Ball hitting posture recognition method, wearable intelligent device and computer storage medium

Abstract

The invention relates to the technical field of motion posture recognition, in particular to a ball hitting posture recognition method, wearable intelligent equipment and a computer storage medium. The method comprises the steps that in a first detection window, ball hitting judgment data are determined, and the ball hitting judgment data comprise the standard deviation of the acceleration of all axes of a user ball hitting hand, the standard deviation of the total angular velocity energy of the user ball hitting hand and the total angular velocity energy of the user ball hitting hand at each sampling point; setting second-level criteria based on the ball hitting judgment data, and determining that the user has a first ball hitting action when any one of the second-level criteria is established; setting a second detection window by taking a moment corresponding to the first ball hitting action as a center moment; and in the second detection window, determining the hitting posture of the user based on the acceleration data and the angular velocity data of each axis of the hitting hand of the user. According to the method, the ball hitting posture of the user is accurately judged through the user related data collected by the sensor in real time.

Description

Technical Field

[0001] This application relates to the technical field of motion posture recognition, and further relates to a method for recognizing a hitting posture, a wearable intelligent device, and a computer storage medium. Background Art

[0002] In the training of ball games such as tennis, recording data such as the hitting postures of users is crucial for evaluating their comprehensive physical fitness and athletic ability. These data can also help formulate targeted training plans to improve the performance of athletes. Therefore, accurately recognizing the hitting posture of athletes during hitting is of great importance.

[0003] Although a variety of intelligent wearable devices have emerged on the market, such as energy bracelets, pedometer shoes, etc., they mainly provide basic health data such as heart rate, number of steps, and moving distance. However, these devices cannot accurately determine the hitting posture of users based on the user-related data collected in real time by sensors. Summary of the Invention

[0004] To solve the above technical problems, this application provides a method for recognizing a hitting posture, a wearable intelligent device, and a computer storage medium, which can accurately determine the hitting posture of a user.

[0005] In a first aspect, this application provides a method for recognizing a hitting posture, including: determining hitting determination data within a first detection window, where the hitting determination data includes the standard deviation of the accelerations of each axis of the user's hitting hand, the standard deviation of the total energy of the angular velocity of the user's hitting hand, and the total energy of the angular velocity of the user's hitting hand at each sampling point; setting a two-level criterion based on the hitting determination data, and when any one of the criteria in the two-level criterion is satisfied, determining that the user has a first hitting action; setting a second detection window with the moment corresponding to the first hitting action as the central moment; and determining the hitting posture of the user within the second detection window based on the acceleration data of each axis and the angular velocity data of each axis of the user's hitting hand.

[0006] This method for recognizing a hitting posture can accurately determine the hitting action of the user by comprehensively analyzing the standard deviation of the accelerations of each axis of the user's hitting hand, the standard deviation of the total energy of the angular velocity, and the total energy of the angular velocity at each sampling point within the first detection window, and then setting a two-level criterion. When any one of the criteria is satisfied, it can be confirmed that the user has a first hitting action. Subsequently, with the moment corresponding to the first hitting action as the center, a second detection window is set, and within this window, the acceleration data of each axis and the angular velocity data of each axis of the user are further analyzed, so as to accurately determine the hitting posture of the user, providing more accurate data support for sports training and technical analysis, and helping to optimize the training plan and improve the performance of athletes.

[0007] In one implementation, within the second detection window, based on the acceleration data and angular velocity data of each axis of the user's batter, determining the batting posture of the user specifically includes: obtaining the time interval between the first batting action and the second batting action immediately before the first batting action; within the second detection window, based on the y-axis angular velocity data and z-axis angular velocity data of each sampling point before the central moment, determining the first-order integral of the y-axis angular velocity and the first-order integral of the z-axis angular velocity of each sampling point before the central moment; within the second detection window, based on the x-axis acceleration data and z-axis acceleration data of each sampling point before the central moment, determining the first-order integral of the x-axis acceleration and the first-order integral of the z-axis acceleration of each sampling point before the central moment; when the time interval is greater than the preset time interval, the first-order integrals of the y-axis angular velocity and the z-axis angular velocity of each sampling point before the central moment are both greater than zero, and the first-order integrals of the x-axis acceleration and the z-axis acceleration of each sampling point before the central moment are both less than zero, determining that the batting posture of the user is serving.

[0008] This batting posture recognition method can accurately obtain the time interval between the first batting action and the previous second batting action, and combine the y-axis angular velocity data and z-axis angular velocity data, as well as the x-axis acceleration data and z-axis acceleration data before the central moment within the second detection window, to determine the first-order integrals of the angular velocity and acceleration of each sampling point, and then accurately judge that the batting posture of the user is serving. At the same time, it helps the user analyze and improve their own strength, as well as predict the game result.

