Abu Dhabi University
College of Electrical and Computer Engineering
Course Syllabus
Electrical/Computer Engineering Program
Spring Semester 23-2024
Course Name: Computer Vision and Machine Learning
Course Code: CEN 454/ECE621
Course Pre-requisites: MTT200, CSC201
Course Co-requisites:
Credit hours: 3
Instructor’s Name: Dr. Syed Omer Gilani
Telephone: E-mail: syed.gilani@adu.ac.ae
Class Location: U-2F-66 Class Time: MW 4:55PM - 6:40PM
Office Location: U-2F-47 Office Hours: MW 6:45 PM - 7:45 PM
1. Course Description
This course will introduce concepts of computer vision including image representation and enhancement,
edge detection, image segmentation, feature extraction and object recognition. It is focused on the
application of machine learning, especially deep learning, to images, video, or 3D signals. It explores
both raw input and feature-space (e.g., shape, appearance, spatial, etc.), imaging data as input to machine
learning for segmentation, object recognition, object detection, and scene understanding. Students will
learn to use Matlab’s Image Processing and Computer Vision toolboxes and Python’s OpenCV in labs
and projects to design image and video processing and understanding pipelines.
2. Learning Outcomes
Expected Learning Outcomes and Associated Assessment Tools.
Course Learning Outcome Assessment Tool(s)*
1. Demonstrate an understanding of the principles of
Exams, Participation Quizzes, Final
computer vision and how they are used to improve
Q1
machine learning performance.
2. Use computer vision and machine learning
algorithms implementations on Matlab and Final Q2, Labs
Python’s OpenCV to solve problems.
3. Demonstrate an understanding the principals of Exams, Participation Quizzes, Final
deep learning for visual data processing. Q3
4. Apply image processing for feature extraction and
machine learning techniques on images and video Final Q4, Labs
signals.
5. Solve real world problems by integrating computer Project Phase 2, Project Phase 3,
vision and machine learning techniques. Labs
6. Investigate contemporary issues in computer vision Project Phase 1
and machine learning of visual information.
1
3. Weekly Outline
Week Topic Readings (Textbook) Labs
1 Introduction to Computer Chapter 2
Vision:
Image Sensing and
Acquisition
Digital Sampling and
Quantization
Pixel Relationships and
Connectivity Analysis
Mathematical Tools in
Image Processing
2 Intensity Transformation and Chapter 3
Spatial Filters:
Intensity Transformation
Histogram Processing:
Stretching, Equalization
Spatial Filtering:
Lowpass, Order-
Statistics, High-pass
Fuzzy Techniques:
Fuzzy Set Theory
3 Frequency Domain Filtering Chapter 4
Review of Fourier
Analysis
Filtering in the
Frequency Domain:
Ideal, Gaussian,
Butterworth filters
Image Sharpening
Notch Filters
4 Image Restoration and Chapter 5
Reconstruction
Noise in Images
Noise Estimation and
Adaptive Filtering
Denoising Techniques
5 Color Image Processing: Chapter 6 Lab 1 – Spatial and
Color Models Frequency Domain Image
Color Smoothing and Enhancement Pipelines on
Sharpening Matlab
Color Segmentation
6 Morphological Image Chapter 9
Processing:
Erosion and Dilation
Hole Filling
Skeletons
2
7 Image Segmentation: Chapter 10
Edge Detection
Global Thresholding
Segmentation with the
K-Means Algorithm
Graph Cut Segmentation
Motion Segmentation
8 Feature Extraction: Chapter 12 Lab 2 – Background
Shape and Region Modeling and Foreground
Features Segmentation using
Corner Detection Gaussian Mixture Models
Scale-Invariant Feature and the EM Algorithm on
Transform (SIFT) Matlab
9 Image Pattern Classification: Chapter 13
Minimum Distance
Classifiers
Correlation and
Template Matching
Feature Matching
Bayes Statistical
Classifiers
10 Neural Networks and Deep Chapter 13 Project Announcement
Learning for Image Processing:
The Perceptron Model Lab 3 – Deep Learning on
Multilayer Feedforward Matlab
Neural Networks
Backpropagation
11 Deep Convolutional
Neural networks
Pre-processing Images
for Classification
12 Introduction to OpenCV on Slides and Papers Lab 4 – Face Detection
Python and Tracking using
Object Tracking: OpenCV
Meanshift and Camshift
Optical Flow
Face Detection
Circle and line detection
13 Multiple Object Slides and Papers
Tracking using Motion
Event Detection
Camera Calibration
14 Final Exam and Project Submission
4. Pedagogy – Learning Process
The three hours teaching per week will consist of lectures and laboratory classes. The course will
require approximately four hours work a week in addition to classes and this should include
laboratory exercises, reading, text-book exercises, and revision.
Lectures will be used to deliver important concepts. Often there is a relationship between ideas
discussed in lectures and practical work in the laboratories.
3
Laboratory classes should be attended by all students. The labs will allow students to put into
practice the theory and concepts presented in readings and lectures. Further practice outside
assigned lab times is encouraged and is essential for many students.
The instructors will not expect students to be passive recipients of information. They will expect
students to participate actively in their own learning. Learning activities may include discussions,
group activities, writing, computer exercises, videos, and some lectures. Both group and
individual work will be expected. Students may also be expected to participate in student-
instructor conferences, small group discussions, audio visual aid presentations, classroom
experiments and in other methods of presenting and analyzing information.
