A real-time animation generation method and generation system

Abstract

The application discloses a real-time animation generation method and a generation system, and relates to the technical field of real-time animation generation.The application solves the technical problem that the data transmission link does not fully consider the data priority and network fluctuation, which easily causes data delay or loss, and seriously affects the real-time performance and fluency of the animation.The application effectively removes noise interference by preprocessing the original data, provides a high-quality data basis for subsequent animation generation, distinguishes the change data from the regular data by statistically analyzing the distribution of the preprocessed data, determines the key data from the change data, analyzes the key data from the aspects of importance and real-time performance, further obtains the priority value of the key data and sorts the key data, performs data transmission according to the priority, ensures the priority and timely transmission of the key data, and improves the quality and fluency of the animation generation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of real-time animation generation, in particular to a real-time animation generation method and system. BACKGROUND

[0002] In the field of real-time animation generation, with the continuous expansion of animation application scenarios such as real-time games, virtual live streaming, immersive interactive experience, etc., higher requirements are put forward for the quality, smoothness and real-time performance of animation generation.

[0003] The patent application with publication number CN107481303A discloses a real-time animation generation method and system, including the following steps: defining a variable set related to a moving object; collecting object motion information and generating first motion data; collecting at least part of the first motion data; generating second motion data according to the variable set and the at least part of the first motion data collected; and synthesizing animation based on the second motion data.

[0004] However, some existing generation methods have many limitations in data collection, processing and transmission, etc. The data processing process lacks effective classification, and cannot accurately distinguish between key data and regular data, making the animation not prominent enough in important details. In the data transmission link, the data priority and network fluctuations are not fully considered, which easily causes data delay or loss, seriously affecting the real-time performance and smoothness of the animation. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a real-time animation generation method and system, which solves the problem that the data transmission link does not fully consider the data priority and network fluctuations, which easily causes data delay or loss, seriously affecting the real-time performance and smoothness of the animation.

[0006] To achieve the above purpose, the present application realizes the following technical scheme: a real-time animation generation method, which specifically includes the following steps:

[0007] Step S1, collecting animation data and recording as original motion data, and pre-processing the original motion data to obtain pre-processed data;

[0008] Step S2, classifying the pre-processed data according to the distribution to obtain change data and regular data, and analyzing the change frequency of the change data to determine the key data and regular data after secondary classification;

[0009] S3, analyzing the importance and real-time performance of the key data, and calculating the priority of the key data to determine the corresponding transmission order, and sorting and marking according to the priority;

[0010] S4, transmission processing is performed on the sorted and marked key data, parallel data is determined by analyzing the parallel relationship between the key data and the regular data, and the parallel data is transmitted asynchronously, transmission analysis processing is performed based on the transmission protocol and the transmission network condition, and corresponding transmission analysis information is generated.

[0011] As a further scheme of the present application, the specific manner of obtaining the preprocessed data in step S1 is:

[0012] The original action data is obtained, and the preprocessed data is obtained by denoising and filtering processing on the original action data. The denoising processing can be performed by mean filtering, median filtering, and wavelet denoising, and the filtering processing can be performed by low-pass filtering, high-pass filtering, and band-pass filtering.

[0013] As a further scheme of the present application, the specific manner of obtaining the preprocessed data in step S2 is:

[0014] All preprocessed data is obtained, and then statistical analysis is performed on the collected preprocessed data to obtain the distribution of the data. The preprocessed data within the normal range of variation is determined as regular data, and the preprocessed data not within the normal range of variation is determined as variation data.

[0015] As a further scheme of the present application, the specific manner of obtaining the preprocessed data in step S2 is:

[0016] All variation data is obtained, and then the variation frequency corresponding to the variation data is calculated. According to the sampling frequency f s The variation data is collected and recorded as a data sequence x(n), where n=1, 2, …, N-1, where n represents the number of sampling points. Then, the time domain data sequence x(n) is converted to the frequency domain using the FFT algorithm to obtain the frequency spectrum X(k), where k=1, 2, …, N-1. According to the sampling theorem, the range of the frequency spectrum is determined as [0, f s / 2], and the actual frequency f

[0017] The actual frequency f k of the variation data is compared with the preset value, and the actual frequency f k greater than the preset value is recorded as key data, and the actual frequency f k less than the preset value is recorded as regular data.

