Ball speed measuring method and device, electronic equipment and readable storage medium

Abstract

The invention discloses a ball speed measuring method and device, electronic equipment and a readable storage medium, and belongs to the technical field of electronics. The method comprises the following steps: acquiring ball hitting motion data and ball hitting sound data when a motion object performs ball hitting motion; acquiring ball hitting action feature information and ball hitting action energy feature information corresponding to the ball hitting motion data, and ball hitting sound feature information corresponding to the ball hitting sound data; based on the ball hitting action feature information, the ball hitting action energy feature information and the sound feature information, the ball hitting speed of the moving object during ball hitting is determined.

Description

Technical Field

[0001] This application belongs to the field of electronic technology, and particularly relates to a ball speed measurement method, device, electronic device, and readable storage medium. Background Art

[0002] With the rise of all-round fitness, ball sports have flourished globally. For example, badminton, tennis, table tennis, etc. Generally, in ball sports using rackets, players usually use rackets combined with hitting techniques to increase the ball speed. For beginners, it is not easy to adjust the way of exerting force, hitting posture, and hitting timing to effectively transmit force and ultimately increase the ball speed. Moreover, since beginners cannot intuitively understand the ball speed, it is difficult to improve the hitting technique targeted in subsequent training. Therefore, how to accurately measure the ball speed is of great significance for assisting ball enthusiasts in improving their hitting technique level.

[0003] In related technologies, radar speed measurement technology, optical speed measurement technology, infrared motion compensation speed measurement technology, etc. are usually used to measure the ball speed. However, the above-mentioned radar speed measurement technology and infrared motion compensation speed measurement technology are not suitable for the daily training of ordinary athletes due to high equipment costs and complex operations. And the above-mentioned optical speed measurement technology usually has high requirements for the lighting environment of the sports venue. When the lighting in the sports venue is insufficient, the accuracy of the ball speed measured by the optical speed measurement technology is relatively low.

[0004] Therefore, how to provide a ball speed measurement method with simple operation and high measurement accuracy is an urgent problem to be solved. Summary of the Invention

[0005] The purpose of the embodiments of this application is to provide a ball speed measurement method, device, electronic device, and readable storage medium, which reduce the cost and improve the stability of ball speed measurement, and can assist players in making more accurate improvements to their hitting techniques.

[0006] In a first aspect, the embodiments of this application provide a ball speed measurement method, which includes: collecting hitting motion data and hitting sound data when a moving object performs a hitting motion; obtaining hitting motion feature information and hitting motion energy feature information corresponding to the hitting motion data, and hitting sound feature information corresponding to the hitting sound data; and determining the hitting ball speed when the moving object performs the hitting motion based on the hitting motion feature information, hitting motion energy feature information, and sound feature information.

[0007] In a second aspect, an embodiment of the present application provides a ball speed measuring device, which includes an acquisition module and a processing module. The acquisition module is configured to acquire the hitting motion data and the hitting sound data when a moving object performs a hitting motion. The processing module is configured to obtain the hitting motion feature information and the hitting motion energy feature information corresponding to the hitting motion data acquired by the acquisition module, and the hitting sound feature information corresponding to the hitting sound data. The processing module is further configured to determine the hitting ball speed when the moving object performs a hitting motion based on the hitting motion feature information, the hitting motion energy feature information, and the sound feature information.

[0008] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor and a memory. The memory stores a program or instruction that can run on the processor. When the program or instruction is executed by the processor, the steps of the method described in the first aspect are implemented.

[0009] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0010] In a fifth aspect, an embodiment of the present application provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor, and the processor is configured to run a program or instruction to implement the method described in the first aspect.

[0011] In a sixth aspect, an embodiment of the present application provides a computer program product, which is stored in a storage medium and is executed by at least one processor to implement the method described in the first aspect.

[0012] In the embodiment of the present application, the hitting motion data and the hitting sound data when a moving object performs a hitting motion are acquired; the hitting motion feature information and the hitting motion energy feature information corresponding to the hitting motion data, and the hitting sound feature information corresponding to the hitting sound data are obtained; and the hitting ball speed when the moving object performs a hitting motion is determined based on the above-mentioned hitting motion feature information, hitting motion energy feature information, and sound feature information. Through this solution, since the hitting sound data can characterize whether the hitting hits the sweet spot of the racket when the moving object performs a hitting motion, when predicting the hitting ball speed of the moving object during the hitting motion, by combining and analyzing the hitting motion feature information and the hitting motion energy feature information corresponding to the hitting motion data of the moving object during the hitting motion with the sound feature information corresponding to the hitting sound data of the moving object during the hitting motion, the analyzed ball speed is more accurate, and the stability of the ball speed measurement quality is improved. Description of the Drawings

[0013] Figure 1 It is a schematic flowchart of the ball speed measurement method provided by some embodiments of the present application;

[0014] Figure 2 It is a schematic diagram of the ball speed range of the ball speed measurement method provided by some embodiments of the present application under a forehand flat hit;

[0015] Figure 3 It is one of the schematic diagrams of the peak data acquisition moment of the ball speed measurement method provided by some embodiments of the present application;

[0016] Figure 4 It is a schematic diagram of the ball speed measurement result of the ball speed measurement method provided by some embodiments of the present application;

[0017] Figure 5 It is a schematic diagram of the ball speed prediction model of the ball speed measurement method provided by some embodiments of the present application;

[0018] Figure 6 It is a schematic flowchart of the ball speed calculation of the ball speed measurement method provided by some embodiments of the present application;

[0019] Figure 7 It is a schematic diagram of the ball speed measurement device provided by some embodiments of the present application;

[0020] Figure 8 It is a schematic diagram of the electronic device provided by some embodiments of the present application;

[0021] Figure 9 It is a schematic diagram of the hardware structure of the electronic device provided by some embodiments of the present application. Detailed implementation manners

[0022] Next, the technical solutions in the embodiments of the present application will be clearly described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, rather than all, of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present application.

[0023] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are usually of the same category, and the number of objects is not limited. For example, the first object can be one or multiple. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally represents an "or" relationship between the associated objects before and after.

[0024] The terms "at least one (item)", "at least one of", etc. in the description and claims of this application refer to any one, any two or more combinations of their included objects. For example, at least one (item) of a, b, and c can represent: "a", "b", "c", "a and b", "a and c", "b and c", and "a, b, and c", where a, b, and c can be single or multiple. Similarly, "at least two (items)" means two or more, and its meaning is similar to that of "at least one (item)".

[0025] The following will combine the accompanying drawings and explain in detail the ball speed measurement method provided by the embodiments of this application through specific embodiments and their application scenarios.

[0026] The ball speed measurement method provided by the embodiments of this application can be applied to scenarios where a racket is used for hitting a ball.

[0027] In the ball speed measurement scenario, a ball speed measurement device is usually installed on the hitting court. When a player hits the ball on the court, this ball speed measurement device can be used to capture the ball speed of the player at different times, and these captured ball speeds can then be used to assist the player in improving the hitting skills. Conventional ball speed measurement devices are bulky, costly, and complex to operate, requiring professional maintenance, and can only be used in venues with professional sports training needs, making it difficult to popularize among ordinary hitting enthusiasts. Moreover, some ball speed measurement devices have high requirements for the site environment. For example, an infrared speedometer that is sensitive to infrared interference signals and an optical speedometer whose speed measurement quality deteriorates in dim light. Therefore, the quality of ball speed measurement is unstable.

