CN116662892A - Method and device for detecting ball hitting point position of ball hitting sports equipment - Google Patents
Method and device for detecting ball hitting point position of ball hitting sports equipment Download PDFInfo
- Publication number
- CN116662892A CN116662892A CN202210141970.8A CN202210141970A CN116662892A CN 116662892 A CN116662892 A CN 116662892A CN 202210141970 A CN202210141970 A CN 202210141970A CN 116662892 A CN116662892 A CN 116662892A
- Authority
- CN
- China
- Prior art keywords
- ball striking
- characteristic information
- vibration
- vibration sensor
- eigenfrequency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
- Length Measuring Devices Characterised By Use Of Acoustic Means (AREA)
Abstract
A method and a device for detecting the position of a hitting point of a hitting sports device. The method comprises the following steps: acquiring a vibration signal generated by the vibration sensor, wherein the vibration signal is generated by detecting the vibration generated by the collision of the batting sports equipment and the ball; performing spectrum analysis on the vibration signal to obtain a plurality of eigenfrequencies of the vibration signal in a frequency domain; and calculating at least one characteristic information by utilizing the amplitudes of the eigenfrequencies and inputting a prediction model which is established in advance by using machine learning so as to estimate the impact position of the ball on the batting sports equipment, wherein the prediction model is trained by using the characteristic information of a plurality of vibration signals and a plurality of corresponding impact positions.
Description
Technical Field
The present disclosure relates to a method and apparatus for detecting a position of a striking point of a striking device, and more particularly, to a method and apparatus for detecting a striking point of a striking device.
Background
Currently, the detection of the position of the contact point between the baseball and the bat (i.e. the batting point) when the baseball player batts is mainly performed by image processing. Shooting the batting image of the player by erecting photographic equipment around the player, analyzing the image by utilizing an intrinsic acquisition and comparison peer-to-peer image processing mode, and finally estimating the batting point position.
However, the above method requires a high-speed photographic apparatus with high cost and adjustment to a specific photographing angle to clearly photograph the striking point. The shooting angle of the shooting equipment cannot be adjusted in real time according to the action because the batting action is changed at any time, and the actual shooting often has shooting dead angles. In addition, the position detection by the image processing method requires a high amount of computation, and particularly when the number of image frames is large, the requirement for the computation capability of the processing device is also increased, and as a result, the cost is increased.
Disclosure of Invention
The present disclosure provides a method and apparatus for detecting a hitting point position of a hitting sports apparatus, which can accurately detect a hitting point position by detecting vibration of the hitting sports apparatus when hitting a ball and analyzing the frequency eigenvalue of the vibration.
The present disclosure provides a method for detecting a hitting point of a hitting sports apparatus, which is suitable for detecting a hitting position by an electronic device with a processor through a vibration sensor configured on the hitting sports apparatus. The method comprises the following steps: acquiring a vibration signal generated by the vibration sensor, wherein the vibration signal is generated by detecting the vibration generated by the collision of the batting sports equipment and the ball; performing spectrum analysis on the vibration signal to obtain a plurality of eigenfrequencies (eigenfrequency) of the vibration signal in a frequency domain; and calculating at least one characteristic information by utilizing the amplitudes of the eigenfrequencies and inputting a prediction model which is established in advance by using machine learning so as to estimate the impact position of the ball on the batting sports equipment, wherein the prediction model is trained by using the characteristic information of a plurality of vibration signals and a plurality of corresponding impact positions.
In an embodiment of the disclosure, the step of calculating the at least one characteristic information using the amplitudes of the eigenfrequencies includes selecting a primary eigenfrequency and at least one secondary eigenfrequency according to the magnitudes of the amplitudes of the eigenfrequencies, and calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency.
In an embodiment of the disclosure, the step of calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency includes calculating a ratio of the primary eigenfrequency to the amplitudes of the respective secondary eigenfrequencies as the characteristic information.
In an embodiment of the disclosure, the step of calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency further includes calculating a ratio of the amplitudes between the secondary eigenfrequencies as the characteristic information.