[0009] In one implementation, it further includes: when the time interval is greater than the preset time interval, the number of positive numbers in the first-order integral results of the y-axis angular velocity and the z-axis angular velocity of all sampling points before the central moment is greater than the number of negative numbers, the first-order integral of the z-axis acceleration of each sampling point before the central moment is less than zero, and the first-order integral of the x-axis acceleration of each sampling point before the central moment is less than the first threshold, determining that the batting posture of the user is serving.

[0010] In one implementation, within the second detection window, based on the acceleration data and angular velocity data of each axis of the user's hitting hand, determining the hitting posture of the user specifically includes: within the second detection window, based on the z-axis acceleration data of each sampling point before the central moment, determining the first-order integral of the z-axis acceleration of each sampling point before the central moment; when the first-order integral of the z-axis acceleration of each sampling point before the central moment is not greater than the minimum value of the first threshold range, determining that the hitting posture of the user is forehand; when the first-order integral of the z-axis acceleration of each sampling point before the central moment is not less than the maximum value of the first threshold range, determining that the hitting posture of the user is backhand; when the first-order integral of the z-axis acceleration of each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data of all sampling points within the second detection window is within the y-axis angular velocity data, determining that the hitting posture of the user is forehand; when the first-order integral of the z-axis acceleration of each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data of all sampling points within the second detection window is not within the y-axis angular velocity data, determining that the hitting posture of the user is backhand.

[0011] This hitting posture recognition method can effectively distinguish whether the user's hitting posture is forehand or backhand by analyzing the z-axis acceleration data before the central moment within the second detection window, calculating the first-order integral of the z-axis acceleration of each sampling point, and combining the first threshold range, while improving the accuracy and reliability of hitting posture recognition.

[0012] In one implementation, within the second detection window, based on the acceleration data and angular velocity data of each axis of the user's hitting hand, determining the hitting posture of the user specifically includes: within the second detection window, based on the z-axis angular velocity data of each sampling point, determining the first-order integral of the z-axis angular velocity of each sampling point; when the user holds the racket with a backhand grip and the minimum value of the angular velocity data of all sampling points within the second detection window is within the x-axis angular velocity data, or when the user holds the racket with a backhand grip and within the second detection window, the first-order integral of the z-axis angular velocity of each sampling point is greater than the first threshold, determining that the hitting posture of the user is topspin; when the user holds the racket with a backhand grip, the minimum value of the first-order integral of the angular velocity data of all sampling points within the second detection window is within the first-order integral of the z-axis angular velocity, and the first-order integral of the z-axis angular velocity of each sampling point is less than the third threshold, determining that the hitting posture of the user is slice.

[0013] In one implementation, it further includes: when the user holds the racket with the right hand and after the first hitting action occurs, if the change amount of the y-axis angular velocity and the change amount of the z-axis angular velocity have different signs, it is determined that the hitting posture of the user is topspin; when the user holds the racket with the right hand and after the first hitting action occurs, if the change amount of the y-axis angular velocity and the change amount of the z-axis angular velocity have the same sign and meet any one of the preset conditions, it is determined that the hitting posture of the user is slice; wherein, the preset conditions include determining the maximum value of the x-axis angular velocity data, the minimum value of the y-axis angular velocity data, and the maximum value of the z-axis angular velocity data corresponding to all sampling points in sequence within the second detection window, the maximum value of the first-order integral of the z-axis corresponding to all sampling points within the second detection window is less than the fourth threshold, and the absolute value of the maximum value of the first-order integral of the z-axis corresponding to all sampling points within the second detection window is less than the absolute value of the minimum value of the first-order integral of the z-axis corresponding to all sampling points within the second detection window.

[0014] For the user who holds the racket with the left hand, this hitting posture recognition method can determine whether the hitting posture is topspin or slice by comparing the relationship between the minimum value of the angular velocity data and the first-order integral of the z-axis angular velocity with the first threshold and the third threshold. For the user who holds the racket with the right hand, by analyzing the sign relationship between the change amounts of the y-axis and z-axis angular velocities and combining the preset conditions, the hitting posture can be further accurately judged, enhancing the understanding and evaluation of the athlete's technical movements, and helping the user optimize the training strategy according to these accurate data and improve the game performance.

[0015] In one implementation, based on the hitting determination data, a secondary criterion is set, which specifically includes: within the first detection window, taking the standard deviation of the accelerations of each axis of the user's hitting hand being greater than the fifth threshold as the first-level criterion; within the first detection window, taking the standard deviation of the accelerations of each axis of the user's hitting hand being greater than the sixth threshold, the standard deviation of the total angular velocity energy of the user's hitting hand being not less than the seventh threshold, and the total angular velocity energy of the user's hitting hand at each sampling point being not less than the eighth threshold as the second-level criterion.

[0016] In one implementation, it further includes: calculating the swing speed of the user based on the user's body data; counting the total number of the user's hits and the number of hits in each hitting posture, and counting the active time of the user based on the acceleration data, angular velocity data, and the central moment of the user during the movement process.

