Image analysis device, image analysis method, and program

Abstract

To reduce a load on a device that executes processing for acquiring, from a video photographing an action, information on the action.SOLUTION: A video analysis device includes a video processing unit that displays a video of a golf swing, which is a video shot at a fixed point, on a display screen 20 of a display device 12, an area setting unit that sets a part of a range shown in the video as a determination area R, and an information acquisition unit that acquires the timing of impact on the basis of a change within the set determination area R. The area setting unit sets the determination area R on the basis of a user's operation of selecting a range surrounding a golf ball B from a still image that constitutes a video and in which the golf ball B is shown.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to a video analysis apparatus, a video analysis method, and a program.

Background Art

[0002] Patent Document 1 discloses a diagnostic apparatus for a swing form that extracts the timing of impact in a swing motion from a video of a golf swing and displays an image. The diagnostic apparatus of Patent Document 1 extracts the timing of impact as follows. First, for each of the consecutive still images constituting the video, using a previously created template of a golf ball, the presence or absence of a golf ball is determined by performing template matching processing of the golf ball template. Then, in the time series of the determination results, the timing when the determination switches from a presence determination to an absence determination is extracted as the timing of impact.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the case of the diagnostic apparatus of Patent Document 1, since it is necessary to repeatedly perform the process of detecting a golf ball by template matching for a large number of still images constituting the video, the load on the device for executing the process increases. In particular, since a video of a golf swing shows the entire body of the golfer, the golf ball shown in the video becomes small. The smaller the golf ball is relative to the size of the video, the larger the amount of calculation for the process of detecting the golf ball for each still image becomes. As a result, the load on the device for executing the process of detecting the golf ball further increases.

Means for Solving the Problems

[0005] The video analysis device for solving the above problems is a video analysis device that acquires information related to the operation based on a video in which the operation of the object is reflected. The video is a moving image in which the process of the stationary object starting to move is shot at a fixed point. The video analysis device includes a video processing unit that displays the video on a display screen of a display device, a region setting unit that sets a part of the range shown in the video as a determination region, and an information acquisition unit that acquires the information based on changes within the set determination region. The region setting unit sets the determination region based on an operation of a user who selects a range surrounding the object from a still image that constitutes the video and in which the object is shown.

[0006] The video is, for example, a video of a golf swing. According to the above configuration, with the determination region specified by the user as the target range, processing for acquiring information related to the operation is executed. Therefore, compared with the case of acquiring information related to the operation with the entire still image, that is, the entire area inside and outside the determination region as the target range, the processing to be executed by the device is lightened. In the case of the above configuration, although processing for setting the determination region is added, since the user selects the determination region, this processing is light processing that does not involve determination by an AI model or the like. Therefore, according to the above configuration, the load on the device for executing the processing for acquiring information related to the operation can be reduced.

[0007] In the above video analysis device, it is preferable that the region setting unit overlays and displays a setting frame for selecting the determination region on the still image displayed on the display screen, and the setting frame is displayed so that its size and position relative to the still image can be changed based on the operation of the user.

[0008] According to the above configuration, the user can easily perform an operation of selecting a range surrounding the object from the still image. In the above-described video analysis device, it is preferable that the information acquisition unit determines whether or not the state in which the object is shown in the determination region is maintained for a plurality of still images constituting the video, and acquires the timing of the start of movement of the object based on the change in the determination result.

[0009] According to the above configuration, the timing of the start of movement of the object can be acquired by a simple determination process. Therefore, the load on the device that executes the process of acquiring information regarding the operation can be further reduced.

[0010] In the above-described video analysis device, it is preferable that the process of the information acquisition unit is executed stand-alone without using a communication network. According to the above configuration, communication costs and processing time can be reduced. Also, the video in which the movement of the object is shown can be personal information. Therefore, it is possible to suppress the user from feeling a sense of aversion regarding transmitting personal information to the outside.

[0011] The video analysis method for solving the above problems is a video analysis method for acquiring information regarding the operation based on a video in which the operation of the object is shown, the video being a moving image in which the process of the stationary object starting to move is shot at a fixed point, the method including a video processing step of displaying the video on a display screen of a display device, a region setting step of setting a part of the range shown in the video as a determination region, and an information acquisition step of acquiring the information based on the change in the set determination region, the region setting step being a step of setting the determination region based on an operation of a user who selects a range surrounding the object from still images constituting the video and in which the object is shown, and in the region setting step, selecting a range surrounding the object such that the area ratio of the object in the determination region is 50% or more.

