Image Processing
1- Discuss Five different application for DIP. you should clearly
explain the application and its importance.
Finger print recognition: Use image enhancement to make images more clearer and extract ridges
and valleys.
Face recognition: Use image detection then use extracted features for recognition.
Plant leaf disease classification: See if leaf has infection or not.
Industrial Machine: Surface defects detection for ceramic tile.
Cancer classification using CNN: Use image sharpening.
2- Images can be classified depending on their colors and bit
representations give a detailed explanation for this classification.
- Gray Scale - Represent shades of gray, with pixel values ranging from 0 (black) to 255 (white).
- RGB - Represent color using three channels: Red, Green, and Blue, requiring 3 matrices.
- Black & White - Each pixel is either 0 (black) or 1 (white), using just 1 bit.
3- DIP is used in industrial machine vision (one example is surface
defects detection for ceramic tile) discuss this application using a
drawing for the system.
Surface defects detection for ceramic tile (examples for pre-processing resizing, RGB to Gray, Edge
detection).
4- discuss the different categorizes for DIP operations (Low level
processes, medium level processes, high level processes) giving an
example for each category
Low level processes: inputs and outputs are images.
Ex: Image sharpening, Noise reduction, contrast enhancement.
Mid level processes: Inputs are images outputs are attributes
Ex: Segmentation, object classification.
Higher-level processing: involves “making sense” of an ensemble of recognized objects
Ex: As in image analysis, performing the cognitive functions associated with vision.
5- give an example for an image that gives an optical illusion for
the eye.
Figure A: square is perceived even though no complete lines form it.
Figure B: A circle appears to be complete with only a few segments present.
Figure C: Two identical lines appear to be of different lengths.
Figure D: Lines at 45° angles, seem non-parallel due to the crosshatch pattern.
6- prove using mathematical equations that adding different
images for the same scene (with noisy images) and averaging
them will give the original image without noise.
7- Discuss how we can use image subtraction to make veins in
medical image clearer.
We take two X-ray pictures. The first is taken before injecting iodine. After the iodine is injected, we
take a second picture that highlights the veins. Then we subtract the first picture from the second
one making veins more visible.
8- Discuss how we can use image multiplication to extract Region
of interest regions from a medical image.
Multiplying given image by mask image that has 1s in ROI and 0s elsewhere. The shape of the ROI can
be arbitrary.
9- there are two types of image processing techniques (spatial
domain, transform domain) give a definition for both using
supporting diagrams.
Spatial domain processing: Image processing methods that direct manipulation of pixels in an image.
Transformation domain processing: Transforming image to transform domain, doing the process
there and obtaining the inverse transform to bring the results back into the spatial domain.
10- Explain the process of sampling and quantization. Explain its
importance in DIP.
Sampling: Digitizing the coordinate values
Quantization: Digitizing the amplitude values
Importance: Improving image quality by samples and discrete intensity levels used.
11- Give a brief description on the effect of the following operation
on a gray scale image Also, what happens if we increase m and n
values.
m and n affects the image's appearance, increasing m and n generally results in greater
smoothing or blurring.
Larger m and n values can amplify edges and enhance image
contrast.
12- Discuss the difference between global and local histogram
equalization.
Global histogram equalization: The equalization is applied to the entire image. (the image is treated
as one)
Local histogram equalization: The equalization is applied on a defined neighbourhood around the
pixel.
13- Give an example by numbers on how the median filter work.
What kind of noise median filter is most famous of enhancing.
When several values in a neighborhood are the same, all equal values are grouped. suppose that a
3*3 neighborhood has values
10 20 20 10 15 20
(10, 20, 20, 20, 15, 20, 20, 25, 100).
Average
These values are sorted as 20 15 20 20 20 20
(10, 15, 20, 20, 20, 20, 20, 25, 100),
which results in a median of 20. 20 25 100 20 25 100
Salts and Pepper noise is the most famous of enhancing.
14- What is the effect of using these mask matrices on a gray scale
image. And what is the difference between the first matrix and
second matrix.
Effect on a Grayscale Image:
- Used for blurring and noise reduction.
- Increasing the number of iterations will make the image more blurred.
Difference Between the Matrices:
- First Matrix (1/9): Box Filter.
- Second Matrix (1/16): Weighted Average Filter
15- Explain the following gray level transformation functions
Image negatives: The negative with intensity levels in the range[0, L-1].
Transformation function s = L-1-r.
Log transformation: Maps a range of low input level values into a wider range of output levels.
Transformation function s = c log(1 + r).
16- Discuss the difference between ideal low pass filter,
Butterworth low pass filter, and Gaussian low pass filter, write
their equation with explanations using graphs.
Ideal low pass filter Butterworth low pass filter Gaussian low pass filter
- Attenuate high-frequency components
Point of transition is called the cutoff
Does not have a sharp discontinuity while preserving low-frequency components.
frequency.
- Based on Gaussian function.
17- Discuss the difference between ideal high pass filter,
Butterworth high pass filter, and Gaussian high pass filter, write
their equation with explanations using graphs.
Ideal high pass filter Butterworth high pass filter Gaussian high pass filter
- Allows high-frequency components of an - Attenuate low-frequency components of an
Allows high-frequency components of
image to pass while attenuating low- image while preserving high-frequency
an image to pass while attenuating low-
frequency components. components.
frequency components.
- Characterized by its smooth transition - Based on Gaussian function.
18- Draw the transformation intensity function for transforming an
image from a gray scale to black and white. (Please add all labels
to the drawing).
X axis = input intensity
Y axis = output intensity
19- Discuss briefly the effect of this transformation intensity
function on a gray scale image.
This transformation highlights intensity range [A, B]
and reduces all other intensities to a lower level.
This transformation highlights range [A, B] and preserves
all other intensity levels.
20- Draw the transformation intensity function for gamma power
transformation. Discuss briefly the effect of different gamma y
values on the output of the gray scale image.
When gamma increases (have big value) – Dark.
When gamma decreases (have low value) – Light.
21- Write a python function that takes an input image and returns
an output image with Edges Detected. (Sheet 1)
ف احتمال ميجيشRevision مش في الـ
22- Write a function that takes an input image and outputs a
Negative Transformation for that image. (Sheet 1 + Mid 24 + Revision)
23- Write a python program for Spatial Filtering (use averaging
matrix). (Final 22 + Revision)
24- Write a python program for Median Filter.(Revision)
The End
