A Study On Impact Of Obstacles On RFID Link State And Its Application In The Field Of Gesture Recognition

With rapid development of information technology, gesture recognition hasreceived much attention in recent years. Till now, many technologies have beendevelopedto recognizehuman’s gestures, such as imagebased method,wirelesssignalsbased method and wearable devices based method. But those methods all have theirlimitations in actual application. In order to solve this problem. This paper proposes aRFID based gesture recognition system that enable gesture recognition without anydevices on human’s body.Firstly, this paper analyzes existed technologies and makes a conclusion.Then, this paper makes a study on impact of obstacles on RFID link state, anddraws a conclusion. According to this conclusion, the paper designs a RFID array, andLSI (Link State Indicator) is proposed to describe whether a tag is obstructed, by usingthe RFID array and LSI, human’s gesture can be described.The paper also designs a link state matrix and a gesture matrix to extract staticgesture’s feature, and then, fisher’s linear discriminant method is used to classify thesefeatures and tell what the gesture is.When recognizing dynamic gestures, a dynamic gesture is regarded as a sequenceof static gestures, in addition, critical gesture and critical switch are proposed to analyzethe sequence.Finally, an experimental system has been built, and an experiment has beenconducted to evaluate the system. The result shows the system can recognize staticgestures with an average accuracy of93.1%, and the accuracy of dynamic gestures’recognition is83.5%.