Monitoring Healthcare Activities toward Automatic Recognition using Accelerometers

Abstract

Intelligent homes should be able to recognize human activities in order to assist people in their daily life. In particular, this is important for healthcare related activities, where people can be reminded of taking some medicine or checking their blood pressure. Not only that, an intelligent system should also be able to verify the correct performance of that activity. Nevertheless, human activity recognition is not trivial and much work still needs to be done. This paper presents a study that show how the activity of measuring blood pressure is performed by professional nurses. This, with the purpose of developing an intelligent system that takes into account the way that the activities are performed by people. A video recording of the activity along with accelerometers attached to manometers allow us to obtain 29 patterns that will help us to code this activity in computer readable form for posterior use by a recognition system.