# Computer Vision

### Second semester 2016

Andrés Marrugo, PhD
*Universidad Tecnológica de Bolívar*

## Aims and Scope

This semester course is an introduction to computer vision. It is aimed at graduate students in the Faculty of Engineering. We will focus on the practical and theoretical aspects of techniques in computer vision.

At the end of the lectures, one would be able to:

- Have clear idea of challenges in computer vision due to increasing use in mobile applications.
- Understand many different computer vision algorithms and approaches.
- Implement computer vision algorithms for mid-level vision tasks.

## Useful Resources

### Tutorials, review materials

- MATLAB tutorial
- More MATLAB tutorials: basic operations, programming, working with images
- Linear algebra review
- Random variables review

### MATLAB reference

## Outline

This is a new course, this website will be updated as we go along.

### Lecture 1: Introduction

We will be discussing the main aspects and motivation for computer vision.

### Lecture 2: Perspective projection

We will be studying the main aspects about perspective projection and the pinhole camera model.

#### Reading

### Assignment 1

In this assignment you will study the basics of projective geometry. You will study the representations of points lines and planes, as well as transformations. **The assignment is due on 2016-09-02 at 11:00 pm.** The assignment and the data:

#### Supporting material

- Lecture notes by Magnus Oskarsson
- Homogeneous Coordinates and Transformations of the Plane
- Projective Geometry and Transformations in 2D

### Lecture 3: Cameras

Cameras with lenses and properties. Thin lens formula, depth of field, field of view, and distorsions.

### Lecture 4: Color

We will discuss the physics of color, human color perception and models of image color.

#### Reading

### Lecture 5: Linear Filtering

Linear filters, convolution kernel, smoothing and sharpening.

#### Reading

### Lecture 6: Frequency representation, pyramids and filter banks.

In this lecture we will discuss the different representation for images and the sampling problems.

Lecture 6 slides - frequency Lecture 6 slides - pyramids

#### Reading

### Questions Lectures 1-6

If you have worked out the lecture questions, please send them to the following link.

### Assignment 2

The goal of this assignment is to learn to work with images in MATLAB. **The assignment is due on 2016-09-24 at 11:59 pm.** The assignment and the data:

### Lecture 7: Edge Detection

We will introduce the general approach towards image edge detection.

#### Reading

### Lecture 8: Corner Detection

We will introduce the general approach towards image edge detection.

Lecture 8 slides

Harris corner detector

#### Reading

### Lecture 9: SIFT

In this lecture we will discuss Scale-Invariant Keypoints.

### Assignment 3

The goal of this assignment is to implement a Laplacian blob detector. **The assignment is due on 2016-10-22 at 11:59 pm.** The assignment and the data:

### Lecture 10: Optical Flow

We will introduce motion estimation in computer vision.

#### Reading

### Lecture 11: Fitting

In this lecture we will discuss the main aspects of fitting data to a parametric model, especially under the assumption of noisy data.

#### Reading

### Lecture 12: Hough Transform

We continue on the topic of fitting, this time via the Hough Transform.

#### Reading

### Lecture 13: Alignment

Registration or alignment is the problem of finding a transformation that takes one dataset to another.

#### Reading

### In class assignment 4

The goal of this assignment is to implement a naive RANSAC line fiting.
**The assignment is due on 2016-10-16 at 11:00 pm.**
The code:

### Lecture 14: Calibration

Calibrating a single camera.

#### Reading

### Lecture 15: Single-view Modeling

Measuring objects from a single image.

#### Reading

### Lecture 16: Epipolar Geometry

Two or more cameras.

#### Reading

### Assignment 5

The goal of this assignment is to implement robust homography and fundamental matrix estimation to register pairs of images separated either by a 2D or 3D projective transformation. **The assignment is due on 2016-12-02 at 12:00 m.** The assignment and the data: