Fundamentals of Image Processing-2181102

Question Bank
The City block distance between pixels p and q for the following image is
2 5 9(q) 1
1 1 0 2
5 4 3 1
1(p) 2 4 7
Set v ={0,1,5,9}

How 2-D image is generated from only one sensor?

Draw the resultant image after quantized (4-bit) the input image as shown below:
16 15.1 12.3
11.9 0.6 2.5
9.9 4.4 13.2
What will be the total size( in Bytes) of an image having 1M pixels quantized with 16 bits

The reflection component for pure white region will be________________

Define: Resolution of a Gray image.

What will be the pixel values after application of the log transformation on an input image
having pixel values equal to 15 for an image region of 5 rows and 5 columns?

What is ringing effect and why it happens?

What is Power Law Transformation? Apply Power Law Transformation on the following
image. Use γ =0.4, C=1
9 9 9 9
1 2 3 4
5 6 7 8
9 9 9 9
What is contrast stretching? Draw piece-wise linear input-output characteristics with
corner points (0,0), (100,110), (140,200) , (185,225) and (255,255). Determine the output
for following input 4x4 image segment.

50 100 120 140

65 120 145 160
175 201 224 255
152 111 160 210
Apply histogram equalization on the image shown below:

1 12 15 2 8 9 9 10
 10 11 1 0 12 2 3 5
8 10 11 12 10 8 5 4
2 3 5 2 0 2 3 1
1 1 8 9 8 9 8 9
15 10 10 8 11 15 0 11
13 11 0 13 10 11 9 0
11 8 13 1 2 3 4 5

Draw the resultant image.

How the digital image will be generated from a continues image?

a. By digitizing the co- b. By digitizing the c. By digitizing both
ordinates Amplitude co-ordinates and
amplitude values
After performing the linear convolution between the two matricesX(m,n) and h(m,n), what
will be the resultant image Y(m,n)?. The centre pixel is h(0,0).
1 2 3
𝑋 𝑚, 𝑛 = 4 5 6 , ℎ 𝑚, 𝑛 = 1 1
7 8 9

𝟑 𝟓 𝟑 11 18 1 3 5
a. 𝒀 𝒎, 𝒏 = 𝟗 𝟏𝟏 𝟔 b. 𝑌 𝑚, 𝑛 = 32 39 c.𝑌 𝑚, 𝑛 = 4 9 11
𝟏𝟓 𝟏𝟕 𝟗 53 60 7 15 17
The image Im(x,y) is uniformly quantized with 3-bits, what will be the image generated
after applying image negative operation on it?
7 7 7
𝐼𝑚 = 4 4 4
1 1 1

a. b. c.
1 1 1 𝟎 𝟎 𝟎 1 4 7
𝐼𝑚 = 4 4 4 𝑰𝒎 = 𝟑 𝟑 𝟑 𝐼𝑚 = 1 4 7
7 7 7 𝟔 𝟔 𝟔 1 4 7

What will be the median valued of the marked pixels shown in following figure using a 3X 3
mask?
18 22 33 25 32 24
𝐼 = 34 𝟏𝟐𝟖 𝟐𝟒 𝟏𝟕𝟐 𝟐𝟔 23
22 19 32 31 28 26
a. 24, 31,31,26 b. 33,32,31,23 c. 22,25,31,26
The Input image and corresponding unsharp mask is shown in following figures. How these
images will be used for obtaining the sharpen image?
 a. Input image + b. Input image - c. Input image -
weighting fraction* weighting fraction* Unsharp mask
mask image mask image image
Let’s consider the 8-bit image is decomposed into 8 no. of bit planes. What will be effect of
higher order bit planes on the visual content of the input image?
a. Details are visible in b. Higher order bit c. Overall picture
higher order bit planes planes are information are
represented as gray found in higher
level values order bit planes
Sharpening is performed using which of the following operations for digital image?
a. spatial integration b. spatial c. spatial averaging
differentiation
The input image of 4X4 is shown below. Use histogram equalization technique on it and
select the correct equalized image (J) from the options.
I= [1 2 3 4; 5 5 6 1; 1 0 2 4;2 3 2 1]
a.J=[ 2 4 4 6; 7 7 7 2; 2 0 4 6; 4 4 b.J=[ 2 4 45; 6 6 7 2; 2 0 4 5; c. J=[ 2 4 5 6; 7 7 7 2; 2 0 4
4 2] 4 4 4 2] 6; 4 5 4 2]
Which colour has maximum wavelength in the Visible spectrum?
a.Violet b. Blue c. Red
The probability Distribution Function (PDF) for the Gaussian noise is
(𝒛−𝝁)𝟐 2 𝑎 𝑏 𝑧 𝑏 −1
𝟏 − b. 𝑃 𝑧 = − 𝑏 (𝑧 − c. 𝑃 𝑧 = − (𝑏−1)!
a. 𝑷 𝒛 = 𝟐
𝒆 𝟐𝝈𝟐
2
𝟐𝝅𝝈 (𝑧 −𝑎 )
𝑎)𝑒 − 𝑏
An operation used for conversion of input image into discrete Fourier transformed function
in Homomorphic filtering is?
a. Logarithmic function b. Exponential function c. Negative
transformation
The image shown in following figure is corrupted by periodic noise. Which filter is required
for removal of this noise?

a. High pass filter b. Band reject filter c. Notch filter

The image shown in following figure is degraded due to motion blur. What kind of
restoration technique will be more suitable for removal of motion blur from the image?

a. High pass Filtering b. Weiner filtering c. Band pass Filtering

Some color map conversion technique is applied on an input image ( Fig 1). After applying
which color map conversion technique the new image (Fig 2) will be generated?

Fig 1: Input image (RGB Format) Fig 2: Resultant image

a. Conversion from RGB b. Conversion from c. Conversion from
format to HSV format RGB format to RGB to CMY format
YCbCr format
___________________ produces a color image from monochrome image.
a. Pseudo color b. Color conversion c. True color
processing technique processing
One of the objective measures of image compression is Peak Signal to Noise Ratio (PSNR) .
For computing the PSNR values, one has to compute the Mean Square Error (MSE) between
input and output images. What will be the value of MSE for the following input image and
reconstructed image after decompression?
1 8 6 6 2 8 8 7
Original Image = 6 3 11 8 ReconstructedImage = 6 3 12 8
8 8 9 10 5 4 9 1
9 10 10 7 15 9 11 9

a. MSE = 9.6875 B. MSE = 31 c. MSE =1.9375

Which type of redundancy is present in the following image?
 a. Pyschovisual b. Interpixel c. Temporal
redundancy redundancy redundancy
An example of Huffman coding is shown below. Compute the Lavg for the given code and
select the correct answer from the given choices.

a. Lavg= 2.7 bits b. Lavg= 2.7 bytes Lavg=3 bits

Apply LZW coding on the given data sequence and chose the correct option for encoding
the sequence : B A A B A A B A A B A A B A A C C D B A F E F F and corresponding dictionary
entries

a. Encoding sequence is b. Encoding sequence is “ 1 c.Encoding sequence is “ 1

“ 1 2 2 7 9 12 12 8 3 15 2 7 9 12 12 13 15 16 18 221221221221223
7 18 20” and 20” and dictionary entries 3 4 1 2 5 6 5 5” and
dictionary entries are are listed below: dictionary entries are
listed below: 9 AB listed below:
9 AB 10 AA 1 B
10 BAA 11 BABA 2 A
11 ABA 12 AAB 3 C
12 AAB 13 BAAC 4 D
13 AAC 14 CC 5 F
14 CC 15 CD 6 E
15 CD 16 DB
16 DB 17 BAF
17 BAF 18 FE
18 FE 19 EEF
 19 EF 20 FF
20 FF
The first step in JPEG compression technique is _____________
a. DCT Transformation b. Quantization c. Run length coding
A source emits four symbols { a, b, c, d } with the probabilities 0.4, 0.2, 0.1 and 0.3
respectively. Construct arithmetic coding to encode the word “dad”
a. Encoding of word b. Encoding of word “dad” c. Encoding of word “dad”
“dad” is (0.784, 0.82] is (0.7, 0.772] is (0.748, 0.772]
The information present in the image, which is ignored by the human eye is known as
a. Coding redundancy b. Interpixel c. Pychovisual
redundancy redundancy
The gray level probability density function of an image is given by 𝑝 𝑟 = 5 exp −5𝑟 , 0 ≤
𝑟 ≤ 1. Which transformations 𝑠 = 𝑟 2 𝑜𝑟 𝑠 = 𝑠𝑞𝑟𝑡(𝑟) would produce a better image?
a. 𝑠 = 𝑟 2 stretches the low gray b.𝒔 = 𝒓𝟐 compresses the c. Both 𝑠 = 𝑟 2 𝑎𝑛𝑑 𝑠 =
level while 𝑠 = 𝑠𝑞𝑟𝑡(𝑟) low gray level while 𝑠𝑞𝑟𝑡(𝑟) are better choice
compresses the low gray level. 𝒔 = 𝒔𝒒𝒓𝒕(𝒓) stretches the for enhanced image.
Hence 𝑠 = 𝑟 2 is a better choice. low gray level. Hence
𝒔 = 𝒔𝒒𝒓𝒕(𝒓) is a
betterchoice.
What is the Full form of DWT?
A. Discrete Wavelet b. Digital Wavelet c. Design Wave
Transform Transform Transform
Identify the operator used for the conversion of an input image into the image shown in
following figure .

