Prof. Dr. Didier Stricker

Dr. Alain Pagani

Keonna Cunningham

Dr. Jason Raphael Rambach

Dr. Oliver Wasenmüller

Dr. Gabriele Bleser

Dr. Bertram Taetz

Rashed Al Koutayni

Jilliam Maria Diaz Barros

Mahdi Chamseddine

Steve Dias da Cruz

Christiano Couto Gava

Tewodros Amberbir Habtegebrial

Jigyasa Singh Katrolia

Andreas Kölsch

Onorina Kovalenko

Muhammad Jameel Nawaz Malik

Nareg Minaskan Karabid

Manthan Pancholi

María Alejandra Sánchez Marín

Dr. Kripasindhu Sarkar

Alexander Schäfer

Ahmet Firintepe

Dr. Vladislav Golyanik

Dr. Aditya Tewari

André Luiz Brandão

Reality models for efficient registration in Augmented Reality

Reality models for efficient registration in Augmented Reality
Alain Pagani
PhD Thesis

Abstract:: The fundamental challenge of vision-based Augmented Reality is to make virtual objects and the real world coincide in a seamless way. In order to solve this registration problem, at least one real object present in the scene has to be detected, recognized and used by a corresponding algorithm to compute the position and orientation of the camera. In this thesis, we take a closer look at the representations of reality that can be used in the context of Augmented Reality. We define a Reality Model as a set of compact information about an object of interest and an algorithm that uses this information for estimating the pose of the camera. We first review existing Augmented Reality systems from the standpoint of these models and propose a classification based on the requirements of the pose estimation. We then present our novel approaches for different types of reality models, with applications in marker-based and texture-based pose estimation, 3D reconstruction, object detection and learning-based methods for local pose estimation.

Back to the overview