[0028] In the embodiments of this application, the hitting motion data and hitting sound data of a moving object during a hitting motion are collected; the hitting motion feature information and hitting motion energy feature information corresponding to the hitting motion data, and the hitting sound feature information corresponding to the hitting sound data are obtained; based on the above-mentioned hitting motion feature information, hitting motion energy feature information, and sound feature information, the hitting ball speed of the moving object during the hitting motion is determined. Through this solution, since the hitting sound data during the hitting motion of the moving object can characterize whether the hitting hits the sweet spot of the racket, when predicting the hitting ball speed of the moving object during the hitting motion in this application, by combining and analyzing the hitting motion feature information and hitting motion energy feature information corresponding to the hitting motion data of the moving object during the hitting motion with the sound feature information corresponding to the hitting sound data of the moving object during the hitting motion, the analyzed ball speed is more accurate, improving the stability of the ball speed measurement quality.

[0029] The execution subject of the ball speed measurement method provided by the embodiments of the present application can be a ball speed measurement device, which can be an electronic device, or a functional module or processing module in the electronic device. The present application does not make any limitations in this regard. In some embodiments of the present application, taking the electronic device as the execution subject to execute the ball speed method as an example, the ball speed measurement method provided by the embodiments of the present application is described.

[0030] Exemplarily, the electronic device can be a portable electronic device, and the portable device can be a wearable device, such as a smart bracelet, a smart watch, etc. The portable device can also be an electronic device including a wrist sensing device. The portable device can also be a portable electronic device that can be flexibly fixed on other sports equipment, such as a portable device that can be installed on a tennis racket, a badminton racket, or a table tennis racket.

[0031] The embodiments of the present application provide a ball speed measurement method. Figure 1 The flowchart of a ball speed measurement method provided by the embodiments of the present application is shown. As Figure 1 shown, the ball speed measurement method provided by the embodiments of the present application may include the following steps 201 to 203.

[0032] Step 201: The electronic device collects the hitting motion data and hitting sound data when the moving object performs a hitting motion.

[0033] In some embodiments of the present application, the above-mentioned hitting motion can be various hitting motions using a racket, such as table tennis, badminton, tennis, softball, baseball, etc.

[0034] In some embodiments of the present application, the hitting motion data when the moving object performs a hitting motion can be collected by the wearable electronic device of the moving object when the moving object performs a hitting motion.

[0035] Exemplarily, the hitting motion data when the moving object performs a hitting motion can be collected by the portable electronic device installed on the racket used by the moving object when the moving object performs a hitting motion. Or, the hitting motion data when the moving object performs a hitting motion can be collected by the portable electronic device worn on the body part of the moving object when the moving object performs a hitting motion. For example, the above-mentioned body part can be the wrist, arm, finger, etc.

[0036] In some embodiments of the present application, the above-mentioned hitting motion data can be collected by the following components in the portable electronic device: an acceleration sensor, a gyroscope sensor, a magnetometer, or the motion data collected by other inertial measurement units. Specifically, it can be determined according to the actual device situation, and the embodiments of the present application do not make any limitations.

[0037] Example 1: The hitting motion data may include ACC data collected by an acceleration sensor (ACC).

[0038] Example 2: The hitting motion data may include GYRO data collected by a gyroscope sensor (GYRO).

[0039] Example 3: The hitting motion data may include motion data in six axial directions collected by a six-axis sensor.

[0040] In some embodiments of the present application, the above-mentioned hitting motion data is the motion data of the moving object during the hitting motion of the moving object.

[0041] In some embodiments of the present application, the above-mentioned hitting sound data may be collected by a sound collection device, such as collected by a microphone in an electronic device; or, collected by a microphone in a portable electronic device for collecting the above-mentioned hitting motion data.

[0042] In some embodiments of the present application, the above-mentioned hitting sound data is collected during the hitting motion of the moving object. It can be understood that since the above-mentioned hitting motion data and the above-mentioned hitting sound data are collected at the same time, and the sound of the sphere hitting the racket can reflect the hitting situation of the moving object during the hitting motion. Therefore, the above-mentioned hitting sound data can characterize the hitting information of the racket during the process of the moving object hitting the ball with the racket, such as missing the racket, hitting the exact center of the sweet spot, hitting the edge of the sweet spot, hitting the edge of the racket, etc.

[0043] In some embodiments of the present application, the above-mentioned hitting motion data includes the hitting motion data of the moving object during the hitting motion in the first time period; the above-mentioned hitting sound data includes the hitting motion data of the moving object during the hitting motion in the first time period. Exemplarily, if the above-mentioned first time period includes M consecutive sampling moments, the above-mentioned hitting motion data includes the motion data collected at each sampling moment among the M consecutive sampling moments.

[0044] Step 202: The electronic device obtains the hitting action feature information and hitting action energy feature information corresponding to the hitting motion data, and the hitting sound feature information corresponding to the hitting sound data.

[0045] In some embodiments of the present application, the above-mentioned hitting action feature information is used to characterize the hitting action features of the moving object during the hitting motion.

[0046] In some embodiments of the present application, the above-mentioned hitting action features include at least one of the following: the way of swinging the racket when the moving object performs the hitting motion, the amplitude frequency of the racket when the moving object performs the hitting motion, the hitting force when the moving object performs the hitting motion.

[0047] In some embodiments of the present application, the above-mentioned hitting action feature information at least includes: the feature information of the racket swing action of the moving object, and the racket swing action feature is used to characterize the racket swing mode when the moving object performs a hitting motion. Exemplarily, the racket swing mode includes at least one of the following: forehand flat hit, forehand topspin, backhand flat hit, backhand topspin, forehand slice, backhand slice.

[0048] It can be understood that since the ball speed has a strong relationship with the racket swing mode when the moving object performs a hitting motion, therefore, different racket swing modes correspond to different ball speed distributions. Generally, serve ball speed > forehand flat hit ball speed > forehand topspin ball speed > backhand flat hit ball speed > backhand topspin ball speed > forehand slice ball speed > backhand slice ball speed. The specific ball speed range distribution corresponding to each racket swing mode is shown in Table 1 below. Thus, after obtaining the racket swing mode of the moving object, the ball speed can be measured with the assistance of the above-mentioned racket swing mode.

[0049]

[0050] Table 1

[0051] In some embodiments of the present application, the above-mentioned hitting action energy feature information is used to characterize the swing amplitude when the moving object performs a hitting motion.

[0052] It can be understood that since the swing amplitude can characterize the combined energy of the hitting action during the hitting process of the moving object, therefore, the above-mentioned hitting action energy feature information can be used to characterize the combined energy of the hitting action when the moving object performs a hitting motion.

[0053] It can be understood that for the same racket swing mode, the ball speeds generated at different swing amplitudes are also different. For example, when ensuring that each hit hits the sweet spot of the racket, the larger the swing amplitude, the easier the ball speed is to be distributed in the range of faster speeds.

[0054] For example, as Figure 2 shown, for the moving object using the forehand flat hit for racket swing, when the combined energy measured by the acceleration sensor is about 100, the ball speed is in the range of 60 - 70; when the combined energy measured by the acceleration sensor is about 200, the ball speed is in the range of 100 - 110; when the combined energy measured by the acceleration sensor is about 250, the ball speed is in the range of 140 - 145.

