Motion analysis system, motion analysis method, and program

Abstract

To provide a motion analysis system, a motion analysis method and a program capable of accurately evaluating a posture of a person performing a motion.SOLUTION: The frame image extraction section 64 extracts a first frame image representing a first scene and a second frame image representing a second scene. An object position specification part 66 specifies a position where the first object is shown in the first frame image and a position where the second object is shown in the second frame image. The evaluation process execution unit 70 evaluates the posture of the person in the second scene based on the position of the first object in the first frame image and the position of the second object in the second frame image.SELECTED DRAWING: Figure 7

Description

[Technical Field]

[0001] The present invention relates to a motion analysis system, a motion analysis method, and a program. [Background technology]

[0002] Patent Document 1 describes a golf swing analysis system that displays images showing the direction of movement of the club head at impact, the orientation of the club face at impact, the direction in which the ball will fly out and curve, and so on. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-104413 Summary of the Invention [Problem to be solved by the invention]

[0004] The motion of a golf swing includes multiple scenes, such as address, take-back, top, and impact. When evaluating the posture of a player performing a golf swing, a more accurate evaluation can be performed by evaluating multiple scenes, such as whether the player's forward lean angle and knee angle at the address are maintained at the top, and whether the club head position at the top is within a V-zone that is set based on the player's posture at the address. However, the technology described in Patent Document 1 was unable to perform such an evaluation.

[0005] In addition, accurate evaluation can be made not only in situations where the posture of a player performing a golf swing is being evaluated, but also in general situations where the posture of a person performing some kind of motion is being evaluated by performing an evaluation involving multiple scenes.

[0006] The present invention has been made in consideration of the above-mentioned problems, and one of its purposes is to provide a motion analysis system, a motion analysis method, and a program that can accurately evaluate the posture of a person performing a motion. [Means for solving the problem]

[0007] (1) The motion analysis system of the present invention includes a first frame image extraction means for extracting a first frame image representing a first scene from a video of a person performing a given motion; a second frame image extraction means for extracting a second frame image representing a second scene that occurs after the first scene from the video; a first position identification means for identifying a position in the first frame image of a first object whose position is determined by the posture of the person in the first scene; a second position identification means for identifying a position in the second frame image of a second object whose position is determined by the posture of the person in the second scene; and an evaluation means for evaluating the posture of the person in the second scene based on the position of the first object whose position is determined by the posture of the person in the first frame image and the position of the second object whose position is determined by the posture of the person in the second scene.

[0008] (2) The motion analysis system described in (1) above may further include an identification display image generation means for generating an identification display image, which is the second frame image in which the position in the second frame image corresponding to the position where the first object is shown in the first frame image and the position where the second object is shown in the second frame image are identified.

[0009] (3) The motion analysis system may be as described in (1) or (2) above, wherein the first object and the second object are the same body part included in the person.

[0010] (4) The motion analysis system may be as described in (1) or (2) above, wherein the first object and the second object are different objects.

[0011] (5) The motion analysis system according to (4) above may be such that the second object is a body part included in the person.

[0012] (6) The motion analysis system according to (4) above may be such that the second object is at least a part of a golf club used by the person.

[0013] (7) The motion analysis method of the present invention includes the steps of: extracting a first frame image representing a first scene from a video of a person performing a given motion; extracting a second frame image representing a second scene that occurs after the first scene from the video; identifying a position in the first frame image at which a first object whose position is determined by the person's posture in the first scene is shown; identifying a position in the second frame image at which a second object whose position is determined by the person's posture in the second scene is shown; and evaluating the person's posture in the second scene based on the position at which the first object is shown in the first frame image and the position at which the second object is shown in the second frame image.

[0014] (8) The program of the present invention is a program for causing a computer to execute the following steps: extracting a first frame image representing a first scene from a video of a person performing a given motion; extracting a second frame image representing a second scene that occurs after the first scene from the video; identifying a position in the first frame image where a first object whose position is determined by the person's posture in the first scene is shown; identifying a position in the second frame image where a second object whose position is determined by the person's posture in the second scene is shown; and evaluating the person's posture in the second scene based on the position where the first object is shown in the first frame image and the position where the second object is shown in the second frame image. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a diagram illustrating an example of the configuration of a motion analysis system according to an embodiment of the present invention. [Figure 2] FIG. 10 is a diagram illustrating an example of an analysis result screen. [Figure 3] FIG. 10 is a diagram illustrating an example of an analysis result screen. [Figure 4] FIG. 10 is a diagram illustrating an example of an analysis result screen. [Figure 5] FIG. 10 is a diagram illustrating an example of an analysis result screen. [Figure 6] FIG. 10 is a diagram illustrating an example of an analysis result screen. [Figure 7] FIG. 2 is a functional block diagram showing an example of functions of a server according to an embodiment of the present invention. [Figure 8] FIG. 10 is a diagram illustrating an example of evaluation management data. [Figure 9] FIG. 1 is a diagram illustrating an example of a skeleton. [Figure 10] FIG. 10 is a flowchart showing an example of the flow of processing performed by the server according to one embodiment of the present invention. [Figure 11]FIG. 10 is a flowchart showing an example of the flow of processing performed by the server according to one embodiment of the present invention. [Figure 12] FIG. 10 is a flowchart showing an example of the flow of processing performed by the server according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0017] FIG. 1 is a diagram showing an example of the configuration of a motion analysis system 1 according to an embodiment of the present invention.

[0018] The motion analysis system 1 according to this embodiment includes a server 10 and a terminal 12. The server 10 and the terminal 12 are connected to a computer network 14 such as the Internet, and the server 10 and the terminal 12 can communicate with each other via the computer network 14.

[0019] The server 10 is, for example, a general-purpose computer equipped with a processor, a storage unit (memory, hard disk drive (HDD), etc.), a communication unit (communication interface, etc.), and the like.

[0020] The terminal 12 is a computer such as a mobile phone (including a smartphone), a personal digital assistant (including a tablet computer), etc. The terminal 12 includes, for example, a processor, a storage unit (memory, solid state drive (SSD), etc.), a communication unit (communication interface, etc.), a sensor unit (camera, touch panel, GPS module, acceleration sensor, motion sensor, etc.), etc.

[0021] In this embodiment, for example, a user operating the terminal 12 takes a picture of a player making a golf swing at a golf course or driving range using a camera provided in the terminal 12. Then, a moving image showing the player making the swing is generated.

