Proposal of fall down detection method using shape of feature quantity obtained by Obrid-Sensor

In recent years, aging society is getting serious as many social problems rising in Japan. One of the problems that we should concern is the daily safety of the elderly people who live alone. Thus, sensors and integrated systems to confirm the safeness of the elder in their daily life have been developed. In our previous research, a system using a sensor named as One-dimensional brightness distribution sensor (Obrid-Sensor), which is able to detect the presence, position, and motion of a subject without privacy offending, was proposed. As an important improvement of previous research, the method to detect the standing/falling-down using Obrid-Sensor is proposed in this research. The obtained vertical feature quantity of subject is used in this method for detection of the standing/falling-down. The method to detect the standing/falling-down with employing a shape change of feature quantity is experimented by using the Obrid-Sensor. According to the experiment result, the settings of Obrid-Sensor were investigated. The system parameters were studied in different settings for high precision detection. Base on the research, the proposed safety confirmation system can be expected to be applied in more conditions for early detection of fall down accidents.