 150 50

8

25

9

25

**********

5

5

10

5

10

5

10

5

5

10

10

10

10

10

10


1. Write short notes on :-

(a) Applications of Image Processing.

(b) Image formation in the eye.

(c) Correspondence between Filtering in the Spatial and Frequency Domains

(d) Hough Transform.


N.B. : (1) Question one is compulsory.(2) Attempt any four questions from remaining six questions.(3) Assume suitable data wherever necessary and state it.

Time : 3 hours Total Marks : 100November - 2006

20

1. Justify the following contentions :

(a) High pass filter is sharpening filter.

(b) Enhancement process does not add any information to the image.

(c) The process of removing phycho-visual redundancy is a lossy process.

(d) Quality of the image depends on the number of grey levels that represent the image.

2. (a) What do you understand by histogram of a digital image ? What information one can get by observing the histogram of images ?

(b) Perform histogram equalization and draw new equalized histogram :

3. (a) What do you understand by negative of an image ?

(b) What is image segmentation ? Explain the following methods of image segmentation : (i) Region growing (ii) Splitting and Merging (iii) Thresholding.

4. What is the difference between the following :

(a) Bit-plane slicing and Grey-level slicing

(b) Spatial-domain processing and Transform-domain processing

(c) Convolution and correlation

(d) Low-pass filter and Median filter.

5. (a) Explain separability property of 2-dimensional DFT.

(b) What are the different types of redundancies in images ?

(c) What do you understand by Discrete Cosine Transform ? Explain its application in Image Compression.

6. (a) What do you understand by Image-Averaging ?

(b) Explain Isopreference curves with respect to digital-image-processing.

(c) Describe the general compression system model for digital images.


(a) Hit or Miss Transform (c) Homomorphic filtering

(b) Dilation and Erosion (d) Hough Transform.


20

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.

Grey level (K)

No. of Pixels (n )K 1023 850 656 329 245 122 81790

1 2 3 4 5 6 70

10

10

5

20

5

5

10

5

5

10

20



**********

Electronics Engineering

N.B. : (1) Question No. 1 is compulsory. (2) Attempt any four questions out of the remaining six questions.


21

12

1. State true or false and justify (Answer any seven) :

(a) Variable length coding procedures can be used to compress a histogram equalized nimage with 2 grey levels.

(b) Segmentation algorithms for monochrome images generally are based on two basic properties of grey level values.

(c) All image compression techniques are invertible.

(d) Enhancement process does not change the information contents of the image.

(e) For digital images having salt-pepper noise, median filter is the best filter.

(f) In general, fine sampling is preferred in smooth regions, while coarse sampling is preferred in regions, where grey level transitions are sharp.

(g) Run length coding always gives data compression.

(h) Poorly illuminated images can be easily segmented.

(i) The discrete histogram equalization technique will not, in general, yield a flat histogram.

2. (a) Generate Huffman code for the given source. Calculate entropy of the source, average length of the code and the compression ratio achieved :

Symbol

Grey level r

Probability

Number of pixels nr

a1 a2

0 1 2 3 4 5 6 7

a3 a4 a5 a6 a7 a8

0.06

2

1

6

1

1

8

0

7

2

0.02

40 60 80 120 140 160 180 220

0.3 0.5 0.04 0.01 0.03 0.04

8

8

6

10

6

(b) Consider an 8-pixel line grey-scale data, {12, 12, 13, 13, 10, 13, 57, 54}, which has been uniformly quantized with 8-bit accuracy. Construct its 3-bit IGS code.

State and explain in brief the type of redundancy which is exploited here to achieve compression.

3. (a) Describe edge detection via graph-theoretic technique.

(b) Describe all possible masks for line detection.

(c) Assume that the edge in the grey-level image starts in the first column and ends in the last column. Find the cost of all possible edges and sketch them. Find the edge corresponding to minimum cost path :

4. (a) Histogram of a digital image with eight quantization level is given below. Perform histogram equalization. Derive the transformation function and the new histogram.

UQP - (Electronics ) - 23 Engineering

10(b) Explain briefly the following operations to digital images along with one application for each :

(i) Median filter. (iv) Image subtraction.

(ii) Contrast stretching. (v) Bit plane slicing.

(iii) Thresholding.

Symbols Probability Code 1 Code 2 Code 3

a1

a2

a3

a4

0.5

0.25

0.125

0.125

0

1

00

11

0

10

110

111

0

01

011

0111

85. (a) For a given source A = {a , a , a , a } the following codes were developed. Check 1 2 3 4for each of them whether it is uniquely decodable or not. Also state which one is the most optimum compared to others and why ?

40 60 55 50

45 62 48 55

20 24 27 34

07 06 40 45

4

5

5

6

4

8

20

(b) Perform zooming operation using linear interpolation as well as replication, on following image. State which method performs generally better.

