Classification device, classification method, and program

Abstract

To perform classification of swings in consideration of the priority of each index using a plurality of indexes in a simpler manner.SOLUTION: A classification device 10 includes: a skeleton position information acquisition unit 131 for acquiring skeleton position information indicating the position of a predetermined skeleton of the body of an object person 5 during a swing; an event acquisition unit 132 for acquiring a detection result of a predetermined event during the swing; an index calculation unit 133 for calculating a plurality of indexes on the basis of the skeleton position information in a detected notable event; and a swing classification unit 134 for classifying the swings of the object person 5 on the basis of the plurality of indexes. The plurality of indexes have hierarchical relationships, and the swing classification unit 134 classifies the swings by sequentially evaluating the calculated indexes according to the hierarchical relationships of the plurality of indexes.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present disclosure relates to a classification device, a classification method, and a program. [Background technology]

[0002] In golf lessons, etc., in order to efficiently improve a swing, an instructor may give advice to a user so that the user's posture is correct during a specific event during the swing (e.g., impact). Patent Document 1 describes a method for calculating a force index acting on a golfer's trunk during a swing and evaluating the stability of the swing based on the force index. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2018-089403 A Summary of the Invention [Problem to be solved by the invention]

[0004] In order to facilitate swing evaluation, swings are sometimes classified into several patterns. There are several indicators for classifying swings, and the importance of the several indicators is not necessarily equal. Therefore, instructors do not evaluate several indicators separately, but may evaluate swings in order of importance, for example, based on their own swing theories, and classify swings. There is a demand for a method that can more easily classify swings using several indicators and taking into account the priority of each indicator, without relying on instructors, etc.

[0005] An object of the present disclosure, made in consideration of the above-mentioned problems, is to provide a classification device, a classification method, and a program that can more easily classify swings using multiple indexes and taking into account the priority of each index. [Means for solving the problem]

[0006] (1) A classification device as one aspect of the present disclosure is a classification device that classifies a swing of a club by a subject based on multiple indices, and includes a skeletal position information acquisition unit that acquires skeletal position information indicating the positions of certain skeletal bones in the body of the subject during the swing, an event acquisition unit that acquires detection results of certain events during the swing, an index calculation unit that calculates the multiple indices based on the skeletal position information of a detected event of interest, and a swing classification unit that classifies the swing of the subject based on the multiple indices, wherein the multiple indices have a hierarchical relationship, and the swing classification unit classifies the swing by sequentially evaluating the calculated indices in accordance with the hierarchical relationship of the multiple indices. According to this configuration, a hierarchical relationship is set for the multiple indices, and the calculated indices are evaluated in order according to the hierarchical relationship, so that the higher the hierarchical level of the index, the greater the influence on the swing classification. Therefore, according to this embodiment, swing classification using multiple indices and taking into account the priority of each index can be performed more easily without relying on an instructor or the like.

[0007] (2) In the classification device of (1), the plurality of indexes include at least a first index calculated based on a predetermined skeletal position of the subject's upper body, and a second index calculated based on a predetermined skeletal position of the subject's lower body. With this configuration, swings can be classified based on the posture of the subject's upper body and lower body.

[0008] (3) In the classification device of (2), the second index is set in a higher hierarchy than the first index. According to such a configuration, it is possible to classify swings by giving priority to lower body movements, which are more important for improving a swing.

[0009] (4) In the classification device of (2) or (3), the first indicator includes a trunk angle of the subject, and the second indicator includes a lower leg angle of the subject. According to such a configuration, it is possible to more accurately classify swings based on the subject's trunk angle and the subject's lower leg angle, which are important indices for classifying swings.

[0010] (5) In any of the classification devices of (1) to (4), the plurality of indices includes an index calculated based on a change in the skeletal position information from an event that occurs temporally before the event of interest to the event of interest. Such a configuration enables more accurate classification of swings by taking into account temporal fluctuations in specific parts of the subject's body.

[0011] (6) In any of the classification devices of (1) to (5), the skeletal position information is information indicating the position of a specific skeletal structure of the subject's body projected onto a two-dimensional plane, the two-dimensional plane being a plane perpendicular to the ground and parallel to the flight path of the ball, or a plane perpendicular to the flight path of the ball. This configuration makes it possible to obtain an index relating to the left / right or front / back inclination of the body of the subject 5. As a result, it becomes possible to more accurately classify swings by taking into account the left / right or front / back inclination of the body of the subject 5.

