REVIEW QUESTIONS & ANSWERS

FOR COMPUTER VISION GROUPS

pc
[COMPANY NAME] [Company address]
 Sadat Academy for Management Sciences Ac. Year: 2024/2025
Faculty of Computers and Information Class: 3 rd AI
Course: Computer Vision and Image Score: 15 marks
computing
Time Allowed: One hour Date: /3/2025
QUESTIONS for MID-TERM Exam
Prof. Dr. Nashaat ElKhameesy ElGhitany
---------------------------------------------------------------------------------------------------------------------

REVIEW QUESTIONS & ANSWERS FOR MIDTERM EXAM

QUESTION 1: MATCH
1 Image A Deals with improving the appearance of an image in an objective way and its
restoration techniques tend to be based on mathematical or probabilistic models of image
degradation.
2 Image B Deals with improving the appearance of an image in a subjective preferences
Enhancement regarding what constitutes a “good” improvement result
3 Feature detection C Refers to finding the features in an image, region, or boundary. Feature
description assigns quantitative attributes to the detected features.
4 Segmentation D Partitions an image into its constituent parts or objects. In general, autonomous
segmentation is one of the most difficult tasks in digital image processing.
5 Image E Interactively select objects, remove them and automatically fill with similar
Completion background (from the same image
6 Morphological F Deals with tools for extracting image components that are useful in the
processing representation and description of shape and boundary of objects.
7 Image G It always follows the output of segmentation step that is, raw pixel data,
Representation constituting either the boundary of an image or points in the region itself.
and Description
8 Image H It is the process that assigns label to an object based on its descriptors. It is the
Recognition: last step of image processing which use artificial intelligence of software.
9 Sampling I The process of mapping continuous function to a discrete one
10 Quantization J The process of mapping continuous variable to a discrete one

11 Image Matting K The process of extracting an object from the original image
12 Image L The process of inserting the object into a different image
Compositing
13 Spatial Grey A measure of the smallest discernible detail in an image, quantitatively, with
Resolution line pairs per unit distance, and dots (pixels) per unit distance being common
measures.

14 Intensity Refers to the smallest discernible change in intensity level

resolution
15 Log We use this type to expand the values of dark pixels in an image, while
transformations compressing the higher-level values. The opposite is true of the inverse log
(exponential) transformation

16 Negative image This type of processing is used to enhance white or gray details embedded in
dark regions of an image, especially when the black areas are dominant in size
17 Power -Law Maps a narrow range of dark input values into a wider range of output values,
Transformation while maps a narrow range of bright input values into a wider range of output
values

18 Contrast Expands the range of intensity levels in an image so that it spans the ideal full
stretching intensity range of the recording medium or display device
 19 Segmentation A technique for dividing a picture into its components or objects
and in general, one of the most difficult tasks in digital image
processing.
20 Gamma A technique for employing the exponent gamma to correct the
Correction response of a Power-law transformation.
22 Contrast Its primary principle is to increase the dynamic range of gray-
stretching levels in an image.
23 Radiance: total amount of energy that flow from the light source, measured in
watts (W)
24 Luminance Refers to the amount of energy an observer perceives from a light
source, measured in lumens(lm)
25 Far infrared light high radiance, but 0 luminance
26 Brightness: A subjective descriptor that is hard to measure, similar to the
achromatic notion of intensity
28 Computer vision Concerned with modeling and replicating human vision using computer
to enable computers gain a high-level understanding from digital images
or videos, similar to human vision.
29 Pattern field of study that focuses on various techniques to classify patterns,
recognition such as images, signals, and data. It involves the development of
algorithms and models to recognize and categorize patterns based on
their features and cha
Stitching : turning overlapping photos into a single seamlessly stitched panorama
3D modeling: converting one or more snapshots into a 3D model of the object or
person you are photographing

QUESTION (3) Answer each:

A) List the Key stages of image processing? Illustrate with a figure??

The key stages are outlined as image acquisition, enhancement, restoration, segmentation, and recognition.
Applications include visual quality improvement and image compression.