6. (a) Write 8 8 Haar transform.

(b) Write Walsh matrix W4. Without using Walsh matrix obtain transform for

f = [ 1 2 3 4 ].

(c) Explain any two properties of 2-dimensional Fourier transform.t(d) If f = [ 1 0 2 3 4 1 3 1 ] ; find its Hadamard transform.

´

(c) Give following masks of size 3 3 and explain their usefulness in image processing (any four) :

(i) Sobel (iv) Frei-chen

(ii) Roberts (v) Low-pass filter

(iii) Prewitt

´

7. Write short notes on any four of the following :

(a) Fidelity criteria.

(b) Homomorphic filtering.

(c) Connectivity of pixels.

(d) Compression using LZW method.

(e) Karhunen-Loeve transform (Hotelling transform).

**********



1. (a) Explain the elements of digital image processing system.

(b) What is fidelity criteria ? How are these used in quantifying the nature and extent of information loss ?

2. (a) Obtain the slant transformation of the image f(x, y) =

(b) Suppose that a digital image is subjected to histogram equilization. Show that second pass of histogram equilization will produce exactly the same result as first pass.

3. (a) Explain the different types of color models and where they used. Also how to convert colors from RGB-to-HSI and HSI-to-RGB models.

(b) Explain with block diagram basic steps for filtering in frequency domain.

4. (a) What is texture ? Explain in brief three principal approaches used in image processing to describe the texture of region.

(b) Write a short note on Log transformation and Power Law transformation.


10

N.B. : (1) Attempt any five questions.(2) Figures to the right indicate full marks.(3) Assume any suitable data wherever required but justify the same.

1 5

6 7

10

10

10

15

5

10

10


7. Write a short note on :

(a) Basic Adaptive thresholding. (c) DPCM/DCT Encoder.

(b) Image Averaging. (d) Pseudocolor Image Processing.

20

**********

(b) Obtained the Hotelling transform given four vectors.T T T Tx = (0,0, 0) , x = (1,0,0) , x = (1,1,0) and x = (1,0,1 ) .l 2 3 4

6. (a) Given 4 ´ 4 image whose grey level, order lexicographically are as follows :

3120203123101302

Calculate the spatial moments upto second order.

(b) Differentiate between Lossy compression and Lossless compression.

(c) Write down the mask and usage of following operator :

(i) Sobel (ii) Prewitt (iii) Laplacian.

12

10

4

6

5. (a) Explain in detail image compression model. 8

1. Justify the following statements :-

(a) The D distance between two points p and q is equal to the shortest 4-Path between 4there points.

(b) A Highpass - filtered image can be obtained in the spatial domain as

Highpass = Original - Lowpass.

(c) Poorly illuminated images can be easily segmented.

(d) Laplacian is not a good edge detector.

(e) Runlength coding is loss-less coding but may not give data compression always.

2. (a) With the help of neat block diagram explain the basic functional elements of an image processing system.

(b) What is Histogram of a digital image ? What information one can get by observing Histogram of an image ?


N.B. : (1) Question No. 1 is compulsory.(2) Answer any four questions out of remaining six questions.(3) Assume suitable data if necessary, state and justify the assumptions made.

20

5

5



(b) Apply Low Pass and High Pass Spatial Mask on the following image matrix :

4. (a) Explain the property of Hadamard Transform, also find the Hadamard Transform of ff(x) = {1, 2, 2, 1}

F =

5

6

3

4

4

2

5

3

7

3

6

7

1

4

8

1

2

1

2

8

10

8

(c) A digital image with eight quantization levels has the following histogram perform histogram equalization and derive transformation function. Give new equalized histogram.

3. (a) Apply the following Image Enhancement techniques on the given Image.

(i) Digital Negative

(ii) Bit Plane Slicing

(iii) Thresholding (T = 4)

Grey level r

No. of Pixels with grey level nr 160 200 150 80 80 100 110120

1 2 3 4 5 6 70

Image

10

102

7

3

0

2

1

4

5

4

1

3

2

1

0

6

4

0

4

2

1

5

1

6

3

4


(c) Explain Region oriented Segmentation.

6. (a) Explain basic data rudendancies. Describe basic compression model used for image compression.

(b) Classify with reasons, the following data compression techniques into lossy and lossless schemes : (i) Run Length Coding (ii) DCT Compression.

4

8

8

(c) How many Unique Huffman Codes are there for three symbol source ? Construct these codes.

4

(b) Three column vectors are given below :-

show that they are orthogonal, also generate all Possible Pattern.

(c) Explain the following Property of Image Transform

(i) Symmetrical (ii) Orthonormal.

