User identification system, method and device based on handwriting recognition

Abstract

This invention provides a user identification system, method, and apparatus based on handwriting recognition. The system includes a device identification component, a feature storage component, and a similarity matching component. The method includes identifying a dot-matrix digital pen to obtain a first dot-matrix digital pen belonging to an administrator and a second dot-matrix digital pen belonging to a user; acquiring the writing trajectory input by the second dot-matrix digital pen, extracting writing features from the writing trajectory to obtain current writing features, and storing these features in a writing feature library on a server; acquiring a new writing trajectory input by the second dot-matrix digital pen, sending the new writing trajectory to the server, and performing similarity matching with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identification of the new writing trajectory is successful; otherwise, it is unsuccessful. The apparatus includes a dot-matrix digital pen, a server, a display, and dot-matrix paper, etc. This invention can not only effectively identify user identity but also prevent identity theft and unauthorized access.

Description

Technical Field

[0001] This invention relates to the field of identity recognition technology, and in particular to a user identity recognition system, method and apparatus based on handwriting recognition. Background Technology

[0002] A dot-matrix digital pen is a type of digital input device that converts handwriting or drawing actions into digital signals. It typically contains electronic components that use sensors to capture the pen's movement on a surface. It can connect to computers or tablets and also features pressure sensing capabilities to detect the user's writing pressure and produce lines of varying thickness. A dot-matrix digital pen works by printing an invisible dot pattern onto ordinary paper. A high-speed camera at the tip of the pen continuously captures the pen's movement, while a pressure sensor transmits the pressure data back to a data processor. The information is then transmitted externally via Bluetooth or USB. Due to the size limitations of dot-matrix digital pens, there are limitations in chip capabilities and battery life, making complex input / output and interactive features impossible.

[0003] In teaching scenarios, dot-matrix digital pens need to be linked to students. Each pen has a unique serial number, which is included in the pen's output handwriting information. This serial number may be printed directly on the pen's casing after production. Administrators can link each student and each pen one-to-one via the backend interface, using the student ID and pen's serial number; one-to-many linking is also possible to support multiple pens under one student's name. However, this method is cumbersome for administrators, and the similar appearance of dot-matrix digital pens leads to low recognition accuracy. When the pen is fully charged and retrieved, it's easy for students to mistakenly use someone else's pen; it also fails to prevent students from intentionally using other people's pens.

[0004] Prior art 1, application number: CN 202210797962.9, discloses a dot-matrix digital pen and its manufacturing method, including: a pen barrel, having a shaft and a mounting head connected to the shaft, a camera disposed within the mounting head, and a processor disposed within the shaft; a pen tip, installed within the shaft; a pen nib, having multiple positioning pins for positioning on the mounting head; and a nib cap for connecting the nib to the mounting head. Both the mounting head and the nib have a partially circular viewfinder, through which the camera captures handwriting to determine the relative coordinates of the handwriting and the pen tip. Although the camera captures images through the partially circular viewfinder on the mounting head and the nib, the image captured by the camera through this partially circular viewfinder is smaller than a full circle. During writing, the image captured when the pen barrel rotates can indicate the direction of the handwriting, and the high-definition camera acquires smaller image data, resulting in faster data processing. However, dot-matrix digital pens cannot be distinguished by appearance, leading to easy confusion during use and hindering personalized configuration.

[0005] Prior art two, application number: CN 202111194655.3, discloses a method for judging pen grip posture when writing with a dot-matrix pen, including the following steps: allowing the user to write with a dot-matrix digital pen; acquiring real-time information of the user's writing and preprocessing it into usable features; inputting the extracted writing feature information into a pre-trained pen grip posture recognition model; judging whether the user used an incorrect pen grip posture when writing, and if so, further judging which incorrect pen grip posture is most likely; uploading the analysis results to an information processing module, such as a smartphone or the dot-matrix pen's own information processing module, via Bluetooth or USB communication module, and then providing feedback to the user through a programmed reminder. Although it achieves the judgment and processing of pen grip adaptation, the dot-matrix pen's function is relatively simple and cannot realize the function of identity recognition, which can still lead to confusion in teaching scenarios.

[0006] Existing technology three, application number: CN 202110961674.8, discloses a method for recognizing emotional states through digital handwriting, including the following steps: a user writes text on dot-matrix paper using a dot-matrix digital pen; real-time information data of the user's writing is acquired; the acquired real-time information data is processed to obtain handwriting features more closely related to emotion tags, wherein the obtained handwriting features more closely related to emotion tags are normalized; the normalized handwriting feature data is input into a pre-trained emotion recognition model to obtain the user's emotional state category. Although this solves the problem of complex and demanding information collection conditions in existing emotion recognition schemes, and uses font feature methods to identify the features of the user's handwriting before classifying the emotional state, thus eliminating most physical hardware devices and physiological signal detection devices; at the same time, digital handwriting technology is more accurate in recognizing handwriting features. However, it lacks the ability to recognize identity features through handwriting, resulting in a low level of functional intelligence.

[0007] Current technologies 1, 2, and 3 suffer from cumbersome administrator operations, similar appearance of dot-matrix digital pens resulting in low recognition accuracy, and an inability to prevent students from intentionally using other people's pens. Therefore, this invention provides a user identification system, method, and apparatus based on handwriting recognition, authenticating user identity based on the content of the handwriting. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides a user identification system based on handwriting recognition, comprising:

[0009] The device identification component is responsible for identifying the dot matrix digital pen, obtaining the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory.

[0010] The feature storage component is responsible for acquiring the writing trajectory of the second dot matrix digital pen input, extracting writing features from the writing trajectory to obtain the current writing features, and storing them in the writing feature library on the server. The current writing features are used to compare with the writing trajectory newly input by the user to obtain the user's login name and establish the association between the login name and the writing trajectory.

[0011] The similarity matching component is responsible for obtaining the new writing trajectory of the second dot matrix digital pen input, sending the new writing trajectory to the server, and matching it with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is successful; otherwise, it is unsuccessful.

[0012] Optional, device identification components, including:

[0013] The first marking operation module is responsible for obtaining the number of dot matrix digital pens to be marked, and at the same time obtaining the running status of the dot matrix digital pens within a preset time period, and sending the running status to the server; randomly selecting one as the first dot matrix digital pen, retrieving the trigger operation corresponding to the administrator's operation permission from the server, and marking the first dot matrix digital pen with the administrator operation mark.

[0014] The second marking operation module is responsible for selecting at least one from other dot matrix digital pens as the second dot matrix digital pen, retrieving the trigger operation corresponding to the writing trajectory from the server, and marking the second dot matrix digital pen with the user operation mark.

[0015] The identifier association operation module is responsible for responding to the first dot matrix digital pen identifier administrator operation mark and the second dot matrix digital pen identifier user operation mark. Upon receiving the identifier parsing request sent by the server, it performs identifier parsing processing on the identifier in the identifier parsing request to obtain the relationship between the dot matrix digital pen and the trigger operation corresponding to the identifier.

[0016] Optional, feature storage component, including:

[0017] The content information acquisition module is responsible for recording the user's writing content in real time through the built-in sensor of the dot matrix digital pen;

[0018] The writing sequence generation module is responsible for converting the writing content into characters and their order, generating a writing sequence, recording the start and end times of writing each character in the writing sequence, and calculating the writing duration of each character in the writing content by combining the timestamp; it calculates the writing speed of the character by the start and end displacement of each character, and at the same time, it cooperates with the data of the pressure sensor to identify the pressure changes applied to each character.

[0019] The association management module is responsible for storing writing feature vectors in a structured manner in the feature repository, forming writing feature records corresponding to users, assigning a unique identifier to each writing feature, and establishing an association relationship with the user's login name.

[0020] Optional, the content information acquisition module includes:

[0021] The device startup submodule is responsible for turning on the dot matrix digital pen and starting the display device; at the same time, it initializes and establishes a connection with the writing surface; the accelerometer and pressure sensor inside the dot matrix digital pen perform self-checks and calibrate the sensed initial state.

[0022] The path acquisition submodule is responsible for establishing a real-time two-dimensional coordinate system within the writing range of the dot matrix digital pen, with the axis position set as the starting writing point. Whenever writing begins, it starts to capture every minute movement of the pen tip, records the coordinates of the pen tip, and generates a set of coordinate data for each writing action, forming a writing trajectory path.

[0023] The pressure acquisition submodule is responsible for monitoring the pressure applied to the pen tip in different areas to form a multi-dimensional pressure dataset. The pressure value is recorded in real time for each writing action, and different levels of writing intensity are monitored in real time to form a pressure waveform. The pressure data and its corresponding timestamps are simultaneously written to the data storage to form a time-pressure sequence.

[0024] Optionally, the sequence generation module includes a vector forming submodule, which contains:

[0025] The character outline forming unit is responsible for identifying the start and end coordinates of strokes. Whenever the dot matrix digital pen touches or leaves the paper, it records a coordinate. The start coordinate is the point where the stroke begins, and the end coordinate is the point where the stroke ends. It draws the shape of each stroke. All strokes are arranged in chronological order according to the timestamp. Adjacent strokes are integrated, and individual strokes are connected into a character dataset.

[0026] The coordinate path calculation unit is responsible for matching the end coordinate of each stroke with the start coordinate of the next stroke. The duration of each stroke is associated with its corresponding spatial path, reconstructing a dynamic writing image on the time axis; forming an ordered series of coordinate paths.

[0027] The feature vector integration unit is responsible for analyzing the geometric features of each stroke, including the curvature, length, width, and closure of the lines, forming a set of geometric parameters; calculating the writing time of each stroke and the time interval between strokes; tracking and recording pressure changes based on the pressure applied by the pen tip during writing; and integrating all information into a comprehensive multidimensional feature vector.

[0028] Optionally, the character dataset of the character contour forming unit contains the start and end coordinates of each stroke and the execution order during the writing process, forming a complete character contour.

[0029] Optional, the association management creation module includes:

[0030] The feature identifier combination submodule is responsible for generating a unique user identifier for each user as a unique identifier; it records in real time through the sensor of the dot matrix digital pen, which includes the user's unique identifier related to handwriting, and combines the newly generated writing features with the user's unique identifier.

[0031] The vector storage submodule is responsible for storing the written feature vectors in the feature repository in the form of structured data; the user unique identifier links the written features to a specific user; and the writing timestamp records the start and end times of the writing process.

[0032] The function definition submodule is responsible for assigning tags to each writing record for subsequent filtering and visualization or personalized recommendations; recording user operations and changes in writing habits, and tracking user writing patterns; and accessing one's own historical writing records through a unique identifier.

[0033] Optional, similarity matching component, including:

[0034] The group lookup module is responsible for using a predefined hash function to input the current writing feature vector into the hash function and calculate a corresponding hash value. The hash value will be used to identify which user group the current writing feature belongs to. After obtaining the generated hash value, the module searches the writing feature database for all user groups that match the hash value.

[0035] The similarity calculation module is responsible for comparing the similarity of all writing features within the user group corresponding to the hash value; normalizing the newly input writing features and the stored writing features; and calculating the similarity score between the current writing feature and the stored features.

[0036] The threshold comparison module is responsible for determining the similarity score between each stored writing feature within a user group and the current writing feature. If the similarity score found is greater than or equal to the preset threshold, the user's identity authentication will be considered successful, and the corresponding user login name will be returned, establishing the association between the writing trajectory and the user's identity. If the similarity scores of all stored features within the group and the current writing feature do not reach the preset threshold, the identity authentication will be marked as failed.

[0037] This invention provides a user identification method based on handwriting recognition, comprising the following steps:

[0038] The dot matrix digital pen is identified to obtain the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory.

[0039] The writing trajectory input by the second dot matrix digital pen is obtained, the writing features are extracted from the writing trajectory to obtain the current writing features, and stored in the writing feature library of the server; the current writing features are compared with the writing trajectory newly input by the user to obtain the user's login name and establish the association between the login name and the writing trajectory;

[0040] The new writing trajectory input by the second dot matrix digital pen is obtained and sent to the server. The new writing trajectory is then matched with the writing features in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is successful; otherwise, it is unsuccessful.

[0041] The present invention provides a user identification device based on handwriting recognition, comprising: a plurality of dot-matrix digital pens, a terminal device, a server, and a writing feature library;

[0042] The dot matrix digital pen connects to the terminal device wirelessly, and the terminal device connects to the server via a network. The writing feature library is installed in the server; the dot matrix digital pen can also connect directly to the server wirelessly.

[0043] The device identification component of this invention identifies two different dot-matrix digital pens: a first dot-matrix digital pen for administrators and a second dot-matrix digital pen for users. Administrators can use their dot-matrix digital pens to perform permission operations on the terminal device, while users use their dot-matrix digital pens to input writing trajectories. This ensures the system can distinguish between administrators and ordinary users, enabling appropriate access control and laying the foundation for system security by preventing unauthorized access. Significance: By effectively identifying different roles, the system's security and manageability are enhanced, ensuring that only specific personnel can perform specific operations. The feature storage component acquires the writing trajectory input by the user using the second dot-matrix digital pen, extracts writing features, and stores the extracted features in the server's writing feature library. Simultaneously, it can compare these features with new writing trajectories input by the user and associate them with the user's login name. This provides a precise way to capture and store user writing features, enabling the association of user identity information with their unique writing characteristics and improving the accuracy of identity recognition. Significance: By storing user writing features, subsequent identity verification can be quickly and effectively compared and identified, providing a user-friendly experience and improving system usability. The similarity matching component can dynamically compare and verify a user's handwriting characteristics to quickly determine their identity, providing a real-time and highly efficient authentication process. Significance: Through effective similarity matching, the system can ensure that only users matching specific handwriting patterns can access their accounts, improving overall security and fraud prevention capabilities.

[0044] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0045] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0046] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0047] Figure 1 This is a block diagram of a user identification system based on handwriting recognition in Embodiment 1 of the present invention;

[0048] Figure 2 This is a block diagram of the device identification component in Embodiment 2 of the present invention;

[0049] Figure 3 This is a block diagram of the feature storage component in Embodiment 3 of the present invention;

[0050] Figure 4 This is a block diagram of the content information acquisition module in Embodiment 4 of the present invention;

[0051] Figure 5 This is a block diagram of the sequence generation module in Embodiment 5 of the present invention;

[0052] Figure 6 This is a block diagram of the vector forming submodule in Embodiment 6 of the present invention;

[0053] Figure 7 A module block diagram for association management in Embodiment 7 of the present invention is provided.

[0054] Figure 8 This is a block diagram of the similarity matching component in Embodiment 8 of the present invention;

[0055] Figure 9 This is a block diagram of the similarity calculation module in Embodiment 9 of the present invention;

[0056] Figure 10 This is a flowchart of the user identification method based on handwriting recognition in Embodiment 10 of the present invention;

[0057] Figure 11 This is a block diagram of a user identification device based on handwriting recognition in Embodiment 11 of the present invention. Detailed Implementation

[0058] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0059] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the embodiments of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0060] In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application. In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0061] Example 1: As Figure 1 As shown, this embodiment of the invention provides a user identification system based on handwriting recognition, comprising:

[0062] The device identification component is responsible for identifying the dot matrix digital pen, obtaining the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory.

[0063] The writing traces include letters, numbers, Chinese characters, or a specific self-created symbol;

[0064] The feature storage component is responsible for acquiring the writing trajectory of the second dot matrix digital pen input, extracting writing features from the writing trajectory to obtain the current writing features, and storing them in the writing feature library on the server. The current writing features are used to compare with the writing trajectory newly input by the user to obtain the user's login name and establish the association between the login name and the writing trajectory.

[0065] The similarity matching component is responsible for obtaining the new writing trajectory of the second dot matrix digital pen input, sending the new writing trajectory to the server, and matching it with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is successful; otherwise, it is unsuccessful.

[0066] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the device identification component identifies the dot matrix digital pen to obtain the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory; the feature storage component obtains the writing trajectory input by the second dot matrix digital pen, extracts writing features from the writing trajectory, obtains the current writing features, and stores them in the writing feature library of the server; the current writing features are compared with the newly input writing trajectory of the user to obtain the user's login name and establish the association between the login name and the writing trajectory; the similarity matching component obtains the new writing trajectory input by the second dot matrix digital pen, sends the new writing trajectory to the server, and performs similarity matching with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is passed; otherwise, it is not passed. The device identification component of the above scheme identifies two different dot-matrix digital pens: the administrator's first dot-matrix digital pen and the user's second dot-matrix digital pen. Administrators can use their dot-matrix digital pens to perform permission operations on the terminal devices, while users use their dot-matrix digital pens to input writing trajectories. This ensures the system can distinguish between administrators and ordinary users, enabling appropriate access control and laying the foundation for system security by preventing unauthorized access. The content of the writing trajectory is not limited to common letters, numbers, and Chinese characters; it can also be a specific symbol created by the student. On the one hand, this attracts users and facilitates product promotion; on the other hand, the greater diversity of personalized symbols among different users can improve recognition accuracy. Significance: By effectively identifying different roles, the system's security and manageability are enhanced, ensuring that only specific personnel can perform specific operations. The feature storage component acquires the writing trajectory input by the user using the second dot-matrix digital pen, extracts writing features, and stores the extracted features in the server's writing feature library. Simultaneously, it can compare this with the user's newly input writing trajectory and associate it with the user's login name. This provides a precise way to capture and store user writing features, enabling the association of user identity information with their unique writing characteristics and improving the accuracy of identity recognition. Significance: By storing users' handwriting characteristics, subsequent authentication can be quickly and effectively compared and identified, providing a user-friendly experience and improving system usability. The similarity matching component can dynamically compare and verify users' handwriting characteristics, quickly determining their identity; providing a real-time and highly efficient authentication process. Significance: Through effective similarity matching, the system can ensure that only users matching specific handwriting patterns can access their accounts, improving overall security and fraud prevention capabilities.

[0067] In summary, this embodiment enhances system security, effectively identifying users and preventing identity theft and unauthorized access. Handwriting recognition-based authentication provides a natural and intuitive user verification method. Compared to traditional passwords or PIN codes, users experience less memory burden and a better user experience when inputting their handwriting. By linking handwriting patterns to user identity, the system can collect and utilize user behavioral data to support further personalized services and information. Through the collaborative efforts of various components, a secure, efficient, and user-friendly identity verification process is achieved, enhancing both system security and reliability while also increasing user engagement and convenience. Dot-matrix digital pens require writing on special dot-matrix paper. To prevent others from forging the authentication process by copying, the following points can be considered: The dots on the dot-matrix paper have a vast numbering space, allowing for unique identification. Each authentication simultaneously records the dot-matrix paper information, preventing duplicate input at the same location. While the writing trajectory can be copied, information such as writing speed and pressure is difficult to replicate. Therefore, the aforementioned trajectory information should not only be a set of visible dot coordinates but should also include information such as writing speed and pressure. For current dot-matrix digital pens, writing speed and pressure information are already readily available. For tablets, only speed information may be available; when the device supports more information, it should use as much information as possible.

[0068] Example 2: Figure 2 As shown, based on Embodiment 1, the device identification component provided in this embodiment of the invention includes:

[0069] The first marking operation module is responsible for obtaining the number of dot matrix digital pens to be marked, and at the same time obtaining the running status of the dot matrix digital pens within a preset time period, and sending the running status to the server. The running status includes writing time, time without writing action, and charging time, etc.; randomly select one as the first dot matrix digital pen, retrieve the trigger operation corresponding to the administrator's operation permission from the server, and mark the first dot matrix digital pen with the administrator operation mark.

[0070] Among them, the running status time is used to automatically trigger actions, and the user login authentication action should occur after charging; when collecting writing trajectory, the user to be collected is determined based on the writing time and the time without writing action;

[0071] The second marking operation module is responsible for selecting at least one from other dot matrix digital pens as the second dot matrix digital pen, retrieving the trigger operation corresponding to the writing trajectory from the server, and marking the second dot matrix digital pen with the user operation mark.

[0072] The identifier association operation module is responsible for responding to the first dot matrix digital pen identifier administrator operation mark and the second dot matrix digital pen identifier user operation mark. Upon receiving the identifier parsing request sent by the server, it performs identifier parsing processing on the identifier in the identifier parsing request to obtain the relationship between the dot matrix digital pen and the trigger operation corresponding to the identifier.

[0073] The working principle and beneficial effects of the above technical solution are as follows: The first marking operation module of this embodiment obtains the number of dot matrix digital pens to be marked, and at the same time obtains the running status of the dot matrix digital pens within a preset time period, and sends the running status to the server. The running status includes writing time, time without writing action, and charging time, etc.; arbitrarily selects one as the first dot matrix digital pen, retrieves the trigger operation corresponding to the administrator operation permission from the server, and assigns an administrator operation mark to the first dot matrix digital pen; the second marking operation module selects at least one from other dot matrix digital pens as the second dot matrix digital pen, retrieves the trigger operation corresponding to the writing trajectory from the server, and assigns a user operation mark to the second dot matrix digital pen; the marking association operation module responds to the first dot matrix digital pen marking administrator operation mark and the second dot matrix digital pen marking user operation mark, receives the marking parsing request sent by the server, performs marking parsing processing on the marking in the marking parsing request, and obtains the relationship between the dot matrix digital pen and the trigger operation corresponding to the marking. The first marking module of the above scheme can acquire the number of dot matrix digital pens to be marked and their operating status (such as writing time, no writing time, and charging time) within a preset time period, providing basic data for the identification and management of dot matrix digital pens. It selects one dot matrix digital pen from the acquired pens as the first dot matrix digital pen and marks it as an administrator operation, retrieving the corresponding trigger operation flag through the server. Significance: Clearly distinguishes between administrator and user devices, ensuring administrators have the necessary operating permissions, facilitating effective management of permissions for different roles within a system, and ensuring system security; real-time monitoring of the dot matrix digital pen's operating status helps ensure normal device operation and maintenance, reduces failure rates, and improves system availability. The second marking module selects at least one other dot matrix digital pen as the second dot matrix digital pen and marks it as a user operation, retrieving the trigger operation of the writing trajectory through the server; it can process multiple dot matrix digital pens simultaneously, enhancing the system's support for different users. Significance: By identifying the user's operation on the second dot-matrix digital pen, the system establishes a unique user identity, ensuring that only authorized users can write and operate the pen. This guarantees reliable user authentication and provides security for subsequent operations and data. The modular design allows the system to adapt to the needs of multiple users, improving its flexibility and scalability. The identifier association operation module responds to the identifiers of the first and second dot-matrix digital pens, processes identifier parsing requests from the server, and determines the relationship between the dot-matrix digital pen and its corresponding trigger operation. It integrates the operation identifiers of administrators and users to form a complete control chain. Significance: By parsing the relationship graph formed by the identifier requests, the system ensures that each operation has appropriate access control, thereby enhancing the overall security of the system. It can generate audit logs based on user and administrator operations, which helps with subsequent monitoring and security auditing, ensuring that all operations can be traced and enhancing transparency and accountability.

[0074] In summary, this embodiment significantly improves system security through device identification and access control, preventing potential unauthorized access and operations. It provides clear identifiers for different operations by administrators and users, facilitating the rational allocation of system permissions and improving management efficiency. Real-time monitoring of device status and user operations enhances system availability and maintainability, reducing the risk of system failures. Furthermore, the data relationships established through identifier resolution support auditing and security tracking, providing a foundation for subsequent decision-making and security incident response.

[0075] Example 3: As Figure 3 As shown, based on Embodiment 1, the feature storage component provided in this embodiment of the invention includes:

[0076] The content information acquisition module is responsible for recording the user's writing content in real time through the built-in sensors of the dot matrix digital pen. The writing content includes writing trajectory, writing content, writing speed and pressure sensing, etc.; the writing content includes letters, numbers, Chinese characters and specific symbols.

[0077] Among them, writing speed is calculated by recording the start and end times of the user's writing to determine the total writing time; pressure sensing records the pressure applied by the user while writing through feedback from the pen tip sensor, reflecting changes in writing force.

[0078] The writing sequence generation module is responsible for converting the writing content into characters and their order, generating a writing sequence, recording the start and end times of writing each character in the writing sequence, and calculating the writing duration of each character in the writing content by combining the timestamp; it calculates the writing speed of the character by the start and end displacement of each character, and at the same time, it cooperates with the data of the pressure sensor to identify the pressure changes applied to each character.

[0079] The association management module is responsible for storing writing feature vectors in a structured manner in the feature repository, forming writing feature records corresponding to users, assigning a unique identifier to each writing feature, and establishing an association relationship with the user's login name.

[0080] Among them, spatial features include constructing the geometric outline of a character based on the stroke combination, the starting point and ending point of the stroke; temporal features include combining writing duration and writing speed to form a time series, providing information on writing rhythm; if the writing speed of a character is high and the writing time is short, it is considered a specific fast writing feature; pressure features include segmenting and analyzing the pressure value applied during the writing process to form a pressure waveform, identifying the user's writing force pattern, such as changes in pressure and speed, corresponding to the user's writing habits;

[0081] Multidimensional writing feature vectors include: geometric features of characters (such as the number of strokes, stroke length, etc.), a list of writing speed and pressure values ​​that reflect the dynamic changes during the writing process; and specific pattern recognition, such as the unique writing rhythm, speed and pressure combination of specific letters or symbols.

[0082] The working principle and beneficial effects of the above technical solution are as follows: The content information acquisition module of this embodiment records the user's writing content in real time through the sensor built into the dot matrix digital pen. The writing content includes writing trajectory, writing content, writing speed, and pressure sensing, etc.; the writing content includes letters, numbers, Chinese characters, and specific symbols; the writing sequence generation module converts the writing content into characters and their order, generates a writing sequence, records the start and end time of writing each character in the writing sequence, and calculates the writing duration of each character in the writing content by combining the timestamp; the writing speed of the character is calculated by the start and end displacement of each character, and at the same time, in conjunction with the data from the pressure sensor, the pressure change applied to each character is identified; the association management establishment module stores the writing feature vector in a structured manner in the feature repository to form the writing feature record corresponding to the user, assigns a unique identifier to each writing feature, and establishes an association relationship with the user's login name. The content information acquisition module of the above solution uses the built-in sensors of the dot-matrix digital pen to record the user's writing dynamics in real time, capturing information such as writing trajectory, content (letters, numbers, Chinese characters, specific symbols), and writing speed and pressure. It can automatically summarize the user's writing behavior to form a dynamic writing record, including feedback data from multiple dimensions (such as speed and pressure). Significance: Ensuring the completeness and accuracy of the recorded writing information is the foundation for subsequent analysis and feature extraction; the information capture capability provides rich raw data for subsequent identification and verification processes; by monitoring writing behavior in real time and providing instant feedback, it can help users better understand their writing habits and improve the user experience. The writing sequence generation module transforms the real-time acquired writing content into an ordered writing sequence, recording the start and end times of each character and calculating its writing duration; based on the start and end positions of each character, it calculates the writing speed and, combined with pressure data, identifies the pressure changes applied to each character. Significance: By constructing a writing sequence, a temporal dimension is provided for subsequent analysis, enabling accurate quantification of writing activities. Writing duration and speed data provide crucial information about the user's writing rhythm. Combining character writing duration and speed reveals more detailed user writing habits and individual characteristics, which is essential for subsequent user identification. The association management module, based on the acquired writing duration, speed, and pressure changes, performs complex feature extraction to generate multi-dimensional writing feature vectors. These extracted feature vectors are stored in a structured feature repository, establishing an association with the user's login name. Significance: By generating multi-dimensional feature vectors containing spatial, temporal, and pressure features, each user's writing characteristics are fully personalized, enhancing the system's recognition capabilities. Storing writing features in a structured manner not only facilitates subsequent identity verification and matching but also provides an effective data foundation for analysis and comparison.By assigning unique identifiers, users' handwriting characteristics data can be effectively managed and retrieved, enhancing the system's performance and reliability.

[0083] In summary, this embodiment forms a complete handwriting feature acquisition and storage system. It integrates multiple handwriting features to ensure the accuracy and reliability of user identification, covering various handwriting styles and habits; it ensures accurate capture of the user's handwriting pattern at every step, providing strong support for subsequent comparisons; personalized feature recording enhances the security of user authentication, while providing a dynamic system capable of recognizing changes in user handwriting, improving the user experience; and structured data storage and management enable fast and efficient retrieval during authentication, improving the overall system performance.

[0084] Example 4: Figure 4 As shown, based on Embodiment 3, the content information acquisition module provided in this embodiment of the invention includes:

[0085] The device startup submodule is responsible for turning on the dot matrix digital pen and starting the display device; at the same time, it initializes and establishes a connection with the writing surface (such as paper or digital tablet); the accelerometer and pressure sensor inside the dot matrix digital pen perform self-tests and calibrate the sensed initial state.

[0086] The path acquisition submodule is responsible for establishing a real-time two-dimensional coordinate system within the writing range of the dot matrix digital pen, with the axis position set as the starting writing point. Whenever writing begins, it starts to capture every minute movement of the pen tip, recording the X and Y coordinates of the pen tip at a frequency of 1000 Hz per second. Each writing action generates a set of coordinate data, forming the writing trajectory path.

[0087] The pressure acquisition submodule is responsible for monitoring the pressure applied to the pen tip in different areas to form a multi-dimensional pressure dataset. Each writing action records the pressure value in real time, and monitors different levels of writing intensity such as light touch (about 200), moderate (about 500), and hard (about 900) to form a pressure waveform. The pressure data and its corresponding timestamp are simultaneously written to the data storage to form a time-pressure sequence.

[0088] The working principle and beneficial effects of the above technical solution are as follows: The device startup submodule of this embodiment turns on the dot matrix digital pen and starts the display device; at the same time, it initializes and establishes a connection with the writing surface (such as paper or digital tablet); the accelerometer and pressure sensor inside the dot matrix digital pen perform self-checks and calibrate the sensed initial state; the path acquisition submodule establishes a real-time two-dimensional coordinate system within the writing range of the dot matrix digital pen, with the axis position set as the starting writing point; whenever writing begins, it starts to capture every minute movement of the pen tip, recording the X and Y coordinates of the pen tip at a frequency of 1000 Hz per second, and each writing action generates a set of coordinate data, forming a writing trajectory path; the pressure acquisition submodule monitors the pressure applied to the pen tip in different areas, forming a multi-dimensional pressure dataset, and records the pressure value in real time for each writing action, monitoring different levels of writing intensity such as light touch (about 200), moderate (about 500), and forceful (about 900), forming a pressure waveform; the pressure data and its corresponding timestamp are simultaneously written into the data storage, forming a time-pressure sequence. The device startup submodule of the above solution activates the dot-matrix digital pen and display device, ensuring normal operation of the device; it establishes a stable connection with the writing surface (such as paper or digital tablet), ensuring efficient and accurate subsequent data acquisition; the accelerometer and pressure sensor inside the dot-matrix digital pen perform self-checks and calibrate the initial state to ensure that any data collected at the start of writing is an accurate baseline. The significance is that self-checks ensure the normal functioning of all devices, reducing data loss or errors caused by hardware failures or connection problems; calibration after device startup ensures the accuracy of subsequent data, providing a stable detection standard for the user's writing behavior, which is crucial for accurately identifying writing features. The path acquisition submodule establishes a two-dimensional coordinate system in real time within the writing area, setting the axis as the starting writing point, allowing writing actions to be accurately captured in a visualized space; it records the X and Y coordinates of the pen tip at a high frequency of 1000 Hz per second, ensuring the capture of every minute movement caused by the pen tip and generating a complete writing trajectory path. Significance Achieved: High-frequency data acquisition captures every detail of the writing process, including minute deviations, providing foundational data for subsequent character analysis. The established two-dimensional coordinate system allows for clear visualization of the writing trajectory, laying the foundation for character recognition, writing style analysis, and personalized feature extraction. The pressure acquisition submodule monitors applied pressure in real time across different areas of the pen tip, forming a comprehensive pressure dataset that captures variations in pressure during writing. It monitors different levels of writing intensity in real time, including light touch, moderate, and forceful strokes, recording pressure values ​​for each writing action to create a pressure waveform, and storing the pressure data along with a timestamp.Significance achieved: By monitoring changes in pressure, we can gain in-depth analysis of users' writing habits and styles, and react quickly to abnormal writing behaviors, such as excessive or insufficient pressure, thereby reducing writing fatigue; the pressure waveform can reflect the changes in the force applied by users during the writing process, which helps to analyze users' emotional and psychological states during writing, and provides data support for personalized teaching or writing style research.

[0089] In summary, the sub-modules of this embodiment collectively constitute the core functionality of the content information acquisition module. It not only accurately captures and records various information during the writing process but also lays a solid foundation for subsequent data processing, feature extraction, and user experience optimization. The dot-matrix digital pen provides more precise and comprehensive writing behavior analysis, offering in-depth insights into users' writing habits, styles, and functionality.

[0090] Example 5: Figure 5 As shown, based on Embodiment 3, the writing sequence generation module provided in this embodiment of the invention includes:

[0091] The sequence presentation submodule is responsible for converting all stroke, time, and pressure data into the outlines of individual characters. The characters are connected into a streamline on a virtual canvas to form a writing sequence, which is presented in the sequence according to the order in which the writing actions occur.

[0092] The speed calculation submodule is responsible for generating timestamps with the built-in clock at the moment the user starts and ends writing, and calculating the time required for each character by combining the writing trajectory of each character; by analyzing the displacement between the initial and end positions and combining the corresponding time difference, the real-time algorithm calculates the writing speed, and stores the writing speed of each character with a timestamp attached, forming a time series of writing speed changes;

[0093] The vector formation submodule is responsible for parsing the implemented stroke combinations and writing order. The start and end coordinate data of each stroke are integrated to form a complete character writing sequence and record the writing process. All captured data are integrated to form a multi-dimensional feature vector including writing trajectory, pressure change, writing speed and duration.

[0094] The working principle and beneficial effects of the above technical solution are as follows: The sequence presentation submodule of this embodiment converts all stroke, time, and pressure data into the outline of each character. The characters are connected into a streamline on the virtual canvas to form a writing sequence, which is presented in the sequence according to the order of writing actions. The speed calculation submodule generates a timestamp with a built-in clock at the moment the user starts and ends writing. Combined with the writing trajectory of each character, it calculates the time required for each character. By analyzing the displacement between the initial and end positions and combining the corresponding time difference, the real-time algorithm calculates the writing speed. The writing speed of each character is added with a timestamp and stored one by one to form a time sequence of writing speed changes. The vector formation submodule parses the implemented stroke combination and writing order. The start and end coordinate data of each stroke are integrated to form a complete character writing sequence and record the writing process. All captured data are integrated to form a multi-dimensional feature vector including writing trajectory, pressure change, writing speed, and time duration. The sequence presentation submodule of the above scheme converts all stroke, time, and pressure data into clear character outlines and connects them into streamlined shapes on a virtual canvas, forming a visual writing sequence. Ensuring characters are presented in the order of writing actions helps preserve the dynamic trajectory of actual writing. Significance: Transforming abstract writing data into visual character images allows users and analysts to intuitively understand the structure and form of writing; presenting in writing order captures the generation process of each character, providing foundational data for further analysis of users' writing habits and styles. The speed calculation submodule automatically generates timestamps at key moments at the beginning and end of writing, providing accurate time data for subsequent speed calculations; by analyzing the writing trajectory of characters, combined with displacement and time difference, it accurately calculates the writing speed of each character and stores this data along with the timestamps, forming a time series of writing speed changes. Significance: Real-time calculation of writing speed allows observation of speed changes during writing, a crucial indicator for analyzing users' writing habits and styles; clearly defining speed characteristics helps identify users' tension and relaxation states or other psychological reactions during writing, thus providing data support for education and psychological analysis. The vector formation submodule performs detailed analysis of the captured stroke combinations and writing order to form a complete writing sequence for each character and records the entire writing process. It integrates multi-dimensional data such as writing trajectory, pressure changes, writing speed, and duration to construct a comprehensive writing feature vector. The significance lies in the fact that multi-dimensional feature vectors can comprehensively reflect various changes in the user's writing process, providing a rich data foundation for subsequent research and analysis. The integrated data helps to establish writing characteristic models for specific users, supporting intelligent teaching, personalized feedback, and writing style analysis.

[0095] In summary, this embodiment forms a complete writing sequence generation process. The sequence presentation submodule visualizes the writing process, providing users with intuitive feedback. The speed calculation submodule monitors changes in character writing speed in real time, providing data support for dynamic features. The vector formation submodule integrates and analyzes all data to form a multi-dimensional feature vector, ultimately constructing a complete graph of the user's handwriting. This not only improves the accuracy of handwriting recognition and analysis but also deepens the understanding and research of writing behavior and its psychological state.

[0096] Example 6: As Figure 6 As shown, based on Embodiment 5, the vector forming submodule provided in this embodiment of the invention includes:

[0097] The character outline forming unit is responsible for identifying the start and end coordinates of strokes. Whenever the dot matrix digital pen touches or leaves the paper, it records a coordinate. The start coordinate is the point where the stroke begins, and the end coordinate is the point where the stroke ends. It draws the shape of each stroke. All strokes are arranged in chronological order according to the timestamp. Adjacent strokes are integrated, and individual strokes are connected into a character dataset, which contains the start and end coordinates of each stroke and the execution order during the writing process, forming a complete character outline.

[0098] The coordinate path calculation unit is responsible for matching the end coordinate of each stroke with the start coordinate of the next stroke. The duration of each stroke is associated with its corresponding spatial path, reconstructing a dynamic writing image on the time axis; forming an ordered series of coordinate paths.

[0099] The feature vector integration unit is responsible for analyzing the geometric features of each stroke, including the curvature, length, width, and closure of the lines, forming a set of geometric parameters; calculating the writing time of each stroke and the time interval between strokes; tracking and recording pressure changes based on the pressure applied by the pen tip during writing; and integrating all information into a comprehensive multidimensional feature vector, which includes the shape of the character, the writing order, and dynamic features.

[0100] The working principle and beneficial effects of the above technical solution are as follows: The character contour forming unit of this embodiment identifies the start and end coordinates of the strokes. Whenever the dot matrix digital pen touches or leaves the paper, it records a coordinate. The start coordinate is the point at which the stroke begins, and the end coordinate is the point at which the stroke ends. The shape of each stroke is drawn. All strokes are arranged in chronological order according to the timestamp. Adjacent strokes are integrated, and individual strokes are connected into a character dataset, which includes the start and end coordinates of each stroke and the execution order during the writing process, forming a complete character contour. The coordinate path calculation unit connects the end coordinate of each stroke with the next... The starting coordinates of each stroke are matched, and the duration of each stroke is associated with its corresponding spatial path to reconstruct a dynamic writing image on the timeline, forming an ordered series of coordinate paths. The feature vector integration unit analyzes the geometric features of each stroke, including the curvature, length, width, and closure of the stroke, forming a set of geometric parameters. It calculates the writing time of each stroke and the time interval between strokes. Based on the pressure applied by the pen tip during writing, it tracks and records changes in pressure. All information is integrated into a comprehensive multi-dimensional feature vector, containing the shape of the character, the writing order, and dynamic features. The character contour forming unit of the above scheme can accurately record the start and end coordinates of strokes when the user writes with a digital pen. The recording not only captures the starting and ending points of each stroke but also considers the trajectory of the stroke throughout the writing process. By introducing timestamps, the strokes can be sorted according to the writing order, reflecting the natural fluency of writing. Adjacent strokes are integrated to form a complete character contour dataset, ensuring the continuity and logic between each stroke. Significance Achieved: By generating complete character outlines, a clear visual expression is provided, enabling accurate capture and reproduction of the shape and structure of the writing. The generated character dataset lays the foundation for subsequent dynamic analysis and feature extraction, providing essential data support for downstream processing and algorithm applications. The coordinate path calculation unit seamlessly connects the end coordinates of each stroke with the start coordinates of the next stroke, ensuring the continuity of writing without interruption. By combining the duration of strokes with the corresponding spatial path, this unit reconstructs the dynamic process of writing, forming a dynamic writing image on the timeline. Ultimately, an ordered series of coordinate paths is formed, clearly reflecting every detail of the writing process. Significance Achieved: The ordered coordinate paths make subsequent analysis more comprehensive, including not only static shape information but also the exploration of dynamic features related to writing rhythm and fluency. It ensures the dynamic capture of the writing process, providing a reference for understanding the temporal characteristics of writing actions and helping to identify the writer's habits and style.The feature vector integration unit performs in-depth analysis of each stroke, including geometric features such as line curvature, length, width, and closure, with parameters describing the character's shape. It calculates the writing time for each stroke and the time interval between strokes, helping to capture changes in writing speed. It monitors and records changes in pen tip pressure, providing crucial information about the user's emotional state and writing posture for subsequent analysis. All captured information is integrated into a multi-dimensional feature vector, encompassing the character's shape, writing order, and dynamic features. Significance: By integrating data from different dimensions, this unit provides a comprehensive and integrated feature representation of character writing behavior, making subsequent pattern recognition and behavior analysis more accurate. It can identify the writer's unique style and characteristics, providing accurate data support and practical value for personalized education, psychological analysis, and other fields.

[0101] In this embodiment, the writing trajectory is modeled as a time-parameterized continuous function P during the writing process. i (t)=(x i (t),y i (t)), where

[0102] Stroke speed V i (t) is defined as:

[0103]

[0104] Stroke length L i Calculated using the following formula:

[0105]

[0106] The relationship between pressure and time during stroke writing, defined as the pressure function P(t) during writing, can be expressed as a parameterized function:

[0107] P(t) = k·t + b

[0108] Here, k and b are constants that combine the writer's personality and writing habits.

[0109] The continuity and start / end conditions between strokes, for continuous strokes P i and P i+1 It is necessary to ensure continuity at the start and end points, that is:

[0110]

[0111] This can be summarized into the following conditions:

[0112]

[0113] The coordinate path calculation unit analyzes the frequency components of writing for each stroke, assuming P... i (t) is the time series of the written trajectory, and its discrete Fourier transform can be expressed as:

[0114] F i (k)=∑t=0 N-1 P i (t)e -j2πkt / N k = 0, 1, ..., N-1

[0115] The frequency components and periodic feature information of each stroke are obtained;

[0116] The spacing and dynamic adjustment of the writing path, taking into account the coordinate differences between adjacent strokes, are represented by a distance function D:

[0117]

[0118] Here, considering the speed during the writing process, Δt is set as the time interval between adjacent strokes to correct the dynamic path:

[0119]

[0120] The feature vector integration unit constructs the following feature vectors:

[0121] V i =(L i CurvatureiΔP i T i Ti,i+1F i (k))

[0122] Curvature can be calculated using the second derivative:

[0123]

[0124] By combining multiple formulas to form a system of simultaneous equations, we can integrate them into a single system of simultaneous equations containing multiple formulas to solve the writing characteristics of characters and their dynamic changes:

[0125]

[0126] The above simultaneous equations are solved using optimization techniques such as the least squares method to obtain the optimal combination of features and their corresponding actual values.

[0127] In summary, the vector formation submodule of this embodiment completes the transformation from simple stroke data to complex feature vectors, advancing the depth of handwriting behavior analysis. Simultaneously, each unit ensures the accuracy, continuity, and analyzability of the handwriting data, providing a solid foundation for subsequent applications. This not only enhances handwriting recognition and understanding capabilities but also expands the application potential of handwriting analysis in fields such as education, psychology, and data science.

[0128] Example 7: Figure 7 As shown, based on Embodiment 3, the association management establishment module provided in this embodiment of the invention includes:

[0129] The feature identifier combination submodule is responsible for generating a unique user identifier for each user as a unique identifier; it records in real time through the sensor of the dot matrix digital pen, which includes the user's unique identifier related to handwriting, and combines the newly generated writing features with the user's unique identifier.

[0130] The vector storage submodule is responsible for storing the written feature vectors in the feature repository in the form of structured data. Each record contains: a unique user identifier, the written content and its identifier, a writing timestamp, and writing characteristic information; the unique user identifier links the written features to a specific user; the written content and its identifier include detailed information on letters, numbers, Chinese characters, and specific symbols; the writing timestamp records the start and end times of writing; and the writing characteristic information includes numerical values ​​such as speed and pressure.

[0131] The function definition submodule is responsible for assigning tags to each writing record for subsequent filtering and visualization or personalized recommendations; recording user operations and changes in writing habits, and tracking user writing patterns; and accessing one's own historical writing records through a unique identifier.

[0132] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the feature identifier combination submodule generates a unique user identifier for the user as a unique identifier; the sensor of the dot matrix digital pen records in real time, including the user's unique identifier related to handwriting, and combines the newly generated writing features with the user's unique identifier; the vector storage submodule stores the writing feature vector in the feature repository in the form of structured data, and each record includes: user's unique identifier, writing content and its identifier, writing timestamp, and writing characteristic information, etc.; the user's unique identifier links the writing features to a specific user; the writing content and its identifier include detailed information on letters, numbers, Chinese characters, and specific symbols; the writing timestamp records the start and end times of writing; the writing characteristic information, such as the values ​​of speed, pressure, etc.; the function definition submodule assigns tags to each writing record for subsequent filtering and visualization or personalized recommendations; records the user's operation and changes in writing habits, and tracks the user's writing patterns; and allows access to one's own historical writing records through the unique identifier. The feature identifier combined with the sub-module of the above scheme can generate a unique user ID for each user, ensuring that data can be clearly attributed to a specific user throughout the entire process of handwriting feature collection and storage. By recording and binding data from each writing action in real time (such as writing trajectory, speed, and pressure), effective connections between data are achieved. Significance: The unique identifier created for each user allows subsequent writing records, habit analysis, and personalized recommendations to accurately correspond to each user, providing tailored services; the introduction of unique user identifiers effectively prevents data confusion and misuse, protecting users' writing data and their privacy. The vector storage sub-module stores writing feature vectors in a clear and structured form, facilitating subsequent retrieval and processing; each record includes a unique user identifier, writing content identifier, timestamp, and writing characteristic information, ensuring data integrity; it can quickly retrieve relevant writing features based on the user identifier, supporting large-scale data processing and analysis. Significance: Effective feature storage allows users to quickly retrieve their own writing records, reflecting their writing habits and providing a basis for analysis; the accumulated large amount of writing feature data provides rich raw data for subsequent behavioral analysis and machine learning, helping to understand the patterns of individual writing behavior. The functional definition submodule assigns specific tags to each writing record, making data grouping and management more systematic and facilitating subsequent analysis and filtering. It also records and analyzes changes in user writing habits in real time, providing in-depth insights into writing patterns. Significance: By fully utilizing tags and historical data, the module can provide users with intelligent recommendations and customized feedback based on their writing habits, enhancing user engagement and satisfaction. Tracking and analyzing user writing habits helps the development team continuously improve system functionality and user experience, striving to provide better services.

[0133] In summary, this embodiment provides robust functional support for the entire feature storage and management system. It enables personalized storage, effective management, and in-depth analysis of user-written data, ultimately enhancing user experience and optimizing system performance. Furthermore, the systematic data management approach lays the foundation for future technological upgrades and expansions, such as intelligent analysis and prediction.

[0134] Example 8: As Figure 8 As shown, based on Embodiment 1, the similarity matching component provided in this embodiment of the invention includes:

[0135] The group lookup module is responsible for using a predefined hash function to input the current writing feature vector into the hash function and calculate a corresponding hash value. The hash value will be used to identify which user group the current writing feature belongs to. After obtaining the generated hash value, the module searches the writing feature database for all user groups that match the hash value.

[0136] The hash function expression is as follows:

[0137] H(F)=(P(F)⊕S(F)⊕T(F))mod M

[0138] Where: F is the input writing feature vector, and P(F) is the direct sum of the pressure features in the writing feature vector F, calculated as follows:

[0139]

[0140] Where p i It is the i-th pressure value in the writing feature vector F, and n is the dimension of the writing feature;

[0141] S(F) is the standard deviation of the writing speed feature in the writing feature vector F, representing the fluctuation of writing speed. The calculation formula is:

[0142]

[0143] Where v i It is the i-th writing speed value of the writing feature vector. It is the average writing speed;

[0144] T(F) is the weighted sum of the writing direction features of the writing feature vector. Taking into account the angle of the writing direction, the calculation formula is:

[0145]

[0146] Where d i w is the angle value of the i-th writing direction. i These are the corresponding weights, which can be adjusted based on the importance of the writing action.

[0147] ⊕ represents the XOR operation, used to combine the results of multiple parts into a unique value, thereby increasing complexity and chaos; M is a prime number used for modulo operation to ensure that the range of hash values ​​is within a certain integer range, reducing the possibility of hash collisions;

[0148] The similarity calculation module is responsible for comparing the similarity of all writing features within the user group corresponding to the hash value; normalizing the newly input writing features and the stored writing features; and calculating the similarity score between the current writing feature and the stored features.

[0149] The threshold comparison module is responsible for determining the similarity score between each stored writing feature within a user group and the current writing feature. If the similarity score found is greater than or equal to the preset threshold, the user's identity authentication will be considered successful, and the corresponding user login name will be returned, establishing the association between the writing trajectory and the user's identity. If the similarity scores of all stored features within the group and the current writing feature do not reach the preset threshold, the identity authentication will be marked as failed.

[0150] The working principle and beneficial effects of the above technical solution are as follows: The group search module in this embodiment uses a predefined hash function to input the current writing feature vector into the hash function and calculate a corresponding hash value. The hash value will be used to identify which user group the current writing feature belongs to. After obtaining the generated hash value, the module searches for all user groups that match the hash value in the writing feature library. The similarity calculation module performs similarity comparisons on all writing features within the user group corresponding to the hash value. The module normalizes the newly input writing features and the stored writing features and calculates the similarity score between the current writing feature and the stored features. The threshold comparison module judges the similarity score between each stored writing feature within the user group and the current writing feature. If the found similarity score is greater than or equal to the preset threshold, the user identity authentication is considered successful, and the corresponding user login name is returned, establishing the association between the writing trajectory and the user identity. If the similarity scores of all stored features within the group and the current writing feature do not reach the preset threshold, the authentication is marked as failed. The grouping search module of the above scheme uses a predefined hash function to process the current writing feature vector and calculate the corresponding hash value. The hash value is used to identify the specific group to which the user's writing feature belongs. In the writing feature database, all matching user groups are retrieved by hash value, significantly reducing the number of features that need to be compared. The significance is that hash grouping allows the system to quickly locate the set of potential users, avoiding a full search of the entire writing feature database, reducing the computational burden, and speeding up the identity authentication process. By narrowing the similarity matching range to a specific user group, the high computational complexity of "brute-force matching" is avoided, making the system more scalable and responsive. Before performing similarity comparison, the similarity calculation module normalizes the current writing feature and each stored writing feature within the group to ensure that the units and ranges of different features remain consistent. For all writing features within the user group corresponding to the hash value, the similarity score between each feature and the current writing feature is calculated. The significance is that standardized feature comparison ensures that the similarity calculation is not affected by a single feature, improving the accuracy of identity verification and reducing the false positive rate. It can flexibly respond to different users' writing habits and continuously optimize the similarity calculation process based on changes in user behavior. The threshold comparison module compares the similarity score of each stored writing feature within a user group with the current writing feature to determine if these scores meet a preset threshold. If at least one feature score meets the threshold, authentication is successful and the user's login name is returned. If all feature scores fail to meet the threshold, authentication is marked as failed. The significance is that by setting thresholds, legitimate and illegitimate users can be effectively identified while ensuring security, enhancing the system's security and credibility. The rapid feedback mechanism allows users to receive authentication results promptly, optimizing the user experience and reducing frustration caused by authentication failures.

[0151] In summary, this embodiment's similarity matching component forms an efficient and accurate processing flow during user authentication. This not only improves recognition efficiency, thus saving computational resources, but also enhances the reliability of identity verification. It can adapt to various users' writing habits, providing a safer and smoother user experience. Furthermore, it makes the system more maintainable and scalable, capable of accommodating potential future increases in user numbers and functional expansions.

[0152] Example 9: As Figure 9 As shown, based on Example 8, the similarity calculation module provided in this embodiment of the invention includes:

[0153] The feature combination submodule is responsible for calculating the minimum and maximum values ​​among all features for each written feature, using a linear transformation to map the feature values ​​to the range of 0 to 1, and combining the normalized features to form the corresponding feature vector for similarity calculation.

[0154] The vector composition submodule is responsible for combining the normalized values ​​of pressure, speed, and direction to form the current feature vector F{curr} and the stored feature vector F{stored}, and calculating the similarity.

[0155] F{curr}=[P{curr,norm},S{curr,norm},T{curr,norm}]

[0156] F{stored}=[P{stored,norm},S{stored,norm},T{stored,norm}]

[0157]

[0158] The scoring processing submodule is responsible for calculating the similarity score between the current written feature and the specific stored feature one by one, saving all the calculated similarity scores into an array, and judging it against the threshold.

[0159] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the feature combination submodule calculates the minimum and maximum values ​​among all features for each writing feature, and uses a linear transformation to map the feature values ​​to the range of 0 to 1. The normalized features are combined to form a corresponding feature vector for similarity calculation. The vector composition submodule combines the normalized values ​​of pressure, speed, and direction to form the current feature vector F{curr} and the stored feature vector F{stored}, and calculates the similarity. The scoring processing submodule calculates the similarity score between the current writing feature and a specific stored feature, saves all calculated similarity scores into an array, and compares them with a threshold. The feature combination submodule of the above solution eliminates the influence of different feature scales by calculating the minimum and maximum values ​​of each writing feature among all features and using a linear transformation to map the feature values ​​to the range of 0 to 1, ensuring that all features are compared under the same standard. The normalized features are combined into a feature vector, providing a unified output for subsequent similarity calculations. Significance Achieved: It ensures that features of different natures (such as pressure, speed, and direction) are compared within the same category, reducing errors and biases caused by inconsistent feature scales, thereby improving the accuracy of the final similarity assessment; by combining normalized features into feature vectors, it provides standardized input for subsequent modules, simplifying the complexity of the similarity calculation process. The vector composition submodule combines the normalized pressure, speed, and direction values ​​to form the current feature vector (F{curr}) and the stored feature vector (F{stored}); it calculates the similarity score between the two vectors using the cosine similarity formula. Significance Achieved: By combining multiple features into a single vector, it can comprehensively reflect the user's handwriting characteristics, enabling similarity calculations to more accurately reflect the relationships between features related to user identity; by defining a clear vector structure and similarity calculation formula, it improves the efficiency and simplicity of the calculation, allowing the system to quickly process large amounts of data for identity recognition. The scoring processing submodule calculates the similarity score between the current handwriting feature and each stored feature one by one, ensuring that all possible comparisons are considered; it saves all calculated similarity scores into an array for subsequent threshold judgment. Significance achieved: By calculating the similarity of different features one by one, the system can accurately assess the matching degree between the current user and each storage user, providing a basis for identity recognition; the storage similarity score provides data support for the subsequent threshold comparison module, making the final identity verification decision more solid and reliable.

[0160] In summary, the similarity calculation module in this embodiment constitutes an efficient and systematic similarity verification process. The various sub-modules complement each other in terms of technical effect and significance, not only improving the system's processing capabilities for handwriting features but also enhancing the accuracy and security of user identification. This ensures that the collected feature vectors can be effectively applied to fast and accurate identity verification, ultimately achieving an optimized user experience and improved system operational efficiency.