[0055] Thus, by combining the swing amplitude for ball speed measurement, the obtained ball speed measurement result is more accurate. In some embodiments of the present application, the "obtaining the hitting action feature information corresponding to the hitting motion data" in step 202 above can be implemented through the following step 202a:

[0056] Step 202a: Obtain the hitting motion feature information corresponding to the hitting motion data based on the hitting behavior information and hitting motion data of the moving object.

[0057] In some embodiments of the present application, the above-mentioned hitting behavior information is used to characterize the racket-holding type of the moving object, and the racket-holding type includes: holding the racket with the left hand, or holding the racket with the right hand.

[0058] It can be understood that since the judgment results of the swinging styles obtained with different hands holding the racket are opposite. For example, for the same motion data, if the swinging style obtained based on holding the racket with the left hand is a forehand flat hit, then the swinging style obtained based on holding the racket with the right hand is a backhand flat hit. Therefore, in the present application, by combining and analyzing the hitting behavior information and hitting motion data of the moving object, the hitting motion feature information that is more in line with the actual hitting motion of the moving object can be obtained, so as to obtain a more accurate ball speed measurement result.

[0059] In this way, the correct hitting motion feature information can be obtained according to the above-mentioned hitting behavior information.

[0060] In some embodiments of the present application, the above-mentioned hitting sound feature information is used to characterize the sound feature of the hitting sound when the moving object performs the hitting motion.

[0061] In some embodiments of the present application, the sound features of the above-mentioned hitting sound include at least one of the following: pitch, timbre, decibel, sound duration.

[0062] It can be understood that since the hitting sound usually includes the sound generated when the ball hits the racket, and the sounds generated when the ball hits different positions of the racket are different, and the sounds generated when the racket is hit by the ball at different ball speeds are also different. Therefore, the hitting ball speed of the moving object can be measured through the hitting sound feature information.

[0063] It should be noted that the above-mentioned hitting sound usually refers to the hitting sound when the moving object hits the ball. The above-mentioned hitting ball speed usually refers to the hitting ball speed when the moving object hits the ball.

[0064] It should be noted that under other unchanged conditions, the ball speed generated when hitting the ball at the exact center of the sweet spot of the racket is greater than the ball speed generated when hitting the ball at the edge of the sweet spot of the racket, and the ball speed generated when hitting the ball at the edge of the sweet spot of the racket is greater than the ball speed generated when hitting the ball with the edge of the racket. And since the sound features generated when different positions of the racket hit the ball are different. For example, when hitting the ball at the exact center of the sweet spot, the generated sound is dull and long, while when hitting the ball with the racket frame, the generated sound is sharp and short. In addition, the louder the hitting sound, the faster the ball speed. In this way, through the sound feature information for ball speed measurement, the obtained ball speed measurement result is more accurate.

[0065] In some embodiments of the present application, the above step 202 can be implemented through the following steps 202b to 202d.

[0066] Step 202b: The electronic device filters the hitting sound data to obtain effective hitting sound data.

[0067] Step 202c: The electronic device converts the effective hitting sound data into a Mel spectrogram.

[0068] Step 202d: The electronic device extracts the hitting action feature information and the hitting action energy feature information corresponding to the effective hitting motion data in the hitting motion data, and extracts the hitting sound feature information corresponding to the hitting sound data from the Mel spectrogram.

[0069] In some embodiments of the present application, the above effective hitting motion data is determined based on the acquisition time of the peak data in the hitting motion data.

[0070] In some embodiments of the present application, before the electronic device extracts the hitting sound feature information corresponding to the hitting sound data, it can perform frequency-domain filtering on the hitting sound data, and extract the hitting sound data within a predetermined frequency band range as the effective hitting sound data. Wherein, the predetermined frequency band range is related to the frequency band range of the sound generated when the racket hits the ball. In this way, by performing frequency-domain filtering on the hitting sound data, interference noises such as wind noise, venue noise, and human voices in the hitting sound can be effectively filtered out, so as to obtain a more accurate hitting sound, and thus a more accurate ball speed prediction result can be obtained.

[0071] In some embodiments of the present application, when the electronic device performs time-domain filtering on the hitting sound data, it can perform windowing processing on the hitting sound data, and perform short-time Fourier transform on the windowed hitting sound data after the windowing processing is completed to complete the time-domain filtering, so as to obtain effective hitting sound data.

[0072] In some embodiments of the present application, since the Mel frequency is a non-linear frequency scale based on the human ear's auditory characteristics, therefore, in the present application, by converting the effective hitting sound data into a Mel spectrogram, that is, converting the linear frequency in the audio signal into the Mel frequency, the obtained Mel spectrogram after conversion can better simulate the human ear's frequency perception. In other words, by converting the original audio signal into a feature representation that is more suitable for human auditory perception, the extracted hitting sound feature information is more accurate.

[0073] In some embodiments of the present application, the electronic device can use the acquisition time of the peak data in the hitting motion data as the center point, and use the acquisition time of the peak data and the hitting motion data corresponding to one or more consecutive acquisition times before and after the acquisition time of the peak data as the effective hitting motion data.

[0074] Exemplarily, after the electronic device detects the acquisition moment of the peak data in the hitting motion data, that is, after detecting the hitting event that satisfies the acceleration signal threshold, it will extract the hitting motion data of 50 sampling points near the acquisition moment of the peak data corresponding to the hitting event as the effective hitting motion data.

[0075] It can be understood that each peak data in the above-mentioned hitting motion data corresponds to a hitting event. That is to say, the motion data corresponding to each hitting event included in the hitting motion data is used as the above-mentioned effective hitting motion data.

[0076] In some embodiments of the present application, the electronic device can determine the acquisition moment of the peak data in the hitting motion data by calculating the resultant acceleration of the hitting motion data at each moment, and then combining the calculated resultant acceleration of the hitting motion data at each moment with a preset threshold. Generally, as Figure 3 shown, when a hitting event occurs, the peak value of the magnitude of the resultant acceleration will increase significantly, generally greater than 5 times the acceleration due to gravity.

[0077] Specifically, the acquisition moment of the peak data in the hitting motion data can be determined by the following formula 1:

[0078] A x 2 +A y 2 +A z 2 >5g (Formula 1)

[0079] Wherein, A x 、A y 、A z are respectively used to represent the measured values of the acceleration sensor in three axial directions, g is the acceleration due to gravity, and the sum of A x 2 、A y 2 、A z 2 is the resultant acceleration.

[0080] In this way, by calculating the resultant acceleration of the hitting motion data at each moment, the occurrence moment of the hitting event can be selected more accurately, so as to achieve a more accurate ball speed measurement result.

[0081] Step 203: Based on the above-mentioned hitting action feature information, hitting action energy feature information, and sound feature information, determine the hitting ball speed when the moving object performs a hitting motion.

[0082] In the embodiments of the present application, when a moving object performs a hitting motion, hitting motion data and hitting sound data are collected; hitting action feature information and hitting action energy feature information corresponding to the hitting motion data, and hitting sound feature information corresponding to the hitting sound data are obtained; based on the above-mentioned hitting action feature information, hitting action energy feature information, and sound feature information, the hitting ball speed when the moving object performs the hitting motion is determined. Through this solution, since the hitting sound data when the moving object performs the hitting motion can characterize whether the hitting hits the sweet spot of the racket, therefore, when predicting the hitting ball speed when the moving object performs the hitting motion in the present application, by combining and analyzing the hitting action feature information and hitting action energy feature information corresponding to the hitting motion data when the moving object performs the hitting motion with the sound feature information corresponding to the hitting sound data when the moving object performs the hitting motion, the analyzed ball speed is more accurate, and the stability of the ball speed measurement quality is improved.