[0022] In this embodiment, before filming a swing, the position and tilt of the terminal 12 are adjusted so that the position of the player relative to the camera position is within a predetermined tolerance range and the orientation of the player relative to the filming direction is within a predetermined tolerance range. Then, in this embodiment, after the terminal 12 has been adjusted, when the user performs a predetermined operation, the camera films the player and a moving image is generated showing the player making the swing. In this embodiment, it is assumed that the number of vertical and horizontal pixels of all frame images included in the generated moving image is the same.

[0023] The moving images thus generated are then uploaded from the terminal 12 to the server 10. Hereinafter, the moving images thus uploaded will be referred to as captured moving images.

[0024] Then, the server 10 executes a motion analysis of a person (here, for example, a player swinging) appearing in the captured video. Then, an analysis result screen 20 (see FIGS. 2 to 6) showing the results of the motion analysis is transmitted from the server 10 to the terminal 12. Then, the analysis result screen 20 is displayed on the touch panel (display) of the terminal 12.

[0025] As shown in FIG. 2, an identification display image 22, a scene icon 24, check item icons 26, and an evaluation icon 28 are arranged on the analysis result screen 20.

[0026] In this embodiment, for example, a plurality of scenes (phases) are defined in advance. Here, for example, nine scenes (address, take-back, halfway back, top, halfway down, pre-impact, impact, halfway follow-through, and finish) are defined in advance. Scene icons 24 (24a to 24i) corresponding to each of the plurality of scenes are arranged on the analysis result screen 20.

[0027] Furthermore, at least one check item is associated with each of these scenes. For example, six check items (forward lean angle, right knee angle, head position, right elbow direction, hand position, and shaft lie angle) are associated with the address scene. For example, five check items (forward lean angle, elbow and hand position, right forearm angle, right knee angle, and club head position) are associated with the top scene. In this way, the number of check items associated with each scene may differ.

[0028] In this embodiment, one of the multiple scenes is determined as a scene of interest. Then, for each of at least one check item associated with the scene of interest, a check item icon 26 corresponding to the check item is arranged on the analysis result screen 20.

[0029] 2 shows the analysis result screen 20 in a situation where the scene of interest is an address. In this situation, the scene icon 24 corresponding to the address (scene icon 24a) is displayed in a manner different from the other scene icons 24 (for example, with a different color or shape).

[0030] In this situation, check item icons 26 (26aa to 26af) corresponding to the six check items associated with the address scene are arranged on the analysis result screen 20.

[0031] Additionally, to the left of each check item icon 26, an evaluation icon 28 is placed, which indicates whether or not the check item corresponding to that check item icon 26 satisfies the evaluation criteria.

[0032] In the example of Fig. 2, evaluation icons 28aa to 28af corresponding to check item icons 26aa to 26af, respectively, are arranged on the analysis result screen 20. Each of evaluation icon 28aa, evaluation icon 28ab, evaluation icon 28ac, and evaluation icon 28ad is displayed with a check mark indicating that the evaluation criteria are met. Each of evaluation icon 28ae and evaluation icon 28af is displayed with a cross mark indicating that the evaluation criteria are not met. That is, in the example of Fig. 2, in the address scene, the forward lean angle, right knee angle, head position, and right elbow direction meet the evaluation criteria, but the hand position and shaft lie angle do not meet the evaluation criteria.

[0033] In this embodiment, one of at least one check item associated with the scene of interest is determined as the check item of interest, and an identification display image 22 generated based on the combination of the scene of interest and the check item of interest is displayed on the analysis result screen 20.

[0034] Fig. 2 shows an analysis result screen 20 when the scene of interest is the address and the check item of interest is the forward lean angle. In this case, as shown in Fig. 2, an identification display image 22 is displayed on the analysis result screen 20, in which a dot representing the position of the neck 32a of the player 30 in the address scene, a solid line segment representing the position of the back 34a, a dot representing the position of the waist 36a, a solid line segment representing the position of the thigh 38a, and a dot representing the position of the right knee 40a are superimposed on a frame image representing the address scene.

[0035] In addition, in this embodiment, when the user performs an operation to select a scene, the selected scene is determined as a scene of interest. That is, the scene of interest is changed to the selected scene. Then, one of at least one check item associated with the changed scene of interest is determined as a check item of interest. Here, for example, a predetermined check item may be determined as a check item of interest.

[0036] Then, the check item icon 26 placed on the analysis result screen 20 is updated to one corresponding to the check item associated with the changed scene of interest, and the identification display image 22 placed on the analysis result screen 20 is updated to one generated based on the combination of the changed scene of interest and the changed check item of interest.

[0037] For example, when the analysis result screen 20 shown in Figure 2 is displayed and a predetermined operation (e.g., a tapping operation) is performed on the scene icon 24 (scene icon 24d) corresponding to the top scene, the displayed analysis result screen 20 switches to the analysis result screen 20 shown in Figure 3.

[0038] 3 to 6 show the analysis result screen 20 when the scene of interest is at the top. In the examples of Fig. 3 to 6, check item icons 26 (26da to 26de) corresponding to the five check items associated with the top scene are arranged on the analysis result screen 20.

[0039] In this embodiment, when the user performs an operation to select a check item, the selected check item is determined to be the check item of interest. That is, the check item of interest is changed to the selected check item. Then, the identification display image 22 arranged on the analysis result screen 20 is updated to one generated based on the combination of the current scene of interest and the changed check item of interest.

[0040] For example, when the analysis result screen 20 shown in FIG. 3 is displayed, if a predetermined operation (e.g., a tapping operation) is performed on the check item icon 26 (check item icon 26db) corresponding to the position of the elbow and hand, the displayed analysis result screen 20 switches to the analysis result screen 20 shown in FIG. 4. Also, when a predetermined operation (e.g., a tapping operation) is performed on the check item icon 26 (check item icon 26dd) corresponding to the angle of the right knee, the displayed analysis result screen 20 switches to the analysis result screen 20 shown in FIG. 5. Also, when a predetermined operation (e.g., a tapping operation) is performed on the check item icon 26 (check item icon 26de) corresponding to the position of the club head, the displayed analysis result screen 20 switches to the analysis result screen 20 shown in FIG. 6.

[0041] 3 to 6, evaluation icons 28 (28da to 28de) corresponding to the check item icons 26 (26da to 26de), respectively, are arranged on the analysis result screen 20. A check mark indicating that the evaluation criteria are met is arranged on each of the evaluation icons 28da, 28db, 28dd, and 28de. A cross mark indicating that the evaluation criteria are not met is arranged on the evaluation icon 28dc. That is, in the example of FIGS. 3 to 6, in the top of the swing scene, the forward lean angle, elbow and hand position, right knee angle, and club head position meet the evaluation criteria, but the right forearm angle does not meet the evaluation criteria.

[0042] 3 shows an analysis result screen 20 in which the scene of interest is the top of the swing and the check item of interest is the forward lean angle. In the example of FIG. 3, an identification display image 22 is arranged on the analysis result screen 20, in which figures representing body parts of the player 30 in the address scene and figures representing body parts of the player 30 in the top of the swing scene are superimposed on a frame image representing the top of the swing scene. For example, the identification display image 22 shown in FIG. 3 includes a dot representing the position of the neck 32a of the player 30 in the address scene, a dashed line segment representing the position of the back 34a, and a dot representing the position of the waist 36a. The identification display image 22 shown in FIG. 3 also includes a dot representing the position of the neck 32b of the player 30 in the top of the swing scene, a solid line segment representing the position of the back 34b, and a dot representing the position of the waist 36b.

[0043] 4 shows an analysis result screen 20 in which the scene of interest is the top of the swing and the check item of interest is the position of the elbows and hands. In the example of FIG. 4, a V-zone whose position is identified based on a frame image showing the address scene, and an identification display image 22 in which figures representing body parts of the player 30 in the top of the swing scene are superimposed on the frame image showing the top of the swing scene, are arranged on the analysis result screen 20. For example, the identification display image 22 shown in FIG. 4 has an upper side 42 and a lower side 44 of the V-zone whose position is identified based on the frame image showing the address scene arranged thereon. In addition, the identification display image 22 shown in FIG. 4 has arranged thereon a circle representing the position of the right hand 46b of the player 30, a circle representing the position of the right elbow 48b, a circle representing the position of the left hand 50b, and a circle representing the position of the left elbow 52b.

[0044] Here, the V zone refers to the area surrounded by a lower edge 44 corresponding to the position where the club shaft is shown in the frame image showing the address scene, and an upper edge 42 corresponding to the position of the line connecting the player's neck 32a and the golf ball in the frame image showing the address scene. Note that the definition of the V zone is not limited to this. For example, the upper edge 42 of the V zone may be the line connecting the player's neck 32a and the club head in the frame image showing the address scene.

[0045] 5 shows an analysis result screen 20 in which the scene of interest is the top of the swing and the check item of interest is the angle of the right knee. In the example of FIG. 5, an identification display image 22 in which figures representing body parts of the player 30 in the address scene and figures representing body parts of the player 30 in the top of the swing scene are superimposed on a frame image representing the top of the swing scene is arranged on the analysis result screen 20. For example, the identification display image 22 shown in FIG. 5 includes a dot representing the position of the hips 36a of the player 30 in the address scene, a dashed line segment representing the position of the right thigh 38a, a dot representing the position of the right knee 40a, a dashed line segment representing the position of the right calf 54a, and a dot representing the position of the right ankle 56a. In addition, the identification display image 22 shown in Figure 5 has a dot representing the position of the player 30's waist 36b in the top scene, a solid line segment representing the position of the right thigh 38b, a dot representing the position of the right knee 40b, a solid line segment representing the position of the right calf 54b, and a dot representing the position of the right ankle 56b.

[0046] 6 shows an analysis result screen 20 when the scene of interest is the top of the swing and the check item of interest is the position of the club head. In this case, an identification display image 22 is displayed on the analysis result screen 20, in which the upper side 42 and lower side 44 of the V-zone, the positions of which are identified based on the frame image showing the address scene, and a circle representing the position of the club head 58b in the top of the swing scene, are superimposed on the frame image showing the top of the swing scene.

[0047] The functions of the server 10 according to this embodiment and the processes executed by the server 10 according to this embodiment will be further described below.

[0048] Fig. 7 is a functional block diagram showing an example of functions implemented in the server 10 according to this embodiment. Note that the server 10 according to this embodiment does not need to implement all of the functions shown in Fig. 7, and functions other than the functions shown in Fig. 7 may also be implemented.

[0049] 7 , the server 10 according to this embodiment functionally includes, for example, a rating management data storage unit 60, a captured video image receiving unit 62, a frame image extraction unit 64, an object position identification unit 66, an identification image generation unit 68, an evaluation process execution unit 70, a selected scene data receiving unit 72, a scene determination unit 74, a selected check item data receiving unit 76, a check item determination unit 78, an evaluation result information generation unit 80, and an evaluation result information transmission unit 82. The rating management data storage unit 60 is implemented primarily in the storage unit of the server 10. The captured video image receiving unit 62, the selected scene data receiving unit 72, the selected check item data receiving unit 76, and the evaluation result information transmission unit 82 are implemented primarily in the communication unit of the server 10. The frame image extraction unit 64, the object position identification unit 66, the identification image generation unit 68, the evaluation process execution unit 70, the scene determination unit 74, the check item determination unit 78, and the evaluation result information generation unit 80 are implemented primarily in the processor of the server 10.

[0050] These functions are implemented by executing a program on the server 10, which is a computer. This program may be stored in various computer-readable information storage media such as semiconductor memory and loaded onto the server 10 from the information storage media. Alternatively, the program may be downloaded to the server 10 via a data communication line such as the Internet.

[0051] In this embodiment, the evaluation management data storage unit 60 stores, for example, the evaluation management data shown in FIG.

[0052] As shown in FIG. 8, the evaluation management data includes an ID, scene data, check item data, reference object data, comparison object data, and evaluation reference data.

[0053] The ID included in the review management data is an identifier for the review management data.

[0054] The scene data included in the evaluation management data is data indicating the scene represented by the identification display image 22 generated based on the evaluation management data.

[0055] The check item data included in the evaluation management data is data indicating check items in the evaluation process executed based on the evaluation management data.

[0056] The reference object data included in the evaluation management data is data indicating an object whose position is determined by the posture of a person in a specific scene such as the first scene (in the above example, the address scene), which is identified and displayed in the identification display image 22 generated based on the evaluation management data.

[0057] The comparison object data included in the evaluation management data is data indicating an object whose position is determined by the posture of a person in a scene indicated by the scene data included in the evaluation management data, and which is identified and displayed in an identified display image 22 generated based on the evaluation management data.

[0058] In this manner, in this embodiment, each of the plurality of objects is associated with a check item.

