This course provides a thorough and well-structured introduction to computer vision, spanning both classical image processing techniques and modern deep learning methods. Students begin with the fundamentals of image formation and move through essential processing topics such as filtering, edge detection, the Hough Transform, and the Fourier Transform, building a strong mathematical and algorithmic foundation. The curriculum then covers camera geometry, stereo vision, homography, and feature extraction techniques including SIFT and RANSAC, equipping students with the tools needed for geometric understanding of visual data. In the latter half of the course, the focus shifts to deep learning, with topics ranging from CNN fundamentals and transfer learning to advanced applications such as object detection (Viola-Jones, R-CNN, YOLO), semantic segmentation, sequential models, and motion analysis. This progression from classical to contemporary methods ensures that students develop a well-rounded expertise in computer vision, capable of tackling both traditional and cutting-edge challenges in the field.
Week Topic
1- Introduction
2- Image Formation
3- Filtering, Edge Detection
4- Hough Transform
5- Fourier Transform
6- Camera Fundamentals & Simple Stereo,
7- Homography
8- Features, SIFT, Ransac
9- Deep Learning Fundamentals
10- CNNs, Transfer Learning
11- Object Detection, Viola Jones, R-CNN, YOLO
12 - Semantic Segmentation
13- Sequential Models, RNN
14- Motion