[0083] In some embodiments of the present application, after the above step 201, the ball speed measurement method provided by the embodiments of the present application further includes the following step 301. In combination with step 301, the above step 202d can be implemented through the following step 302, and the above step 203 can be implemented through the following step 303a and step 302b:

[0084] Step 301: The electronic device inputs the effective hitting motion data and the Mel spectrogram into the ball speed prediction model.

[0085] In some embodiments of the present application, the above ball speed prediction model includes a convolutional layer, a feature fusion module, and a fully connected layer.

[0086] Step 302: The electronic device extracts the hitting action feature information and hitting action energy feature information corresponding to the effective hitting motion data through the convolutional layer in the ball speed prediction model, and extracts the hitting sound feature information corresponding to the hitting sound data from the Mel spectrogram.

[0087] Step 303a: The electronic device fuses the effective hitting motion data, hitting action feature information, hitting action energy feature information, and sound feature information through the feature fusion module in the ball speed prediction model to obtain fused feature information.

[0088] In some embodiments of the present application, the above feature fusion module only fuses the hitting action feature information, hitting action energy feature information, and sound feature information to obtain fused feature information.

[0089] Step 303b: The electronic device calculates and outputs the hitting ball speed when the moving object performs the hitting motion based on the fused feature information through the fully connected layer in the ball speed prediction model.

[0090] In some embodiments of the present application, the above ball speed prediction model can be a neural network model.

[0091] In some embodiments of the present application, the above ball speed prediction model can be a two-dimensional convolutional neural network (CNN) model.

[0092] In some embodiments of the present application, after the above effective hitting motion data and Mel spectrogram are input into the ball speed prediction model, they will be input into different convolutional layers in the ball speed prediction model, and corresponding feature information will be extracted through different convolutional layers.

[0093] In one example, the above hitting motion feature information and hitting motion energy feature information are extracted through different convolutional layers in the ball speed prediction model.

[0094] In one example, the above hitting motion feature information and hitting motion energy feature information are extracted through the same convolutional layer in the ball speed prediction model.

[0095] In some embodiments of the present application, by using a combination of convolutional layers, pooling layers, feature fusion modules, and fully connected layers in the ball speed prediction model, statistical features in the data can be automatically extracted, so as to extract the hitting motion feature information and hitting motion energy feature information corresponding to the hitting motion data, and the hitting sound feature information corresponding to the hitting sound data. Exemplarily, as Figure 4 shown, taking the input of ACC signals and GYRO signals with a length of 6×50 as an example, high-level features in the ACC signals and GYRO signals, such as hitting motion feature information and hitting motion energy feature information, are extracted through the convolutional layer in the ball speed prediction model. The convolutional kernel size of this convolutional layer is 1×3, the first channel of the convolutional kernel is set to 1 to keep the ACC signals and GYRO signals from interfering with each other, the stride is 1×1, and zero padding is used to keep the signal size unchanged. Then, the pooling layer in the ball speed prediction model is used to obtain multi-scale information of the ACC signals and GYRO signals. The size of this pooling layer is 1×2, and the first channel is also set to 1 to prevent signal interference. Next, the feature fusion module fuses all the multi-scale information to obtain fused feature information. Finally, the fully connected layer is used to map the extracted fused feature information to the classification prediction result, so as to output the hitting ball speed when the moving object performs the hitting motion.

[0096] It should be noted that during the training phase of the ball speed prediction model, a large amount of collected data containing different hitting scenarios and player levels can be used. Optimization algorithms such as Stochastic Gradient Descent (SGD) and its variants like Adagrad, Adadelta, Adam, etc., and the cross-entropy function can be used as the loss function to continuously adjust the parameters of the ball speed prediction model to minimize the error between the predicted ball speed and the actual ball speed obtained by the radar speedometer, thereby training a high-precision ball speed prediction model.

[0097] Among them, the formula 3 corresponding to the above cross-entropy function is as follows:

[0098]

[0099] In a possible example, a ball speed data set containing 120 athletes, 9 hitting actions, and 25,196 hitting data can be established. Based on the ball speed measurement data set, a convolutional neural network is used for training. The input is a combined data size of 9×50, including the original inertial measurement unit (IMU) data size of 6×50, the swing mode classification label data size of 1×50, the combined energy data size of the swing amplitude of 1×50, and the Mel spectrogram of 1×50. The relevant neural network is fine-tuned, and the overall structure remains unchanged. It should be noted that the above IMU includes an ACC sensor and a GYRO sensor.

[0100] As Figure 5 shown, Figure 5 shows the comparison between the ball speed measurement results obtained by the ball speed measurement method provided in this application and the actual ball speed under different swing modes. Through Figure 5 it can be clearly seen that the ball speed measurement results obtained by the ball speed measurement method provided in this application are more consistent with the actual ball speed, improving the user experience. It should be noted that Figure 5 the 4 figures in

[0101] In some embodiments of the present application, the above hitting motion data may include the motion data collected by a six-axis sensor (i.e., an ACC sensor and a GYRI sensor) at each of N consecutive sampling times. The motion data collected at each sampling time includes the sampling data of each axis of the six-axis sensor. Further optionally, based on this, the "extracting the hitting action energy feature information corresponding to the effective hitting motion data" in step 202d above can be implemented through the following step 202e:

[0102] Step 202e: The electronic device accumulates the squared values of the sampling data of the i-th axis of the six-axis sensor corresponding to each of the N sampling moments to obtain the energy eigenvalue corresponding to the i-th axis, and accumulates the energy eigenvalues corresponding to each axis of the six-axis sensor to obtain the hitting motion energy characteristic information, where i ∈ {1, 2, 3, 4, 5, 6}.

[0103] In an embodiment of the present application, the electronic device can extract the hitting motion energy characteristic information through the following formula 2.

[0104]

[0105] Among them, x(n) i is used to represent the sampling data of the i-th axis of the six-axis sensor at the n-th sampling moment among the N sampling moments.

[0106] In this way, by quantifying the hitting motion energy characteristic information and predicting the ball speed based on the quantization value, a more accurate ball speed measurement result can be obtained.

[0107] In some embodiments of the present application, the hitting ball speed during the hitting motion of the above-mentioned moving object includes the hitting ball speed of each hitting motion of the moving object. Based on this, after the above step 203, the ball speed measurement method provided by the present application further includes the following steps 204 and 205, or steps 204 and 206, or steps 204, 205 and 206:[[]]

[0108] Step 204: The electronic device obtains hitting motion information based on the above-mentioned hitting motion characteristic information, hitting motion energy characteristic information, and sound characteristic information, and the hitting motion information is used to indicate the hitting motion corresponding to each hitting motion of the moving object.

[0109] Step 205: The electronic device statistically analyzes the hitting ball speed and hitting motion information of each hitting motion of the moving object to generate an analysis report.

[0110] Step 206: The electronic device displays a data statistics interface, and the data statistics interface includes information related to the hitting motion of the moving object.

[0111] In some embodiments of the present application, the above-mentioned information related to the hitting motion includes at least one of the following: information related to the hitting ball speed, the action identifier of the hitting motion corresponding to each hitting motion of the moving object, and the number of swings when the moving object performs the hitting motion.

[0112] In some embodiments of the present application, the above-mentioned information related to the hitting ball speed includes at least one of the following: the average hitting ball speed of the moving object during the hitting motion, the highest hitting ball speed of the moving object during the hitting motion, the lowest hitting ball speed of the moving object during the hitting motion, and the ball speed distribution information of the moving object during the hitting motion.