[0059] The evaluation criteria data included in the evaluation management data indicates the evaluation criteria for the evaluation process executed based on the evaluation management data. Whether a check mark or a cross mark is placed on the evaluation icon 28 is determined depending on whether the evaluation criteria are met.

[0060] In this embodiment, the captured video receiving unit 62 receives, from the terminal 12, a captured video of a person performing a given motion (a golf swing in the above example), for example.

[0061] In this embodiment, for example, the frame image extraction unit 64 extracts frame images representing each of a plurality of scenes from the captured video. Here, the frame image extraction unit 64 may identify a frame image representing each of a plurality of predefined scenes from among a plurality of frame images included in the captured video using scene detection technology using a machine learning model such as Transformer. Then, the frame image representing each of the identified scenes may be extracted from among the plurality of frame images included in the captured video. In the above example, frame images representing each of nine scenes (address, take-back, halfway back, top, halfway down, pre-impact, impact, halfway follow-through, and finish) are extracted.

[0062] In this embodiment, for example, the object position identifying unit 66 identifies a position within a frame image representing a certain scene (e.g., a scene of interest) at which an object whose position is determined by the posture of a person in the scene is displayed. Here, if multiple check items are associated with the scene, the object position identifying unit 66 identifies the position of each of multiple objects corresponding to the multiple check items associated with the scene. An example of an object whose position is determined by the posture of a person is a body part of the person. Alternatively, the object position identifying unit 66 may acquire a frame image representing a scene extracted by the frame image extraction unit 64 and determined by the scene determination unit 74 (described later), and identify the position within the frame image at which each of multiple objects corresponding to the multiple check items associated with the scene, which are objects whose positions are determined by the posture of a person in the scene, are displayed.

[0063] For example, the object position identifying unit 66 may identify the position in a frame image where the skeleton of a person appearing in the frame image is shown by using a skeletal analysis technique (Openpose, etc.) that uses a machine learning model, etc. Here, the position in each frame image where the skeleton is shown may be identified by applying the skeletal analysis technique to a moving image, or the position in the frame image where the skeleton is shown may be identified by applying the skeletal analysis technique to a still image (frame image).

[0064] In this way, the position of an object (part) whose position is determined by the posture of the person, which is part of the skeleton whose position in the frame image is identified, may be identified.

[0065] Fig. 9 is a diagram showing an example of a skeleton whose position within a frame image depicting a shot at address is identified by applying a skeletal analysis technique to the frame image. The example in Fig. 9 shows the positions of the neck 32a, back 34a, waist 36a, right thigh 38a, right knee 40a, right calf 54a, right ankle 56a, head 90a, nape 92a, right upper arm 94a, right elbow 48a, right lower arm 96a, right hand 46a, right foot 98a, and right toe 100a in the shot at address. The distinguishable display image 22 shown in Figs. 2, 3, and 5 distinguishably displays parts of the skeleton whose positions within the frame image are shown in Fig. 9. Furthermore, Figs. 4 and 6 distinguishably display the top edge 42 that passes through a part of the skeleton whose position within the frame image is shown in Fig. 9 (e.g., neck 32a). Note that the body parts included in the skeleton shown in Figure 9 are merely examples, and the positions of parts other than these parts may be identified, or the positions of at least some of these parts may not be identified.

[0066] Furthermore, by applying skeletal analysis technology to frame images showing top scenes in a similar manner, the positions of a person's skeleton within the frame images may be identified. The identification display images 22 shown in Figures 3, 4, and 5 identify and display parts that are part of the skeleton whose positions are identified in this manner. For example, the neck 32b, back 34b, waist 36b, right thigh 38b, right knee 40b, right hand 46b, right elbow 48b, left hand 50b, left elbow 52b, right calf 54b, and right ankle 56b correspond to parts of the skeleton whose positions are identified within the frame images showing top scenes.

[0067] Furthermore, the object position identifying unit 66 may apply object recognition technology to identify a position within the frame image where an object held or used by the person moves in accordance with the person's movement. For example, the object position identifying unit 66 may identify a position within the frame image where a golf club is indicated (a position where the club head is indicated and a position where the club shaft is indicated). FIGS. 4 and 6 show a bottom edge 44, which is a line along the club shaft, whose position is identified within the frame image representing the address scene. FIG. 6 also shows a club head 58b, whose position is identified within the frame image representing the top of the swing scene. In this embodiment, for example, the identifying image generating unit 68 generates an identifying image 22, which is a frame image in which the position identified by the object position identifying unit 66 is identified. Here, the identifying image generating unit 68 may generate the identifying image 22 based on evaluation management data. For example, the identification image generating unit 68 may generate an identification image 22 in which the position of the object indicated by the reference object data included in the evaluation management data and the position of the object indicated by the comparison object data included in the evaluation management data are identified within a frame image representing a scene indicated by the scene data included in the evaluation management data. Also, the identification image generating unit 68 may generate a plurality of identification images 22 each associated with a respective one of a plurality of evaluation management data.

[0068] For example, based on evaluation management data with ID 001, an identification display image 22 (see Figure 2) may be generated in which the positions of the neck 32a, back 34a, waist 36a, right thigh 38a, and right knee 40a within a frame image representing an address scene are identified and displayed.

[0069] Furthermore, in this embodiment, the frame image extraction section 64 may extract a first frame image representing a first scene (for example, the address scene) from the captured video, and then extract a second frame image representing a second scene (for example, the opening scene) from the captured video.

[0070] The object position identifying section 66 may then identify the position in the first frame image where a first object, whose position is determined by the posture of the person in the first scene, is shown, and the object position identifying section 66 may then identify the position in the second frame image where a second object, whose position is determined by the posture of the person in the second scene, is shown.

[0071] The identification display image generating unit 68 may then generate an identification display image 22, which is a second frame image in which the position in the second frame image corresponding to the position where the first object is shown in the first frame image and the position where the second object is shown in the second frame image are identified.

[0072] Here, the above-mentioned reference object corresponds to an example of a first object, and the above-mentioned comparison object corresponds to an example of a second object. Here, the first object and the second object may be distinguishably displayed in different ways (for example, with different colors or shapes). For example, as shown in Figures 3 and 5, the first object may be represented by a dashed line, and the second object may be represented by a solid line.

