Automatic Recognition of Emergencies with the Help of Optical and Acoustic Sensors

Abstract

Within this article a sensor-based safety system is presented which enables an automatic recognition of emergency situations in every day domestic ambiance and triggers autonomously alerting measures without explicit user interaction. The development is performed in the research project sens@home [1] which is funded by BMBF (German Federal Ministry of Education and Research). Within the scope of this project a requirements’ analysis had been accomplished first by doing workshops, interviews and questionnaires with potential user groups. Considering the determined demands and requirements, potential adequate sensors have been selected. Here, the main focus was on reasonably priced and contactless sensors which assure a quick recognition. To meet these demands acoustic and optical 3D sensors turned out to be a promising solution. In the second step, by means of practical laboratory experiments, appropriate evaluation algorithms have been developed in order to combine data from several sensors and to elaborate scene-based accident recognition. The advantage of this solution based on low priced optical and acoustic sensors is, that typical emergencies can be detected automatically within a few seconds and there is neither a need for any physical contact or interaction of its users nor any expensive alterations as it can be build in cheaply in every kind of living space. In addition, a large part of the evaluation can already be accomplished in the sensor whereby no data will have to be saved or transmitted.