Modulcode:	Inf-CV
Englische Bezeichnung:	Image-based 3D Scene Reconstruction
Modulverantwortliche(r):	Prof. Dr.-Ing. Reinhard Koch
Turnus:	unregelmäßig (WS16/17 SS18 SS20 SS21)
Präsenzzeiten:	4V 2Ü
ECTS:	8
Workload:	60 h lectures, 30 h exercises, 150 h self studies
Dauer:	ein Semester
Modulkategorien:	MSc-Inf-WP (MSc Inf (21)) 2F-MEd-Inf-WP (MEd-Hdl Inf (21)) 2F-MA-Inf-WP (2F-MA Inf (21)) MSc-WInf-WP-Inf (MSc WInf (21)) WI (MSc Inf (15)) WI (MEd Inf) WPI (MEd Inf) IG (MSc Inf) IS (MSc Inf) MV (MSc Inf)
Lehrsprache:	Englisch
Voraussetzungen:	Inf-EinfBV Inf-Math-A Inf-Math-B Inf-Math-C

Kurzfassung:

Teaching computers to see and to interpret the surrounding environment is an important step towards intelligent and autonomous systems. The goal of this lecture is to introduce the students into the concepts and algorithms of computer vision and 3D modelling for robotics, autonomous vehicles, drone vision, and underwater vision from the GEOMAR Marine Vision Group.

Computer vision methods in image sequence analysis are presented. Aim is the geometrical and visual surface reconstruction of 3-D objects as well as object tracking and camera motion tracking from image sequences.

Lernziele:

The students can handle entities of projective geometry and image-based geometric transformations and implement these in the context of image-based 3-D scene reconstruction. They have programming skills in OCTAVE and MATLAB.

Lehrinhalte:

The following topics are discussed:

• Image processing, correspondence analysis
• Basics of projective geometry
• Homographies and panoramic images from rotating cameras
• Multi-view geometry from a moving camera
• Epipolar geometry and depth estimation
• Camera tracking and pose estimation
• Application in the field of image-based modeling and underwater vision (GEOMAR)

All lecture videos, slides and course material will be in English. The lecture discussions will be held in English if at least one student does not speak German. Otherwise the course students may choose to have the lecture discussion language either in German or in English.

Weitere Voraussetzungen:

Mathematical knowledge from Bachelor courses in linear algebra, geometry, analysis, and solving of linear equations is needed. Prior knowledge in image processing, like the Bachelor lecture InfEinfBV (Introduction to Image Processing) is required.

Prüfungsleistung:

Oral exam

Lehr- und Lernmethoden:

The course will be held as online course with weekly lecture videos and slides applying the flipped classroom concept. Each week, the lecture material (videos, slides) of the previous week will be discussed in interactive zoom meetings. It is expected that the students actively participate in the weekly discussions. Please be prepared to switch on your video and microphone during the discussions.

For the exercise, weekly homework is handed out and solved in teams of 2. During the weekly interactive ZOOM exercise sessions, active participation of the students is expected. Each team should present homework solutions at least once to be accepted to the exams.

Verwendbarkeit:

Literatur:

Szeliski, Rick: Computer Vision: Algorithms and Applications. Springer 2010. Elektronische Version: http://szeliski.org/Book/

Hartley, Zissermann: Multiple View Geometry, Cambridge 2003.

Schreer: Stereoanalyse und Bildsynthese, Springer 2004

Verweise:

Kommentar:

The Lecture will be held jointly by Prof. Reinhard Koch (Institut für Informatik) and Dr. Kevin Köser (GEOMAR).