A handwriting trajectory tracking method based on RFID

Abstract

A handwriting trajectory tracking method based on radio frequency identification (RFID) uses a combination of a reader, a small microstrip antenna, and a commercial passive RFID tag to track handwriting trajectories in a specific area. Specifically, it includes system construction and data acquisition, establishing a trajectory tracking model, and generating final trajectory results. RFID technology is used to obtain the coordinate position of the tag in the plane and the direction relationship in the 3D space to obtain the motion trajectory of the stylus tip on the handwriting board. This method uses an antenna array to collect the phase changes of the tag during the three-dimensional motion process to restore the handwriting trajectory in the two-dimensional plane, and constructs a miniaturized handwriting trajectory tracking system that can achieve centimeter-level trajectory tracking effects and obtain handwriting trajectory results in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of high-precision radio frequency positioning, and in particular to a handwriting trajectory tracking method based on RFID. Background Art

[0002] In recent years, wireless positioning and tracking technology has attracted widespread attention in the academic community. Most existing work uses different signal features for trajectory tracking, such as Angle of Arrival (AOA), Channel State Information (CSI), Time of Flight (TOF) or a combination of them to improve the accuracy to sub-centimeter level. However, these methods usually require large phased array antennas, relatively large spectrum bandwidths, dedicated deployment areas, and environments with relatively little interference to achieve good performance. This will lead to problems such as increased deployment costs of the system and limitations on deployment areas.

[0003] With the development of radio frequency identification technology, RFID tags have been widely recognized as a lightweight passive positioning sensor. In RFID systems, passive RFID tags can backscatter wireless signals from the reader antenna, so the tags are usually attached to the tracked objects. The backscattered signals read by the reader can accurately track the movement trajectory of the target object. For example, using objects with RFID tags to write and draw on a virtual plane. Early work on RFID positioning was based on the received signal strength (RSS) method, but RSS is not a reliable position indicator, especially for ultra-high frequency RFID tags. It is highly related to the direction of the tag and the antenna gain, but in a mobile environment, the tag direction is often unknown. In recent years, people have become more and more interested in using phase information to locate tags, which has shown good performance in both positioning and motion detection. Although some solutions can achieve centimeter-level accuracy for tag positioning, they usually require complex deployment or lack real-time positioning capabilities, which is impractical in real-life application scenarios.

[0004] Therefore, based on the above issues, the use of miniaturized RFID devices to achieve handwriting tracking is crucial for the promotion of RFID-based sensing technology to smart homes, classrooms, and offices. Summary of the invention

[0005] The present invention provides a method for tracking handwriting trajectories using RFID phase values, which can perform dynamic real-time positioning and is suitable for tracking and identifying handwriting trajectories on a plane. The method uses a dual-tag four-antenna array to collect RF phases, and establishes a relationship model between 2D trajectories and 3D directions and RF phases from both time and space aspects, aiming to accurately track the real-time motion trajectory of RFID tags. The system has the advantages of simple deployment and strong real-time performance.

[0006] A handwriting trajectory tracking method based on RFID, characterized in that the method comprises the following steps:

[0007] Step 1: System construction and data collection: The system includes a handwriting board constructed with an antenna array composed of multiple directional gain RFID antennas, and a stylus constructed with a passive RFID tag as the core; write with the pen in the handwriting board area; the reader continuously collects real-time phase data of the RFID tag during the movement of the stylus through the antenna array;

[0008] Step 2: Data preprocessing: including noise reduction and unified sampling rate;

[0009] Step 3: Establish a trajectory tracking model: This model is divided into two parts: 2D trajectory tracking and 3D direction estimation. The 2D trajectory refers to the motion trajectory of the stylus tip on the writing plane, and the 3D direction refers to the direction of the stylus relative to the writing plane during the movement.

[0010] Step 4: Generate the final trajectory result: Combine the coordinate positions of the two tags in the plane and the direction relationship in the 3D space, further expand to the determination of the pen tip coordinates, and finally output the motion trajectory of the pen tip on the handwriting tablet.

[0011] Furthermore, in step 1, a stylus is constructed with a passive RFID tag as the core; a double tag is attached to the surface of the pen body, wherein the first tag is located near the tip of the pen, and the second tag is located near the tail of the pen; the two RFID tags are attached to both sides of the pen body, that is, the tags are placed in opposite directions so that they disperse and receive RFID signals from all directions.