[0073] For example, the identification display image 22 shown in FIG. 3 may be generated based on the evaluation management data with ID 041. Here, for example, positions showing the neck, back, and waist may be identified in the frame image showing the address scene. Then, positions in the frame image showing the top of the swing scene that correspond to these positions in the frame image showing the address scene may be identified (see neck 32a, back 34a, and waist 36a in FIG. 3). Here, for example, positions in the frame image showing the top of the swing scene that have the same coordinates as the positions in the frame image showing the address scene may be identified. Then, positions showing the neck, back, and waist may be identified in the frame image showing the top of the swing scene (see neck 32b, back 34b, and waist 36b in FIG. 3). Then, the identification display image 22 shown in FIG. 3 may be generated, showing these positions.

[0074] Furthermore, the identification display image 22 shown in FIG. 4 may be generated based on the evaluation management data with ID 042. Here, for example, in a frame image representing the address scene, a position where the club shaft is shown may be identified. Then, a position in the frame image representing the top of the swing scene that corresponds to this position in the frame image representing the address scene may be identified (see bottom side 44 in FIG. 4). Also, in the frame image representing the address scene, a position of a line connecting the player's neck 32a and the golf ball may be identified. Then, a position in the frame image representing the top of the swing scene that corresponds to this position in the frame image representing the address scene may be identified (see top side 42 in FIG. 4). Then, in the frame image representing the top of the swing scene, positions where the right hand, right elbow, left hand, and left elbow are shown may be identified (see right hand 46b, right elbow 48b, left hand 50b, and left elbow 52b in FIG. 4). Then, the identification display image 22 shown in FIG. 4 may be generated, indicating these positions.

[0075] Furthermore, the identification display image 22 shown in FIG. 5 may be generated based on the evaluation management data with ID 044. Here, for example, in a frame image representing the address scene, positions showing the hips, right thigh, right knee, right calf, and right ankle may be identified. Then, positions in a frame image representing the top of the swing scene that correspond to these positions in the frame image representing the address scene may be identified (see hips 36a, right thigh 38a, right knee 40a, right calf 54a, and right ankle 56a in FIG. 5). Then, positions showing the hips, right thigh, right knee, right calf, and right ankle may be identified in the frame image representing the top of the swing scene (see hips 36b, right thigh 38b, right knee 40b, right calf 54b, and right ankle 56b in FIG. 5). Then, the identification display image 22 shown in FIG. 5 may be generated, showing these positions.

[0076] Furthermore, the identification display image 22 shown in FIG. 6 may be generated based on the evaluation management data with ID 045. Here, for example, the positions of the top side 42 and bottom side 44 in FIG. 6 may be identified in the same manner as the identification display image 22 shown in FIG. 4. Then, the position where the club head is shown may be identified in the frame image showing the top scene (see club head 58b in FIG. 6). Then, the identification display image 22 shown in FIG. 6 may be generated, showing these positions.

[0077] As shown in FIGS. 3 and 5, the first object and the second object may be the same body part included in a person.

[0078] 4 and 6, the first object and the second object may be different objects. In this case, the second object may be a body part of a person. The second object may also be at least a part of a golf club used by the person.

[0079] In this embodiment, for example, the evaluation process execution unit 70 executes an evaluation process associated with a scene represented by an extracted frame image on the frame image, thereby evaluating the posture of a person appearing in the frame image. In this embodiment, a different evaluation process may be associated with each of a plurality of scenes. Then, for each of the plurality of scenes, the evaluation process associated with that scene may be executed.

[0080] Here, the evaluation processing execution unit 70 may evaluate the posture of a person appearing in a frame image, for example, in accordance with evaluation criteria indicated by evaluation criteria data included in the evaluation management data. Alternatively, the evaluation processing execution unit 70 may evaluate the posture of a person appearing in a frame image based on the position indicated in the frame image representing a certain scene of an object whose position is determined by the posture of the person in the scene.

[0081] In addition, the evaluation processing execution unit 70 may evaluate the posture of a person in a second scene based on the position where a first object is shown in a first frame image representing the first scene and the position where a second object is shown in a second frame image representing the second scene.

[0082] For example, the forward lean angle of a person appearing in a frame image depicting an address scene may be evaluated based on the evaluation management data with ID 001 and the identification display image 22 shown in Fig. 2. For example, if the angle between the back 34a and the right thigh 38a in the identification display image 22 shown in Fig. 2 is 135 degrees or more and 150 degrees or less, it may be determined that the forward lean angle in the address scene meets the evaluation criteria. If not, it may be determined that the forward lean angle in the address scene does not meet the evaluation criteria.

[0083] Furthermore, for example, the forward lean angle of a person appearing in a frame image representing a top shot may be evaluated based on the evaluation management data with ID 041 and the identification display image 22 shown in FIG. 3. For example, if the difference between the angle between the back 34a in the address shot and a given reference line (e.g., a line in the left-right direction) and the angle between the back 34b in the top shot and the reference line (e.g., a line in the left-right direction) in the identification display image 22 shown in FIG. 3 is 10 degrees or less, it may be determined that the forward lean angle of the person appearing in the frame image representing the top shot meets the evaluation criteria. If not, it may be determined that the forward lean angle of the person appearing in the frame image representing the top shot does not meet the evaluation criteria. In this way, even if the check item is the same (e.g., forward lean angle), the evaluation criteria may differ depending on the scene.

[0084] Furthermore, for example, the positions of the elbows and hands of a person appearing in a frame image representing a top scene may be evaluated based on the evaluation management data with ID 042 and the identification display image 22 shown in Fig. 4. For example, if the positions of the right hand 46b, right elbow 48b, left hand 50b, and left elbow 52b in the top scene in the identification display image 22 shown in Fig. 4 are all within the V zone (i.e., within the area sandwiched between the upper side 42 and the lower side 44), it may be determined that the positions of the elbows and hands of the person appearing in the frame image representing a top scene meet the evaluation criteria. Otherwise, it may be determined that the positions of the elbows and hands of the person appearing in the frame image representing a top scene do not meet the evaluation criteria.

[0085] Furthermore, for example, the angle of the right knee of the person appearing in the frame image representing the top scene may be evaluated based on the evaluation management data with ID 044 and the identification display image 22 shown in Fig. 5. For example, if the difference between the angle between the right thigh 38a and the right calf 54a in the address scene and the angle between the right thigh 38b and the right calf 54b in the top scene in the identification display image 22 shown in Fig. 5 is 10 degrees or less, it may be determined that the angle of the right knee of the person appearing in the frame image representing the top scene meets the evaluation criteria. Otherwise, it may be determined that the angle of the right knee of the person appearing in the frame image representing the top scene does not meet the evaluation criteria.