[0161] Example 10: As Figure 10 As shown, based on Examples 1-9, the user identification method based on handwriting recognition provided in this embodiment of the invention includes the following steps:

[0162] S100: Identify the dot matrix digital pen to obtain the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory.

[0163] S200: Obtain the writing trajectory input by the second dot matrix digital pen, extract writing features from the writing trajectory to obtain the current writing features, and store them in the writing feature library of the server; the current writing features are used to compare with the writing trajectory newly input by the user to obtain the user's login name and establish the association between the login name and the writing trajectory;

[0164] S300: Obtain the new writing trajectory input by the second dot matrix digital pen, send the new writing trajectory to the server, and perform similarity matching with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is successful; otherwise, it is unsuccessful.

[0165] The working principle and beneficial effects of the above technical solution are as follows: This embodiment first identifies the dot matrix digital pen to obtain the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory; secondly, the writing trajectory input by the second dot matrix digital pen is obtained, and writing features are extracted from the writing trajectory to obtain the current writing features, which are stored in the writing feature library of the server; the current writing features are compared with the newly input writing trajectory of the user to obtain the user's login name and establish the association between the login name and the writing trajectory; finally, the new writing trajectory input by the second dot matrix digital pen is obtained, and the new writing trajectory is sent to the server for similarity matching with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is passed; otherwise, it is not passed. Step S100 of the above scheme involves the identification and use of the dot-matrix digital pen. By identifying the dot-matrix digital pen, the system can distinguish between the administrator's pen and the user's pen, ensuring the security and accuracy of subsequent operations. The administrator uses the first dot-matrix digital pen to perform terminal device permission operations, while users use the second dot-matrix digital pen to input their writing trajectory, ensuring the separation of functions for different roles. Significance: By distinguishing identities, unauthorized users are prevented from performing authorized operations, enhancing system security and reducing the risk of tampering or misuse; ensuring that the roles of each person are clear during operation, maintaining the effective separation of system management and user operation, and helping to standardize management processes. Step S200 involves the extraction and storage of writing features. Features are extracted from the writing trajectory input by the user, which can analyze the user's writing habits from multiple dimensions (such as pressure, speed, writing direction, etc.) and store relevant features; the current writing features are stored in the writing feature library on the server, providing database support for subsequent recognition. Significance: Extracting and storing handwriting features lays the foundation for subsequent identity verification or matching, allowing the system to use these features for analysis and judgment during user authentication; establishing a handwriting feature library enables the system to have stronger individual recognition capabilities, gradually adapting to users' handwriting habits and styles, thereby improving the accuracy and effectiveness of recognition. Step S300 User Identity Authentication and Similarity Matching: Obtain the new handwriting trajectory input by the user with the second-dot-matrix digital pen and send it to the server; perform similarity matching with features in the handwriting feature library; calculate the similarity score, if the score is less than a predetermined threshold, authentication is successful, otherwise it fails. Significance: By comparing with the real-time input handwriting trajectory, the system allows for real-time identity verification during the user's writing process, improving security responsiveness; through efficient similarity matching, user identity can be quickly determined, improving user experience, reducing unnecessary verification delays, thereby enhancing user trust in the system.

[0166] In summary, the steps in this embodiment collectively support a handwriting recognition-based user identification method, achieving a secure, flexible, and efficient authentication mechanism. Each step forms a complete workflow, complementing each other on multiple levels, including security, user experience, personalized feature recognition, and system effectiveness, ensuring the final system meets high-level identity verification requirements. Simultaneously, this approach reduces the risks of identity theft and data leakage, enhancing system reliability and user confidence.

[0167] Example 11: As Figure 11 As shown, based on embodiments 1-10, the user identification device based on handwriting recognition provided in this embodiment of the invention includes: a plurality of dot matrix digital pens, a terminal device, a server, and a writing feature library.

[0168] The dot matrix digital pen connects to the terminal device wirelessly, and the terminal device connects to the server via a network. The writing feature library is installed in the server; the dot matrix digital pen can also connect directly to the server wirelessly.

[0169] The working principle and beneficial effects of the above technical solution are as follows: This embodiment identifies two types of dot-matrix digital pens: the first pen is used by the administrator to manage terminal device permissions; the second pen is used by the user to input the user's writing trajectory. Both pens establish a wireless connection with the terminal device, ensuring clear functions and permissions for different roles. Writing feature extraction and storage: The user inputs a writing trajectory using the second pen. The system acquires and analyzes the features of these trajectories (such as pressure, speed, and writing direction) and stores the extracted features in a writing feature library on the server. During this step, a correlation is established between the current writing features and the user's login name for future identity verification. Identity authentication and similarity matching: The user inputs a new writing trajectory again using the second pen. The system acquires this trajectory and sends it to the server. The server performs a similarity match between the new writing features and the features stored in the writing feature library. If the similarity score is below a set threshold, identity authentication is successful; otherwise, authentication fails.

[0170] The technical effects and significance of each structure in this embodiment are as follows: The connection structure between the dot-matrix digital pen and the terminal device / server allows the pen to wirelessly transmit the input writing trajectory to the terminal device in real time, avoiding data transmission delays and inconvenience. Significance: Real-time interaction and connection enhance the user experience, allowing users to write and authenticate at any time without worrying about limitations of the connection method. The writing feature extraction and storage structure accurately reflects the user's writing habits, and the features are stored in the server's feature library for easy access and management. Significance: Building a feature library based on personalized writing features enables the system to accurately identify individuals, thereby improving the accuracy and reliability of identity authentication and providing crucial support for user identity protection. The similarity matching structure matches new writing trajectories with existing features in the feature library, providing real-time feedback for quick and effective user identity determination. Significance: This provides automated and efficient means of identity authentication, allowing users to quickly complete security verification when using the device, reducing authentication time and cumbersome steps; furthermore, the threshold setting mechanism flexibly adapts to changes in different users' writing styles, increasing the robustness of recognition.

[0171] In summary, this embodiment, by combining the aforementioned structures, enables a handwriting recognition-based user identification device that not only achieves efficient and accurate identity verification but also significantly enhances user experience and system security. The entire solution, through wireless connectivity, feature extraction, and real-time similarity matching, ensures convenience and flexibility in the identification process. Furthermore, as users' writing habits change and the system continuously accumulates data, the device can adapt to the personalized needs of different users, laying a solid foundation for future development and applications.

[0172] In this embodiment, the authentication interface does not display traditional input methods like username and password. Instead, a handwriting box is used where the user directly writes their name. This name does not need to be the user's real name; it can be a login nickname or a numeric account, as long as it meets two conditions: it must be able to identify the user in the user management system and not conflict with other users; and it cannot be too simple to ensure that the written result has sufficient information for identification. After the user finishes writing, the program on the terminal device, such as a PAD or touch screen, sends the user's writing trajectory to the server. The server performs recognition and authentication and sends the authentication result to the terminal device. If the authentication is successful, the user enters the system; if the authentication fails, the user is prompted. Since there is no screen, the following methods can be used: 1. Pre-print a piece of paper with special information and an input box printed on it, specifically for the user to write their login username; 2. Print a fixed-format box in a fixed position on all papers for the user to write their login username; 3. No fixed position. When the dot matrix digital pen is detected to have been inactive for an extended period (the data it collects can include time information, and the dot matrix digital pen can determine how long the trajectory data has been uncollected), any login username written by the user in any location will be used as login authentication input.

[0173] After the user finishes writing, the dot-matrix digital pen sends the user's writing trajectory to the server. The server recognizes and authenticates the writing and sends the authentication result to the terminal device. Due to functional limitations, the dot-matrix digital pen does not connect directly to the server, but instead uses a proxy device. However, the existence of the proxy device does not affect the overall process. The proxy device merely relays information. Because of the dot-matrix digital pen's limited functionality, the authentication result can only be output through the pen's accompanying prompt component. For example, if authentication is successful, the dot-matrix digital pen's indicator light displays blue; or it flashes blue several times and then turns off to conserve energy. If authentication fails, the dot-matrix digital pen's indicator light displays red, and generally does not turn off, to prompt the user to rewrite their login name; the user's writing trajectory needs to be collected beforehand. Based on a similar approach, data collection can be very convenient. Users simply need to write their login username on any dot-matrix paper. During actual data collection, the location can be limited to a specific spot, such as the administrator's seat, to observe the system's results. A specific dot-matrix digital pen can also be used; the administrator can select this pen by its number in the management interface as the dedicated dot-matrix digital pen for trajectory collection. The advantage of this method is that once the system is designed, during multi-user data collection, the administrator no longer needs to navigate the complex management interface. The administrator only needs to verify that the login name written by the user is indeed that user's login name.

[0174] The data collection process for each user is as follows: The user writes their login name on paper using a designated dot-matrix digital pen, and the system collects the trajectory; the trajectory is identified to confirm the user to be collected; writing features are extracted from the trajectory to confirm their validity and ensure their quality; the association between the user and the features is saved; user switching is seamless because the system's current time is automatically obtained each time data is collected. Based on the writing time interval, it can be determined that a user has changed.

[0175] If identifying the login username is omitted, then there is no need to write the login username. Trajectory features can be extracted from any of the user's input; the advantage is convenience, as it eliminates the need to explicitly write the login username.

[0176] Before using this embodiment, all the dot matrix digital pens for students in a class were charged together simultaneously. Since the dot matrix digital pen and user were fixedly bound one-to-one, no pen was available while charging. Providing two pens for each student significantly increased costs. With this embodiment, a small number of extra dot matrix digital pens can be prepared, solving the charging problem. Detailed explanation follows: Assume a class has N students; assume each dot matrix digital pen has a writing time of W after being fully charged; assume the time required to fully charge each pen is C (generally, C is less than W); assume P pens are needed (P is greater than N); assume the duration of the test is T. Then the following formulas can be derived: (MN)·T·(W / C)=N·T, (MN)·(W / C)=N, MN=N·(C / W), where MN represents the extra dot matrix digital pens used. If a 2-hour charge provides 8 hours of battery life, then an additional 2 / 8 (1 / 4) of the number of students' dot matrix digital pens would be needed. Considering that all pens can be charged after school, the actual number of additional dot matrix digital pens required can be reduced. When a student's dot matrix digital pen runs out of power, they can simply exchange it for a new one at the charging station. After the exchange, the user's identity verification and binding process is completed as described above. Simultaneously, the backend resolves the binding relationship between the user and the previous dot matrix digital pen.

[0177] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of equivalents of this invention, this invention is also intended to include these modifications and variations.

Claims

1. A user identification system based on handwriting recognition, characterized in that, Include: The device identification component is responsible for identifying the dot matrix digital pen, obtaining the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory. The feature storage component is responsible for acquiring the writing trajectory of the second dot matrix digital pen input, extracting writing features from the writing trajectory to obtain the current writing features, and storing them in the writing feature library on the server. The current writing features are used to compare with the writing trajectory newly input by the user to obtain the user's login name and establish the association between the login name and the writing trajectory. The similarity matching component is responsible for acquiring the new writing trajectory input by the second-dot matrix digital pen, sending the new writing trajectory to the server, and performing similarity matching with the writing features existing in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is successful; otherwise, it fails. A hash value is calculated using a predefined hash function, and the hash value is used to search for all user groups that match the hash value in the writing feature library. For all writing features within the user group corresponding to the hash value, a similarity comparison is performed one by one. Hash function The expression is: ； in: It is the input written feature vector. It is writing feature vectors The direct sum of the pressure characteristics in the middle, It is writing feature vectors The standard deviation of the writing speed characteristic in It is a prime number used in modular arithmetic; It is writing feature vectors The weighted sum of the writing direction features.

2. The user identification system based on handwriting recognition as described in claim 1, characterized in that, Device identification components, including: The first marking operation module is responsible for obtaining the number of dot matrix digital pens to be marked, and at the same time obtaining the running status of the dot matrix digital pens within a preset time period, and sending the running status to the server; randomly selecting one as the first dot matrix digital pen, retrieving the trigger operation corresponding to the administrator's operation permission from the server, and marking the first dot matrix digital pen with the administrator operation mark. The second marking operation module is responsible for selecting at least one from other dot matrix digital pens as the second dot matrix digital pen, retrieving the trigger operation corresponding to the writing trajectory from the server, and marking the second dot matrix digital pen with the user operation mark. The identifier association operation module is responsible for responding to the first dot matrix digital pen identifier administrator operation mark and the second dot matrix digital pen identifier user operation mark. Upon receiving the identifier parsing request sent by the server, it performs identifier parsing processing on the identifier in the identifier parsing request to obtain the relationship between the dot matrix digital pen and the trigger operation corresponding to the identifier.

3. The user identification system based on handwriting recognition as described in claim 1, characterized in that, Feature storage component, comprising: The content information acquisition module is responsible for recording the user's writing content in real time through the built-in sensor of the dot matrix digital pen; The writing sequence generation module is responsible for converting the written content into characters and their order, generating a writing sequence, recording the start and end times of writing each character in the writing sequence, and calculating the writing duration of each character in the written content by combining the timestamp; The writing speed of each character is calculated by measuring the start and end displacements of each character, and the pressure changes applied to each character are identified by combining the data from the pressure sensor. The association management module is responsible for storing writing feature vectors in a structured manner in the feature repository, forming writing feature records corresponding to users, assigning a unique identifier to each writing feature, and establishing an association relationship with the user's login name.

4. The user identification system based on handwriting recognition as described in claim 3, characterized in that, The content information acquisition module includes: The device startup submodule is responsible for turning on the dot matrix digital pen and starting the display device; at the same time, it initializes and establishes a connection with the writing surface; the accelerometer and pressure sensor inside the dot matrix digital pen perform self-checks and calibrate the sensed initial state. The path acquisition submodule is responsible for establishing a real-time two-dimensional coordinate system within the writing range of the dot matrix digital pen, with the axis position set as the starting writing point. Whenever writing begins, it starts to capture every minute movement of the pen tip, records the coordinates of the pen tip, and generates a set of coordinate data for each writing action, forming a writing trajectory path. The pressure acquisition submodule is responsible for monitoring the pressure applied to the pen tip in different areas to form a multi-dimensional pressure dataset. The pressure value is recorded in real time for each writing action, and different levels of writing intensity are monitored in real time to form a pressure waveform. The pressure data and its corresponding timestamps are simultaneously written to the data storage to form a time-pressure sequence.

5. The user identification system based on handwriting recognition as described in claim 3, characterized in that, The sequence generation module contains a vector forming submodule, which includes: The character outline forming unit is responsible for identifying the start and end coordinates of strokes. Whenever the dot matrix digital pen touches or leaves the paper, it records a coordinate. The start coordinate is the point where the stroke begins, and the end coordinate is the point where the stroke ends. It draws the shape of each stroke. All strokes are arranged in chronological order according to the timestamp. Adjacent strokes are integrated, and individual strokes are connected into a character dataset. The coordinate path calculation unit is responsible for matching the end coordinate of each stroke with the start coordinate of the next stroke. The duration of each stroke is associated with its corresponding spatial path, reconstructing a dynamic writing image on the time axis; forming an ordered series of coordinate paths. The feature vector integration unit is responsible for analyzing the geometric features of each stroke, including the curvature, length, width, and closure of the lines, forming a set of geometric parameters; calculating the writing time of each stroke and the time interval between strokes; tracking and recording pressure changes based on the pressure applied by the pen tip during writing; and integrating all information into a comprehensive multidimensional feature vector.

6. The user identification system based on handwriting recognition as described in claim 5, characterized in that, The character dataset of the character contour forming unit contains the start and end coordinates of each stroke and the execution order during the writing process, forming a complete character contour.

7. The user identification system based on handwriting recognition as described in claim 3, characterized in that, The association management module includes: The feature identifier combination submodule is responsible for generating a unique user identifier for each user as a unique identifier; it records in real time through the sensor of the dot matrix digital pen, which includes the user's unique identifier related to handwriting, and combines the newly generated writing features with the user's unique identifier. The vector storage submodule is responsible for writing feature vectors and storing them in the feature repository in the form of structured data; A user's unique identifier links writing characteristics to a specific user; The writing timestamp records the start and end times of the writing process; The function definition submodule is responsible for assigning tags to each writing record for subsequent filtering and visualization or personalized recommendations; recording user operations and changes in writing habits, and tracking user writing patterns; and accessing one's own historical writing records through a unique identifier.

8. The user identification system based on handwriting recognition as described in claim 1, characterized in that, Similarity matching component, including: The group lookup module is responsible for using a predefined hash function. The current writing feature vector is input into the hash function to calculate a corresponding hash value. The hash value will be used to identify which user group the current writing feature belongs to. The similarity calculation module is responsible for normalizing the newly input writing features and the stored writing features, and calculating the similarity score between the current writing features and the stored features. The threshold comparison module is responsible for determining the similarity score between each stored writing feature within a user group and the current writing feature. If the similarity score found is greater than or equal to the preset threshold, the user's identity authentication will be considered successful, and the corresponding user login name will be returned, establishing the association between the writing trajectory and the user's identity; if the similarity scores of all stored features in the group and the current writing feature do not reach the preset threshold, the identity authentication will be marked as failed.

9. A user identification method based on handwriting recognition, implemented using the user identification system based on handwriting recognition according to any one of claims 1-8, characterized in that, Includes the following steps: The dot matrix digital pen is identified to obtain the administrator's first dot matrix digital pen and the user's second dot matrix digital pen; the first dot matrix digital pen is used to enable the administrator to perform permission operations on the terminal device, and the second dot matrix digital pen is used to input the user's writing trajectory. The writing trajectory input by the second dot matrix digital pen is obtained, the writing features are extracted from the writing trajectory to obtain the current writing features, and stored in the writing feature library of the server; the current writing features are compared with the writing trajectory newly input by the user to obtain the user's login name and establish the association between the login name and the writing trajectory; The new writing trajectory input by the second dot matrix digital pen is obtained and sent to the server. The new writing trajectory is then matched with the writing features in the writing feature library. If the similarity is less than a predetermined threshold, the user identity authentication of the new writing trajectory is successful; otherwise, it is unsuccessful.

10. A user identification device based on handwriting recognition, implemented using the user identification system based on handwriting recognition according to any one of claims 1-8, characterized in that, Includes: several dot-matrix digital pens, terminal devices, servers, and a writing feature library; The dot matrix digital pen connects to the terminal device wirelessly, and the terminal device connects to the server via a network. The writing feature library is installed in the server. The dot matrix digital pen connects directly to the server wirelessly.

Images