Markerless-based Augmented Reality Kernel

Markerless is a software platform of about 30,000 lines of code for Linux and Windows. I was the leading author of this code-base during my thesis.

This software implements an algorithm for localising a monocular camera in 3D using visual information and 3D CAD models of the world. Pose estimation is performed by a non-linear minimisation approach known as virtual visual servoing. In this way pose estimation is considered as the dual problem of visual control. The software has many assets including: high precision, robustness, real-time efficiency and accuracy. A detailed description can be found in my thesis.

The particularity of this software is that is does not require the placement of physical markers on target objects in the real environment to function.

This software is currently used by various enterprises including:

• Total Immersion,
• The European Space Agency,
• France Telecom,

A licence of this software was yielded to the company Total Immersion.

The software has also been demonstrated at various exhibitions such as at Siggraph'05, "à la fêtes de la science aux Jardins du Luxembourg à Paris", at the "Journées GDR ISIS" and others.

Version 1.0 of the software has been submitted to and is protected by the "Agence pour la Protection des Programmes (APP)".

Videos and publications related to this software are available on these demo pages:

• 3D Model-based Tracking
• Augmented Reality

Visual Odometer

This is a software platform that I have developed for performing real-time 6 degrees of freedom pose estimation from multiple cameras. The code is related to published work on Quadrifocal Visual Odomery.

If you have any questions please send me an This e-mail address is being protected from spambots. You need JavaScript enabled to view it .