[0017] In a second aspect, the present application further provides a wearable intelligent device, including a memory and a processor, where a computer program is stored on the memory, and when the computer program is executed by the processor, the steps of the hitting posture recognition method in any of the above implementations are implemented.

[0018] In a third aspect, the present application also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of implementing the above-mentioned batting posture recognition method are realized.

[0019] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0020] 1. By comprehensively analyzing the standard deviation of the accelerations of each axis of the user's batting hand, the standard deviation of the total energy of the angular velocity, and the total energy of the angular velocity at each sampling point within the first detection window, the user's batting action can be accurately determined. Then, a secondary criterion is set. When any one of the criteria is met, it can be confirmed that the user has a first batting action. Subsequently, with the moment corresponding to the first batting action as the center, a second detection window is set, and within this window, the acceleration data of each axis and the angular velocity data of each axis of the user are further analyzed, so as to accurately determine the user's batting posture, providing more accurate data support for sports training and technical analysis, and helping to optimize the training plan and improve the performance of athletes.

[0021] 2. By accurately obtaining the time interval between the first batting action and the previous second batting action, and combining the y-axis angular velocity data and z-axis angular velocity data before the central moment within the second detection window, as well as the x-axis acceleration data and z-axis acceleration data, the first integral of the angular velocity and acceleration at each sampling point can be determined, and then it can be accurately judged that the user's batting posture is serving. At the same time, it helps the user analyze and improve their own strength, as well as predict the game result.

[0022] 3. By analyzing the z-axis acceleration data before the central moment within the second detection window and calculating the first integral of the z-axis acceleration at each sampling point, and combining the first threshold range, it can effectively distinguish whether the user's batting posture is forehand or backhand, and at the same time improve the accuracy and reliability of batting posture recognition.

[0023] 4. For users with a backhand grip, by comparing the minimum value of the angular velocity data and the relationship between the first integral of the z-axis angular velocity and the first threshold and the third threshold, it can be determined whether the batting posture is topspin or slice. For users with a forehand grip, by analyzing the sign relationship of the change in the y-axis and z-axis angular velocities and combining the preset conditions, the batting posture can be further accurately judged, enhancing the understanding and evaluation of the athlete's technical movements, and helping the user optimize the training strategy according to these accurate data and improve the game performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above characteristics, technical features, advantages and their implementation manners of the present invention will be further described below in a clear and understandable manner in combination with the drawings in the preferred embodiments.

[0025] Figure 1Shows a flowchart of a batting posture recognition method provided by an embodiment of the present application;

[0026] Figure 2 Shows a flowchart of recognizing a batting posture as a serve in an embodiment of the present application;

[0027] Figure 3 Shows a flowchart of recognizing a batting posture as forehand or backhand in an embodiment of the present application;

[0028] Figure 4 Shows a flowchart of recognizing a batting posture as topspin or slice in an embodiment of the present application. Detailed implementation manners

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the specific implementation manners of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts, and other implementation manners can also be obtained.

[0030] To make the drawings concise, only the parts related to the invention are schematically shown in each drawing, and they do not represent the actual structure of the product. In addition, to make the drawings concise and easy to understand, in some drawings, components with the same structure or function are only schematically shown for one of them, or only one of them is marked. In this document, "one" not only means "only this one", but also means "more than one" situation.

[0031] It should also be further understood that the term "and / or" used in the specification and claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0032] In this document, it should be noted that, unless otherwise clearly defined and limited, the terms "install", "connect", and "couple" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0033] In addition, in the description of the present application, the terms "first", "second", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0034] It should be noted that the above embodiments can be freely combined as needed. The above is only the preferred embodiment of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

[0035] Wearable intelligent devices integrate a variety of sensors to provide functions such as health monitoring, motion tracking, and environmental perception. These sensors include, but are not limited to, biosensors (heart rate sensors such as PPG and ECG, blood glucose sensors, blood pressure sensors, body temperature sensors, electroencephalogram sensors) for real-time monitoring of physiological parameters and health status; environmental sensors (temperature and humidity sensors, ultraviolet sensors, PM2.5 sensors, gas sensors, pH sensors, barometric pressure sensors) to monitor the external environment and provide health reminders; global positioning system (GPS) for positioning and navigation; optical heart rate monitoring for convenient heart rate monitoring; pressure sensors to measure changes in skin surface pressure and evaluate mental state and life stress; motion sensors (accelerometers, gyroscopes) to monitor steps, sleep quality, motion trajectories, and posture changes.