[0012] According to the above configuration, the determination area is small, so that the effect of reducing the processing load for acquiring information about the motion is more pronounced. Also, according to the above configuration, the largest object present in the determination area is the target object. In this case, the object detected in the determination area can be regarded as the target object. Therefore, when determining whether or not the target object exists in the determination area, it is only necessary to determine whether or not the target object is detected in the determination area, and it is not necessary to determine whether or not the detected object is the target object.

[0013] A program for solving the above problem is a program for causing a computer to function as a video analysis device that acquires information about an object's movement based on a video showing the movement of the object, the video being a video shot at a fixed point showing the process of the object starting to move from a stationary state, and the program causes the computer to execute a video display process that displays the video on a display screen of a display device, an area setting process that sets a portion of the area shown in the video as a judgment area, and an information acquisition process that acquires the information based on changes within the set judgment area, the area setting process being a process of setting the judgment area based on a user's operation to select an area surrounding the object from a still image that constitutes the video and that contains the object. Effect of the Invention

[0014] According to the present invention, it is possible to reduce the load on a device that executes a process for acquiring information about a specific action from a video in which the action is captured. [Brief description of the drawings]

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

[0016] <Video Analysis System> Hereinafter, an embodiment of a video analysis system including a video analysis device will be described. The video analysis device of this embodiment acquires the timing of impact in a golf swing based on a video in which the golf swing is photographed.

[0017] As shown in FIG. 1, the video analysis system 10 includes a photographing device 11, a display device 12, and a video analysis device 13. [Photographing Device] The photographing device 11 is not particularly limited as long as it can acquire a video obtained by photographing a region surrounding an actor performing a golf swing and a golf ball at a fixed point. Examples of the photographing device 11 include cameras mounted on mobile phones such as video cameras, smartphones, and tablet terminals. The photographing device 11 is configured to output the data of the video photographed at a fixed point to the video analysis device 13.

[0018] The video photographed by the photographing device 11 is a moving image composed of continuous still images photographed at a predetermined frame rate. The frame rate of the above moving image is, for example, 30 FPS or 60 FPS. Also, the photographing direction of the above video is not particularly limited as long as it is a direction in which the golf ball is reflected, and it may be from the front side of the actor, from the rear side in the traveling direction of the hit ball, or from the front side in the traveling direction of the hit ball. In the drawings, as an example, the case of photographing from the front side of the actor is illustrated.

[0019] [Display Device] The display device 12 is a display device such as a liquid crystal display or an organic EL display. The display device 12 is preferably a smartphone or a tablet terminal having a display screen.

[0020] As shown in FIGS. 1, 4(a), and 4(b), it includes a display screen 20 for displaying the display content controlled by the video analysis device 13. At the upper part of the display screen 20, a video display screen 21 for displaying a golf swing video is provided. On the video display screen 21, a video display unit 22 for displaying a golf swing video and an operation unit 23 for displaying operation icons for operating the video displayed on the video display unit 22 are provided. The operation icons displayed on the operation unit 23 are a play / stop button 23a, a rewind button 23b, a fast-forward button 23c, and a seek bar 23d. The operation icons are not particularly limited, and those provided in known video playback software can be applied.

[0021] At the lower part of the display screen 20, an information display screen 24 for displaying information based on the video displayed on the video display screen 21 is provided. On the information display screen 24, a timing icon 25 indicating the position of the impact on the seek bar 23d is displayed. Also, the timing icon 25 is a timing designation button for causing the video at the impact time to be displayed on the video display screen 21 based on an operation input such as a click.

[0022] Operations on the operation icons displayed on the operation unit 23 and the timing icon 25 displayed on the information display screen 24 are performed, for example, by an input operation using a pointing device (not shown) or a touch operation on a touch panel. The pointing device is, for example, a keyboard, a touch panel, or a mouse.