[0012] (7) In any of the classification devices of (1) to (6), the swing classification unit compares the calculated index with a threshold value corresponding to the index to determine whether the item indicated by the index is good or bad. According to such a configuration, it is possible to easily determine, for each index, whether the item indicated by the index is good or bad.

[0013] (8) A classification method as one aspect of the present disclosure is a classification method using a classification device that classifies a swing of a club by a subject based on a plurality of indexes, the classification method including: acquiring skeletal position information indicating positions of certain skeletal structures in the body of the subject during the swing; acquiring detection results of a noteworthy event during the swing; calculating the plurality of indexes based on the skeletal position information in the noteworthy event; and classifying the swing of the subject based on the plurality of indexes, wherein the plurality of indexes have a hierarchical relationship, and the swing is classified by sequentially evaluating the calculated indexes in accordance with the hierarchical relationship of the plurality of indexes. According to this configuration, a hierarchical relationship is set for the multiple indices, and the calculated indices are evaluated in order according to the hierarchical relationship, so that the higher the hierarchical level of the index, the greater the influence on the swing classification. Therefore, according to this embodiment, swing classification using multiple indices and taking into account the priority of each index can be performed more easily without relying on an instructor or the like.

[0014] (9) A program as one aspect of the present disclosure causes a computer to operate as any one of the classification devices (1) to (7). Effect of the Invention

[0015] According to the classification device, classification method, and program of the present disclosure, it is possible to more easily classify swings using multiple indexes and taking into account the priority of each specification. [Brief description of the drawings]

[0016] [Figure 1] FIG. 1 is a diagram illustrating a configuration example of an information processing system according to an embodiment of the present disclosure. [Figure 2A] FIG. 13 is a diagram showing a state in which a subject performing a swing is imaged from the front. [Figure 2B] 13 is a diagram showing a state in which a subject taking a swing is imaged from behind in the direction of the ball flight. [Diagram 3] FIG. 2 is a diagram illustrating an example of the configuration of the classification device illustrated in FIG. [Figure 4]FIG. 13 is a diagram illustrating an example of a hierarchical relationship between a plurality of indices. [Diagram 5] 4 is a flowchart showing an example of the operation of the classification device shown in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0018] 1 is a diagram showing an example of the configuration of an information processing system 1 according to an embodiment of the present disclosure. The information processing system 1 according to the present embodiment classifies a club swing by a subject based on a plurality of indices.

[0019] As shown in FIG. 1, the information processing system 1 includes an imaging device 2, a user terminal 3, an image analysis device 4, and a classification device .

[0020] As shown in Figures 2A and 2B, the imaging device 2 captures an image of a subject 5 swinging a club 6. The moving image captured by the imaging device 2 is, for example, a moving image captured from the front of the subject 5 making a swing as shown in Figure 2A, or a moving image captured from behind the subject 5 making a swing in the direction of the ball flight as shown in Figure 2B. The imaging device 2 transmits the captured moving image of the subject 5 to the user terminal 3 via wired communication or wireless communication.

[0021] The user terminal 3 is a terminal used by a user (for example, a subject 5 or an instructor), such as a smartphone, a tablet terminal, etc. The user terminal 3 includes a communication unit 31 and an output unit 32.

[0022] The communication unit 31 includes one or more communication modules. The communication unit 31 may include a communication module compatible with a mobile communication standard such as 4G (4th Generation) or 5G (5th Generation). The communication unit 31 may include a communication module compatible with a wireless LAN standard (IEEE802.11 as an example). The communication unit 31 may also include a communication module compatible with a wired LAN standard. In this embodiment, the communication unit 31 includes a communication module for performing wired and / or wireless communication with the imaging device 2, the image analysis device 4, and the classification device 10. The communication unit 31 receives a moving image transmitted from the imaging device 2, and transmits the received moving image to the image analysis device 4. The communication unit 31 also receives a classification result of the swing by the subject 5 from the classification device 10.

[0023] The output unit 32 outputs the classification result of the swing by the subject person 5 received by the communication unit 31. The output unit 32 outputs the classification result, for example, by displaying it on a display. The output unit 32 also outputs the classification result, for example, by audio output from a speaker.

[0024] The image analyzing device 4 is, for example, a server device provided on a network. The image analyzing device 4 performs image analysis of a moving image transmitted from the user terminal 3. As shown in FIG. 1, the image analyzing device 4 includes a skeleton position analyzing unit 41 and an event analyzing unit 42.