[0036] As a competitive sport, the diversity of the technical movements in tennis is mainly demonstrated through forehand, backhand, topspin, slice, serve and other techniques. In order to comprehensively evaluate the comprehensive sports ability of users during tennis movement, it is necessary to accurately identify the movement postures of users during tennis movement, so that users can view the recognition results in a timely manner during or after the game to improve their own strength. In the embodiments of the present application, the accelerometer and gyroscope on the wearable intelligent device are used to collect the acceleration data and angular velocity data of the user during tennis movement in real time, and these acceleration data and angular velocity data are processed and analyzed, and at least one of the following beneficial effects can be achieved: accurately judging the user's hitting posture; or, helping the user analyze and improve their own strength, and predicting the game result.

[0037] The following is elaborated in conjunction with the accompanying drawings:

[0038] Refer to the attached Figure 1 , which shows a flowchart of a hitting posture recognition method provided by an embodiment of the present application. As Figure 1 shown, it includes:

[0039] S100, within a first detection window, determine hitting determination data, where the hitting determination data includes the standard deviation of the accelerations of each axis of the user's hitting hand, the standard deviation of the total angular velocity energy of the user's hitting hand, and the total angular velocity energy of the user's hitting hand at each sampling point.

[0040] S110, based on the hitting determination data, set secondary criteria, and when any one of the secondary criteria is satisfied, determine that the user has a first hitting action.

[0041] S120, set a second detection window with the moment corresponding to the first hitting motion as the central moment.

[0042] S130, within the second detection window, determine the user's hitting posture based on the acceleration data and angular velocity data of each axis of the user's hitting hand.

[0043] Before the user starts tennis exercise, first determine the user's hitting hand, and wear the wearable smart device on the user's hitting hand, so that the wearable smart device can obtain the acceleration data and angular velocity data of the user during the exercise in real time through the accelerometer and gyroscope.

[0044] The first detection window (or the first sliding detection window) is essentially a preset time period, which is adjustable. Essentially, when analyzing whether the user has a hitting motion, each adjacent first detection window is analyzed one by one. For example, when the user does not have a hitting motion in the current first detection window, the acceleration data and angular velocity data of the user in the next adjacent first detection window are analyzed in real time to determine whether the user has a hitting motion.

[0045] After the user wears the wearable smart device on the user's hitting hand, in the first detection window, obtain the acceleration data and angular velocity data of each axis corresponding to each sampling point in real time according to the preset sampling frequency. Based on the acceleration data of each axis (x-axis acceleration data, y-axis acceleration data, and z-axis angular velocity data) corresponding to each sampling point in the first detection window, calculate the standard deviation of the acceleration of each axis (x-acceleration standard deviation, y-axis acceleration standard deviation, and z-axis acceleration standard deviation) within the first detection window. Based on the angular velocity data of each axis corresponding to each sampling point in the first detection window, calculate the algebraic sum of the angular velocity data corresponding to each sampling point (the sum of the x-axis angular velocity data, y-axis angular velocity data, and z-axis angular velocity data), that is, obtain the total angular velocity energy corresponding to each sampling point. Furthermore, based on the total angular velocity energy corresponding to each sampling point, calculate the standard deviation of the total angular velocity energy within the first detection window.

[0046] Set the first-level criterion based on the standard deviation of the accelerations of each axis within the first detection window, and set the second-level criterion based on the standard deviation of the accelerations of each axis within the first detection window, the standard deviation of the total angular velocity energy within the first detection window, and the total angular velocity energy corresponding to each sampling point within the first detection window. When any one of the second-level criteria is satisfied, it is determined that the user has a first hitting action. Meanwhile, with the moment corresponding to the first hitting action as the central moment and extending a preset time period before and after (the time periods before and after the central moment can be the same or different), construct a second detection window (or a second sliding detection window). Further, based on the acceleration data of each axis and the angular velocity data of each axis of the user's hitting hand corresponding to each sampling point within the second detection window, determine the user's hitting posture.

[0047] In the embodiment of the present application, by comprehensively analyzing the standard deviation of the accelerations of each axis of the user's hitting hand, the standard deviation of the total angular velocity energy, and the total angular velocity energy of each sampling point within the first detection window, the hitting action of the user can be accurately determined. Then, the second-level criterion is set. When any one of the criteria is satisfied, it can be confirmed that the user has a first hitting action. Subsequently, with the moment corresponding to the first hitting action as the center, a second detection window is set, and within this window, the acceleration data of each axis and the angular velocity data of each axis of the user are further analyzed, so as to accurately determine the user's hitting posture, provide more accurate data support for sports training and technical analysis, and help optimize the training plan and improve the performance of athletes.

[0048] In an embodiment of the present application, it further includes: setting a second-level criterion based on the hitting determination data, specifically including: within the first detection window, taking the standard deviation of the accelerations of each axis of the user's hitting hand being greater than a fifth threshold as the first-level criterion; within the first detection window, taking the standard deviation of the accelerations of each axis of the user's hitting hand being greater than a sixth threshold, the standard deviation of the total angular velocity energy of the user's hitting hand being not less than a seventh threshold, and the total angular velocity energy of the user's hitting hand at each sampling point being not less than an eighth threshold as the second-level criterion.