In an embodiment of the disclosure, the step of calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency includes calculating a first characteristic value using the amplitudes of the primary eigenfrequency, calculating at least one second characteristic value using the amplitudes of the secondary eigenfrequencies, and calculating a ratio of the first characteristic value to each second characteristic value as the characteristic information.
In an embodiment of the disclosure, the method further obtains a plurality of vibration signals generated by the vibration sensor detecting vibrations generated by the ball striking sports equipment at a plurality of preset striking positions, performs a frequency spectrum analysis on the vibration signals to obtain a plurality of eigenfrequencies of the vibration signals in a frequency domain, calculates the characteristic information by using the amplitudes of the eigenfrequencies, uses the characteristic information as an input of a prediction model, uses the corresponding striking positions as an output of the prediction model, trains the prediction model, and records a plurality of learning parameters of the trained prediction model.
The present disclosure provides a ball striking point position detection device for a ball striking sports apparatus, which comprises a data acquisition device, a storage device and a processor. The data acquisition device is configured to be coupled to a vibration sensor disposed on the ball striking piece of athletic equipment, the vibration sensor detecting vibrations of the ball striking piece of athletic equipment to generate a vibration signal. The storage device is used for storing a plurality of learning parameters of a prediction model which is built in advance by machine learning, wherein the prediction model is trained by using characteristic information of a plurality of vibration signals and a plurality of corresponding impact positions, and the impact positions are positions where the batting sports equipment impacts with the ball. The processor is coupled to the data acquisition device and the storage device, and is configured to acquire a vibration signal generated by the vibration sensor detecting vibration generated by the impact of the ball striking device with the ball, perform spectrum analysis on the vibration signal to obtain a plurality of eigenvrequency of the vibration signal in a frequency domain, calculate at least one piece of characteristic information by using the amplitude of each eigenvrequency, and input the at least one piece of characteristic information into the prediction model to estimate the impact position of the ball on the ball striking device.
In an embodiment of the disclosure, the processor includes selecting a primary eigenfrequency and at least one secondary eigenfrequency according to the amplitude of the eigenfrequency, and calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency.
In an embodiment of the present disclosure, the processor includes calculating a ratio of the amplitude of the primary eigenfrequency to the amplitude of each of the secondary eigenfrequencies as the characteristic information.
In an embodiment of the disclosure, the processor further calculates a ratio of amplitudes between the secondary eigenfrequencies as the characteristic information.
In an embodiment of the disclosure, the processor includes calculating a first eigenvalue using the amplitudes of the primary eigenfrequencies, calculating at least one second eigenvalue using the amplitudes of the respective secondary eigenfrequencies, and calculating a ratio of the first eigenvalue to the respective second eigenvalue as the eigenvalue.
In an embodiment of the disclosure, the processor further obtains a plurality of vibration signals generated by detecting vibrations generated by the ball striking sport equipment at a plurality of preset striking positions by using the data obtaining device, performs a frequency spectrum analysis on the vibration signals to obtain a plurality of eigenfrequencies of the vibration signals in a frequency domain, calculates the characteristic information by using the amplitudes of the eigenfrequencies, uses the characteristic information as an input of a prediction model, uses the corresponding striking positions as an output of the prediction model, trains the prediction model, and records learning parameters of the trained prediction model in the storage device.
In an embodiment of the present disclosure, the machine learning includes decision tree (SVM), convolutional neural network (Convolutional Neural Network, CNN), deep neural network (Deep Neural Networks, DNN), or support vector machine (Support Vector Machine, SVM).
In an embodiment of the disclosure, the vibration sensor includes one or a combination of a piezoelectric vibration sensor, an electric vibration sensor, an eddy current vibration sensor, an inductive vibration sensor, a capacitive vibration sensor, a resistive vibration sensor and a photoelectric vibration sensor.
In order to make the present disclosure more comprehensible, embodiments accompanied with figures are described in detail below.
Drawings
FIG. 1 is a block diagram of a ball striking point position detection apparatus for a ball striking sports device according to one embodiment of the present disclosure.
Fig. 2 is a schematic diagram illustrating a configuration of a vibration sensor according to an embodiment of the disclosure.
