J. Krauss, J. Solà, P. Renevey, U. Maeder, and H. Buchholz (Switzerland)
Human kinetics, physical activity classification, portable biomedical device, accelerometry, indirect calorimetry
Together with the Swiss Federal Office for Sports and the company ActiSmile, CSEM has designed and developed a portable biofeedback device to monitor continuously the physical activity of its user. The key technology resides in sophisticated signal processing methods for multi-axis accelerometer sensor systems, including modern feature extraction, classification algorithms and low-power implementation on real-time platforms. A set of classification features has been defined for the following human activities: standing/sitting, lying, walking and running. The features are extracted from the sensor signals of a multi-axis accelerometer and are based on a single stride detection algorithm. A binary classification tree, requiring a low processing complexity, is used for the activity classification tasks. The cumulated time spent in the different activities used to inform the user in real time via the LCD about the progress of the completed physical activity according the WHO guidelines. Additionally, the stored activity features can be sent either via USB or via a wireless Bluetooth link to a PC or a mobile phone for a more detailed analysis. The present work introduces a novel biofeedback device and its underlying technology. The so-called ActiSmile is a novel health-care tool to motivate its user to be physical active and to lead a modern and healthy life.
Important Links:
Go Back