1

1

1

x = ,1

-2

1

1

x = ,2

0

-1

1

x =3

8

4

5. (a) Explain Segmentation based on discontinuity and Segmentation based on similarities.

(b) Design compare gradient operators of size 3 ´ 3 to measure gradients of edges oriented in eight directions.

E, NE, N, NW, W, SW, S and SE.

Give the form of these eight operators using coefficients valued 0, 1, or -1 also specify the gradient direction of each mark.

8

8


(a) Image Sampling (d) Hough Transform

(b) K-L Transform (e) Transform Coding.

(c) Image acquisition method

**********

20


201. Reason the following contentions :m(a) For an N ´ N image with z grey levels, points lying on the isopreference curve in

N-m plane correspond to images of equal subjective quality.

(b) Brightness discrimination is poor at low levels of Illumination.

(c) Quality of the picture depends on the number of pixels and the number of grey level that represent the picture.

(d) Moments are used as a similarity measure in comparison of two regions of an image with any other region.

Biomedical Engineering

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) All questions carry equal marks.

2. (a) Assuming that the edge starts on the first column and ends in the last column for the following grey level image.

15


j

k 0 1 2

0

1

2

2

1

6

1

1

8

0

7

2

(b) Write a note on Image Subtraction. 5

3. (a) Write note on Uniform and Non Uniform sampling.

(b) Plot the histogram for the following image. Equalize the histogram and then plot the equalized histogram and the image corresponding to the

equalized histogram. The 4 ´ 5 image is represented by 3 bits/pixel.

8

12 1 3 3 3 2

3 2 3 1 2

5 1 1 1 4

5 6 6 7 0

4. State and explain the difference between the following :-

(i) Image processing and Computer Graphics

(ii) Smoothing Filters and Sharpening Filters

(iii) Lossy Compression and Lossless Compression

(iv) Linear Contrast stretching and Histogram Equalization.

20

UQP - (Biomedical ) - 28 Engineering

5. (a) Give an ALU procedure to compute the derivative in

x-direction at each point in an image using the 3 ´ 3 mask shown in the figure.

8 -1

0 0 0

1 2 1

-2 -1

7. (a) What is morphology ? Name and explain the basic operations in morphology. 5

(b) Write notes on : (i) Texture (ii) Point Detection 15

**********

(b) (i) Write the Haar Transform matrix of size 8 ´ 8.

(ii) Sketch the butterfly diagram for the flow graph for the fast algorithm of the above Haar transform

(iii) Write the properties of Haar transform.

12

6. (a) What are the different types of redundancies that can be present in a digital image and state which method can be used to remove/reduce them ?

8

(b) Write short notes on : (i) Expert systems. (ii) Distance measures. 12

1. (a) Discuss briefly the process of image formation in the eye.

(b) Discuss the following image enhancement techniques and give their applications : (i) Log transformations

(ii) Power - law transformations

(iii) Contrast stretching.

2. (a) Explain the following histogram modification techniques and state their advantages and disadvantages :

(i) Linear stretching

(ii) Histogram equalization

(iii) Histogram specification.

(b) What are the steps required to perform filtering in the frequency domain. Also, explain the Butterworth low pass filter.

3. (a) Explain the following Edge Extraction operators with their advantages and disadvantages if any :

(i) Sobel (ii) Laplacian.

May - 2007

Time : 3 hours Total Marks : 100

8

Biomedical Engineering

N.B. : (1) Question No. 1 is compulsory.(2) Attempt any four questions out of remaining six questions.(3) Figures to the right indicate full marks.

12

12

8

10


5. (a) Explain different image redundancies along with a compression technique to reduce them.

(b) Explain the Huffman coding procedure with a suitable example.

6. (a) Discuss the different reconstruction techniques used in computed tomography.

(b) Explain opening and closing operations with suitable examples.


(a) Median Filtering (d) Homomorphic Filtering

(b) Connectivity of Pixels (e) Sampling and Quantization.

(c) Mask Mode Radiography

1 3 4 5

3 2 6 4

4 6 2 3

5 4 3 1

10

10

12

8

20

**********

(b) For the given image, perform region based segmentation by split and merge technique.

4. (a) State and explain the differences between 2-D DFT and DCT.

(b) Generate Haar Matrix of size 4 and comment on the result.

(c) Compare the DFT of the following image :

13

13

11

13

12

62

62

61

12

13

12

13

11

62

61

62

13

63

63

62

62

63

61

63

12

63

62

63

63

61

62

11

11

61

61

62

62

61

13

12

12

11

62

61

11

62

12

11

11

12

12

13

12

13

13

12

12

11

12

13

11

13

11

12

10

6

6

8


Image-Uni Q Papers

Documents

Transcript of Image-Uni Q Papers