Chargement...
 

Virtual and Augmented Reality

Domaine
Virtual and Augmented Reality
Domain - extra
Image processing
Année
2012
Starting
October 2012
État
Open
Sujet
Ivora (Image and Computer Vision for Augmented Reality): Color invariance and correspondences for the definition of a camera/video-projector system.
Thesis advisor
JACQUEMIN Christian
Co-advisors
BOUCHAFA Samia, IEF (Institut d'Electronique Fondamentale), Image Processing
GOUIFFES Michele, IEF (Institut d'Electronique Fondamentale), Image Processing

Laboratory
Collaborations
VIXID company http://www.vixid.com/
Other collaborations are possible with the partners of the LIMSI and IEF laboratories according to the development of the work
Abstract
Spatial Augmented Reality (SAR) adds virtual visual elements to the physical world, immersing the user in a mixed environment. SAR is used, among others, in digital art, architecture (heritage digitization and augmentation), or urban planning. The purpose of the work is to ensure the best video-projection of virtual images on the most "appropriate" surfaces with geometrical and colorimetric compensations.
The thesis work is divided into four parts: 1) automatic identification of best projection surfaces in terms of geometry (ideally flat or piecewise flat surfaces), color, texture and roughness, and morphology, 2) to ensure undistorted rendering, compensation of color and geometric pattern projected taking in consideration the Human Visual System sensitivity. 3) development of algorithms on graphics card (GPU) to ensure interactive frame rates. 4) experiment in partnership with Vixid company on architecture and art applications.

Context
The project aims at developing a video projection system on heterogeneous surfaces in terms of geometry and color. In standard projection systems, the highest quality is obtained by the use of a rectangular white opaque and diffuse screen, in a fronto-parallel position and at a suitable distance of the projection system. To get closer to such quality with non standard surfaces, the video will be projected geometrically and radiometrically compensated to offer a best view according to the Human Visual System sensitivity.
This work will elaborate on the recent results in Spataial Augmented Reality such as (Von Itzstein et al. 2011), (Wang et al. 2010) or (Thomas et al. 2011) about relighting, environment design, on the fly geometrical acquisition...
Objectives
Two types of scenarios have been chosen:
1) the case of indoor scenes (exhibition hall, auditorium, meeting room, office). Initially, the system and the surface will be assumed fixed. Considering a real scene, the system will detect an ideal surface projection. In a second step, we will consider a mobility of the projection surface and / or system.
2) For an augmentation of architectural heritage, the projection surface is predefined. It may be a facade or an architecture of any shape. A mapping between image and real world will be used to identify reference patterns in the real scene.
Work program
The thesis work is divided into four parts: 1) automatic identification of best projection surfaces in terms of geometry (ideally flat or piecewise flat surfaces), color, texture and roughness, and morphology, 2) to ensure undistorted rendering, compensation of color and geometric pattern projected taking in consideration the Human Visual System sensitivity. 3) development of algorithms on graphics card (GPU) to ensure interactive frame rates. 4) experiment in partnership with Vixid company on architecture and art applications.

Extra information
Prerequisite
Background in image processing and an ability to develop realtime applications in C++/OpenCV and other libraries.
Détails
Expected funding
Institutional funding
Status of funding
Expected
Candidates
Utilisateur
Créé
Mercredi 11 avril 2012 09:16:37 CEST
dernière modif.
Mercredi 11 avril 2012 12:43:03 CEST

Fichiers joints

 filenamecrééhitsfilesize 
Aucun fichier joint à cette fiche


Ecole Doctorale Informatique Paris-Sud


Directrice
Nicole Bidoit
Assistante
Stéphanie Druetta
Conseiller aux thèses
Dominique Gouyou-Beauchamps

ED 427 - Université Paris-Sud
UFR Sciences Orsay
Bat 650 - aile nord - 417
Tel : 01 69 15 63 19
Fax : 01 69 15 63 87
courriel: ed-info à lri.fr