[0012] Furthermore, when setting two RFID tags, it is necessary to determine the optimal position of the tag pair; while ensuring that the tags can be read by all antennas, the distance between the first tag and the pen tip is made as small as possible; while the tag distance does not exceed the length of the stylus, the distance between the tags is made as large as possible, so as to determine the distance between the tag pairs and minimize the mutual interference between the tags caused by electromagnetic backscatter coupling.

[0013] Furthermore, in step 1, a handwriting board is constructed with an RFID antenna array; the RFID antenna array is composed of small microstrip antennas to collect real-time phase information during the movement of the RFID tag; the tracking area of ​​the handwriting board is a rectangular plane preset in the antenna array.

[0014] Furthermore, the present invention integrates the RFID microstrip antenna into the handwriting board, and the pen with the label will write on the plane above the antenna. Since the distance between the label and the antenna is too close, the success rate and accuracy of phase data acquisition will be affected. Therefore, there is a certain distance between the label and the pen tip, which will cause the movement trajectory of the label and the movement trajectory of the pen tip to be different during actual writing. The dual-label array can be used to expand the obtained label coordinates to the pen tip coordinate position to determine the handwriting trajectory.

[0015] Furthermore, in step 2, the phase data collected in step 1 is resampled to align the phase data collected by different antennas; then, the timing RF phase of each tag is phase unfolded to remove the influence of the 2π jump of the phase; finally, the phase data is smoothed and filtered to remove environmental noise and interference.

[0016] Furthermore, in step 3, the input of the 2D trajectory tracking model is the difference in phase data of the same tag collected by two antenna pairs in the antenna array. For the two tags in the antenna pair, a hyperbola-based differential hologram positioning method is used to determine their positions in the coordinate system respectively.

[0017] Furthermore, a differential hologram positioning method based on a hyperbola is provided; the positioning area is divided into grid points of a certain size according to the requirements of positioning accuracy, and the difference between the theoretical phase difference obtained from the distance from each pair of antennas to each grid point and the measured value is used as input, and based on the characteristic that the distance difference from a point on the hyperbola to the focus is a constant, two pairs of antennas in the antenna array are selected to generate a pair of hyperbolas, and then the probability density function of appropriate parameters is used to map the intersection of multiple pairs of hyperbolas as the differential result in the form of a hologram.

[0018] Furthermore, in step 3, the input of the 3D direction estimation model is to obtain the coordinates of the tag in the 2D trajectory tracking model, and to establish a geometric model in space with the projection of the stylus on the writing plane and the known coordinate position relationship, and finally obtain the elevation angle and azimuth angle in space.

[0019] Furthermore, in step 4, the tag position and the elevation and azimuth of the tag in 3D space are utilized to extend the tracking of the tag position to the real-time position of the pen tip on the writing plane when the distance between the tag and the pen tip and the distance between the tags are known, thereby finally obtaining the handwriting trajectory.

[0020] The beneficial effects achieved by the present invention are: providing a handwriting trajectory tracking method based on RFID, the method and system can simultaneously provide the 2D trajectory and 3D direction information of the stylus, and can achieve the purpose of real-time tracking of the user's writing trajectory. Compared with the prior art, the present invention has the advantages of convenient deployment, strong real-time performance, and can achieve high-precision handwriting trajectory tracking effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of a handwriting recognition trajectory tracking method in an implementation example of the present invention.

[0022] Figure 2 It is a physical picture of the stylus design in the implementation example of the present invention.

[0023] Figure 3 It is a physical picture of the handwriting board design in the implementation example of the present invention.

[0024] Figure 4 It is a trajectory diagram generated in the implementation example of the present invention. DETAILED DESCRIPTION

[0025] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings.

[0026] See also Figure 1 , a flow chart of a handwriting trajectory tracking method based on RFID, which first takes the phase of the timing RF signal of the dual tags attached to the surface of the stylus during the movement read by the reader as input. Then data preprocessing is performed, and the timing RF phase is resampled to solve the random sampling problem based on the time-slot ALOHA protocol in the RFID system, while removing problems such as device diversity and RF phase periodicity. Next, the 2D trajectory tracking module uses the phase change data of the tag during the movement of continuous time points to restore the movement trajectory of the tag in the plane; at the same time, the 3D direction estimation module is used to obtain the change in the direction angle of the entire pen in space. Finally, the displacement of the pen tip is calculated based on the estimated 3D direction and 2D trajectory to restore the movement of the pen tip on the 2D plane, which specifically includes the following steps:

[0027] Step 1: System construction and data collection: Use multiple directional gain RFID antennas to form an antenna array to build a handwriting board, and use a passive RFID tag as the core to build a stylus. Write in the handwriting board area. The reader continuously collects real-time phase data of the RFID tag during the movement of the stylus.