Input Image Output Image

a. Contrast Stretching b. Image Negative c. Thresholding
The 2D DFT of the image [1, 1; 1, 1] is_______________
[4,0; 0,0] [0,4; 0,0] [0,0;4,0]
The transform which is widely used to detect 'lines' in an image is______________
Fourier transform Hough transform Cosine transform
Which one of the following is a lossy coding?
Run length coding Uniform quantiser Huffman Coding
Which of the following filters will in general have the best performance in enhancing edges
in an image?
laplacian filter Median filter Mean filter
For the binary image shown below, calculate the number of white pixels left if it eroded by
the structuring element(shown in figure)

1 4 2
The minimum D4 and D8 distances between the pixels I(1,1) and I(3,4) are I= [ 1 2 4 8; 2 6
4 2; 1 3 4 5; 2 2 1 5]
3,5 6,4 5,3
You have an image f(m,n). on taking two times the Fourier transform, the result will be
f(2m,2n) f(-m,-n) 1/2f(m,n)
what does the definition of entropy tell us?
the lower bound to the upper bound to the average number
encode a source without encode a source of bits to encode a
distortion without distortion source given a
certain distortion
The edge detector which uses non-maxima suppression to thin edges is the
Canny operator Sobel operator Roberts operator
The parameter that may change if all the pixels in an image are shuffled is
histogram covariance entropy
An example of a dictionary-based coding technique is
Huffman coding Predictive coding LZW coding
In an image compression system, 16384 bits are used to represent a 128 X 128 image with
256 gray levels. what is the compression ratio for this system?
4 8 12
What is the full form of DCT?
Discrete Cos Transform Discrete Cosine Digital Cosine
Transform Transform
the conversion between RGB and CMY colour model is given by
[ R ; G ; B] = [C; M; Y] [R ; G ; B] = [ 1; 1; 1 [R ; G ; B] = [ 2; 2; 2
]- [ C; M; Y ] ]- [ C; M; Y ]
A 2X2 iamge f(m,n) = [2 4; 6 8] is passed through the linear filter h(m,n)= 1/2 [0 0 0 ; 1 0 1;
0 0 0 ]. Then the resultant image is (assume zero padding )
[6 8 ; 2 4] [ 2 6; 4 8] [2 1 ; 4 3]
the colour model which is more relevant to a display system is the
CMY Model RGB Model HSI Model
The mask which can be used to perform smoothing operation of an input image is
[ 1 -1 1; 1 -1 1; 1 -1 1] [1 1 1 ; -1 -1 -1; 1 1 1 ] [1 1 1; 1 1 1; 1 1 1]
 Below are the pixel values in a 5 X 5 gray-level image: [ 1 2 3 1 2; 4 5 2 3 3 ; 3 3 5 4 4 ; 1 3 2
3 5; 2 1 3 1 3]. What is the value of the 3rd row and 3rd column pixel after applying a 3 X 3
median filter?
2 4 3
A 4 X 4 image is given by [ 2 3 4 5; 1 2 4 6; 2 3 2 4; 1 5 7 6]. If this image is filtered by a Min
filter with a mask given below then the resultant image is (assume zero padding)
[0 0 0 0 ; 0 2 1 0 ; 0 4 2 0; 0 0 0 [ 0 0 0 0; 0 2 1 0; 0 3 2 0; 0 [0 0 0 0; 0 1 2 0 ; 0 2 3 0; 0
0] 0 0 0] 0 0 0]

Which of the following operations is idempotent?

opening dilation Median filter
You have a digital image that takes up 240kb. The spatial resolution of the image is given by
600 X 200. What is the bit depth?
5 4 2
Given an image of size 3 X 3 as f(m,n)= [ 128 212 255; 54 62 124; 140 152 156]. Determine
the output image g(m,n) using Logarithmic transformation by choosing C=1
g(m,n) =[224 247 256; g(m,n) = [3 3 3; 2 2 g(m,n)= [2 2 2; 2 2
185 191 223; 228 232 3; 3 3 3] 3; 3 3 3]
233]
The impulse response of a filter is given by

the filer will act like a

Low pass filter High pass filter Band pass filter
Let f(m,n) and h(m,n) represent 2D signals. If h(m,n) is a two-dimensional impulse then the
convolution of f(m,n) with h(m,n) will result in
f(m,n) h(m,n) f(m,n)+h(m,n)
the number of shades of gray in a six-bit image is
256 128 64
Answer the following questions( Any Ten)

1. Define m-adjacency between two pixels P and Q.

2. Define 4-connectivity between pixels.
3. Show how an average of a 2D function can be obtained using 2D Fourier Transform?
4. Explain the illumination and reflection components related to digital image.
5. Define the term thresholding.
6. What do you mean by a histogram of an image?
7. Explain Gaussian Noise. Also mention its mean and variance with respect to the gray level of an
image.
8. Explain contraharmonic mean filter. How it will be used for removal of noise?
9. Draw the mask of a box filter.
 10. Explain sampling and quantization for image data.
11. Define the Euclidean Distance between two points P and Q with coordinates of (x,y) and (s,t).

For the image segment shown in Figure 1 below, apply the zooming algorithm for increase the size of
the image segment by the factor of 2. Use the nearest neighbor interpolation technique for assigning a
gray value to the new pixels.
8 7 6 0
1 2 3 5
3 4 6 8
8 5 1 0
Figure 1 : image segment of 4 * 4
What is the advantage of homomorphic filtering; explain with the help of suitable derivations.
Apply the high pass mask on the image shown in Figure 2. Conclude about the result.
20 20 20 20 20
20 30 30 20 30
20 30 30 20 30
20 30 30 20 30
20 20 20 20 30
Figure 2 : image segment of 5 * 5
Consider the binary image given in figure 3 below. Show the result of 3 * 3 median filtering if the
following masks are used. (A ‘zero’ in a mask position means that the corresponding pixel is not used
for median calculation).

1 1 1 0 1 0
(a) 1 1 1 ,(b) 1 1 1
1 1 1 0 1 0

Figure 3: Binary Image of 7 * 10.

Explain the Wiener Filtering for image Restoration and discuss the problems associated with it. Also
discuss the remedy available in the form of Constrained Least Square Filtering Technique.
Answer the following questions( Any Ten)

 Define 4-connectivity between pixels.

 Show how an average of a 2D function can be obtained using 2D Fourier Transform?
 Explain the illumination and reflection components related to digital image.
 Define the term thresholding.
 What do you mean by a histogram of an image?
 Explain Gaussian Noise. Also mention its mean and variance with respect to the gray level of an
image.
 Explain contraharmonic mean filter. How it will be used for removal of noise?
 Draw the mask of a box filter.
 Explain sampling and quantization for image data.
 Define the Euclidean Distance between two points P and Q with coordinates of (x,y) and (s,t).

For the image segment shown in Figure 1 below, apply the zooming algorithm for increase the size of
the image segment by the factor of 2. Use the nearest neighbor interpolation technique for assigning a
gray value to the new pixels.
8 7 6 0
1 2 3 5
3 4 6 8
8 5 1 0
Figure 1 : image segment of 4 * 4
------------------------OR------------------------
What is the advantage of homomorphic filtering; explain with the help of suitable derivations.
Apply the high pass mask on the image shown in Figure 2. Conclude about the result.
20 20 20 20 20
20 30 30 20 30
20 30 30 20 30
20 30 30 20 30
20 20 20 20 30
Figure 2 : image segment of 5 * 5
Consider the binary image given in figure 3 below. Show the result of 3 * 3 median filtering if the
following masks are used. (A ‘zero’ in a mask position means that the corresponding pixel is not used
for median calculation).

1 1 1 0 1 0
(b) 1 1 1 ,(b) 1 1 1
1 1 1 0 1 0

Figure 3: Binary Image of 7 * 10.

Explain the Wiener Filtering for image Restoration and discuss the problems associated with it. Also
discuss the remedy available in the form of Constrained Least Square Filtering Technique.

Answer the following questions( Any Ten)

 Define m-adjacency between two pixels P and Q.

 Define 4-connectivity between pixels.
 Show how an average of a 2D function can be obtained using 2D Fourier Transform?
 Explain the illumination and reflection components related to digital image.
 Define the term thresholding.
 What do you mean by a histogram of an image?
 Explain Gaussian Noise. Also mention its mean and variance with respect to the gray level of an
image.
 Explain contraharmonic mean filter. How it will be used for removal of noise?
 Draw the mask of a box filter.
 Explain sampling and quantization for image data.
 Define the Euclidean Distance between two points P and Q with coordinates of (x,y) and (s,t).

For the image segment shown in Figure 1 below, apply the zooming algorithm for increase the size of
the image segment by the factor of 2. Use the nearest neighbor interpolation technique for assigning a
gray value to the new pixels.
8 7 6 0
 1 2 3 5
3 4 6 8
8 5 1 0
Figure 1 : image segment of 4 * 4
What is the advantage of homomorphic filtering; explain with the help of suitable derivations.
Apply the high pass mask on the image shown in Figure 2. Conclude about the result.
20 20 20 20 20
20 30 30 20 30
20 30 30 20 30
20 30 30 20 30
20 20 20 20 30
Figure 2 : image segment of 5 * 5
Explain spatial filtering technique in detail. Apply the box filter of size 3X3 on the input image shown
in Fig 1.

0 0 0 0 0 0

0 1 1 1 1 0

0 1 2 2 1 0

0 1 2 2 1 0

0 1 1 1 1 0

0 0 0 0 0 0

Fig 1: 6 X 6 Image segment

What do you mean by “Salt & Pepper” Noise? Explain the Median filtering in detail with suitable
example.

Compare median filtering and a low pass filtering on a noisy image. Which performs best (i.e. removal
of most noise with least blurring)? How? Explain with suitable example.

Explain the steps for Filtering of image in Frequency Domain. Perform Fourier Transform on the
following 2 X 2 image shown in Fig 2.

2 2
 2 2

Fig 2: image segment of 2 X 2

Explain the Wiener Filtering for image Restoration and discuss the problems associated with it. Also
discuss the remedy available in the form of Constrained Least Square Filtering Technique.

Apply the high pass mask (Laplacian operator of size 3X 3) on the image shown in Fig 3. Conclude
about the result.
20 20 20 20 20
20 30 30 30 20
20 30 60 30 20
20 30 30 30 20
20 20 20 20 20
Fig 3 : Image segment of 5 * 5

Answer the following:

1. Define D4-Distance between two pixels P and Q of coordinates (x,y) and (s,t) respectively.
2. Show that an edge spatial filter is equivalent to subtracting the original image from its
smoothed image.
3. Compare median filtering and a low pass filtering on a noisy image. Which performs best (i.e.
removal of most noise with least blurring)? How?
4. Explain Rayleigh noise. Also mention its mean and variance with respect to the gray level of an
image.
Explain Interpolation techniques in detail with suitable example.

Explain bit-plane slicing in detail. (i) What would be the effect on the histogram if we set to zero the
lower-order bit planes? (ii)What would be the effect on the histogram if we set to zero the higher-
order bit planes instead?

The input image f(x,y) is passed through a linear shift-invariant system h(x,y). determine the output
image if f(x,y) and h(x,y) are given below:

12 10 8 4 1 2 1
1
f(x,y) = 8 14 6 9 And ℎ 𝑥, 𝑦 = 16
2 4 2
5 9 13 8 1 2 1
14 5 7 9
Assume zero padding of the original image.

For the given image patch shown in following figure, compute the 2D DFT for the image patch. Then
reconstruct the original image patch by neglecting the last two coefficients in 2D DFT. Comment on the
observed result.
 12 4
5 10
Figure: 2 X 2 Image patch

Explain Histogram Equalization in detail.

Write a short note on homomorphic filtering. Explain the advantages of homomorphic filtering with
suitable derivations.

For the input image shown, obtain the output image after application of (1) A 3X3 Median filter ,(2) A 3
X 3 Max filter

10 10 10 10 10 10 10
10 20 20 255 20 20 10
10 20 0 50 50 0 10
10 20 50 50 50 20 10
10 20 20 20 20 255 10
10 10 10 10 10 10 10
If the template T and the image I is defined as follows:

1 0 1 1 3 3

T= 0 1 I= 1 1 4 4

2 1 3 3

1 1 1 4

Find the resulting convolved image, O= T * I. Where, *- convolution operator

In a given application, an averaging mask is applied to input images to reduce noise and then a
Laplacian mask is applied to enhance small details. Would mathematics predict that the result should
be the same if the order of the operations were reversed?

Do the Following:

1) What is the value of the marked pixel after convolution with 3 X 3 point operator?
2) What is the value of the marked pixel after convolution with 3 X 3 horizontal line operator?
3) What is the value of the marked pixel after convolution with 3 X 3 operator for 45 degree line
detection?
4) What is the value of the marked pixel after convolution with 3 X 3 operator for 45 degree line
detection and thresholding with the maximum difference?
2 2 2 2 2

1 12 1 2 3

12 12 12 12 12

5 5 5 12 1
 2 2 2 2 2

List down different types of noises which can be found in an image. How can we estimate mean and
variance of a noisy image? Explain the MMSE filter to remove the noise from the image.