Fig. 3 is a flow chart of a method of detecting a hitting point position of a hitting sports device according to an embodiment of the present disclosure.
FIG. 4 is a flow chart illustrating a method of training a predictive model using machine learning in accordance with an embodiment of the disclosure.
Fig. 5 is a flowchart of a method of detecting a ball striking point location of a ball striking piece according to one embodiment of the present disclosure.
Fig. 6A-6G illustrate examples of a club striking point position detection method according to an embodiment of the present disclosure.
Symbol description
10: batting point position detecting device for batting sports equipment
12: data acquisition device
14: storage device
16: processor and method for controlling the same
20: piezoelectric vibration sensor
22: baseball bat
24: ball striking point position
S302 to S306, S402 to S406, S502 to S510: step (a)
Detailed Description
The embodiments of the present disclosure provide a method and apparatus for detecting a hitting point of a hitting sports apparatus, which detect vibration of the hitting sports apparatus when hitting a ball by installing a vibration sensor on the hitting sports apparatus, and perform a spectrum analysis on a vibration signal to obtain a plurality of eigenfrequencies in a frequency domain, and then input relative features of the eigenfrequencies into a prediction model previously established and trained using machine learning, thereby estimating the hitting point position. Thus, regardless of the magnitude of the striking force, the disclosed embodiments can accurately calculate the striking point position.
In detail, fig. 1 is a block diagram of a striking point position detection apparatus of a striking sports apparatus according to an embodiment of the present disclosure. Referring to fig. 1, a ball striking point position detection apparatus 10 of the present disclosure is a computer device such as a file server, a database server, an application server, a workstation or a personal computer, or a mobile device such as a mobile phone or a tablet computer, and includes components such as a data acquisition device 12, a storage device 14 and a processor 16, and functions of the components are as follows:
the data acquisition device 12 is, for example, any wired or wireless interface device connectable to a vibration sensor disposed on the ball striking piece of athletic equipment for acquiring a vibration signal generated by the vibration sensor detecting vibration of the ball striking piece of athletic equipment. For wired mode, the data acquisition device 12 may be, but not limited to, a universal serial bus (universal serial bus, USB), RS232, a universal asynchronous receiver transmitter (universal asynchronous receiver/transmitter, UART), an internal integration circuit (I2C), a serial peripheral interface (serial peripheral interface, SPI), a display port (display port), or a thunderbolt port (thunderbolt). For wireless mode, the data acquisition device 12 may be a device supporting wireless fidelity (wireless fidelity, wi-Fi), RFID, bluetooth, infrared, near-field communication (NFC) or device-to-device (D2D) communication protocols, but is not limited thereto. The vibration sensor is, for example, one or a combination of a piezoelectric vibration sensor, an electric vibration sensor, an eddy current vibration sensor, an inductance vibration sensor, a capacitance vibration sensor, a resistance vibration sensor and a photoelectric vibration sensor, which are disposed or attached to the tail end, the club head or any other position of the ball striking sports equipment, and the present embodiment is not limited in kind and disposition.
For example, fig. 2 is a schematic diagram illustrating a configuration of a vibration sensor according to an embodiment of the disclosure. Referring to fig. 2, for example, a piezoelectric vibration sensor 20 is disposed at the tail end of a baseball bat 22 to detect the vibration generated when the baseball bat 22 hits a ball and generate a vibration signal, and the vibration sensor 20 transmits the vibration signal to a bat hitting point position detecting device 10, so that the bat hitting point position detecting device 10 calculates a hitting point position 24. In the present embodiment, the piezoelectric vibration sensor 20 is made of a thinner piezoelectric material, so that the vibration of the bat can be detected without affecting the grip and the characteristics of the bat. It should be noted that, although the present embodiment is described taking a baseball bat as an example, the present embodiment is not limited thereto, and the present disclosure may be applied to various batting sports apparatuses such as a softball bat, a badminton racket, a tennis racket, a table racket, a hockey stick, a golf club, and the like, without limitation.
