Course Learning Outcomes (CLOs) are those describing the knowledge, skills and competencies that students are expected to achieve upon successful completion of the course. In this context, Course Learning Outcomes defined for this course unit are as follows:
Week |
Subject |
Materials Sharing * |
|
Related Preparation |
Further Study |
1) |
What is Computer Vision?
image representation using math.
image matrix repreparation.
Image transformations
Advanced Application of Image Processing
Why is computer vision so hard?
Fun with Color
Image Resizing and Filtering
Panorama Stitching
Neural Networks
PyTorch
Deep Learning/Machine Learning
Advanced Application of Image Processing |
prepare some demos, program code of examples.
Materyal
|
n/a |
2) |
What is Computer Vision?
image representation using math.
image matrix repreparation.
Image transformations
Advanced Application of Image Processing
Why is computer vision so hard?
Fun with Color
Image Resizing and Filtering
Panorama Stitching
Neural Networks
PyTorch
Deep Learning/Machine Learning
Advanced Application of Image Processing |
prepare some demos, and program code examples.
Materyal
|
n/a |
3) |
map of color
The 2D array of color
Grayscale images
Geometric HSV to RGB
More Details on Color Spaces
Color histograms can represent an image.
Local Binary Pattern (LBP) Measure
|
prepare some demos, and program code examples.
Materyal
|
n/a |
4) |
An image is a matrix of light.
Addressing pixels
Color representation.
Image interpolation and resizing
A note on coordinates in images
Nearest-Neighbor Interpolation
Bilinear Interpolation
Image resizing
|
prepare some demos, and program code examples.
Materyal
|
n/a |
5) |
Convolution operation
Box filters smooth image.
Gaussians
smoothing with Gaussians
Highpass Kernel
Identity Kernel
Sharpen Kernel
Emboss Kernel
Sobel Kernels
|
prepare some demos, and program code examples.
Materyal
|
n/a |
6) |
Cross-Correlation vs Convolution
What’s an edge?
Finding edges
Image derivatives
Laplacian (2nd derivative)
LoG filter
Difference of Gaussian (DoG)
Gradient magnitude
Canny Edge Detection
Non-maximum suppression
Hough Transform for lines and circles.
|
prepare some demos, and program code examples.
Materyal
|
n/a |
7) |
Corner Detection
Harris Matrix
Estimating Response
Harris Corner Detector
Properties of the Harris corner detector
The SIFT
Key point localization
|
prepare some demos, and program code examples.
Materyal
|
n/a |
8) |
Descriptors
Simple Normalized Descriptor
Orientation Normalization
SIFT descriptor.
Properties of SIFT
Matching with Features
Compute Transformations
Image reprojection
Image Warping
Solving for homographies
Direct Linear Transforms (n points)
RANSAC for estimating the homography
|
prepare some demos, and program code examples.
Materyal
|
n/a |
9) |
Filtering in the Frequency Domain
(2D continuous and discrete Fourier transform and its inverse, relationship between spatial and frequency intervals.)
|
|
|
9) |
Panorama algorithm
Stitching panoramas
Very bad for big panoramas (Triangle)
How do we fix it? Cylinders
build a panorama from two (or more) images.
RANSAC for Homography
Image Blending
Feathering
Effect of window (ramp-width) size
Pyramid blending.
Alpha Blending
Gain Compensation: Getting rid of artifacts.
Blending Comparison
Recognizing Panoramas
Finding the panoramas
Creating Panoramas
|
prepare some demos, and program code examples.
Materyal
|
n/a |
10) |
Panorama algorithm
Stitching panoramas
Very bad for big panoramas (Triangle)
How do we fix it? Cylinders
build a panorama from two (or more) images.
RANSAC for Homography
Image Blending
Feathering
Effect of window (ramp-width) size
Pyramid blending.
Alpha Blending
Gain Compensation: Getting rid of artifacts.
Blending Comparison
Recognizing Panoramas
Finding the panoramas
Creating Panoramas
|
prepare some demos, and program code examples.
Materyal
|
n/a |
10) |
Filtering in the Frequency Domain
(2D discrete convolution theorem, frequency domain filtering fundamentals.)
|
|
|
11) |
Filtering in the Frequency Domain
(Low-pass, high-pass, band-reject and band-pass filters in the frequency domain.)
|
|
|
11) |
Queries
Commercial Systems
Retrieval Features
Indexing in the FIDS System
Lead-in to Object Recognition
Weakness of the EM Classifier Approach
|
prepare some demos, and program code examples.
Materyal
|
n/a |
12) |
Image Restoration and Reconstruction
(Noise models and estimating noise parameters, mean, order statistic and adaptive filters.)
|
|
|
12) |
CNN
Convolution Operation
Learning
Pooling
CNN Structures
Image Classification
Semantic Segmentation
Convolutional Neural Networks In PyTorch
|
prepare some demos, and program code examples.
Materyal
|
n/a |
13) |
Unsupervised Learning: Autoencoders
Unsupervised Learning: Variational Autoencoders
Distributions during training
GAN: Sample Architecture (DC-GAN)
Bidirectional GAN (BiGAN)
Conditional GAN (cGAN)
Progressive Growing of GANs
|
prepare some demos, and program code examples.
Materyal
|
n/a |
14) |
Course revision all the chapters |
Course revision
|
n/a |
(KPLOs and SPLOs are the abbreviations for Key & Sub- Programme Learning Outcomes, respectively. )
WORKLOAD OF TEACHING & LEARNING ACTIVITIES |
Teaching & Learning Activities |
# of Activities per semester |
Duration (hour) |
Total Workload |
Course |
14 |
3 |
42 |
Laboratory |
0 |
0 |
0 |
Application |
0 |
0 |
0 |
Special Course Internship (Work Placement) |
0 |
0 |
0 |
Field Work |
0 |
0 |
0 |
Study Hours Out of Class |
0 |
0 |
0 |
Presentations / Seminar |
0 |
0 |
0 |
Project |
3 |
0 |
0 |
Homework Assignments |
0 |
0 |
0 |
Total Workload of Teaching & Learning Activities |
- |
- |
42 |
WORKLOAD OF ASSESMENT & EVALUATION ACTIVITIES |
Assesment & Evaluation Activities |
# of Activities per semester |
Duration (hour) |
Total Workload |
Quizzes |
0 |
0 |
0 |
Midterms |
1 |
3 |
3 |
Semester Final Exam |
1 |
3 |
3 |
Total Workload of Assesment & Evaluation Activities |
- |
- |
6 |
TOTAL WORKLOAD (Teaching & Learning + Assesment & Evaluation Activities) |
48 |
ECTS CREDITS OF THE COURSE (Total Workload/25.5 h) |
6 |