In the course project, students will design an image processing pipeline to detect when two
people as seen from a surveillance camera violate social distancing guidelines. They will have to
calibrate their cameras and use bird-eye transformation to measure the distance between detected
moving objects and then label violators for warning.
In the research assignment, students will research and write a survey paper on cutting edge image
and video processing techniques for feature extractions in machine learning applications.
5. Textbooks and Learning Material
Textbook: Gonzalez, Woods, and Eddins, Digital Image Processing (2018), 4 th Edition, Pearson.
ISBN-13: 9780133356724
Software: Python IDE, Matlab Image Processing, Neural Networks, and Computer Vision
Toolboxes, and OpenCV.
6. Assessment of Student Learning
7.
Learning
Percen
Assessment Tool Description Outcome
t
s
Term Exams Used to progressively assess the
15 %
students’ understanding of the CLO1,
Participation Quizzes / course material CLO3
5%
Assignments
Used to test students’ ability to CLO2,
20%
apply the theory to practical and CLO4,
Labs (Group)
communicate their experience CLO5
Phase 1: State of the
10%
Art Review
Projec
Phase 2: Computer Used to assess the overall skills CLO5,
t 10%
Vision Design acquired in the course CLO6
Phase 3: Testing and
10%
Results
Will examine students’
understanding of all course CLO1-
Final Exam 30 %
material on completion of the CLO4
course
4
Students will receive the same grade for the group assignments reports, but performance-discriminatory
grading for demos depending on their contribution and response to oral exam questions.
Delay Submission Policy
Before deadline, maximum grade is up to 100% of assigned weight.
After the deadline and until the last day of classes, 50% of the assigned weight.
After the last day of classes, 0%.
8. Grading Key
Grade Percentage Grade Point
A 90-100 4
B+ 85-89 3.5
B 80-84 3
C+ 75-79 2.5
C 70-74 2
D+ 65-69 1.5
D 60-64 1
F Less than 60 0.0
I N/A 0.0
9. Attendance
For all students, absence is marked at a maximum of 30% of total class time. Final dismissals (WA)
will be posted on the student’s academic record. The calculation of the absence percentage is based
on the total number of class meetings, regardless of the length of the meetings. For example, if a
class meets once a week for 15 weeks, then a student would need 5 absences to reach the 30%
threshold.
Students will be considered absent if they do not arrive on time for a lesson. Taking attendance will
start on the first day of classes and will continue until the last day of classes in the semester.
Final Dismissal will not be waived under any circumstances. The Registrar's Office will accept
excuses only from students missing an exam/major assignment due to absence. Students will be
permitted to take a make-up exam, if its weight is at least 10% of the course total mark upon
approval of a legitimate excuse.
Exams/Tests will be repeated when serious reasons similar to the following take place and
documented evidence are submitted to the Office of the Registrar on the day student returns to class.
Hospitalization,
Contagious Disease,
Death, may God forbid, of an immediate family member (father, mother, sibling, husband,
son, daughter, grandparent),
Car Accident,
Job Requirement (special assignment), with prior written approval from the Office of the
Registrar
Al Haj
Al Umra
5
Al Umra trips are not valid excuses for students to be absent.
In the case of excused absence for a final exam, the student has to apply for an Incomplete (I) grade
at the Office of the Registrar within 48 hours of the exam.
10. Academic Integrity
The mission of ADU includes developing students’ sense of ethics, morality, and social
responsibility. Students at ADU have the fundamental obligation to conduct themselves with utmost
integrity. To ensure an education environment characterized with Academic Integrity, ADU has
revised its Academic Integrity Policy (AIP).
The revised Academic Integrity Policy (AIP) – copies available on the intranet under the link of the
Office of Academic Integrity (OAI) – has identified violations of Academic Integrity punishable by
serious penalties, including “dismissal” from ADU. The types of violations identified in AIP, along
with the description and examples of each violation, are:
1. Cheating
2. Plagiarism
3. Fabrication of data
4. Presenting false credentials
5. Collusion
6. Free Riding
At the beginning of each semester, your instructor will review the essential elements of AIP in class
and will describe the intrinsic values of compliance, and the consequences of non-compliance (i.e.,
penalties) with the Policy. In addition, OAI undertakes a widespread Academic Integrity Awareness
Campaign at the beginning of each semester to educate students about Academic Integrity issues and
the ways students can avoid violations.
In view of resources provided by ADU in this regard, all students are expected to strictly comply
with the tenet of AIP and conduct themselves with highest degree of integrity and ethical conduct.
Students in need of additional information or assistance are requested to seek help from OAI.
Failure to comply with the provisions of AIP will have severe consequences, including “dismissal”
from ADU.
11. Miscellaneous
Mobile phones: A student whose mobile phone rings during class will be asked to leave the
classroom and will receive a half absence. Should this happen during an exam, the student
will not be allowed to retake the exam at another time, while at the same time receiving a full
absence.
Eating/Drinking/Smoking: Students will be requested to refrain from engaging in these
activities while in class.
Only emails from university-provided email accounts will be recognized and used in this
course.
No email or hardcopy submissions are allowed. All student work is to be submitted by the
due date and time using Blackboard only.