[0025] The skeletal position analysis unit 41 performs image analysis of a moving image of the subject 5 performing a swing, and analyzes the position of a predetermined skeletal structure of the subject 5's body. The skeletal position analysis unit 41 detects the positions of the subject's ankles, knees, hip joints, shoulders, elbows, wrists, head, and the like. The skeletal position analysis unit 41 analyzes the position of a predetermined skeletal structure of the subject 5's body, for example, by image analysis using AI (Artificial Intelligence). The skeletal position analysis unit 41 outputs skeletal position information indicating the detected positions of the predetermined skeletal structure of the subject 5's body during a swing to the classification device 10.

[0026] The skeletal position information is, for example, information indicating the position of a specific skeletal part of the body of the subject 5 projected onto a two-dimensional plane. When the video of the subject 5 is a video of the subject 5 taking a swing from the front as shown in FIG. 2A, the two-dimensional plane is a plane perpendicular to the ground and parallel to the flight line of the ball. When the video of the subject 5 is a video of the subject 5 taking a swing from the front as shown in FIG. 2B, the two-dimensional plane is a plane perpendicular to the flight line of the ball. In other words, the two-dimensional plane onto which the skeletal position information is projected is a plane perpendicular to the ground and parallel to the flight line of the ball, or a plane perpendicular to the flight line of the ball.

[0027] In this embodiment, an example will be described in which the position of a predetermined skeleton of the body of the subject 5 is analyzed from a moving image captured by the imaging device 2, but the present disclosure is not limited to this. For example, the subject 5 performing a swing may be captured by a depth camera capable of measuring the distance (depth) to an object, and the skeleton position analysis unit 41 may analyze the position of a predetermined skeleton of the body of the subject 5 from the image captured by the depth camera.

[0028] Also, for example, subject 5 may be photographed swinging with markers attached to specific skeletal positions on the subject's 5 body, and the skeletal position analysis unit 41 may analyze the positions of specific skeletal positions on the subject's 5 body by detecting the positions of the markers from the photographed image.

[0029] The event analysis unit 42 detects a predetermined event during a swing (for example, address, top, impact, follow-through, and finish, etc.) and outputs the detection result to the classification device 10. The event analysis unit 42 detects an event based on, for example, the similarity between a predetermined skeletal position of the body of the subject 5 detected by the skeletal position analysis unit 41 and a predetermined skeletal position in a predetermined event. For example, when detecting an impact as a predetermined event, a sensor that detects a collision with a ball may be attached to the club, and the event analysis unit 42 may detect the impact based on the detection result of the sensor. For example, the event analysis unit 42 may receive an input of the timing of a predetermined event in a video image of the subject 5 during a swing, and detect the event based on the input. In this embodiment, the swing of the subject 5 is classified according to the posture of the subject 5 during a predetermined event (for example, impact) during the swing. Hereinafter, an event when classifying a swing is referred to as a target event.

[0030] The classification device 10 calculates a plurality of indices for classifying the swing of the subject 5 based on the skeleton position information output from the skeleton position analysis unit 41 in the event of interest detected by the event analysis unit 42. The classification device 10 classifies the swing of the subject 5 based on the calculated plurality of indices, and outputs the classification result to the user terminal 3.

[0031] FIG. 3 is a diagram illustrating an example of the configuration of the classification device 10. As shown in FIG.

[0032] As shown in FIG. 3, the classification device 10 according to this embodiment includes a communication unit 11, a storage unit 12, and a control unit 13.

[0033] The communication unit 11 is configured to include one or more communication modules. The communication unit 11 may include a communication module compatible with mobile communication standards such as 4G and 5G. The communication unit 11 may include a communication module compatible with a wireless LAN standard (IEEE802.11 as an example). The communication unit 11 may also include a communication module compatible with a wired LAN standard. In this embodiment, the communication unit 11 includes a communication module for performing wired communication and / or wireless communication with the user terminal 3 and the image analyzing device 4.

[0034] The storage unit 12 is one or more memories. The memory may be, for example, a semiconductor memory, a magnetic memory, or an optical memory, but is not limited to these, and may be any memory. The storage unit 12 is, for example, built into the classification device 10, but may also be configured to be accessed from outside the classification device 10 via any interface.

[0035] The storage unit 12 stores various data used in various processes executed by the control unit 13. The storage unit 12 may also store results and intermediate data of various processes executed by the control unit 13.

