Modeling of the positioning and 3D acquisition system of the brain.
Philippe Lambert
Patrick Hébert (Supervisor)
Problem: The systems which nature has conceived are always an inspiration. My project on the positioning of a camera in an unknown environment and the 3D modeling of environments will thus attempt to apply the methods which the brain itself uses for positioning. The hypothesis which I would like to verify is that positioning and modeling are in fact connected and inter-related in the methods which the brain uses for positioning.
Motivation: This research therefore implies an analysis of the human positioning system to enable an improvement and proposal of new methods for the positioning of cameras. Moreover, since our hypothesis is that 3D modeling plays an essential role in positioning, this area will also be studied. The possibility of carrying out a treatment in real time is also an important objective for all of the algorithms which will be developed. In fact, by using the human system as a model, we are directing our research towards real time treatment.
Approach: We propose to use an electronic set-up which resembles that of the human brain, i.e. a stereo pair which is coupled with a gyroscope and accelerometers so as to imitate the human eyes and the vestibular system. This set-up is based on work carried out by Martin Labrie (see: This research, which will involve psychology and biology in addition to computer systems, will enable me to use the electronic system to experiment on new approaches using this configuration. The 3D representation used will also be of great importance. The most appropriate representation is probably the one developed by Dragan Tubic (see: ). The platform of applications used will involve augmented reality tools so as to communicate the information acquired by the system to the user. For example, the model of the environment which is created by the system could be superimposed on a real environment so as to allow for comparisons. The application to augmented reality will also enable a link to the Lantiss project which could benefit from the research performed.
Applications: This project will lead to a better knowledge of the human brain and may lead to a source of inspiration for different representations and algorithms than those presently used.
Calendar: September 2003 - September 2005
Last modification: 2007/09/26 by plambert


©2002-. Computer Vision and Systems Laboratory. All rights reserved