The storage device 14 is, for example, any type of fixed or removable random access Memory (Random Access Memory, RAM), read-Only Memory (ROM), flash Memory (Flash Memory), hard disk, or the like, or a combination thereof, for storing a computer program executable by the processor 16. In some embodiments, the storage device 14 also stores learning parameters of a predictive model that is pre-established and trained by the processor 16 using machine learning. In some embodiments, the storage device 14 may be used to temporarily store learning parameters downloaded by the processor 16 from a cloud server or remote device using the data acquisition apparatus 12 using machine learning pre-established and trained predictive models. The machine learning includes, but is not limited to, decision tree (decision tree), convolutional neural network (Convolutional Neural Network, CNN), deep neural network (Deep Neural Networks, DNN), or support vector machine (Support Vector Machine, SVM).
The processor 16 is, for example, a central processing unit (Central Processing Unit, CPU), or other programmable general purpose or special purpose Microprocessor (Microprocessor), microcontroller (Microcontroller), digital signal processor (Digital Signal Processor, DSP), programmable controller, application specific integrated circuit (Application Specific Integrated Circuits, ASIC), programmable logic device (Programmable Logic Device, PLD), or other similar device, or combination of devices, as the disclosure is not limited in this regard. In this embodiment, the processor 16 may load a computer program from the storage device 14 to perform the machine learning-based shot point position detection method of the embodiments of the present disclosure.
Fig. 3 is a flow chart of a method of detecting a hitting point position of a hitting sports device according to an embodiment of the present disclosure. Referring to fig. 1 and 3, the method of the present embodiment is applicable to the above-mentioned ball striking point position detecting device 10 of ball striking sports equipment, and the following describes the detailed steps of the ball striking point position detecting method of the present embodiment with respect to the components of the ball striking point position detecting device 10 of ball striking sports equipment.
In step S302, a vibration signal generated by the processor 16 of the ball striking point position detection apparatus 10 of the ball striking device detecting vibration of the ball striking device due to the impact of the ball striking device with the ball is acquired by the vibration sensor using the data acquisition device 12. Wherein the vibration signal is, for example, normalized energy in the time domain measured by the vibration sensor as the ball striking piece vibrates.
In step S304, the processor 16 performs a spectrum analysis on the vibration signal to obtain a plurality of eigenfrequencies (eigenfrequencies) of the vibration signal in the frequency domain. The processor 16 performs a frequency domain analysis method such as fourier transform (Fourier transform) or wavelet transform on the vibration signal, but is not limited thereto, and in other embodiments, the processor 16 may use any kind of frequency domain analysis method to transform the vibration signal into a frequency domain signal, and measure multiple peaks (peaks) in the frequency domain signal as eigenfrequencies.
In step S306, the processor 16 calculates at least one piece of characteristic information using the amplitudes of the respective eigenfrequencies and inputs a predictive model that is pre-trained using machine learning to estimate the impact location of the ball on the ball striking piece of athletic equipment. The processor 16 selects a primary eigenfrequency and at least one secondary eigenfrequency according to the amplitude of the eigenfrequency, for example, and calculates the characteristic information by using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency. The characteristic information is, for example, a relative characteristic of the eigenfrequency, such as the ratio of the amplitude of the primary eigenfrequency to the secondary eigenfrequency, thereby excluding the impact of the impact force.
For example, in some embodiments, the processor 16 may perform normalization (normalization) operations on the amplitudes of the plurality of eigenfrequencies (i.e., divide each amplitude by a maximum value therein), select the first 3 of the amplitudes from the eigenfrequencies (M in turn 1 、M 2 、M 3 ) And calculates the maximum amplitude M thereof 1 With other amplitude M 2 、M 3 Ratio M of (2) 1 /M 2 、M 1 /M 3 As characteristic information of the ball striking device. In some embodiments, processor 16 may use only maximum amplitude M 1 And the next largest amplitude M 2 Ratio M of (2) 1 /M 2 As characteristic information of the ball striking piece. In some embodiments, processor 16 may further calculate amplitude M 2 、M 3 Ratio M of (2) 2 /M 3 And to the aforementioned ratio M 1 /M 2 、M 1 /M 3 Together as characteristic information of the ball striking piece of athletic equipment. The above examples are only illustrative of possible embodiments of the invention and are not intended to limit the invention.
