A writing ability evaluation method based on key point detection and related device

Abstract

The application discloses a writing ability evaluation method based on key point detection and related devices, which is used to improve the authenticity and effectiveness of the writing ability evaluation result. The application includes: acquiring a video of a writing area in real time through a camera device, and establishing a virtual coordinate system for a practice book in the video; when a user writes on the practice book using a writing pen, detecting the nib of the writing pen through a CenterNet key point detection algorithm, and tracking the motion trajectory of the nib on the practice book through a deepsort algorithm to obtain trajectory information; generating writing data according to the trajectory information and the coordinate system, wherein the writing data includes a plurality of coordinate points of the nib on the practice book and a plurality of time points corresponding to the plurality of coordinate points; when the user finishes writing, analyzing the writing data to obtain an analysis result; and determining the writing speed of the user according to the analysis result.

Description

Technical Field

[0001] This application relates to the field of writing ability assessment technology, and in particular to a writing ability assessment method and related apparatus based on key point detection. Background Technology

[0002] Nowadays, people are paying more and more attention to the all-round development of children's morality, intelligence, physical fitness and aesthetics. Various assessments are conducted in the process of children's growth to understand their development in these aspects. For example, assessments are conducted on children's height, weight, drawing, thinking and writing abilities. However, current assessments of children are mostly limited to height and weight, and feedback on assessments of other dimensions is relatively lacking.

[0003] Current technologies for assessing children's writing abilities typically employ smart writing tablets and pens. Children use these tablets to write, and the tablets provide feedback on their writing skills. However, because smart writing pens differ from ordinary pens and writing tablets from regular writing paper, and because these devices are novel, children's curiosity can lead to unusual actions during writing. Therefore, current methods for assessing children's writing abilities deviate from their actual writing experiences, affecting the accuracy and validity of the assessment results. Summary of the Invention

[0004] To address the aforementioned technical issues, this application provides a writing ability assessment method and related apparatus based on key point detection, thereby improving the authenticity and effectiveness of the writing ability assessment results.

[0005] The first aspect of this application provides a writing ability assessment method based on key point detection, including:

[0006] The video of the writing area, which includes a notebook and a writing pen, is captured in real time by a camera device, and a virtual coordinate system is established for the notebook in the video.

[0007] When a user writes on the exercise book with the writing pen, the pen tip is detected by the CenterNet key point detection algorithm, and the movement trajectory of the pen tip on the exercise book is tracked by the deepsort algorithm to obtain trajectory information;

[0008] Writing data is generated based on the trajectory information and the coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to the coordinate points.

[0009] When the user finishes writing, the writing data is analyzed to obtain the analysis results;

[0010] The user's writing speed is determined based on the analysis results.

[0011] Optionally, the analysis of the written data includes:

[0012] The written data is analyzed using an analysis algorithm to obtain the analysis results. The analysis algorithm includes wavelet filtering and Hampel filtering.

[0013] Optionally, the analysis of the written data using the analysis algorithm includes:

[0014] Based on the written data, determine the waveform showing the relationship between the X-axis coordinate change and the time point change;

[0015] The target waveform is obtained by performing wavelet filtering and Hampel filtering on the aforementioned waveform.

[0016] Determining the user's writing speed based on the analysis results includes:

[0017] The user's writing speed is determined based on the target waveform.

[0018] Optionally, determining the user's writing speed based on the target waveform includes:

[0019] The target time point of the pen tip in the target writing grid is determined according to the target waveform. The exercise book contains several writing grids, and each writing grid contains one character.

[0020] The writing speed of the user in the target writing grid is calculated based on the target time point.

[0021] Optionally, after acquiring the video of the writing area in real time via a camera device, and before establishing a virtual coordinate system for the exercise book in the video, the writing ability assessment method further includes:

[0022] A perspective transformation is performed on the side view of the exercise book in the video to obtain a top view of the exercise book.

[0023] Optionally, before acquiring video of the writing area in real time via a camera device, the writing ability assessment method further includes:

[0024] The camera is controlled to aim at the writing area, and the position of the exercise book is determined by clicking on the four corners of the exercise book.

[0025] The second aspect of this application provides a writing ability assessment system based on key point detection, including:

[0026] The acquisition unit is used to acquire video of the writing area in real time through a camera device. The writing area includes a notebook and a writing pen, and a virtual coordinate system is established for the notebook in the video.

[0027] The processing unit is used to detect the pen tip of the writing pen using the CenterNet key point detection algorithm and track the movement trajectory of the pen tip on the exercise book using the deepsort algorithm when the user writes on the exercise book with the writing pen to obtain trajectory information.

[0028] A generation unit is used to generate writing data based on the trajectory information and the coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to the several coordinate points.

[0029] The analysis unit is used to analyze the writing data when the user finishes writing and obtain analysis results;

[0030] A determining unit is used to determine the user's writing speed based on the analysis results.

[0031] Optionally, the analysis unit includes:

[0032] The analysis module is used to analyze the written data using analysis algorithms to obtain analysis results. The analysis algorithms include wavelet filtering and Hampel filtering.

[0033] Optionally, the analysis module includes:

[0034] The first determining module is used to determine the waveform of the relationship between the change in X-axis coordinate and the change at time points based on the written data;

[0035] The processing submodule is used to perform wavelet filtering and Hampel filtering on the relational waveform to obtain the target waveform;

[0036] The determining unit includes:

[0037] The second determining module is used to determine the user's writing speed based on the target waveform.

[0038] Optionally, the second determining module includes:

[0039] The determination submodule is used to determine the target time point of the pen tip in the target writing grid based on the target waveform. The exercise book contains several writing grids, and each writing grid contains one character.

[0040] The calculation submodule is used to calculate the user's writing speed in the target writing grid based on the target time point.

[0041] Optionally, the writing ability assessment system based on key point detection further includes:

[0042] The transformation unit is used to perform perspective transformation on the side view of the exercise book in the video to obtain a top view of the exercise book.

[0043] Optionally, the writing ability assessment system based on key point detection further includes:

[0044] The control unit is used to control the camera device to align with the writing area and to determine the position of the exercise book by selecting the four corners of the exercise book.

[0045] A third aspect of this application provides a writing ability assessment device based on key point detection, comprising:

[0046] Central processing unit, memory, input / output interfaces, wired or wireless network interfaces, and power supply;

[0047] The memory is either a short-term storage memory or a persistent storage memory;

[0048] The central processing unit is configured to communicate with the memory and execute instructions in the memory to perform any of the first aspect and any of the alternative methods of the first aspect.

[0049] A fourth aspect of this application provides a computer-readable storage medium including instructions that, when executed on a computer, cause the computer to perform any of the methods described in the first aspect and its alternatives.

[0050] As can be seen from the above technical solutions, this application has the following effects:

[0051] By acquiring real-time video of the user's writing process, a virtual coordinate system is established on the exercise book within the video. The CenterNet keypoint detection algorithm is used to detect the pen tip, and the deepsort algorithm is used to track the pen tip's movement trajectory on the exercise book to obtain trajectory information. Then, writing data is generated based on the trajectory information and the coordinate system. When the user finishes writing, the writing data is analyzed, and the user's writing speed is determined based on the analysis results. In this way, the writing ability assessment process does not require the use of professional smart writing pens or smart writing tablets; ordinary pens and exercise books are sufficient. This reduces the influence of testing tools on children's writing assessment process and can more closely simulate children's real writing environment, thereby improving the authenticity and effectiveness of the assessment results. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a schematic diagram of the writing ability assessment method based on key point detection in this application;

[0054] Figure 2 This is another schematic diagram of the writing ability assessment method based on key point detection in this application;

[0055] Figure 3 This is a schematic diagram of the writing ability assessment system based on key point detection in this application;

[0056] Figure 4 This is another schematic diagram of the writing ability assessment system based on key point detection in this application;

[0057] Figure 5 This is a schematic diagram of the writing ability assessment device based on key point detection according to this application;

[0058] Figure 6 This is a schematic diagram of the relationship waveform obtained in the writing ability assessment method based on key point detection in this application. Detailed Implementation

[0059] This application provides a writing ability assessment method and related apparatus based on key point detection, which can improve the authenticity and effectiveness of writing ability assessment results.

[0060] It should be noted that the writing ability assessment method based on key point detection provided in this application can be applied to terminals, systems, and servers. For example, terminals can be smartphones, computers, tablets, smart TVs, smartwatches, portable computer terminals, or fixed terminals such as desktop computers. For ease of explanation, this application uses a system as the implementing entity for illustration.

[0061] Please see Figure 1 , Figure 1 This is a schematic diagram of a writing ability assessment method based on key point detection provided in this application. The writing ability assessment method includes:

[0062] 101. The system acquires real-time video of the writing area through a camera device. The writing area includes exercise books and writing pens, and a virtual coordinate system is established for the exercise books in the video.

[0063] There is a connection between the camera equipment and the system, which allows the camera equipment and the system to transmit data to each other, such as wireless connection, wired connection, etc.

[0064] The writing area is the area where the user's hand rests during the writing process, and this area is larger than the area of ​​the exercise book. The camera is always pointed at this writing area and captures video of it in real time, ensuring that the exercise book is clearly visible and completely in the video. A virtual coordinate system is established with the top left corner of the exercise book as the origin, the X-axis from left to right, and the Y-axis from top to bottom.

[0065] It is understood that the users described in this application are children, and this application is used in the process of assessing children's writing ability.

[0066] 102. When a user writes on a notebook with a writing pen, the system detects the pen tip using the CenterNet key point detection algorithm and tracks the movement trajectory of the pen tip on the notebook using the deepsort algorithm to obtain trajectory information.

[0067] The system is equipped with CenterNet keypoint detection algorithm and deepsort algorithm. When the user is taking a writing ability assessment, that is, when the user is writing on the exercise book with a writing pen, the system sets the keypoint as the pen tip and uses CenterNet keypoint detection algorithm to detect the pen tip in the video. Then, the deepsort algorithm tracks the movement trajectory of the pen tip on the exercise book to obtain the trajectory information of the pen tip movement.

[0068] The video captured by the camera device includes other objects such as the desktop, the user's hand, the notebook, and the pen. The CenterNet keypoint detection algorithm can accurately identify the pen tip, facilitating its tracking. The deepsort algorithm can accurately track the pen tip, and the system records the tracked path, forming trajectory information. This trajectory information is continuously recorded during the writing process. This embodiment utilizes keypoint detection technology to track and locate the pen tip trajectory.

[0069] 103. The system generates writing data based on trajectory information and coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to these coordinate points.

[0070] After the system obtains the trajectory information, the system generates writing data based on the trajectory information and the established virtual coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book, and several time points corresponding to the several coordinate points. When the pen tip touches the exercise book, there will be coordinate points. The system records the coordinate points and also records the time points corresponding to the recorded coordinate points. For example, the pen tip is at coordinate point A corresponding to time point a, and at coordinate point B corresponding to time point b. The movement trajectory of the pen tip on the exercise book has several coordinate points, and there is a time point corresponding to each coordinate point.

[0071] It should be noted that there are several writing grids arranged neatly on the exercise book. The writing grids are used for users to write words. One writing grid can accommodate one word. During the writing ability evaluation process, the test words in some writing grids are fixed. For example, the words in the writing grids of the first three rows are all "中", and the words from the fourth row to the sixth row are all "东", etc.

[0072] 104. When the user finishes writing, the system analyzes the writing data to obtain an analysis result.

[0073] When the user finishes writing, the system records the writing data from the start of writing to the end of writing. The system analyzes this writing data and then obtains an analysis result. The analysis process is to convert the writing data into other forms that are easy to understand intuitively. The analysis process is specifically described in the embodiments. Figure 2 as described in the embodiments.

[0074] 105. The system determines the user's writing speed according to the analysis result.

[0075] Writing speed is one of the user's writing abilities. Generally, writing abilities can also include the user's concentration during writing, the smoothness of the pen movement during writing, etc. [[ID=!18]]

[0076] In this embodiment, the system determines the user's writing speed according to the analysis result obtained from the analysis. In addition, other writing abilities of the user can be directly or indirectly determined through the analysis result and the writing speed.

[0077] In this embodiment, there is no need to use a smart writing pen or a smart writing board to evaluate the writing ability. Instead, it is carried out by using a common writing pen and an exercise book. The trajectory information of the pen tip during writing is obtained through video shooting, in cooperation with the CenterNet key point detection algorithm and the deepsort algorithm. Writing data is generated based on the trajectory information and the virtual coordinates of the exercise book. Finally, the analysis result is obtained according to the analysis of the writing data, and the writing speed is determined through the analysis result. Thus, the influence of test tools such as smart writing pens or smart writing boards on the evaluation process can be reduced, and the real writing environment of children can be simulated to a greater extent, thereby improving the authenticity and effectiveness of the evaluation results.

[0078] Please continue reading. Figure 2 , Figure 2 Another schematic diagram of a writing ability assessment method based on key point detection provided in this application, the writing ability assessment method includes:

[0079] 201. The system controls the camera to aim at the writing area and determines the position of the exercise book by clicking on the four corners of the exercise book.

[0080] The system is interconnected with the camera equipment. The camera is angled downwards towards the exercise book to avoid obstruction by the user. The camera angle can be adjusted and is controlled by the system. When the camera cannot be fully aligned with the writing area of ​​the exercise book, the system controls the camera to align the shooting direction with the writing area. The exercise book's position is then determined by clicking on its four corners. For example, the captured video is transmitted to the system in real time. Operators or staff can locate the exercise book on the system and select its area by clicking on its four corners with the mouse, facilitating the establishment of a virtual coordinate system for the exercise book.

[0081] 202. The system acquires real-time video of the writing area through a camera device. The writing area includes exercise books and writing pens, and establishes a virtual coordinate system for the exercise books in the video.

[0082] Step 202 in this embodiment and Figure 1 Step 101 in the embodiment is similar and will not be repeated here.

[0083] Optionally, in one feasible embodiment, this embodiment further includes the system performing perspective transformation on the side view of the exercise book in the video to obtain a top view of the exercise book. The system establishes a virtual coordinate system for this top view of the exercise book. Generally, the camera device captures the exercise book at an angle, resulting in an angled view. Establishing a coordinate system for an angled view has a significant deviation. Therefore, the system transforms the angled view into a top view through perspective transformation and then establishes a virtual coordinate system for the top view. Specifically, the algorithm used by the system for perspective transformation mainly includes the `perspectiveTransform()` and `warpPerspective()` methods in OpenCV. In this embodiment, perspective transformation is used to avoid interference from the camera device's view of the person being tested writing, thus maximizing the reproduction of the most realistic writing scene.

[0084] 203. When a user writes on a notebook with a writing pen, the system detects the pen tip using the CenterNet key point detection algorithm and tracks the movement trajectory of the pen tip on the notebook using the deepsort algorithm to obtain trajectory information.

[0085] 204. The system generates writing data based on trajectory information and coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to these coordinate points.

[0086] Steps 203 and 204 in this embodiment are Figure 1 Steps 102 and 103 in the embodiments are similar and will not be repeated here.

[0087] 205. The system analyzes the written data using algorithms to obtain analysis results. The analysis algorithms include wavelet filtering and Hampel filtering.

[0088] The system uses wavelet filtering and Hampel filtering algorithms to analyze the written data. In practice, the written data is filtered using two methods: wavelet filtering filters the Y-axis coordinates and only analyzes the X-axis coordinates and time points, while Hampel filtering replaces outliers with the median of adjacent non-outliers when outliers are found.

[0089] 206. The system determines the waveform of the relationship between the change in X-axis coordinate and the change at time points based on the written data.

[0090] Please continue to participate. Figure 6 The relationship waveform determined by the system is as follows: Figure 6 As shown, Figure 6 This represents the waveform obtained when writing three lines, each with five characters, during the writing process. The length represents the time spent writing the corresponding grid for each line. Anomalies include... Figure 6 As indicated by number 1, anomalies are caused by the user's hand movements during writing, especially in children who have difficulty concentrating and are prone to inattention and hand movements during writing. For example, if each line takes a relatively slow and even time, it indicates that the pen strokes are not smooth enough; if writing a particular character in a line takes an exceptionally long time, it indicates that the user is not concentrating while writing that character. Figure 6 There are two waveforms: the more tortuous one refers to the waveform without wavelet filtering and Hampel filtering, while the smoother one refers to the waveform obtained after wavelet filtering and Hampel filtering.

[0091] 207. The system performs wavelet filtering and Hampel filtering on the relational waveform to obtain the target waveform.

[0092] Target waveform as Figure 6As shown in the diagram, a relatively smooth waveform is obtained by removing outliers through wavelet filtering and then smoothing it through Hampel filtering. Thus, a smoother target waveform is obtained through wavelet filtering and Hampel filtering.

[0093] In this embodiment, the method of recording, filtering, cleaning, and then analyzing the data cleverly solves the problem of unnecessary actions during the user's writing process interfering with writing speed, thereby further improving the accuracy and effectiveness of the evaluation results.

[0094] 208. The system determines the user's writing speed based on the target waveform.

[0095] The writing time for each cell and each line can be obtained from the target waveform. The writing time for each cell can be used to calculate the user's writing speed within that cell. The exercise book contains several writing cells, each accommodating one character.

[0096] Optionally, in one feasible approach, the system determines the target time point of the pen tip in the target writing grid based on the target waveform.

[0097] The system calculates the user's writing speed in the target writing grid based on the target time point.

[0098] Please see Figure 3 , Figure 3 A schematic diagram of a handwriting ability assessment system based on key point detection provided in this application, the handwriting ability assessment system comprising:

[0099] The acquisition unit 301 is used to acquire video of the writing area in real time through a camera device. The writing area includes a notebook and a writing pen, and a virtual coordinate system is established for the notebook in the video.

[0100] The processing unit 302 is used to detect the pen tip of the writing pen through the CenterNet key point detection algorithm and track the movement trajectory of the pen tip on the exercise book through the deepsort algorithm when the user writes on the exercise book with the writing pen, so as to obtain trajectory information.

[0101] The generation unit 303 is used to generate writing data based on trajectory information and coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to the coordinate points.

[0102] The analysis unit 304 is used to analyze the writing data when the user finishes writing and obtain the analysis results;

[0103] Unit 305 is used to determine the user's writing speed based on the analysis results.

[0104] In this embodiment, the acquisition unit 301 first acquires a video of the writing area in real time through a camera device and establishes a virtual coordinate system for the exercise book in the video. Then, when the user writes on the exercise book with a writing pen, the processing unit 302 detects the pen tip using the CenterNet keypoint detection algorithm and tracks the pen tip's movement trajectory on the exercise book using the deepsort algorithm, thus obtaining trajectory information. Next, the generation unit 303 generates writing data based on the trajectory information and the virtual coordinate system. This writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to those coordinate points. Then, when the user finishes writing, the analysis unit 304 analyzes the writing data, obtaining the analysis results. Finally, the determination unit 305 determines the user's writing speed based on these analysis results. This eliminates the need for a smart writing pen or smart writing tablet for writing ability assessment, reducing the influence of testing tools on the child's writing assessment process. It can simulate the child's writing environment to a greater extent, thereby improving the authenticity and effectiveness of the assessment results.

[0105] Please continue reading. Figure 4 , Figure 4 Another schematic diagram of a handwriting ability assessment system based on key point detection provided in this application, the handwriting ability assessment system includes:

[0106] The control unit 401 is used to control the camera device to aim at the writing area and to determine the position of the exercise book by clicking on the four corners of the exercise book;

[0107] The acquisition unit 402 is used to acquire video of the writing area in real time through a camera device. The writing area includes a notebook and a writing pen, and a virtual coordinate system is established for the notebook in the video.

[0108] Transformation unit 403 is used to perform perspective transformation on the side view of the exercise book in the video to obtain the top view of the exercise book;

[0109] The processing unit 404 is used to detect the pen tip of the writing pen through the CenterNet key point detection algorithm and track the movement trajectory of the pen tip on the exercise book through the deepsort algorithm when the user writes on the exercise book with the writing pen, so as to obtain trajectory information.

[0110] The generation unit 405 is used to generate writing data based on trajectory information and coordinate system. The writing data includes several coordinate points of the pen tip on the exercise book and several time points corresponding to the coordinate points.

[0111] Analysis unit 406 is used to analyze the writing data when the user finishes writing and obtain the analysis results;

[0112] Analysis unit 406 includes:

[0113] Analysis module 4061 is used to analyze the written data through analysis algorithms to obtain analysis results. The analysis algorithms include wavelet filtering and Hampel filtering.

[0114] Analysis module 4061 includes:

[0115] The first determining submodule 40611 is used to determine the waveform of the relationship between the change of the X-axis coordinate and the change at time points based on the written data;

[0116] Processing submodule 40612 is used to perform wavelet filtering and Hampel filtering on the relational waveform to obtain the target waveform;

[0117] Unit 407 is used to determine the user's writing speed based on the analysis results;

[0118] Determining unit 407 includes:

[0119] The second determining module 4071 is used to determine the user's writing speed based on the target waveform;

[0120] The second determining module 4071 includes:

[0121] The determination submodule 40711 is used to determine the target time point of the pen tip in the target writing grid based on the target waveform. The exercise book contains several writing grids, and each writing grid contains one character.

[0122] The calculation submodule 40712 is used to calculate the user's writing speed in the target writing grid based on the target time point.

[0123] Please see Figure 5 , Figure 5 This is a schematic diagram of a writing ability assessment device based on key point detection provided in this application. The writing ability assessment device includes:

[0124] Central processing unit 502, memory 501, input / output interface 503, wired or wireless network interface 504, and power supply 505;

[0125] Memory 501 is either a short-term storage memory or a persistent storage memory;

[0126] The central processing unit 502 is configured to communicate with the memory 501 and execute instructions stored in the memory 501 to perform the aforementioned operations. Figures 1 to 2 The steps in the illustrated embodiment.

[0127] This application provides a computer-readable storage medium, including instructions that, when executed on a computer, cause the computer to perform the aforementioned... Figures 1 to 2 The steps in the illustrated embodiment.

[0128] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0129] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0130] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0131] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0132] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for evaluating writing ability based on key point detection, characterized in that, The method comprises the following steps: real-time acquisition of a video of a writing area by a camera device, the writing area containing a practice book and a writing pen, and a virtual coordinate system is established for the practice book in the video; when a user writes on the practice book using the writing pen, the tip of the writing pen is detected by a CenterNet key point detection algorithm, and the motion trajectory of the tip on the practice book is tracked by a deepsort algorithm to obtain trajectory information; writing data is generated according to the trajectory information and the coordinate system, the writing data including a plurality of coordinate points of the tip on the practice book and a plurality of time points corresponding to the plurality of coordinate points; when the user finishes writing, the writing data is analyzed to obtain an analysis result; the writing speed of the user is determined according to the analysis result; after the real-time acquisition of the video of the writing area by the camera device and before the virtual coordinate system is established for the practice book in the video, the method further comprises the following steps: perspective transformation is performed on a side view of the practice book in the video to obtain a top view of the practice book; the analysis of the writing data comprises the following steps: the writing data is analyzed by an analysis algorithm to obtain an analysis result, the analysis algorithm including wavelet filtering and hampel filtering; the analysis of the writing data by the analysis algorithm comprises the following steps: a relationship waveform between X-axis coordinate change and time point change is determined according to the writing data; wavelet filtering and hampel filtering are performed on the relationship waveform to obtain a target waveform; the determination of the writing speed of the user according to the analysis result comprises the following steps: the writing speed of the user is determined according to the target waveform.

2. The method of claim 1, wherein the determination of the writing speed of the user according to the target waveform comprises the following steps: a target time point of the tip in a target writing grid is determined according to the target waveform, the practice book containing a plurality of writing grids, one writing grid accommodating one character; the writing speed of the user in the target writing grid is calculated according to the target time point.

3. The writing ability assessment method according to claim 1 or 2, characterized by, before the real-time acquisition of the video of the writing area by the camera device, the method further comprises the following steps: the camera device is controlled to be aligned with the writing area, and the position of the practice book is determined by clicking on the four corners of the practice book.

4. A keypoint detection based writing ability assessment system, characterized in that, The method comprises the following steps: an acquisition unit is configured to acquire a video of a writing area by a camera device, the writing area containing a practice book and a writing pen, and a virtual coordinate system is established for the practice book in the video; a transformation unit is configured to perform perspective transformation on a side view of the practice book in the video to obtain a top view of the practice book; a processing unit is configured to, when a user writes on the practice book using the writing pen, detect the tip of the writing pen by a CenterNet key point detection algorithm, and track the motion trajectory of the tip on the practice book by a deepsort algorithm to obtain trajectory information; The generating unit is configured to generate writing data according to the trajectory information and the coordinate system, the writing data comprising a plurality of coordinate points of the pen nib on the exercise book and a plurality of time points corresponding to the plurality of coordinate points; The analyzing unit is configured to analyze the writing data when the user finishes writing, and obtain an analysis result; The determining unit is configured to determine a writing speed of the user according to the analysis result; The analyzing unit comprises: An analysis module configured to analyze the writing data by an analysis algorithm to obtain the analysis result, the analysis algorithm comprising wavelet filtering and hampel filtering; The analysis module comprises: A first determining module configured to determine a relationship waveform between an X-axis coordinate change and a time point change according to the writing data; A processing submodule configured to perform wavelet filtering and hampel filtering on the relationship waveform to obtain a target waveform; The determining unit comprises: A second determining module configured to determine the writing speed of the user according to the target waveform.

5. A keypoint detection based writing ability assessment device, characterized in that, The computer readable storage medium comprises instructions, when the instructions are executed on the computer, the computer executes the method of any one of claims 1 to 3.

6. A computer readable storage medium comprising instructions which, when executed on a computer, cause the computer to carry out the method of any one of claims 1 to 3. ​ ​ ​

Images