2023-24 First Sem - Computer Vision Mid Sem Regular
2023-24 First Sem - Computer Vision Mid Sem Regular
Note to Students:
1. Answer all the questions.
2. Write your name and sign at the end of all the pages.
3. Assumptions made, if any, should be stated clearly at the beginning of your answer.
2 2 4 4
4 4 2 2
3 3 5 5
5 5 3 3
Page (1/2)
Q4) [Write the answer in A4 Sheet, Scan & Upload]
4 4 2 2
3 3 5 5
5 5 3 3
(a) Compute the direction of the edge at location [2,2 ], that is the highlighted location with the
pixel value 5. Use Prewitt’s operator to find the required gradients. Also, write down the
strength of the edge at the highlighted pixel. [3.0 M]
(b) Write down ways of computing approximation of edge gradients and magnitude if efficiency
is of prime requirement. [2.0 M]
(a) How does canny’s edge detection algorithm compute the derivative of convolution? What is
the reason for this approach? What is the benefit? [2 M]
(b) Explain the Non-maxima suppression algorithm as used in Canny’s edge detection
algorithm with all the steps. Why do we perform this? [2 M]
(c) What is the purpose of double thresholding in Cannys Edge Detection? How to choose two
thresholds? [1 M]
[2.0 + 2.0 +1.0 = 5 Marks]
(a) What are local features? What are the benefits of using local features in the computer
vision pipeline? In what way are global features and local features different? [3 M]
(b) What are the steps in Harris Corner Detector? Explain briefly. [2 M]
[3.0 + 2.0 = 5 Marks]
Page (2/2)