Vimal Jyothi Engineering College

(Autonomous)
DEPARTMENT OF
VIIIth Semester - [Link]
Course : CST474 - COMPUTER VISION

QNo QUESTION CO MARK

Module 1
Submodule 1

1. Discuss rotation matrices and translation vectors with suitable examples. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )
Topic 1

2. Explain the extrinsic camera parameters and derive the transformation from world to
CO1 (14.00)
camera coordinates.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

3. With diagrams, explain how sensor geometry and pixel size influence intrinsic
CO1 (14.00)
parameters.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

4. Discuss the limitations of the pinhole camera model. How do thin lenses overcome
CO1 (14.00)
these limitations?
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Question Paper ) (Question Paper 2019 )

5. With a neat diagram, explain the working of a pinhole camera and derive the
CO1 (14.00)
perspective projection equations.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

6. Define intrinsic camera parameters with examples. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Medium ) (Ref : Book ) (Textbook )

7. Explain the concept of perspective projection in a pinhole camera. CO1 (3.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

8. Define the pinhole camera model and list any three of its characteristics. CO1 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 2

9. Explain world coordinate system vs. camera coordinate system. CO1 (3.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

10. What are rotation and translation vectors? CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Topic 1

11. Explain homogeneous coordinate transformation used in camera geometry. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

12. Define extrinsic camera parameters. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 3
Topic 1

1 of 15
13. With diagrams, describe calibration using planar patterns. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

14. Explain the steps involved in estimating intrinsic and extrinsic parameters. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

15. Differentiate between intrinsic and extrinsic calibration. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

16. Explain the linear approach to camera calibration with mathematical formulation. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

17. What is the purpose of calibration targets? CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textboook )

18. Define camera calibration. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 4
Topic 1

19. Explain 2D convolution with suitable examples and diagrams. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

20. Discuss the effects of using different linear filters (average, Gaussian, sharpening) on
CO1 (14.00)
images.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

21. Explain linear filtering and derive the convolution operation for images. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

22. State the properties of linear filters. CO1 (3.00)

( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) (Textbook )

23. What is convolution? Provide its mathematical expression. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

24. Define linear filtering in image processing. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 5
Topic 1

25. Explain discrete convolution for 1D and 2D signals with examples. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

26. Discuss the relationship between impulse response, convolution, and system output. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

27. Explain shift invariant systems and derive the discrete convolution operation. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

28. Write a note on impulse response of LSI systems.

CO1 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

29. What is a linear shift invariant (LSI) system? CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

30. Define shift invariance. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 6
Topic 1

31. With diagrams, illustrate how correlation-based pattern detection works. CO1 (14.00)
( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) (Textbook )

2 of 15
32. Discuss the steps involved in template matching using correlation measures. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textook )

33. Explain normalized cross-correlation and its application in pattern matching. CO1 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

34. State the difference between correlation and convolution. CO1 (3.00)
( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) (Textbook )

35. What is template matching? CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

36. Define normalized correlation. CO1 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Module 2
Submodule 1
Topic 1

37. Explain the mathematical formulation of image gradients and analyze their role in
CO2 (14.00)
feature extraction.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

38. Explain how image gradients are used to detect edges, including preprocessing and
CO2 (14.00)
postprocessing steps.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

39. (a) Explain the principle of gradient-based edge detection. (7)

CO2 (14.00)
(b) Explain Sobel edge detection algorithm. (7)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

40. Why is gradient important in feature detection? CO2 (3.00)

( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

41. Write the formula for gradient magnitude and direction. CO2 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

42. Define image gradient. CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 2
Topic 1

43. How does epipolar geometry reduce correspondence search? CO2 (14.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

44. Derive the epipolar constraint and explain its role in stereo image rectification. CO2 (14.00)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) (Textbook )

45. a) Explain binocular camera geometry. (7)

CO2 (14.00)
(b) Explain epipolar geometry with diagram. (7)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

46. What is the Harris corner detector? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

47. Explain the concept of corner detection in computer vision. CO2 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