[0113] In some embodiments of the present application, the above-mentioned ball speed distribution information of the moving object during the hitting motion can be displayed on the data statistics interface in the form of a histogram, a bar chart, a line chart, etc.

[0114] In some embodiments of the present application, the above-mentioned action identifier of the hitting action includes at least one of the following: the action name of the hitting action, and the action picture of the hitting action.

[0115] In some embodiments of the present application, after calculating the ball speed, the electronic device can display the racket swing result of the moving object on the screen in real time.

[0116] For example, different ball speed intervals are displayed in different colors to facilitate the user to quickly identify.

[0117] For example, at the end of the game or training, the electronic device automatically counts all the racket swing speed parameters, such as the average hitting ball speed, the highest hitting ball speed, the lowest hitting ball speed, and the ball speed interval distribution, and displays them on the screen in the form of a bar chart. For example, the horizontal axis of the bar chart can be the ball speed interval, and the vertical axis can be the number of racket swings. The user can understand their own hitting characteristics through this bar chart.

[0118] Furthermore, the user can click on each data bar in the bar chart to trigger the electronic device to display the hitting action corresponding to the ball speed interval of the data bar.

[0119] In this way, the electronic device generates an analysis report by counting the hitting ball speed and hitting action information of each hitting motion of the moving object, provides training suggestions for the user, and helps to improve the training.

[0120] Hereinafter, taking the electronic device as a smart watch as an example, an exemplary description will be given of the ball speed measurement method provided by the embodiments of the present application.

[0121] As Figure 6 shown, the ball speed measurement method provided by the present application may include the following steps 401 to 406:

[0122] Step 401: The electronic device takes the ACC signal collected by the ACC sensor and the GYRO signal collected by the GYRO sensor as hitting motion data, and inputs the audio signal collected by the MIC sensor into the audio processing module.

[0123] Exemplarily, the above ACC signal contains ACC data, the above GYRO signal contains GYRO data, and the above audio signal contains hitting sound data.

[0124] Exemplarily, the above hitting motion data is six-axis IMU data.

[0125] Exemplarily, before performing a hitting motion, the moving object needs to correctly wear a smart watch containing an Inertial Measurement Unit (IMU) and provide hitting behavior information, such as the way of holding the racket, etc., to ensure effective data collection and accurate data analysis.

[0126] Step 402: The electronic device performs peak seeking on the hitting motion data based on the combined acceleration threshold to obtain effective hitting motion data.

[0127] Exemplarily, the above effective hitting motion data is six-axis IMU data during the occurrence of the hitting action.

[0128] Exemplarily, after the ACC sensor and GYRO sensor of the electronic device are turned on and before entering the ball speed measurement process, the primary task is to accurately identify the user's racket swing behavior. This process is implemented based on ACC.

[0129] Exemplarily, the racket swing behavior needs to satisfy: the signal threshold needs to be greater than a certain set value. The three-axis data of ACC can effectively represent the user's action intensity. The electronic device can calculate the combined acceleration of the ACC data at each moment, and then combine the calculated combined acceleration of the ACC data at each moment with a preset threshold to determine the acquisition moment of the peak data in the above six-axis IMU signal. For other descriptions of the acquisition moment of the peak data, refer to the relevant descriptions in the above steps 202c - 202d. To avoid repetition, it will not be elaborated here.

[0130] Exemplarily, the electronic device can use the acquisition moment of the peak data in the six-axis IMU signal as the center point, and use the six-axis IMU signals corresponding to the acquisition moment of the peak data and one or more consecutive acquisition moments before and after the acquisition moment of the peak data as effective hitting motion data.

[0131] It can be understood that when a hitting action that meets the combined acceleration threshold is detected, the electronic device will quickly extract the six-axis IMU signals of N sampling points near the acquisition moment of the peak data. These signals are crucial for the subsequent deep learning model analysis and will be input into the ball speed prediction model for feature convolution and result output.

[0132] Exemplarily, the ball speed prediction model uses a deep learning framework to process the hitting action feature information for subsequent auxiliary ball speed recognition. The deep learning training of the ball speed prediction model can be divided into the following steps: construction, annotation, and cleaning of the dataset, model selection, training, and fine-tuning, model deployment and inference. The specific steps of the above deep learning training are as follows:

[0133] 1) Construction, annotation, and cleaning of the dataset:

[0134] Build a private tennis hitting dataset that comprehensively covers multiple hitting types and ball speed ranges. This dataset contains 9 types of hitting data from 120 people. When obtaining the hitting action feature information, considering that most hitting movements are carried out on outdoor courts, based on cost and practicality considerations, the high-precision radar speed measurement principle is selected to obtain the true value of the hitting action feature information. The radar speedometer emits two linear signals with a time interval of T c Since there is a phase difference between the phases corresponding to the two received peaks, when the sphere moves at a speed of v, the phase difference of the received wave can be seen in Formula 5:

[0135]

[0136] Based on this, the maximum unambiguous speed of the sphere can be deduced as:

[0137]

[0138] Select manual ball speed annotation. There are 25,196 hitting data in total, and data cleaning is performed on the simulation model. Based on the condition that the absolute value difference between the model prediction result and the annotation result is greater than the threshold, further review and screening are carried out. The screening conditions are as follows:

[0139]

[0140] The above dataset contains the hitting action feature information of the hitting, which is used for the training of the classification model.

[0141] 2) Model selection, training, and fine-tuning:

[0142] Due to the limitation of the size of the electronic device, both the storage problem of multi-parameter variables caused by larger models and the limitation of the hardware inference ability lead to limitations in the selection of deep learning models. In this application, a 2D convolutional network is used for hitting action feature information recognition. The specific network model structure is as Figure 4 shown. For other descriptions of Figure 4 , please refer to the relevant descriptions in Step 303 - Step 302 and Step 303a, Step 303b above. To avoid repetition, it will not be elaborated here.

[0143] 3) Model deployment and inference:

[0144] Export the model using TensorFlow, then convert the model operators into binary code using Tflite, and finally deploy it to the hardware side using C language.

[0145] Step 403: Perform Mel spectrogram transformation on the audio signal collected by the MIC sensor to obtain the hitting sound feature information.

[0146] In some embodiments of the present application, after the MIC sensor of the electronic device collects an audio signal, the audio signal is input into the audio processing module in the electronic device. The audio processing module performs Mel spectrogram transformation on the audio signal and outputs a Mel spectrogram. The Mel spectrogram is input into the ball speed prediction model of the electronic device. Thus, the ball speed prediction model obtains the hitting sound feature information based on the Mel spectrogram.

[0147] Exemplarily, the above-mentioned hitting sound feature information is used to characterize the sound feature of the hitting sound when the moving object performs a hitting motion.

[0148] Exemplarily, the above-mentioned hitting sound feature information includes at least one of the following: pitch, timbre, decibel, and sound duration.

[0149] Step 404a: The ball speed prediction model performs hitting energy spectrum transformation on the effective hitting motion data to obtain the hitting action energy feature information.

[0150] Exemplarily, the above-mentioned hitting action energy feature information can be used to characterize the combined energy of the hitting action when the moving object performs a hitting motion. For other descriptions of the hitting action energy feature information, see the relevant descriptions in the above step 202c. To avoid repetition, it will not be elaborated here.