[0028] Specifically, a RFID-based handwriting trajectory tracking system includes: an Impinj Speedway R420 reader, four RFID microstrip antennas and a combination of two commercial passive RFID tags. Figure 2The tags are attached to a stylus with a total length of 18 cm. The first tag is located 4 cm away from the tip of the stylus, and the second tag is located 16 cm away from the first tag. The tags are placed in opposite directions to disperse and receive RFID signals reflected from all directions. Figure 3 The antenna array consists of four 5.2×5.2cm small RFID microstrip antennas. The distance between the centers of the two antennas is 16cm, which is a half-wavelength distance and can eliminate phase ambiguity. The antennas can be embedded in the four corners of a square handwriting board with a side length of 20cm. The use of small microstrip antennas can greatly reduce the deployment space of the system. The four antennas are connected to an Impinj Speedway R420 reader, which can provide fine-grained signal characteristics to collect the phase information of the tag when the user uses a stylus with a label attached to the writing process. When the system is started, the reader sends high-power continuous radio frequency waves. The RFID tags with microchips and antennas in the tracking area can absorb the energy from the continuous radio frequency waves to drive their chips and respond to the reader's reply command. The reader then receives the backscattered signal from the tag to obtain the phase data of the tag during movement, and the collected phase data is input into the established handwriting trajectory tracking prototype system. When a user writes on a handwriting board with a stylus attached to an RFID tag, the user can obtain the trajectory tracking results in real time, ultimately transforming an ordinary stylus into an intelligent human-computer interaction device.

[0029] Step 2: Data preprocessing: Improve the reliability of RF phase by reducing the noise in the actual environment and unifying the sampling rate of the RFID system. First, smooth and filter the phase data collected in step 1 to remove environmental noise and some interference; then interpolate the phase data to align the phase data collected by different antennas; finally, phase unfold the timing RF phase of each tag to remove the impact of the 2π phase jump.

[0030] Specifically, step 2-1: smoothing of phase data. The phase measurement data collected from commercial RFID devices will be interfered by various noises in the environment, such as random white noise and interference from surrounding electromagnetic devices. Most of these noises follow a typical Gaussian distribution. The Kalman filter algorithm can be used to perform corresponding noise filtering on the measured phase value to obtain smoothed data.

[0031] Step 2-2: Phase data interpolation processing. Commercial RFID readers use the Slotted ALOHA algorithm (S-ALOHA algorithm) to query tags within the reading range. When the reader sends a query command, each tag randomly selects a time point to reply. Linear interpolation is used to align the two-phase sequence:

[0032]

[0033] Among them, θ(t k ) and θ(t k+1 ) represents two phase data at any adjacent time points. After data interpolation, the change of the phase diagram can more clearly reveal its implicit periodicity and reduce time deviation.

[0034] Step 2-3: Phase data unwrapping process. The measured phase can be calculated as:

[0035]

[0036] Where λ is the wavelength of the RF signal. For an RF carrier with a frequency of f (Hz), c is the speed of electromagnetic waves, which is approximately equal to the speed of light (≈3×10 8 m / s). d represents the distance between the reader and the tag, θ dev represents the phase shift introduced by the diversity of hardware devices, θ r Represents other noise interference. Since the phase is a periodic function, the value range of θ is [0,2π), so the phase jump needs to be eliminated. Assuming the phase change sequence is [θ(1),θ(2),…,θ(t)], then:

[0037]

[0038] Where, t>1.

[0039] Step 3: Establish a trajectory tracking model: This model is divided into two parts: 2D trajectory tracking and 3D direction estimation. The 2D trajectory refers to the motion trajectory of the pen tip on the writing plane, and the 3D direction refers to the direction of the pen relative to the writing plane during movement, which together reflect the user's writing content.

