SETHU INSTITUTE OF TECHNOLOGY, KARIAPATTI

(An Autonomous Institution Affiliated to Anna University, Chennai)

Regulation 2019(Question Bank)
Department: Biomedical Engineering Subject name : Image Processing Techniques

Subject code : 19UBM703 Question Pattern : Part A: (2 x 10 = 20 marks)

Part B: (5 x 16 = 80 marks)
Course Coordinator : Mrs. Mary James Time Duration : 3 hours

Approved by: HoD/BME

Font size : Times new roman :12 Line spacing: 1.15

PART – B ( 2 Marks)
UNIT - I (Minimum 8 Questions)
1. State the primary application of the DCT (Discrete Cosine Transform) in image
CO2 A -
compression?
2. Explain the significance of the HSI color model in color image processing. CO1-U -
3. Define image sampling and explain its role in digital image processing. CO1-U -
4. Compute the distance between the two pixels q(1,1) p (2,2) using the city block CO2 A -
distance measurement.
5. Describe the distribution of cones and rods in the human retina. Are they CO2 A -
evenly distributed, or is there a specific pattern?
6. Use a step-by-step diagram to depict the progression of the Mach band effect as CO3-An -
it manifests in an image with varying levels of intensity.
7. How is image acquisition done in a digital image processing? CO1-U -
8. Write the transform pair for 3D-DFT. CO1-U -
9. Explain the significance of image acquisition in biomedical image processing CO1-U -
in two sentences.
UNIT - II (Minimum 8 Questions)
1. Demonstrate the formula for the transfer function of a Butterworth high-pass CO1-U -
filter.

2. Specify the objective of an image enhancement technique in two sentences. CO1-U -

3. Demonstrate the two categories of image enhancement and briefly explain CO1-U -
each.
4. Define contrast stretching in the context of image processing. CO1-U -
5. Spatial filters are also called averaging filters. Provide a reason for this CO3-An -
nomenclature in two sentences.
6. Define intensity level slicing and explain its purpose in image processing. CO1-U -
7. Sketch a block diagram for homomorphic filtering, highlighting its key CO3-An -
components.
 8. Discuss the concept of histogram equalization and its role in enhancing image CO2-A -
details and improving visualization.

UNIT - III (Minimum 8 Questions)

1. Sketch the basic components of the model of the image degradation/restoration CO3-An -
process in digital image processing.
2. Describe the different types of noise commonly found in images. CO3-An 3
3. Why is it important to estimate the degradation function in image restoration? CO2-A 2
4. Write down the mathematical formula for Salt and pepper noise. CO1-U -
5. Provide a definition of inverse filtering in the context of image restoration. CO1-U -
6. Discuss the characteristics of uniform noise and its impact on image quality. CO1-U -
7. State the significance of the signal-to-noise ratio (SNR) in the Wiener filter? CO1-U -
8. State the probability density function of exponential noise. CO1-U -
9. Define Rayleigh noise. CO1-U -
UNIT - IV (Minimum 8 Questions)
1. Explain thresholding in image segmentation. CO1-U -
2. Classify the types of discontinuity in digital image? CO2-A 2
3. Demonstrate the steps involved in splitting and merging? CO1-U -
4. Define medical image segmentation and explain why it is essential CO1-U -
in the field of healthcare.
5. Explain the concept of image representation schemes and their importance in CO1-U -
image analysis.
6. Compare an edge with line in an image. CO3-An 3
7. Define region growing? CO1-U -
8. Illustrate the steps in polygonal approximation method. CO2-A 2
9. How are the derivatives obtained in edge detection during CO3-An 3
formulation?
UNIT - V (Minimum 8 Questions)
1. Define image compression? CO1 U 2
2. Illustrate the need for Compression? CO3 An 3
3. Classify the different Compression Methods? CO2 A -
4. Define Huffman coding CO1 U 2
5. Define bit plane coding. CO1 U -
6. Define run length coding. CO1 U 2
7. Classify the types of decoder? CO2 A 3
8. Sketch the two equivalent representations of the perception model for two CO3 An 3
pattern classes.

PART – C
UNIT - I (Minimum 5 Questions either or choice )
1. (a) Classify the different transforms used in digital image CO3- An PO 1 (16)
processing. Explain in detail.

Or
(b) Illustrate the conversion of RGB color model to HSI model and CO3- An PO 1 (16)
vice versa in detail.
2
2. (a) Define digital image processing. Explain various functional block CO1- U PO 1 (16)
of digital image processing with diagram.

Or
(b) Explain the concept of spatial domain sampling in digital images CO1- U PO 1 (16)
and how it relates to pixel resolution and image size.Describe the
process of quantization in image processing, highlighting its role in
reducing continuous intensity values to discrete levels.

3. (a) Analyze the concept of pixel connectivity in digital images, and CO3 - An PO 3 (16)
discuss how it influences the representation and processing of
objects in an image. Explain the basic relationships between
pixels in detail.

Or
(b) Analyze the elements of visual perception in the context of human CO3 - An PO 3 (16)
vision and describe the process of image formation in the eye.

4. (a) Explain the color model which is suitable for hardware CO1- U PO 1 (16)
implementation with neat diagram.
Or
(b) i) Explain Discrete Fourier Transform (DFT) of an image in detail. CO1- U PO 1 (16)
ii) Explain Discrete Cosine Transform of an image in detail.

5. (a) Analyze the fundamental differences between the RGB and HSI CO3 - An PO 3 (16)
color models, explaining how they represent color information and
their respective advantages in different applications.
Or
(b) Sketch and illustrate the elements of visual perception and also CO3 - An PO 3 (16)
describe image formation in the eye.

UNIT - II (Minimum 5 Questions either or choice )

1. (a) Evaluate the impact of different histogram equalization algorithms CO3 - An PO 3 (16)
on images with varying statistical distributions and assess their
ability to improve contrast.
Or

(b) Utilize a point processing technique of your choice to enhance the CO3 - PO 3 (16)
An
visibility of faint structures in astronomical images. Discuss the various
point processing techniques in detail.

2. (a) Illustrate the design of image sharpening and smoothing filters in CO3 - An PO 3 (16)
frequency domain.
3
 Or
(b) Analyze the basic intensity transformation functions along with CO3 An (16) (16)
diagram.

3. (a) Write short notes on the following: CO1 - U PO 1 (16)

i) Contrast stretching
ii) Intensity level slicing
iii) Bit level slicing
iv) Image negatives
Or
(b) Discuss the image smoothing and sharpening filters with its model in CO1- U (16) (16)
the spatial domain.

4. (a) Illustrate the mathematical principles behind homomorphic filtering CO3 - An PO 3 (16)
and how it enables the separation of illumination and reflectance
components in an image. How is the Fourier transform utilized in this
process?
Or
(b) Illustrate the design of image sharpening and smoothing filters in CO3 - An PO (16)
spatial domain. 3

5. (a) Explain histogram and Illustrate the steps in histogram equalization of CO1 - U PO (16)
the image. 1
Or

(b) With the help of a block diagram, illustrate the principle of CO1 - U PO (16)
homomorphic filtering. 1
UNIT - III (Minimum 5 Questions either or choice )

1. (a) Evaluate the impact of inverse filtering advancements on fields CO3 - An PO 3 (16)
such as medical diagnostics, forensics, and remote sensing,
highlighting both the benefits and potential risks.
Or
(b) Analyze how frequency domain filtering techniques can be used to CO3 - An PO 1 (16)
reduce periodic noise in images? Provide a brief overview of the
approach.

2. (a) Evaluate the advantages and disadvantages of using order statistics CO3 - An PO 1 (16)
filters in image processing applications, comparing them with
other filtering methods such as median filters and Gaussian filters.
Or
(b) Illustrate the minimum mean square error filtering in detail. CO3 - An PO 3 (16)

3. (a) Analyze the different types of noise models along with their CO3 - An PO 3 (16)
probability density functions.
Or
4
 (b) Design weiner filtering approach for segmentation in detail. CO3 - An PO 3 (16)

4. (a) Analyze the statistical properties of mean filters, including their CO3 - An PO 3 (16)
ability to reduce different types of noise (e.g., Gaussian, salt-and-
pepper) and their impact on image features and texture. Compare
and contrast these properties with other noise reduction
techniques.
Or
(b) Classify the different types of order statistics filtering in detail. CO3 - An PO 3 (16)

5. (a) Demonstrate the different types of mean filtering in detail. CO3 - An PO 3 (16)
Or
(b) Illustrate the minimum mean square error filtering in detail. CO3 - An PO 3 (16)

UNIT - IV (Minimum 5 Questions either or choice )

1. (a) Illustrate how hough transforms can be used to link edges. CO3 - An PO 3 (16)
Or
(b) Illustrate the techniques used for region representation in detail. CO3 - An PO 3 (16)

2. (a) Examine region based segmentation with relevant examples. CO3 - An PO 3 (16)
Or
(b) Illustrate with suitable examples how detection of edges done in CO3 - An PO 3 (16)
medical images are done.

3. (a) Demonstrate how edge detection is performed? Explain hough CO3 - An PO 3 (16)
transform and discuss how the edge points are linked.
Or
(b) Classify and explain the various image representation approaches in CO3 - An PO 3 (16)
detail.

4. (a) Examine image segmentation based on various thresholding CO1 - U PO 1 (16)

techniques.
Or
(b) Assess how an image is segmented using region splitting and merging CO1 - U PO 1 (16)
algorithm in detail and how the segmented object is represented by
chain codes.

5. (a) Illustrate how hough transforms can be used to link edges. CO1 - U PO 1 (16)
Or
(b) Examine region based segmentation and region growing with an CO1 - U PO 1 (16)
example.
5
 UNIT - V (Minimum 5 Questions either or choice )

1. (a) Explain the fundamental concepts of image recognition, emphasizing CO1 - U PO 1 (16)
the role of patterns and pattern classes. Provide examples to illustrate
your explanation.
Or
(b) Explain the fundamental principles of image compression. Differentiate CO1 - U PO 1 (16)
between lossless and lossy compression techniques, focusing on their
applications and trade-offs.

2. (a) Critically assess the optimality of variable-length coding methods such CO3 - An PO 3 (16)
as Huffman and Arithmetic coding in terms of compression efficiency.
Discuss scenarios where one might be preferred over the other.
Or
(b) Propose a scenario where an optimum statistical classifier would be the CO3 - An PO 3 (16)
most suitable image recognition method.

3. (a) Compare and contrast decision theoretic methods, minimum distance CO3 - An PO 3 (16)
classifiers, and matching by correlation in the context of image
recognition. Assess the strengths and weaknesses of each approach.
Or
(b) Develop a step-by-step explanation of how transform coding works in CO3 - An PO 3 (16)
image compression. Use an example to illustrate the transformation
process and its impact on the image.

4. (a) Write short notes on the following in detail, CO1 - U PO 3 (16)

i . Patterns and Pattern classes
ii . minimum distance classifier
Or
(b) Explain the need of image compression. How is run length encoding CO1 - U PO 3 (16)
approach used for compression?it is lossy? Justify.

5. (a) Explain the need of image compression. How are bit plane coding and CO1 - U PO 3 (16)
constant area coding approaches used for compression?
Or
(b) Write short notes on the following in detail, CO1 - U PO 3 (16)
i . Matching by correlation,
ii . Optimum statistical classifier.