[0049] Reference attached Figure 2 shows a flowchart for identifying a hitting posture as a serve provided by an embodiment of the present application. As Figure 2 shown, it includes:

[0050] S200, obtain the time interval between the first hitting action and the second hitting action immediately before the first hitting action.

[0051] S210, within the second detection window, based on the y-axis angular velocity data and z-axis angular velocity data of each sampling point before the central moment, determine the first-order integral of the y-axis angular velocity and the first-order integral of the z-axis angular velocity of each sampling point before the central moment.

[0052] S220. Within the second detection window, based on the x-axis acceleration data and z-axis acceleration data of each sampling point before the central moment, determine the first-order integral of the x-axis acceleration and the first-order integral of the z-axis acceleration of each sampling point before the central moment.

[0053] S230. When the time interval is greater than the preset time interval, the first-order integral of the y-axis angular velocity and the first-order integral of the z-axis angular velocity of each sampling point before the central moment are both greater than zero, and the first-order integral of the x-axis acceleration and the first-order integral of the z-axis acceleration of each sampling point before the central moment are both less than zero, determine that the user's hitting posture is a serve.

[0054] In tennis, a serve is often the first hit (by the serving side) in each intensive hitting segment of a game, and there are usually links such as the player taking a short break to adjust, changing the ball, and getting ready between each intensive hitting segment. That is, there will be a certain time interval between two serving actions (for example, this time interval can be 5 seconds). Therefore, after determining that the user has a first hitting action, it is necessary to determine the time interval between the first hitting action and the previous second hitting action, and use this time interval as a judgment basis.

[0055] Most tennis serves are high-pressure serves with an obvious downward pressing action of the wrist. Therefore, use the first-order integral of the y-axis angular velocity and the first-order integral of the z-axis angular velocity of each sampling point before the central moment within the second detection window as a judgment basis.

[0056] When serving in tennis, the hand makes a rapid swinging motion from top to bottom with a fixed movement trajectory. Therefore, use the first-order integral of the x-axis acceleration and the first-order integral of the z-axis acceleration of each sampling point before the central moment within the second detection window as a judgment basis.

[0057] When the time interval between the first hitting action and the previous second hitting action is greater than the preset time interval, the first-order integral of the y-axis angular velocity of each sampling point before the central moment is greater than zero and the first-order integral of the z-axis angular velocity of each sampling point before the central moment is greater than zero, and the first-order integral of the x-axis acceleration of each sampling point before the central moment is less than zero and the first-order integral of the z-axis acceleration of each sampling point before the central moment is less than zero, determine that the user's hitting posture is a serve.

[0058] In the embodiments of the present application, by accurately obtaining the time interval between the first hitting action and the previous second hitting action, and combining the y-axis angular velocity data and z-axis angular velocity data, as well as the x-axis acceleration data and z-axis acceleration data, before the central moment within the second detection window, the first-order integral of the angular velocity and acceleration of each sampling point can be determined, and then it can be accurately judged that the user's hitting posture is a serve. This method helps users analyze and improve their own strength, as well as predict the game results.

[0059] In an embodiment of the present application, it further includes: when the time interval is greater than a preset time interval, the number of positive numbers in the first-order integral results of the y-axis angular velocity and the first-order integral results of the z-axis angular velocity at all sampling points before the central moment is greater than the number of negative numbers, the first-order integral of the z-axis acceleration at each sampling point before the central moment is less than zero, and the first-order integral of the x-axis acceleration at each sampling point before the central moment is less than a first threshold, it is determined that the user's hitting posture is a serve.

[0060] Count the number of positive numbers and the number of negative numbers in the results of the first-order integral of the y-axis angular velocity at all sampling points before the central moment within the second detection window, and count the number of positive numbers and the number of negative numbers in the results of the first-order integral of the z-axis angular velocity at all sampling points before the central moment within the second detection window. Compare the sum of the number of positive numbers on the y-axis and the z-axis with the sum of the number of negative numbers on the y-axis and the z-axis as a supplementary judgment basis.

[0061] Due to factors such as personal hitting habits, when serving, there may be obvious movements of the racket from top to bottom and then up, as well as movements from right to left. Therefore, it can be used as a supplementary judgment basis that the first-order integral of the x-axis acceleration at each sampling point before the central moment within the second detection window is less than the first threshold.

[0062] Furthermore, when the time interval is greater than a preset time interval, the sum of the number of positive numbers on the y-axis and the z-axis is greater than the sum of the number of negative numbers on the y-axis and the z-axis, the first-order integral of the z-axis acceleration at each sampling point before the central moment is less than zero, and the first-order integral of the x-axis acceleration at each sampling point before the central moment is less than the first threshold, it is determined that the user's hitting posture is a serve.

