Computer Vision and Virtual Reality

The course is a substitute for:

Computer Vision and Virtual Reality (XE33PVR)

Lecturer:

Tutor:

Supervisor:

Department of Cybernetics

Synopsis:

The aim of subject is to explain basic approaches used in analyses of digital images. First, the student will learn about image capturing, its digitization, image hardware, and procedures for processing two-dimensional images. The methods for analysis spatial three-dimensional images will follow. Model-based vision. Active vision. Indrustrial applications. Practical recommendations. Computer vision as data source for virtual reality.

Requirements:

Syllabus of lectures:

1. What are computer vision, concepts, and summary of relevant parts of ZSO

2. Different computer vision theories, namely Marr's one

3. Relation to human vision, physiological and psychological considerations

4. Image segmentation, invariants and scale space

5. Basics of projective geometry. Geometrical camera calibration

6. Image capturing from radiometric point of view. Shape from shading

7. Bilinear and trilinear relations among views

8. Devices and techniques for capturing 3D images

9. Stereo vision. Motion detection and analysis.

10. Analysis of 3D images. Bottom up approach. Shape from X

11. 3D reconstruction, image reconstruction. Point clouds and surface

12. Model-based vision. Active vision

13. Industrial applications. Practical recommendations

14. Computer vision as data source for virtual reality

Syllabus of tutorials:

The exercises are conducted in the laboratory. Teams of two students solve, defend and present two small projects. The aim is to introduce real industrial topics to students. Projects are changed each semester. The project consists of its specification, solution proposal, experiments, realization, demonstration of functionality, and defense of a report. The solution is presented to other students on a seminar.

1. Assignment of projects. Introduction to tools available for solving projects.

2. - 4. Students solve simple industrial task consisting of scene capturing, image preprocessing, objects segmentation, description and classification. Ready-made industrial image analysis tools are used.

5. Defense of assignment No. 1

6. - 12. Solution of more complex task that would be hardly solvable using ready made software and without theory presented in lectures. A ready made environment of MATLAB is used for solution including small modules written by students themselves

13. Defense of assignment No. 2

Study Objective:

Study materials:

[1] Jain, R., Kasturi, R., Schunk, B. G.: Machine vision. McGraw-Hill, New York 1995

[2] Šonka, M., Hlaváč, V., Boyle, R.: Image processing, analysis and machine vision. PWS, Boston 1998

Note:

Further information:

No time-table has been prepared for this course

The course is a part of the following study plans: