Learn more about cookies at : http://www.aboutcookies.org/Default.aspx?page=1
Augmented reality
List institute develops readjustment tools for 3D objects models on video stream for augmented reality applications. These tools exploit a video camera signals in order to localise these objects in real time on a specific scenario. The video signal can be merged with information received from GPS or inertial sensors.
Our researchers rely on their expertise on image treatment, Simultaneous Localization And Mapping (SLAM), data fusion and on their know-how on mobile architecture real-time calculation. They collaborate with industries from smart manufacturing and automotive sectors.
Among our academic partners
Pascal Institute in Clermont-Ferrand (France) – University Clermont II
Assets
- No prior marking step,
- High stability on virtual information real-time display, no floating or shaking,
- Use of standard sensors for wider access to the technology,
- Innovating vision systems exploitation (glasses and semi-transparent screens).
Major technologies
Augmented reality on 3D objects
Description
The objective is to resolve in a non-linear system the geometric constraints brought by the object 3D model simultaneously with the constraints linked to the scene 3D structure (Simultaneous Localization and Mapping or SLAM). Despite the complexity of forms, diversity of textures, occultation, lighting conditions and changes, large-scale movements, etc, the application remains robust. The application works in real-time on a simple pad (electronic tablet) or with semi-transparent augmented reality (AR) glasses.
Applications
e-learning, assistance to technicians on their work station (maintenance, assembling), sales assistance.
Major projects
- FUI project Seemake
- FUI project Siforas
Publications
- M. Tamaazousti, V. Gay-Bellile, S. Naudet Collette, S. Bourgeois, M. Dhome. NonLinear refinement of structure from motion reconstruction by taking advantage of a partial knowledge of the environment. CVPR 2011,Colorado Springs, USA 21-23/06/2011.
- Vincent Gay-Bellile, Steve Bourgeois, Mohamed Tamaazousti, Sylvie Naudet-Collette. A mobile markerless Augmented Reality system for the automotive field. ISMAR Workshop 2012
Augmented reality for indoor and outdoor navigation assistance
Description
The approach here consists in a precise and large scale vehicle localisation based on the combination of data from a camera, from a GPS system and from an odometer merged with data from geo-referenced points extracted from a geographic information system (GIS); we talk here about constrained SLAM. (Simultaneous Localization and Mapping or SLAM). This technology brings back into the driver’s or passenger’s attention and field of vision extremely localised information of context (road signs, historical heritage elements or military applications spots).The technology has also been proved efficient indoor for pedestrian navigation in large premises (malls, train stations etc.)
Applications
Driving assistance, autonomous and connected vehicle, pedestrian navigation…
Projets majeurs
Publications
- Dorra Larnaout, Vincent Gay-Bellile, Steve Bourgeois, Benjamin Labbé, Michel Dhome. Driving in an augmented-city: from fast and automatic large scale environment modeling to on-line 6DOF vehicle localization. VRCAI 2013: 9-16
- Vincent Gay-Bellile, Mohamed Tamaazousti, Romain Dupont, Sylvie Naudet-Collette. A Vision-Based Hybrid System for Real-time Accurate Localization in an Indoor Environment. VISAPP (1) 2010: 216-222
"We readjust real images with 3D virtual images avoiding markings and with a high level of stability."