[0036] The control unit 13 is one or more processors. The processor may be, for example, a general-purpose processor or a dedicated processor specialized for a specific process, but is not limited to these and may be any processor. The control unit 13 controls the overall operation of the classification device 10.

[0037] The classification device 10 may have the following software configuration. One or more programs according to the present disclosure used to control the operation of the classification device 10 are stored in the storage unit 12. When the programs stored in the storage unit 12 are loaded by the processor of the control unit 13, they cause the control unit 13 to function as a skeleton position information acquisition unit 131, an event acquisition unit 132, an index calculation unit 133, and a swing classification unit 134.

[0038] The bone structure position information acquisition section 131 acquires bone structure position information indicating the positions of predetermined bone structures of the body of the subject person 5 during a swing, the information being output from the image analysis device 4 (bone structure position analysis section 41).

[0039] The event acquisition unit 132 acquires the detection result of a predetermined event during a swing output from the image analysis device 4 (event analysis unit 42). From the event detection result, the timing of a target event can be identified.

[0040] The index calculation unit 133 calculates a plurality of indexes for classifying swings based on the skeletal position information in a target event (for example, impact). The indexes calculated by the index calculation unit 133 include, but are not limited to, the lower leg angle (the angle of the part from the knee to the ankle), the waist position, and the trunk angle (the inclination angle of the line segment connecting the midpoints of both shoulders and the midpoints of both hip joints). The index calculation unit 133 calculates various indexes that can be calculated based on the skeletal position information and are useful for classifying swings.

[0041] As described above, the skeletal position information may be information indicating the position of a specific skeletal part of the subject's 5 body projected onto a two-dimensional plane perpendicular to the ground and parallel to the flight line of the ball. In this case, the skeletal position information indicates the position of a specific skeletal part when the subject's 5 body is viewed from the front while swinging. By using such skeletal position information, an index regarding the left and right inclination of the subject's 5 body when the subject's 5 body is viewed from the front while swinging can be obtained. As a result, the influence of the left and right inclination of the subject's 5 body can be taken into consideration in classifying the swing.

[0042] As described above, the skeletal position information may be information indicating the position of a specific skeletal structure of the subject's 5 body projected onto a two-dimensional plane perpendicular to the flight line of the ball. In this case, the skeletal position information indicates the position of a specific skeletal structure when the subject's 5 body performing the swing is viewed from the side. By using such skeletal position information, an index regarding the forward and backward inclination of the subject's 5 body can be obtained. As a result, the influence of the forward and backward inclination of the subject's 5 body can be taken into consideration in classifying the swing.

[0043] The swing classification section 134 classifies the swing of the subject person 5 based on the multiple indexes calculated by the index calculation section 133, and outputs the classification result to the user terminal 3.

[0044] In this embodiment, the multiple indexes have a hierarchical relationship as shown in Fig. 3. The swing classification unit 134 classifies the swings by sequentially evaluating the calculated indexes in accordance with the hierarchical relationship of the multiple indexes.

[0045] In the example shown in FIG. 4, the swing classification unit 134 first classifies the index A set in the highest hierarchy and the threshold T A In other words, the swing classification unit 134 compares the calculated index with a threshold value corresponding to the index to determine whether the item indicated by the index is good or bad. In this way, it is possible to easily determine whether the item indicated by the index is good or bad for each index. When it is determined that the item indicated by index A is good, the swing classification unit 134 compares the calculated index with a threshold value corresponding to the index, and determines whether the item indicated by the index is good or bad. B By comparing with this, it is determined whether the item indicated by index B is good or bad.

[0046] When it is determined that the item indicated by the index B is good, the swing classification unit 134 classifies the index C set in a lower hierarchy than the index B and the threshold value T C and judges whether the item indicated by index C is good or bad. If it is judged that the item indicated by index C is good, swing classification unit 134 outputs classification a as the swing classification result. On the other hand, if it is judged that the item indicated by index C is bad, swing classification unit 134 outputs classification b as the swing classification result.

[0047] When it is determined that the item indicated by the index B is poor, the swing classification unit 134 classifies the index D set in a lower hierarchy than the index B and the threshold value T D and judges whether the item indicated by index D is good or bad. If it is judged that the item indicated by index D is good, swing classification unit 134 outputs classification c as the swing classification result. On the other hand, if it is judged that the item indicated by index D is bad, swing classification unit 134 outputs classification d as the swing classification result.

