Device-free Gesture Recognition Technology Combining Radio Frequency Identification Method

Abstract

With the development of computer technology, the methods of human-computer
interaction (HCI) have also been innovated continuously. From the traditional mouse
and keyboard to the popular touch screen, and the more advanced voice control, HCI
technology has become increasingly humane. Compared with the conventional HCI
methods, gesture interaction has the advantages of contactless operation, easy to learn,
and massive information carried. Therefore, the study of gesture-based HCI technology
is of great significance. Compared with the existing gesture recognition system based on
radio frequency identification (RFID), the RFID-based device-free gesture recognition
method can track the hand area during movement effectively by estimating and judging
the position of the hand centroid when users are not required to carry any device. During
dynamic gesture trajectory tracking, signal interference by gesture is considered as a
fingerprint feature. Multipath is used to increase the matching difficulty and ensure the
gesture recognition accuracy. The synthetic aperture radar (SAR) algorithm is used to
obtain the fingerprint feature matrix corresponding to each gesture. The priori gesture
fingerprint database is matched by the dynamic time warping (DTW) for gesture
recognition. The experimental results suggest that the RFID-based device-free gesture
recognition method has significantly improved the accuracy with excellent robustness