On the other hand, the above-described prediction model is, for example, a machine learning model trained in advance using the characteristic information of a plurality of vibration signals and a corresponding plurality of impact positions. In detail, the predictive model requires a decision criterion to determine which features and regression equations to use. Because the ratio of the amplitudes from different characteristic information has different distributions, the embodiments of the disclosure can implement a high-precision prediction model through advanced modes such as decision tree regressor (regressor) or deep machine learning.
In detail, fig. 4 is a flowchart illustrating a method for training a predictive model using machine learning according to an embodiment of the disclosure. Referring to fig. 1 and 4, the method of the present embodiment is applicable to the above-mentioned striking point position detection apparatus 10 of the striking sports apparatus.
In step S402, the processor 16 of the batting sports equipment batting point position detection apparatus 10 acquires a plurality of vibration signals generated by the vibration sensor detecting vibrations generated by the ball striking a plurality of impact positions preset on the batting sports equipment using the data acquisition device 12. In this embodiment, for example, 40 impact points (for example, one impact point is set every 1 cm) are evenly distributed in 40 cm from the club head of the ball striking sports apparatus (as shown in fig. 2), and the balls are sequentially used to impact the impact points and the vibration sensor is used to detect the vibration generated when the ball striking sports apparatus is impacted so as to generate a vibration signal. The processor 16 acquires these vibration signals using the data acquisition device 12.
In step S404, the processor 16 performs spectrum analysis on the vibration signals to obtain a plurality of eigenfrequencies of the vibration signals in the frequency domain, and calculates characteristic information using the amplitudes of the eigenfrequencies. Similar to the calculation of the characteristic information, the processor 16 selects a primary eigenfrequency and at least one secondary eigenfrequency according to the magnitude of the amplitudes of the eigenfrequencies, and calculates the characteristic information by using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency, for example, calculates the ratio of the amplitudes of the primary eigenfrequency and each secondary eigenfrequency as the characteristic information, and calculates the ratio of the amplitudes of the secondary eigenfrequencies as the characteristic information.
In step S406, the processor 16 takes the calculated feature information as an input of the prediction model, takes the corresponding impact position as an output of the prediction model, trains the prediction model, and records a plurality of learning parameters of the trained prediction model in the storage device 14.
After the characteristic distribution of the batting sports equipment is constructed by the above-mentioned prediction model, the batting sports equipment batting point position detecting device 10 according to the embodiments of the present disclosure can accurately locate the batting point by converting the detected vibration signal into the characteristic information and inputting the characteristic information each time the batting sports equipment is used for batting. In some embodiments, the striking point position detection apparatus 10 may, for example, individually build a prediction model for different types of striking sports apparatuses (e.g., wood sticks, aluminum sticks) and store the prediction model in the storage device 14, so that when an actual striking is performed, the corresponding prediction model may be accessed from the storage device 14 to detect the striking point position by identifying the serial number or type of the striking sports apparatus. Thereby, the accuracy of the detected position of the ball striking point can be increased.
Fig. 5 is a flowchart of a method of detecting a ball striking point location of a ball striking piece according to one embodiment of the present disclosure. Referring to fig. 1 and 5, the method of the present embodiment is applicable to the above-mentioned ball striking point position detecting device 10 of ball striking sports equipment, and the following describes the detailed steps of the ball striking point position detecting method of the present embodiment with respect to the components of the ball striking point position detecting device 10 of ball striking sports equipment.
In step S502, a vibration signal generated by the processor 16 of the ball striking point position detection apparatus 10 of the ball striking device detecting vibration of the ball striking device due to the impact of the ball striking device with the ball is acquired by the vibration sensor using the data acquisition device 12. In step S504, the processor 16 performs a spectrum analysis on the vibration signal to obtain a plurality of eigenfrequencies of the vibration signal in the frequency domain. The steps S502 to S504 are the same as or similar to the steps S302 to S304 in the foregoing embodiment, so the details thereof are not repeated here.