[0023] [Video Analysis Device] The video analysis device 13 is, for example, a server, a PC (Personal Computer), a smartphone, a tablet terminal, or a combination thereof. The video analysis device 13 is preferably a device such as a PC (Personal Computer), a smartphone, or a tablet terminal that can execute various processes performed by the device in a stand-alone manner without using a communication network. Further, the video analysis device 13 is preferably a smartphone or a tablet terminal having a function as the imaging device 11 and a function as the display device 12.

[0024] As shown in FIG. 1, the video analysis device 13 includes a transmission / reception unit 30, a video processing unit 31, a region setting unit 32, an information acquisition unit 33, and a storage unit 34. The transmission / reception unit 30 performs data transmission and reception between the video analysis device 13 and the imaging device 11, and between the video analysis device 13 and the display device 12. The transmission / reception unit 30 has wired or wireless communication means and performs mutual communication using a known communication method. Examples of the known communication method include short-range wireless communication such as Bluetooth (registered trademark) communication.

[0025] The video processing unit 31 performs video display processing for displaying the golf swing video input from the imaging device 11 on the video display unit 22 of the video display screen 21 on the display device 12. The above processing is executed at the timing when the golf swing video is input from the imaging device 11, or at the timing when a predetermined operation by the user is performed on the video analysis device 13, for example, an operation of reading the golf swing video stored in the storage unit 34.

[0026] The video processing unit 31 performs processing for displaying an operation icon on the operation unit 23 of the video display screen 21. When an operation is performed on the operation icon, the video processing unit 31 performs processing for displaying a video corresponding to the content of the operation on the video display unit 22, for example, video playback, stop, frame backward, frame forward, and change of the playback position.

[0027] The area setting unit 32 performs an area setting process of setting a part of the range shown in the video as a determination area based on a user operation, targeting the video of the golf swing being displayed on the video display unit 22. As shown in FIG. 2, the area setting unit 32 performs a process of displaying an area setting button 26 on the video display unit 22. The display of the area setting button 26 is performed at the timing when a video of a golf swing in which the process of acquiring the impact timing has not been performed is displayed on the video display unit 22. At this time, the area setting button 26 is displayed as "Detection". Also, in the state of FIG. 2, the timing icon 25 is not displayed on the information display screen 24.

[0028] As shown in FIG. 3(a), when an operation is performed on the area setting button 26 with the "Detection" display, the area setting unit 32 overlays and displays a setting frame 27 for selecting a determination area R on the still image (hereinafter referred to as the target still image) currently being displayed on the video display unit 22. Note that the operation on the area setting button 26 with the "Detection" display is performed when a still image showing the golf ball B before hitting is being displayed.

[0029] The display position, shape, and size of the setting frame 27 are fixed according to the screen size of the video display unit 22. The display position, shape, and size of the setting frame 27 are not particularly limited. The shape of the setting frame 27 is, for example, rectangular, polygonal, or circular. The size of the setting frame 27 is, for example, 40% or more and 80% or less of the screen size of the video display unit 22. In the drawing, as an example, the case where a square setting frame 27 is displayed at the lower center of the video display unit 22 is illustrated.

[0030] Also, when an operation is performed on the area setting button 26 with the "Detection" display, the area setting unit 32 enables the target still image on the video display unit 22 to be enlarged, reduced, and translated in parallel. That is, it enables the user to perform an enlargement / reduction operation and a parallel translation operation on the target still image.

[0031] In this state, the user can arbitrarily adjust the range included within the setting frame 27 in the target still image by operating the target still image. Although details will be described later, as shown in FIG. 3(b), when the golf ball B is included in the setting frame 27 and the main object located within the setting frame 27 is the golf ball B, or the largest object excluding the background is the golf ball B, the size and position of the target still image are adjusted by the user.

[0032] Also, when an operation is performed on the region setting button 26 with the "Detection" display, the display of the region setting button 26 is changed from "Detection" to "Detection Start". When an operation is performed on the region setting button 26 with the "Detection Start" display, the region setting unit 32 sets the determination region R to the same range as the range included within the setting frame 27 in the target still image for a plurality of still images constituting the golf swing video. For example, the region setting unit 32 obtains the range (X1 to X2, Y1 to Y2) of the XY coordinates where the range included within the setting frame 27 in the target still image is located. Then, for a plurality of still images constituting the golf swing video, the portion corresponding to the obtained range (X1 to X2, Y1 to Y2) of the XY coordinates is set as the determination region R. Note that the plurality of still images may be all the still images constituting the golf swing video, or may be a part of the still images extracted at a preset cycle.

[0033] Also, when an operation is performed on the region setting button 26 with the "Detection Start" display, the target still image in the video display unit 22 returns to the display state before the operation on the region setting button 26 with the "Detection" display. Note that, from the time when an operation is performed on the region setting button 26 with the "Detection" display until an operation is performed on the region setting button 26 with the "Detection Start" display, the region setting unit 32 is set to a state where it does not accept input of an operation on the operation unit 23.

[0034] When the determination area R is set by the area setting unit 32, the information acquisition unit 33 performs information acquisition processing for acquiring the timing of the impact. The information acquisition unit 33 determines whether the state in which the golf ball B is reflected in the determination area R is maintained for a plurality of still images in which the determination area R is set.

[0035] FIG. 5 is a diagram in which still images at respective timings important in a golf swing are extracted from each still image constituting the video of the golf swing and arranged in time series. Also shown is the determination area R in each still image. The above timings are address (t0), top (t1), impact (t2, t3), and finish (t4). From the address (t0) to the impact (t2), the state in which the golf ball B is reflected in the determination area R is maintained. On the other hand, after the impact (t3), since the golf ball B has been hit, the state in which the golf ball B is reflected in the determination area R is not maintained.

[0036] The above determination uses an object tracking technique for tracking the position of an object. Specifically, object tracking is performed using a learned AI model only for the determination area R of a plurality of still images constituting the video of the golf swing. At this time, by setting the largest object existing in the determination area R as the golf ball B, the object to be tracked becomes the golf ball B without performing a matching process.

[0037] The AI model outputs a tracking result for a still image in which the golf ball B is reflected in the determination area R, and outputs untrackable for a still image in which the golf ball B is not reflected in the determination area R. When a tracking result is output, the information acquisition unit 33 determines that the state is one in which the golf ball B is reflected, and when untrackable is output, the information acquisition unit 33 determines that the state is one in which the golf ball B is not reflected. Note that the above AI model may be provided in the information acquisition unit 33, or a model installed in a PC, smartphone, tablet terminal, etc. that can be used as the video analysis device 13 may be used.

[0038] As shown in FIG. 5, the information acquisition unit 33 detects the timing (t3) of the still image that has switched from the presence determination (◯) to the absence determination (×) in the time series of the determination results, and acquires the detected timing as the impact timing. As shown in FIG. 4, after acquiring the impact timing, the information acquisition unit 33 displays the timing icon 25 on the information display screen 24. The timing icon 25 is displayed below the point corresponding to the acquired impact timing on the seek bar 23d.

[0039] The video processing unit 31, the region setting unit 32, and the information acquisition unit 33 can be configured as circuitry including one or more processors that operate according to a computer program (software), one or more dedicated hardware circuits such as an application specific integrated circuit (ASIC) that executes at least a part of various processes, or a combination thereof.

[0040] The storage unit 34 stores the data of the golf swing video input from the imaging device 11, and stores the acquired impact timing linked to the data of the original video. Further, the storage unit 34 stores a program 34a for controlling the execution of each process in the video processing unit 31, the region setting unit 32, and the information acquisition unit 33. Note that the program 34a corresponds to a program for causing a computer to function as the video analysis device 13.

[0041] <Video Analysis Method> Next, with reference to FIG. 6, a video analysis method using the video analysis system 10 of the present embodiment will be described. In the video analysis method, the following steps S101 to S105 are performed in order.

[0042] Step S101 is a shooting step of shooting a video of a golf swing using the shooting device 11. The video to be shot is a video obtained by shooting at a fixed point the area surrounding the golfer performing the golf swing and the golf ball B. Step S102 is a video processing step of displaying the video of the golf swing shot by the shooting device 11 on the video display unit 22 of the display device 12. At this time, the video displayed on the video display unit 22 is a video in which the process of acquiring the impact timing is not performed.

[0043] Step S103 is an area setting step of setting a part of the range shown in the video displayed on the video display unit 22 as the determination area R based on the user's operation. Step S104 is an information acquisition step of acquiring the impact timing from the video of the golf swing based on the change within the determination area R set in step S103. Details of the area setting step and the information acquisition step will be described later.

[0044] Step S105 is an information display step of displaying the timing icon 25 on the information display screen 24 of the display device 12. As shown in FIG. 4(a), the information acquisition unit 33 causes the timing icon 25 to be displayed on the information display screen 24. Also, as shown in FIG. 4(b), when the timing icon 25 is operated, the information acquisition unit 33 causes a still image of the impact timing to be displayed on the video display screen 21.

[0045] Next, with reference to FIG. 7, details of the area setting step and the information acquisition step will be described. In the area setting step which is step S103, first, a target still image is selected by the user (step S201). The user operates the operation icon displayed on the operation unit 23 of the display device 12 to make the still image in which the golf ball B before hitting is shown be displayed on the video display unit 22.

[0046] Thereafter, the user operates the area setting button 26 of the "detection" display shown on the video display unit 22 (step S202). When the area setting button 26 is operated, the area setting unit 32 overlays and displays the setting frame 27 on the target still image shown on the video display unit 22 and changes the display state of the target still image (step S203). In step S203, the target still image is changed to a state where it can be enlarged, reduced, and translated parallel on the video display unit 22.

[0047] Next, the user adjusts the size and position of the target still image so that the entire golf ball B is included within the setting frame 27 (step S204). At this time, the size and position of the target still image are adjusted so that the main object located within the setting frame 27 is the golf ball B, or so that the largest object excluding the background is the golf ball B. In particular, it is preferable to adjust so that the area ratio of the golf ball B occupying within the adjusted setting frame 27 is 50% or more. Also, it is preferable to adjust so that the background of the golf ball B is included within the adjusted setting frame 27. For example, adjust so that the area ratio of the background occupying within the setting frame 27 is 40% or more. By adjusting to satisfy one or both of the above requirements regarding the area ratio of the golf ball B and the area ratio of the background, the accuracy of the determination process by the information acquisition unit 33 is improved. Note that the operation for adjusting the size and position of the target still image is performed, for example, by an input operation using a pointing device or a touch operation on a touch panel.

[0048] Thereafter, the user operates the area setting button 26 of the "detection start" display shown on the video display unit 22 (step S205). When the area setting button 26 is operated, the area setting unit 32 sets the same range as the range included within the setting frame 27 in the target still image as the determination area R for a plurality of still images constituting the golf swing video. Thereafter, the information acquisition step, which is step S104, is performed.

[0049] In the information acquisition step, first, a determination process is performed by the information acquisition unit 33. The information acquisition unit 33 determines whether the state in which the golf ball B is reflected in the determination area R is maintained for a plurality of still images in which the determination area R is set (step 301). Then, the information acquisition unit 33 detects the timing of the still image that has switched from the affirmative determination to the negative determination in the time series of the determination result, and acquires the detected timing as the timing of the impact (step S302). After step S302, the information display step which is step S105 is performed.

[0050] Next, the operations and effects of the present embodiment will be described. (1) The video analysis device 13 includes a video processing unit 31 that displays on the display screen 20 of the display device 12 a video of a golf swing that is a video taken at a fixed point, a region setting unit 32 that sets a part of the range shown in the video as the determination region R, and an information acquisition unit 33 that acquires the timing of the impact based on the change within the set determination region R. The region setting unit 32 sets the determination region R based on an operation of the user who selects the range surrounding the golf ball B from the still images that constitute the video and in which the golf ball B is reflected.

[0051] According to the above configuration, a process for acquiring the timing of the impact is executed with the determination region R specified by the user as the target range. Therefore, compared with the case where the timing of the impact is acquired with the entire still image, that is, the entire inside and outside of the determination region R as the target range, the process to be executed by the device is lightened. In the case of the above configuration, a process for setting the determination region R is added, but since the user selects the position and size of the determination region R, the process is a light process that does not involve determination by an AI model or the like.

[0052] Therefore, according to the above configuration, the load on the device for executing the process of acquiring the timing of the impact can be reduced. As a result, the time required for the process of acquiring the timing of the impact can be reduced, and heat generation of the device associated with the process can be suppressed.

[0053] In addition, according to the above configuration, for each video, even when the size thereof, particularly the size of the golf ball B in the video, is different, the impact timing can be obtained in the same manner. Therefore, it is possible to realize obtaining the impact timing corresponding to videos of various sizes.

[0054] (2) The area setting unit 32 superimposes and displays a setting frame 27 for selecting a determination area R on the target still image displayed on the display screen 20. The setting frame 27 can be relatively changed in size and position with respect to the target still image based on a user operation. In the present embodiment, the position and size of the setting frame 27 are fixed, and the target still image can be enlarged, reduced, and translated.

[0055] According to the above configuration, the user can easily perform an operation of selecting a range surrounding the golf ball B with respect to the target still image. In particular, since the configuration allows the target still image to be enlarged, reduced, and translated, even when the golf ball B is shown small, an appropriate range surrounding the golf ball B can be easily selected.

[0056] (3) The information acquisition unit 33 determines whether or not the state where the golf ball B is shown within the determination area R is maintained for a plurality of still images constituting the golf swing video, and acquires the impact timing based on a change in the determination result.

[0057] According to the above configuration, the impact timing can be obtained by a simple determination process. Therefore, the load on the device for executing the process of obtaining the impact timing can be further reduced.

[0058] (4) Select a range surrounding the golf ball B such that the main object located in the determination area R becomes the golf ball B, or such that the area ratio of the golf ball B in the determination area R, that is, the area ratio of the golf ball B in the setting frame 27, becomes 50% or more.

[0059] According to the above configuration, since the determination area R becomes smaller, the effect that the processing to be executed by the device is light, which is the effect of (1) above, can be obtained more remarkably. Further, according to the above configuration, the largest object existing in the determination area R becomes the golf ball B. In this case, the object detected in the determination area R can be regarded as the golf ball B as it is. Therefore, when determining whether the golf ball B exists in the determination area R, it is only necessary to determine whether an object is detected in the determination area R, and it is not necessary to determine whether the detected object is the golf ball B.

[0060] (5) The processing of the information acquisition unit 33 is executed in a stand-alone manner without using a communication network. According to the above configuration, communication costs and processing time can be reduced. In addition, the video of the golf swing may be personal information that can identify the actor in some cases. Therefore, it is possible to suppress the user from feeling a sense of aversion to transmitting personal information to the outside.

[0061] Note that this embodiment can be implemented with the following modifications. This embodiment and the following modification examples can be implemented in combination with each other within a technically non - conflicting range. · The configuration for making the size and position of the setting frame 27 for the target still image relatively changeable is not limited to the above embodiment. For example, the target still image may be fixed, and the size and position of the setting frame 27 may be configured to be changeable. Further, instead of the configuration in which the size and position of the setting frame 27 displayed overlaid on the target still image can be relatively changed, it may be configured to be able to draw a setting frame 27 of an arbitrary shape based on the user's operation.

[0062] · In the determination process of the information acquisition step, the timing of the still image several frames before the still image switched from the presence determination to the absence determination may be used as the impact timing. · The impact analysis device 13 may perform further analysis based on the timing of the acquired impact. For example, based on the timing of the acquired impact as a reference point, a process of cutting out the images of several frames before and after the reference point as a single swing image may be performed. Further, the timings of several frames before and after the reference point may be acquired as the top timing and the finish timing, respectively.

[0063] · The method for determining whether or not the state in which the golf ball B is shown in the determination area R is maintained is not limited to the method using object tracking technology. Other determination methods include, for example, a determination method using a color histogram. Specifically, for a plurality of still images in which the determination area R is set, a color histogram within the determination area R is created, and it is determined whether or not the amount of change in the color histogram is equal to or less than a preset threshold value as compared with the still image one frame before. Then, for a still image in which the amount of change in the color histogram is equal to or less than the threshold value, it is determined that the golf ball B is shown, and for a still image in which the amount of change in the color histogram exceeds the threshold value, it is determined that the golf ball B is not shown.

[0064] · The information obtained from the golf swing video is not limited to the timing of the impact of the golf ball B. For example, when using object tracking technology, other information regarding the start of the movement of the golf ball B such as the initial velocity when the golf ball B is hit and the hitting angle of the golf ball B can also be obtained.

[0065] · The video to be analyzed by the video analysis device 13 is not limited to the golf swing video, and can be applied to the entire video of the process in which a stationary object starts to move. Other videos include, for example, a free kick video in soccer, a gate ball swing video, a tee batting video in baseball, and a start video using a starting block in a short distance race.

[0066] In addition, as a reference example, it is also conceivable to apply the video analysis device 13 to the analysis of videos other than the video of the process in which a stationary object starts to move. For example, it is conceivable to obtain information regarding the movement of a part of the body of the performer as the object, such as when obtaining information on whether the head of the performer is moving based on the video of a golf swing. In this case, the determination region R is set so as to select a range surrounding a part of the body of the performer. Also, in the video of a basketball shot, it is also conceivable to obtain information regarding the operation on a stationary object, such as when obtaining information on the timing at which the ball passes through the basket. In this case, the determination region R is set so as to select a range surrounding the stationary object.

[0067] · The determination process in the information acquisition step may be performed on a server or in the cloud using a communication network. · The video to be analyzed by the video analysis device 13 may be a video captured by a capturing device other than the capturing device 11 constituting the video analysis system 10.

Explanation of Signs

[0068] B… Golf ball R… Determination region 10… Video analysis system 11… Capturing device 12… Display device 13… Video analysis device 20… Display screen 31… Video processing unit 32… Region setting unit 33… Information acquisition unit 34… Storage unit 34a… Program

Claims

1. A video analysis device that acquires information regarding the start of movement of an object based on a video reflecting the movement of the object, comprising: the video is a moving picture in which the process of the stationary object starting to move is fixed-point photographed, the information is the timing of the start of movement of the object, the initial velocity of the start of movement of the object, or the angle of the start of movement of the object, a video processing unit that displays the video on a display screen of a display device, a region setting unit that sets a part of the range shown in the video as a determination region, a determination unit that determines the presence or absence of the object within the set determination region, and acquires the information based on the determination result, or tracks the object within the determination region and acquires the information based on the tracking result, The region setting unit sets the determination region based on an operation of a user who selects a range surrounding the object from a still image that constitutes the video and in which the object is shown. The video analysis device is characterized by this.

2. The video analysis device according to claim 1, wherein the video is a video of a golf swing.

3. The region setting unit overlays and displays a setting frame for selecting the determination region on the still image displayed on the display screen, The video analysis device according to claim 1 or claim 2, wherein the setting frame is displayed so that its size and position relative to the still image can be changed based on an operation of the user.

4. The information acquisition unit determines, for a plurality of still images constituting the video, whether or not the state in which the object is shown within the determination region is maintained, and acquires the timing of the start of movement of the object based on a change in the determination result. The video analysis device according to claim 1 or claim 2.

5. The video analysis device according to claim 1 or claim 2, wherein the processing of the information acquisition unit is executed stand-alone without using a communication network.

6. A video analysis method for acquiring information regarding the start of movement of an object based on a video reflecting the movement of the object, comprising: the video is a moving picture in which the process of the stationary object starting to move is fixed-point photographed, the information is the timing of the start of movement of the object, the initial velocity of the start of movement of the object, or the angle of the start of movement of the object, a video processing step of displaying the video on a display screen of a display device, a region setting step of setting a part of the range shown in the video as a determination region, An information acquisition step of determining the presence or absence of the object within the set determination area and acquiring the information based on the determination result, or tracking the object within the determination area and acquiring the information based on the tracking result. The area setting step is a step of setting the determination area based on an operation of a user who selects a range surrounding the object from a still image constituting the video and in which the object is shown. The video analysis method according to claim 2, wherein, in the area setting step, a range surrounding the object is selected so that the area ratio of the object in the determination area is 50% or more.

7. A program for causing a computer to function as a video analysis apparatus that acquires information regarding the start of movement of an object based on a video showing the movement of the object, The video is a moving image in which a process of the stationary object starting to move is shot at a fixed point. The information is the timing of the start of movement of the object, the initial velocity of the start of movement of the object, or the angle of the start of movement of the object. A video display process for displaying the video on a display screen of a display device, An area setting process for setting a part of the range shown in the video as a determination area, Causing the computer to execute a determination of the presence or absence of the object within the set determination area and acquiring the information based on the determination result, or tracking the object within the determination area and acquiring the information based on the tracking result. The area setting process is a process of setting the determination area based on an operation of a user who selects a range surrounding the object from a still image constituting the video and in which the object is shown.

Images