Swing evaluation device, swing evaluation method, and computer program

Abstract

The present invention provides a device capable of evaluating the magnitude of kinetic energy during a swing. [Solution] This is a device for evaluating a swing. This device includes an acquisition unit that acquires motion values ​​measured from the player's swing, an expansion unit that expands the motion values ​​into first data showing the coordinated motion of the swing and second data showing the magnitude of the coordinated motion in time series for each of a plurality of modes, and a calculation unit that calculates third data D3 in which each of the magnitudes of the coordinated motion included in the second data is converted into mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing, for at least one of the plurality of modes.

Description

[Technical Field]

[0001] This invention relates to a swing evaluation device, a swing evaluation method, and a computer program. [Background technology]

[0002] Patent Document 1 below describes a swing evaluation device. This device includes an acquisition unit that acquires first motion values ​​measured from the first swing, which is the golf swing of a first player, and an extraction unit that extracts first coordinated motion data showing the characteristic behavior of the first swing by singular value decomposition of the first motion values. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2024-79159 [Overview of the project] [Problems that the invention aims to solve]

[0004] Generally, the quality of a swing should be evaluated not only by the posture during the swing, but also by the magnitude of the kinetic energy. Therefore, it would be useful to be able to evaluate the magnitude of the kinetic energy during the swing.

[0005] This invention was devised in view of the above-described circumstances, and its main objective is to provide a device capable of evaluating the magnitude of kinetic energy during a swing. [Means for solving the problem]

[0006] The present invention relates to a swing evaluation device, comprising: an acquisition unit that acquires motion values ​​measured from a player's swing; an expansion unit that expands the motion values ​​into first data showing the coordinated motion of the swing and second data showing the magnitude of the coordinated motion in time series for each of a plurality of modes; and a calculation unit that calculates third data in which each of the magnitudes of the coordinated motion included in the second data is converted into mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing, for at least one of the plurality of modes. [Effects of the Invention]

[0007] By adopting the above configuration, the swing evaluation device of the present invention makes it possible to evaluate the magnitude of kinetic energy during a swing. [Brief explanation of the drawing]

[0008] [Figure 1] This is a diagram illustrating an example of a swing evaluation system including a swing evaluation device. [Figure 2] This is a block diagram showing an example of a swing evaluation device. [Figure 3] This flowchart shows an example of the processing procedure for a swing evaluation method. [Figure 4] This diagram illustrates an example of positional data for various parts of a player's body and for various parts of a golf club. [Figure 5] (a) is a diagram showing an example of a swing model based on the first mode of coordinated motion, and (b) is a diagram showing an example of a swing model based on the second mode of coordinated motion. [Figure 6] (a) is a graph of the second data D2 showing the relationship between the magnitude of the coordinated operation in the first mode and time, and (b) is a graph of the second data D2 showing the relationship between the magnitude of the coordinated operation in the second mode and time. [Figure 7] This is a flowchart showing an example of the processing steps for the calculation process. [Figure 8]This flowchart shows an example of the processing procedure for the calculation step of another embodiment of the present invention. [Figure 9] This is a diagram illustrating an example of a dyad product. [Figure 10] (a) is a graph of the third data set showing an example of the relationship between the mode energy of the first mode and time, and (b) is a graph of the third data set showing an example of the relationship between the mode energy of the second mode and time. [Figure 11] This graph shows an example of time-series data for total mode energy. [Figure 12] This figure shows an example of time-series data of total mode energy displayed on a display device. [Figure 13] This flowchart shows an example of the processing procedure for a swing evaluation method according to another embodiment of the present invention. [Figure 14] This graph shows the time-series data of the contribution of each mode's mode energy. [Figure 15] This figure shows an example of time-series data of contributions displayed on a display device for each of the multiple modes. [Modes for carrying out the invention]

[0009] Embodiments of the present invention will be described below with reference to the drawings. It should be understood that the drawings contain exaggerations and representations that differ from the actual dimensional ratios of the structures in order to aid in understanding the content of the invention. Furthermore, the same or common elements are denoted by the same reference numerals throughout each embodiment, and redundant explanations are omitted. Moreover, the specific configurations shown in the embodiments and drawings are for the purpose of understanding the content of the present invention, and the present invention is not limited to the specific configurations shown in the drawings.

[0010] Figure 1 is a configuration diagram showing an example of a swing evaluation system 1 including a swing evaluation device 2. In this embodiment of the swing evaluation system 1 (swing evaluation device 2), the swing of player P is evaluated. The player P being evaluated may be one person or multiple people. Furthermore, while the swing in this embodiment is exemplified as a golf swing, it is not limited to this form, and may be a swing performed in sports such as baseball or tennis, for example.

[0011] In this embodiment, when a golf swing is evaluated, a golf club 3 and a golf ball 4 are used. These golf clubs 3 and golf balls 4 may be the same as those used in the above-mentioned Patent Document 1, but are not particularly limited, and various types may be used depending on the purpose of evaluation, for example. Also, when a baseball swing is evaluated, for example, a bat and baseball (not shown) may be used, and when a tennis swing is evaluated, for example, a racket and tennis ball (not shown) may be used.

[0012] [Swing Evaluation System] The swing evaluation system 1 of this embodiment comprises a swing evaluation device 2 and a measuring device 5. The swing evaluation system 1 (swing evaluation device 2) is used to execute the swing evaluation method described later.

[0013] [Measuring device] The measuring device 5 is for measuring the swing of player P. In this embodiment, the measuring device 5 is configured as a motion capture system 5A, but it is not particularly limited, and various devices capable of measuring a swing can be used.

[0014] The measuring device 5 of this embodiment (in this example, the motion capture system 5A) includes a plurality of markers 6, a plurality of cameras 7, and a computer 8.

[0015] Multiple markers 6 are attached to various parts of the player P's body (e.g., head, wrists, fingertips, elbows, shoulders, waist, knees, ankles, and toes) as described in Patent Document 1, and are assigned predetermined identification numbers (e.g., numbers 1 to 53). Furthermore, multiple markers 6 are attached to predetermined parts of the golf club 3 (e.g., shaft 3a, club head 3b, and grip 3c), and are assigned predetermined identification numbers (e.g., numbers 54 to 57).

[0016] Multiple cameras 7 are used to capture video data of a player P, to whom multiple markers 6 are attached. A computer 8 is used to perform three-dimensional measurement of the swing based on the video data captured by the multiple cameras 7. These markers 6, cameras 7, and computer 8 may be the same as those used in the above-mentioned Patent Document 1, but are not particularly limited as long as they can measure the swing.

[0017] The measuring device 5 in this embodiment is connected to the swing evaluation device 2 via a wired or wireless communication line (not shown) so as to be able to communicate. This allows the measured values ​​measured by the measuring device 5 to be transmitted to the swing evaluation device 2.

[0018] [Swing evaluation device] The swing evaluation device 2 is comprised of, for example, a computer 10. Examples of computer 10 include a desktop computer, a notebook computer, a tablet computer, a smartphone, and a cloud server. In this embodiment, a desktop computer is used as computer 10. Figure 2 is a block diagram showing an example of the swing evaluation device 2.

[0019] The swing evaluation device 2 of this embodiment is configured to include, for example, an input device 11, a display device 12, a communication device 13, and a calculation processing device 14.

[0020] [Input devices and display devices] The input device 11 may include, for example, the keyboard 11a or mouse 11b shown in Figure 1. The display device 12 may include, for example, the display 12a shown in Figure 1.

[0021] [communication equipment] As shown in Figure 2, the communication device 13 of this embodiment is connected to the measuring device 5 (motion capture system 5A) via a communication line 20 so as to be able to communicate with it. This allows the communication device 13 (swing evaluation device 2) to acquire (receive) the measured values ​​of the measuring device 5 via the communication line 20. Furthermore, the communication device 13 (swing evaluation device 2) of this embodiment can, for example, transmit signals for controlling the measuring device 5 to the measuring device 5 via the communication line 20.

[0022] [Arithmetic Processing Unit] The arithmetic processing unit 14 of this embodiment is configured to include, for example, an arithmetic unit (CPU) 15 that performs various calculations, a storage unit 16 in which data, programs, etc. are stored, and a working memory 17.

[0023] [Storage] The storage unit 16 is a non-volatile information storage device, such as a magnetic disk, optical disk, or SSD. The storage unit 16 in this embodiment includes a data unit 18 and a program unit 19.

[0024] [Data Section] The data unit 18 stores data (information) necessary for evaluating the swing of player P as shown in Figure 1, as well as evaluation results. In this embodiment, the data unit 18 includes an operating value input unit 18a, a first data input unit 18b, a second data input unit 18c, a third data input unit 18d, a fourth data input unit 18e, and a fifth data input unit 18f. Furthermore, the data unit 18 includes a total input unit 18g, a grand total input unit 18h, and a contribution input unit 18i. Note that the data unit 18 is not limited to this configuration; for example, some of these may be omitted, or a data unit for storing other data may be included. Details of the data input to these data units 18 will be described later.

[0025] [Programming Department] The program unit 19 is a program (computer program) necessary for evaluating the swing of player P as shown in Figure 1. The program unit (program) 19 is executed by the calculation unit 15, thereby enabling the computer 10 to function as a specific means.

[0026] The program unit 19 of this embodiment includes an acquisition unit 19a, an expansion unit 19b, a calculation unit 19c, a total calculation unit 19d, a total display unit 19e, a comprehensive calculation output unit 19f, a contribution calculation unit 19g, and a contribution display unit 19h. The calculation unit 19c also includes a first calculation unit 21, a second calculation unit 22, and a third calculation unit 23. The program unit 19 is not limited to this embodiment, and may further include program units with other functions. Details of the functions of these program units 19 will be described later.

[0027] [Swing evaluation method (first embodiment)] Next, the swing evaluation method of this embodiment will be described. In the swing evaluation method of this embodiment, the swing of player P shown in Figure 1 is evaluated. Generally, it is important to evaluate the quality of a swing not only by the posture of player P during the swing, but also by the magnitude of the kinetic energy. This is because the magnitude of the kinetic energy has a significant effect on the distance the ball (for example, the golf ball 4 shown in Figure 1) travels. In the swing evaluation method of this embodiment, the magnitude of the kinetic energy during the swing is evaluated.

[0028] Figure 3 is a flowchart showing an example of the processing procedure for the swing evaluation method. In this embodiment, each step of the swing evaluation method is performed by the swing evaluation device 2 (computer 10) shown in Figures 1 and 2.

[0029] [Obtain the player's swing motion data] In the swing evaluation method of this embodiment, first, motion values ​​are obtained by measuring the swing of player P (step S1). The motion values ​​can be obtained as appropriate, as long as they can quantitatively represent the swing of player P. The motion values ​​of this embodiment include position data of each part of player P's body. Furthermore, the motion values ​​of this embodiment may also include position data of each part of the golf club 3.

[0030] In step S1 of this embodiment, first, the acquisition unit 19a included in the program unit 19 shown in Figure 2 is loaded into the working memory 17. The acquisition unit 19a is a program for acquiring the motion values ​​measured from the swing of player P. When this acquisition unit 19a is executed by the calculation unit 15, the computer 10 can function as a means for acquiring motion values.

[0031] Step S1 of this embodiment can be carried out in the same procedure as step S1 of acquiring the first operating value in Patent Document 1 described above. That is, in step S1 of this embodiment, first, multiple markers 6 are attached to various parts of the player P's body and various parts of the golf club 3 as shown in Figure 1. Then, from the start to the end of the swing by player P (for example, from address to finish), the swing is continuously filmed by multiple cameras 7. As a result, a group of images (video data) is acquired that includes multiple images of the swing continuously filmed in chronological order (multiple time points). Note that the setting of the sampling frequency and image processing can be carried out in the same manner as in Patent Document 1 described above, for example, but can be carried out as appropriate depending on the performance of the camera 7, etc.

[0032] In this embodiment, the position data of each of the multiple markers 6 is acquired as operating values. These position data allow for the identification of the positions of various body parts of the player P to which the multiple markers 6 are attached (position data of each body part), and the positions of various parts of the golf club 3 (position data of the shaft 3a, club head 3b, and grip 3c).

[0033] The positional data for each body part of player P (53 in this example) may be aggregated into a subset of positional data (22 in this example) by averaging the joint centers of those body parts. This reduces the amount of positional data, preventing the number of modes in the coordinated motion data obtained by singular value decomposition, described later, from increasing unnecessarily. Alternatively, even without such aggregation, the positional data for each body part of player P that is of interest when analyzing the swing may be limited, for example.

[0034] Figure 4 illustrates an example of positional data for each body part 25 of player P and positional data for each club part 26 of golf club 3. In Figure 4, (a) is the positional data viewed from the X-axis direction (front of player P), and (b) is the positional data viewed from the Y-axis direction (side of player P). The positional data for each body part 25 of player P and the positional data for each club part 26 of golf club 3 are acquired as three-dimensional (Cartesian coordinate system) coordinate values. The three-dimensional coordinate values ​​are determined, for example, in the front view of player P shown in Figure 4(a), based on the depth direction (X-axis direction), left-right direction (Y-axis direction), and height direction (Z-axis direction). By determining these positional data in a time series, motion values ​​are obtained.

[0035] The coordinate values ​​of each body part 25 and each club part 26 may be transformed (moved) so that, for example, the coordinate value of one body part 25 selected from a plurality of body parts 25 coincides with a predetermined reference position 24. By obtaining motion values ​​based on such coordinate values, it becomes easier to compare, for example, the motion values ​​of other players with different physiques or the motion values ​​of other players whose motions were measured at different locations. Details of the reference position 24 are as described in Patent Document 1 above. Furthermore, the coordinate values ​​of each body part 25 and each club part 26 may be transformed so that, for example, the vector (not shown) from the position of the right toe 25b of player P to the position of the left toe 25a is parallel to the Y-axis direction (left-right direction). This ensures that motion values ​​with the orientation of player P's body are aligned are obtained, making it easier to compare them with the motion values ​​of other players P.

[0036] The operating values ​​of this embodiment are defined by the following equations (1) to (3), similar to those in Patent Document 1 described above.

[0037]

number

[0038]

number

[0039]

number

[0040] In equation (1) above, the position data (position vector relative to the origin) of the i-th marker 6 (each body part 25) is identified at any time t from the first time (e.g., the start of the swing) to the Nth time (e.g., the end of the swing). Also, x in equation (1) above i (t), y i (t) and zi r(t) represents the coordinates in the X-axis direction, Y-axis direction, and Z-axis direction of the i-th marker 6 (each body part 25) at time t. This position data r i (t) is created from the first time to the N-th time, and the matrix [r i arranged row by row is defined by the above formula (2).

[0041] The matrix [r i in the above formula (2) is a matrix with N rows and 3 columns. In order along the time series from the first row to the N-th row, the three-dimensional coordinates r i (1), r i (2), ···, r i (N) are arranged. And for the 22 matrices [r1], [r2]... [r 22 for the marker 6 (each body part 25), the matrix [R] arranged column by column (horizontally) in this order is defined by the above formula (3). Note that the matrix [R] is a matrix with N rows and 66 (= 3×22) columns.

[0042] The matrix [R] in the above formula (3) is time-series behavior data representing the positions of 22 markers 6 (each body part 25) from the first time to the N-th time in time series. Note that the t-th row of the matrix [R] contains the three-dimensional coordinates of 22 markers 6 (each body part 25) on the player P's body at the t-th time.

[0043] In the t-th row of the matrix [R] in the above formula (3), the position data (swing by the player P) of each body part 25 of the player P at the t-th time is represented. With such a matrix [R], the motion value can be specified. Also, in step S1 of the present embodiment, a swing model (stick picture) M including bones 27 that connect a plurality of body parts 25 specified by the matrix [R] as joints may be set. The motion value and the swing model M are stored in the motion value input unit 18a shown in FIG. 2.

[0044] [Expand from motion value to first data and second data] Next, in the swing evaluation method of this embodiment, the operating values ​​are expanded into first data and second data for each of the multiple modes (step S2). The first data shows the coordinated movement of the swing. The second data shows the magnitude of the coordinated movement of the swing in chronological order.

[0045] In step S2 of this embodiment, first, the operating values ​​input to the operating value input unit 18a shown in Figure 2 are read into the working memory 17. Furthermore, the expansion unit 19b included in the program unit 19 is read into the working memory 17. The expansion unit 19b is a program for expanding the operating values ​​into first data and second data for each of multiple modes. When this expansion unit 19b is executed by the calculation unit 15, the computer 10 can function as a means for expanding into first data and second data.

[0046] The expansion of the operating values ​​(matrix [R] in equation (3) above) into the first and second data can be performed as appropriate. Such expansion is preferably performed based on known singular value decomposition or modal analysis, for example.

[0047] In this embodiment, the operating values ​​(the matrix [R] in equation (3) above) are subjected to singular value decomposition to obtain singular values, right singular vectors, and left singular vectors. Of these singular values, right singular vectors, and left singular vectors, the right singular vectors are spatial basis vectors that represent the positional information of each marker 6 (each body part 25) in player P shown in Figure 4. Since such right singular vectors indicate the movement (coordinated movement) of each body part 25 in each mode, they can be developed as first data of coordinated movement that shows the characteristic behavior of the swing. On the other hand, the left singular vectors are temporal basis vectors that represent the temporal information of the right singular vectors. Since such left singular vectors indicate the magnitude of the movement of each body part 25 (i.e., coordinated movement that shows the characteristic behavior of the swing) in time series in each mode, they can be developed as second data.

[0048] In step S2, singular value decomposition may be performed on the operating values ​​(matrix [R] in equation (3) above), but it is preferable that the operating values ​​are expanded prior to the singular value decomposition, similar to Patent Document 1 above. By expanding the operating values ​​(matrix) in this way, it is possible to bring the reference point for singular value decomposition closer to the initial swing position. The expansion of the operating values ​​is performed based on the following equations (4) to (6).

[0049]

number

[0050]

number

[0051]

number

[0052] The matrix extension first involves the coordinate values ​​(position vectors) of the initial posture of the i-th marker (25 body parts shown in Figure 4) r i (1) is obtained based on the above equation (1). Next, the obtained coordinate value (position vector) r i (1) A matrix ([r i (1)]) is created. At this time, [r i (1) All elements in each row are r i (1) Next, the above matrix ([r i (1)]) is created for markers 1 to 22 (25 body parts), and the matrix ([R(1)]) obtained by arranging them column by column is defined by equation (4) above.

[0053] Furthermore, in order to increase the number of data points in the matrix and improve the resolution, the operating value (matrix [R] in equation (3) above) is inverted in the time series direction to form matrix [R t] is defined by equation (5) above. Then, the operating value (matrix [R] from equation (3) above is the matrix [R(1)] from equation (4) above, and the matrix [R] from equation (5) below. t The observation matrix [R] is a matrix formed by concatenating the two ]. a ] is defined by equation (6) above.

[0054] In step S2 of this embodiment, similar to Patent Document 1 described above, the time series data of the positions of the 22 markers 6 (each body part 25) are obtained using the observation matrix [R] of the above equation (6). a The matrix [R0] represents the reference initial posture (average posture), and singular value decomposition is performed based on the following equation (7). Note that the matrix [R0] is the observation matrix [R a It is defined as a matrix (6N rows x 66 columns = 3 x 22) obtained by taking the average in the time direction (row direction) of ] and arranging 6N of these averages in each row.

[0055]

number

[0056] Based on equation (7) above, the observation matrix [R a [R0] is the difference between [R0] and the initial pose matrix. a ]-[R0] is decomposed into singular value decomposition, and for n=1,2,···,J (where J is an integer less than or equal to 66), Γ (n) , V (n) and U (n) This can be obtained.

[0057] The Γ in equation (7) above (n) This is the singular value of the nth mode (the nth one), and [R a This represents the proportion of the nth mode's (n-th) contribution to [the given state].

[0058] V in equation (7) above (n) is the observation matrix [R a This is the right singular vector of ]. The right singular vector V of this embodiment (n) This is 66-dimensional. This right singular vector V (n)This is a spatial basis that represents the positional information of each marker 6 (each body part 25) in the first player P1.

[0059] Spatial basis (right singular vector V) (n) ) is a vector indicating the direction of motion of the nth mode (the nth) relative to the reference initial posture matrix [R0]. With these spatial basis values, the movement (direction of movement) of each body part 25 of player P during the swing can be quantified for multiple modes (in this example, modes 1 to 66). Therefore, the right singular vector V (n) This can be derived from the motion values ​​as the first data of the coordinated movement that shows the characteristic behavior of the swing in each mode (n-th mode).

[0060] Figure 5(a) shows an example of a swing model M based on the first mode of coordinated operation. Figure 5(b) shows an example of a swing model M based on the second mode of coordinated operation. The swing model M is, for example, based on the procedure for creating a stick diagram described in Japanese Patent Application Publication No. 2020-185370, and the first data D1 (right singular vector V) is expanded from the operating values. (n) ) can be generated using ). In this embodiment, swing models M for the address, back 9 o'clock, top, down 9 o'clock, impact, follow 3 o'clock, and finish in a golf swing are shown, but are not particularly limited.

[0061] In the first mode swing model M shown in Figure 5(a), the arms are swung up at the top of the swing. At impact, the arms are swung down, and at the finish, the arms are swung up again. Meanwhile, the postures at 9 o'clock during the backswing, 9 o'clock during the downswing, and 3 o'clock during the follow-through converge to approximately the same posture (address posture). Therefore, in the first mode, the up-and-down movement of the arms can be extracted as a coordinated swing motion (characteristic behavior) shown in the first data D1.

[0062] On the other hand, in the second mode swing model M shown in Figure 5(b), the arms swing from side to side at 9 o'clock in the backswing, 9 o'clock in the downswing, 3 o'clock in the follow-through, and at the finish. Meanwhile, the top and impact positions converge to approximately the same position (address position). Therefore, in the second mode, the left-right movement of the arms can be extracted as a coordinated movement (characteristic behavior) of the swing shown in the first data D1. In this way, the movement values ​​are obtained for each of the multiple modes, using the first data D1 (right singular vector V (n) By expanding the data in this way, characteristic behaviors of the swing can be extracted. These first data D1s are input to the first data input unit 18b (shown in Figure 2) for each of the multiple modes (in this example, each of the 1st to 66th modes).

[0063] Note that the first data D1 (right singular vector V (n) While this method extracts characteristic swing behaviors (direction of movement of each body part), it does not include the characteristics of their time-series changes.

[0064] Next, U in equation (7) above (n) is the observation matrix [R a This is the left singular vector of ]. The left singular vector U of this embodiment (n) This will have 6N rows and 66 columns (=3 × 22). This left singular vector U (n) This is the right singular vector V (n) It is a time basis that represents the time information of [the system].

[0065] Time basis (left singular vector U) (n) The left singular vector U is a time-series vector that holds the magnitude (phase information) of the nth mode (the nth) of motion relative to the reference initial posture matrix [R0]. Based on the value of this time basis vector, the magnitude of the player P's movement (the movement of each body part 25) during the swing can be quantified in a time series (from the 1st time to the 6Nth time) for each of the multiple modes (in this example, the 1st mode to the 66th mode). Therefore, the left singular vector U (n) This can be derived from the operating values ​​as a second set of data showing the magnitude of coordinated operation over time for each of the multiple modes.

[0066] Figure 6(a) is a graph of the second data D2, showing the relationship between the magnitude of the coordinated motion in the first mode and time. This graph shows the magnitude of the coordinated motion in the first mode, as shown in Figure 5(a), in a time series. In this embodiment, the coordinated motion in the first mode (up and down movement of the arm in this example) is large from -2 seconds to 0 seconds (impact).

[0067] Figure 6(b) is a graph of the second data D2, showing the relationship between the magnitude of the coordinated movement in the second mode and time. This graph shows the magnitude of the coordinated movement in the second mode, as shown in Figure 5(b), in a time series. In this embodiment, the magnitude of the coordinated movement in the second mode (in this example, the left-right movement of the arms) peaks immediately after -1.5 seconds, -0.4 seconds, and 0 seconds (impact).

[0068] In this way, the operating values ​​are expanded into the second data D2 for each of the multiple modes, allowing the time at which the coordinated operation (characteristic behavior) of each mode becomes larger during the swing to be identified. Such second data D2 is input to the second data input unit 18c (shown in Figure 2) for each of the multiple modes (in this example, each of the 1st to 66th modes).

[0069] In this embodiment, for each of the multiple modes, the posture during the swing can be evaluated by referring to first data D1 (shown in Figures 5(a) and (b)) which shows the coordinated motion of the swing, and second data D2 (shown in Figures 6(a) and (b)) which shows the magnitude of the coordinated motion in time series. Of these first data D1 and second data D2, second data D2 differs from first data D1 in that although it shows the magnitude of the coordinated motion (phase information) in time series, it cannot identify (evaluate) the swing speed or the magnitude of the kinetic energy. This is because the magnitude of the coordinated motion shown in second data D2 is the motion value (observation matrix [R a ]) and the observation matrix [R a The difference between the initial pose matrix [R0] and the matrix [R] averaged over time (row direction) (i.e., [R aSince ]-[R0]) is obtained through singular value decomposition, it shows the time-series change in magnitude in the direction of the first data D1, and does not represent the actual swing speed or kinetic energy.

[0070] In the evaluation method of this embodiment, in the following calculation step S3, in at least one of the multiple modes, each of the magnitudes of the coordinated motion included in the second data D2 shown in Figure 6 is converted into mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing. By calculating such mode energy, it becomes possible to evaluate the magnitude of the kinetic energy of the swing in each mode.

[0071] [Calculation Process (First Embodiment)] Next, in the swing evaluation method of this embodiment, a third data is calculated by converting each of the magnitudes of the coordinated motion included in the second data D2 into mode energy in at least one of the multiple modes (calculation step S3).

[0072] In the calculation step S3 of this embodiment, third data converted to mode energy is calculated for at least two of the multiple modes (the first mode and the second mode in this example). The modes for which the third data is calculated are appropriately selected according to the characteristic behavior of the swing to be evaluated (for example, "up and down movement of the arms" in the first mode). In this embodiment, an example is given of how the third data converted to mode energy is calculated in the first mode and the second mode, but the embodiment is not limited to this example.

[0073] Mode energy is a physical quantity with the same dimensions as the kinetic energy of the swing. The third data set is the time-series data of mode energy. Details of mode energy will be described later.

[0074] In the calculation step S3 of this embodiment, first, the second data D2 (shown in Figure 6) input to the second data input unit 18c shown in Figure 2 is loaded into the working memory 17. Furthermore, the calculation unit 19c (in this example, including the first calculation unit 21, the second calculation unit 22, and the third calculation unit 23) included in the program unit 19 is loaded into the working memory 17. The calculation unit 19c is a program for calculating third data, which is obtained by converting each of the magnitudes of the cooperative operation included in the second data D2 into mode energy in at least one of the multiple modes. By executing this calculation unit 19c by the arithmetic unit 15, the computer 10 can be made to function as a means for calculating the third data.

[0075] The third data can be calculated as appropriate by converting the magnitude of the coordinated motion included in the second data D2 shown in Figures 6(a) and (b) into mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing. In general, it is known that kinetic energy F can be calculated by multiplying 1 / 2 by the mass m and the square of the velocity v (hereinafter sometimes referred to as the "square of the velocity") v2, based on the following equation (8).

[0076]

number

[0077] When calculating the kinetic energy F of the swing, one can substitute the swing velocity v and the mass m of the player P into equation (8) above. Note that the swing velocity v can be obtained by differentiating the swing displacement. As mentioned above, the magnitude of the coordinated motion shown in the second data D2 in Figure 6 (left singular vector U) (n) ) does not represent the swing velocity v. However, the second data D2 correlates with the swing displacement in that it shows the magnitude of the coordinated movement over time.

[0078] In this embodiment, in any mode, the second data D2 (left singular vector U) at each time step (n)A fourth data point is obtained by differentiating each of the second data points D2 with respect to time. This fourth data point shows the time change (rate of change) of the magnitude of the coordinated motion during the swing at each time point. Note that by differentiating the second data point D2 with respect to time at each time point, the magnitude of the coordinated motion in the initial position (constant at each time point) becomes zero. As a result, the influence of the initial position (derivative value) is removed from the fourth data point. The singular value Γ decomposed by singular value decomposition is then obtained from this fourth data point. (n) and the right singular vector V (n) By multiplying by this, a fifth data point can be obtained that shows a time-series representation of the physical quantity relating to the swing velocity of any mode.

[0079] In the fifth data of this embodiment, physical quantities relating to the velocity of multiple markers 6 (each body part 25) shown in Figure 4 are acquired in a time series for any mode. By substituting these physical quantities relating to velocity into the velocity v in equation (8) above, the kinetic energy F of the swing in any mode can be calculated in a time series for each of the multiple markers 6 (each body part 25).

[0080] Furthermore, since the fifth data set contains physical quantities related to velocity acquired in time series for each of the multiple markers 6, substituting these physical quantities into the velocity v in equation (8) above tends to complicate the calculation of the kinetic energy of the swing. Therefore, it is preferable to simplify the calculation of the kinetic energy of the swing.

[0081] Here, in the swing by player P shown in Figure 1, it can be assumed that the mass of player P does not change from the start to the end of the swing. Also, even when evaluating the swings of different players P, the difference in their masses may be small. In these cases, if we assume that the mass m (including the mass of player P) is the same in equation (8) above, the kinetic energy F increases proportionally only to the square of the velocity v2. From this viewpoint, by omitting 1 / 2 and mass m in equation (8) above, a physical quantity consisting of the square of the velocity v2 is obtained, which, although different from the actual value of kinetic energy F, makes it possible to evaluate the relative magnitude of kinetic energy F. Therefore, in this embodiment, the square of the velocity v2 is treated as mode energy, which is a physical quantity of the same dimension as kinetic energy F (a physical quantity with 1 / 2 and mass m omitted).

[0082] In the calculation step S3 of this embodiment, a fifth data point representing a physical quantity related to the swing velocity in time series is acquired, and by squaring this fifth data point, a third data point is calculated that is converted into mode energy, a physical quantity of the same dimension as the kinetic energy of the swing. As a result, it is not necessary to multiply the fifth data point by mass m (including the mass of player P), and the calculation of the third data point can be made less complex. Figure 7 is a flowchart showing an example of the processing procedure in the calculation step S3.

[0083] [Acquired the fifth data point for the physical quantity related to swing speed] In the calculation step S3 of this embodiment, first, in at least one of the multiple modes, a fourth data point is obtained by differentiating the magnitude of the coordinated motion included in the second data point D2 shown in Figure 6 with respect to time. Then, by multiplying the fourth data point, the singular value, and the first data point, a fifth data point is obtained that shows a physical quantity related to the swing velocity in time series (step S31). In step S31 of this embodiment, the fourth data point and the fifth data point are calculated in at least two modes (in this example, the first mode and the second mode).

[0084] In step S31 of this embodiment, the first calculation unit 21 included in the program unit 19 shown in Figure 2 is executed by the calculation unit 15. This first calculation unit 21 is a program for acquiring fourth data, which is obtained by differentiating the magnitude of the coordinated motion included in the second data D2 with respect to time, and fifth data, which shows the physical quantity related to the swing velocity in a time series. By executing this first calculation unit 21 by the calculation unit 15, the computer 10 can be made to function as a means for acquiring the fourth data and the fifth data.

[0085] In step S31 of this embodiment, first, in at least one of the multiple modes, the magnitude of the cooperative action included in the second data D2 is differentiated with respect to time to obtain the fourth data. In this embodiment, the fourth data is calculated in at least two of the multiple modes (in this example, the first mode and the second mode).

[0086] As described above, the second data point D2 is the left singular vector U in equation (7) above. (n) It is composed of the following. Furthermore, the second data D2 consists of a vector (in this example, 6N rows and 66 columns (=3×22)) that shows the magnitude of the coordinated movement over time for at least two of the multiple modes (in this example, the first mode and the second mode). In this embodiment, in each mode (in this example, the first mode and the second mode), the multiple vectors (vectors at each time point) that make up the second data D2 are each differentiated with respect to time. As a result, in each mode (in this example, the first mode and the second mode), a fourth data showing the time change in the magnitude of the coordinated movement during the swing can be obtained in a time series. In each mode (in this example, the first mode and the second mode), the fourth data is obtained as a matrix with the same dimensions as the second data (left singular vector) D2. The fourth data for each mode is input to the fourth data input unit 18e shown in Figure 2.

[0087] Next, in step S31 of this embodiment, for at least one nth mode (in this example, the first mode and the second mode) among the multiple modes, the fourth data (U (n)(Display "·" above it) and the singular value Γ (n) And, the first data (right singular vector V (n) ) is multiplied by each time step. This yields a fifth data point that shows the physical quantity related to the swing speed in a time series. The fifth data point is obtained based on the following equation (9).

[0088]

number

[0089] As mentioned above, the fourth data (U (n) The symbol "·" above it indicates that for each of the multiple modes (in this example, the first mode and the second mode), the first data (right singular vector V) is generated. (n) The time series shows the change in the magnitude of the coordinated action indicated by ). The first data (right singular vector V) that shows the coordinated action is shown in this fourth data. (n) ) and the singular value Γ (n) This is multiplied by . As a result, a fifth data point (indicated by "·" above [R(n)]) is obtained for each of the multiple modes (in this example, the first mode and the second mode), which shows the physical quantity related to the swing speed in time series.

[0090] In the fifth data of this embodiment, the physical quantity relating to the swing speed is the speed of each of the multiple markers 6 (each body part 25) shown in Figure 4 during the swing. The fifth data, in which the speeds of these multiple markers 6 are acquired in time series, is the observation matrix [R] of equation (6) above. a It is obtained as a matrix of the same dimension as ]. The fifth data for each mode is input to the fifth data input unit 18f shown in Figure 2.

[0091] [Calculate the third data by squaring the fifth data point] Next, in calculation step S3 of this embodiment, in at least one of the multiple modes, the third data is calculated by squaring the fifth data to convert it into mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing (step S32). In step S32 of this embodiment, the third data is calculated in at least two modes (in this example, the first mode and the second mode).

[0092] In step S32 of this embodiment, the second calculation unit 22, included in the program unit 19 shown in Figure 2, is executed by the arithmetic unit 15. This second calculation unit 22 is a program for calculating the third data by squaring the fifth data. By executing this second calculation unit 22 by the arithmetic unit 15, the computer 10 can function as a means for acquiring the third data.

[0093] In step S32 of this embodiment, for each mode (in this example, the first mode and the second mode), the fifth data (indicated by "·" above [R(n)]) in equation (9) above is squared. As described above, the fifth data represents a physical quantity related to the swing velocity (in this example, the velocities of multiple markers 6) in time series. Furthermore, by omitting 1 / 2 and mass m from equation (8) above, a physical quantity consisting of the squared value of the velocity v2 is obtained, which, although different from the actual value of the kinetic energy F, makes it possible to evaluate the relative magnitude of the kinetic energy F. From these viewpoints, by squaring the fifth data, which is obtained as a time series of velocities of multiple markers 6, a third data can be calculated, which is converted into mode energy, a physical quantity of the same dimension as the kinetic energy of the swing.

[0094] In this embodiment, the third data is obtained in time series for mode energy for each of the multiple markers 6 (each body part 25) in at least two of the multiple modes (in this example, the first mode and the second mode). The third data for each mode is input to the third data input unit 18d shown in Figure 2.

[0095] Thus, in the calculation step S3 of this embodiment, a third data can be calculated by converting at least one of the multiple modes (in this example, the first mode and the second mode) into a mode energy of the same dimension as the kinetic energy of the swing. With such mode energy, the magnitude of the kinetic energy of a characteristic behavior exhibited by any mode during the swing (for example, the up-and-down movement of the arm in the first mode) can be evaluated. Furthermore, in this embodiment, since it is not necessary to consider the mass of the player P in the calculation of the mode energy (third data), the third data can be easily calculated.

[0096] As described above, the fifth data in this embodiment acquires the velocity of each of the multiple markers 6 (each body part 25) during the swing in a time series for at least one of the multiple modes (in this example, the first mode and the second mode). For this reason, the size of the matrix showing the velocity of each marker 6 in a time series in the fifth data tends to increase in proportion to the number of modes selected. If such fifth data is squared as is, although it is not necessary to consider the mass of the player P, the calculation of the third data (mode energy) may become complicated or overestimated due to the product of vectors. Furthermore, in the third data, the mode energy is acquired in a time series for each of the multiple markers 6 (each body part 25) in any mode. For this reason, a separate calculation is required to evaluate the magnitude of the kinetic energy of the entire swing (i.e., the sum of the kinetic energy of all markers 6).

[0097] [Calculation Process (Second Embodiment)] As a result of diligent research, the inventors discovered that by multiplying the fifth data by the transpose matrix of the fifth data, the calculation of the third data can be simplified while simultaneously calculating the mode energy for the entire swing (i.e., aggregating all markers 6). Based on this finding, in calculation step S3 of this embodiment, the third data is calculated by multiplying the fifth data by the transpose matrix of the fifth data. Figure 8 is a flowchart showing an example of the processing procedure for calculation step S3 of another embodiment of the present invention.

[0098] [Acquired the fifth data point for the physical quantity related to swing speed] In the calculation step S3 of this embodiment, as in previous embodiments, in at least one of the multiple modes, the fourth data (U) is obtained by differentiating the magnitude of the cooperative action included in the second data D2 shown in Figure 6 with respect to time. (n) The following is obtained: (a "·" is displayed above it). Then, the fourth data and the singular value Γ (n) And, the first data (right singular vector V (n) By multiplying by ), a fifth data point is obtained that shows a physical quantity related to the swing speed in a time series (step S31). In step S31, as in step S31 of previous embodiments, the fourth data point and the fifth data point are obtained in at least two modes (in this example, the first mode and the second mode) based on the above equation (9).

[0099] [Calculate the third data point by multiplying the fifth data point by its transpose matrix.] Next, in the calculation step S3 of this embodiment, the third data is calculated by multiplying the fifth data by the transpose matrix of the fifth data in at least one of the multiple modes (step S33). In step S33 of this embodiment, the third data is calculated in at least two modes (in this example, the first mode and the second mode).

[0100] In step S33 of this embodiment, the third calculation unit 23, included in the program unit 19 shown in Figure 2, is executed by the arithmetic unit 15. This third calculation unit 23 is a program for calculating the third data by multiplying the fifth data by the transpose matrix of the fifth data. By executing this third calculation unit 23 by the arithmetic unit 15, the computer 10 can function as a means for acquiring the third data.

[0101] In step S33 of this embodiment, for at least one nth mode (in this example, the first mode and the second mode) among the multiple modes, the fifth data (indicated by "·" above [R(n)]) and the transpose matrix of the fifth data are multiplied based on the following equation (10). As a result, the third data is calculated for each nth mode (in this example, the first mode and the second mode).

[0102]

number

[0103] As shown in equation (10) above, for each nth mode (in this example, the first and second modes), the fifth data and the transpose matrix of the fifth data are multiplied. Of the above equation (10), the first data (right singular vector V (n) ) and the transpose matrix of the first data (i.e., V (n) T) and are multiplied together to form the product of orthogonal bases, thus becoming the identity matrix. This eliminates the need to multiply the first data points together. And the singular value Γ (n) The squared value of and the fourth data (U (n) The third data can be calculated by multiplying the product of (displaying a "·" above) and the transpose matrix of the fourth data by this product. Therefore, in this embodiment, the calculation of the third data can be simplified compared to previous embodiments in which the fifth data shown in equation (9) above was simply squared.

[0104] In equation (10) above, it is preferable that the product of the fourth data point and the transpose matrix of the fourth data point be performed based on the dyad product (dyadic calculation). This can simplify the calculation.

[0105] Figure 9 illustrates an example of a dyad product. Figure 9 shows the calculation result of the dyad product between the fourth data point of the first mode and the transpose matrix of this fourth data point. The fourth data point of the first mode is represented by equation (11) below. In this fourth data point, the physical quantity obtained by differentiating the magnitude of the cooperative action in the first mode with respect to time is shown in a time series for each time t.

[0106]

number

[0107] In the dyad product calculation results shown in Figure 9, the diagonal component 29, colored gray, yields the multiplication value obtained by multiplying physical quantities at the same time t. On the other hand, the off-diagonal components 30, excluding the diagonal component 29, yield the multiplication value obtained by multiplying physical quantities at different times t, which is unsuitable for calculating mode energy (third data) obtained in time series. Therefore, in the product of the fourth data and the transpose matrix of the fourth data, it is preferable that the multiplication value in the diagonal component 29 is extracted in time series.

[0108] And for the first mode, the singular value Γ (n) The squared value of is multiplied by the time series data of the product of the diagonal element 29 (i.e., the product of the fourth data and the transpose matrix of the fourth data). This allows for the calculation of the third data for the first mode, which is converted into mode energy, a physical quantity of the same dimension as the kinetic energy of the swing. Furthermore, for other modes among the multiple modes, the third data converted into mode energy can be calculated based on the same procedure as for the first mode.

[0109] In this embodiment, the third data is calculated in time series for the mode energy of the entire swing (i.e., aggregated for all markers 6) in at least one nth mode (in this example, the first mode and the second mode) among the multiple modes. The third data for each mode is input to the third data input unit 18d shown in Figure 2.

[0110] Thus, in the calculation step S3 of this embodiment, as in previous embodiments, a third data can be calculated by converting at least one of the multiple modes (in this example, the first mode and the second mode) into a mode energy of the same dimension as the kinetic energy of the swing. With such mode energy, the magnitude of the kinetic energy for any mode during the swing (for example, the up-and-down movement of the arm in the first mode) can be evaluated.

[0111] Furthermore, in the calculation step S3 of this embodiment, as in previous embodiments, it is not necessary to consider the mass of player P when calculating the mode energy (third data), so the third data can be easily calculated. Also, in the calculation step S3 of this embodiment, unlike previous embodiments, the fifth data and the transpose matrix of the fifth data are multiplied, thereby obtaining the first data (right singular vector V (n) The squared value of ) is converted to the identity matrix. This omits the product of the first data, further simplifying the calculation of the third data, and making it easier to calculate the mode energy for the entire swing (i.e., the sum of all markers 6).

[0112] Figure 10(a) is a graph of the third data point D3 showing an example of the relationship between the mode energy of the first mode and time. Figure 10(b) is a graph of the third data point D3 showing an example of the relationship between the mode energy of the second mode and time. In Figures 10(a) and (b), the line types for the backswing, downswing, and follow-through swings are different. Also, in Figures 10(a) and (b), the mode energy for multiple swings is shown by separate lines. Note that in Figures 10(a) and (b), the time of impact for the swing is 0 seconds.

[0113] As mentioned above, the third data set D3 acquires mode energy, a physical quantity of the same dimension as the kinetic energy of the swing, in a time series. Using this third data set D3, the magnitude of the kinetic energy for the characteristic behavior exhibited by each mode (for example, the up-and-down motion of the arm in the first mode) can be evaluated in a time series.

[0114] In Figure 10(a), the mode energy is high from the downswing to just before impact (time: 0 seconds), and also high from immediately after impact to the follow-through. This trend in mode energy differs from the trend of coordinated movement in the first mode, which is high from -2 seconds to 0 seconds (impact), as shown in Figure 6(a), and coincides with the timing when player P applies force in the characteristic behavior of the first mode (up and down movement of the arms in this example).

[0115] In Figure 10(b), the mode energy increases immediately before and after impact (time: 0 seconds), which is a different trend from the coordinated movement of the second mode shown in Figure 6(b), where it increases at -1.5 seconds, -0.4 seconds, and immediately after 0 seconds (impact). This trend in mode energy coincides with the timing when player P applies force in the characteristic behavior of the second mode (in this example, the lateral movement of the arm).

[0116] Thus, in this embodiment, by calculating the third data D3, the magnitude of the kinetic energy for the characteristic behavior exhibited by each mode (for example, the up-and-down movement of the arm in the first mode) can be evaluated over time.

[0117] [Calculate time-series data of total mode energy] Next, as shown in Figure 3, in the swing evaluation method of this embodiment, time-series data of the total mode energy is calculated by adding up the mode energies contained in the third data D3 of at least two modes (step S4). As described above, mode energy is treated as a physical quantity of the same dimension as the kinetic energy of the swing. Therefore, unlike the magnitude of the coordinated motion shown in the second data D2 in Figure 6, it is possible to directly add up the mode energies of multiple modes.

[0118] In step S4 of this embodiment, as shown in Figures 10(a) and (b), a third data D3, which is time-series data of mode energy, is calculated for each of at least two modes (the first mode and the second mode in this example) out of a plurality of modes. Therefore, in step S4 of this embodiment, the mode energy included in the third data D3 of the first mode (shown in Figure 10(a)) and the mode energy included in the third data D3 of the second mode (shown in Figure 10(b)) are added together at each time step. This calculates the time-series data of the total mode energy for the first and second modes.

[0119] In step S4 of this embodiment, first, the third data D3 of at least two modes (in this example, the first mode and the second mode) input to the third data input unit 18d shown in Figure 2 is loaded into the working memory 17. Furthermore, the total calculation unit 19d included in the program unit 19 is loaded into the working memory 17. This total calculation unit 19d is a program for calculating time-series data of the total mode energy, which is obtained by adding up the mode energies included in the third data D3 of at least two modes. When this total calculation unit 19d is executed by the calculation unit 15, the computer 10 can function as a means for calculating time-series data of the total mode energy.

[0120] As described above, the mode energy included in the third data (shown in Figures 10(a) and (b)) differs from the magnitude of the coordinated swing motion included in the second data D2, which is derived from the swing motion values ​​(shown in Figures 6(a) and (b)), and can identify the magnitude of the kinetic energy. Therefore, by calculating the time-series data of the total mode energy, which is the sum of the mode energies of these modes (in this example, the first mode and the second mode), the magnitude of the kinetic energy in the combined behavior, which combines the characteristic behaviors of these modes (the "up-and-down movement of the arm" in the first mode and the "left-and-right movement of the arm" in the second mode), can be evaluated in a time series.

[0121] Figure 11 is a graph showing an example of time-series data 31 of total mode energy. In Figure 11, the line types of mode energy are different for the backswing, downswing, and follow-through, and the time of impact of the swing is set to 0 seconds.

[0122] In step S4 of this embodiment, the mode energy of the third data D3 of the first mode shown in Figure 10(a) and the mode energy of the third data D3 of the second mode shown in Figure 10(b) are added together at each time step. As a result, time-series data 31 of the total mode energy can be obtained, as shown in Figure 11. This time-series data 31 of the total mode energy allows for the evaluation, over time, of the magnitude of the kinetic energy in a composite behavior that combines the characteristic behaviors of at least two of the multiple modes included in the swing (the "up-and-down movement of the arm" of the first mode and the "left-and-right movement of the arm" of the second mode). By evaluating the magnitude of the kinetic energy of this composite behavior, the individuality and challenges of the player P's swing can be evaluated. The time-series data 31 of the total mode energy is input to the total input unit 18g shown in Figure 2.

[0123] [Display time-series data of total mode energy] Next, in the swing evaluation method of this embodiment, the time-series data 31 of the total mode energy shown in Figure 11 is displayed (step S5). The time-series data of this embodiment is displayed on the display device 12 shown in Figure 1, but is not particularly limited and may be output to a printer or the like, for example.

[0124] In step S5 of this embodiment, first, the time-series data 31 of the total mode energy (shown in Figure 11), which is input to the total input unit 18g shown in Figure 2, and the total display unit 19e included in the program unit 19 are loaded into the working memory 17. This total display unit 19e is a program for displaying the time-series data 31 of the total mode energy. When this total display unit 19e is executed by the calculation unit 15, the computer 10 can be made to function as a means for displaying the time-series data 31 of the total mode energy.

[0125] Figure 12 shows an example of time-series data 31 of total mode energy displayed on the display device 12. In step S5, the time-series data 31 of total mode energy displayed on the display device 12 allows the player P, operator, etc., to understand the magnitude of the kinetic energy in the composite behavior, which combines the characteristic behaviors of at least two of the multiple modes included in the swing (the "up-and-down movement of the arm" of the first mode and the "left-and-right movement of the arm" of the second mode).

[0126] In step S5, the mode energies of the third data D3 of at least two modes (in this example, the first mode and the second mode) used to calculate the time-series data 31 of the total mode energy may be displayed (shown in Figures 10(a) and (b)). This allows for a comparison between the magnitude of the total mode energy and the magnitude of the mode energy of each mode.

[0127] In step S5, a spatial basis (right singular vector V) is created for at least two modes. (n) A swing model M obtained by summing up the ) may also be displayed. By comparing such a swing model M with the time series data 31 of the total mode energy, the magnitude of the kinetic energy in the combined behavior can be easily evaluated. The swing model M can be appropriately generated, for example, based on the procedure for creating a stick diagram described in Japanese Patent Application Publication No. 2020-185370.

[0128] In step S5, the motion values ​​of the swings of multiple players P are acquired, and the third data D3 calculated from these motion values ​​and the time-series data 31 of the total mode energy may be displayed on the display device 12. This makes it easy to compare the magnitude of the kinetic energy of multiple players P and identify areas for improvement in each player P's swing.

[0129] [Evaluate the swing] Next, in the swing evaluation method of this embodiment, the swing is evaluated to determine whether it is good or not (step S6). In step S6, for example, it is determined whether the third data D3 converted to mode energy and the time-series data 31 of the total mode energy meet predetermined criteria. The swing evaluation may be performed by an operator or by a computer 10. The criteria may be set, for example, according to the level of player P.

[0130] If the swing is judged to be good ("Yes" in step S6), the series of processes for the swing evaluation method is completed. On the other hand, if the swing is judged to be not good ("No" in step S6), steps S1 to S6 are performed again. In the repeated step S1, it is preferable that the swing of player P, which has been advised based on the evaluation result, is measured. This may improve player P's swing.

[0131] [Swing evaluation method (second embodiment)] In the swing evaluation method of this embodiment, the third data D3 of at least two modes (in this example, the first mode and the second mode) out of a plurality of modes is calculated, and time series data 31 of the total mode energy is calculated by adding up the mode energies of these modes. However, the method is not limited to this configuration. For example, the third data D3 of all modes out of a plurality of modes (e.g., the first mode to the 66th mode) may be calculated, and time series data of the grand total mode energy is calculated by summing the mode energies included in each of the third data D3 of all modes. Furthermore, time series data of the contribution of each mode to the grand total mode energy may be calculated for each of the plurality of modes. By calculating such time series data of the contribution of mode energy, it is possible to easily identify the mode (characteristic behavior) in which the influence of the player P's swing on the kinetic energy is relatively large at each time point during the swing.

[0132] Figure 13 is a flowchart showing an example of the processing procedure for a swing evaluation method according to another embodiment of the present invention. In the calculation step S3 of this embodiment, for all modes among the multiple modes, a third data D3 is calculated by converting the magnitude of the coordinated motion included in the second data D2 into mode energy.

[0133] In this embodiment, similar to step S2 of previous embodiments, the operating value is a singular value Γ for each of the multiple modes (in this example, the 1st mode to the 66th mode). (n) , 1st data (right singular vector V (n) ) and second data D2 (left singular vector U (n) ) is then expanded. In the calculation step S3 of this embodiment, the third data D3 for all modes (in this example, the 1st mode to the 66th mode) is calculated. In this embodiment, the third data is calculated based on the processing procedure of the calculation step S3 shown in Figure 8, but it may also be calculated based on the processing procedure of the calculation step S3 shown in Figure 7. The third data D3 for each of the modes is input to the third data input unit 18d shown in Figure 2.

[0134] [Calculate the total mode energy by summing up all mode energies] Next, in the swing evaluation method of this embodiment, time-series data of the total mode energy is calculated by summing the mode energies included in the third data D3 of each mode (step S7).

[0135] In step S7 of this embodiment, first, the third data D3 of all modes (in this example, the 1st mode to the 66th mode) input to the third data input unit 18d shown in Figure 2 is loaded into the working memory 17. Furthermore, the comprehensive calculation output unit 19f included in the program unit 19 is loaded into the working memory 17. This comprehensive calculation output unit 19f is a program for calculating time-series data of the total mode energy, which is the sum of the mode energies included in the third data D3 of all modes. When this comprehensive calculation output unit 19f is executed by the calculation unit 15, the computer 10 can function as a means for calculating time-series data of the total mode energy.

[0136] In step S7 of this embodiment, based on the following equation (12), the mode energy E(n) of all modes (in this example, the 1st mode to the 66th mode) is calculated for each time (t=0 to L) from the start to the end of the swing. t The total mode energy Eall t This is calculated.

[0137]

number

[0138] Next, in step S7 of this embodiment, the total mode energy Eall for each time (t=0 to L) is calculated based on the following equation (13). t However, they are arranged in chronological order (ascending). As a result, in step S7 of this embodiment, the total mode energy Eall t The time series data Eall is calculated. This time series data Eall is input into the total input unit 18h shown in Figure 2.

[0139]

number

[0140] [Calculate time-series data of the contribution of mode energy] Next, in the swing evaluation method of this embodiment, the total mode energy Eall is calculated for each of the multiple modes. t Mode energy E(n) for t Time-series data of the contribution of is calculated (step S8). As described above, mode energy E(n) t This is included in the third data set in a time series for each of the multiple modes.

[0141] In step S8 of this embodiment, first, the third data for each of the modes (in this example, the 1st mode to the 66th mode) input to the third data input unit 18d shown in Figure 2 is read into the working memory 17. Furthermore, the time-series data of the total mode energy Eall (shown in equation (13) above) input to the total input unit 18h and the contribution calculation unit 19g included in the program unit 19 are read into the working memory 17. This contribution calculation unit 19g calculates the total mode energy Eall for each of the multiple modes. t Mode energy E(n) for t This is a program for calculating time-series data of the contribution. This contribution calculation unit 19g is executed by the calculation unit 15, which enables the computer 10 to function as a means for calculating time-series data of the contribution.

[0142] In step S8 of this embodiment, based on the following formula (14), at each time (t = 0 to L) from the start to the end of the swing, the mode energy E(n) of the n-th mode (for example, the first mode) t is divided by the total mode energy Eall t . As a result, for the n-th mode, time-series data CE(n) of the contribution degree (E(n) t / Eall t ) of the mode energy to the total mode energy is obtained.

[0143]

Equation

[0144] Next, in step S8 of this embodiment, for each of the remaining modes (for example, the second mode to the 66th mode), time-series data CE(n) of the contribution degree (E(n) t / Eall t ) of the mode energy to the total mode energy is obtained.

[0145] At each time t, when the contribution degrees (E(n) t / Eall t ) of all modes (in this example, the first mode to the 66th mode) are added together, the value becomes "1". Also, at each time t, in the n-th mode where the contribution degree (E(n) t / Eall t ) is relatively large, the characteristic behavior of that n-th mode (for example, "up and down movement of the arm" in the first mode) has a relatively large influence on the kinetic energy of the swing. Thus, the contribution degree (E(n) t / Eall tBy obtaining the time-series data CE(n) of ), characteristic behaviors where the influence of the swing on the kinetic energy becomes relatively larger at each time t can be easily identified.

[0146] Figure 14 is a graph showing the time series data CE(n) of the mode energy contribution of each mode. In Figure 14, among multiple modes (modes 1 to 66 in this example), the time series data CE(1) of the contribution of mode 1 to CE(4) of mode 4, which have relatively large contributions, are shown as representative. Although the contributions are expressed as percentages (%), this is not particularly limited.

[0147] As shown in Figure 14, the contribution of the first mode (characteristic behavior "up and down arm movement") is relatively large from the address position to the backswing at 9 o'clock. On the other hand, the contribution of the second mode (characteristic behavior "left and right arm movement") is relatively large from the backswing at 9 o'clock to the top of the swing. Thus, the influence of each mode on the kinetic energy of the swing (contribution of mode energy) changes moment by moment. By calculating the time-series data CE(n) of the contribution of each mode's mode energy, characteristic behaviors that have a relatively large influence on the kinetic energy of the swing can be easily identified at each time t of the swing. The time-series data CE(n) of the contributions calculated for each of the multiple modes are input to the contribution input unit 18i shown in Figure 2.

[0148] [Display time-series data of contribution] Next, in the swing evaluation method of this embodiment, time-series data CE(n) of the contribution is displayed for each of the multiple modes (step S9). The time-series data CE(n) of this embodiment (shown in Figure 14) is displayed on the display device 12 shown in Figure 1, but is not particularly limited and may be output to a printer or the like, for example.

[0149] In step S9 of this embodiment, first, the time-series data CE(n) of the contribution level input unit 18i shown in Figure 2 and the contribution display unit 19h included in the program unit 19 are loaded into the working memory 17. This contribution display unit 19h is a program for displaying the time-series data CE(n) of the contribution level for each of the multiple modes. When this contribution display unit 19h is executed by the calculation unit 15, the computer 10 can be made to function as a means for displaying the time-series data CE(n) of the contribution level for each of the multiple modes.

