Module and method for inputting three-dimensional motion data

Abstract

A module and a method for inputting three-dimensional motion data belong to the technical field of inputting three-dimensional motion data, and are characterized in that the real-time three-dimensional motion data is acquired and pre-processed, and the acquisition speed is actively changed according to the reliable condition of the real-time three-dimensional motion data, so that the accurate analysis of the three-dimensional motion data is ensured. According to the method, the task load of accurate analysis of the three-dimensional motion data is reduced, the reliability of the accurate analysis of the three-dimensional motion data is ensured, and the more accurate acquisition reference score of the three-dimensional motion data is obtained by calibrating each numerical point in the three-dimensional motion data and removing the disturbance of the acquisition speed to the acquisition reference score of the numerical point, so that the accuracy of the three-dimensional motion data is improved. The higher the acquisition reference score is, the outlier of the three-dimensional motion data is generated, and the acquisition speed of the three-dimensional motion data is actively changed, so that the accurate analysis of the three-dimensional motion data is more accurately met.

Description

Technical Field

[0001] The present invention belongs to the technical field of inputting three-dimensional motion data, and in particular relates to a module and method for inputting three-dimensional motion data. Background Art

[0002] Through the small sensors worn on athletes as motion collection devices, the athletes' three-dimensional motion data can be collected in real time, including position, speed, acceleration, posture and other information.

[0003] Specifically, as mentioned in the prior art solution with patent publication number "CN111587809A", the motion acquisition device is connected to the central control device. The motion acquisition device is used to instantly acquire the three-dimensional motion data of the object to be measured and transmit it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured. The motion acquisition device is a module for inputting the three-dimensional motion data.

[0004] However, currently, when analyzing the three-dimensional motion data of the object to be measured, a pre-set constant acquisition speed is generally used to collect the three-dimensional motion data. When the motion state data obtained is detected, it cannot meet the analysis purpose of being efficient, accurate and reliable. If the acquisition speed is set too high, the amount of information to be processed will be too much, and a huge amount of calculation will be required; and if the acquisition speed is too low, it will be impossible to obtain accurate motion state data when analyzing the collected three-dimensional data, that is, it cannot meet the reliability requirement. Summary of the invention

[0005] In order to solve the defects existing in the prior art, the present invention proposes a module and method for inputting three-dimensional motion data, which performs pre-processing on real-time three-dimensional motion data, and actively changes the acquisition speed according to the reliability of the real-time three-dimensional motion data to ensure accurate analysis of the three-dimensional motion data, thereby reducing the workload of accurate analysis of the three-dimensional motion data and ensuring the reliability of accurate analysis of the three-dimensional motion data. By calibrating each numerical point in the three-dimensional motion data and removing the disturbance of the acquisition speed on the acquisition benchmark score of the numerical point, a more accurate acquisition benchmark score of the three-dimensional motion data is obtained. The higher the acquisition benchmark score, the more outliers the three-dimensional motion data will produce. By actively changing the acquisition speed of the three-dimensional motion data, a more accurate and accurate analysis of the three-dimensional motion data can be achieved.

[0006] The present invention uses the following technical solutions.

[0007] A module for inputting three-dimensional motion data, comprising:

[0008] The motion acquisition device is connected to the central control device. The motion acquisition device is used to instantly acquire the three-dimensional motion data of the object to be measured and transmit it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured. The motion acquisition device is a module that inputs the three-dimensional motion data.

[0009] The unit running on the motion acquisition device includes:

[0010] A standard unit, which is used to obtain real-time 3D motion data for pre-processing, perform outlier detection on the 3D motion data using the Z-score method, and obtain a starting benchmark score for the 3D motion data;

[0011] An amplitude unit is used to obtain the acquisition speed of each numerical point in the three-dimensional motion data and the numerical point in the near zone, obtain the acquisition benchmark score of each numerical point in the near zone according to the initial benchmark score and the acquisition speed, calculate the variance of all real-time three-dimensional motion data as the global fluctuation amplitude of the three-dimensional motion data, and obtain the partial fluctuation amplitude of each numerical point according to the acquisition speed of each numerical point and the acquisition speed of the numerical point in the near zone and the average of each numerical point and the three-dimensional motion data;

[0012] A calculation unit, which is used to calculate the acquisition speed variation coefficient of each numerical point according to the acquisition reference score and the partial fluctuation amplitude;

[0013] A refresh unit is used to treat the acquisition speed variation coefficient as a variation parameter of the acquisition speed corresponding to each numerical point, refresh the acquisition speed of each numerical point, perform acquisition according to the refreshed acquisition speed, obtain outliers of the three-dimensional motion data and remove the outliers, and perform analysis on the three-dimensional motion data with the outliers removed to obtain motion state data.

[0014] A method for inputting three-dimensional motion data, comprising:

[0015] The motion acquisition device collects the three-dimensional motion data of the object to be measured in real time and transmits it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured;

[0016] The period during which the motion acquisition device instantly acquires the three-dimensional motion data of the object to be measured also includes:

[0017] Step 1, obtain real-time 3D motion data for pre-processing, use the Z-score method to perform outlier detection on the 3D motion data, and obtain the initial benchmark score of the 3D motion data;

[0018] Step 2, obtaining the acquisition speed of each numerical point in the three-dimensional motion data and the numerical point in the near zone, obtaining the acquisition benchmark score of each numerical point in the near zone according to the initial benchmark score and the acquisition speed, calculating the variance of all real-time three-dimensional motion data as the global fluctuation amplitude of the three-dimensional motion data, and obtaining the partial fluctuation amplitude of each numerical point according to the acquisition speed of each numerical point and the acquisition speed of the numerical point in the near zone and the average of each numerical point and the three-dimensional motion data;

[0019] Step 3, calculating the acquisition speed variation coefficient of each numerical point according to the acquisition benchmark score and the partial fluctuation amplitude;

[0020] Step 4, taking the acquisition speed variation coefficient as the variation parameter of the acquisition speed corresponding to each numerical point, refreshing the acquisition speed of each numerical point, performing acquisition according to the refreshed acquisition speed, obtaining the outliers of the three-dimensional motion data and removing the outliers, performing analysis on the three-dimensional motion data with the outliers removed, thereby obtaining the motion state data.

[0021] Preferably, in step 1, a minimum acquisition speed is defined in advance, and a minimum acquisition speed T is defined in advance. 0 .

[0022] Preferably, in step 1, the three-dimensional motion data is collected, and the collected three-dimensional motion data is real-time three-dimensional motion data. The j-th three-dimensional motion data collected is defined as E j , define the number of all three-dimensional motion data collected as p.

[0023] Preferably, in step 1, the collected three-dimensional motion data is benchmarked using the Z-score method, that is, any three-dimensional motion data in the three-dimensional motion data is taken as the target value point, the Pearson coefficient between the target value point and the other collected three-dimensional motion data is calculated, and each Pearson coefficient is arranged in order from high to low, and then the Pearson coefficients arranged in the first x are divided by 1 to obtain each quotient value, and then the average of each quotient value is calculated, so as to take the average as the partial aggregation degree of the target value point;

[0024] The quotient obtained by dividing the partial aggregation degree of the target numerical point by the average of the partial aggregation degrees of other three-dimensional motion data corresponding to the previous x Pearson coefficients is used as the starting benchmark score, thereby obtaining the starting benchmark score of the three-dimensional motion data. When the starting benchmark score is higher than one, the target numerical point is an outlier.

[0025] Preferably, in step 1, a near zone of each numerical point in the three-dimensional motion data is defined in advance, and the near zone of the numerical point contains a numerical point one of which the number is set in advance to y. The numerical point one of which the number is set in advance to y is the first y numerical points whose Pearson coefficients between other numerical points in the three-dimensional motion data other than the numerical point to which the near zone belongs and the numerical point to which the near zone belongs are arranged in order from high to low. The benchmark score of the j-th numerical point obtained based on the above is the starting benchmark score B. j .

[0026] Preferably, step 2 specifically comprises:

[0027] Step 2-1: Calculate the quotient of the cumulative amount of the acquisition speed of each numerical point in the near area corresponding to each numerical point and the acquisition speed of the numerical point, and multiply the quotient by the initial reference score as the acquisition reference score;

[0028] Step 2-2: Select an arbitrary numerical point as the target numerical point, calculate the subtraction of the mean of the three-dimensional motion data of the target numerical point and other numerical points in the nearby area, and obtain the quotient of the acquisition speed of the target numerical point and the acquisition speed of an arbitrary numerical point in the nearby area corresponding to the target numerical point; take the square root of the average of the cumulative amount of the square of the amount obtained by multiplying the subtraction and the quotient as the partial fluctuation amplitude of the target numerical point.

[0029] Preferably, in step 2-1, the calculation equation for collecting the benchmark score is:

[0030]

[0031] In the equation, H j represents the acquisition benchmark score of the jth value point, B j represents the starting benchmark score of the j-th numerical point, l represents the number of numerical points contained in the numerical point cluster formed by all numerical points in the corresponding nearby area of ​​the j-th numerical point, T j,k represents the acquisition speed of the kth value point in the value point cluster corresponding to the jth value point, T j Represents the acquisition speed of the jth numerical point.

[0032] Preferably, in step 2-2, the calculation equation of the partial fluctuation amplitude is:

[0033]

[0034] In the equation, represents the partial fluctuation amplitude at the jth three-dimensional motion data, l represents the number of numerical points in the corresponding near-area numerical point cluster of the jth numerical point, T 0 represents the predefined starting data acquisition speed, T j,krepresents the acquisition speed of the kth value point in the cluster of value points in the vicinity corresponding to the jth value point, E j,k represents the value of the three-dimensional motion data of the kth value point in the cluster of value points near the jth value point, ν j Represents the average of all 3D motion data in the near-area numerical point cluster of the j-th 3D motion data.

[0035] Preferably, in step 3, the calculation equation of the acquisition speed variation coefficient is:

[0036]

[0037] In the equation, G j Represents the acquisition speed variation coefficient of the jth value point, H j represents the acquisition benchmark score of the jth value point, H 0 Represents a predefined collection benchmark score benchmark, Represents the global fluctuation amplitude of the 3D motion data, Represents the partial fluctuation amplitude at the jth three-dimensional motion data.

[0038] Preferably, in step 3, the benchmark amount of the acquisition benchmark score is ninety percent.

[0039] Preferably, in step 4, the variable parameter is multiplied by the current numerical point acquisition speed to obtain the value obtained as the refreshed numerical point acquisition speed. Based on the refreshed numerical point acquisition speed, the three-dimensional motion data is immediately acquired, and the acquisition benchmark score of each acquired three-dimensional motion data is calculated. When there are a pre-defined number of numerical point acquisition benchmark scores that are higher than the acquisition benchmark score benchmark amount, it is determined that the three-dimensional motion data has outliers, and the Z-score method is used to obtain the outlier value of the three-dimensional motion data.

[0040] Preferably, in step 4, the variable parameter is multiplied by the current value point acquisition speed as the refreshed value point acquisition speed, and the next three-dimensional motion data is acquired according to the refreshed acquisition speed;

[0041] It is assumed that the predefined number U is seventeen and the collection benchmark score benchmark is ninety percent.

[0042] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include:

[0043] By obtaining real-time three-dimensional motion data for pre-processing, the acquisition speed is actively changed according to the reliability of the real-time three-dimensional motion data to ensure accurate analysis of the three-dimensional motion data, reducing the workload of accurate analysis of the three-dimensional motion data and ensuring the reliability of accurate analysis of the three-dimensional motion data. By calibrating each numerical point in the three-dimensional motion data and removing the disturbance of the acquisition speed on the acquisition benchmark score of the numerical point, a more accurate acquisition benchmark score of the three-dimensional motion data is obtained. The higher the acquisition benchmark score, the more outliers the three-dimensional motion data will produce. By actively changing the acquisition speed of the three-dimensional motion data, more accurate analysis of the three-dimensional motion data can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of the method for inputting a module of three-dimensional motion data in the present invention;

[0045] Figure 2 It is a partial structural diagram of the module for inputting three-dimensional motion data in the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the present invention clearer, the following will be combined with the drawings in the embodiments of the present invention to clearly and completely express the technical solution of the present invention. The embodiments expressed in this application are only some embodiments of the present invention, not all embodiments. According to the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the protection scope of the present invention.

[0047] like Figure 2 As shown, a module for inputting three-dimensional motion data according to the present invention comprises:

[0048] The motion acquisition device is connected to the central control device. The motion acquisition device is used to instantly acquire the three-dimensional motion data of the object to be measured and transmit it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured. The motion acquisition device is a module that inputs the three-dimensional motion data.

[0049] The unit running on the motion acquisition device includes:

[0050] A standard unit, which is used to obtain real-time 3D motion data for pre-processing, perform outlier detection on the 3D motion data using the Z-score method, and obtain a starting benchmark score for the 3D motion data;

[0051] An amplitude unit is used to obtain the acquisition speed of each numerical point in the three-dimensional motion data and the numerical point in the near zone, obtain the acquisition benchmark score of each numerical point in the near zone according to the initial benchmark score and the acquisition speed, calculate the variance of all real-time three-dimensional motion data as the global fluctuation amplitude of the three-dimensional motion data, and obtain the partial fluctuation amplitude of each numerical point according to the acquisition speed of each numerical point and the acquisition speed of the numerical point in the near zone and the average of each numerical point and the three-dimensional motion data;

[0052] A calculation unit, which is used to calculate the acquisition speed variation coefficient of each numerical point according to the acquisition reference score and the partial fluctuation amplitude;

[0053] A refresh unit is used to treat the acquisition speed variation coefficient as a variation parameter of the acquisition speed corresponding to each numerical point, refresh the acquisition speed of each numerical point, perform acquisition according to the refreshed acquisition speed, obtain outliers of the three-dimensional motion data and remove the outliers, and perform analysis on the three-dimensional motion data with the outliers removed to obtain motion state data.

[0054] like Figure 1 As shown, a method for inputting three-dimensional motion data according to the present invention comprises:

[0055] The motion acquisition device collects the three-dimensional motion data of the object to be measured in real time and transmits it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured;

[0056] The period during which the motion acquisition device instantly acquires the three-dimensional motion data of the object to be measured also includes:

[0057] Step 1, obtain real-time 3D motion data for pre-processing, use the Z-score method to perform outlier detection on the 3D motion data, and obtain the initial benchmark score of the 3D motion data;

[0058] However, a too low three-dimensional motion data acquisition speed often makes it impossible to obtain accurate motion state data when analyzing the collected three-dimensional data. Therefore, in a preferred but non-limiting embodiment of the present invention, in step 1, a minimum acquisition speed is defined in advance, and the acquisition speed of the three-dimensional motion data should not be lower than the minimum acquisition speed no matter how it changes; a minimum acquisition speed T is defined in advance. 0 , T 0 The three-dimensional motion data can be collected one hundred times per second.

[0059] In a preferred but non-limiting embodiment of the present invention, in step 1, the three-dimensional motion data can be collected according to the minimum collection speed and another collection speed. The collected three-dimensional motion data is real-time three-dimensional motion data. The motion state of the current object to be measured can be obtained through the real-time change of the three-dimensional motion data. When the three-dimensional motion data suddenly increases or decreases, it often indicates that the motion amplitude of the object to be measured will increase or has increased. It is necessary to perform reliable analysis on the three-dimensional motion data more effectively, and define the jth three-dimensional motion data collected as E j , define the number of all three-dimensional motion data collected as p.

[0060] In a preferred but non-restrictive embodiment of the present invention, in step 1, the Z-score method is used to perform benchmarking on the collected three-dimensional motion data, that is, any three-dimensional motion data in the three-dimensional motion data is taken as the target numerical point, and the Pearson coefficient between the target numerical point and the other collected three-dimensional motion data is calculated, and the various Pearson coefficients are arranged in order from high to low, and then the Pearson coefficients arranged in the first x are divided by 1 to obtain each quotient value, and then the average of the various quotient values ​​is calculated, so as to use the average as the partial aggregation degree of the target numerical point; x can be set according to specific requirements.

[0061] The quotient obtained by dividing the partial aggregation degree of the target numerical point by the average of the partial aggregation degrees of other three-dimensional motion data corresponding to the previous x Pearson coefficients is taken as the starting benchmark score, thereby obtaining the starting benchmark score of the three-dimensional motion data. When the starting benchmark score is higher than one, the target numerical point is an outlier; when the starting benchmark score is close to one, the target numerical point is a reasonable point; when the starting benchmark score is lower than one, the target numerical point is a hidden outlier.

[0062] In this application, the Z-score method is used, and when faced with the collected three-dimensional motion data, the benchmark score can be calculated based on the Pearson coefficient between the numerical points. The three-dimensional motion data benchmark score obtained by the Z-score method can reflect the corresponding aggregation degree between the nearby numerical points, so as to determine the benchmark score of the numerical point based on the corresponding aggregation degree.

[0063] In a preferred but non-limiting embodiment of the present invention, in step 1, a near zone of each numerical point in the three-dimensional motion data is defined in advance, and the near zone of the numerical point contains a numerical point one of which the number is set in advance to y. The numerical point one of which the number is set in advance to y is the first y numerical points whose Pearson coefficients between other numerical points in the three-dimensional motion data other than the numerical point to which the near zone belongs and the numerical point to which the near zone belongs are arranged in order from high to low. The value of y can be thirty, so the benchmark score of the j-th numerical point obtained based on this is the starting benchmark score B. j .

[0064] Step 2, obtaining the acquisition speed of each numerical point in the three-dimensional motion data and the numerical point in the near zone, obtaining the acquisition benchmark score of each numerical point in the near zone according to the initial benchmark score and the acquisition speed, calculating the variance of all real-time three-dimensional motion data as the global fluctuation amplitude of the three-dimensional motion data, and obtaining the partial fluctuation amplitude of each numerical point according to the acquisition speed of each numerical point and the acquisition speed of the numerical point in the near zone and the average of each numerical point and the three-dimensional motion data;

[0065] When faced with three-dimensional motion data, in this application, the acquisition speeds of different numerical points are different. When the acquisition speed is not low, the distribution aggregation degree of the numerical points is not high, which will make the benchmark score of the numerical point higher. However, when faced with such numerical points, it is still impossible to determine whether they are outliers, so it is necessary to calibrate the benchmark scores of the numerical points to remove the effect of the acquisition speed on the benchmark scores of the numerical points in order to obtain a benchmark score that can better characterize the outliers of the three-dimensional motion data.

[0066] The three-dimensional motion data collected for the first time is obtained through the same collection speed, so the collection speed of all numerical points is the same. With the successive collection and analysis of the three-dimensional motion data, the collection speed of the three-dimensional motion data will be changed, so the collection speed of the three-dimensional motion data will change thereafter.

[0067] In a preferred but non-limiting embodiment of the present invention, step 2 specifically comprises:

[0068] Step 2-1: Calculate the quotient of the cumulative amount of the acquisition speed of each numerical point in the near area corresponding to each numerical point and the acquisition speed of the numerical point, and multiply the quotient by the initial reference score as the acquisition reference score;

[0069] In a preferred but non-limiting embodiment of the present invention, in step 2-1, the calculation equation for collecting the benchmark score is:

[0070]

[0071] In the equation, H j represents the acquisition benchmark score of the jth value point, B j represents the starting benchmark score of the j-th numerical point, l represents the number of numerical points contained in the numerical point cluster formed by all numerical points in the corresponding nearby area of ​​the j-th numerical point, T j,k represents the acquisition speed of the kth value point in the value point cluster corresponding to the jth value point, T j Represents the acquisition speed of the jth numerical point.

[0072] Only when the collection speed between the numerical point and other numerical points in the nearby area is different can the benchmark scores of the surrounding numerical points be disturbed. When the collection speed of the numerical point is not low compared to the collection speed of other data in the nearby numerical point cluster, the distribution aggregation degree of the numerical point will also be not low, making the benchmark score higher. Therefore, the benchmark score of the numerical point should be changed to a lower level. When the collection speed of the numerical point is not high compared to the collection speed of other numerical points in the nearby numerical point cluster, the distribution aggregation degree of the numerical point will also be not high. Therefore, the benchmark score of the numerical point should be changed to a higher level.

[0073] Step 2-2: Select an arbitrary numerical point as the target numerical point, calculate the subtraction of the mean of the three-dimensional motion data of the target numerical point and other numerical points in the nearby area, and obtain the quotient of the acquisition speed of the target numerical point and the acquisition speed of an arbitrary numerical point in the nearby area corresponding to the target numerical point; take the square root of the average of the cumulative amount of the square of the amount obtained by multiplying the subtraction and the quotient as the partial fluctuation amplitude of the target numerical point.

[0074] In a preferred but non-limiting embodiment of the present invention, in step 2-2, the calculation equation of the partial fluctuation amplitude is:

[0075]

[0076] In the equation, represents the partial fluctuation amplitude at the jth three-dimensional motion data, l represents the number of numerical points in the corresponding near-area numerical point cluster of the jth numerical point, T 0 represents the predefined starting data acquisition speed, T j,k represents the acquisition speed of the kth value point in the cluster of value points in the vicinity corresponding to the jth value point, E j,k represents the value of the three-dimensional motion data of the kth value point in the cluster of value points near the jth value point, ν j Represents the average of all 3D motion data in the near-area numerical point cluster of the j-th 3D motion data.

[0077] and It means that when facing random three-dimensional motion data, the distance between the three-dimensional motion data of each numerical point in the numerical point cluster in the corresponding near area and the average of the three-dimensional motion data can be obtained, and the acquisition speed of each numerical point is set with additional parameters, which fully takes into account the differences in the acquisition speeds of each numerical point and obtains more accurate partial fluctuation amplitude.

[0078] Step 3, calculating the acquisition speed variation coefficient of each numerical point according to the acquisition benchmark score and the partial fluctuation amplitude;

[0079] In a preferred but non-limiting embodiment of the present invention, in step 3, the calculation equation of the acquisition speed variation coefficient is:

[0080]

[0081] In the equation, G j Represents the acquisition speed variation coefficient of the jth value point, H j represents the acquisition benchmark score of the jth value point, H 0 Represents a predefined collection benchmark score benchmark, Represents the global fluctuation amplitude of the 3D motion data, Represents the partial fluctuation amplitude at the jth three-dimensional motion data.

[0082] In a preferred but non-limiting embodiment of the present invention, in step 3, the benchmark amount of the acquisition benchmark score is ninety percent.

[0083] If the partial fluctuation amplitude is lower than the global fluctuation amplitude, the partial fluctuation amplitude will appropriately reduce the benchmark score of the value point to make the benchmark score more accurate. Then, the acquisition benchmark score after the change of the attached amount is compared with the acquisition benchmark score benchmark amount to obtain the acquisition speed change coefficient.

[0084] According to the pre-defined acquisition benchmark score baseline, if the acquisition benchmark score in the three-dimensional motion data is higher than the acquisition benchmark score baseline, a variation coefficient higher than one shall be set to increase the acquisition speed of the three-dimensional motion data; otherwise, if the acquisition benchmark score in the three-dimensional motion data is not higher than the acquisition benchmark score baseline, the acquisition speed of the three-dimensional motion data shall be reduced.

[0085] Step 4, taking the acquisition speed variation coefficient as the variation parameter of the acquisition speed corresponding to each numerical point, refreshing the acquisition speed of each numerical point, performing acquisition according to the refreshed acquisition speed, obtaining the outliers of the three-dimensional motion data and removing the outliers, performing analysis on the three-dimensional motion data with the outliers removed, thereby obtaining the motion state data.

[0086] However, a minimum acquisition speed is defined in advance in this application. When the acquisition speed of the three-dimensional motion data is lower than the minimum acquisition speed, it cannot be guaranteed that the acquired three-dimensional motion data can be parsed into accurate motion state data. That is, no matter how the acquisition speed of the three-dimensional motion data is changed, it cannot be lower than the minimum acquisition speed.

[0087] In a preferred but non-restrictive implementation of the present invention, in step 4, the variable parameter is multiplied by the current numerical point acquisition speed to obtain the value obtained as the refreshed numerical point acquisition speed. Based on the refreshed numerical point acquisition speed, the three-dimensional motion data is immediately acquired, and the acquisition benchmark score of each numerical point of the acquired three-dimensional motion data is calculated. When there are a pre-defined number of numerical point acquisition benchmark scores that are continuously higher than the acquisition benchmark score benchmark amount, it is determined that the three-dimensional motion data has outliers, and the Z-score method is used to obtain the outlier value of the three-dimensional motion data.

[0088] In a preferred but non-limiting embodiment of the present invention, in step 4, the variable parameter is multiplied by the current value point acquisition speed as the refreshed value point acquisition speed, and the next three-dimensional motion data is acquired according to the refreshed acquisition speed;

[0089] It is assumed that the predefined number U is seventeen and the collection benchmark score benchmark is ninety percent.

[0090] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include:

[0091] By obtaining real-time three-dimensional motion data for pre-processing, the acquisition speed is actively changed according to the reliability of the real-time three-dimensional motion data to ensure accurate analysis of the three-dimensional motion data, reducing the workload of accurate analysis of the three-dimensional motion data and ensuring the reliability of accurate analysis of the three-dimensional motion data. By calibrating each numerical point in the three-dimensional motion data and removing the disturbance of the acquisition speed on the acquisition benchmark score of the numerical point, a more accurate acquisition benchmark score of the three-dimensional motion data is obtained. The higher the acquisition benchmark score, the more outliers the three-dimensional motion data will produce. By actively changing the acquisition speed of the three-dimensional motion data, more accurate analysis of the three-dimensional motion data can be achieved.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not deviate from the spirit and scope of the present invention should be included in the protection scope of the rights requirements of the present invention.

Claims

1. A module for inputting three-dimensional motion data, characterized in that: include: The motion acquisition device is connected to the central control device. The motion acquisition device is used to instantly acquire the three-dimensional motion data of the object to be measured and transmit it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured. The motion acquisition device is a module that inputs the three-dimensional motion data. The unit running on the motion acquisition device includes: A standard unit, which is used to obtain real-time 3D motion data for pre-processing, perform outlier detection on the 3D motion data using the Z-score method, and obtain a starting benchmark score for the 3D motion data; An amplitude unit is used to obtain the acquisition speed of each numerical point in the three-dimensional motion data and the numerical point in the near zone, obtain the acquisition benchmark score of each numerical point in the near zone according to the initial benchmark score and the acquisition speed, calculate the variance of all real-time three-dimensional motion data as the global fluctuation amplitude of the three-dimensional motion data, and obtain the partial fluctuation amplitude of each numerical point according to the acquisition speed of each numerical point and the acquisition speed of the numerical point in the near zone and the average of each numerical point and the three-dimensional motion data; A calculation unit, which is used to calculate the acquisition speed variation coefficient of each numerical point according to the acquisition reference score and the partial fluctuation amplitude; A refresh unit is used to treat the acquisition speed variation coefficient as a variation parameter of the acquisition speed corresponding to each numerical point, refresh the acquisition speed of each numerical point, perform acquisition according to the refreshed acquisition speed, obtain outliers of the three-dimensional motion data and remove the outliers, and perform analysis on the three-dimensional motion data with the outliers removed to obtain motion state data.

2. A method for inputting three-dimensional motion data, characterized in that: include: The motion acquisition device collects the three-dimensional motion data of the object to be measured in real time and transmits it to the central control device. The central control device is used to analyze the collected three-dimensional motion data to obtain motion state data. The three-dimensional motion data includes the speed, acceleration or angular velocity of the object to be measured; The period during which the motion acquisition device instantly acquires the three-dimensional motion data of the object to be measured also includes: Step 1, obtain real-time 3D motion data for pre-processing, use the Z-score method to perform outlier detection on the 3D motion data, and obtain the initial benchmark score of the 3D motion data; Step 2, obtaining the acquisition speed of each numerical point in the three-dimensional motion data and the numerical point in the near zone, obtaining the acquisition benchmark score of each numerical point in the near zone according to the initial benchmark score and the acquisition speed, calculating the variance of all real-time three-dimensional motion data as the global fluctuation amplitude of the three-dimensional motion data, and obtaining the partial fluctuation amplitude of each numerical point according to the acquisition speed of each numerical point and the acquisition speed of the numerical point in the near zone and the average of each numerical point and the three-dimensional motion data; Step 3, calculating the acquisition speed variation coefficient of each numerical point according to the acquisition benchmark score and the partial fluctuation amplitude; Step 4, taking the acquisition speed variation coefficient as the variation parameter of the acquisition speed corresponding to each numerical point, refreshing the acquisition speed of each numerical point, performing acquisition according to the refreshed acquisition speed, obtaining the outliers of the three-dimensional motion data and removing the outliers, performing analysis on the three-dimensional motion data with the outliers removed, thereby obtaining the motion state data.

3. The method for inputting three-dimensional motion data according to claim 1, characterized in that: In step 1, a minimum acquisition speed is defined in advance, and a minimum acquisition speed T0 is defined in advance; In step 1, the three-dimensional motion data is collected. The collected three-dimensional motion data is the real-time three-dimensional motion data. The j-th three-dimensional motion data collected is defined as E j , define the number of all three-dimensional motion data collected as p.

4. The method for inputting three-dimensional motion data according to claim 3, characterized in that: In step 1, the collected three-dimensional motion data is benchmarked using the Z-score method, that is, any three-dimensional motion data in the three-dimensional motion data is taken as the target value point, the Pearson coefficient between the target value point and the other collected three-dimensional motion data is calculated, and each Pearson coefficient is arranged in order from high to low, and then the Pearson coefficients arranged in the first x are divided by 1 to obtain each quotient value, and then the average of each quotient value is calculated, so as to take the average as the partial aggregation degree of the target value point; The quotient obtained by dividing the partial aggregation degree of the target numerical point by the average of the partial aggregation degrees of other three-dimensional motion data corresponding to the previous x Pearson coefficients is taken as the initial benchmark score, thereby obtaining the initial benchmark score of the three-dimensional motion data. When the initial benchmark score is higher than one, the target numerical point is an outlier; In step 1, the near area of ​​each numerical point in the three-dimensional motion data is defined in advance, and the near area of ​​the numerical point contains a numerical point one of which the number is set in advance to y. The numerical point one of which the number is set in advance to y is the first y numerical points whose Pearson coefficients between other numerical points in the three-dimensional motion data other than the numerical point belonging to the near area and the numerical point belonging to the near area are arranged in order from high to low. The benchmark score of the j-th numerical point obtained based on this is the starting benchmark score B j .

5. The method for inputting three-dimensional motion data according to claim 4, characterized in that: Step 2 specifically includes: Step 2-1: Calculate the quotient of the cumulative amount of the acquisition speed of each numerical point in the near area corresponding to each numerical point and the acquisition speed of the numerical point, and multiply the quotient by the initial reference score as the acquisition reference score; Step 2-2: Select an arbitrary numerical point as the target numerical point, calculate the subtraction of the mean of the three-dimensional motion data of the target numerical point and other numerical points in the nearby area, and obtain the quotient of the acquisition speed of the target numerical point and the acquisition speed of an arbitrary numerical point in the nearby area corresponding to the target numerical point; take the square root of the average of the cumulative amount of the square of the amount obtained by multiplying the subtraction and the quotient as the partial fluctuation amplitude of the target numerical point.

6. The method for inputting three-dimensional motion data according to claim 5, characterized in that: In step 2-1, the calculation equation for collecting the benchmark score is: In the equation, H j represents the acquisition benchmark score of the jth value point, B j represents the starting benchmark score of the j-th numerical point, l represents the number of numerical points contained in the numerical point cluster formed by all numerical points in the corresponding nearby area of ​​the j-th numerical point, T j,k represents the acquisition speed of the kth value point in the value point cluster corresponding to the jth value point, T j Represents the acquisition speed of the jth value point; In step 2-2, the calculation equation of the partial fluctuation amplitude is: In the equation, represents the partial fluctuation amplitude at the jth three-dimensional motion data, l represents the number of numerical points in the corresponding near-area numerical point cluster of the jth numerical point, T0 represents the predefined starting data acquisition speed, T j,k represents the acquisition speed of the kth value point in the cluster of value points in the vicinity corresponding to the jth value point, E j,k represents the value of the three-dimensional motion data of the kth value point in the cluster of value points near the jth value point, ν j Represents the average of all three-dimensional motion data in the near-area numerical point cluster of the j-th three-dimensional motion data.

7. The method for inputting three-dimensional motion data according to claim 6, characterized in that: In step 3, the calculation equation of the acquisition speed variation coefficient is: In the equation, G j Represents the acquisition speed variation coefficient of the jth value point, H j represents the collection benchmark score of the jth value point, H0 represents the pre-defined collection benchmark score benchmark, Represents the global fluctuation amplitude of the 3D motion data, Represents the partial fluctuation amplitude at the jth three-dimensional motion data.

8. The method for inputting three-dimensional motion data according to claim 7, characterized in that: In step 3, the base score is collected with a base amount of ninety percent.

9. The method for inputting three-dimensional motion data according to claim 8, characterized in that: In step 4, the variable parameter is multiplied by the current value point acquisition speed to obtain the value obtained as the refreshed value point acquisition speed. Based on the refreshed value point acquisition speed, the three-dimensional motion data is collected in real time, and the acquisition benchmark score of each value point of the collected three-dimensional motion data is calculated. When there are a pre-defined number of value point acquisition benchmark scores that are higher than the acquisition benchmark score benchmark, it is determined that the three-dimensional motion data has outliers, and the Z-score method is used to obtain the outlier value of the three-dimensional motion data.

10. The method for inputting three-dimensional motion data according to claim 9, characterized in that: In step 4, the variable parameter is multiplied by the current value point acquisition speed, which is used as the refreshed value point acquisition speed, and the next three-dimensional motion data is acquired according to the refreshed acquisition speed; It is assumed that the predefined number U is seventeen and the collection benchmark score benchmark is ninety percent.

Images