Scribd
Upload
128 views

Lab Report Experiment 3

This document provides a summary of a digital image processing class test consisting of 4 questions worth a total of 50 marks. Question 1 involves applying contrast stretching and histogram normalization to images. Question 2 applies filters to images and analyzes the results. Question 3 makes deductions about images based on their Fourier transforms. Question 4 derives an edge detection filter from the Gaussian function.

Uploaded by

Jake Dyson
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
128 views

Lab Report Experiment 3

This document provides a summary of a digital image processing class test consisting of 4 questions worth a total of 50 marks. Question 1 involves applying contrast stretching and histogram normalization to images. Question 2 applies filters to images and analyzes the results. Question 3 makes deductions about images based on their Fourier transforms. Question 4 derives an edge detection filter from the Gaussian function.

Uploaded by

Jake Dyson
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

DIGITAL IMAGE PROCESSING

CLASS TEST 16 March 2016

1 hour
Answer ALL the questions. Marks are indicated for each question. The test is out of 50
marks. You may use your notes.

QUESTION ONE:
(a) The image below has pixels with dynamic range 0 to 15. One pixel should be
regarded as an outlier. Deal with that pixel, and then apply a contrast stretch
to the image. [5]

12 10 9 8
11 11 13 13
12 10 0 8
8 12 10 13

(b) Apply a histogram normalization to the image in (a). [3]


(c) It is desired to compare structural features in two images independent of
illumination effects. Would you recommend that a contrast stretch or a
histogram normalization be applied to the images before comparison? Explain
your answer. [3]

TOTAL 11 MARKS.

1
Page 2 of 3
QUESTION TWO:
(a) Apply the 3x3 filter below:

0 -1 0

-1 5 -1

0 -1 0

to the toy image:

0 1 3 1 3
2 1 2 6 4
0 2 7 6 4
3 7 6 5 7
7 7 6 6 7

What is this filter supposed to do to an image?


Comment on the result. [6]
(b) Find the result of applying the Kuwahara filter to the image in (a) above for
the central pixel in the image.
Was the result what you expected? Explain. [8]

TOTAL 14 MARKS.

QUESTION THREE:
(a) You suspect that a scanner has blurred an image in some direction. The
Fourier transform of the image consists of the usual bright cloud around the
origin, but the cloud is elliptical with the long axis of the ellipse at 45 degrees.
What do you deduce? [3]
(b) When you look at the Fourier transform of a certain image, you are surprised
to see that it consists entirely of values along the x axis. What do you deduce
about the image? [3]

2
Page 3 of 3
(c) An image contains a uniform area of beach sand. The variance of the whole
image is 2 500 units while the variance of the sand is 25 units. Estimate the
SNR of the image. [3]
(d) Below is an image of Lena and the results of certain processing. Explain what
is happening in each of the four images (A) – (D) [4]

Original (A) (B) (C) (D)

TOTAL 13 MARKS

QUESTION FOUR:
Let g(x,y) be the two dimensional Gaussian function with σ = 1 pixel and g(0,0) = 100.
𝝏𝒈
For edge detection we wish to find the filter corresponding to .
𝝏𝒙
𝝏𝒈
(a) Show that (x,y) = -100 x exp[-(x 2 + y 2)/2]. [4]
𝝏𝒙
𝝏𝒈
(b) Find the 3x3 edge detection filter corresponding to the function . [8]
𝝏𝒙
TOTAL 12 MARKS

You might also like