[0063] Refer to the appendix Figure 3 , which shows a flowchart for identifying forehand and backhand hitting postures provided by an embodiment of the present application. As Figure 3 shown, it includes:

[0064] S300, within the second detection window, based on the z-axis acceleration data at each sampling point before the central moment, determine the first-order integral of the z-axis acceleration at each sampling point before the central moment.

[0065] S310, when the first-order integral of the z-axis acceleration at each sampling point before the central moment is not greater than the minimum value of the first threshold range, determine that the user's hitting posture is a forehand.

[0066] S320, when the first-order integral of the z-axis acceleration at each sampling point before the central moment is not less than the maximum value of the first threshold range, determine that the user's hitting posture is a backhand.

[0067] S330. When the first-order integral of the z-axis acceleration at each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data at all sampling points within the second detection window is within the y-axis angular velocity data, it is determined that the user's hitting posture is a forehand.

[0068] S340. When the first-order integral of the z-axis acceleration at each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data at all sampling points within the second detection window is not within the y-axis angular velocity data, it is determined that the user's hitting posture is a backhand.

[0069] When the user holds the racket with the right hand and makes a forehand swing, the z-axis of the wearable smart device is reversed; when the user holds the racket with the right hand and makes a backhand swing, the z-axis of the wearable smart device is forward. Therefore, the first-order integral of the z-axis acceleration at each sampling point before the central moment within the second detection window can be used as the basis for judging forehand and backhand. That is, when the first-order integral of the z-axis acceleration at each sampling point before the central moment within the second detection window is not greater than the minimum value of the first threshold range, it is determined that the user's hitting posture is a forehand; when the first-order integral of the z-axis acceleration at each sampling point before the central moment within the second detection window is not less than the maximum value of the first threshold range, it is determined that the user's hitting posture is a backhand.

[0070] When the result of the first-order integral of the z-axis acceleration at each sampling point before the central moment within the second detection window is not obvious (that is, the first-order integral of the z-axis acceleration at each sampling point before the central moment within the second detection window is within the first threshold range), considering the wrist will have an obvious inward bending movement during the racket-receiving stage of various forehand movements. Therefore, the minimum value of the angular velocity data at all sampling points within the second detection window can be used as the basis for judgment. That is, the x-axis angular velocity data, y-axis angular velocity data, and z-axis angular velocity data at each sampling point within the second detection window are obtained through the gyroscope, and all the angular velocity data are compared to determine the minimum value. When the first-order integral of the z-axis acceleration at each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data at all sampling points within the second detection window is within the y-axis angular velocity data, it is determined that the user's hitting posture is a forehand; when the first-order integral of the z-axis acceleration at each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data at all sampling points within the second detection window is not within the y-axis angular velocity data, it is determined that the user's hitting posture is a backhand.

[0071] By analyzing the z-axis acceleration data before the central moment within the second detection window, calculating the first-order integral of the z-axis acceleration at each sampling point, and combining the first threshold range, the embodiments of the present application can effectively distinguish whether the user's hitting posture is a forehand or a backhand, and at the same time improve the accuracy and reliability of hitting posture recognition.

[0072] Refer to the appendixFigure 4 , which shows a flowchart for identifying a forehand or slice stroke provided by an embodiment of the present application. As Figure 4 shown, it includes:

[0073] S400, within the second detection window, based on the z-axis angular velocity data of each sampling point, determine the first integral of the z-axis angular velocity of each sampling point.

[0074] S410, when the user holds the racket in a backhand grip and the minimum value of the angular velocity data of all sampling points within the second detection window is within the x-axis angular velocity data, or when the user holds the racket in a backhand grip and the first integral of the z-axis angular velocity of each sampling point within the second detection window is greater than the first threshold, determine that the user's stroke is a forehand.

[0075] S420, when the user holds the racket in a backhand grip, the minimum value of the first integral of the angular velocity data of all sampling points within the second detection window is within the first integral of the z-axis angular velocity, and the first integral of the z-axis angular velocity of each sampling point is less than the third threshold, determine that the user's stroke is a slice.

[0076] After determining the first stroke action and setting the second detection window, calculate the first integral of the z-axis angular velocity of each sampling point within the second detection window, and determine which axis the minimum value of the angular velocity data of all sampling points within the second detection window and the minimum value of the first integral of the angular velocity data of all sampling points within the second detection window are respectively in. Furthermore, when the user holds the racket in a backhand grip and the minimum value of the angular velocity data of all sampling points within the second detection window is within the x-axis angular velocity data, determine that the user's stroke is a forehand; or when the user holds the racket in a backhand grip and the first integral of the z-axis angular velocity of each sampling point within the second detection window is greater than the first threshold, determine that the user's stroke is a forehand.

[0077] When the user holds the racket in a backhand grip, the minimum value of the first integral of the angular velocity data of all sampling points within the second detection window is within the first integral of the z-axis angular velocity, and the first integral of the z-axis angular velocity of each sampling point is less than the third threshold, determine that the user's stroke is a slice.

