Embedded System used for Classifying Motor Activities of Elderly and Disabled People

Nicolas Fourty (1, 2) , David Guiraud (3) , Philippe Fraisse (3) , Guillaume Perolle (4) , Igone Etxeberria (5) , Thierry Val (2)

1 LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
2 LATTIS - Laboratoire Toulousain de Technologie et d'Ingénierie des Systèmes
3 DEMAR - Artificial movement and gait restoration
4 Fatronik
5 INGEMA

Our modern societies are confronted to a new growing problem: the global ageing of population. In order to find ways to encourage elderly people to live longer in their own home, ensuring the necessary vigilance and security at the lowest cost, some tele-assistance systems are already available commercially. This article presents a specific neural network used on a portable system for classifying activities in ambulatory monitoring. After more precisions about this specific neural network, we present in the second part some experimental results from our prototype stemming from gerontologic institute Ingema.

Mots clés: Activity monitoring; automatic fall detection; classification; embedded systems; neural networks

Domaines: Automatique / Robotique

Nicolas Fourty, David Guiraud, Philippe Fraisse, Guillaume Perolle, Igone Etxeberria, et al.. Embedded System used for Classifying Motor Activities of Elderly and Disabled People. Computers & Industrial Engineering, 2009, 57 (1), pp.419-432. ⟨10.1016/j.cie.2009.01.011⟩. ⟨lirmm-00370444⟩