[0086] Furthermore, for example, the position of the club head in the frame image representing the top scene may be evaluated based on the evaluation management data with ID 045 and the identification display image 22 shown in FIG. 6. For example, if the position of the club head 58b in the top scene in the identification display image 22 shown in FIG. 6 is within the V zone (i.e., within the area between the upper side 42 and the lower side 44), it may be determined that the position of the club head in the frame image representing the top scene satisfies the evaluation criteria. If not, it may be determined that the position of the club head in the frame image representing the top scene does not satisfy the evaluation criteria. In this way, the posture of a person in the frame image may be evaluated based on objects other than the person's body parts.

[0087] As described above, the evaluation process may be associated with a combination of a scene and at least one check item associated with the scene. The evaluation process execution unit 70 may then execute, for an extracted frame image, the evaluation process associated with the combination of the scene represented by the frame image and the check item of interest, thereby evaluating the posture of a person appearing in the frame image.

[0088] In this embodiment, for example, the selected scene data receiving unit 72 receives selected scene data indicating a scene selected on the terminal 12 from the terminal 12. For example, when the user performs a predetermined operation such as a tap operation on any of the scene icons 24 arranged on the analysis result screen 20, the terminal 12 transmits selected scene data including an identifier of the scene corresponding to the tapped scene icon 24 to the server 10. Then, the selected scene data receiving unit 72 receives the selected scene data.

[0089] In this embodiment, the scene determination unit 74 determines, for example, one of a plurality of predefined scenes. The scene determined in this manner corresponds to the scene of interest described above. Here, for example, a predetermined scene (e.g., an address scene) may be determined as the scene of interest. Alternatively, a scene indicated by the selected scene data received by the selected scene data receiving unit 72 may be determined as the scene of interest. The scene determination unit 74 may hold an identifier of the scene of interest.

[0090] In this embodiment, for example, the selected check item data receiving unit 76 receives selected check item data indicating a check item selected on the terminal 12 from the terminal 12. For example, when the user performs a predetermined operation such as a tap operation on one of the check item icons 26 arranged on the analysis result screen 20, the terminal 12 transmits selected check item data including the identifier of the check item corresponding to the tapped check item icon 26 to the server 10. Then, the selected check item data receiving unit 76 receives the selected check item data.

[0091] In this embodiment, the check item determination unit 78 determines, for example, one check item from at least one check item associated with the scene of interest. The check item determined in this manner corresponds to the check item of interest described above. Here, for example, a predetermined check item (e.g., forward lean angle) may be determined as the check item of interest. Alternatively, a check item indicated by the selected check item data received by the selected check item data receiving unit 76 may be determined as the check item of interest. The check item determination unit 78 may hold an identifier of the check item of interest.

[0092] In this embodiment, for example, the evaluation result information generating unit 80 generates evaluation result information indicating the evaluation result by the evaluation process executing unit 70. Here, for example, an analysis result screen 20 may be generated as the evaluation result information.

[0093] In this embodiment, the generated evaluation result information is associated with the evaluation management data. For example, on the analysis result screen 20 associated with the evaluation management data, a scene icon 24 corresponding to the scene indicated by the scene data included in the evaluation management data is identifiable and displayed. On the analysis result screen 20 associated with the evaluation management data, a check item icon 26 corresponding to the check item indicated by the check item data included in the evaluation management data is identifiable and displayed. On the analysis result screen 20 associated with the evaluation management data, an identifiable display image 22 generated based on the evaluation management data is disposed. On the analysis result screen 20 associated with the evaluation management data, an evaluation icon 28 indicating the evaluation result (e.g., whether or not the evaluation criteria are met) for at least one check item associated with the scene indicated by the scene data included in the evaluation management data is disposed.

[0094] In this embodiment, for example, the evaluation result information transmitting unit 82 transmits the evaluation result information generated by the evaluation result information generating unit 80 to the terminal 12. Here, for example, the evaluation result information transmitting unit 82 may transmit to the terminal 12 an analysis result screen 20 associated with evaluation management data including scene data indicating a scene of interest and check item data indicating check items of interest. Then, the analysis result screen 20 transmitted to the terminal 12 in this manner may be displayed on a touch panel (display) of the terminal 12.

[0095] In addition, in this embodiment, for example, as described above, the evaluation result information sending unit 82 displays on the display of the terminal 12 an identification display image 22, which is a frame image in which the position of an object corresponding to one of the multiple check items associated with the scene of interest is identified.

[0096] Furthermore, in this embodiment, for example, as described above, in response to receiving a check item change request, the evaluation result information sending unit 82 switches the identification display image 22 displayed on the display of the terminal 12 from a frame image in which the position of an object corresponding to one of the check items is identified to a frame image in which the position of an object corresponding to a check item other than the check item is identified. Here, for example, the reception of the selected check item data described above corresponds to an example of receiving a check item change request.

[0097] For example, suppose a predetermined operation is performed on the check item icon 26db when the analysis result screen 20 shown in Fig. 3 is displayed. Then, the identification display image 22 arranged on the displayed analysis result screen 20 switches from a frame image representing a top scene in which the positions of the objects (neck, back, and waist) corresponding to the check item "forward lean angle" are identified to a frame image representing a top scene in which the positions of the objects (right hand, right elbow, left hand, left elbow, and V-zone) corresponding to the check item "elbow / hand position" are identified (see Fig. 4).

[0098] 3 is displayed, the identification display image 22 arranged on the displayed analysis result screen 20 is switched from a frame image representing a top scene in which the positions of the objects (neck, back, and waist) corresponding to the check item "forward lean angle" are identified to a frame image representing a top scene in which the positions of the objects (waist, right thigh, right knee, right calf, and right ankle) corresponding to the check item "right knee angle" are identified (see FIG. 5).

[0099] 3 is displayed, the identification display image 22 arranged on the displayed analysis result screen 20 changes from a frame image showing a top of the swing scene in which the positions of the objects (neck, back, and waist) corresponding to the check item "forward lean angle" are identified to a frame image showing a top of the swing scene in which the positions of the objects (club head and V zone) corresponding to the check item "head position" are identified (see FIG. 6).

[0100] Here, an example of the flow of processing performed by the server 10 according to this embodiment when a captured video is generated in the terminal 12 will be described with reference to the flow diagram shown in FIG.

