A human motion recognition using data mining for a service robot

Abstract

This research presents a system that provides useful information by data mining. The system collects human motions data to process data mining, and uses the result to detect human intention, and then provide service. The system includes three parts: the input devices part which is to collect data, the server part which is to analyze and process the data mining, and the client part which is to provide a service. The way that set the devices on human to get motions always makes human feel uncomfortable. Therefore, the authors consider using the cameras to obtain human movements rather than setting the devices on human. The collected data can be classed by the server part to find the motion data. The motion data are analyzed to process data mining and find the relation between motion and intention. The result of mining can be applied to provide information on the mobile device or to use for human-robot interaction. The system shows the possibility of applying the data mining to human motion to detect human intention. And the results can offer useful data to assist robotic system or show useful information to user.