Intelligent sensing writing brush and calligraphy learning method and system thereof

Abstract

The invention provides an intelligent sensing writing brush and a calligraphy learning method and system thereof, and relates to the technical field of intelligent calligraphy learning. A first data collection module is arranged at the root of the pen point and is used for collecting pressure data of contact between the pen point and a paper surface as first writing data; a data processing module, a second data collection module, a power supply module and a data transmission module are arranged in the pen holder; the second data collection module is used for collecting second writing data; the second writing data comprises grip data and acceleration data, speed data, angular velocity data and font data of the writing brush main body; and the data processing module is used for obtaining a comprehensive writing style recognition result and a comprehensive writing quality evaluation result according to the first writing data, the second writing data and a multi-modal writing data analysis method. According to the intelligent sensing writing brush provided by the invention, accurate writing feedback and personalized guidance are provided through multi-modal data acquisition and analysis.

Description

Technical Field

[0001] The present application relates to the field of intelligent calligraphy learning technology, in particular to an intelligent sensing brush and a calligraphy learning method and system thereof. Background Art

[0002] Smart brush calligraphy learning tools combine traditional calligraphy with modern information technology. Built-in sensors collect data such as writing pressure, angle, and speed, and leverage data analysis to provide instant feedback and personalized guidance. Existing smart brush calligraphy learning tools, such as digital pressure-sensitive brushes, sensor brushes based on image recognition technology, and brushes based on software platforms, are becoming increasingly popular, offering a certain degree of convenience and instant feedback. However, these existing tools each have limitations: digital pressure-sensitive brushes, constrained by electronic platforms, struggle to simulate the real writing experience; sensor brushes based on image recognition focus on displaying results, lacking guidance on key elements of the writing process; and brushes based on software platforms perform poorly in simulating hand feel and providing personalized adjustments. Summary of the Invention

[0003] The present application provides an intelligent sensing brush and a calligraphy learning method and system thereof to solve one or more technical problems existing in the prior art, and at least provide beneficial options or create conditions.

[0004] In one aspect, the present application provides a smart sensing brush, comprising a brush body; The brush body comprises a pen tip and a pen holder; The base of the pen tip is provided with a first data collection module, and the first data collection module is used to collect pressure data of the contact between the pen tip and the paper as first writing data; The pen holder is provided with a data processing module, a second data collection module, a power supply module and a data transmission module; the second data collection module is used to collect second writing data of the brush body; the second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body; The first writing data and the second writing data are used as initial writing data of the brush body; The data processing module is used to obtain a comprehensive writing style recognition result and a comprehensive writing quality evaluation result based on the initial writing data and the multimodal writing data analysis method.

[0005] Further, the first data collection module includes a first pressure sensor; the second data collection module includes a second pressure sensor, an accelerometer, a gyroscope and a scanner; The second pressure sensor is used to collect the grip force data; The accelerometer is used to collect the velocity data and the acceleration data; The gyroscope is used to collect the angular velocity data; The scanner is used to collect the font data.

[0006] Furthermore, the pen tip is made of animal hair or artificial synthetic material to ensure the water absorption and ink distribution characteristics of the pen tip.

[0007] On the other hand, the present application provides a calligraphy learning method using an intelligent sensing brush, comprising the following steps: collecting first writing data and second writing data of a brush body as initial writing data of the brush body; The first writing data includes pressure data of the tip of the brush body in contact with the paper surface; The second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body; Based on the initial writing data and multimodal writing data analysis method, a comprehensive writing style recognition result and a comprehensive writing quality evaluation result are obtained.

[0008] Furthermore, the comprehensive writing style recognition result and comprehensive writing quality evaluation result obtained based on the initial writing data and the multimodal writing data analysis method include: Optimizing the initial writing data to obtain third writing data; performing feature extraction on the third writing data to obtain a writing feature set; According to the writing feature set and the multimodal writing data analysis method, the comprehensive writing style recognition result and the comprehensive writing quality evaluation result are obtained.

[0009] Furthermore, the optimizing the initial writing data to obtain the third writing data includes: filtering the pressure data to obtain filtered pressure data; filtering the grip force data, the angular velocity data, the velocity data, and the acceleration data to obtain filtered grip force data, filtered angular velocity data, filtered velocity data, and filtered acceleration data; Performing fusion filtering on the acceleration data and the angular velocity data to obtain filtered attitude angle data; performing cleaning and segmentation processing on the font data to obtain font structure information of each font in the font data as font structure data; The filtered grip force data, the filtered pressure data, the filtered posture angle data, the filtered angular velocity data, the filtered velocity data, the filtered acceleration data and the font structure data serve as third writing data.

[0010] Furthermore, the extracting features of the third writing data to obtain a writing feature set includes: Performing feature extraction on the filtered pressure data and the filtered grip force data in the third writing data to obtain pressure features; the pressure features include grip force features, pressure change frequency features, and pressure change amplitude features; Performing feature extraction on the filtered posture angle data and the filtered angular velocity data in the third writing data to obtain posture features; the posture features include posture angle features and angular velocity features; Performing feature extraction on the filtered velocity data and filtered acceleration data in the third writing data to obtain motion features; the motion features include velocity features and acceleration features; Extracting features from the font structure data in the third writing data to obtain font structure features; the font structure features include center of gravity position features and center position features; The pressure feature, the posture feature, the movement feature and the font structure feature serve as a writing feature set.

[0011] Furthermore, the comprehensive writing style recognition result and the comprehensive writing quality evaluation result are obtained based on the writing feature set and the multimodal writing data analysis method, including: Obtaining a first writing style recognition result and a first writing quality evaluation result based on the pressure feature in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining a second writing style recognition result and a second writing quality evaluation result based on the posture features in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining a third writing style recognition result and a third writing quality evaluation result based on the motion features in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining a fourth writing style recognition result and a fourth writing quality evaluation result based on the font structure features in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining the comprehensive writing style recognition result according to the first writing style recognition result, the second writing style recognition result, the third writing style recognition result, and the fourth writing style recognition result; The comprehensive writing quality evaluation result is obtained according to the first writing quality evaluation result, the second writing quality evaluation result, the third writing quality evaluation result and the fourth writing quality evaluation result.

[0012] On the other hand, the present application provides a calligraphy learning system for an intelligent sensing brush, comprising a first data acquisition module, a second data acquisition module, and a data analysis module; The first data acquisition module is used to collect first writing data of the brush body; the first writing data includes pressure data of the contact between the tip of the brush body and the paper surface; The second data acquisition module is used to collect second writing data of the brush body; the second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body; The data analysis module is used to use the first writing data and the second writing data as the initial writing data of the brush body; based on the initial writing data and the multimodal writing data analysis method, a comprehensive writing style recognition result and a comprehensive writing quality evaluation result are obtained.

[0013] Furthermore, the data analysis module includes a data optimization unit, a feature extraction unit and a feature analysis unit; The data optimization unit is used to optimize the initial writing data to obtain third writing data; The feature extraction unit is used to extract features from the third writing data to obtain a writing feature set; The feature analysis unit is used to obtain the comprehensive writing style recognition result and the comprehensive writing quality evaluation result based on the writing feature set and the multimodal writing data analysis method.

[0014] The beneficial effects of the present application are as follows: the present application provides an intelligent sensing brush, comprising a brush body; the brush body comprises a nib and a pen holder; a first data collection module is provided at the base of the nib, the first data collection module being used to collect pressure data of the contact between the nib and the paper as first writing data; a data processing module, a second data collection module, a power supply module and a data transmission module are provided inside the pen holder; the second data collection module is used to collect second writing data of the brush body; the second writing data comprises grip data and acceleration data, velocity data, angular velocity data and font data of the brush body; the first writing data and the second writing data serve as the initial writing data of the brush body; the data processing module is used to obtain a comprehensive writing style recognition result and a comprehensive writing quality evaluation result based on the initial writing data and the multimodal writing data analysis method. The intelligent sensing brush provided by the present application provides accurate writing feedback and personalized guidance through multimodal data collection and analysis. The present application also provides a calligraphy learning method and system corresponding to the intelligent sensing brush, which will not be repeated here.

[0015] Other features and advantages of the present application will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present application. The purposes and other advantages of the present application can be achieved and obtained through the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation to the technical solution of the present invention.

[0017] Figure 1 This is a structural diagram of the intelligent sensor brush provided by this application; Figure 2 This is a flow chart of the calligraphy learning method using the smart sensing brush provided by this application; Figure 3 is a schematic diagram of the principle of determining the comprehensive writing style recognition result and the comprehensive writing quality evaluation result provided by this application; Figure 4 This is a structural diagram of the calligraphy learning system of the intelligent sensing brush provided by this application; Figure 5 This is a schematic diagram of the pen tip pressure change curve for different practice times and standards when writing official script provided by this application; Figure 6 This is a schematic diagram of the pen tip pressure change curve for different practice times and standards when writing seal script provided by this application; Figure 7 This is a schematic diagram of the pen tip pressure change curve for different practice times and standards when writing regular script provided by this application; Figure 8 This is a schematic diagram of the pen tip pressure change curve for different practice times and standards when writing running script fonts provided by this application; Figure 9 This is a schematic diagram of the pen tip pressure change curve for different practice times and standards when writing cursive fonts provided by this application; Figure 10 This is a schematic diagram of the curve of the posture angle changing over time when writing the official script font provided by this application; Figure 11 This is a schematic diagram of the curve of the posture angle changing over time when writing seal script provided by this application; Figure 12 This is a schematic diagram of a curve showing the change of posture angle over time when writing regular script provided by the present application; Figure 13 This is a schematic diagram of the curve of the posture angle changing over time when writing running script fonts provided by this application; Figure 14This is a schematic diagram of the curve of the posture angle changing over time when writing cursive fonts provided by this application; Figure 15 This is a schematic diagram of the relationship between pen tip pressure and grip force when writing five fonts provided by this application; Figure 16 This is a schematic diagram of the relationship between pen tip pressure and pen speed when writing regular script provided by this application; Figure 17 This is a schematic diagram of a curve showing the change in pen speed over time when writing regular script provided in this application. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0019] The present application is further described below in conjunction with the accompanying drawings and specific embodiments. The described embodiments should not be considered as limiting the present application. All other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0020] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0022] Traditional calligraphy learning relies on pen and paper and teacher guidance, emphasizing the refinement of fundamental skills. However, this practice suffers from long feedback cycles, uneven teaching resources, and a lack of dynamic demonstrations. With the advancement of technology, digital learning methods are becoming increasingly popular. While existing intelligent brush calligraphy learning tools, such as digital pressure-sensitive brushes, sensor brushes based on image recognition technology, and software-based brushes, offer a certain degree of convenience and immediate feedback, they still fall short in simulating a realistic writing experience, providing comprehensive guidance throughout the writing process, and enabling personalized adjustments. These tools fail to fully meet the diverse needs of calligraphy learners.

[0023] In response to the problems and defects in related technologies, the embodiments of the present application provide an intelligent sensing brush and its calligraphy learning method and system. By setting a first data collection module at the base of the pen tip and configuring a second data collection module inside the pen shaft, the embodiments of the present application can comprehensively capture multimodal data such as pressure, grip, acceleration, speed, angular velocity and font during writing, providing detailed basic information for subsequent analysis. The built-in data processing module uses a multimodal writing data analysis method to optimize and process the collected initial writing data, extract features, and ultimately derive comprehensive writing style recognition results and comprehensive writing quality evaluation results, providing users with instant and accurate feedback to help them adjust their writing habits in a timely manner and improve their learning efficiency. By analyzing the writing feature sets of different writers, targeted learning suggestions are provided based on the actual situation of each learner, promoting the effective improvement of calligraphy skills and the cultivation of personal style. Despite the introduction of advanced sensing technology and data analysis algorithms, the intelligent sensing brush provided by the embodiments of the present application still retains the material and design of the traditional brush, ensuring that users can enjoy digital assistance without losing the real paper and pen touch and writing environment.

[0024] The following will describe in detail an intelligent sensing brush provided by an embodiment of the present application with reference to the accompanying drawings.

[0025] Reference Figure 1 , Figure 1 This is a structural diagram of an intelligent sensor brush provided in this application. The specific structure and function of the intelligent sensor brush provided in the embodiment of this application are described as follows.

[0026] The smart sensing writing brush comprises a writing brush body 100. The writing brush body 100 comprises a pen tip 101 and a pen holder 200.

[0027] A first data collection module 102 is provided at the base of the pen tip 101 . The first data collection module 102 is used to collect pressure data of the contact between the pen tip 101 and the paper surface as first writing data.

[0028] The penholder 200 is internally equipped with a data processing module 201, a second data collection module 202, a power supply module 203, and a data transmission module 204. The second data collection module 202 is used to collect second writing data of the brush body 100. The second writing data includes grip force data, acceleration data, velocity data, angular velocity data, and font data. The first and second writing data serve as the initial writing data of the brush body 100.

[0029] The data processing module 201 is used to obtain comprehensive writing style recognition results and comprehensive writing quality evaluation results based on the initial writing data and the multimodal writing data analysis method.

[0030] The intelligent sensing brush realizes comprehensive monitoring and analysis of the writing process through the collaborative work of multiple modules within it. The first data collection module 102 of the pen tip 101 and the second data collection module 202 of the pen body 200 jointly collect a wealth of writing data, including pressure, grip, acceleration, speed, angular velocity and font data. The data processing module 201 conducts a comprehensive analysis of these data to provide writing style recognition and quality evaluation results. The power supply module 203 ensures the stable operation of the system, and the data transmission module 204 realizes real-time transmission and remote access of data. Through these functions, the intelligent sensing brush provides users with personalized calligraphy learning support, improving learning efficiency and writing quality.

[0031] In some embodiments of the present application, the first data collection module 102 includes a first pressure sensor.

[0032] It should be noted that the pressure sensor works based on the changes in its internal physical properties to sense pressure. When the pen tip 101 contacts the paper and is subjected to pressure, the sensing surface of the sensor will be squeezed or deformed. For common piezoresistive pressure sensors, the internal resistance value will change with the change of pressure; while the capacitive pressure sensor reflects the magnitude of the pressure through the change of capacitance value. In this design, the sensing surface of the sensor is in full contact with the pen tip 101, ensuring that the pressure exerted on the pen tip 101 can be directly and accurately sensed. Any slight pressure change can be converted into a change in the electrical signal through the physical mechanism inside the sensor, thereby providing accurate raw data for subsequent data collection and analysis.

[0033] In the first data collection module 102, the pressure data of the pen tip 101 in contact with the paper is used as the first writing data, which is one of the important indicators for evaluating writing effects. The first writing data reflects the amount of pressure applied by the user on the paper during the writing process, which directly affects the thickness, thickness, and stability of the strokes. By evaluating whether the pressure of the pen tip 101 in contact with the paper is uniform and stable, and whether it meets the basic requirements of calligraphy, it helps users understand their pen tip 101 pressure control during writing and make corresponding adjustments. This provides valuable feedback for both beginners and advanced learners, helping them improve their writing skills and enhance their calligraphy skills.

[0034] Optionally, in some embodiments of the present application, the design concept of the first pressure sensor may include but is not limited to the following.

[0035] First, set the first pressure sensor to be a cylinder with a bottom radius of , the height is The bottom radius of the groove for mounting the first pressure sensor is set to , the height is , the length is In order to ensure the stability and safety of the first pressure sensor installation, a certain safety margin needs to be left around the first pressure sensor. , The general value range of is between 0.05 mm and 0.15 mm. Therefore, the bottom radius of the groove for mounting the first pressure sensor is Should meet In addition, a safety margin is also required in the height direction. , The general value range of is between 0.1 mm and 0.25 mm, so the height of the groove Should meet .

[0036] Then, set the radius of the cylinder at the base of the brush to , the length is In order to ensure the structural strength and writing performance of the brush root, the groove bottom radius Should meet At the same time, in order to avoid excessive weakening of the brush root structure, the groove length Should meet .

[0037] Furthermore, to ensure that the sensing surface of the first pressure sensor is in full contact with the base of the pen tip 101 and accurately senses pressure, it is necessary to ensure that the base of the pen tip 101 naturally covers the sensing surface of the first pressure sensor after the first pressure sensor is installed. If the base of the pen tip 101 is thick, the height of the groove may need to be appropriately increased to ensure that the first pressure sensor is properly positioned and that pressure is transmitted between the pen tip 101 and the sensor.

[0038] Finally, the first pressure sensor and the pen tip 101 are connected using a special elastic material and structure, which does not affect the original writing experience of the pen tip 101 while effectively transmitting the pressure signal.

[0039] In some embodiments of the present application, the second data collection module 202 includes a second pressure sensor, an accelerometer, a gyroscope, and a scanner. The second pressure sensor is used to collect grip force data. The accelerometer is used to collect velocity data and acceleration data. The gyroscope is used to collect angular velocity data. The scanner is used to collect font data.

[0040] The second pressure sensor collects grip data while writing. This data reflects the strength with which the user holds the brush during writing. By analyzing this data, we can assess whether the user's grip is appropriate during different stages of writing (such as starting, moving, and ending the stroke). This is crucial for calligraphy learning, as proper grip directly impacts writing stability and stroke quality. Grip data can also help users understand and improve their writing technique, thereby enhancing their writing performance.

[0041] The accelerometer is used to collect speed and acceleration data of the brush body 100 during writing. The speed and acceleration data reflect the movement of the brush during the writing process. By analyzing this data, the user's writing speed and rhythm can be understood. Writing quickly may result in less detailed strokes, while writing slowly can better control the shape of the strokes. Acceleration data can also reveal dynamic changes during the writing process, such as the rapid start and end of the stroke. This data is very useful for assessing writing fluency and coherence, helping users adjust their writing speed to achieve better writing results.

[0042] The gyroscope is used to collect angular velocity data of the brush body 100 during writing. The angular velocity data reflects the rotation and tilt of the brush during the writing process. By analyzing the angular velocity data, the user's pen movement and gestures during writing can be understood.

[0043] The scanner is used to collect font data of the brush body 100 when writing. The font data includes the shape, structure, and details of the characters written by the user. The scanner can obtain the actual strokes and font images written by the user. This data can be used to evaluate the writing effect, such as the aesthetics of the font, the uniformity of the stroke thickness, the stability of the structure, etc. The font data can also be compared with standard fonts to help users identify problems in writing and make targeted improvements. In addition, the scanner can also be used to record and save the user's handwriting for subsequent review and analysis.

[0044] Therefore, through the combination of a second pressure sensor, accelerometer, gyroscope, and scanner, the second data collection module 202 is able to comprehensively collect various data about the user during the writing process, including grip strength, speed, acceleration, angular velocity, and font data. This data provides users with a multi-dimensional writing analysis framework, helping them gain a deeper understanding of their writing habits and techniques. Combined with data analysis, users can obtain personalized writing feedback and suggestions, allowing for more targeted practice and improvement. This data is not only helpful for individual learning, but also provides valuable information for calligraphy teaching and research. Teachers can use this data to guide students more scientifically, and researchers can use it to conduct more in-depth artistic and technical research.

[0045] In some embodiments of the present application, the pen tip 101 is made of animal hair or artificial synthetic materials to ensure the water absorption and ink distribution properties of the pen tip.

[0046] It should be noted that water absorption refers to the ability of the pen tip 101 to absorb and retain water. Good water absorption enables the brush to remain moist for a long time after dipping in ink, thereby ensuring that the ink color is uniform and smooth during the writing process. This is especially important when writing longer paragraphs or applying to a large area, which can avoid frequent dipping in ink and improve writing efficiency. The ink distribution characteristics refer to how the pen tip 101 distributes the ink evenly on the paper. The ideal ink distribution should be able to show the depth and lightness of the ink color while maintaining the clarity and smoothness of the lines when writing. This not only affects the aesthetics of the written work, but also affects whether the writer can accurately express the subtle differences in the art of calligraphy. Using materials with good ink distribution characteristics can better simulate the effect of traditional brushes, allowing writers to be closer to the actual calligraphy creation experience when practicing.

[0047] Furthermore, natural animal hair, due to its unique fiber structure, offers excellent water absorption and ink distribution capabilities. For example, goat hair is soft and has a high water retention capacity, making it suitable for writing with softer strokes; wolf hair, on the other hand, is stiffer and suitable for writing situations that require a stronger pen force. However, animal hair is relatively expensive and requires careful care to maintain its performance. Synthetic materials are generally less expensive, easier to maintain, and in some cases can achieve performance comparable to natural hair. Modern technology allows the manufacture of synthetic nibs 101 with varying degrees of hardness, water absorption, and wear resistance to suit different types of writing needs. Furthermore, synthetic materials can be designed to be more durable, requiring less frequent replacement, thereby reducing the cost of long-term use.

[0048] Therefore, the main purpose of selecting animal hair or synthetic materials as the pen tip 101 is to ensure that the brush can exhibit excellent water absorption and ink distribution characteristics when writing, which is crucial for improving the writing experience, ensuring writing quality, and inheriting traditional calligraphy culture. Whether it is natural or synthetic materials, they need to be selected according to the actual application scenario and personal preferences to achieve the best writing effect. Through such a design, the learning-assisted multimodal sensing brush can not only provide the writing texture required for traditional calligraphy, but also combine modern scientific and technological means to provide users with richer and more personalized learning support.

[0049] Secondly, the embodiment of the present application provides a calligraphy learning method using an intelligent sensing brush, comprising the following steps: Step 101 : collecting first writing data and second writing data of the brush body as initial writing data of the brush body.

[0050] It should be noted that the first writing data includes pressure data of the brush tip in contact with the paper surface, and the second writing data includes grip data and acceleration data, velocity data, angular velocity data, and font data of the brush body.

[0051] In step 101, the first writing data and the second writing data of the brush body are collected as the initial writing data of the brush body, providing a data basis for subsequent calligraphy learning.

[0052] Step 102: Obtain comprehensive writing style recognition results and comprehensive writing quality evaluation results based on the initial writing data and the multimodal writing data analysis method.

[0053] In step 102, based on the initial writing data and the multimodal writing data analysis method, comprehensive writing style recognition results and comprehensive writing quality evaluation results are obtained to provide personalized feedback suggestions for calligraphy learners, helping them understand their strengths and weaknesses, and achieve more targeted practice and effective improvement of calligraphy skills.

[0054] In some embodiments of the present application, in step 102, the implementation process of obtaining the comprehensive writing style recognition result and the comprehensive writing quality evaluation result based on the initial writing data and the multimodal writing data analysis method includes but is not limited to the following steps.

[0055] Step 201: Optimize the initial writing data to obtain third writing data.

[0056] In step 201, the initial writing data is optimized to eliminate interference and errors in the initial writing data, thereby obtaining third writing data, thereby enhancing the consistency and accuracy of data analysis and providing high-quality data for subsequent analysis.

[0057] Step 202: extract features from the third writing data to obtain a writing feature set.

[0058] In step 202, a set of writing features closely related to writing style and quality is extracted from the third writing data to facilitate subsequent analysis and recognition. In addition, feature extraction can reduce the dimensionality of the data, simplify the subsequent analysis process, and improve computational efficiency.

[0059] Step 203: Obtain a comprehensive writing style recognition result and a comprehensive writing quality evaluation result based on the writing feature set and the multimodal writing data analysis method.

[0060] In step 203, the extracted writing feature set is combined with the multimodal writing data analysis method for comprehensive analysis to obtain comprehensive writing style recognition results and comprehensive writing quality evaluation results, providing users with personalized and scientific writing feedback, thereby effectively improving the effect of calligraphy learning.

[0061] In some embodiments of the present application, in step 201, the initial writing data is optimized to obtain the third writing data, including but not limited to the following steps.

[0062] Step 301: Filter the pressure data to obtain filtered pressure data.

[0063] In step 301, a filtering algorithm is used to remove high-frequency noise and transient fluctuations in the pressure data. The filtered pressure data is more stable and reliable, and can more accurately reflect the actual pressure changes during writing.

[0064] In some embodiments of the present application, a low-pass filtering algorithm and a Kalman filter are used to filter the pressure data in the first writing data to obtain filtered pressure data.

[0065] Low-pass filtering algorithms simply allow low-frequency signals to pass while attenuating high-frequency signals. This makes them suitable for removing rapidly changing noise. Different pressure sensors may have different measurement ranges and output ranges. To facilitate subsequent data analysis and algorithm processing, the collected data must be normalized. Common normalization methods include minimum and maximum normalization.

[0066] The Kalman filter is an optimal estimation algorithm based on a linear system state-space model. It assumes that both system noise and observation noise are Gaussian and independent of each other. The algorithm uses two main steps—prediction and update—to reduce the impact of noise on the signal. In the prediction step, the system's state transition equation is used to predict the state and covariance matrix at the next moment. In the update step, the system's measured values are compared with the predicted state to correct the state estimate and covariance matrix.

[0067] Optionally, a low-pass filtering algorithm and a Kalman filter are used to filter the pressure data in the first writing data to obtain filtered pressure data, and the implementation process includes but is not limited to the following steps.

[0068] First, for the collected pressure signal sequence , Represents the index of the sampling point. Filtering is achieved through the differential equation to obtain the signal sequence after low-pass filtering. , Satisfies the following formula (1): (1); In formula (1), when ,set up ; is the feedback coefficient, which controls the memory effect of the filter. A larger value will result in a faster response time but may introduce more high-frequency noise; a larger A low value will increase the memory effect, making the filter react more slowly to rapidly changing signals, but it can more effectively suppress high-frequency noise; is the gain factor, which affects the gain of the filter; and Together they determine the behavior of the low-pass filter.

[0069] Secondly, Use the minimum-maximum normalization method to normalize and obtain the normalized pressure signal sequence , Mapped to interval, Satisfies the following formula (2): (2); In formula (2), for Minimum value of ; for The maximum value of , , , then we can calculate .

[0070] Furthermore, the parameters of the Kalman filter algorithm are initialized. Specifically, first, according to the signal sequence after low-pass filtering, The measurement noise covariance of the Kalman filter algorithm is updated by the normalized pressure signal sequence , satisfy ,in yes The standard deviation of The following formula (3) is satisfied: (3); In formula (3), Is the signal sequence Secondly, assuming that the pressure value at the initial moment is , the initial pressure change rate is 0, the initial state estimation Satisfy the formula , the initial covariance matrix Satisfy the formula ,in is the variance of the initial pressure estimated based on prior knowledge, The variance of the pressure change rate estimated based on prior knowledge. Assume that at a certain moment in the writing process , sampling time interval is 0.01 seconds, and the observation vector corresponding to the electrical signal collected by the first pressure sensor is , then the process noise covariance matrix , measurement noise covariance .

[0071] Then, the state prediction and covariance prediction are performed according to the Kalman filter algorithm. Specifically, the state prediction is performed first. According to the parameters of the Kalman filter algorithm initialized above, it can be known that at a certain moment in the writing process The state estimate at the next moment The following formula (4) is satisfied: (4); In formula (4), It's a certain moment State estimation, assuming ,So ,in For a certain moment The pressure value, For a certain moment The initial pressure change rate. Then the covariance prediction is performed, at a certain moment The covariance matrix of the next moment The following formula (5) is satisfied: (5); In formula (5), it is assumed that at a certain moment The covariance matrix of , then at some point The covariance matrix of the next moment .

[0072] Then, the state update and covariance update are performed according to the Kalman filter algorithm. Specifically, the Kalman gain is first calculated , The following formula (6) is satisfied: (6); In formula (6), is the observation matrix, assuming , measurement noise covariance Then we can calculate . Secondly, update the status. The following formula (7) is satisfied: (7): In formula (7), it is assumed that at a certain moment The observation vector at the next moment is , then we can calculate . Furthermore, the covariance is updated and the updated covariance matrix is The following formula (8): (8); In formula (8), is the unit matrix. According to the above assumptions, we can calculate .

[0073] Finally, by continuously repeating the above steps, the Kalman filter can process the electrical signal of the pressure sensor in real time during the contact between the pen tip and the paper, effectively reducing noise interference and obtaining a more accurate pressure estimate, providing a more reliable data basis for subsequent calligraphy data analysis and learning assistance.

[0074] Step 302 : Filter the grip force data, angular velocity data, velocity data, and acceleration data to obtain filtered grip force data, filtered angular velocity data, filtered velocity data, and filtered acceleration data.

[0075] In step 302, according to the aforementioned formulas (1) to (8), a low-pass filtering algorithm and a Kalman filter are used to filter the grip force data, angular velocity data, velocity data, and acceleration data to obtain filtered grip force data, filtered angular velocity data, filtered velocity data, and filtered acceleration data.

[0076] Step 303: Perform fusion filtering on the acceleration data and the angular velocity data to obtain filtered attitude angle data.

[0077] In step 303, the acceleration and angular velocity data are filtered and fused, enabling the system to obtain more accurate, stable, and reliable attitude angle data. This not only improves data quality and accuracy, but also enhances the system's real-time and robustness, providing users with intuitive feedback and personalized guidance, while also supporting calligraphy teaching and research. Ultimately, this step effectively enhances the overall performance and user experience of the intelligent sensing brush system.

[0078] Optionally, an extended Kalman filter algorithm and a complementary filter algorithm are used to filter and fuse the acceleration data and angular velocity data in the second writing data to obtain filtered attitude angle data. The implementation process of this step includes but is not limited to the following steps.

[0079] First, the parameters of the extended Kalman filter algorithm are initialized. Specifically, the state vector of the brush posture is set to ,in is the pitch angle, for The corresponding angular velocity; is the roll angle, for The corresponding angular velocity; is the yaw angle, for The corresponding angular velocity. Then set the initial state value Assuming that the initial posture angle is 0 and the initial angular velocity is also 0, the initial state value is At the same time, initialize the state covariance matrix , is a diagonal matrix, assuming , indicating that there is a certain estimation error for the initial attitude angle and angular velocity. Process noise covariance matrix The uncertainty of the brush motion model is determined. Since the brush motion is relatively stable, but there are uncertainties such as slight shaking of the writer's hand, for the rate of change of the angle state (angular velocity), the standard deviation of the process noise is assumed to be , for the angle state, the process noise standard deviation is . Then the process noise covariance matrix Can be set to: ; Measurement noise covariance matrix It is determined by the measurement accuracy of the accelerometer and gyroscope. Assume that the noise standard deviation of the accelerometer in measuring acceleration in all directions is , the standard deviation of the noise of the gyroscope measuring angular velocity in each axis is The measurement vector is ,in Accelerometer measurements, is the gyroscope measurement value), then the measurement noise covariance matrix Can be set to: ; Secondly, perform state prediction and covariance prediction. Specifically, first perform state prediction and establish the state equation based on the kinematic model of the brush. ,in is the process noise. Assuming that there is no external torque to change the angular velocity, for a simple angle and angular velocity model that ignores complex external forces and coupling effects, the pitch angle and The corresponding angular velocity The update equation is , ; Similarly, the roll angle and the corresponding angular velocity ; Yaw angle and the corresponding angular velocity There is also a corresponding update equation. Then the state prediction equation is . Secondly, perform covariance prediction. Specifically, the covariance matrix of the predicted state Satisfy the formula ,in is the Jacobian matrix of the state equation with respect to the state vector. Can be set to: ; Next, perform state update and covariance update. First, calculate the measurement equation ,in is a function that maps the state vector to the measurement space, is the measurement noise. For the accelerometer part, the measurement equation is related to the attitude angle. For example, in the body coordinate system, the accelerometer measurement value contains the components of gravity acceleration in all directions, which are related to the attitude angle. For the gyroscope part, the measurement value is directly related to the angular velocity. Then calculate the Kalman gain Satisfy the formula , where Jacobian matrix of the measurement equation with respect to the state vector. Update the state estimate based on the measurement value , and at the same time, update the covariance matrix of the state estimate ,in is the identity matrix.

[0080] Then, the complementary filtering algorithm is used to acquire and initialize data. Specifically, linear acceleration data is acquired from the accelerometer. , obtain angular velocity data from the gyro , use the initial measurement value of the accelerometer to initialize the attitude angle estimation. For example, the initial attitude angle is calculated based on the component of the gravity acceleration measured by the accelerometer in the body coordinate system. Let is the magnitude of the gravitational acceleration, then the initial pitch angle The following formula (9) is satisfied: (9); In formula (9), 、 and is the accelerometer measurement at the initial moment. Similarly, the initial roll angle and the initial yaw angle It can also be calculated through similar geometric relationships.

[0081] Furthermore, the attitude angle is calculated and fused. Specifically, first at each sampling moment , the attitude angle is calculated based on the accelerometer measurement value, where the pitch angle The following formula (10) is satisfied: (10); Similarly, at each sampling moment , roll angle and yaw angle It can also be calculated by similar geometric relationships. Then, the attitude angle is obtained by integrating the angular velocity measured by the gyroscope, and the sampling time interval is set to , then the pitch angle Satisfy the formula ,in is the estimated value of the initial pitch angle. Similarly, the estimated values of the roll angle and yaw angle are updated. Finally, the attitude angle calculated based on the accelerometer measurement value and the attitude angle obtained by integrating the angular velocity measured by the gyroscope are filtered and fused. The filter coefficient is , The value of is between 0 and 1, and the pitch angle after fusion is estimated Satisfy the formula , the fused roll angle estimation Satisfy the formula , the yaw angle estimation after fusion Satisfy the formula .

[0082] Determined by accelerometer and gyroscope performance If the accelerometer data is relatively stable at low frequencies, you can increase the If you want to rely more on the high frequency response of the gyroscope, reduce In this way, the attitude angle estimates from the accelerometer and gyroscope are fused at each sampling moment to obtain the brush's attitude estimate. This fusion process continues during the writing process to obtain accurate brush attitude information in real time.

[0083] Step 304 : Cleaning and segmenting the font data to obtain font structure information of each font in the font data as font structure data.

[0084] In step 304, the font data is cleaned and segmented to extract the font structure information for each font. The following steps can be followed. These steps are designed to clean up noise in the data, standardize the data format, and accurately extract the structural information of each character, laying the foundation for subsequent data analysis, visualization, and learning feedback.

[0085] In some embodiments of the present application, the implementation process of cleaning and segmenting font data to obtain font structure information of each font in the font data as font structure data includes but is not limited to the following steps.

[0086] First, the font data is cleaned to remove non-character data points that appear during the writing process (such as irrelevant data points generated by accidental touches and jitter), ensuring that subsequent analysis is based only on valid character data. A smoothing algorithm is then used to reduce glitches caused by slight hand shaking during writing, making the data more similar to the actual writing trajectory and facilitating accurate analysis of stroke details. Finally, normalization is performed to standardize the scale of different handwriting samples, ensuring fair comparison even with varying handwriting sizes. This helps establish a consistent standard for analyzing handwriting differences between different users.

[0087] Next, the cleaned font data is segmented. Stroke segmentation is performed first, automatically dividing the entire character into individual strokes based on stroke breaks or significant speed changes. This allows for individual analysis of stroke characteristics, such as length and curvature, which is crucial for understanding the writer's style. Character segmentation is then performed, isolating individual characters from a continuous writing stream. This is crucial for assessing the quality of individual characters and comparing the consistency of writing across different characters.

[0088] Then, based on the segmented character data, the basic structural information of each character is extracted, such as the number of strokes, stroke order, stroke intersections, closed areas, etc. This information is the key element that constitutes the unique appearance of the character and is also the focus of calligraphy learning.

[0089] Through the above cleaning and segmentation processes, high-quality font structure data can be obtained, which in turn supports more refined calligraphy analysis and learning guidance. This not only helps users better understand and improve their own writing skills, but also provides valuable reference materials for calligraphy educators.

[0090] Step 305 : filtering the grip force data, filtering the pressure data, filtering the posture angle data, filtering the angular velocity data, filtering the velocity data, filtering the acceleration data and the font structure data as the third writing data.

[0091] In step 305, the integrated filtered grip data, filtered pressure data, filtered posture angle data, filtered angular velocity data, filtered velocity data, filtered acceleration data and font structure data are used as the third writing data, which can provide more accurate and useful information for calligraphy learning, thereby supporting personalized learning guidance and feedback, and improving learning efficiency and writing level.

[0092] In some embodiments of the present application, in step 202, the feature extraction of the third writing data to obtain a writing feature set includes but is not limited to the following steps.

[0093] Step 401 : extract features from the filtered pressure data and the filtered grip strength data in the third writing data to obtain pressure features.

[0094] It should be noted that the pressure characteristics include grip force characteristics, pressure change frequency characteristics and pressure change amplitude characteristics.

[0095] In step 401, by calculating the statistical quantities such as the mean, median and standard deviation of the pen grip force, the grip force characteristics are obtained, which can be used to understand the overall trend and stability of the force applied by the writer during the entire writing process. The grip force characteristics help to assess whether the writer has solid basic skills and his or her ability to control the writing tools. By calculating the number of pressure changes per unit time, the pressure change frequency characteristics are obtained, which can reflect the writer's writing rhythm and fluency. Rapid and frequent pressure changes may indicate fast writing speed and rapid switching between strokes, while conversely, it may mean that the writing is relatively slow or cautious. By analyzing the difference between the maximum and minimum pressure values, the pressure change amplitude characteristics are obtained, which can reveal the writer's ability to control the force in different strokes. Large pressure changes may mean that the writing style is more unrestrained or expressive, while smaller pressure changes may imply that the writer pursues delicacy and stability.

[0096] Step 402: extract features from the filtered posture angle data and filtered angular velocity data in the third writing data to obtain posture features.

[0097] It should be noted that the posture features include posture angle features and angular velocity features.

[0098] In step 402, by analyzing the magnitude and direction of changes in angular velocity, angular velocity features are generated, providing insight into the speed and intensity of the writing action. For example, at a turning point or when quickly starting to write, angular velocity changes significantly, which is crucial for analyzing the writer's agility and fluidity. By calculating the mean, range, and frequency of changes in the posture angles (pitch, roll, and yaw), posture angle features are generated, which can depict the writer's habitual posture during writing, as well as the magnitude and frequency of posture changes. This helps identify the writer's individual style and provide suggestions for improving their writing posture.

[0099] Step 403: extract features from the filtered velocity data and filtered acceleration data in the third writing data to obtain motion features.

[0100] It should be noted that motion characteristics include velocity characteristics and acceleration characteristics.

[0101] In step 403, by measuring the speed of the pen tip during writing, a velocity feature is generated, which can be used to assess the writer's writing efficiency and fluency. The velocity feature also helps identify whether the writer is writing too quickly, resulting in imprecise strokes, or writing too slowly, affecting efficiency. By analyzing the changes in acceleration during pen tip movement, an acceleration feature is generated, which helps understand the writer's speed regulation during different stroke stages. The acceleration feature can effectively capture key movements during writing, such as rapid starting and ending of the pen.

[0102] Step 404: Extract the font structure data in the third handwriting data to obtain font structure features. The font structure features include center of gravity position features and center position features.

[0103] In step 404, a coordinate system is constructed to measure the center of gravity and center position of each Chinese character. These center of gravity and center position features are generated to assess the writer's mastery of the font's structural balance. The degree to which the center of gravity deviates from the center position reflects the writer's ability to control the font's layout. The center position feature determines the center of the overall character layout and helps assess the writer's performance in maintaining the overall symmetry and aesthetics of the font.

[0104] Optionally, by constructing a coordinate system, the center of gravity position and center position of a single Chinese character are measured to obtain the center of gravity position feature and center position feature. The implementation process includes but is not limited to the following steps. Specifically, for scanning single characters written with a brush, draw a virtual square to frame the character to be analyzed. For fonts with strict structure such as regular script, the reasonable range can be relatively small, for example, with the center of the square as the center of the circle and a radius of 1 / 6 or 1 / 8 of the length of the square side. For fonts with strong dynamics such as running script and cursive script, the reasonable range can be appropriately relaxed. After practicing a certain number of single characters (such as 30 or 50), measure the center of gravity position of each character one by one, and determine whether it is within a reasonable range.

[0105] Step 405: The pressure feature, posture feature, motion feature and font structure feature are taken as a writing feature set.

[0106] In step 405, pressure characteristics, posture characteristics, motion characteristics, and font structure characteristics are integrated into a set of writing features. This set provides a comprehensive view of the writer's technique, style, and personal habits. Using these characteristics, the system can provide users with personalized feedback and suggestions to help them improve their writing skills and calligraphy. Furthermore, these characteristics can help calligraphy educators better understand students' performance and adopt more effective teaching strategies.

[0107] In some embodiments of the present application, in step 203, the implementation process of obtaining the comprehensive writing style recognition result and the comprehensive writing quality evaluation result based on the writing feature set and the multimodal writing data analysis method includes but is not limited to the following steps.

[0108] Step 501 : Obtain a first writing style recognition result and a first writing quality evaluation result based on the pressure feature in the writing feature set and in combination with a multimodal writing data analysis method.

[0109] In step 501, pressure characteristics (such as grip strength, pressure variation frequency, and amplitude) can reflect the force characteristics applied by the writer to the brush. Combined with multimodal data analysis methods, these characteristics can be used to infer the writer's writing style, such as whether they tend to use heavier pressure (possibly indicative of regular script) or lighter pressure (possibly in running or cursive script). Furthermore, the uniformity, stability, and variation pattern of pressure characteristics are also used to assess writing quality. For example, excessively unstable pressure may indicate a lack of proficiency or poor control.

[0110] Step 502 : Based on the posture features in the writing feature set and in combination with a multimodal writing data analysis method, a second writing style recognition result and a second writing quality evaluation result are obtained.

[0111] In step 502, posture features (such as posture angle and angular velocity) reveal the dynamic posture of the brush during the writing process. By analyzing these features, the writer's unique posture patterns in different fonts (such as regular script and running script) can be identified. Furthermore, posture stability, adjustment frequency, and range of variation are also important indicators for evaluating writing quality. Good writing posture is generally associated with higher writing quality.

[0112] Step 503 : Obtain a third writing style recognition result and a third writing quality evaluation result based on the motion features in the writing feature set and in combination with a multimodal writing data analysis method.

[0113] In step 503, motion characteristics (such as velocity and acceleration) demonstrate the motion characteristics of pen tip 101 during writing. Different writing styles have different velocity and acceleration profiles; for example, running script is generally faster than regular script. Motion characteristics help distinguish different writing styles, and changes in velocity and acceleration directly affect writing quality. For example, excessive velocity may result in rough strokes, while appropriate acceleration can make writing smooth and natural.

[0114] Step 504 : Obtain a fourth writing style recognition result and a fourth writing quality evaluation result based on the font structure features in the writing feature set and in combination with a multimodal writing data analysis method.

[0115] In step 504, font structural features (such as the center of gravity and center position) describe the visual balance and composition of the written characters. Fonts of different styles have unique structural characteristics, and by analyzing these characteristics, the writer's style can be identified. Furthermore, the structural rationality and proportion harmony of characters are also important components of handwriting quality. Well-structured handwriting generally creates an aesthetically pleasing experience.

[0116] Step 505 : Obtain a comprehensive writing style recognition result based on the first writing style recognition result, the second writing style recognition result, the third writing style recognition result, and the fourth writing style recognition result.

[0117] In step 505, by integrating the first to fourth writing style recognition results from the four dimensions of pressure, posture, motion, and font structure, a comprehensive and accurate writing style recognition model can be formed. This helps to more finely distinguish different writing styles and also provides a basis for personalized calligraphy learning, helping users discover and develop their own unique writing style.

[0118] Step 506 , obtaining a comprehensive writing quality evaluation result based on the first writing quality evaluation result, the second writing quality evaluation result, the third writing quality evaluation result, and the fourth writing quality evaluation result.

[0119] In step 506, the first through fourth writing quality evaluation results are combined to produce a comprehensive writing quality score. This not only considers the physical skills involved in the writing process but also incorporates the artistic aesthetics of the final work, providing comprehensive and specific feedback to the user. This evaluation helps users understand where they excel and where they need improvement, promoting the overall improvement of their calligraphy skills.

[0120] Furthermore, an embodiment of the present application provides a calligraphy learning system for an intelligent sensing brush, comprising a first data acquisition module, a second data acquisition module and a data analysis module.

[0121] The first data acquisition module is used to collect first writing data of the brush body. The first writing data includes pressure data of the contact between the tip of the brush body and the paper surface.

[0122] The second data acquisition module is used to collect second writing data of the brush body. The second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body.

[0123] The data analysis module is used to use the first writing data and the second writing data as initial writing data of the brush body, and obtain a comprehensive writing style recognition result and a comprehensive writing quality evaluation result based on the initial writing data and the multimodal writing data analysis method.

[0124] In some embodiments of the present application, the data analysis module includes a data optimization unit, a feature extraction unit, and a feature analysis unit.

[0125] The data optimization unit is used to optimize the initial writing data to obtain third writing data.

[0126] The feature extraction unit is used to extract features from the third writing data to obtain a writing feature set.

[0127] The feature analysis unit is used to obtain comprehensive writing style recognition results and comprehensive writing quality evaluation results based on a writing feature set and a multimodal writing data analysis method.

[0128] In some embodiments of the present application, the calligraphy learning system of the smart sensor brush provided by the embodiment of the present application also includes a visualization module, which is used to present third writing data, writing feature set, writing style recognition results and writing quality evaluation results.

[0129] The visualization module presents the results of the data analysis module to the user in an intuitive and easy-to-understand manner. It displays third-party writing data, a collection of writing features, writing style recognition results, and writing quality assessment results. Through charts, animations, and other forms, users can more clearly understand their own writing performance, identify shortcomings, and conduct targeted training based on the visual feedback.

[0130] Through data analysis and visualization, the data collected by the sensor brush can be more deeply understood and utilized, providing strong support for calligraphy learning, teaching, and research. Through intuitive visualizations and animations, users can more clearly understand their own writing habits and characteristics, identify existing problems, and implement targeted improvements and enhancements. Furthermore, these data analysis and visualization results can provide teaching references for calligraphy educators, helping them develop more scientific teaching plans and methods, and improving teaching effectiveness.

[0131] The visualization module includes a pen tip pressure change curve interface, a posture angle change curve interface, a pen tip pressure and grip force relationship interface, and a speed analysis interface.

[0132] Reference Figures 5 to 9The pen tip pressure change curve interface is used to display the different numbers of practice times and the standard pen tip pressure change curve diagram when writing five different fonts. The pen tip pressure change curve interface can intuitively show how the pressure applied by the pen tip changes over time when the user writes different fonts. By comparing with the standard pen tip pressure change curve, the user can clearly see the gap between themselves and the professional level. By recording the changes in pen tip pressure under different numbers of practice, it can help users track their progress. For example, as the number of practice increases, if the user's pen tip pressure changes gradually approach the standard curve, it means that their writing skills are improving. In addition, by observing the pen tip pressure change curve, users can more easily find their own problems. For example, the pressure at certain moments is too high or too low, or the pressure change is not smooth enough. These are areas that need improvement.

[0133] Reference Figures 10 to 14 The posture angle change curve interface is used to display a schematic diagram of the posture angle change curve over time when writing five different fonts. This interface can intuitively show the changes in the posture angle of the brush (including pitch angle, roll angle, and yaw angle) when the user writes different fonts. This helps users understand their pen holding posture and pen movement during the writing process. By comparing with the standard posture angle change curve, users can clearly see whether their writing posture meets the standards and the specific differences that exist. This comparison can help users identify areas that need improvement, such as incorrect or unstable posture angles of certain strokes. By recording the changes in posture angles under different numbers of practice, users can track their progress and development trends. If the user's posture angle changes gradually approach the standard curve with increasing number of practices, it means that their writing skills are improving.

[0134] Reference Figure 15 , the pen tip pressure and grip force relationship interface is used to display a schematic diagram of the relationship between pen tip pressure and grip force when writing five fonts. This interface can intuitively show the relationship between the pressure applied by the pen tip and the pen grip strength when the user writes different fonts. Through scatter plots or other forms of charts, you can clearly see the mutual influence and change pattern between the two. By analyzing the relationship between pen tip pressure and grip force, users can identify possible problems. For example, if the grip force is too strong but the pen tip pressure is too low, the strokes may not be full enough; conversely, if the grip force is small and the pen tip pressure is too high, the writing may be unsmooth or the strokes may be too thick. This visual tool makes problems easier to detect.

[0135] Reference Figure 16 and Figure 17The speed analysis interface is used to display a diagram of the relationship between the pen tip pressure and speed and a diagram of the pen speed changing over time. The diagram of the relationship between the pen tip pressure and speed can intuitively show the relationship between the pressure applied by the pen tip and the pen speed during the writing process. Through this correlation diagram, the user can clearly see the changing pattern of pressure at different speeds. By observing the relationship between speed and pressure, the user can more easily discover potential problems in writing. For example, if the pressure is too high or too low at high speeds, it may cause the stroke quality to deteriorate; and if the pressure is unstable at low speeds, it may cause unsmooth writing. The diagram of the pen speed changing over time shows how the pen speed changes over time, helping users understand their ability to accelerate, decelerate, and maintain a constant speed during the writing process. It can help identify problems with the writing rhythm, such as whether the speed can be smoothly controlled and whether there are sudden accelerations or decelerations.

[0136] In some embodiments of the present application, the calligraphy learning system of the smart sensor brush provided in the embodiment of the present application also includes an interaction module, and the interaction module includes an alarm prompt unit and a user interaction unit.

[0137] The interactive module enables users to interact with the system more intuitively and conveniently. The interactive module mainly includes the alarm prompt unit and the user interaction unit, each of which has its specific function and significance.

[0138] The alarm prompt unit is used to issue alarm messages and prompt messages based on the initial writing data and the preset standard writing data range. When the user's writing data exceeds the preset standard writing data range, the alarm prompt unit will issue an alarm message. For example, if the user's grip strength is too strong or too weak, or the writing speed is too fast or too slow, the system will remind the user through an alarm message. This helps users to promptly identify and correct problems in writing and improve writing quality. In addition to alarm messages, the alarm prompt unit can also issue prompt messages and provide suggestions on how to improve writing. For example, if the user's stroke order is incorrect, the system can prompt the correct stroke order. This instant feedback can help users master correct writing techniques more quickly.

[0139] The user interaction unit allows users to perform data query operations, adjust control operations, provide feedback and suggestions, and configure and define operations. Through the user interaction unit, users can query their own writing data, including historical writing records, writing style recognition results, and writing quality evaluation results. This helps users understand their progress and writing performance. Users can also adjust some system control parameters, such as data collection frequency and alarm thresholds. This allows users to customize system behavior according to their needs and preferences.

[0140] In summary, the smart sensing brush and calligraphy learning method and system provided in the embodiments of the present application provide the following beneficial effects.

[0141] The embodiments of this application utilize advanced sensor technology and data analysis methods to effectively improve the efficiency and quality of calligraphy learning. The system monitors the user's writing data, such as grip force, pressure, and posture angle, in real time and provides instant feedback. For example, if the user's grip force is too strong or too weak, the system will immediately issue an alert, prompting the user to make adjustments. This real-time feedback mechanism helps users promptly identify and correct writing problems, allowing them to master correct writing techniques more quickly. Furthermore, the system provides personalized learning suggestions and guidance based on each user's specific situation, making the learning process more targeted and efficient. Through an intuitive visualization module, users can view their writing data in the form of charts and animations, such as a graph showing pressure changes over time and a graph showing changes in posture angle. This visualization makes it easier for users to understand and improve their writing habits, enhancing the user experience. The interactive module allows users to query data, adjust control parameters, and provide feedback and suggestions, improving the system's ease of use and user engagement. Teachers can use the data analysis and visualization results provided by the system to understand students' writing performance and develop more effective teaching plans and methods. The system also supports distance learning. Teachers can view students' writing data through the Internet and provide remote guidance, breaking geographical restrictions and making calligraphy teaching more flexible and convenient.

[0142] Although the embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

[0143] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. Intelligent sensor writing brush, characterized in that, Including the brush body; The brush body comprises a pen tip and a pen holder; The base of the pen tip is provided with a first data collection module, and the first data collection module is used to collect pressure data of the contact between the pen tip and the paper as first writing data; The pen holder is internally provided with a data processing module, a second data collection module, a power supply module and a data transmission module; the second data collection module is used to collect second writing data of the brush body; the second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body; The first writing data and the second writing data are used as initial writing data of the brush body; The data processing module is used to obtain a comprehensive writing style recognition result and a comprehensive writing quality evaluation result based on the initial writing data and the multimodal writing data analysis method.

2. The intelligent sensor writing brush according to claim 1, characterized in that: The first data collection module includes a first pressure sensor; the second data collection module includes a second pressure sensor, an accelerometer, a gyroscope and a scanner; The second pressure sensor is used to collect the grip force data; The accelerometer is used to collect the velocity data and the acceleration data; The gyroscope is used to collect the angular velocity data; The scanner is used to collect the font data.

3. The intelligent sensor writing brush according to claim 1, characterized in that: The pen tip is made of animal hair or artificial synthetic material to ensure the water absorption and ink distribution characteristics of the pen tip.

4. The calligraphy learning method using the intelligent sensing brush is characterized by: The steps include: collecting first writing data and second writing data of a brush body as initial writing data of the brush body; The first writing data includes pressure data of the tip of the brush body in contact with the paper surface; The second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body; Based on the initial writing data and multimodal writing data analysis method, a comprehensive writing style recognition result and a comprehensive writing quality evaluation result are obtained.

5. The calligraphy learning method using the intelligent sensing brush according to claim 4, characterized in that: The method for analyzing the initial writing data and the multimodal writing data to obtain a comprehensive writing style recognition result and a comprehensive writing quality evaluation result includes: Optimizing the initial writing data to obtain third writing data; performing feature extraction on the third writing data to obtain a writing feature set; According to the writing feature set and the multimodal writing data analysis method, the comprehensive writing style recognition result and the comprehensive writing quality evaluation result are obtained.

6. The calligraphy learning method using the intelligent sensing brush according to claim 5, characterized in that: The optimizing the initial writing data to obtain the third writing data includes: filtering the pressure data to obtain filtered pressure data; Filtering the grip force data, the angular velocity data, the velocity data, and the acceleration data to obtain filtered grip force data, filtered angular velocity data, filtered velocity data, and filtered acceleration data; Performing fusion filtering on the acceleration data and the angular velocity data to obtain filtered attitude angle data; performing cleaning and segmentation processing on the font data to obtain font structure information of each font in the font data as font structure data; The filtered grip force data, the filtered pressure data, the filtered posture angle data, the filtered angular velocity data, the filtered velocity data, the filtered acceleration data and the font structure data serve as third writing data.

7. The calligraphy learning method using the intelligent sensing brush according to claim 5, characterized in that: The extracting features of the third writing data to obtain a writing feature set includes: Performing feature extraction on the filtered pressure data and the filtered grip force data in the third writing data to obtain pressure features; the pressure features include grip force features, pressure change frequency features, and pressure change amplitude features; Performing feature extraction on the filtered posture angle data and the filtered angular velocity data in the third writing data to obtain posture features; the posture features include posture angle features and angular velocity features; Performing feature extraction on the filtered velocity data and filtered acceleration data in the third writing data to obtain motion features, wherein the motion features include velocity features and acceleration features; Extracting features from the font structure data in the third writing data to obtain font structure features; the font structure features include center of gravity position features and center position features; The pressure feature, the posture feature, the movement feature and the font structure feature serve as a writing feature set.

8. The calligraphy learning method using the intelligent sensing brush according to claim 5, characterized in that: Obtaining the comprehensive writing style recognition result and the comprehensive writing quality evaluation result based on the writing feature set and the multimodal writing data analysis method includes: Obtaining a first writing style recognition result and a first writing quality evaluation result based on the pressure feature in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining a second writing style recognition result and a second writing quality evaluation result based on the posture features in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining a third writing style recognition result and a third writing quality evaluation result based on the motion features in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining a fourth writing style recognition result and a fourth writing quality evaluation result based on the font structure features in the writing feature set and in combination with the multimodal writing data analysis method; Obtaining the comprehensive writing style recognition result according to the first writing style recognition result, the second writing style recognition result, the third writing style recognition result, and the fourth writing style recognition result; The comprehensive writing quality evaluation result is obtained according to the first writing quality evaluation result, the second writing quality evaluation result, the third writing quality evaluation result and the fourth writing quality evaluation result.

9. The calligraphy learning system of the intelligent sensing brush is characterized by: It includes a first data acquisition module, a second data acquisition module and a data analysis module; The first data acquisition module is used to collect first writing data of the brush body; the first writing data includes pressure data of the contact between the tip of the brush body and the paper surface; The second data acquisition module is used to collect second writing data of the brush body; the second writing data includes grip force data and acceleration data, velocity data, angular velocity data and font data of the brush body; The data analysis module is used to use the first writing data and the second writing data as the initial writing data of the brush body; Based on the initial writing data and multimodal writing data analysis method, a comprehensive writing style recognition result and a comprehensive writing quality evaluation result are obtained.

10. The calligraphy learning system of the intelligent sensing brush according to claim 9 is characterized in that: The data analysis module includes a data optimization unit, a feature extraction unit and a feature analysis unit; The data optimization unit is used to optimize the initial writing data to obtain third writing data; The feature extraction unit is used to extract features from the third writing data to obtain a writing feature set; The feature analysis unit is used to obtain the comprehensive writing style recognition result and the comprehensive writing quality evaluation result based on the writing feature set and the multimodal writing data analysis method.

Images