Research On RFID-based Handwriting Trajectory Tracking Technology

With the widespread popularization of intelligent devices,an increasing number of applications are providing various services through human-computer interaction.Achieving more natural interaction has become the original intention of human-computer interaction related application design.Radio Frequency Identification(RFID)devices can provide more fine-grained tag location information,allowing the position and motion trajectory of objects to be obtained by deploying tags on them.This capability opens up the possibility of transforming ordinary objects into intelligent human-computer interaction devices.Although there are relatively well-established devices(such as touch screen,Apple Pencil,etc.)to realize handwriting trajectory tracking,most of them are complex in design and can only work on specific intelligent devices.Conversely,RFID-based handwriting trajectory tracking schemes offer advantages such as low cost,strong scalability and easy deployment.Therefore,it is crucial to extend such systems to intelligent homes,classrooms and offices.This thesis focuses on studying the handwriting trajectory tracking technology based on RFID,and designs a writing pad system RF-Board.Based on the theoretical research on the relationship between RFID phase information and distance variation,a prototype system is developed using the existing RFID devices.Firstly,four small microstrip antennas are positioned beneath the wooden board.When the user moves an object with dual tags as a pen on the board,the reader collects the phase information of tag movement through the antennas.Secondly,two positioning algorithms are utilized to locate the tags near the pen tip and pen tail,respectively.Finally,by combining the geometric relationship between tag position and pen tip position in space,the motion trajectory of the pen tip is reconstructed.This thesis proposes the Hyperbolic Differential Hologram algorithm to locate the moving trajectory of the tag near the pen tip.It effectively weakens the dynamic interference of the arm to the RF signal,thus improving the robustness of the system.Experimental evaluation is conducted to evaluate the tag near the pen tip tracking effect.The experimental results show that the average trajectory error is 2.75 cm after eliminating the initial offset.After obtaining the coordinates of the tag near the pen tip at each time point,the motion trajectory of the tag near the pen tail is tracked.Further,this thesis presents a Surface Hologram Positioning algorithm in 3D space to obtain the orientation coordinates of the tag near the pen tail.The accuracy of the tag near the pen tail in 3D orientation is evaluated.After eliminating the initial offset,the experimental results show the average angle error between the elevation angle and direction angle is 6.4° and 7.8°,respectively.After obtaining the position information of the tag near the pen tip and the orientation information of the tag near the pen tail,this thesis designs the pen tip coordinate recovery algorithm.It combines the position and orientation of the tag with the position of the pen tip on the writing plane to obtain the accurate pen tip writing trajectory.To evaluate the system performance,a series of experiments are conducted and the results are compared with the classical RFID-based handwriting trajectory tracking system.The experimental results show that the average trajectory error is 2.8cm,and the average character recognition accuracy is 93.6% after eliminating the initial offset,both of which are better than the compared RFID-based handwriting track tracking system.