Week 1 >>> Basic image processing >>> Deep Learning in Computer Vision

1.

Question 1

Consider linear contrast correction. Given y_min = 10 and y_max = 207, compute value of pixel with y = 54 after correction. Round result to the nearest integer.

For example, if you get 90.5, type 91

1 point

57

——————————————————————————————–

2.

Question 2

Convolve a small (4×4 pixels) image I with a 2×2 kernel K and write result. Don’t use any padding for convolution. Input integer output elements only. Place result of convolution in the variable answer.

K = [[4, 3], [7, 2]]

1 point

answer = [[84, 48, 69], [73,76,38], [75, 63, 41]]

——————————————————————————————–

3.

Question 3

You are given images with unknown gamma correction parameters. Choose appropriate gamma for each image. Available gamma values are 0.5, 0.75, 1, 1.5, 2. Type in a comma-separated list of gamma values corresponding to the following list of images. Example of answer: 2, 1.5, 1, 0.75, 0.5

1 point

0.5, 1.5, 2, 1, 0.75

——————————————————————————————–

4.

Question 4

How many convolution operations are needed for canny edge detector?

1 point

3

——————————————————————————————–

5.

Question 5

Imagine that you want to store color images with alpha (opacity) channel. Four numbers are used for every pixel: three color values (R, G, B) from range [0..31] and one opacity value from range [0..63]. How many bits are used for every pixel?