Image Morphological
Restoration Processing

Image
Enhancement IMAGE AER Segmentation

MOR
Image SEG Object
Acquisition Recognition
RECO
REPRE & DESCR
Representation
& Description
 B) Explain briefly each of the following

i) Image acquisition
ii) Image enhancement Vs. Image Restoration

i) Image Acquisition: The initial step in digital image processing, where an image is captured in digital
form using sensors and a light source. The light reflects (e.g., visible light) or transmits (e.g., X-rays)
from the object and is recorded by the sensor, producing an output in voltage waveform.

The image captured is analog image as the output is continuous.

To digitize the image, we use sampling and quantization where discretize the image. Sampling is discretizing
the image spatial coordinates whereas quantization is discretizing the image am plitude values.

ii) Image Enhancement vs. Image Restoration

• Image Enhancement: Modifies an image to improve its visual quality for specific purposes, such
as noise reduction and contrast adjustment, often using filters. This process can be done in the
frequency domain using Fourier transforms.
• Image Restoration: Focuses on recovering an image from degradation by estimating and
reversing distortions. It is an objective process that relies on observation, experimentation, or
mathematical modeling to restore the original image.

.
Image Restoration Block Diagram:
The restoration process uses one of three filters:

• Inverse Filter: Simplest method but ineffective with noise.

• Wiener Filter: Minimizes mean square error for better restoration.
• Constrained Least Squares Filter: Uses constraints for optimal results, making it the most
effective method.

C) Explain basic difference between each of the following:

i) Sampling Vs. Quantization ( in imaging)
ii) Scotopic Vs, Photopic Vs, Mesopic vision
iii) Image enhancement Vs. Image restoration
iv) Radiance Vs. Illumination
v) Segmentation Vs. Classification Vs. Pattern Recognition

i) Sampling vs. Quantization (in imaging)

• Sampling: Converts a continuous image into discrete spatial points.

• Quantization: Converts continuous intensity values into discrete levels.
Both processes enable digital image storage and manipulation.

ii) Scotopic vs. Photopic vs. Mesopic Vision

• Scotopic: Uses rods for black-and-white vision in low light.

• Photopic: Uses cones for color vision in bright light.
• Mesopic: A mix of both, used in intermediate lighting conditions.

iii) Image Enhancement vs. Image Restoration

• Enhancement: Subjective process improving visual appeal based on human perception.

• Restoration: Objective process correcting image degradation using mathematical models.
iv) Radiance vs. Luminance

• Radiance: Total energy emitted by a light source (measured in watts).

• Luminance: Perceived brightness by an observer (measured in lumens).

v) Segmentation vs. Classification vs. Pattern Recognition

• Segmentation: Divides an image into meaningful regions based on similarity or discontinuity.

• Classification: Assigns labels to segmented regions based on features.
• Pattern Recognition: Identifies patterns in classified data using AI or statistical models.

vi) Feature Extraction

After segmentation, feature extraction identifies key attributes from images, regions, or boundaries.
Examples include corner detection and boundary descriptors (e.g., shape numbers, chain codes).
Extracted features should remain invariant to scaling, rotation, translation, and illumination for
reliable analysis.

Main Levels of Image Processing

1. Low-Level Processing
o Definition: Basic operations where both input and output are images.
o Examples: Image enhancement, restoration, and sharpening.
o Comparison: Like adjusting brightness or removing noise in a photo.
2. Mid-Level Processing
o Definition: Extracts meaningful features from images.
o Examples: Image segmentation, object detection, and classification.
o Comparison: Identifying edges in a medical scan or detecting objects in a security
camera.
3. High-Level Processing
o Definition: Interprets recognized objects for decision-making.
o Examples: Face recognition, autonomous driving, scene understanding.
o Comparison: A self-driving car recognizing pedestrians and road signs.

