TU Darmstadt / ULB / tuprints

Efficient Line and Patch Feature Characterization and Management for Real-time Camera Tracking

Wuest, Harald :
Efficient Line and Patch Feature Characterization and Management for Real-time Camera Tracking.
TU Darmstadt
[Ph.D. Thesis], (2008)

[img]
Preview
PDF
DissHaraldWuest.pdf
Available under Creative Commons Attribution Non-commercial No Derivatives.

Download (9Mb) | Preview
Item Type: Ph.D. Thesis
Title: Efficient Line and Patch Feature Characterization and Management for Real-time Camera Tracking
Language: English
Abstract:

One of the key problems of augmented reality is the tracking of the camera position and viewing direction in real-time. Current vision-based systems mostly rely on the detection and tracking of fiducial markers. Some markerless approaches exist, which are based on 3D line models or calibrated reference images. These methods require a high manual preprocessing work step, which is not applicable for the efficient development and design of industrial AR applications. The problem of the preprocessing overload is addressed by the development of vision-based tracking algorithms, which require a minimal workload of the preparation of reference data. A novel method for the automatic view-dependent generation of line models in real-time is presented. The tracking system only needs a polygonal model of a reference object, which is often available from the industrial construction process. Analysis-by-synthesis techniques are used with the support of graphics hardware to create a connection between virtual model and real model. Point-based methods which rely on optical flow-based template tracking are developed for the camera pose estimation in partially known scenarios. With the support of robust reconstruction algorithms a real-time tracking system for augmented reality applications is developed, which is able to run with only very limited previous knowledge about the scene. The robustness and real-time capability is improved with a statistical approach for a feature management system which is based on machine learning techniques.

Alternative Abstract:
Alternative AbstractLanguage
Eines der Hauptprobleme von Anwendungen im Bereich der erweiterten Realität ist die Bestimmung der Kameraposition und Blickrichtung in Echtzeit. Viele derzeitige Anwendungen beruhen auf speziell konzipierten Markern, die sich leicht in einem Bild erkennen lassen. Markerlose Verfahren zur Bestimmung der Kamerapose basieren häufig auf gegebenen Linienmodellen oder kalibrierten Referenzbildern. Für solche Methoden ist eine aufwändige Vorverarbeitung nötig, was für eine einfache Erstellung von industriellen Anwendungen sehr hinderlich sein kann. Diese Arbeit beschreibt Methoden zur Bestimmung der Kamerapose, welche nur einen minimalen Aufwand an Vorverarbeitung benötigen. Die Ziele der in dieser Arbeit entwickelten Trackingalgorithmen sind das Akquirieren von Linien- und Punktmerkmalen in Kamerabildern, die Erstellung einer genauen Beschreibung und Charakteristik dieser Bildmerkmale und das Management von Bildmerkmalen in einer Sammlung von erstellten Korrespondenzen zwischen Bildpunkten und zugehöriger 3D-Geometrie.German
Uncontrolled Keywords: computer vision, augmented reality, camera tracking
Alternative keywords:
Alternative keywordsLanguage
computer vision, augmented reality, camera trackingEnglish
Classification DDC: 000 Allgemeines, Informatik, Informationswissenschaft > 004 Informatik
Divisions: Fachbereich Informatik > Graphisch-Interaktive Systeme
Date Deposited: 22 Dec 2008 08:26
Last Modified: 07 Dec 2012 11:54
URN: urn:nbn:de:tuda-tuprints-11906
License: Creative Commons: Attribution-Noncommercial-No Derivative Works 3.0
Referees: Fellner, Prof. Dr. Dieter and Stricker, Prof. Dr. Didier
Refereed: 5 November 2008
URI: http://tuprints.ulb.tu-darmstadt.de/id/eprint/1190
Export:

Actions (login required)

View Item View Item