[0078] In an embodiment of the present application, it further includes: when the user holds the racket with the right hand and after the first hitting action occurs, if the change amount of the y-axis angular velocity and the change amount of the z-axis angular velocity have different signs, it is determined that the user's hitting posture is topspin; when the user holds the racket with the right hand and after the first hitting action occurs, if the change amount of the y-axis angular velocity and the change amount of the z-axis angular velocity have the same sign and satisfy any one of the preset conditions, it is determined that the user's hitting posture is slice; wherein, the preset conditions include determining the maximum value of the x-axis angular velocity data, the minimum value of the y-axis angular velocity data, and the maximum value of the z-axis angular velocity data corresponding to all sampling points in sequence within the second detection window, the maximum value of the first-order integral of the z-axis corresponding to all sampling points within the second detection window is less than the fourth threshold, and the absolute value of the maximum value of the first-order integral of the z-axis corresponding to all sampling points within the second detection window is less than the absolute value of the minimum value of the first-order integral of the z-axis corresponding to all sampling points within the second detection window.

[0079] In the embodiment of the present application, for a user holding the racket with the left hand, by comparing the minimum value of the angular velocity data and the relationship between the first-order integral of the z-axis angular velocity and the first threshold and the third threshold, it can be determined that the hitting posture is topspin or slice. For a user holding the racket with the right hand, by analyzing the sign relationship of the change amounts of the y-axis and z-axis angular velocities and combining the preset conditions, the hitting posture can be further accurately judged, enhancing the understanding and evaluation of the athlete's technical movements, and helping the user optimize the training strategy according to these accurate data and improve the competition performance.

[0080] In an embodiment of the present application, it further includes: calculating the user's racket swing speed based on the user's body data; counting the total number of the user's hits and the number of hits in each hitting posture, and counting the user's active time based on the acceleration data, angular velocity data, and central moment during the user's movement.

[0081] Based on the user's body data, such as gender and height, etc., the user's racket swing speed can be calculated. Count the total number of the user's hits during the movement and the number of hits in each hitting posture, and based on the acceleration data, angular velocity data, and central moment during the user's movement, determine whether the user is in an active state and count the corresponding active time.

[0082] Among them, characteristics such as the resultant acceleration modulus, the algebraic sum of angular velocities, and the resultant acceleration standard deviation can be obtained through the acceleration data and angular velocity data, and corresponding empirical thresholds can be set to determine whether the user is in an active state. After determining the active state, the active time can be counted according to the sampling frequencies of the accelerometer and gyroscope. If it is detected that the user has the first hitting action, then the user must be in an active state at this time, and the duration of the second detection window can be directly counted into the active time. After the user finishes the movement, the statistical results and recognition results are displayed to the user through the wearable intelligent device.

[0083] An embodiment of the present application further provides a wearable intelligent device, including a memory and a processor. A computer program is stored on the memory, and when the computer program is executed by the processor, the steps of the batting posture recognition method described in any one of the above embodiments are implemented.

[0084] An embodiment of the present application further provides a computer storage medium, on which a computer program is stored, and when the computer program is executed by the processor, the steps of the batting posture recognition method described in any one of the above embodiments are implemented.

[0085] It should be noted that the above embodiments can be freely combined as needed. The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A batting posture recognition method, characterized in that: include: Determining, within the first detection window, hitting determination data, the hitting determination data including standard deviations of accelerations of each axis of the user hitter, standard deviations of the total angular velocity energy of the user hitter, and the total angular velocity energy of the user hitter at each sampling point; Setting secondary criteria based on the hitting determination data, and determining that the user has a first hitting action when any primary criterion in the secondary criteria is satisfied; Setting a second detection window with the time corresponding to the first hitting action as the center time; In the second detection window, the user's batting posture is determined based on the acceleration data of each axis and the angular velocity data of each axis of the user's batting hand.

2. The batting posture recognition method according to claim 1, characterized in that: Determining the batting posture of the user based on the acceleration data of each axis and the angular velocity data of each axis of the batting hand of the user within the second detection window specifically includes: Acquire a time interval between the first hitting action and a second hitting action before the first hitting action; In the second detection window, based on the y-axis angular velocity data and the z-axis angular velocity data of each sampling point before the center moment, determine the first-order integral of the y-axis angular velocity and the first-order integral of the z-axis angular velocity of each sampling point before the center moment; In the second detection window, based on the x-axis acceleration data and the z-axis acceleration data of each sampling point before the center moment, determine the first-order integral of the x-axis acceleration and the first-order integral of the z-axis acceleration of each sampling point before the center moment; When the time interval is greater than the preset time interval, the first-order integral of the y-axis angular velocity and the first-order integral of the z-axis angular velocity of each sampling point before the center moment are both greater than zero, and the first-order integral of the x-axis acceleration and the first-order integral of the z-axis acceleration of each sampling point before the center moment are both less than zero, it is determined that the user's batting posture is a serve.

