703610 VO Visual Geometry

summer semester 2020 | Last update: 13.03.2020 Place course on memo list
703610
VO Visual Geometry
VO 1
2
weekly
annually
English
The extraction of information from images is a fascinating topic with diverse practical applications. This course conveys basic knowledge in fundamental techniques and competencies in computer vision, using the rich example of geometric scene reconstruction from video sequences. Students who have successfully completed this course possess in-depth knowledge and skills in geometric aspects of computer vision. They are able to apply these to complex problems in computer vision such as the reconstruction of camera and scene geometries from videos or multiple images (703612) or the stitching of individual images to panoramas.
Cameras and their calibration; stereo geometry; linear filters, finding point correspondences; generalization to multiple cameras; 3D reconstruction from stereo pairs, 3D reconstruction from image sequences
The unifying theme of this course is the task of inserting a virtual object into a video sequence. To this end, this course covers basic methods underlying this and releated special effects, as they are used in the movie and TV industries. This task is put into practice in the complementary Programming Project Computer Vision (703612).
Oral Exam
Multiple View Geometry in Computer Vision by Hartley and Zisserman is highly recommended.

Basic familiarity with geometry and linear algebra will be helpful.

see dates
Group 0
Date Time Location
Mon 2020-03-02
10.15 - 12.00 HS 11 HS 11 Barrier-free
Mon 2020-03-09
10.15 - 12.00 HSB 4 HSB 4 Barrier-free
Mon 2020-03-16
CANCELED
10.15 - 12.00 HS 11 HS 11 Barrier-free
Mon 2020-03-23
CANCELED
10.15 - 12.00 HS 11 HS 11 Barrier-free
Mon 2020-03-30
CANCELED
10.15 - 12.00 HS 11 HS 11 Barrier-free
Mon 2020-04-20
10.15 - 12.00 HS 11 HS 11 Barrier-free
Mon 2020-04-27
10.15 - 12.00 HS 11 HS 11 Barrier-free
Mon 2020-05-04
10.15 - 12.00 HS 11 HS 11 Barrier-free