Unlike the foregoing embodiment, in the present embodiment, the processor 16 selects the primary eigenfrequency and at least one secondary eigenfrequency according to the magnitude of the amplitude of the eigenfrequency in step S506, calculates a first eigenvalue using the amplitude of the primary eigenfrequency, calculates at least one second eigenvalue using the amplitude of each secondary eigenfrequency in step S508, and calculates the ratio of the first eigenvalue to each second eigenvalue as the eigenvalue. For example, the processor 16 may calculate the square, square root or other power of the amplitude of the primary eigenfrequency as the first eigenvalue, and correspondingly calculate the square, square root or other power of the amplitude of the secondary eigenfrequency as the second eigenvalue, and then calculate the ratio of the first eigenvalue to each of the second eigenvalues as the eigenvalue.
In step S510, the processor 16 inputs the calculated characteristic information into a predictive model trained in advance using machine learning to estimate the impact location of the ball on the ball striking piece of athletic equipment. Wherein the processor 16 calculates, for example, in the same manner, characteristic information of a plurality of vibration signals and is used to train a predictive model to obtain accurate shot point positions.
With the above-described predictive model, the batting sports equipment batting point position detection device 10 of the embodiments of the present disclosure can also accurately locate the batting point by converting the detected vibration signal into characteristic information and inputting the predictive model each time the batting sports equipment is used for batting.
Fig. 6A to 6G illustrate examples of a method for detecting a ball striking point position of a ball bat according to an embodiment of the present invention. In this embodiment, for example, the ball is impacted at a plurality of impact positions d (where d is a position of every 1 cm within 40 cm from the club head) on the club with three different forces, and vibration of the club is detected by using the vibration sensor to obtain a vibration signal pd in the time domain as shown in fig. 6A, where each point in the vibration signal pd represents the impact strength obtained by normalizing the vibration signal pd to a value between 0.0 and 1.0, where the black point represents the impact strength obtained by impacting the club with a strong force, the dark gray point represents the impact strength obtained by impacting the club with a medium force, and the light gray point represents the impact strength obtained by impacting the club with a weak force.
Then, by the vibration signal Pd]After spectral analysis to obtain a plurality of eigenfrequencies in the frequency domain and normalization of the amplitudes of these eigenfrequencies, a distribution in which the amplitudes of the first 3 large eigenfrequencies are distributed with the impact position d is obtained, including FIG. 6BMaximum amplitude M shown 1 [d]The distribution of the 2 nd maximum amplitude M shown in FIG. 6C 2 [d]And the maximum 3 maximum amplitude M shown in FIG. 6D 3 [d]Is a distribution of (a).
Then by adjusting amplitude M 1 [d]、M 2 [d]、M 3 [d]After the ratio is calculated, the distribution of the characteristic information of the bat along with the impact position d can be obtained, wherein the characteristic information M shown in FIG. 6E 1 [d]/M 2 [d]Distribution of characteristic information M shown in FIG. 6F 1 [d]/M 3 [d]And the distribution of the characteristic information M shown in FIG. 6G 2 [d]/M 3 [d]Is a distribution of (a). As can be seen from fig. 6E to fig. 6G, the impact force on the characteristic information can be eliminated by calculating the ratio.
Finally, by combining these characteristic information M 1 [d]/M 2 [d]、M 1 [d]/M 3 [d]、M 2 [d]/M 3 [d]And the corresponding impact position d is respectively used as input and output for training a prediction model established by machine learning, so that the prediction model recorded with the club characteristics can be obtained and used as the basis for detecting the batting point position subsequently.
In summary, in the method and apparatus for detecting the position of a striking point of a striking device according to the embodiments of the present disclosure, a vibration sensor is mounted on the striking device to detect the vibration of the striking device when striking the striking device, and a predictive model that can respond to the characteristics of the striking device is pre-established by using the relative characteristics of the vibration signal in the frequency domain. Thereby, each time a ball is struck with the same or similar type of ball striking sports equipment, the position of the ball striking point can be accurately calculated by inputting the relative characteristics of the vibration signals into the corresponding predictive model.
While the present disclosure has been described with reference to the exemplary embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present disclosure as defined by the appended claims and their equivalents.
Claims (16)
1. A method of detecting a position of a ball striking point of a ball striking sports apparatus, adapted to detect an impact position by an electronic device having a processor using a vibration sensor disposed on the ball striking sports apparatus, the method comprising the steps of:
acquiring a vibration signal generated by the vibration sensor when the batting sports equipment and the ball are impacted;
performing spectrum analysis on the vibration signal to obtain a plurality of eigenfrequencies of the vibration signal in a frequency domain; and
at least one characteristic information is calculated using the amplitudes of each of the eigenfrequencies and input into a predictive model pre-established using machine learning to estimate the impact location of the ball on the ball striking piece of athletic equipment, wherein the predictive model is trained using the characteristic information of a plurality of vibration signals and a corresponding plurality of impact locations.
2. The method of claim 1, wherein the step of calculating at least one characteristic information using the amplitude of each of the eigenfrequencies comprises:
selecting a primary eigenfrequency and at least one secondary eigenfrequency according to the magnitude of the amplitude of the eigenfrequency, and calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency.
3. The method of claim 2, wherein calculating the characteristic information using the amplitudes of the primary eigenfrequencies and the secondary eigenfrequencies comprises:
calculating a ratio of the amplitude of the primary eigenfrequency to each of the secondary eigenfrequencies as the characteristic information.
4. The method of claim 3, wherein the step of calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency further comprises:
the ratio of the amplitudes between the secondary eigenfrequencies is calculated as the characteristic information.
5. The method of claim 2, wherein calculating the characteristic information using the amplitudes of the primary eigenfrequencies and the secondary eigenfrequencies comprises:
and calculating a first eigenvalue by using the amplitude of the main eigenvalue, calculating at least one second eigenvalue by using the amplitude of each secondary eigenvalue, and calculating the ratio of the first eigenvalue to each second eigenvalue as the eigenvalue information.
6. The method of claim 1, further comprising:
acquiring a plurality of vibration signals generated by the vibration sensor, wherein the vibration signals are generated by detecting vibration generated by the ball striking a plurality of preset striking positions on the batting sports equipment;
respectively carrying out frequency spectrum analysis on the vibration signals to obtain a plurality of eigenvectors of the vibration signals on a frequency domain, and calculating the characteristic information by utilizing the amplitude of each eigenvector; and
and taking the characteristic information as an input of the prediction model, taking the corresponding impact position as an output of the prediction model, training the prediction model, and recording a plurality of learning parameters of the trained prediction model.
7. The method of claim 1, wherein the machine learning comprises a decision tree, a convolutional neural network, a deep neural network, or a support vector machine.
8. The method of claim 1, wherein the vibration sensor comprises one or a combination of a piezoelectric vibration sensor, an electrodynamic vibration sensor, an eddy current vibration sensor, an inductive vibration sensor, a capacitive vibration sensor, a resistive vibration sensor, and a photoelectric vibration sensor.
9. A ball striking point position detection device for a ball striking sports apparatus, comprising:
a data acquisition device coupled to a vibration sensor disposed on the ball striking piece of athletic equipment, the vibration sensor detecting vibrations of the ball striking piece of athletic equipment to generate a vibration signal;
a storage device storing a plurality of learning parameters of a prediction model established in advance using machine learning, wherein the prediction model is trained using characteristic information of a plurality of vibration signals and a corresponding plurality of impact positions, the impact positions being positions at which the ball striking piece collides with a ball; and
a processor, coupled to the data acquisition apparatus and the storage device, configured to:
acquiring, by the data acquisition device, the vibration signal generated by the vibration sensor detecting vibrations generated by the impact of the ball striking piece with the ball;
performing spectrum analysis on the vibration signal to obtain a plurality of eigenfrequencies of the vibration signal in a frequency domain; and
at least one characteristic information is calculated using the amplitudes of each of the eigenfrequencies and input into the predictive model to estimate the impact location of the ball on the ball striking piece of athletic equipment.
10. The ball striking point position detection apparatus of ball striking equipment according to claim 9, wherein the processor includes selecting a primary eigenfrequency and at least one secondary eigenfrequency according to a magnitude of the amplitude of the eigenfrequency, and calculating the characteristic information using the amplitudes of the primary eigenfrequency and the secondary eigenfrequency.