48. What is a corner point? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 3
Topic 1

3 of 15
49. a) Explain the conditions under which the Essential Matrix can be derived from the
Fundamental Matrix. (7) CO2 (14.00)
b) Explain the role of camera calibration parameters in forming the Essential Matrix. (7)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

50. a) Derive the Fundamental Matrix for a pair of uncalibrated stereo cameras.
(7 Marks)
CO2 (14.00)
b) Explain how the Fundamental Matrix establishes the relationship between points in
two images. (7 Marks)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) (Textbook )

51. (a) Derive essential matrix. (7)

CO2 (14.00)
(b) Explain its role in stereo vision. (7)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

52. Write the epipolar constraint equation. CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

53. What is a fundamental matrix? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

54. What is an essential matrix? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 4
Topic 1

55. Discuss challenges in binocular reconstruction. CO2 (14.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

56. a) Explain disparity computation. (7)

CO2 (14.00)
b) Explain depth estimation. (7)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

57. Explain the 3D reconstruction process using stereo images.

CO2 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

58. Define disparity. CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

59. What is binocular reconstruction? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 5
Topic 1

60. a) Explain the role of similarity measures in local stereo matching. (7 Marks)

b) Discuss the advantages and limitations of local stereo matching methods. (7 Marks) CO3 (14.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

61. a) Explain the impact of window size selection in local stereo matching methods. (7

Marks)
CO2 (14.00)
b) Explain how occlusion and depth discontinuities affect local stereo matching. (7
Marks)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

4 of 15
62. Explain window-based stereo matching. CO2 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

63. Give one example of a local fusion method. CO2 (3.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

64. What is a local matching method? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

65. What is binocular fusion? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

66. a) Explain local methods for binocular fusion. (7 Marks)

CO2 (14.00)
b) Explain block matching. (7 Marks)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )
Submodule 6
Topic 1

67. a) Demonstrate the use of local binocular fusion methods for disparity estimation in a
stereo vision system. (7 Marks)
CO2 (14.00)
b) Illustrate how global binocular fusion methods generate a consistent disparity map

using optimization techniques. (7 Marks)

( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) (Textbook )

68. Explain Energy-Based Stereo Matching in Global Methods. CO2 (14.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

69. Explain Global Methods for Binocular Fusion. CO2 (14.00)

( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

70. What is the objective of global binocular fusion? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

71. Explain energy function in global stereo vision? CO2 (3.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

72. What is meant by global stereo matching? CO2 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Module 3
Submodule 1
Topic 1

73. a) Discuss different image segmentation techniques. (7 Marks)

b) Explain the criteria used to evaluate the quality of image segmentation results. (7 CO3 (14.00)
Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) (Textbook )

74. Illustrate the complete image segmentation pipeline for a given image, explain each
CO3 (14.00)
stage is applied in practice.
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) (Textbook )

75. Explain how image segmentation acts as a bridge between low-level image processing
CO3 (14.00)
and high-level vision tasks.
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook )

76. List applications of image segmentation. CO3 (3.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook )

5 of 15
77. What is the goal of segmentation? CO3 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )

78. What is image segmentation? CO3 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 2
Topic 1

79. a) Explain the working principle of background subtraction models used in video
analysis. (7 Marks)

b) Explain interactive segmentation techniques and the role of user input in improving CO3 (14.00)

segmentation accuracy. (7 Marks)

( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

80. Explain interactive segmentation techniques.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

81. a) Explain background subtraction for motion detection. (7 Marks)

b) Explain the challenges faced in background subtraction and the methods used to
CO3 (14.00)
handle them. (7 Marks)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean
Ponce. Computer vision: A modern approach. Prentice hall, 2011. )

82. Give one application of background subtraction. CO3 (3.00)

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook 1 )