[0040] Specifically, step 3-1: determine the position of the first tag (the tag 2 cm away from the pen tip). Assume that the experimenter holds a pen with a tag and writes on a square handwriting board. The tag moves in 3D space as the pen moves. Deploy four small RFID microstrip antennas at the four vertices of the handwriting board, represented by A = {A1, A2, A3, A4}. In the four-antenna array, any three antennas can be combined to obtain Then, take the actual phase difference between any two antenna pairs among the three antennas measured by the reader. Divide the rectangular tracking area into 100×100 grid points and obtain the theoretical phase difference of the antenna pair at each grid point:

[0041]

[0042] The difference between the theoretical phase difference and the actual phase measured by the reader is input into the probability density function, which is expressed as:

[0043]

[0044] The smaller Δθ is, the greater the probability that the tag will appear on a certain associated grid. Among them, a controls the rate of descent of the function, and b controls the position of the zero point of the function. The position where the tag has the highest probability of appearing in the tracking area at a certain moment can be obtained. Finally, the tag position obtained by the phase difference of each group of antennas is averaged to obtain the final tag position. Since the entire movement of the pen is composed of a series of instantaneous postures at different time points, the position at each moment can be superimposed to obtain the trajectory shape of the tag movement.

[0045] Step 3-2: Determine the position of the second tag (16 cm away from the first tag). First, obtain the phase difference between the two tags measured by the same antenna at a certain moment, construct a hologram positioning method on a curved surface in 3D space, and take the average of the tag position coordinates finally output by the four antennas to obtain the final tag position.

[0046] Step 3-3: Get the trajectory of the pen tip. After knowing the positions of the first label and the second label, the elevation and azimuth of the pen in three-dimensional space can be obtained according to a specific geometric relationship. The displacement of the pen tip along the X-axis and Y-axis on a two-dimensional plane can be expressed as:

[0047]

[0048] Among them, (x B ,y B ,0) represents the absolute position of the pen tip, l i,x represents the projection of the length from the i-th label to the pen tip in the X-axis direction, l i,y represents the projection of the length from the i-th label to the pen tip in the Y-axis direction. The direction of the pen can be uniquely determined by the elevation angle and azimuth angle. The elevation angle β e Indicates the inclination angle between the pen and the XY plane (the writing plane of the pen), the azimuth angle β a It represents the angle between the projection of the pen on the XY plane and the X axis. When the pen moves on the XY plane, β e The value range is β a The value range of is (0,2π).

[0049] Step 4: Output the final trajectory result: Input the preprocessed phase data into the constructed trajectory tracking prototype system and output the result, which can be regarded as a real-time interaction process between the user and the computer.

[0050] Specifically, when a user holds a pen with a label and writes on a flat surface, the label will move or rotate as the user writes, but the pen tip is always on the flat surface of the tablet. The entire writing action is composed of a series of instantaneous gestures at different time points. Each instantaneous gesture can be represented by the absolute position of the pen tip on the two-dimensional plane and the direction of the pen in three-dimensional space. Finally, combined with the position of the pen tip (x B ,y B ,0) and the direction of the pen shaft <β e ,β a >Can accurately describe the pen's instantaneous posture.

[0051] The positioning method based on the hyperbola differential hologram is specifically to divide the tracking area into W×L grid points, and the mapped hologram is also a W×L pixel map. The pixel value represents the possibility that the associated grid is the tag position. The difference between the theoretical phase value difference and the measured phase value difference between two pairs of tags is used to generate the RF hologram by constructing an appropriate probability density function, and finally locate the position of the tag.

[0052] The present invention proposes a handwriting trajectory tracking system based on RFID, which models the relationship between the RF phase and the 2D motion and 3D direction of a tagged pencil, and uses four fixed antennas to achieve high-precision real-time tracking of passive RFID tags. First, the phase difference of different antennas for the same tag is extracted and input into the trajectory tracking and positioning algorithm based on the hyperbola differential hologram to determine the position of the tag close to the pen tip. Then, the phase difference between the two tags collected by the same antenna is extracted, and a hologram on an aerial surface is constructed to determine the position of the second label. Finally, when the position and distance of the two tags are known, the movement of the two tags in three-dimensional space is extended to the tracking of the position of the pen tip on a two-dimensional plane. On this basis, we implemented the prototype system of the present invention and tested its performance in a practical environment. Reference Figure 4 ,The experimental results confirm the effectiveness of the system in 2D motion and 3D ,direction, and the final trajectory tracking result can achieve centimeter-level ,accuracy.