11. The ball striking point location detection apparatus of ball striking equipment according to claim 10, wherein the processor includes calculating a ratio of the primary eigenfrequency to the amplitude of each of the secondary eigenfrequencies as the characteristic information.
12. The ball striking point position detection apparatus of ball striking equipment of claim 11, wherein the processor further calculates a ratio of the amplitudes between the secondary eigenfrequencies as the characteristic information.
13. The ball striking point position detection apparatus of ball striking equipment according to claim 10, wherein the processor includes calculating a first eigenvalue using the amplitude of the primary eigenfrequency, calculating at least one second eigenvalue using the amplitude of each secondary eigenfrequency, and calculating a ratio of the first eigenvalue to each second eigenvalue as the eigenvalue.
14. The batting sports equipment batting point position detection device according to claim 9, the processor further obtains a plurality of vibration signals generated by detecting vibrations generated by the ball striking a plurality of impact positions preset on the batting sports equipment by using the data acquisition equipment, performs spectrum analysis on the vibration signals respectively to obtain a plurality of eigenfrequencies of the vibration signals in a frequency domain, calculates the characteristic information by using the amplitudes of the eigenfrequencies, takes the characteristic information as an input of the prediction model, takes the corresponding impact positions as an output of the prediction model, trains the prediction model, and records the learning parameters of the trained prediction model in the storage device.
15. The ball striking point position detection apparatus of ball striking sports equipment of claim 9, wherein the machine learning comprises a decision tree, a convolutional neural network, a deep neural network, or a support vector machine.
16. The ball striking point position detection apparatus of ball striking sports equipment of claim 9, wherein the vibration sensor comprises one of a piezoelectric vibration sensor, an electrodynamic vibration sensor, an eddy current vibration sensor, an inductive vibration sensor, a capacitive vibration sensor, a resistive vibration sensor, and a photoelectric vibration sensor, or a combination thereof.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210141970.8A CN116662892A (en) | 2022-02-16 | 2022-02-16 | Method and device for detecting ball hitting point position of ball hitting sports equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210141970.8A CN116662892A (en) | 2022-02-16 | 2022-02-16 | Method and device for detecting ball hitting point position of ball hitting sports equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116662892A true CN116662892A (en) | 2023-08-29 |
Family
ID=87719359
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210141970.8A Pending CN116662892A (en) | 2022-02-16 | 2022-02-16 | Method and device for detecting ball hitting point position of ball hitting sports equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116662892A (en) |
-
2022
- 2022-02-16 CN CN202210141970.8A patent/CN116662892A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10369409B2 (en) | Dynamic sampling in sports equipment | |
| CN104524755B (en) | Swing analyzing apparatus | |
| JP3624761B2 (en) | Swing measurement method and golf swing analysis method | |
| AU2005302549B2 (en) | Golf club waggle shaking moving vibrating weight mass computer process | |
| US11173362B2 (en) | Analysis apparatus, analysis method, and recording medium | |
| JP6235692B2 (en) | Impact and acoustic analysis for golf equipment | |
| US9217753B2 (en) | Impact and sound analysis for golf equipment | |
| JP6268274B2 (en) | Impact and acoustic analysis for golf equipment | |
| US9245457B2 (en) | Impact and sound analysis for golf equipment | |
| US8808105B2 (en) | Fitting system for a golf club | |
| CN116662892A (en) | Method and device for detecting ball hitting point position of ball hitting sports equipment | |
| TWI788226B (en) | Method and apparatus for impact position detection of impact sport equipment | |
| US20180169471A1 (en) | Selection support apparatus, selection support system, and selection support method | |
| CN115867956A (en) | Grip analysis system and method | |
| US20180229079A1 (en) | Data processing method, program, storage medium and motion analysis device | |
| JP7058435B1 (en) | Golf club with built-in 9-axis sensor and swing evaluation system using it | |
| US20220203180A1 (en) | Information processing apparatus and non-transitory computer-readable storage medium | |
| KR20160111269A (en) | System for service of Golf swing motion correction for standard position and method thereof | |
| CN106132492A (en) | Information providing method, information provider unit, information providing system and information provision procedure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |