AI-based Intangible Cultural Heritage Inheritance and Display Method and System

Abstract

The present invention discloses a method and system for the inheritance and display of intangible cultural heritage based on artificial intelligence, specifically related to the field of artificial intelligence technology. By obtaining dynamic detail time series data in real time from a capture system, analyzing the trajectory smoothness using Fourier transform, generating a lossless reference model using an ultra-high-resolution 3D scanning and imaging device, and combining point cloud registration and differential heat map to evaluate the spatial deviation of the trajectory, performing texture consistency analysis on significantly different regions to judge the fidelity of dynamic details, comprehensively evaluating the accuracy of capture by combining smoothness and fidelity, classifying the capture results into two categories: accurate capture and inaccurate capture, archiving the accurately captured data for efficient reuse; for inaccurately captured data, dynamically complementing the trajectory through an autoregressive integrated moving average model, significantly improving the accuracy and integrity of the capture of intangible cultural heritage dynamic details, ensuring that the display effect is real and coherent, and contributing to the accurate inheritance and wide dissemination of the cultural connotation of intangible cultural heritage.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and particularly to a method and system for inheriting and displaying intangible cultural heritage based on artificial intelligence. Background Art

[0002] The inheritance and display of intangible cultural heritage based on artificial intelligence refers to using artificial intelligence technology to digitally and intelligently display and spread the content, form, and cultural connotations of intangible cultural heritage (ICH), so as to promote the protection and inheritance of ICH. Through AI technologies such as machine learning, computer vision, and natural language processing, this method reconstructs the forms of ICH to make them more vivid, modern, and interactive. For example, through generated digital models, voice synthesis, and virtual reality (VR) technology, users can feel the charm of traditional handicrafts, dances, or music as if they were on the scene. This display method not only improves the dissemination efficiency of ICH culture but also reduces geographical and time limitations, enabling more people to access and understand ICH culture.

[0003] The existing technologies have the following deficiencies:

[0004] ICH projects often have complex dynamic characteristics, such as the delicate movements of dances or the fine details of the process of handicraft production. When performing 3D modeling and dynamic rendering, if the algorithm fails to accurately capture these details, it will lead to a lack of authenticity in the virtual experience and lose the meaning of ICH display. Especially during the real-time interaction process, the distortion of movements or changes will cause a sense of disconnection between the user and the content. In addition, the value of ICH projects lies in their cultural authenticity and uniqueness. For example, each movement in a dance may contain specific cultural meanings, and the production steps of handicrafts reflect the essence of traditional skills. If the details are distorted, what the user encounters is an incomplete or incorrect form of expression, which may lead to a misunderstanding of the cultural connotations of ICH and even result in incorrect cultural inheritance. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for inheriting and displaying intangible cultural heritage based on artificial intelligence to solve the deficiencies in the background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions: A method for inheriting and displaying intangible cultural heritage based on artificial intelligence, including the following steps:

[0007] S1: Real-time obtain time series data of dynamic details from a capture system, where the time series data includes trajectory data of dance movements in ICH handicraft production;

[0008] S2: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components;

[0009] S3: When the smoothness of the dance movement trajectory in space is inconsistent, model the real scene through an ultra-high-resolution 3D scanning device and imaging device to generate a lossless reference model. Through the point cloud registration algorithm, compare the difference between the trajectory data of the dance movement and the data of the reference model, and map the difference value to a heat map;

[0010] S4: Divide the heat map into regions with significant differences and regions with insignificant differences. For regions with significant differences, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the regions with significant differences, and judge the fidelity of the dynamic details;

[0011] S5: Evaluate the accuracy of dynamic detail capture according to the smoothness consistency of the dance movement trajectory in space and the fidelity of the dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture;

[0012] S6: Archive the accurately captured dynamic detail data for repeated use in different intangible cultural heritage display projects; for inaccurate capture, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectories.

[0013] Preferably, in S2, use Fourier transform to decompose the trajectory data into frequency components and calculate the proportion of high-frequency noise components, specifically:

[0014] Obtain the time series trajectory data of the dance movement from the capture system, apply the discrete Fourier transform to the trajectory data sequence, and set the frequency threshold , the low-frequency component is , the high-frequency component is , calculate the high-frequency energy and the total spectral energy , the expression is: ; where X(f) is the complex value of frequency f, f represents the frequency component, N is the length of the time series, and the high-frequency noise ratio is: .

[0015] Preferably, in S3, judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the high-frequency noise component ratio, specifically:

[0016] After analyzing the abnormal fluctuation of the high-frequency noise component ratio, generate a high-frequency noise ratio fluctuation index. The acquisition method of the high-frequency noise ratio fluctuation index is: the high-frequency noise ratio time series , where t = 1, 2,..., Q represents the time point, is the high-frequency noise ratio at the t-th moment. Set the sliding window length W, which represents the length of the time series for calculating the standard deviation each time, and the sliding step Δt, which represents the time interval for each movement of the sliding window. Extract multiple sliding windows from the time series: ; where: is the high-frequency noise ratio at the time point in the time series, , representing the starting time point of the k-th window. For each sliding window , calculate the local standard deviation , and the expression is: ; where, is the mean value within the window. Calculate the standard deviation of all windows in sequence to obtain the volatility time series: ; M is the total number of windows. The high-frequency noise ratio fluctuation index is defined as the mean value of the standard deviations of all sliding windows, and the expression is: ; in the formula, is the high-frequency noise ratio fluctuation index.