3. The batting posture recognition method according to claim 2, characterized in that: Also includes: When the time interval is greater than the preset time interval, the number of positive numbers in the first-order integration results of the y-axis angular velocity and the first-order integration results of the z-axis angular velocity of all sampling points before the center moment is greater than the number of negative numbers, the first-order integral of the z-axis acceleration of each sampling point before the center moment is less than zero, and the first-order integral of the x-axis acceleration of each sampling point before the center moment is less than a first threshold, it is determined that the user's batting posture is a serve.

4. The batting posture recognition method according to claim 1, characterized in that: Determining the batting posture of the user based on the acceleration data of each axis and the angular velocity data of each axis of the batting hand of the user within the second detection window specifically includes: In the second detection window, based on the z-axis acceleration data of each sampling point before the center moment, determine the first-order integral of the z-axis acceleration of each sampling point before the center moment; When the first-order integral of the z-axis acceleration of each sampling point before the center moment is not greater than the minimum value of the first threshold range, determining that the user's batting posture is a forehand; When the first-order integral of the z-axis acceleration of each sampling point before the center moment is not less than the maximum value of the first threshold range, determining that the user's batting posture is a backhand; When the first-order integral of the z-axis acceleration of each sampling point before the central moment is within the first threshold range and the minimum value of the angular velocity data of all sampling points in the second detection window is within the y-axis angular velocity data, determining that the user's batting posture is a forehand; When the first-order integral of the z-axis acceleration of each sampling point before the center moment is within the first threshold range and the minimum value of the angular velocity data of all sampling points in the second detection window is not in the y-axis angular velocity data, it is determined that the user's batting posture is backhand.

5. The batting posture recognition method according to claim 1, characterized in that: Determining the batting posture of the user based on the acceleration data of each axis and the angular velocity data of each axis of the batting hand of the user within the second detection window specifically includes: In the second detection window, based on the z-axis angular velocity data of each sampling point, determine the first-order integral of the z-axis angular velocity of each sampling point; When the user holds the racket with a backhand and the minimum value of the angular velocity data of all sampling points in the second detection window is in the x-axis angular velocity data, or when the user holds the racket with a backhand and the first-order integral of the z-axis angular velocity of each sampling point in the second detection window is greater than a first threshold, it is determined that the user's hitting posture is topspin; When the user holds the racket with his backhand, the minimum value of the first-order integral of the angular velocity data of all sampling points in the second detection window is in the first-order integral of the z-axis angular velocity, and the first-order integral of the z-axis angular velocity of each sampling point is less than the third threshold, it is determined that the user's batting posture is a slice.

6. The method for identifying a batting posture according to claim 5, characterized in that: Also includes: When the user holds the racket in the forehand and the first hitting action occurs, the y-axis angular velocity change and the z-axis angular velocity change have different signs, and it is determined that the user's hitting posture is topspin; When the user holds the racket with his forehand and the first hitting action occurs, the y-axis angular velocity change and the z-axis angular velocity change have the same sign and meet any one of the preset conditions, determining that the user's hitting posture is a slice; Among them, the preset condition includes determining the maximum value of the x-axis angular velocity data, the minimum value of the y-axis angular velocity data, and the maximum value of the z-axis angular velocity data corresponding to all sampling points in the second detection window in sequence, the maximum value of the z-axis first-order integral corresponding to all sampling points in the second detection window is less than a fourth threshold, and the absolute value of the maximum value of the z-axis first-order integral corresponding to all sampling points in the second detection window is less than the absolute value of the minimum value of the z-axis first-order integral corresponding to all sampling points in the second detection window.

7. The method for identifying a batting posture according to any one of claims 1 to 6, characterized in that: The second-level criteria are set based on the hitting determination data, specifically including: In the first detection window, the standard deviation of the acceleration of each axis of the user's batsman is greater than a fifth threshold value, which is used as a first-level criterion; In the first detection window, the standard deviation of the acceleration of each axis of the user batter is greater than the sixth threshold, the standard deviation of the total angular velocity energy of the user batter is not less than the seventh threshold, and the total angular velocity energy of the user batter at each sampling point is not less than the eighth threshold, as the second-level judgment criteria.

8. The method for identifying a batting posture according to claim 7, characterized in that: Also includes: Calculating the user's swing speed based on the user's physical data; The total number of shots of the user and the number of shots in each shot posture are counted, and the active time of the user is counted based on the acceleration data, angular velocity data and the center time of the user during the movement.

9. A wearable smart device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the steps of the batting posture recognition method according to any one of claims 1 to 8 are implemented.

10. A computer storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the steps of the batting posture recognition method according to any one of claims 1 to 8 are implemented.

Images