Capture de mouvement multi senseurs mobiles pour l'optimisation biomecanique

par Housseyne Nadour

Projet de thèse en Automatique - productique

Sous la direction de Nicolas Marchand et de Lionel Reveret.

Thèses en préparation à Grenoble Alpes , dans le cadre de Electronique, Electrotechnique, Automatique, Traitement du Signal (EEATS) , en partenariat avec Grenoble Images Parole Signal Automatique (laboratoire) et de Systèmes Non Linéaires et Complexes (SYSCO) (equipe de recherche) depuis le 01-10-2018 .

  • Titre traduit

    Motion capture for biomechanical optimization by means of multiple mobile sensors


  • Résumé

    The main goal of the PhD will be the development of safe and cooperative control algorithms to control the trajectories of the sensors (cameras) and their orientation in order to optimize the musculoskeletal biomechanics observation. Controlling a sensor fleet is a very popular subject and there are numerous works. Remains however a lot of open problems, especially the obstacle avoidance problem for a coordinated fleet or the position and orientation control, existing works focusing only on position control. The GIPSA-lab laboratory is working on position and orientation control since more than 15 years with many publications and three start-ups. GIPSA-lab has its own motion capture arena (EQUIPEX ROBOTEX) enabling an easy measure of position and orientation and therefore fast experimental validations. Within the AIRCAP project, a generic formalization of the interaction between the control of the sensor's position and orientation and the 3D morphological reconstruction has to be developed (real time characterization of the quality of the 3D reconstruction as a function of the position and orientation) in order to allow the development of efficient position control algorithm with obstacle avoidance. The final goal is to have control laws that dynamically optimize the positioning and orientation of each sensor according to a 3D reconstruction index, while avoiding the other sensors and obstacles (like walls, trees).