[0048] By providing a hierarchical relationship between the multiple indices and sequentially evaluating the calculated indices according to the hierarchical relationship, the higher the hierarchical level of the index, the greater the influence on the swing classification. Therefore, according to this embodiment, swing classification using multiple indices and taking into account the priority of each index can be performed more easily without relying on an instructor or the like.

[0049] The multiple indices may include an index (first index) calculated based on a predetermined skeletal position of the upper body of the subject 5, and an index (second index) calculated based on a predetermined skeletal position of the lower body of the subject 5. In this way, swings can be classified based on the posture of the upper body and the posture of the lower body of the subject 5.

[0050] The index (first index) calculated based on a predetermined skeletal position of the upper body of the subject 5 may include a trunk angle of the subject 5. The index (second index) calculated based on a predetermined skeletal position of the lower body of the subject 5 may include a lower leg angle of the subject. These indexes are important indexes for classifying swings. By performing classification including these indexes, more accurate classification of swings is possible.

[0051] Generally, it is said that lower body-centered movement is important for efficient swinging. Therefore, it is desirable that the index (second index) calculated based on the position of a predetermined skeleton of the lower body of the subject 5 is set in a higher hierarchy than the index (first index) calculated based on the position of a predetermined skeleton of the upper body of the subject 5. In this way, it is possible to classify swings by giving priority to lower body movements, which are more important for improving swings.

[0052] The multiple indices may include an index calculated based on a change in skeletal position information from an event (e.g., address) that occurs prior to the event of interest (e.g., impact) in time to the event of interest. By including such an index, it becomes possible to classify swings more accurately, taking into account the temporal fluctuation of a specific part of the subject's 5 body.

[0053] The hierarchical relationship of the multiple indicators may be set in advance, or may be set by an evaluator such as an instructor. By allowing the evaluator to set the hierarchical relationship of the multiple indicators, it becomes possible to prioritize the multiple indicators according to the evaluator's knowledge, etc.

[0054] Furthermore, when the skeletal position information is information that two-dimensionally indicates the position of a specific skeleton of the subject 5, a hierarchical relationship of multiple indices may be set according to differences in the viewpoint (imaging direction of the imaging device 2) from which the subject 5 making a swing is viewed. Differences in the viewpoint from which the subject 5 making a swing is viewed change the obtained skeletal position information, and the indices calculated from the skeletal position information also change. By setting a hierarchical relationship of multiple indices according to the viewpoint from which the subject 5 making a swing is viewed, it is possible to classify the swings of the subject 5 from various viewpoints.

[0055] Next, a description will be given of the operation of the classification device 10 according to this embodiment. Fig. 5 is a flowchart showing an example of the operation of the classification device 10 according to this embodiment, and is a diagram for explaining a classification method by the classification device 10 according to this embodiment.

[0056] The skeletal position information acquisition unit 131 acquires skeletal position information indicating the positions of predetermined skeletal parts of the body of the subject 5 during a swing from the image analysis device 4 (skeletal position analysis unit 41) (step S11). In addition, the event acquisition unit 132 acquires the detection result of a predetermined event during a swing from the image analysis device 4 (event analysis unit 42) (step S12).

[0057] The index calculation unit 133 calculates a plurality of indexes based on the skeleton position information in the detected event of interest (step S13).

[0058] The swing classification unit 134 classifies the swing of the subject 5 based on the multiple indices calculated by the index calculation unit 133 (step S14). Here, as described with reference to Fig. 4, the multiple indices have a hierarchical relationship. The swing classification unit 134 classifies the swing by sequentially evaluating the calculated indices according to the hierarchical relationship of the multiple indices.

[0059] As described above, the classification device 10 according to this embodiment includes a skeletal position information acquisition unit 131, an event acquisition unit 132, an index calculation unit 133, and a swing classification unit 134. The skeletal position information acquisition unit 131 acquires skeletal position information indicating the position of a specific skeletal part of the body of the subject 5 during a swing. The event acquisition unit 132 acquires a detection result of a specific event during a swing. The index calculation unit 133 calculates a plurality of indexes based on the skeletal position information in the detected event of interest. The swing classification unit 134 classifies the swing of the subject 5 based on the plurality of indexes. Here, the plurality of indexes have a hierarchical relationship. The swing classification unit 134 classifies the swing by sequentially evaluating the calculated indexes according to the hierarchical relationship of the plurality of indexes.

[0060] By providing a hierarchical relationship between the multiple indices and sequentially evaluating the calculated indices according to the hierarchical relationship, the higher the hierarchical level of the index, the greater the influence on the swing classification. Therefore, according to this embodiment, swing classification using multiple indices and taking into account the priority of each index can be performed more easily without relying on an instructor or the like.

[0061] In the present embodiment, the classification device 10 is described using an example in which the skeleton position information acquisition unit 131, the event acquisition unit 132, the index calculation unit 133, and the swing classification unit 134 are provided, but the present disclosure is not limited to this. A part or all of the skeleton position information acquisition unit 131, the event acquisition unit 132, the index calculation unit 133, and the swing classification unit 134 may be provided in the user terminal 3 or the image analysis device 4. Therefore, the user terminal 3 or the image analysis device 4 may operate as the classification device 10 according to the present disclosure. In addition, a part or all of the skeleton position analysis unit 41 and the event analysis unit 42 may be provided in the user terminal 3. In addition, the user terminal 3 may have a function of capturing an image of the subject 5 performing a swing.

[0062] Although not specifically mentioned in the embodiment, a program for causing a computer to operate as the classification device 10 may be provided. The program may be recorded in a computer-readable medium. The computer-readable medium can be used to install the program in a computer. Here, the computer-readable medium on which the program is recorded may be a non-transient recording medium. The non-transient recording medium is not particularly limited, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM.

[0063] Although the above-mentioned embodiment has been described as a representative example, it is clear to those skilled in the art that many modifications and substitutions are possible within the spirit and scope of the present invention. Therefore, the present invention should not be interpreted as being limited by the above-mentioned embodiment, and various modifications and changes are possible without departing from the scope of the claims. For example, it is possible to combine multiple building blocks shown in the configuration diagram of the embodiment into one, or to divide one building block. [Explanation of symbols]

[0064] 1. Information Processing Systems 2. Imaging device 3. User terminal 4. Image analysis equipment 5. Target Audience 6. Clubs 10 Classifier 11 Communications Department 12 Storage section 13 Control section 31 Communications Department 32 Output section 41 Skeletal Position Analysis Department 42 Event Analysis Section 131 Skeleton position information acquisition unit 132 Event Acquisition Department 133 Indicator calculation section 134 Swing Classification Section

Claims

1. A classification device that classifies a club swing by a subject based on a plurality of indices, a skeletal position information acquiring unit that acquires skeletal position information indicating a position of a predetermined skeletal part of the subject's body during a swing; An event acquisition unit that acquires a detection result of a predetermined event during the swing; an index calculation unit that calculates the plurality of indexes based on the skeleton position information in the detected event of interest; a swing classification unit that classifies the swing of the subject based on the plurality of indexes, The plurality of indices have a hierarchical relationship; The swing classification unit classifies the swing by sequentially evaluating the calculated indexes in accordance with a hierarchical relationship of the multiple indexes.

2. 2. The classification device according to claim 1, The plurality of indices include at least a first index calculated based on a predetermined skeletal position of the subject's upper body, and a second index calculated based on a predetermined skeletal position of the subject's lower body.

3. 3. The classification device according to claim 2, A classification device, wherein the second index is provided at a higher level in a hierarchy than the first index.

4. 3. The classification device according to claim 2, the first indicator includes a trunk angle of the subject; A classification device, wherein the second indicator includes a lower leg angle of the subject.

5. 2. The classification device according to claim 1, A classification device, wherein the multiple indices include an index calculated based on a change in the skeleton position information from an event temporally preceding the event of interest to the event of interest.

6. 2. The classification device according to claim 1, The skeletal position information is information indicating positions of predetermined skeletal structures of the subject's body projected onto a two-dimensional plane, The two-dimensional plane is a plane perpendicular to the ground and parallel to the flight path of the ball, or a plane perpendicular to the flight path of the ball.

7. 2. The classification device according to claim 1, The swing classification unit compares the calculated index with a threshold value corresponding to the index to determine whether the item indicated by the index is good or bad.

8. A classification method using a classification device that classifies a club swing by a subject based on a plurality of indices, acquiring skeletal position information indicating positions of predetermined skeletal members of the subject's body during a swing; Obtaining a detection result of a noteworthy event during the swing; calculating the plurality of indices based on the skeleton position information in the event of interest; and classifying the swing of the subject based on the plurality of indices; The plurality of indices have a hierarchical relationship; The method classifies the swings by sequentially evaluating the calculated indices according to a hierarchical relationship of the multiple indices.

9. A program for causing a computer to operate as the classification device according to claim 1.

Images