[0151] Step 404b: The ball speed prediction model extracts the hitting action feature information from the effective hitting motion data.

[0152] Exemplarily, the above-mentioned hitting action feature information is used to characterize the hitting action feature when the moving object performs a hitting motion. For other descriptions of the hitting action energy feature information, see the relevant descriptions in the above step 202. To avoid repetition, it will not be elaborated here.

[0153] Step 405: The ball speed prediction model fuses the hitting action feature information, the hitting action energy feature information, the hitting sound feature information, and the hitting motion data to obtain the ball speed measurement result.

[0154] Step 406: Output a motion analysis report according to the fused feature information.

[0155] Exemplarily, the above-mentioned output motion analysis report includes at least one of the following: display the maximum ball speed, average ball speed, and minimum ball speed; display the ball speed distribution range; display the number of swings and swing actions.

[0156] Exemplarily, the above output analysis report includes: statistically analyzing the hitting ball speed and hitting action information of each hitting motion of the moving object to generate an analysis report; displaying a data statistics interface, which includes information related to the hitting action of the moving object during the hitting motion.

[0157] In this way, by collecting the hitting motion data and hitting sound data of the moving object during the hitting motion; obtaining the hitting action feature information and hitting action energy feature information corresponding to the hitting motion data, and the hitting sound feature information corresponding to the hitting sound data; based on the above-mentioned hitting action feature information, hitting action energy feature information, and sound feature information, determining the hitting ball speed of the moving object during the hitting motion. Through this solution, since the hitting sound data of the moving object during the hitting motion can represent whether the hitting hits the sweet spot of the racket, therefore, when predicting the hitting ball speed of the moving object during the hitting motion in this application, by combining and analyzing the hitting action feature information and hitting action energy feature information corresponding to the hitting motion data of the moving object during the hitting motion with the sound feature information corresponding to the hitting sound data of the moving object during the hitting motion, the analyzed ball speed is more accurate, and the stability of the ball speed measurement quality is improved.

[0158] It should be noted that the above-mentioned various method embodiments, or various possible implementation manners in each method embodiment, can be executed independently, or any two or more of them can be combined with each other. Specifically, it can be determined according to actual usage requirements, and the embodiments of this application do not limit this.

[0159] For the ball speed measurement method provided by the embodiments of this application, the execution subject can be a ball speed measurement device. In the embodiments of this application, taking the ball speed measurement device executing the ball speed measurement method as an example, the ball speed measurement device provided by the embodiments of this application is described.

[0160] Attached Figure 7 shows a possible structural schematic diagram of the ball speed measurement device involved in the embodiments of this application. As Figure 7 shown, the ball speed measurement device 700 may include: a collection module 701, a processing module 702.

[0161] Among them, the collection module 701 is used to collect the hitting motion data and hitting sound data of the moving object during the hitting motion.

[0162] The processing module 702 is used to obtain the hitting action feature information and hitting action energy feature information corresponding to the hitting motion data collected by the collection module, and the hitting sound feature information corresponding to the hitting sound data;

[0163] The processing module 702 is further configured to determine the ball hitting speed of the moving object during the ball hitting motion based on the ball hitting motion feature information, the ball hitting motion energy feature information, and the sound feature information.

[0164] In some embodiments of the present application, the above-mentioned ball hitting motion information is the feature information of the racket swinging motion feature of the moving object, and the racket swinging motion feature is used to characterize the racket swinging manner of the moving object during the ball hitting motion. The racket swinging manner includes at least one of the following: forehand flat hit, forehand topspin, backhand flat hit, backhand topspin, forehand slice, and backhand slice.

[0165] In some embodiments of the present application, the above-mentioned ball hitting motion energy feature information is used to characterize the racket swinging amplitude of the moving object during the ball hitting motion.

[0166] In some embodiments of the present application, the processing module 702 is specifically configured to: filter the ball hitting sound data to obtain effective ball hitting sound data; convert the effective ball hitting sound data into a Mel spectrogram; extract the ball hitting motion feature information and the ball hitting motion energy feature information corresponding to the effective ball hitting motion data in the ball hitting motion data, and extract the ball hitting sound feature information corresponding to the ball hitting sound data from the Mel spectrogram.

[0167] In some embodiments of the present application, the above-mentioned effective ball hitting motion data is determined based on the acquisition time of the peak data in the ball hitting motion data.

[0168] In some embodiments of the present application, the processing module 702 is specifically configured to: input the effective ball hitting motion data and the Mel spectrogram into a ball speed prediction model. The ball speed prediction model includes a convolutional layer, a feature fusion module, and a fully connected layer; extract the ball hitting motion feature information and the ball hitting motion energy feature information corresponding to the effective ball hitting motion data through the convolutional layer, and extract the ball hitting sound feature information corresponding to the ball hitting sound data from the Mel spectrogram; fuse the effective ball hitting motion data, the ball hitting motion feature information, the ball hitting motion energy feature information, and the ball hitting sound feature information through the feature fusion module to obtain fused feature information; calculate and output the ball hitting speed of the moving object during the ball hitting motion based on the fused feature information through the fully connected layer.

[0169] In some embodiments of the present application, the above-mentioned effective hitting motion data includes the motion data collected by the six-axis sensor at each of the N consecutive sampling moments; the motion data collected at each sampling moment includes the sampling data of each axis of the six-axis sensor. Based on this, the above-mentioned processing module 702 is specifically configured to accumulate the squared values of the sampling data of the i-th axis of the six-axis sensor corresponding to each sampling moment among the N sampling moments through a ball speed prediction model to obtain the energy eigenvalue corresponding to the i-th axis, and accumulate the energy eigenvalues corresponding to each axis of the six-axis sensor to obtain the hitting action energy characteristic information, where i ∈ {1, 2, 3, 4, 5, 6}.

[0170] In some embodiments of the present application, the processing module 702 is specifically configured to: based on the hitting behavior information and hitting motion data of the motion object, obtain the hitting action characteristic information corresponding to the hitting motion data, where the hitting behavior information is used to characterize the racket-holding type of the motion object, and the racket-holding type includes: left-handed racket holding, or right-handed racket holding.

[0171] In the embodiments of the present application, the hitting motion data and hitting sound data of the motion object during the hitting motion are collected; the hitting action characteristic information and hitting action energy characteristic information corresponding to the hitting motion data, and the hitting sound characteristic information corresponding to the hitting sound data are obtained; based on the above-mentioned hitting action characteristic information, hitting action energy characteristic information, and sound characteristic information, the hitting ball speed of the motion object during the hitting motion is determined. Through this solution, since the hitting sound data during the hitting motion of the motion object can characterize whether the hitting hits the sweet spot of the racket, when predicting the hitting ball speed of the motion object during the hitting motion in the present application, by combining and analyzing the hitting action characteristic information and hitting action energy characteristic information corresponding to the hitting motion data of the motion object during the hitting motion with the sound characteristic information corresponding to the hitting sound data of the motion object during the hitting motion, the analyzed ball speed is more accurate, and the stability of the ball speed measurement quality is improved.

[0172] The beneficial effects of various implementation manners in this embodiment can be specifically referred to the beneficial effects of the corresponding implementation manners in the above method embodiments. To avoid repetition, they will not be elaborated here.

[0173] The ball speed measurement device in the embodiments of the present application may be an electronic device or a component in an electronic device, such as an integrated circuit or a chip. The electronic device may be a terminal or other devices other than terminals. Exemplarily, the electronic device may be a mobile phone, a tablet computer, a laptop computer, a handheld computer, an in-vehicle electronic device, a Mobile Internet Device (MID), an augmented reality (AR) / virtual reality (VR) device, a robot, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc., or may also be a server, a Network Attached Storage (NAS), a personal computer (PC), a television (TV), a teller machine, or a self-service machine, etc. The embodiments of the present application do not make specific limitations.

[0174] The ball speed measurement device in the embodiments of the present application may be a device with an operating system. The operating system may be an Android operating system, an iOS operating system, or other possible operating systems. The embodiments of the present application do not make specific limitations.

[0175] Optionally, as Figure 8 shown, the embodiments of the present application further provide an electronic device 900, including a processor 901 and a memory 902. A program or instruction that can run on the processor 901 is stored on the memory 902. When the program or instruction is executed by the processor 901, it implements each step of the above-mentioned ball speed measurement method embodiments and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0176] It should be noted that the electronic devices in the embodiments of the present application include the above-mentioned mobile electronic devices and non-mobile electronic devices.

[0177] Figure 9 It is a schematic hardware structure diagram of an electronic device for implementing the embodiments of the present application.

[0178] The electronic device 100 includes, but is not limited to: a radio frequency unit 101, a network module 102, an audio output unit 103, an input unit 104, a sensor 105, a display unit 106, a user input unit 107, an interface unit 108, a memory 109, and a processor 110 and other components.

[0179] Those skilled in the art can understand that the electronic device 100 may further include a power source (such as a battery) for supplying power to each component. The power source can be logically connected to the processor 110 through a power management system, so as to manage functions such as charging, discharging, and power consumption management through the power management system. Figure 9 The electronic device structure shown in Figure 9 does not limit the electronic device. The electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0180] Among them, the sensor 105 is used to collect the hitting motion data and hitting sound data when the moving object performs a hitting motion.

[0181] The processor 110 is used to obtain the hitting action feature information and hitting action energy feature information corresponding to the hitting motion data collected by the acquisition module, and the hitting sound feature information corresponding to the hitting sound data.

[0182] The processor 110 is further used to determine the hitting ball speed when the moving object performs a hitting motion based on the hitting action feature information, hitting action energy feature information, and sound feature information.

[0183] In some embodiments of the present application, the above-mentioned hitting action information is the feature information of the swing action feature of the moving object. The swing action feature is used to characterize the swing method when the moving object performs a hitting motion. The swing method includes at least one of the following: forehand flat hit, forehand topspin, backhand flat hit, backhand topspin, forehand slice, backhand slice.

[0184] In some embodiments of the present application, the above-mentioned hitting action energy feature information is used to characterize the swing amplitude when the moving object performs a hitting motion.

[0185] In some embodiments of the present application, the processor 110 is specifically used to: filter the hitting sound data to obtain effective hitting sound data; convert the effective hitting sound data into a Mel spectrogram; extract the hitting action feature information and hitting action energy feature information corresponding to the effective hitting motion data in the hitting motion data, and extract the hitting sound feature information corresponding to the hitting sound data from the Mel spectrogram.

[0186] In some embodiments of the present application, the above-mentioned effective hitting motion data is determined based on the acquisition time of the peak data in the hitting motion data.

[0187] In some embodiments of the present application, the processor 110 is specifically configured to: input the effective hitting motion data and the Mel spectrogram into a ball speed prediction model. The ball speed prediction model includes a convolutional layer, a feature fusion module, and a fully connected layer; extract the hitting motion feature information and the hitting motion energy feature information corresponding to the effective hitting motion data through the convolutional layer, and extract the hitting sound feature information corresponding to the hitting sound data from the Mel spectrogram; fuse the effective hitting motion data, the hitting motion feature information, the hitting motion energy feature information, and the hitting sound feature information through the feature fusion module to obtain fused feature information; calculate and output the hitting ball speed when the moving object performs a hitting motion based on the fused feature information through the fully connected layer.

[0188] In some embodiments of the present application, the above-mentioned effective hitting motion data includes the motion data collected by the six-axis sensor at each of N consecutive sampling moments; the motion data collected at each sampling moment includes the sampling data of each axis of the six-axis sensor. Based on this, the above-mentioned processor 110 is specifically configured to accumulate the squared values of the sampling data of the i-th axis of the six-axis sensor corresponding to each of the N sampling moments through the ball speed prediction model to obtain the energy feature value corresponding to the i-th axis, and accumulate the energy feature values corresponding to each axis of the six-axis sensor to obtain the hitting motion energy feature information, where i ∈ {1, 2, 3, 4, 5, 6}.

[0189] In some embodiments of the present application, the processor 110 is specifically configured to: obtain the hitting motion feature information corresponding to the hitting motion data based on the hitting behavior information of the moving object and the hitting motion data. The hitting behavior information is used to characterize the racket-holding type of the moving object, and the racket-holding type includes: left-handed racket holding, or right-handed racket holding.

[0190] In some embodiments of the present application, the hitting motion data includes the motion data collected by the six-axis sensor at each of N consecutive sampling moments. The motion data collected at each sampling moment includes the sampling data of each axis of the six-axis sensor. The processor 110 is specifically configured to:

[0191] Accumulate the squared values of the sampling data of the i-th axis of the six-axis sensor corresponding to each of the N sampling moments to obtain the energy feature value corresponding to the i-th axis, and accumulate the energy feature values corresponding to each axis of the six-axis sensor to obtain the hitting motion energy feature information, where i ∈ {1,..., 6}.

[0192] In some embodiments of the present application, the processor 110 is specifically configured to: filter the hitting sound data to obtain effective hitting sound data; convert the effective hitting sound data into a Mel spectrogram; extract the hitting sound feature information corresponding to the hitting sound data from the Mel spectrogram.

[0193] In some embodiments of the present application, the processor 110 is specifically configured to: based on the hitting behavior information of the moving object and the hitting motion data, obtain the hitting motion feature information corresponding to the hitting motion data, where the hitting behavior information is used to characterize the racket-holding type of the moving object, and the racket-holding types include: left-handed racket holding, or right-handed racket holding.

[0194] In the embodiments of the present application, the hitting motion data and the hitting sound data of the moving object during the hitting motion are collected; the hitting motion feature information and the hitting motion energy feature information corresponding to the hitting motion data, and the hitting sound feature information corresponding to the hitting sound data are obtained; based on the above-mentioned hitting motion feature information, hitting motion energy feature information, and sound feature information, the hitting ball speed of the moving object during the hitting motion is determined. Through this solution, since the hitting sound data during the hitting motion of the moving object can characterize whether the hitting hits the sweet spot of the racket, therefore, when predicting the hitting ball speed of the moving object during the hitting motion in the present application, by combining and analyzing the hitting motion feature information and the hitting motion energy feature information corresponding to the hitting motion data of the moving object with the sound feature information corresponding to the hitting sound data of the moving object during the hitting motion, the analyzed ball speed is more accurate, and the stability of the ball speed measurement quality is improved.

[0195] The beneficial effects of various implementation manners in this embodiment can be specifically referred to the beneficial effects of the corresponding implementation manners in the above method embodiments. To avoid repetition, they will not be elaborated here.

[0196] It should be understood that in the embodiments of the present application, the input unit 104 may include a Graphics Processing Unit (GPU) 1041 and a microphone 1042. The graphics processor 1041 processes the image data of static pictures or videos obtained by an image capture device (such as a camera) in the video capture mode or the image capture mode. The display unit 106 may include a display panel 1061, and the display panel 1061 may be configured in the form of a liquid crystal display, an organic light-emitting diode, etc. The user input unit 107 includes at least one of a touch panel 1071 and other input devices 1072. The touch panel 1071 is also called a touch screen. The touch panel 1071 may include two parts: a touch detection device and a touch controller. The other input devices 1072 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, and a joystick, which will not be elaborated here.

[0197] The memory 109 can be used to store software programs and various data. The memory 109 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data. Among them, the first storage area can store an operating system, application programs or instructions required for at least one function (such as a sound playback function, an image playback function, etc.). In addition, the memory 109 can include a volatile memory or a non-volatile memory, or the memory 109 can include both a volatile and a non-volatile memory. Among them, the non-volatile memory can be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable PROM (EEPROM), or a flash memory. The volatile memory can be a Random Access Memory (RAM), a Static RAM (SRAM), a Dynamic RAM (DRAM), a Synchronous DRAM (SDRAM), a Double Data Rate SDRAM (DDR SDRAM), an Enhanced SDRAM (ESDRAM), a Synch link DRAM (SLDRAM), and a Direct Rambus RAM (DRRAM). The memory 109 in the embodiments of the present application includes but is not limited to these and any other suitable types of memories.

[0198] The processor 110 may include one or more processing units; optionally, the processor 110 integrates an application processor and a modem processor. Among them, the application processor mainly processes operations related to the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communication signals, such as a baseband processor. It can be understood that the above-mentioned modem processor may not be integrated into the processor 110 either.

[0199] The embodiments of the present application also provide a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, it implements each process of the above-mentioned ball speed measurement method embodiment and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0200] Among them, the processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory ROM, random access memory RAM, magnetic disks, or optical discs, etc.

[0201] Another embodiment of the present application provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement each process of the above embodiment of the ball speed measurement method, and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0202] It should be understood that the chip mentioned in the embodiments of the present application may also be referred to as a system-on-chip, system chip, chip system, or system-on-chip, etc.

[0203] The embodiments of the present application provide a computer program product. The program product is stored in a storage medium and is executed by at least one processor to implement each process of the above embodiment of the ball speed measurement method, and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0204] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article, or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article, or device including that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, the features described with reference to certain examples may be combined in other examples.

[0205] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, in essence or the part that contributes to the prior art, can be embodied in the form of a computer software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present application.

[0206] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms without departing from the purpose of the present application and the scope protected by the claims, and all of them belong to the protection scope of the present application.

Claims

1. A ball speed measurement method, characterized in that: The method comprises: Collecting ball-hitting motion data and ball-hitting sound data when the sports object performs a ball-hitting motion; Acquire the batting action characteristic information and the batting action energy characteristic information corresponding to the batting motion data, and the batting sound characteristic information corresponding to the batting sound data; The hitting ball speed of the moving object when performing the hitting motion is determined based on the hitting motion characteristic information, the hitting motion energy characteristic information and the sound characteristic information.

2. The method according to claim 1, characterized in that: The batting action characteristic information at least includes characteristic information of the swing action characteristics of the sports object; The swing action feature is used to characterize the swinging manner of the sports object when performing the hitting motion; The swinging method includes at least one of the following: forehand flat shot, forehand topspin, backhand flat shot, backhand topspin, forehand slice, backhand slice; The batting motion energy characteristic information is used to characterize the swing amplitude of the sports object when performing the batting motion.

3. The method according to claim 1 or 2, characterized in that: The step of obtaining the batting action characteristic information and the batting action energy characteristic information corresponding to the batting motion data, and the batting sound characteristic information corresponding to the batting sound data, comprises: filtering the ball hitting sound data to obtain effective ball hitting sound data; Converting the effective hitting sound data into a mel spectrogram; The hitting motion feature information and the hitting motion energy feature information corresponding to the effective hitting motion data in the hitting motion data are extracted, and the hitting sound feature information corresponding to the hitting sound data is extracted from the mel spectrogram, wherein the effective hitting motion data is determined based on the collection time of the peak data in the hitting motion data.

4. The method according to claim 3, characterized in that: The step of extracting the hitting action characteristic information and the hitting action energy characteristic information corresponding to the effective hitting action data in the hitting action motion data, and extracting the hitting sound characteristic information corresponding to the hitting sound data from the mel-sound spectrogram includes: Inputting the effective hitting motion data and the Mel-spectrogram into a ball speed prediction model, wherein the ball speed prediction model includes a convolutional layer, a feature fusion module, and a fully connected layer; Extracting the hitting action feature information and the hitting action energy feature information corresponding to the effective hitting motion data through the convolution layer, and extracting the hitting sound feature information corresponding to the hitting sound data from the mel spectrogram; The determining of the hitting ball speed of the moving object when performing the hitting motion based on the hitting motion characteristic information, the hitting motion energy characteristic information and the sound characteristic information comprises: fusing the effective hitting motion data, the hitting action feature information, the hitting action energy feature information and the hitting sound feature information through the feature fusion module to obtain fused feature information; The fully connected layer calculates and outputs the ball speed of the moving object when performing the hitting motion based on the fused feature information.

5. The method according to claim 4, characterized in that The effective ball-hitting motion data includes motion data collected by the six-axis sensor at each sampling moment in N consecutive sampling moments; the motion data collected at each sampling moment includes sampling data of each axial direction of the six-axis sensor; The ball speed prediction model extracts the energy characteristic information of the hitting action corresponding to the effective hitting motion data, including: The ball speed prediction model accumulates the square values ​​of the sampling data of the i-th axis of the six-axis sensor corresponding to each sampling moment in the N sampling moments to obtain the energy characteristic value corresponding to the i-th axis, and accumulates the energy characteristic values ​​corresponding to each axis of the six-axis sensor to obtain the energy characteristic information of the hitting action, i∈{1, 2, 3, 4, 5, 6}.

6. The method according to any one of claims 1 to 5, characterized in that: The step of obtaining the batting action characteristic information corresponding to the batting motion data includes: Based on the batting behavior information of the sports object and the batting motion data, batting action feature information corresponding to the batting motion data is obtained, wherein the batting behavior information is used to characterize the racket holding type of the sports object, and the racket holding type includes: left-handed racket holding or right-handed racket holding.

7. A ball speed measuring device, characterized in that: The device comprises a collection module and a processing module; The acquisition module is used to collect the hitting motion data and hitting sound data of the moving object when the moving object performs the hitting motion; The processing module is used to obtain the hitting action characteristic information and the hitting action energy characteristic information corresponding to the hitting motion data collected by the collection module, and the hitting sound characteristic information corresponding to the hitting sound data; The processing module is further used to determine the hitting ball speed of the moving object when performing the hitting motion based on the hitting motion characteristic information, the hitting motion energy characteristic information and the sound characteristic information.

8. The device according to claim 7, characterized in that The batting action information is feature information of the swing action feature of the sports object; The swing action feature is used to characterize the swinging manner of the sports object when performing the hitting motion; The swinging method includes at least one of the following: forehand flat shot, forehand topspin, backhand flat shot, backhand topspin, forehand slice, backhand slice; The batting motion energy characteristic information is used to characterize the swing amplitude of the sports object when performing the batting motion.

9. An electronic device, characterized in that: It comprises a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the ball speed measurement method as described in any one of claims 1-6 are implemented.

10. A storage medium for reading, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps of the ball speed measurement method according to any one of claims 1 to 6 are implemented.

Images