[0053] The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiment. Any equivalent modifications or changes made by ordinary technicians in this field based on the contents disclosed by the present invention should be included in the protection scope recorded in the claims.

Claims

1. A handwriting trajectory tracking method based on RFID, characterized in that: The method comprises the following steps: Step 1: System construction and data collection: The system includes a handwriting board constructed with an antenna array composed of multiple directional gain RFID antennas, and a stylus constructed with a passive RFID tag as the core; Hold the pen and write in the writing area; The reader continuously collects real-time phase data of the RFID tag during the movement of the stylus through the antenna array; Step 2: Data preprocessing: including noise reduction and unified sampling rate; In step 2, the phase data collected in step 1 is resampled to align the phase data collected by different antennas; then the timing RF phase of each tag is phase unfolded to remove the influence of the 2π phase jump; finally, the phase data is smoothed and filtered to remove environmental noise and interference; Step 3: Establish a trajectory tracking model: This model is divided into two parts: 2D trajectory tracking and 3D direction estimation. The 2D trajectory refers to the motion trajectory of the stylus tip on the writing plane, and the 3D direction refers to the direction of the stylus relative to the writing plane during the movement. In step 3, the input of the 2D trajectory tracking model is the difference in phase data of the same tag collected by two antenna pairs in the antenna array. For the two tags in the antenna pair, the hyperbola-based differential hologram positioning method is used to determine their positions in the coordinate system respectively; Hyperbola-based differential hologram positioning method: The positioning area is divided into grid points of a certain size according to the positioning accuracy requirements. The difference between the theoretical phase difference and the measured value obtained from the distance between the two antenna pairs to each grid point is used as input. According to the characteristic that the distance difference from the point on the hyperbola to the focus is a constant, two pairs of antennas in the antenna array are selected to generate a pair of hyperbolas. Then, the probability density function is used to map the intersection of multiple pairs of hyperbolas as the differential result in the form of a hologram. The probability density function expression is as follows: Where a controls the rate of descent of the function, and b controls the zero position of the function; In step 3, the input of the 3D direction estimation model is the coordinates of the tag obtained in the 2D trajectory tracking model, and a geometric model in space is established based on the projection of the stylus on the writing plane and the known coordinate position relationship, and finally the elevation angle and azimuth angle in space are obtained; Step 4: Generate the final trajectory result: Combine the coordinate positions of the two tags in the plane and the direction relationship in the 3D space, further expand to the determination of the pen tip coordinates, and finally output the motion trajectory of the pen tip on the handwriting tablet.

2. The RFID-based handwriting trajectory tracking method according to claim 1, characterized in that: In step 1, a stylus is constructed with a passive RFID tag as the core; a double tag is attached to the surface of the pen body, wherein the first tag is located near the tip of the pen, and the second tag is located near the tail of the pen; two RFID tags are attached to both sides of the pen body, that is, the tags are placed in opposite directions so that they can disperse and receive RFID signals from all directions.

3. The RFID-based handwriting trajectory tracking method according to claim 2, characterized in that: When setting up two RFID tags, it is necessary to determine the optimal position of the tag pair; make the distance between the first tag and the pen tip as small as possible while ensuring that the tags can be read by all antennas; make the distance between the tags as large as possible while ensuring that the tag distance does not exceed the length of the stylus, so as to determine the distance between the tag pairs and minimize the mutual interference between the tags caused by electromagnetic backscatter coupling.

4. The RFID-based handwriting trajectory tracking method according to claim 1, characterized in that: In step 1, a handwriting board is constructed with an RFID antenna array; the RFID antenna array is composed of small microstrip antennas to collect real-time phase information during the movement of the RFID tag; the tracking area of ​​the handwriting board is a rectangular plane preset in the antenna array.

5. The RFID-based handwriting trajectory tracking method according to claim 3 or claim 4, characterized in that: The RFID microstrip antenna is integrated into the handwriting tablet, and the pen with a tag will write on the plane above the antenna. The dual-tag array is used to expand the obtained tag coordinates to the pen tip coordinate position to determine the handwriting trajectory.

6. The RFID-based handwriting trajectory tracking method according to claim 1, characterized in that: In step 4, using the tag position and the elevation and azimuth of the tag in 3D space, and given the known distance between the tag and the pen tip and the distance between the tags, the tracking of the tag position is extended to the real-time position of the pen tip on the writing plane, and finally the handwriting trajectory is obtained.

Images