3D Computer Vision

The course is not on the list Without time-table
Code Completion Credits Range Language
AE4M33TDV Z,ZK 6 2+2c
The course cannot be taken simultaneously with:
3D Computer Vision (A4M33TDV)
The course is a substitute for:
3D Computer Vision (A4M33TDV)
Department of Cybernetics

This course introduces methods and algorithms for 3D geometric scene reconstruction from images. The student will understand these methods and their essence well enough to be able to build variants of simple systems for reconstruction of 3D objects from a set of images or video, for inserting virtual objects to video-signal source, or for computing ego-motion trajectory from a sequence of images. The labs will be hands-on, the student will be gradually building a small functional 3D scene reconstruction system.


Knowledge equivalent to Geometry for Computer Vision and Graphics and Computer Vision Methods.

Detailed up-to-date information on the course at http://cw.felk.cvut.cz/doku.php/courses/a4m33tdv/start

Syllabus of lectures:

1. 3D computer vision, goals and applications, the course overview

2. Real perspective camera

3. Calibration of real perspective camera

4. Epipolar geometry

5. Computing camera matrices and 3D points from sparse correspondences

6. Autocalibration

7. Consistent multi-camera reconstruction

8. Optimal scene reconstruction

9. Epipolar image rectification

10. Stereoscopic vision

11. Algorithms for binocular stereoscopic matching, multi-camera


12. Shape from shading and contour

13. Shape from texture, defocus, and color

14. Surface reconstruction

Syllabus of tutorials:

1. Labs introduction and overview, experimental data, entrance test

2. Camera calibration without radial distortion from a known scene

3. Camera calibration with radial distortion from a known scene

4. Computing epipolar geometry from 8 points

5. Computing epipolar geometry from 7 points, RANSAC

6. Constructing projection matrices from epipolar geometry, computing camera motion and scene structure

7. Autocalibration of intrinsic camera parameters

8. Consistent reconstruction of a many-camera system

9. Accuracy improvement by bundle adjustment

10. Time slot to finish all pending assignments

11. Epipolar rectification for stereoscopic vision

12. Stereoscopic matching by dynamic programming

13. 3D point cloud reconstruction

14. 3D sketch reconstruction

Study Objective:

To master conceptual and practical knowledge of the basic methods in 3D computer vision.

Study materials:

R. Hartley and A. Zisserman. Multiple View Geometry. 2nd ed. Cambridge

University Press 2003.

Y. Ma, S. Soatto, J. Kosecka, S.S. Sastry. An Invitation to 3D

Vision. Springer 2004.

Further information:
No time-table has been prepared for this course
The course is a part of the following study plans:
Data valid to 2019-03-20
For updated information see http://bilakniha.cvut.cz/en/predmet12823604.html