E)
i) What are the components of a digital image processing system?
Physical sensors and digitizers are required for image acquisition. Image displays like flat, colour screen
monitors, hardcopy devices such as laser printers, computers with specialised hardware and software for
digital image processing systems, mass storage and networking and cloud communication are the
components of digital image processing systems .

ii)Why is digital image processing more famous than analog image

 processing?
D) Analog image processing is a costlier and time consuming process whereas digital image processing is a
cheaper and fast process.
F)
What does each of the following images demonstrate?

Figure-1

1. Viewpoint Variation – Objects appear different from various angles.

2. Scale Variation – Objects differ in real-world size.
3. Deformation – Non-rigid objects change shape.
4. Occlusion – Objects may be partially hidden.
5. Illumination – Lighting affects pixel appearance.
6. Background Clutter – Objects blend with surroundings.
7. Intra-Class Variation – Broad categories (e.g., chairs) have diverse appearances.

Figure -2
 Stitching: turning overlapping photos into a single seamlessly stitched panorama (Fig-
ure 1.5a), as described in Section 8.2;

Exp osure bracketing: merging multiple exposures taken under challenging lighting conditions (strong sunlight
and shadows) into a single perfectly exposed image

(c)

• morphing: turning a picture of one of your friends into another, using a seamless
morph transition → blending between two photographs

figure 3
• Linear (Negative Transform) – Reverses intensity levels [0:L−1]→[L−1:0][0: L-1] \rightarrow [L-1:
0][0:L−1]→[L−1:0], creating a negative image for enhanced details.
• Logarithmic Transform – Expands dark pixel values while compressing bright ones. Its inverse
(exponential transform) does the opposite.
• Power Law Transform – Adjusts contrast by expanding dark or bright pixel ranges based on an
exponent.
C)
Original Image → Negative Image Transformation

• Used for enhancement or subjective analysis (e.g., medical imaging).

• Highlights fine details by enhancing white or gray regions.
• Formula: s=L−1−rs = L - 1 - rs=L−1−r, where LLL is the maximum intensity level.
• Though the visual content remains the same, negatives can improve contrast for better
interpretation.

QUESTION(2) MCQ
1 Spatial resolution is identified by
A) smallest discernible B) number of line pairs C) Number of dots D) A LL
details along space per unit distance (pixels) per unit distance
coordinates
2 Computer vision is a discipline that studies how to reconstruct, interrupt and understand a 3d scene
from its
A)1D -images B) 2D- images C) 3D- images D) ALL
3 Which is a Valid category of Computer vision?
A) image acquisition B) image understanding, C) image analysis, D) ALL
and processing.
4 Intensity Resolution
A) Smallest discernible B) Number of dots C) Using the number of D) A+C
change in intensity levels (pixels) per unit distance levels of intensities

5 Which is the first step in image processing

• A) Image pre-processing B) Image Enhancement C) Image acquisition.. D) NOT ALL
...

6 Which of the following is a TRUE definition/examples for feature detection

 A) Refers to finding the B) Assigns quantitative C) Identifying corners in D)A +C
features in an image, attributes to the detected an image region
region, or boundary. features.
Whi ch i s a Va l i d sta tement a bo ut huma n ey e v i si o n?
A)Th e n u m b er o f B) several rods are C) Rods capture an D_ALL
ro d s is ~ So m e 7 5 t o connected to a single overall image of the field
1 5 0 m illio n a re nerve ending, to reduce of view but are not
7
d ist rib u t ed o v er t h e the number of details involved in color vision,
ret in a . discernible by these and are sensitive to low
receptors. levels of illumination.

8 Whi ch i s NOT a Va l i d sta tement a bo ut huma n ey e v i si o n?

A) C o n es recep t o rs B) Humans can resolve C) Rods vision is called D_ALL

b et ween 6 a n d 7 fine details because each photopic or bright-light
m illio n a re lo ca t ed cone is connected to its vision.
in t h e cen t ra l p o rt io n
own nerve end. Muscles
o f t h e ret in a , f o v ea ,
rotate the eye until the
a n d a re h igh ly image of a region of
sen sit iv e t o co l o r interest falls on the
fovea
Which phenomena corresponds to the case when objects that appear brightly colored in daylight
appear as colorless forms in moonlight because only the rods are stimulated. This phenomenon is
9 known as scotopic or dim-light vision.
A) Photopic or bright- B) Scotopic or dim-light C) vison illusion D) B+ C
light vision vision.
Which phenomena corresponds to the case when the eye takes few seconds to discriminate between
changes in light intensity at any specific adaptation level
10
A) photopic or bright- B) scotopic or dim- C) brightness adaption D) B+ C
light vision light vision.
Which phenomena corresponds to the case when the perceived brightness is not a simple function of
intensity as the visual system tends to undershoot or overshoot around the boundary of regions of different
11 intensities
A) photopic or bright- B) scalloped bands are C) brightness adaption D) B+ C
light vision called Mach bands.
Which phenomena in which the eye fills in non-existing details or wrongly perceives geometrical properties
of objects
12
A) The Optical Illusion B) scalloped bands are C) brightness adaption D) B+ C
called Mach bands.
Which of the following is TRUE about Human Visual Perception?
A) Human visual B) Cones and rods are C) Cones are highly D) B+ C
perception plays No role two receptors in the eye sensitive to color while
in selecting a retina part Rods are insensitive but
13
visualization technique give a general overall
picture of view as they
are sensitive to low level
of illumination
Which of the following is TRUE about Scotopic?
A) Vision under low B) Vision under good C) Vision under good D) ALL
14
illumination and rod cells illumination and cone illumination and rod cells
are dominant cells are dominant are dominant
Which of the following is a TRUE about the use of negative image
A) It’s obtained using the B) Assigns quantitative C) some viewers find it D) A+ C
15
equation s=L-1-r, where attributes to the detected easier to analyze the fine
r is the gray level of features details of the breast tissue
 source image and s is the using the negative image
gray level of the Enhance white or gray
processed image details
16 Which is TRUE about Contrast ratio?
A) The ratio between B) The ratio between C) The ratio between D) A +C
maximum detectable minimum detectable detectable saturation
intensity to the minimum intensity to the intensity to the noise
detectable intensity maximum detectable detectable intensity
intensity
17 Which of the following statements is TRUE regarding intensity and contrast appearance
A) Image could appear B) Image could appear C) Image could appear D) A +B
darker as background darker as background darker as background gets
gets lighter even with gets darker even with lighter but must also make
same intensity level same intensity level higher intensity levels
18 Which of the following is TRUE about Human Visual Perception
A) Human visual B) Cones and rods are C) Cones are highly D) B+ C
perception plays No role two receptors in the eye sensitive to color while
in selecting a retina part Rods give a general
visualization technique overall picture of view,
are insensitive to color
and are sensitive to low
level of illumination
19 Which of the following is TRUE about Scotopic?
A) Vision under low B) Vision under good C) Vision under good D) ALL
illumination and rod illumination and cone illumination and rod cells
cells are dominant cells are dominant are dominant

Which is TRUE about Digital Image Processing?

A) It’s an application B) ) It’s a software C) It’s a system that D It’s a machine
20 that alters digital that allows altering manipulates digital that allows altering
videos digital pictures medias digital images

What is the commonly used language for image processing?

21 A) Only C++ B) Python, C/C++ with C) Python only D ALL

OpenCV, Matlab, Java.

How does picture formation in the eye vary from image formation in a camera?

22 A) Fixed focal length B) Varying distance C) Variable focal DNOT ALL

between lens and length
imaging plane
Which of the following statement describe the term pixel depth?

A) It is the number of B) It is the number of C) ) It is the number D) ALL

24 units used to mm used to of bits used to
represent each pixel represent each pixel represent each pixel
in RGB space in RGB space in RGB space

25
___________ determines the quality of a digital image.
 A) The discrete gray B) The number of C) discrete gray D) NOT ALL
levels samples levels & number of
samples

Which of the following is the abbreviation of JPEG?

A) Joint B) Joint C) Joint Photographic D) NOT ALL

26
Photographic Photographs Expanded Group
Experts Group Expansion Group

Which is TRUE about function of segmentation in image processing?

