Seminar: Recovering 3D Structure from Projected 2D Motion - Details

Seminar: Recovering 3D Structure from Projected 2D Motion - Details

Sie sind nicht in Stud.IP angemeldet.

Allgemeine Informationen

Veranstaltungsname Seminar: Recovering 3D Structure from Projected 2D Motion
Untertitel
Veranstaltungsnummer 8.3286
Semester WiSe 2016/17
Aktuelle Anzahl der Teilnehmenden 20
Heimat-Einrichtung LE Cognitive Science
Veranstaltungstyp Seminar in der Kategorie Offizielle Lehrveranstaltungen
Erster Termin Freitag, 04.11.2016 14:00 - 16:00, Ort: 93/E12
Art/Form
Teilnehmende ab 5. Semester
SWS 2
Sprache Englisch
ECTS-Punkte 4

Räume und Zeiten

93/E12
Freitag: 14:00 - 16:00, wöchentlich (10x)

Kommentar/Beschreibung

Basic knowledge about computer vison and programming experience (e.g. MATLAB, C++, openCV), strong interest in 3D reconstruction using structure from x (SFx) approaches.
</p>

<p>
How to recover 3D structures or models from projected 2D motion in a set of photos or videos? This course will attempt to give an overview of? computer vision based 3D reconstruction methods from 2D projections. (cf. Szeliski, Richard. Computer vision: algorithms and applications. Springer Science & Business Media, 2010. Chapter 7 Structure from motion / Chapter 12 3D reconstruction)<br>
By implementing several algorithms, students attending this course will reconstruct real objects e.g. a coffee mug, out of a set of photos and discuss the advantages and drawbacks of state-of-the-art reconstruction approaches.