[0018] As a further scheme of the present application, the specific manner of analyzing the importance of the key data in step S3 is:

[0019] Establish animation feature vectors Simultaneously establish key data feature vectors Next, according to the cosine similarity formula Calculate the animation feature vector With key data feature vectors The cosine similarity value is calculated and used as the key data importance value Qi, where i = 1, 2, ..., j, and j represents the number of key data.

[0020] As a further aspect of the present invention, the specific method for analyzing the real-time performance of key data in step S3 is as follows:

[0021] The real-time performance of key data is analyzed to obtain the maximum latency corresponding to key data i. This maximum latency is then used as the latency tolerance value Gi for key data i. Next, the matching degree between key data i and the animation frame rate is calculated to obtain the timestamp ti corresponding to key data i. The animation frame rate is also obtained and denoted as y. Simultaneously, the frame interval is calculated based on the frame rate y. Next, for each key data point i, find the animation frame time t that is closest to it. j = j × Δt, where j is an integer, then calculate the time difference Δt. i =|t i -t j |, then according to the formula Calculate the average time difference corresponding to the key data i and record it as the animation frame rate matching value Hi. Then, sum the calculated latency tolerance value Gi with the animation frame rate matching value Hi to obtain the real-time value Si of the key data.

[0022] As a further aspect of the present invention, the specific method for calculating the priority of key data and determining the corresponding transmission order in step S3 is as follows:

[0023] The priority value Wi of the key data is calculated by summing the importance value Qi and the real-time value Si of the key data, and Wi = Qi + Si. The key data are then sorted from largest to smallest according to their priority values ​​Wi, and the timestamps of the key data are marked.

[0024] As a further aspect of the present invention, the specific method for transmitting the sorted and marked key data in step S4 is as follows:

[0025] Acquire all key and routine data, identify parallel data based on data characteristics, and asynchronously transmit the obtained parallel data.

[0026] Then the transmission protocol is acquired, the transmission format in the transmission protocol is acquired, and the transmission format is matched with the format of the key data; if they are the same, the current transmission protocol is used for transmission; otherwise, if they are different, the transmission protocol is selected according to the format of the key data as a standard, a standard transmission protocol is generated, the transmission network condition is acquired, the maximum transmission speed peak value and the minimum transmission speed peak value corresponding to the transmission network condition within a time T are acquired, then the difference between the two is calculated as a peak difference value, and the obtained peak difference value is compared with a preset difference value;

[0027] If the peak difference value is greater than the preset difference value, it indicates that the transmission network condition is unstable, and an unstable transmission signal is generated; otherwise, if the peak difference value is less than the preset difference value, it indicates that the transmission network condition is stable, and a stable transmission signal is generated, and transmission processing is performed according to the generated signal.

[0028] As a further scheme of the application, the specific manner of the step S4 of performing transmission processing according to the generated signal is:

[0029] The generated unstable transmission signal is processed, the minimum fluctuation range value corresponding to the transmission network condition within a time t1 is acquired, and the minimum value corresponding to the minimum fluctuation range value is used as a standard transmission speed, then the key data is segmented and transmitted according to the standard transmission speed.

[0030] The generated stable transmission signal is processed, the transmission speed average corresponding to the transmission network condition within a time T is acquired, and the transmission speed average is used as a standard to segment and transmit the key data, and transmission analysis information is generated.

[0031] A real-time animation generation system, comprising a data acquisition module, a data preprocessing and classification module, a priority determination and analysis module, a transmission analysis module, and a transmission information output module.

[0032] The data acquisition module is used to acquire animation data and transmit it to the data preprocessing and classification module.

[0033] The data preprocessing and classification module is used to preprocess the acquired animation data to obtain preprocessed data, classify the preprocessed data according to the distribution of the preprocessed data to obtain change data and regular data, analyze the change frequency of the change data to determine the key data and regular data after secondary classification, and transmit the key data and regular data to the priority determination and analysis module.

[0034] The priority determination and analysis module analyzes the importance and real-time performance of the key data, calculates the priority of the key data to determine the transmission order, sorts and marks the key data according to the priority, generates sorted key data, and transmits it to the transmission analysis module.

[0035] The transmission analysis module is used for analyzing the obtained sorting key data, determining parallel data by analyzing the parallel relationship between the key data and the regular data, and performing asynchronous transmission on the parallel data, and performing transmission analysis processing based on the transmission protocol and the transmission network condition to generate corresponding transmission analysis information, and transmitting the transmission analysis information to the transmission information output module;

[0036] The transmission information output module is used for transmitting the animation data according to the obtained transmission analysis information.

[0037] The application provides a real-time animation generation method and a generation system.

[0038] The application effectively removes noise interference by preprocessing the original data, provides a high-quality data basis for subsequent animation generation, accurately distinguishes change data and regular data by statistical analysis of the distribution of the preprocessed data and combination of data change frequency calculation, further determines key data from the change data, analyzes the key data from the aspects of importance and real-time, calculates the importance value by calculating the cosine similarity of the key data and the animation theme style, calculates the real-time value by combining the delay tolerance value and the animation frame rate matching value, and then obtains the priority value of the key data and sorts, performs data transmission according to the priority, ensures the priority and timely transmission of the key data, and improves the quality and smoothness of the animation generation. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 The application provides a real-time animation generation method and a generation system.

[0040] Figure 2 The application provides a real-time animation generation method and a generation system. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0042] Embodiment one, please refer to Figure 1 The application provides a real-time animation generation method, which specifically includes the following steps:

[0043] Step S1, collecting animation data, selecting a motion capture device with excellent performance, such as OptiTrack and other professional optical motion capture systems, which have high-precision sensors and fast data acquisition capabilities, can obtain accurate motion data in a short time, and are recorded as original motion data, and the original motion data is preprocessed to obtain preprocessed data.

[0044] Obtain the original motion data, and obtain the preprocessed data by denoising and filtering the original motion data. The specific denoising processing can be processed by mean filtering, median filtering and wavelet denoising, and the filtering processing can be processed by low-pass filtering, high-pass filtering and band-pass filtering. The specific processing method is selected by the operator.

[0045] Step S2, classify the preprocessed data according to the distribution to obtain change data and regular data, and analyze the change frequency of the change data to determine the key data and regular data after secondary classification.

[0046] S21, obtain all preprocessed data, then statistically analyze the collected preprocessed data to understand the distribution of the data, such as the average value and standard deviation of the joint angle, etc. Assuming that in a record of a person's walking motion data, the right knee joint angle values at multiple sampling points are 120°, 125°, 118°, 122°, etc. After calculation, the average value is 121°, which represents the average angle state of the right knee joint in the walking motion. At the same time, according to the data distribution, the preprocessed data within the normal range of change is determined as regular data, and the preprocessed data not within the normal range of change is determined as change data. For example, for the angle change of the left hip joint in the normal walking motion of an adult, according to a large number of existing normal walking motion data statistics, the angle usually fluctuates between 90°-130°. Therefore, the left hip joint angle data within this range is identified as regular data. For example, in the above left hip joint angle data, if the angle value of a certain sampling point is 150°, which is obviously beyond the normal range of 90°-130°, then this data point and a series of data associated with it can be marked as change data.

[0047] S22, obtain all change data, then calculate the change frequency corresponding to the change data according to the sampling frequency f s Collect the change data and record it as data sequence x(n), where n=1, 2, …, N-1, where n represents the number of sampling points. Then, the time domain data sequence x(n) is converted to frequency domain using the FFT algorithm to obtain the frequency spectrum X(k), where k=1, 2, …, N-1. According to the sampling theorem, the range of the frequency spectrum is determined as [0, f s / 2], and the actual frequency corresponding to the frequency point is calculated

[0048] S23、the actual frequency f corresponding to the change data is compared with the preset value, and the specific value of the preset value is set by the operator, and the actual frequency f is screened k k The change data greater than the preset value is recorded as key data, and the actual frequency f is recorded as key data k The change data less than the preset value is recorded as regular data.

[0049] Step S3, by analyzing the importance and real-time of the key data, and calculating the priority of the key data to determine the transmission order, and sorting and marking according to the priority.

[0050] The key data is obtained, and then the key data is analyzed from the importance and real-time;

[0051] S31, the importance of the key data is analyzed, and the animation feature vector is established And the feature vector here represents the vector information corresponding to the theme and style information of the animation, and the key data feature vector is established For example, for a martial arts style animation, the feature vector of the theme and style of the animation may include the stiffness of the action, the speed, the smoothness of the move, etc. Dimension; and the key data may be the motion trajectory of the character's limbs, the joint angle change, etc. These key data are converted into corresponding feature vectors, such as the curvature of the motion trajectory, the amplitude of the joint angle change, etc. Then according to the cosine similarity formula The animation feature vector is calculated And the cosine similarity value of the key data feature vector The cosine similarity value calculated is taken as the key data importance value Qi, and i=1, 2, …, j, wherein j represents the number of key data;

[0052] S32, the real-time of the key data is analyzed, the maximum delay corresponding to the key data i is obtained, and the maximum delay is taken as the delay tolerance value Gi of the key data i. Then the matching degree value of the key data i and the animation frame rate is calculated, the time stamp ti corresponding to the key data i is obtained, and the frame rate of the animation is recorded as y, and the frame time interval Then for each key data point i, find the closest animation frame time t j =j×Δt, where j is an integer, and then calculate the time difference Δt i =|t i -t j |, and then according to the formula ​The average time difference value corresponding to the key data i is calculated and recorded as the animation frame rate matching value Hi, and the delay tolerance value Gi obtained by calculation is summed with the animation frame rate matching value Hi to obtain the key data real-time value Si, and the specific calculation formula is Si = Gi x a1 + Hi x a2, wherein a1 and a2 are corresponding weight coefficients, and the specific values are set by the operator.

[0053] S33, the key data importance value Qi and the key data real-time value Si are summed to obtain the priority value Wi of the key data, and Wi = Qi + Si, and the key data timestamp is marked according to the key data priority value Wi from large to small.

[0054] Step S4, the sorted and marked key data is transmitted, the parallel data is determined by analyzing the parallel relationship between the key data and the regular data, and the parallel data is transmitted asynchronously, and transmission analysis information is generated based on the transmission protocol and the transmission network condition.

[0055] All key data and regular data are obtained, and parallel data is determined based on data characteristics. The specific parallel data usually has similar data structure and format. For example, in text data, parallel sentences or paragraphs may have the same syntax structure, part of speech distribution, or follow similar patterns. For example, in English-Chinese parallel corpus, the corresponding sentences may have certain similarity in word order and sentence component arrangement. Parallel data are closely related in content and theme, often describing the same or similar things, concepts, events, etc. The obtained parallel data are transmitted asynchronously;

[0056] Then, the transmission protocol is obtained, and the transmission format in the transmission protocol is obtained, and the transmission format is matched with the format of the key data. If they are the same, the current transmission protocol is used for transmission. Otherwise, if they are not the same, the transmission protocol is selected based on the format of the key data to generate a standard transmission protocol. The transmission network condition is obtained, and the maximum transmission speed peak value and the minimum transmission speed peak value corresponding to the transmission network condition in the time T are obtained. Then, the difference between the two is calculated and recorded as the peak value difference. The obtained peak value difference is compared with the preset difference value, and the specific value of the preset difference value is set by the operator.

[0057] If the peak value difference is greater than the preset difference value, it indicates that the transmission network condition is unstable, and an unstable transmission signal is generated. Otherwise, if the peak value difference is less than the preset difference value, it indicates that the transmission network condition is stable, and a stable transmission signal is generated. Transmission is performed according to the generated signal respectively.

[0058] The generated unstable transmission signal is subjected to transmission processing, the minimum fluctuation range value corresponding to the transmission network condition within time t1 is obtained, and the minimum value corresponding to the minimum fluctuation range value is taken as the standard transmission speed, and then the key data is subjected to segmented transmission according to the standard transmission speed;

[0059] The generated stable transmission signal is subjected to transmission processing, the transmission speed average corresponding to the transmission network condition within time T is obtained, and the transmission speed average is taken as the standard to segment the key data for transmission, and transmission analysis information is generated; specifically, the segmented transmission is performed according to the priority of the key data, and the processing mode of the normal data is the same as that of the key data.

[0060] Embodiment two, please refer to Figure 2 The application provides a real-time animation generation system, which comprises a motion data acquisition module, a data preprocessing and classification module, a priority determination and analysis module, a transmission analysis module and a transmission information output module, and combines Figure 2 It can be known that the above functional modules are unidirectionally electrically connected.

[0061] The data acquisition module is used for acquiring animation data and transmitting the animation data to the data preprocessing and classification module.

[0062] The data preprocessing and classification module is used for preprocessing the acquired animation data to obtain preprocessing data, classifying the preprocessing data according to the distribution of the preprocessing data to obtain change data and normal data, analyzing the change frequency of the change data to determine key data and normal data after secondary classification, and transmitting the key data and the normal data to the priority determination and analysis module.

[0063] The priority determination and analysis module analyzes the importance and real-time performance of the key data, calculates the priority of the key data to determine the transmission order, sorts and marks the key data according to the priority, generates sorted key data, and transmits the sorted key data to the transmission analysis module.

[0064] The transmission analysis module is used for analyzing the sorted key data, determining parallel data by analyzing the parallel relationship between the key data and the normal data, performing asynchronous transmission on the parallel data, performing transmission analysis processing based on a transmission protocol and a transmission network condition, generating corresponding transmission analysis information, and transmitting the transmission analysis information to the transmission information output module.

[0065] The transmission information output module is configured to transmit the animation data according to the obtained transmission analysis information.

[0066] Some data in the above formula are taken as dimension for numerical calculation, and the contents not described in detail in the specification all belong to the prior art known by those skilled in the art.

[0067] The above embodiments are only used to illustrate the technical method of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. A method for real-time animation generation, characterized by, The method specifically comprises the following steps: Step S1, collecting animation data and recording as original motion data, and preprocessing the original motion data to obtain preprocessed data; Step S2, classifying according to the distribution of the preprocessed data to obtain change data and regular data, and analyzing the change frequency of the change data to determine the key data and regular data after secondary classification, and the specific processing method is: All the change data is acquired, then the change frequency corresponding to the change data is calculated, and the sampling frequency f s The change data is collected and recorded as a data sequence x(n), where n=1, 2, …, N-1, where N represents the number of sampling points, then the time domain data sequence x(n) is converted into frequency domain by using the FFT algorithm to obtain the frequency spectrum X(k), where k=1, 2, …, N-1, and according to the sampling theorem, the range of the frequency spectrum is determined as [0, f s / 2], and the actual frequency corresponding to the frequency point is calculated ; Corresponding to the change data, the actual frequency f k Compare with the preset value, and filter the actual frequency f k The change data greater than the preset value is recorded as key data, and the actual frequency f k The change data less than the preset value is recorded as regular data; S3, analyze the importance and real-time of the key data, calculate the priority of the key data to determine the transmission order, and sort and mark according to the priority; S4, transmit the sorted and marked key data, determine parallel data by analyzing the parallel relationship between key data and regular data, and perform asynchronous transmission on the parallel data, and perform transmission analysis and processing based on the transmission protocol and transmission network conditions to generate corresponding transmission analysis information, and the specific processing method is as follows: Get all the key data and regular data, and determine the parallel data based on the recognition of data characteristics, and perform asynchronous transmission on the obtained parallel data; Then get the transmission protocol, get the transmission format in the transmission protocol, match the transmission format with the format of the key data, if they are the same, transmit with the current transmission protocol, otherwise if they are not the same, select the transmission protocol based on the format of the key data, generate a standard transmission protocol, get the transmission network condition, and get the maximum transmission speed peak value and the minimum transmission speed peak value corresponding to the transmission network condition within time T, then calculate the difference value of the two as peak difference value, and compare the obtained peak difference value with the preset difference value; If the peak difference value is greater than the preset difference value, it indicates that the transmission network condition is unstable, and an unstable transmission signal is generated, otherwise if the peak difference value is less than the preset difference value, it indicates that the transmission network condition is stable, and a stable transmission signal is generated, and transmission processing is performed according to the generated signal respectively; For the generated unstable transmission signal, get the minimum fluctuation range value corresponding to the transmission network condition within time t1, and take the minimum value corresponding to the minimum fluctuation range value as the standard transmission speed, then divide and transmit the key data according to the standard transmission speed; For the generated stable transmission signal, get the transmission speed average value corresponding to the transmission network condition within time T, and divide and transmit the key data according to the transmission speed average value to generate transmission analysis information, and transmit the animation data according to the transmission analysis information.

2. The method of claim 1, wherein, The specific way of obtaining preprocessed data in step S1 is: Get the original motion data, and get the preprocessed data by denoising and filtering the original motion data, wherein the denoising processing can be performed by mean filtering, median filtering and wavelet denoising, and the filtering processing can be performed by low-pass filtering, high-pass filtering and band-pass filtering.

3. The method of claim 1, wherein, The specific way of classifying the preprocessed data to obtain change data and regular data in step S2 is: All pre-processing data is acquired, and then statistical analysis is performed on the collected pre-processing data to obtain the distribution of the data, and pre-processing data within the normal range of variation is determined as normal data and pre-processing data not within the normal range of variation is determined as variation data according to the distribution of the data.

4. The method of claim 1, wherein, The step S3 analyzes the importance of the key data in the following specific manner: Establish animation feature vectors At the same time, establish key data feature vectors Then, according to the cosine similarity formula Calculate the animation feature vector With key data feature vectors The cosine similarity value is calculated and used as the key data importance value Qi, where i = 1, 2, ..., j, and j represents the number of key data.

5. The method of claim 1, wherein, The step S3 analyzes the real-time performance of the key data in the following specific manner: The key data real-time performance is analyzed, the maximum delay corresponding to the key data i is obtained, and the maximum delay is taken as the delay tolerance value Gi of the key data i. Then, the matching degree value of the key data i and the animation frame rate is calculated, the timestamp ti corresponding to the key data i is obtained, the frame rate of the animation is recorded as y, and the frame time interval is calculated according to the frame rate y . Then, for each key data point i, the closest animation frame time t j is found , where j is an integer, and then the time difference is calculated. Then, the average time difference value corresponding to the key data i is calculated according to the formula , and is recorded as the animation frame rate matching value Hi. The delay tolerance value Gi and the animation frame rate matching value Hi are summed to obtain the key data real-time performance value Si.

6. The method of claim 1, wherein, The step S3 calculates the priority of the key data to determine the corresponding transmission order in the following specific manner: The key data importance value Qi and the key data real-time performance value Si are summed to obtain the priority value Wi of the key data, and Wi=Qi+Si, and the key data is sorted in descending order of the priority value Wi, and the time stamp of the key data is marked.

7. A real-time animation generation system for performing a real-time animation generation method according to any one of claims 1 to 6, characterized in that The system comprises a data acquisition module, a data pre-processing classification module, a priority determination analysis module, a transmission analysis module and a transmission information output module. The data acquisition module is configured to acquire animation data and transmit the animation data to the data pre-processing classification module. The data pre-processing classification module is configured to pre-process the acquired animation data to obtain pre-processing data, classify the pre-processing data according to the distribution of the pre-processing data to obtain variation data and normal data, analyze the variation frequency of the variation data to determine key data and normal data after secondary classification, and transmit the key data and the normal data to the priority determination analysis module. The priority determination analysis module is configured to analyze the importance and real-time performance of the key data, calculate the priority of the key data to determine the corresponding transmission order, sort and mark the key data according to the priority, generate sorted key data, and transmit the sorted key data to the transmission analysis module. The transmission analysis module is configured to analyze the sorted key data, determine parallel data by analyzing the parallel relationship between the key data and the normal data, perform asynchronous transmission on the parallel data, perform transmission analysis and processing based on a transmission protocol and a transmission network, generate corresponding transmission analysis information, and transmit the transmission analysis information to the transmission information output module. The transmission information output module is configured to transmit the animation data according to the acquired transmission analysis information.

Images