[0101] First, when a captured video is transmitted from the terminal 12 to the server 10, the captured video receiving unit 62 receives the captured video (S101).

[0102] Then, the frame image extraction unit 64 executes scene detection on the captured video received in the process shown in S101, and extracts, for each of a plurality of predefined scenes, a frame image representing that scene from among the plurality of frame images included in the captured video (S102). Here, for example, nine frame images are extracted.

[0103] Then, for each of the nine frame images extracted by the process shown in S102, the object position identification unit 66 identifies the position within the frame image of the object (e.g., the above-mentioned reference object and comparison object) associated with that frame image (S103).

[0104] Then, the identification display image generating unit 68 generates an identification display image 22 associated with each of the plurality of review management data stored in the review management data storage unit 60 (S104).

[0105] Then, the review processing execution unit 70 executes a review process associated with each of the plurality of review management data stored in the review management data storage unit 60 (S105).

[0106] Then, the evaluation result information generating unit 80 generates evaluation result information indicating the evaluation results in the evaluation process shown in S105 for each of the multiple evaluation management data stored in the evaluation management data storage unit 60 (S106). Here, for each of the multiple evaluation management data stored in the evaluation management data storage unit 60, an analysis result screen 20 associated with the evaluation management data may be generated.

[0107] Then, the evaluation result information transmitting unit 82 transmits any one of the evaluation result information generated in the process shown in S106 to the terminal 12 (S107), and the process shown in this processing example is terminated. The evaluation result information transmitted in the process shown in S107 is displayed on a display such as a touch panel of the terminal 12. Here, it is assumed that the scene of interest in the initial state is the address, and the check item of interest in the initial state is the forward lean angle. In this case, in the process shown in S107, for example, evaluation result information associated with evaluation management data whose ID is 001 is transmitted.

[0108] Next, an example of the processing flow performed by the server 10 according to this embodiment when a user performs a predetermined operation such as a tap operation on any of the scene icons 24 displayed on the analysis result screen 20 will be described with reference to the flow chart illustrated in FIG. 11.

[0109] When a predetermined operation is performed on a scene icon 24 and selected scene data including an identifier of the scene corresponding to the scene icon 24 is transmitted from the terminal 12 to the server 10, the selected scene data receiving unit 72 receives the selected scene data (S201).

[0110] Then, the scene determination unit 74 changes the scene of interest to the scene indicated by the selected scene data received in the process shown in S201 (S202). Here, the scene identifier held by the scene determination unit 74 may be updated to the scene identifier indicated by the selected scene data received in the process shown in S201.

[0111] Then, the check item determination unit 78 determines the check item of interest to be at least one of the check items associated with the scene indicated by the selected scene data received in the process shown in S201 (S203). Here, for example, a predetermined check item may be determined as the check item of interest. Also, the identifier of the check item held by the check item determination unit 78 may be updated to the identifier of the changed check item.

[0112] Then, the evaluation result information sending unit 82 identifies evaluation result information that corresponds to evaluation management data including scene data indicating the scene of interest and check item data indicating the check item of interest from the evaluation result information generated by the processing shown in S106 (S204).

[0113] Then, the evaluation result information transmitting unit 82 transmits the evaluation result information identified in the process shown in S204 to the terminal 12 (S205), and the process shown in this processing example is terminated. The evaluation result information transmitted in the process shown in S205 is displayed on a display such as a touch panel of the terminal 12. In this way, the evaluation result information displayed on the display of the terminal 12 is switched.

[0114] Next, an example of the processing flow performed by the server 10 according to this embodiment when a user performs a predetermined operation such as a tap operation on any of the check item icons 26 located on the analysis result screen 20 will be described with reference to the flow diagram illustrated in FIG. 12.

[0115] When a predetermined operation is performed on a check item icon 26 and selected check item data including an identifier of the check item corresponding to the check item icon 26 is sent from the terminal 12 to the server 10, the selected check item data receiving unit 76 receives the selected check item data (S301).

[0116] Then, the check item determination unit 78 changes the target check item to the check item indicated by the selected check item data received in the process shown in S301 (S302). Here, the identifier of the check item held by the check item determination unit 78 may be updated to the identifier of the check item indicated by the selected check item data received in the process shown in S301.

[0117] Then, the evaluation result information sending unit 82 identifies evaluation result information that corresponds to evaluation management data including scene data indicating the scene of interest and check item data indicating the check item of interest from the evaluation result information generated by the processing shown in S106 (S303).

[0118] Then, the evaluation result information transmitting unit 82 transmits the evaluation result information identified in the process shown in S303 to the terminal 12 (S304), and the process shown in this processing example is terminated. The evaluation result information transmitted in the process shown in S304 is displayed on a display such as a touch panel of the terminal 12. In this way, the evaluation result information displayed on the display of the terminal 12 is switched.

[0119] In this embodiment, as in the processing example shown in Figure 10, the server 10 does not need to generate evaluation result information associated with each of all evaluation management data stored in the evaluation management data storage unit 60 in response to receiving captured video images.

[0120] For example, in the process shown in S102, a frame image of an initial scene of interest may be extracted. Then, in the process shown in S103, the position of an object related to at least one check item associated with the scene of interest within the frame image may be identified. Then, in the process shown in S104, an identification display image 22 associated with the combination of the scene of interest and the initial check item of interest may be generated. Then, in the process shown in S105, an evaluation process may be performed for each of at least one check item associated with the scene of interest, associated with the combination of the scene of interest and the check item. Then, in the process shown in S106, evaluation result information associated with the scene of interest and the check item of interest may be generated.

[0121] Then, every time the scene of interest is changed, the processes shown in S102 to S106 may be executed for the changed scene of interest. Also, every time the check item of interest is changed, an identification display image 22 associated with the changed check item of interest may be generated.

[0122] In this embodiment, as described above, the identification image 22, which is a frame image in which the position of an object corresponding to one of the multiple check items associated with a scene is identified, is displayed on the display of the terminal 12. Then, in response to reception of a check item change request, the identification image 22 displayed on the display of the terminal 12 is switched from a frame image in which the position of an object corresponding to one of the check items is identified to a frame image in which the position of an object corresponding to a check item other than the check item is identified. In this way, according to this embodiment, it is possible to clearly display the status of each of the multiple check items in a scene included in a motion.

[0123] Furthermore, in this embodiment, it is possible to perform an evaluation of the posture of a person performing a motion that involves multiple scenes included in the motion, such as the address scene and the top scene, etc. Therefore, according to this embodiment, it is possible to perform an accurate evaluation of the posture of a person performing a motion.