A) Deals with B) Deals with C) Deals with Deals

D) with
property in which partitioning an extracting attributes
techniques for
27 images are image into its that result in some
reducing the
subdivided constituent parts or quantitative storage required
successively into objects information of interest
saving an image, or
smaller regions the bandwidth
required
Which of the following tool is used in tasks such as zooming, shrinking, rotating,
etc.?
28 D NOT ALL
A) Filters B) Sampling C) Interpolation

___________ is a commercial use of Image Subtraction.

29 A) MRI scan B) CT scan C) Mask mode D) NOT ALL

radiography

Region of Interest (ROI) operations is generally known as _______

30 A) Masking B) Dilation C) Shading correction D) NOT ALL

Which of the following illustrates three main types of image enhancing functions?

31 A) Linear, logarithmic B) Linear, C) Power law, D) NOT ALL

and power law exponential and logarithmic and
inverse law inverse law

Which of the following transform operation is done on the pixels in sharpening the
image, in the spatial domain?
32
A) Differentiation B) Median C) Integration D) Average

Which of the following transform operation is done on the pixels in blurring the image,
in the spatial domain?
33
A) Differentiation B) Median C) Integration D) Average
 Which of the following fact is an accurate Model for an image?
A) An image is the B) An image is the C) An image is the D) An image is the
multiplication of subtraction of subtraction of addition of
34 illumination and reflectance illumination illumination and
reflectance component from component from reflectance
component illumination reflectance component
component component

Gamma Correction is defined as __________

35 A) Light brightness B) A Power-law C) Inverted Intensity D) NOT ALL

variation response curve
phenomenon
Which is Valid for applying Log transformation on an input image?
A) Compress the B) expand the values of C) maps a narrow range D) A+ B
dynamic range of dark pixels in an image, of dark input values into a
36 images with large while compressing the wider range of output
variations in pixel v higher-level values. values, while maps a
narrow range of bright
input values into a wider
range of output values
Which vision includes object recognition and 3D scene Interpretation?
37 A) Low-level vision B) Intermediate-level High-level vision D) ALL
vision
Low-level vision includes process image for feature extraction.
38
A) True B) False C) A+ B D) ALL

Which is Valid for Mach Band?

A) It’s is an optical B) It’s a high level C) It emphasizes the D) A+ C

39 illusion. vision differentiation between
edges with varying shades
of grey when they reach
each other.
Which is Valid for applying power transformation on an input image?
A) Compress the B) expand the values of C) maps a narrow range D) A+ B
dynamic range of images dark pixels in an image, of dark input values into a
40 with large variations in while compressing the wider range of output
pixel v higher-level values. values, and vice versa for
bright input values

The transformation s = T(r) producing a gray level s for each pixel value r of input
41 image. Then, if the T(r) is single valued in interval 0 ≤ r ≤ 1, what does it signifies?
 A) It guarantees the B) It is needed to C) It guarantees that D) ALL
existence of inverse restrict producing of the output gray level
transformation as some inverted gray and the input gray
T(r) is single valued levels in output level will be in same
in interval 0 ≤ r ≤ 1, range
guarantees the
existence of inverse
transformation.
The transformation s = T(r) producing a gray level s for each pixel value r of input
image. Then, if the T(r) is monotonically increasing in interval 0 ≤ r ≤ 1, what does it
signifies?
A) It guarantees the B) It is needed to C) It guarantees that D) ALL
existence of inverse restrict producing of the output gray level
42 transformation as some inverted gray and the input gray
T(r) is single valued levels in output, level will be in same
in interval 0 ≤ r ≤ 1, range
guarantees the
existence of inverse
transformation.
Common applications of the multiplication function in image processing is/ are
43
A) Contrast stretching B) Masking C) Shading correction D) B+ C

Which type of image transforms processing is particularly suited for enhancing white or gray detail
44 embedded in dark regions of an image
A) Negative s= L-1-r B) Multiply C) Logarithmic D) NOT ALL