[0150] Figure 15 shows an example of time-series data CE(n) of contributions displayed on the display device 12 for each of multiple modes. By displaying such time-series data CE on the display device 12, players P and operators can grasp the characteristic behavior in which the influence on the kinetic energy of the swing becomes relatively larger at each time t of the swing.

[0151] In step S9, time-series data CE(n) of contributions calculated from the swings of two different players P may be displayed. In Figure 15, the time-series data CE(n) of contributions for two players P (first player Pa, second player Pb) are displayed. Of the two players P, one player P may be more advanced than the other player P.

[0152] In player 1, Pa, the contributions of the second and third modes are significant before and after the top of the stroke. Therefore, in player 1, Pa, it can be evaluated that the movements (kinetic energy) associated with the characteristic behavior of the second mode (lateral arm movement) and the characteristic behavior of the third mode (stepping / kicking motion) are significant before and after the top of the stroke.

[0153] In the second player Pb, the contribution of the first mode is significant before and after the top of the stroke. Therefore, in the second player Pb, it can be evaluated that the movement (kinetic energy) related to the up-and-down motion of the arms is large before and after the top of the stroke.

[0154] In this way, the time-series data CE(n) of contributions calculated from the swings of the first player Pa and the second player Pb, which are different from each other, is displayed, allowing for a comparison of the characteristic behaviors (movements) of the swings of the first player Pa and the second player Pb.

[0155] [Evaluate the swing] Next, in the swing evaluation method of this embodiment, as in the previous embodiments, it is evaluated whether the swing is good or not (step S6). In step S6, for example, it is determined whether the time-series data CE(n) of the contribution meets a predetermined criterion. The criterion may be set, for example, according to the level of player P.

[0156] If the swing is judged to be good ("Yes" in step S6), the series of processes for the swing evaluation method is completed. On the other hand, if the swing is judged to be not good ("No" in step S6), steps S1 to S6 are repeated. This allows player P to improve their swing.

[0157] Although particularly preferred embodiments of the present invention have been described in detail above, the present invention is not limited to the illustrated embodiments and can be implemented in various modified forms.

[0158] [Note] The present invention includes the following embodiments.

[0159] [Invention 1] A device for evaluating a swing, An acquisition unit that acquires motion values ​​measuring the player's swing, From the aforementioned operating values, the expansion unit expands the data into first data showing the coordinated movement of the swing for each of the multiple modes, and second data showing the magnitude of the coordinated movement in time series. A calculation unit that calculates a third data set in which each of the magnitudes of the coordinated motion included in the second data is converted into a mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing, in at least one of the multiple modes, is included. Swing evaluation device. [2nd Invention] The swing evaluation device according to the present invention 1, wherein the expansion unit expands the right singular vector as the first data and the left singular vector as the second data from among the singular values, right singular vector and left singular vector obtained by singular value decomposition of the operating value. [Invention 3] The calculation unit obtains a fourth data obtained by differentiating the magnitude of the coordinated motion included in the second data with respect to time, and obtains a fifth data representing a physical quantity related to the swing velocity in a time series by multiplying the fourth data, the singular value, and the first data. A swing evaluation device according to the present invention, comprising: a second calculation unit that calculates the third data by squaring the fifth data. [4th Invention] The calculation unit obtains a fourth data obtained by differentiating the magnitude of the coordinated motion included in the second data with respect to time, and obtains a fifth data representing a physical quantity related to the swing velocity in a time series by multiplying the fourth data, the singular value, and the first data. The swing evaluation device according to the present invention, comprising: a third calculation unit that calculates the third data by multiplying the fifth data by the transpose matrix of the fifth data. [5th ​​Invention] The calculation unit calculates the third data in at least two of the multiple modes, A total calculation unit calculates time-series data of the total mode energy obtained by adding up the mode energies included in the third data for each of the at least two modes, A swing evaluation device according to any one of the present invention 1 to 4, comprising a total display unit that displays the time-series data of the total mode energy. [Invention 6] The calculation unit calculates the third data for all of the multiple modes, A comprehensive calculation unit calculates time-series data of the total mode energy obtained by summing the mode energies included in the third data for each of the modes, A contribution calculation unit calculates time-series data of the contribution of each of the multiple modes to the total mode energy, A swing evaluation device according to any one of the present invention 1 to 5, comprising a contribution display unit that displays time-series data of the contribution for each of the multiple modes. [7th Invention] A method for evaluating a swing, Computers The process involves acquiring motion values ​​by measuring the player's swing, From the aforementioned operating values, a process is performed to extract, for each of the multiple modes, first data showing the coordinated movement of the swing and second data showing the magnitude of the coordinated movement in time series. The process includes calculating a third data set in which each of the magnitudes of the coordinated motion included in the second data is converted into a mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing, in at least one of the multiple modes. Swing evaluation methods. [8th Invention] A computer program for evaluating a swing, Computers, A means of acquiring motion values ​​that measure the player's swing, A means for extracting, from the aforementioned operating values, first data showing the coordinated movement of the swing for each of the multiple modes, and second data showing the magnitude of the coordinated movement in time series, In at least one of the aforementioned multiple modes, the means for calculating a third data set is provided, which is obtained by converting each of the magnitudes of the coordinated motion included in the second data into a mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing. Computer program. [Explanation of Symbols]

[0160] D3 Third Data

Claims

1. A device for evaluating a swing, An acquisition unit that acquires motion values ​​measuring the player's swing, From the aforementioned operating values, an expansion unit expands the data into first data showing the coordinated movement of the swing for each of the multiple modes, and second data showing the magnitude of the coordinated movement in time series. A calculation unit that calculates a third data in which each of the magnitudes of the coordinated motion included in the second data is converted into a mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing, in at least one of the multiple modes, is included. Swing evaluation device.

2. The swing evaluation device according to claim 1, wherein the expansion unit expands the right singular vector as the first data and the left singular vector as the second data from among the singular values, right singular vector and left singular vector obtained by singular value decomposition of the operating value.

3. The calculation unit obtains a fourth data obtained by differentiating the magnitude of the coordinated motion included in the second data with respect to time, and obtains a fifth data which shows the physical quantity related to the swing speed in a time series by multiplying the fourth data, the singular value, and the first data. The swing evaluation device according to claim 2, further comprising: a second calculation unit that calculates the third data by squaring the fifth data.

4. The calculation unit obtains a fourth data obtained by differentiating the magnitude of the coordinated motion included in the second data with respect to time, and obtains a fifth data which shows the physical quantity related to the swing speed in a time series by multiplying the fourth data, the singular value, and the first data. The swing evaluation device according to claim 2, further comprising: a third calculation unit that calculates the third data by multiplying the fifth data by the transpose matrix of the fifth data.

5. The calculation unit calculates the third data in at least two of the multiple modes, A total calculation unit calculates time-series data of the total mode energy obtained by adding up the mode energies included in the third data for each of the at least two modes, The swing evaluation device according to claim 1, further comprising a total display unit that displays time-series data of the total mode energy.

6. The calculation unit calculates the third data for all of the multiple modes, A comprehensive calculation output unit calculates time-series data of the total mode energy obtained by summing the mode energies included in the third data for each of the modes, A contribution calculation unit calculates time-series data of the contribution of each of the multiple modes to the total mode energy, The swing evaluation device according to claim 1, further comprising a contribution display unit that displays time-series data of the contribution for each of the aforementioned multiple modes.

7. A method for evaluating a swing, Computers The process involves acquiring motion values ​​by measuring the player's swing, From the aforementioned operating values, a process is performed to extract, for each of the multiple modes, first data showing the coordinated movement of the swing and second data showing the magnitude of the coordinated movement in time series. The process includes calculating a third data in which each of the magnitudes of the coordinated motion included in the second data is converted into a mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing, in at least one of the multiple modes. Swing evaluation methods.

8. A computer program for evaluating a swing, Computers, A means of acquiring motion values ​​that measure the player's swing, A means for extracting from the aforementioned operating values, for each of the multiple modes, first data showing the coordinated movement of the swing and second data showing the magnitude of the coordinated movement in time series, In at least one of the aforementioned multiple modes, the means for calculating a third data set is provided, which is obtained by converting each of the magnitudes of the coordinated motion included in the second data into a mode energy, which is a physical quantity of the same dimension as the kinetic energy of the swing. Computer program.

Images