83. What is interactive segmentation? CO3 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook 1 )

84. What is background subtraction? CO3 (3.00)

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook )
Submodule 3
Topic 1

85. Compare region-based and edge-based segmentation.

CO3 (14.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

86. a) Illustrate the region growing approach for image segmentation by applying it to a
sample image. (7 Marks)
b) Apply suitable homogeneity criteria to justify region formation in region-based CO3 (14.00)

segmentation. (7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

87. Explain region-based image segmentation.

( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean
CO3 (3.00)
Ponce. Computer vision: A modern approach. Prentice hall, 2011. )

88. Explain the concept of region merging?

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

89. What is region growing?

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

6 of 15
90. What is a region in an image?
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 4
Topic 1

91. a) Explain the Watershed Segmentation Algorithm with suitable illustrations. (7 Marks)
b) Explain the role of marker-based watershed segmentation in improving
CO3 (14.00)
segmentation accuracy. (7 Marks)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

92. a) Illustrate the step-by-step working of the Watershed Segmentation algorithm for a

given image. (7 Marks)

b) Demonstrate how over-segmentation arises in watershed segmentation and apply CO3 (14.00)

suitable techniques to control it. (7 Marks)

( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

93. Explain the Watershed algorithm for image segmentation.

CO3 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean
Ponce. Computer vision: A modern approach. Prentice hall, 2011. )

94. Why does over-segmentation occur?

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

95. What is a catchment basin?

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

96. Explain the topographic interpretation used in watershed segmentation.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 5
Topic 1

97. Derive the Optical Flow Constraint Equation and explain the Aperture Problem.
CO3 (14.00)
( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

98. Explain the assumptions used in optical flow estimation.

CO3 (14.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

99. Explain different Optical Flow Models and their assumptions.

CO3 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean
Ponce. Computer vision: A modern approach. Prentice hall, 2011. )

100. Explain the brightness constancy assumption used in optical flow.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

101. What is motion segmentation? Explain how optical flow helps in it.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

102. Define optical flow and explain its significance in motion analysis.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 6
Topic 1

7 of 15
103. a) Compare Single-Motion Segmentation and Layered-Motion Segmentation.
(7 Marks)
b) Explain the advantages of layered motion segmentation in complex dynamic scenes. CO3 (14.00)

(7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

104. a) Illustrate the process of motion segmentation using layered motion representation
for a dynamic scene. (7 Marks)
b) Demonstrate how layered motion representation is used to handle occlusion and CO3 (14.00)
multiple moving objects in motion analysis. (7 Marks)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

105. a) Illustrate the use of flow models in motion analysis for representing object motion. (7
Marks)
b) Apply the assumptions of flow models to real-world motion scenarios and analyze CO3 (14.00)

their limitations. (7 Marks)

( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

106. Differentiate between single motion model and layered motion model.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

107. What is layered motion segmentation? Why is it needed?

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

108. Explain the concept of flow models in motion analysis.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Textbook1 :Forsyth, David, and Jean Ponce.
CO3 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Module 4
Submodule 1
Topic 1

109. Illustrate the steps involved in building an image classification system.

CO3 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

110. a) Formulate a linear dynamical model for object motion. (7 Marks)

b) Explain its use in prediction during tracking. (7 Marks) CO4 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

111. a) Illustrate how linear dynamical models represent motion of objects. (7 Marks)
b) Explain how prediction using a linear dynamical model helps estimate the future
CO4 (14.00)
position of a moving object. (7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

112. State any three real-world applications of image classification. CO4 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) (Application 1 (1), App 2 (1), App 3 (1) )

8 of 15
113. Differentiate between two-class and multiclass classification.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

114. Define classification in computer vision and state its purpose.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 2
Topic 1

115. Apply suitable techniques to control overfitting in image classification tasks. CO4 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

116. a) Demonstrate how cross validation helps in selecting the best classifier model. (7
Marks)
b) Explain how training, validation and testing datasets are used to evaluate model CO4 (14.00)

performance. (7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

117. a) Illustrate different types of classification errors. (7 Marks)

b) Apply regularization to reduce overfitting. (7 Marks) CO4 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

118. Explain the purpose of cross validation in model selection.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

119. What is regularization and why is it used?

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

120. Define overfitting in classification problems.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 3
Topic 1

121. a) Illustrate how tracking systems handle motion variation in practical videos. (7 Marks)
b) Explain how adaptive models help maintain tracking accuracy during scale and
CO4 (14.00)
appearance changes. (7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

122. a) Apply suitable tracking strategies for monitoring a moving object in a scene. (7
Marks)
b) Explain how appearance and motion cues are used together to improve tracking CO4 (14.00)

performance. (7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

123. a) Illustrate the complete tracking pipeline in video sequences. (7 Marks)

b) Demonstrate tracking by detection framework. (7 Marks) CO4 (14.00)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

9 of 15
124. Explain the concept of tracking by detection.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

125. State three simple tracking strategies used in video analysis.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce. CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

126. Define object tracking in computer vision.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 4
Topic 1

127. a) Apply linear dynamical models for motion prediction in video tracking systems. (7
Marks)
b) Explain how model parameters are estimated and updated during tracking. (7 CO4 (14.00)

Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

128. a) Formulate a linear dynamical model for object motion. (7 Marks)

b) Demonstrate prediction in tracking using the model. (7 Marks) CO4 (14.00)

( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

129. a) Illustrate how linear dynamical models represent object motion. (7 Marks)
b) Explain the role of prediction in tracking the future position of a moving object. (7
CO4 (14.00)
Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

130. Mention two applications of linear dynamical models.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

131. State the role of state space representation in tracking.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

132. What is a linear dynamical system model?

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 5

133. Explain the difference between the prediction step and update step in the Kalman filter.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Topic 1

134. a) Apply Kalman filtering framework in real-time tracking applications. (7 Marks)

b) Explain how process noise and measurement noise influence Kalman filter
CO4 (14.00)
performance. (7 Marks)
( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

10 of 15
135. Illustrate how the Kalman filter balances prediction and measurement to estimate
object position. CO4 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

136. a) Illustrate Kalman filtering concept in object tracking. (7 Marks)

b) Explain the assumptions of the Kalman filter in a practical tracking environment. (7
CO4 (14.00)
Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

137. Explain the roles of the state vector and covariance matrix in the Kalman filter.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

138. What is meant by prediction and correction in Kalman filtering?

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 6
Topic 1

139. Demonstrate Kalman filtering algorithm flow with suitable tracking scenario.
CO4 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

140. Illustrate the role of state transition and measurement models in Kalman filter based

motion estimation. CO4 (14.00)

( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

141. a) Illustrate Kalman filter prediction step with equations. (7 Marks)

b) Derive the Kalman filter update (correction) step using the measurement equations.
CO4 (14.00)
(7 Marks)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

142. Mention the role of covariance matrix in Kalman filtering.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

143. Define state vector and measurement vector.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO4 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

144. What are the main steps in Kalman filter algorithm?

CO4 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )
Module 5
Submodule 1
Topic 1

145. Apply multi-scale sliding window detection for pedestrian detection using HOG
features. CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

11 of 15
146. a) Demonstrate object detection using sliding window approach for detecting
pedestrians in images. (7 Marks)
b) Explain how multi-scale scanning and Non-Maximum Suppression improve detection CO5 (14.00)
accuracy in sliding window methods. (7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

147. a) Illustrate the working of the sliding window technique in object detection. (7 Marks)
b) Demonstrate how multi-scale detection is achieved using sliding windows. (7 Marks) CO5 (14.00)
( Theory ) ( Applying(P) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

148. Mention two challenges of sliding window object detection.

CO5 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

149. State the role of window size in sliding window detection.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

150. Define the sliding window method for object detection.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 2

151. What is invariance in object recognition? Give an example.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Topic 1

152. Apply object recognition methods in an autonomous vehicle scenario and explain how
real-time models like YOLO support recognition. CO5 (14.00)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

153. a) Illustrate the difference between feature-based recognition and deep learning based
recognition. (7 Marks)
b) Demonstrate the use of CNN models such as VGG or ResNet for object recognition. CO5 (14.00)

(7 Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

154. a) Illustrate the complete object recognition pipeline (features → classifier). (7 Marks)
b) Demonstrate the role of invariance properties. (7 Marks) CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

155. State the role of training datasets in object recognition systems.

( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

156. Mention two feature descriptors used in object recognition.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce. CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

Submodule 3
Topic 1

12 of 15
157. Apply binocular stereo vision for obstacle avoidance and path planning in an
autonomous mobile robot. CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

158. a) Illustrate stereo matching algorithms used for robot navigation. (7 Marks)

b) Demonstrate the role of epipolar geometry in improving stereo matching accuracy. (7

CO5 (14.00)
Marks)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

159. a) Illustrate the steps involved in disparity map computation in stereo vision. (7 Marks)
b) Demonstrate how depth estimation is obtained from disparity for robot navigation. (7
CO5 (14.00)
Marks)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

160. Mention three practical challenges in using stereo vision for navigation.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

161. What is a disparity map and how does it help in depth estimation?
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

162. Define binocular stereo vision and state its role in robot navigation.
CO5 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 4
Topic 1

163. Explain how a face detection system works in a real-time attendance or security
system. CO5 (14.00)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean
Ponce. Computer vision: A modern approach. Prentice hall, 2011. )

164. a) Illustrate the use of HOG features for face detection. (7 Marks)
b) Demonstrate how SVM classifiers are applied for detecting faces. (7 Marks) CO5 (14.00)
( Theory ) ( Applying(P) ) ( Hard ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

165. a) Illustrate the Viola–Jones face detection algorithm with its main stages. (7 Marks)
b) Demonstrate how cascade classifiers improve detection speed. (7 Marks) CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

166. Mention any three challenges faced in real-world face detection systems.
CO5 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

167. What is the role of Haar features in the Viola–Jones face detection method?
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

168. Define face detection and state its importance in computer vision applications.
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 5
Topic 1

13 of 15
169. Mention two cues used for tracking people in videos.
CO5 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

170. Define human activity recognition in computer vision.

CO5 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

171. a) Explain how a face recognition system works in a smartphone unlocking system. (7
Marks)
b) Explain the role of feature extraction and matching in ensuring secure face CO5 (14.00)

verification. (7 Marks)
( Theory ) ( Understanding(U) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean
Ponce. Computer vision: A modern approach. Prentice hall, 2011. )

172. a) Apply deep learning-based face recognition using embeddings for access control
systems. (7 Marks)

b) Explain how training data and threshold selection influence the performance of CO5 (14.00)

embedding-based face recognition. (7 Marks)

( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

173. a) Illustrate the Eigenfaces approach for face recognition. (7 Marks)

b) Demonstrate how feature matching is performed for recognizing faces. (7 Marks) CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

174. What is the Eigenfaces method used in face recognition?

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

175. Differentiate between face detection and face recognition.

( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
CO5 (3.00)
Computer vision: A modern approach. Prentice hall, 2011. )

176. Define face recognition and state its purpose.

CO5 (3.00)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )
Submodule 6
Topic 1

177. a) Explain how tracking people is used in surveillance or smart home monitoring
systems. (7 Marks)
b) Explain the challenges involved in tracking multiple people in real-world video CO5 (14.00)
scenes. (7 Marks)
( Theory ) ( Understanding(U) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

178. Apply optical flow and trajectory-based features for recognizing human actions in
videos. CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

179. a) Illustrate activity recognition using spatio-temporal features. (7 Marks)

b) Demonstrate how tracking helps improve activity recognition accuracy.(7 Marks) CO5 (14.00)
( Theory ) ( Applying(P) ) ( Medium ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )
14 of 15
180. What is multi-object tracking?
CO5 (3.00)
( Theory ) ( Remembering(R) ) ( Easy ) (Ref : Book ) ( Textbook1 :Forsyth, David, and Jean Ponce.
Computer vision: A modern approach. Prentice hall, 2011. )

15 of 15