[0017] Preferably, in S3, through the point cloud registration algorithm, compare the difference between the trajectory data of the dance movement and the data of the reference model, and map the difference value to a heat map, specifically:

[0018] Use a motion capture system to obtain the trajectory point cloud of the dance movement: Pmotion={q1,q2,...,qi}, where qi=(xi,yi,zi) is the three-dimensional coordinate of the trajectory point. Align the dance movement trajectory point cloud Pmotion to the reference model point cloud Pbaseline, and the expression is: ; where, R is the rotation matrix and t is the translation vector, represents the result after initial transformation of the dance movement trajectory point cloud Pmotion through the rotation matrix R and the translation vector t, that is, the new position after performing rotation and translation operations on Pmotion;

[0019] Use the point-to-point ICP algorithm for fine registration: matching point pairs: ; find the closest point pair between the trajectory point and the reference point , and optimize the rigid transformation: ; represents the new coordinates obtained after applying the rigid transformation to the trajectory point qi, and m is the number of matching point pairs;

[0020] The registered trajectory point cloud , and the expression is: ; in the formula, Denote the optimal rigid transformation matrix, which is a rotation matrix obtained through a minimization process and is used to transform the trajectory point cloud of the dance movement to a position aligned with the reference model; for each point , calculate its Euclidean distance to the nearest reference point : , where represents the difference between the trajectory point and the reference model

[0021] Map the difference value di of each point to a color value Ci to construct a heat map: Ci = f(di); where f(di) is a color mapping function, and apply the generated color value Ci to the trajectory point cloud Paligned to generate a colored point cloud model

[0022] Preferably, in S4, divide the heat map into a significantly different region and a non-significantly different region, specifically:

[0023] Set the high difference threshold as δthreshold and the low difference threshold as δlow, and mark the points whose difference values exceed the threshold; divide the difference values into three levels and distinguish them by colors: low difference green: di ≤ δlow; medium difference yellow: δlow < di ≤ δthreshold; high difference red: di > δthreshold

[0024] Conduct spatial clustering analysis on the difference points, mark the spatial regions where the high difference points are located, and the clustering result is expressed as: C = {C1, C2,..., Ck}; where Ck is a significantly different region

[0025] Preferably, in S4, for the significantly different region, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the significantly different region, and judge the fidelity of the dynamic details, specifically:

[0026] Map each trajectory point qi of the dance movement to the surface of the reference model to obtain the surface texture data Tmotion(qi) of the dance movement, extract the texture map Tbaseline(sp) from the reference model, and map it to the corresponding 3D model surface according to the coordinates of the surface point cloud. The expression is: Tbaseline(sp) = Texture Mapping(Pbaseline, sp); where sp is a certain surface point in the reference model, and Tbaseline(sp) is the corresponding texture map data

[0027] For each pair of corresponding points qi and sp in the significantly different region, calculate the similarity of their texture images , and the expression is: ; where represents the texture image of the dance movement Represents the reference model texture image, is the image mean of local luminance, is the image mean of local luminance, is the image standard deviation of local luminance, representing the contrast of the image, is the image standard deviation of local luminance, representing the contrast of the image, represents the image and the local covariance of the image and are constants; calculate the texture similarity of each point in the significantly different region to obtain the consistency index of regional texture , and the expression is: ; where represents the significantly different region, is the number of points within the region;

[0028] The dynamic detail fidelity refers to the consistency and true restoration degree of the texture of the dance movement and the texture of the reference model within the significantly different region, and the calculation expression is: ; in the formula, is the dynamic detail fidelity, is the mean square error, and the calculation expression is: ; in the formula, B is the number of comparison points.

[0029] Preferably, in S5, evaluate the accuracy of dynamic detail capture according to the smoothness consistency of the dance movement trajectory in space and the fidelity of dynamic details;

[0030] Convert the high-frequency noise ratio fluctuation index and the dynamic detail fidelity into a comprehensive feature vector, use the comprehensive feature vector as the input of the machine learning model, take the prediction of the accuracy value label of dynamic detail capture for each group of comprehensive feature vectors by the machine learning model as the prediction target, and take minimizing the sum of prediction errors of all accuracy value labels of dynamic detail capture as the training target, train the machine learning model until the sum of prediction errors reaches convergence and then stop the model training, and determine the accuracy value of dynamic detail capture according to the model output result, where the machine learning model is a polynomial regression model.

[0031] Preferably, according to the evaluation results, the accuracy of dynamic detail capture is divided into two categories: accurate capture and inaccurate capture. Specifically: compare the obtained accuracy value of dynamic detail capture with the reference threshold of the accuracy value of dynamic detail capture preset according to historical data. If the accuracy value of dynamic detail capture is greater than or equal to the reference threshold of the accuracy value, it indicates that the accuracy of dynamic detail capture is high. At this time, no warning signal is generated, and the accuracy of dynamic detail capture is classified as accurate capture; if the accuracy value of dynamic detail capture is less than the reference threshold of the accuracy value, it indicates that the accuracy of dynamic detail capture is low. At this time, a warning signal is generated, and the accuracy of dynamic detail capture is classified as inaccurate capture.

[0032] Preferably, in S6, for inaccurate capture, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectories. Specifically:

[0033] Obtain the trajectory time series from the capture system: ; where t = 1, 2,..., D, t is the time frame, is the three-dimensional coordinate of the t-th frame, marking the time frame of inaccurate capture; ; Split the complete trajectory, where the known point sequence: ; Missing point sequence: ; Autoregressive is obtained by linearly regressing the current value from the values of the previous few time frames: ; In the formula, is the trajectory value of the current time frame, is the autoregressive coefficient, used to represent the contribution of the previous p trajectory values, is the trajectory value of the time frame, is the error term; Moving average corrects the current value through the residuals of the previous few time frames: ; is the moving average coefficient, used to correct the predicted value, is the error value at time ; The overall ARIMA model is expressed as: ARIMA(p, d, q); where: p is the autoregressive order, indicating how many previous values are used, d is the differencing order, indicating the number of differencing operations, and q is the moving average order, indicating how many residual values are used;

[0034] For non-stationary sequences, perform d differencing operations to obtain a stationary sequence , and the expression is: ; Generate the initial prediction input data: ; Train the model for each trajectory dimension separately: ; Optimize the model parameters ϕ, θ through maximum likelihood estimation, and based on the trained model, predict the values of the missing points: ; For each missing point Make a point-by-point prediction, replace the missing points in the original data with the predicted values, and the expression is: , represents the finally generated point cloud, is the newly added trajectory point, which represents the points supplemented during the dynamic detail capture process.

[0035] The present invention also provides a non-genetic inheritance display system based on artificial intelligence, including a dynamic detail capture module, a trajectory smoothness analysis module, a difference analysis module, a dynamic detail fidelity analysis module, a dynamic detail capture accuracy evaluation module, and a data archiving and display module;

[0036] Dynamic detail capture module: Obtain the time series data of dynamic details from the capture system in real time. The time series data includes the trajectory data of dance movements in the non-genetic craft production;

[0037] Trajectory smoothness analysis module: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components;

[0038] Difference analysis module: When the smoothness of the dance movement trajectory in space is inconsistent, model the real scene through an ultra-high-resolution 3D scanning device and an imaging device to generate a lossless reference model. Through the point cloud registration algorithm, compare the data differences between the trajectory data of the dance movement and the reference model, and map the difference values to a heat map;

[0039] Dynamic detail fidelity analysis module: Divide the heat map into regions with significant differences and regions with insignificant differences. For regions with significant differences, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the regions with significant differences, and judge the fidelity of the dynamic details;

[0040] Dynamic detail capture accuracy evaluation module: Evaluate the accuracy of dynamic detail capture according to the consistency of the smoothness of the dance movement trajectory in space and the fidelity of the dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture;

[0041] Data archiving and display module: Archive the accurately captured dynamic detail data for repeated use in different projects of non-genetic display; for inaccurate capture, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectories.

[0042] In the above technical solution, the technical effects and advantages provided by the present invention:

[0043] 1. The present invention captures the trajectory data of intangible cultural heritage dances in real time and analyzes the smoothness using Fourier transform. Further, for trajectories with inconsistent smoothness, a reference model is generated through ultra-high-resolution 3D scanning, and a point cloud registration algorithm is used to locate the differences and generate a difference heat map. The texture consistency calculation and dynamic detail fidelity evaluation of significantly different regions ensure the authenticity of the capture results and the restoration of cultural connotations. By comprehensively analyzing the trajectory smoothness and fidelity using a polynomial regression model, the accuracy of dynamic detail capture is accurately evaluated, thereby achieving accurate data archiving and intelligent completion of inaccurate data, significantly improving the overall quality of dynamic detail capture and display.

[0044] 2. While ensuring high-fidelity dynamic details of intangible cultural heritage projects, the present invention realizes the systematic storage and efficient reuse of captured data, meeting the needs of intangible cultural heritage display in multiple scenarios. Moreover, the present invention repairs inaccurate captured data through intelligent completion technology, making it more coherent and complete, providing technical support for the protection and inheritance of intangible cultural heritage. At the same time, the accuracy classification and warning mechanism for dynamic detail capture enhance the intelligence and reliability of the system, providing an important guarantee for the virtualization and digital application of intangible cultural heritage display, and promoting the innovative development of intangible cultural heritage projects in the field of digital technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0046] Figure 1 It is a flowchart of the method of the present invention.

[0047] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0049] Example 1. Please refer to Figure 1 and Figure 2 As shown, the method for intangible cultural heritage inheritance display based on artificial intelligence in this embodiment includes the following steps:

[0050] S1: Obtain the time series data of dynamic details in real time from the capture system, where the time series data includes the trajectory data of dance movements in the intangible cultural heritage craft production;

[0051] S2: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components;

[0052] S3: When the smoothness of the dance movement trajectory in space is inconsistent, model the real scene through an ultra-high-resolution 3D scanning device and imaging device to generate a lossless reference model. Through the point cloud registration algorithm, compare the data differences between the trajectory data of the dance movement and the reference model, and map the difference values into a heat map;

[0053] S4: Divide the heat map into regions with significant differences and regions with insignificant differences. For regions with significant differences, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the regions with significant differences, and judge the fidelity of the dynamic details;

[0054] S5: Evaluate the accuracy of dynamic detail capture according to the consistency of the smoothness of the dance movement trajectory in space and the fidelity of the dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture;

[0055] S6: Archive the accurately captured dynamic detail data for repeated use in different intangible cultural heritage display projects; for inaccurate capture, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectories.

[0056] In S1, obtaining the time series data of dynamic details in real time from the capture system, where the time series data includes the trajectory data of dance movements in the intangible cultural heritage craft production, specifically:

[0057] The capture system needs to meet the requirements of high precision, real-time performance, and adaptability to complex dynamic characteristics. Common systems include: Optical motion capture system: such as Vicon, OptiTrack. By arranging reflective marker points (such as reflective balls) on the dancer's body and using multiple high-speed cameras to capture the three-dimensional coordinate changes of these marker points. It is suitable for high-precision dance motion capture. Inertial motion capture system: such as Xsens, Perception Neuron. By using inertial sensors to capture the limb movement data of the dancer, it does not rely on external cameras and is suitable for scenarios with lower space requirements. Depth camera and RGB camera: such as Microsoft Kinect, Intel RealSense. Using depth sensors and image processing algorithms to directly capture human skeleton data, it is suitable for low-cost real-time capture applications.

[0058] Determine the key parts of the dance movement (such as joints, fingertips, toes) and arrange marker points at these parts. The more marker points, the more detailed the captured trajectory data. For markerless capture systems, ensure that the field of view of the device can completely cover the movement area of the dancer. Surround and arrange multiple cameras to form a three-dimensional capture space covering the stage. Calibrate the position and angle of the cameras to ensure that there are no blind spots in the capture area. Use calibration tools (such as calibration rods or calibration plates) to calibrate the capture system to ensure the consistency of the coordinate systems of each camera.

[0059] The dancer performs according to the established movements, and the capture system records the three-dimensional coordinates or skeleton data of each marker point in real time. The data format is usually a time series, and each frame contains the position data of all marker points. Example: ; where represents the timestamp, is the three-dimensional coordinate of the nth marker point. Convert the time series data into trajectory data and use interpolation algorithms to supplement the missing parts of the data. Generate trajectory curves for visually displaying the movement paths of dance movements.

[0060] Preprocess the data, use filtering algorithms (such as Kalman filtering) to remove high-frequency noise and ensure smooth trajectories. Convert the captured coordinates into a unified reference coordinate system for subsequent analysis and modeling. Through data transfer protocols (such as UDP or WebSocket), transfer the captured time series data to the backend server or rendering engine in real time. Ensure low latency in data transfer (such as within 5 milliseconds) to support real-time dynamic rendering and analysis.

[0061] Verify the accuracy of the capture system for key movements (such as complex limb changes, rapid movement switches). Check whether there are obvious breakpoints or missing parts in the time series data to ensure continuous trajectories. By comparing the captured data with the benchmark model of the dance movement, calculate the position error and trajectory deviation, and adjust the parameters of the capture device.

[0062] The trajectory data of dance movements is output in real time and provided to subsequent analysis and rendering modules for dynamic modeling, motion evaluation, and cultural display. Data files are usually stored in standard formats (such as BVH, FBX, or JSON) for easy cross-platform use.

[0063] S2: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuations of the proportion of high-frequency noise components.

[0064] Obtain the time series trajectory data of dance movements from the capture system, represented as the trajectory of key points in three-dimensional space, for example: ; where t is the time frame and n is the number of key points. Denoise the trajectory data, for example, using moving average or Kalman filter, to eliminate outliers and invalid frames. Convert the trajectory into a single-dimensional data sequence in each direction for easy Fourier transform processing.

[0065] Apply the discrete Fourier transform (DFT) to the trajectory data sequence: ; where f represents the frequency component, N is the length of the time series, X(f) is the complex value at frequency f, is the kernel function of the Fourier transform, calculate the amplitude (modulus value) of the spectrum , and the expression is: ; in the formula, are the real and imaginary parts of the spectrum amplitude respectively, and draw the spectrum curve to show the energy distribution at different frequencies.

[0066] Set the frequency threshold according to the sampling rate and the characteristics of the captured trajectory , low-frequency components: , representing the main motion characteristics of the motion trajectory. High-frequency components: , usually representing the noise and rapid jitter in the trajectory data. Calculate the proportion of high-frequency noise, calculate the high-frequency energy and the total spectrum energy respectively, and the expression is: ; in the formula, X(f) is the complex value at frequency f, f represents the frequency component, N is the length of the time series, and the proportion of high-frequency noise is: . The higher the , the more noise components in the trajectory and the worse the trajectory smoothness.

[0067] After analyzing the abnormal fluctuations of the proportion of high-frequency noise components, generate the high-frequency noise proportion fluctuation index. The method for obtaining the high-frequency noise proportion fluctuation index is:

[0068] After analyzing the abnormal fluctuations in the proportion of high-frequency noise components, a high-frequency noise proportion fluctuation index is generated. The method for obtaining the high-frequency noise proportion fluctuation index is as follows: the high-frequency noise proportion time series , where t = 1, 2,..., Q represents time points, is the high-frequency noise proportion at the t-th moment. Set the sliding window length W, which represents the length of the time series for calculating the standard deviation each time, and the sliding step Δt, which represents the time interval for each movement of the sliding window. Extract multiple sliding windows from the time series: ; where: is the high-frequency noise proportion at the time point in the time series, , representing the starting time point of the k-th window. For each sliding window , calculate the local standard deviation , and the expression is: ; where is the mean value within the window. Calculate the standard deviations of all windows in sequence to obtain the volatility time series: ; M is the total number of windows. The high-frequency noise proportion fluctuation index is defined as the mean value of the standard deviations of all sliding windows, and the expression is: ; in the formula, is the high-frequency noise proportion fluctuation index.

[0069] The larger the high-frequency noise proportion fluctuation index, the more significant the fluctuation amplitude of the high-frequency noise proportion, the greater the change in the smoothness of the trajectory in different time periods, and there are sudden or irregular trajectory perturbations. This indicates that the smoothness of the dance movement trajectory in space is inconsistent, and obvious jitters, jumps, or abnormal curves of the trajectory may occur in some time periods. For the display of intangible cultural heritage dances, such trajectories will weaken the coherence of the movements and the accurate expression of the cultural connotations.

[0070] On the contrary, when the high-frequency noise proportion fluctuation index is smaller, it indicates that the change amplitude of the high-frequency noise proportion is lower, the smoothness of the trajectory in the time series tends to be consistent, and the overall movement trajectory shows stable and coherent dynamic characteristics. This indicates that the smoothness of the dance movement trajectory in space is relatively high and consistent. Such data can better restore the cultural connotations of the dance movements and provide high-quality dynamic detail support for the display of intangible cultural heritage.

[0071] S3: When the smoothness of the dance movement trajectory in space is inconsistent, use an ultra-high-resolution 3D scanning device and imaging device to model the real scene, generate a lossless reference model, and through the point cloud registration algorithm, compare the difference between the trajectory data of the dance movement and the data of the reference model, and map the difference value to a heat map.

[0072] When the smoothness of the dance movement trajectory in space is inconsistent, use a structured light scanner, a laser scanner, or a multi-view photogrammetry device to capture high-resolution 3D data of the dancer. Use a high-frame-rate camera to synchronously record the video of the dance movement and capture dynamic texture data.

[0073] Apply a point cloud filtering algorithm (such as voxel grid filtering) to remove noise points. Ensure that the coordinate system of the reference model is consistent with the coordinate system of the dance movement capture for subsequent registration. Save the generated point cloud in a standard format (such as PLY or OBJ).

[0074] Use a motion capture system to obtain the trajectory point cloud of the dance movement: Pmotion={q1,q2,...,qi}, where qi=(xi,yi,zi) is the three-dimensional coordinate of the trajectory point. Filter and time-synchronize the trajectory data to make its sampling frequency consistent with that of the reference model.

[0075] Align the dance movement trajectory point cloud Pmotion to the reference model point cloud Pbaseline to compare the differences. Use a rigid transformation algorithm (such as ICP initial alignment) to perform a rough alignment of the point cloud: ; where R is the rotation matrix and t is the translation vector, represents the result of the initial transformation of the dance movement trajectory point cloud Pmotion through the rotation matrix R and the translation vector t, that is, the new position after performing rotation and translation operations on Pmotion.

[0076] Use the point-to-point ICP (Iterative Closest Point) algorithm for fine registration: matching point pairs: ; find the trajectory point to the reference point of the closest point pair. Optimize the rigid transformation: ; represents the new coordinates obtained after applying the rigid transformation to the trajectory point qi, and m is the number of matching point pairs.

[0077] The registered trajectory point cloud , the expression is: ; in the formula, represents the optimal rigid transformation matrix, the rotation matrix calculated through the minimization process, used to transform the dance movement trajectory point cloud to the position aligned with the reference model; for each point , calculate its Euclidean distance to the closest reference point : , in the formula, represents the difference between this trajectory point and the reference model.

[0078] Map the difference value di of each point to a color value Ci to construct a heat map: Ci = f(di); where f(di) is a color mapping function, and common color gradients (such as blue-green-red) represent the magnitude of differences: blue: small difference. Red: large difference. Apply the color value Ci to the trajectory point cloud Paligned to generate a colored point cloud model.

[0079] S4: Divide the heat map into regions with significant differences and regions with insignificant differences. For regions with significant differences, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the regions with significant differences, and judge the fidelity of the dynamic details.

[0080] Observe the difference heat map and locate the regions where the trajectory differs significantly from the reference model. For regions with larger difference values, further analyze the accuracy of the captured data or recalibrate the reference model. The difference heat map is output in 3D visualization form, which is convenient for dynamic visualization in the display of intangible cultural heritage. At the same time, save the difference statistical data (such as mean, standard deviation) for subsequent optimization analysis.

[0081] Set the high difference threshold as and mark the points whose difference values exceed the threshold: It can be set in the following ways:

[0082] Such as the mean μd plus k times the standard deviation σd: ; Usually k takes 2 or 3, or a fixed value is selected according to specific application requirements.

[0083] Divide the difference values into multiple levels (such as low, medium, and high differences) and distinguish them by colors: low difference (green): di ≤ δlow; medium difference (yellow): δlow < di ≤ δthreshold; high difference (red): di > δthreshold;

[0084] Perform spatial clustering analysis on the difference points (such as DBSCAN clustering), mark the spatial regions where the high difference points are located, and the clustering result is expressed as: C = {C1, C2,..., Ck}; where Ck is a region with significant differences.

[0085] Map each trajectory point qi of the dance movement to the surface of the reference model to obtain the surface texture data Tmotion(qi) of the dance movement. Extract the texture map Tbaseline(sp) from the reference model and map it to the corresponding 3D model surface according to the coordinates of the surface point cloud. The expression is: Tbaseline(sp) = Texture Mapping(Pbaseline, sp); where sp is a certain surface point in the reference model, and Tbaseline(sp) is the corresponding texture map data;

[0086] For each pair of corresponding points qi and sp in the significantly different region, calculate the similarity of their texture images , and the expression is: ; where represents the texture image of the dance movement, represents the texture image of the reference model, is the mean of the local brightness of the image , is the mean of the local brightness of the image , is the standard deviation of the local brightness of the image , representing the contrast of the image , is the standard deviation of the local brightness of the image , representing the contrast of the image , represents the local covariance of the image and the image , and are constants; calculate the texture similarity of each point in the significantly different region to obtain the consistency index of the regional texture , and the expression is: ; where represents the significantly different region, is the number of points within the region;

[0087] The dynamic detail fidelity refers to the consistency and true restoration degree of the texture of the dance movement and the texture of the reference model within the significantly different region, and the calculation expression is: ; in the formula, is the dynamic detail fidelity, is the mean square error, and the calculation expression is: ; in the formula, B is the number of comparison points.

[0088] High fidelity: If the fidelity index is close to 1, it indicates that the texture of the dance movement is highly consistent with the texture of the reference model within the significantly different region, and the dynamic details are restored with high fidelity.

[0089] Medium fidelity: If the fidelity index is between 0.6 and 1, it indicates that the texture consistency is good, but there are still a few differences, and the dynamic details may be lost in some parts.

[0090] Low fidelity: If the fidelity index is below 0.6, it indicates that there are significant differences between the texture of the dance movement and the reference model, and the dynamic details are severely distorted, affecting the user experience and cultural inheritance.

[0091] S5: Evaluate the accuracy of dynamic detail capture based on the spatial smoothness consistency of the dance movement trajectory and the fidelity of dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture.

[0092] Convert the high-frequency noise ratio fluctuation index and dynamic detail fidelity into a comprehensive feature vector. Use the comprehensive feature vector as the input of a machine learning model. The machine learning model takes predicting the accuracy value label of dynamic detail capture for each group of comprehensive feature vectors as the prediction target, and minimizing the sum of prediction errors for all accuracy value labels of dynamic detail capture as the training target. Train the machine learning model until the sum of prediction errors reaches convergence and then stop the model training. Determine the accuracy value of dynamic detail capture according to the model output result. Among them, the machine learning model is a polynomial regression model.

[0093] The method for obtaining the accuracy value of dynamic detail capture is: obtain the corresponding function expression from the comprehensive feature vector training data of the trained machine learning model: ; In the formula, is the output function of the model, is the high-frequency noise ratio fluctuation index, is the dynamic detail fidelity, is the accuracy value of dynamic detail capture.

[0094] Compare the obtained accuracy value of dynamic detail capture with the reference threshold of the accuracy value of dynamic detail capture preset according to historical data. If the accuracy value of dynamic detail capture is greater than or equal to the reference threshold of the accuracy value, it indicates that the accuracy of dynamic detail capture is high. At this time, no warning signal is generated, and the accuracy of dynamic detail capture is divided into accurate capture; if the accuracy value of dynamic detail capture is less than the reference threshold of the accuracy value, it indicates that the accuracy of dynamic detail capture is low. At this time, a warning signal is generated, and the accuracy of dynamic detail capture is divided into inaccurate capture.

[0095] S6: Archive the accurately captured dynamic detail data for repeated use in different intangible cultural heritage display projects; for inaccurate capture, use an autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectories.

[0096] Classify the dynamic detail data according to the attributes of intangible cultural heritage projects for quick retrieval and application: Intangible cultural heritage category: such as dance, handicraft production, music performance, etc. Action type: such as gestures, dance steps, handicraft process steps, etc. Dynamic characteristics: distinguish stable actions (such as rotation) and complex actions (such as rapid jumps).

[0097] Stored in a unified standardized data format for easy cross-platform and multi-device use: Trajectory data: Use JSON, BVH, or FBX format to record time series trajectories and skeletal animation information. Point cloud data: Use PLY or OBJ format to store high-resolution point cloud models. Texture data: Use PNG or JPEG format to store texture maps, ensuring high-fidelity images.

[0098] Add detailed meta-information to the archived dynamic detail data for easy subsequent retrieval and understanding: Project description: Name of the intangible cultural heritage project. Source of the action and cultural background (such as a specific dance or ceremony). Capture device: Model and parameters of the capture device used (such as frame rate, resolution). Data attributes: Timestamp: Generation time of the data. Data size and precision (such as point cloud density, texture resolution). Quality assessment: High-frequency noise ratio fluctuation index. Dynamic detail fidelity index.

[0099] Establish a database for intangible cultural heritage dynamic details to store archived data categorically: Relational database: Use MySQL or PostgreSQL to store trajectory data and meta-information. Non-relational database: Use MongoDB, which is suitable for storing unstructured data such as point clouds and texture data. Store large-scale point cloud and texture files in a distributed file system (such as HDFS) to improve data reading efficiency.

[0100] Establish an efficient indexing mechanism for quick query and invocation: Establish a primary index by intangible cultural heritage category and action type. Establish auxiliary indexes for the features of dynamic details (such as trajectory complexity, texture area). Regularly check the integrity of the archived data to ensure that the files are not damaged or lost. Verify the consistency between the meta-information and the dynamic detail data.

[0101] Dynamically update the archived data by adding new capture results or optimized versions: Version management: Establish version numbers for different versions of the same action (such as the initial version and the optimized version). Differential storage: Use incremental storage technology to only record the parts that differ from the baseline data. Provide a convenient retrieval tool for the archived dynamic detail data: Keyword-based search: Such as "dance", "handicraft steps", etc. Feature-based search: Filter data according to trajectory smoothness or action complexity.

[0102] Develop standardized APIs to support the invocation of dynamic detail data in different scenarios: Educational platform: Provide detailed action breakdowns and demonstrations for intangible cultural heritage teaching. Virtual display: Provide high-resolution dynamic data for AR / VR applications. Game or animation production: Provide skeletal animation data for developers to quickly invoke.

[0103] For inaccurate captures, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details based on the captured front and back trajectories. Specifically:

[0104] Obtain the trajectory time series from the capture system ; where \(t = 1, 2, \cdots, D\), \(t\) is the time frame, is the three-dimensional coordinate of the \(t\)-th frame, marking the time frame of inaccurate capture; ; Split the complete trajectory. Among them, the known point sequence: ; Missing point sequence: ; The autoregression is obtained by linearly regressing the current value from the values of the previous few time frames: ; In the formula, is the trajectory value of the current time frame, is the autoregressive coefficient, used to represent the contribution of the previous \(p\) trajectory values, is the trajectory value of the \(t - i\) time frame, is the error term; The moving average corrects the current value by the residuals of the previous few time frames: ; is the moving average coefficient, used to correct the predicted value, is the error value at time ; The overall ARIMA model is expressed as: ARIMA(\(p,d,q\)); where: \(p\) is the autoregressive order, indicating how many previous values are used, \(d\) is the differencing order, indicating the number of differencing operations, and \(q\) is the moving average order, indicating how many residual values are used;

[0105] For non-stationary sequences, perform \(d\) differencing operations to obtain a stationary sequence , and the expression is: ; Generate the initial prediction input data: ; Train the model for each trajectory dimension separately: ; Optimize the model parameters \(\phi\), \(\theta\) through maximum likelihood estimation. Based on the trained model, predict the values of the missing points: ; For each missing point predict point by point, and replace the missing points in the original data with the predicted values. The expression is: , represents the finally generated point cloud, is the newly added trajectory point, indicating the points supplemented during the dynamic detail capture. Save the completed trajectory in a standardized format (such as JSON, BVH). Used for the dynamic detail reproduction of intangible cultural heritage exhibitions. Provide complete data support for intangible cultural heritage teaching and research.

[0106] In this embodiment, after the trajectory data of dance movements in the intangible cultural heritage craftsmanship is obtained in real time through the capture system, Fourier transform is used to decompose the trajectory data into frequency components, the proportion of high-frequency noise components is calculated, and whether the smoothness of the trajectory in space is consistent is judged through abnormal fluctuation analysis; if the trajectory smoothness is inconsistent, a lossless reference model is generated by a super-high-resolution 3D scanning device, the differences between the trajectory data and the reference model are compared using a point cloud registration algorithm, and the difference values are visualized in the form of a heat map; further, the heat map is divided into significantly different regions and insignificantly different regions, the trajectory data and texture maps are extracted from the significantly different regions, and the texture matching degree is calculated to evaluate the dynamic detail fidelity; combining the trajectory smoothness consistency and fidelity, the accuracy of dynamic detail capture is evaluated, and the capture results are divided into two categories: accurate capture and inaccurate capture; for the accurately captured data, systematic archiving is performed for repeated use, while for the inaccurately captured data, an autoregressive integrated moving average model is used to predict and dynamically complete the missing details through the front and back trajectories, so as to comprehensively improve the dynamic detail quality and reuse value of intangible cultural heritage display.

[0107] Embodiment 2. The intangible cultural heritage inheritance display system based on artificial intelligence described in this embodiment includes a dynamic detail capture module, a trajectory smoothness analysis module, a difference analysis module, a dynamic detail fidelity analysis module, a dynamic detail capture accuracy evaluation module, and a data archiving and display module;

[0108] Dynamic detail capture module: Obtain the time series data of dynamic details in real time from the capture system, and the time series data includes the trajectory data of dance movements in the intangible cultural heritage craftsmanship;

[0109] Trajectory smoothness analysis module: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components;

[0110] Difference analysis module: When the smoothness of the dance movement trajectory in space is inconsistent, model the real scene through a super-high-resolution 3D scanning device and imaging device to generate a lossless reference model, compare the data differences between the trajectory data of the dance movement and the reference model through a point cloud registration algorithm, and map the difference values into a heat map;

[0111] Dynamic detail fidelity analysis module: Divide the heat map into significantly different regions and insignificantly different regions. For the significantly different regions, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the significantly different regions, and judge the fidelity of the dynamic details;

[0112] Dynamic Detail Capture Accuracy Evaluation Module: Evaluate the accuracy of dynamic detail capture based on the spatial smoothness consistency of the dance movement trajectory and the fidelity of dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture;

[0113] Data Archiving and Display Module: Archive the dynamically captured detail data for accurate capture for repeated use in different intangible cultural heritage display projects; for inaccurate capture, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details based on the captured front and rear trajectories.

[0114] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0115] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Among them, A and B can be singular or plural. In addition, the character " / " in this article generally represents an "or" relationship between the front and rear associated objects, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context.

[0116] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0117] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application.

Claims

1. An intangible cultural heritage inheritance and display method based on artificial intelligence, characterized in that: It includes the following steps: S1: Obtain time series data of dynamic details in real time from a capture system, where the time series data includes trajectory data of dance movements in the intangible cultural heritage craftsmanship; S2: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components; Obtain the time series trajectory data of the dance movement from the capture system, apply the discrete Fourier transform to the trajectory data sequence, and set the frequency threshold , the low-frequency component is , the high-frequency component is , calculate the high-frequency energy and the total spectral energy respectively, and the expressions are: ; where X(f) is the complex value of frequency f, f represents the frequency component, N is the time series length, and the high-frequency noise ratio is: ; S3: When the smoothness of the dance movement trajectory in space is inconsistent, model the real scene through an ultra-high-resolution 3D scanning device and imaging device to generate a lossless reference model. Through the point cloud registration algorithm, compare the data differences between the trajectory data of the dance movement and the reference model, and map the difference values to a heat map; S4: Divide the heat map into regions with significant differences and regions with insignificant differences. For regions with significant differences, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the regions with significant differences, and judge the fidelity of the dynamic details; S5: Evaluate the accuracy of dynamic detail capture according to the consistency of the smoothness of the dance movement trajectory in space and the fidelity of the dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture; S6: Archive the accurately captured dynamic detail data for repeated use in different projects of intangible cultural heritage display; for inaccurate capture, use the autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectories.

2. The method for inheriting and displaying intangible cultural heritage based on artificial intelligence according to claim 1, wherein: In S3, judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components, specifically: After analyzing the abnormal fluctuations in the proportion of high-frequency noise components, a high-frequency noise proportion fluctuation index is generated. The method for obtaining the high-frequency noise proportion fluctuation index is as follows: the high-frequency noise proportion time series , where t = 1, 2,..., Q represents time points, is the proportion of high-frequency noise at the t-th moment. Set the sliding window length W, which represents the length of the time series for calculating the standard deviation each time, and the sliding step Δt, which represents the time interval for each movement of the sliding window. Extract multiple sliding windows from the time series: ; where: is the proportion of high-frequency noise at the time point in the time series, , representing the starting time point of the k-th window. For each sliding window , calculate the local standard deviation , and the expression is: ; where, is the mean value within the window. Calculate the standard deviations of all windows in sequence to obtain the volatility time series: ; M is the total number of windows. The high-frequency noise proportion fluctuation index is defined as the mean value of the standard deviations of all sliding windows, and the expression is: ; in the formula, is the high-frequency noise proportion fluctuation index.

3. The method for displaying the inheritance of intangible cultural heritage based on artificial intelligence according to claim 2, wherein: In S3, compare the data differences between the trajectory data of the dance movement and the reference model through the point cloud registration algorithm, and map the difference values to a heat map, specifically: Use a motion capture system to obtain the trajectory point cloud of the dance movement: Pmotion={q1,q2,...,qi}, where qi=(xi,yi,zi) is the three-dimensional coordinate of the trajectory point. Align the dance motion trajectory point cloud Pmotion to the reference model point cloud Pbaseline, and the expression is: ; where R is the rotation matrix and t is the translation vector, represents the result after initial transformation of the dance motion trajectory point cloud Pmotion by the rotation matrix R and the translation vector t, that is, the new position after performing rotation and translation operations on Pmotion; Use the point-to-point ICP algorithm for fine registration: matching point pairs: ; Find the trajectory points to the reference points of the nearest point pairs, and optimize the rigid transformation: ; represents the new coordinates obtained after applying the rigid transformation to the trajectory point qi, and m is the number of matching point pairs; Registered trajectory point cloud , the expression is: ; in the formula, represents the optimal rigid transformation matrix, a rotation matrix obtained through a minimization process, used to transform the trajectory point cloud of the dance movement to a position aligned with the reference model; for each point , calculate its Euclidean distance to the nearest reference point : , in the formula, represents the difference between the trajectory point and the reference model; Map the difference value di of each point to a color value Ci to construct a heat map: Ci=f(di); where f(di) is a color mapping function, and apply the generated color value Ci to the trajectory point cloud Paligned to generate a colored point cloud model.

4. The method for inheriting and displaying intangible cultural heritage based on artificial intelligence according to claim 3, characterized in that: In S4, divide the heat map into regions with significant differences and regions with insignificant differences, specifically: Set the high difference threshold as δthreshold and the low difference threshold as δlow, and mark the points whose difference values exceed the threshold; divide the difference values into three levels and distinguish them by colors: low difference green: di≤δlow; medium difference yellow: δlow<di≤δthreshold; high difference red: di>δthreshold; Conduct spatial clustering analysis on the difference points, mark the spatial regions where the high difference points are located, and the clustering result is expressed as: C={C1,C2,...,Ck}; where Ck is a region with significant differences.

5. The method for displaying the inheritance of intangible cultural heritage based on artificial intelligence according to claim 4, wherein: In S4, for significantly different regions, extract the trajectory data of the dance movements and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the significantly different regions, and judge the fidelity of the dynamic details. Specifically: Map each trajectory point qi of the dance movement to the surface of the reference model to obtain the surface texture data Tmotion(qi) of the dance movement. Extract the texture map Tbaseline(sp) from the reference model and map it to the corresponding 3D model surface according to the coordinates of the surface point cloud. For each pair of corresponding points qi and sp in the significantly different region, calculate the similarity of their texture images , and the expression is: ; where represents the texture image of the dance movement, represents the texture image of the reference model, is the mean of the local brightness of the image , is the mean of the local brightness of the image , is the standard deviation of the local brightness of the image , representing the contrast of the image , is the standard deviation of the local brightness of the image , representing the contrast of the image , represents the local covariance of the image and the image , and are constants; calculate the texture similarity of each point in the significantly different region to obtain the consistency index of the regional texture , and the expression is: ; where represents the significantly different region, is the number of points in the region; Dynamic detail fidelity refers to the consistency and true restoration degree of the texture of dance movements with the texture of the reference model within significantly different regions, and the calculation expression is: ; In the formula, is the dynamic detail fidelity, is the mean square error, and the calculation expression is: ; In the formula, B is the number of comparison points.

6. The method for displaying the inheritance of intangible cultural heritage based on artificial intelligence according to claim 5, wherein: In S5, evaluate the accuracy of dynamic detail capture based on the smoothness consistency of the dance movement trajectory in space and the fidelity of the dynamic details; Convert the high-frequency noise ratio fluctuation index and the dynamic detail fidelity into a comprehensive feature vector. Use the comprehensive feature vector as the input of the machine learning model. The machine learning model takes predicting the accuracy value label of dynamic detail capture for each group of comprehensive feature vectors as the prediction target, and minimizing the sum of prediction errors for all accuracy value labels of dynamic detail capture as the training target. Train the machine learning model until the sum of prediction errors reaches convergence and then stop the model training. Determine the accuracy value of dynamic detail capture according to the model output result. Among them, the machine learning model is a polynomial regression model.

7. The method for displaying the inheritance of intangible cultural heritage based on artificial intelligence according to claim 6, characterized in that: According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture. Specifically: Compare the obtained accuracy value of dynamic detail capture with the reference threshold of the accuracy value of dynamic detail capture preset according to historical data. If the accuracy value of dynamic detail capture is greater than or equal to the reference threshold of the accuracy value, it indicates that the accuracy of dynamic detail capture is high. At this time, no warning signal is generated, and the accuracy of dynamic detail capture is classified as accurate capture; if the accuracy value of dynamic detail capture is less than the reference threshold of the accuracy value, it indicates that the accuracy of dynamic detail capture is low. At this time, a warning signal is generated, and the accuracy of dynamic detail capture is classified as inaccurate capture.

8. The method for displaying the inheritance of intangible cultural heritage based on artificial intelligence according to claim 1, wherein: In S6, for inaccurate capture, use an autoregressive integrated moving average model to dynamically complete the missing dynamic details according to the captured front and back trajectory. Specifically: Obtain the trajectory time series from the capture system: ; where t = 1, 2, ..., D, t is the time frame, is the three-dimensional coordinate of the t-th frame, marking the time frame with inaccurate capture; ; Split the complete trajectory, where the known point sequence ; the missing point sequence ; the autoregression is obtained by linearly regressing the current value from the values of the previous few time frames: ; in the formula, is the trajectory value of the current time frame, is the autoregression coefficient, used to represent the contribution of the previous p trajectory values, is the trajectory value of the time frame, is the error term; the moving average corrects the current value through the residuals of the previous few time frames: ; is the moving average coefficient, used to correct the predicted value, is the time the error value at; the overall ARIMA model is expressed as: ARIMA(p,d,q); where: p is the autoregressive order, indicating how many previous values are used, d is the differencing order, indicating the number of differencing operations, and q is the moving average order, indicating how many residual values are used; For non-stationary sequences, perform d difference operations to obtain stationary sequences , and generate initial prediction input data , the expression is: ; Train the model for each trajectory dimension respectively: ; Optimize the model parameters ϕ and θ through maximum likelihood estimation. Based on the trained model, predict the values of the missing points: ; For each missing point predict point by point, and replace the missing points in the original data with the predicted values. The expression is: , represents the finally generated point cloud, is the newly added trajectory point, indicating the points supplemented during the dynamic detail capture process.

9. An AI-based intangible cultural heritage inheritance display system for implementing the AI-based intangible cultural heritage inheritance display method according to any one of claims 1-8, characterized in that: It includes a dynamic detail capture module, a trajectory smoothness analysis module, a difference analysis module, a dynamic detail fidelity analysis module, a dynamic detail capture accuracy evaluation module, and a data archiving and display module; Dynamic detail capture module: Obtain the time series data of dynamic details from the capture system in real time. The time series data includes the trajectory data of the dance movements in the intangible cultural heritage craftsmanship. Trajectory smoothness analysis module: Use Fourier transform to decompose the trajectory data into frequency components, calculate the proportion of high-frequency noise components, and judge whether the smoothness of the dance movement trajectory in space is consistent according to the abnormal fluctuation of the proportion of high-frequency noise components. Difference analysis module: When the smoothness of the dance movement trajectory in space is inconsistent, model the real scene through an ultra-high-resolution 3D scanning device and imaging device to generate a lossless reference model. Through the point cloud registration algorithm, compare the data differences between the trajectory data of the dance movement and the reference model, and map the difference value to a heat map. Dynamic detail fidelity analysis module: Divide the heat map into significantly different regions and insignificantly different regions. For significantly different regions, extract the trajectory data of the dance movement and the texture map of the reference model, compare the texture matching degree to calculate the texture consistency of the significantly different regions, and judge the fidelity of the dynamic details. Dynamic Detail Capture Accuracy Evaluation Module: Evaluate the accuracy of dynamic detail capture based on the spatial smoothness consistency of the dance movement trajectory and the fidelity of dynamic details. According to the evaluation results, divide the accuracy of dynamic detail capture into two categories: accurate capture and inaccurate capture; Data Archiving and Display Module: Archive the dynamically captured detail data that is accurately captured for repeated use in different intangible cultural heritage display projects; for inaccurate capture, use an autoregressive integrated moving average model to dynamically complete the missing dynamic details based on the captured front and back trajectories.

Images