[0124] The present invention is not limited to the above-described embodiment.

[0125] For example, in response to receiving a scene change request, the evaluation result information transmitting unit 82 may switch the displayed identification display image 22 from a frame image representing a scene in which the position of an object corresponding to one of the check items associated with that scene is identified to a frame image representing a scene different from the scene in which the position of the object is identified. Here, for example, the reception of the selected scene data described above corresponds to an example of receiving a scene change request.

[0126] 3 is displayed, a predetermined operation such as a tap operation is performed on the scene icon 24c. In this case, an analysis result screen 20 may be generated in which a dot representing the position of the neck 32a of the player 30, a dashed line segment representing the position of the back 34a, and a dot representing the position of the waist 36a are superimposed in the address scene, and an identification display image 22 is arranged in which a dot representing the position of the neck of the player 30, a dashed line segment representing the position of the back, and a dot representing the position of the waist are superimposed in the halfway back scene. Then, the analysis result screen 20 displayed on the display of the terminal 12 may be switched to the analysis result screen 20 generated in this manner.

[0127] Furthermore, for example, information related to the evaluation process described above may be placed on the identification display image 22. For example, a number representing the angle between the back 34a and the right thigh 38a in the identification display image 22 shown in Fig. 2 may be placed on the identification display image 22 shown in Fig. 2. Furthermore, a number representing the difference between the angle between the back 34a and a reference line (e.g., a line in the left-right direction) in the address scene and the angle between the back 34b and a reference line (e.g., a line in the left-right direction) in the top of the swing scene in the identification display image 22 shown in Fig. 3 may be placed on the identification display image 22 shown in Fig. 3. Furthermore, in the identification display image 22 shown in Fig. 5, numbers representing the angle between the right thigh 38a and the right calf 54a in the address scene and the angle between the right thigh 38b and the right calf 54b in the top of the swing scene may be placed on the identification display image 22 shown in Fig. 5. In addition, a number representing the difference between the angle between the right thigh 38a and the right calf 54a in the address scene and the angle between the right thigh 38b and the right calf 54b in the top scene in the identification display image 22 shown in Figure 5 may be placed on the identification display image 22 shown in Figure 5.

[0128] Furthermore, the same analysis result screen 20 as the analysis result screen 20 shown in FIGS. 2 to 6 may be displayed for scenes other than the address scene and the top scene.

[0129] The scope of application of the present invention is not limited to analyzing golf swings, but can also be applied to analyzing swings and motions in other sports, and can also be applied to other general motion analysis such as dancing.

[0130] 7 may be implemented in the terminal 12. In addition, all or part of the processing shown in FIG.

[0131] Furthermore, the specific character strings and numerical values ​​described above and the specific character strings and numerical values ​​in the drawings are examples, and the present invention is not limited to these character strings and numerical values. [Explanation of symbols]

[0132] 1 Motion analysis system, 10 Server, 12 Terminal, 14 Computer network, 20 Analysis result screen, 22 Identification display image, 24, 24a, 24b, 24c, 24d, 24e, 24f, 24g, 24h, 24i Scene icon, 26, 26aa, 26ab, 26ac, 26ad, 26ae, 26af, 26da, 26db, 26dc, 26dd, 26de Check item icon, 28, 28aa, 28ab, 28ac, 28ad, 28ae, 28af, 28da, 28db, 28dc, 28dd, 28de Evaluation icon, 30 Player, 32a, 32b Neck, 34a, 34b Back, 36a, 36b Waist, 38a, 38b Right thigh, 40a, 40b Right knee, 42 Upper side, 44 lower side, 46a, 46b right hand, 48a, 48b right elbow, 50b left hand, 52b left elbow, 54a, 54b right calf, 56a, 56b right ankle, 58b club head, 60 evaluation management data storage unit, 62 captured video image receiving unit, 64 frame image extraction unit, 66 object position identification unit, 68 identification display image generation unit, 70 evaluation processing execution unit, 72 selected scene data receiving unit, 74 scene determination unit, 76 selected check item data receiving unit, 78 check item determination unit, 80 evaluation result information generation unit, 82 evaluation result information transmission unit, 90a head, 92a neck, 94a right upper arm, 96a right lower arm, 98a right foot, 100a right toe.

Claims

1. a first frame image extracting means for extracting a first frame image representing a first scene from a moving image of a person performing a given motion; a second frame image extracting means for extracting a second frame image representing a second scene that occurs after the first scene from the moving image; a first position specifying means for specifying a position in the first frame image where a first object, the position of which is determined by the posture of the person in the first scene, is shown; a second position specifying means for specifying a position in the second frame image where a second object, the position of which is determined by the posture of the person in the second scene, is shown; an evaluation means for evaluating the posture of the person in the second scene based on the position where the first object is indicated in the first frame image and the position where the second object is indicated in the second frame image; Motion analysis system including.

2. an identification display image generating means for generating an identification display image, which is the second frame image in which a position in the second frame image corresponding to a position where the first object is shown in the first frame image and a position where the second object is shown in the second frame image are identified and displayed, The motion analysis system of claim 1 .

3. the first object and the second object are the same body part included in the person; The motion analysis system of claim 1 .

4. the first object and the second object are different objects; The motion analysis system of claim 1 .

5. the second object is a body part included in the person; The motion analysis system of claim 4 .

6. the second object is at least a part of a golf club used by the person; The motion analysis system of claim 4 .

7. extracting a first frame image representing a first scene from a video of a person performing a given motion; extracting a second frame image representing a second scene that occurs after the first scene from the moving image; Identifying a position in the first frame image where a first object whose position is determined by the posture of the person in the first scene is shown; Identifying a position in the second frame image where a second object whose position is determined by the posture of the person in the second scene is shown; Evaluating a posture of the person in the second scene based on a position where the first object is shown in the first frame image and a position where the second object is shown in the second frame image; Motion analysis methods including:

8. extracting a first frame image representing a first scene from a video of a person performing a given motion; extracting a second frame image representing a second scene that occurs after the first scene from the moving image; Identifying a position in the first frame image where a first object whose position is determined by the posture of the person in the first scene is shown; Identifying a position in the second frame image where a second object whose position is determined by the posture of the person in the second scene is shown; Evaluating a posture of the person in the second scene based on a position where the first object is shown in the first frame image and a position where the second object is shown in the second frame image; A program that causes a computer to execute the following.

Images