KR100955539B1 - Motion analysis system for golf swing and motion analysis method adapted for the same - Google Patents

Abstract

Disclosed is a posture analysis system of a golf swing and a posture analysis method using the same. The posture analysis method of the golf swing may include acquiring continuous images capturing the swing motions of the trainer from an image pickup device installed on the front, side, and top of the tutor wearing the garment having a photosensitive identification mark. Image recognition algorithms are applied to the cameras to recognize a high or low luminance pattern including an identification mark, and to extract an image of an address posture, a back swing posture, and a hitting posture using the recognized luminance pattern. And detecting a rotation angle of the left shoulder from the extracted image, acquiring data on an ideal sample position from the detected rotation angle, and displaying the sample position.

According to the present invention, there is provided a posture analysis system and posture analysis method using the same, which corrects an incorrect posture and induces posture correction by providing graphic information comparing and comparing the swing posture of an instructor with an ideal posture.

Description

Posture analysis system for golf swing and posture analysis method using the same {motion analysis system for golf swing and motion analysis method adapted for the same}

The present invention relates to a golf swing posture analysis system and a posture analysis method using the same, a posture analysis system for correcting an incorrect posture and inducing posture correction by providing graphic information comparing and analyzing the swing posture of an instructor with an ideal posture. It relates to a posture analysis method using the same.

In recent years, the popularity of golf has been gradually increasing due to the improvement of living standards and the increase of leisure time, and the number of beginners to golf is increasing as part of leisure utilization. The industrial and service sectors are expanding greatly. However, the fact that beginners who do not have the correct swing posture receive private lessons from professional golfers, there are many golf driving ranges that do not belong to professional golfers and only practice facilities. Or limited use of personal time to meet the teaching time, and sufficient time for a large number of beginners in charge of one professional golfer to conduct intensive private lessons. There are some aspects that are not easy from a golfer's point of view, and there is a limit in that a satisfactory teaching effect cannot be achieved due to the characteristics of posture correction acquired through repeated learning.

An object of the present invention is to provide a posture analysis system and posture analysis method of a golf swing that induces the instructor to correct the posture by analyzing the posture of the teacher and the ideal sample posture.

In order to achieve the above object and other objects, the golf swing attitude analysis method of the present invention,

Acquiring continuous images capturing the swing motion of the teacher from an imaging apparatus provided on the front, side, and top of the tutor wearing the garment having the photosensitive identification mark on the garment;

Recognizing a high luminance or low luminance pattern including the identification mark by applying an image recognition algorithm to the acquired continuous images;

Extracting an image in which an address posture, a back swing posture, and a hitting posture are taken using the recognized luminance pattern;

Detecting a rotation angle of the left shoulder from the extracted image;

Acquiring data on an ideal sample position from the detected rotational angle; And

And displaying the sample position.

Preferably, in the step of displaying the sample position,

The posture of the instructor is added to the captured image, and the sample posture is synthesized and displayed together.

In the present invention,

The angle formed by the rotational angle of the head and the knuckles, when defining the angle variables θ1, θ2, θ3, .., θn,

The data relating to the sample position includes the values of the angle variables θ2, θ3, .., θn outputted from the data table or a functional relationship with the rotation angle θ1 of the left shoulder as an input.

In the present invention,

In the step of extracting the image of the posture,

The posture taken at the moment when the rotational angle of the left shoulder starts to rise is determined as the address posture,

The posture taken at the moment when the rotation angle of the left shoulder is maximized is determined as the back swing posture,

The posture taken at the moment of contact between the club head and the golf ball is determined as the hitting posture.

In the present invention,

The identification mark is attached to the top, bottom and hat worn by the instructor.

On the other hand, the attitude analysis system of the golf swing according to another aspect of the present invention

Clothing having a photosensitive identification mark attached to each part of the body;

An imaging device for capturing images of the front, side, and top of the tutor wearing the garment;

Recognizing a high luminance or low luminance pattern including the identification mark from the image of the teacher,

By using the recognized luminance pattern, the image of the address posture, back swing posture, and hitting posture is extracted.

A host computer for detecting a rotational angle of the left shoulder from the extracted image, and acquiring data on an ideal sample posture from the detected rotational angle; And

And a display device for displaying the sample position under the control of the host computer.

According to the present invention, it is possible to automatically extract important classification operations, for example, address posture, back swing posture, and striking posture, from continuous images capturing full swing motion, and to detect a specific reference angle detected in the extracted image. By comparing and matching with the sample posture, the teacher instructs the student to recognize and correct the wrong part of his posture.

Figure 1 shows the configuration of the golf swing attitude analysis system according to a preferred embodiment of the present invention. In the illustrated posture analysis system, a platform 100 having a location mark 110 indicating a seating position of a golf ball and a scaffold pad 120 for aligning a foot of a tutor with respect to the location mark 110 are provided. . In addition, a plurality of imaging devices C10, C20, and C30 fixedly installed on the front, the side, and the upper side are provided to photograph the teacher's swing motion in a designated direction. In addition, the posture analysis system includes a host computer including a memory device for storing the flow of image data output from the imaging devices C10, C20, and C30, and a digital signal processor that controls the data flow and controls the shooting operation. (PC) is provided. And, ultimately, the display panel 150 which graphically displays the result of comparing and analyzing the teacher's posture and the ideal sample posture, the so-called self-taught function to exercise the self-taughter to check and correct his / her wrong posture. .

The platform 100 is preferably provided with a golf ball detection sensor (LD, PD) for detecting the presence of the golf ball to detect the moment of hitting. For example, the golf ball detection sensors LD and PD may automatically capture the presence or absence of a golf ball by applying an optical mechanism. For example, the golf ball detection sensors LD and PD may be formed of a pair of a laser diode LD installed directly under a position display and emitting a laser, and a photodiode PD collecting a laser on an optical path. In this case, the photodiode PD may output a first signal S1 having a high level and a second signal S2 having a low level as a binarized sensor signal, and indicate a time point t1 at which a signal change is detected. The golf ball can be detected as a hitting time when it is separated. Based on the detected viewpoint, it is possible to accurately extract the frame of the hitting point from the plurality of image frames that are continuously stored on the time axis.

The scaffold pad 120 allows both feet of the instructor to be aligned to maintain the correct body angle as well as to position the instructor. This is not only necessary to match the optimal position to the golf ball and the optimized hitting direction to the target point, but also to capture the swing motion of the tutor in a predetermined direction with the imaging devices fixedly installed at a certain position. However, since the hitting position of the golf ball may vary according to the personal physique conditions of the instructor, the scaffold pad 120 may be installed in such a way that the position can be changed to personalize the hitting position.

The posture analysis system captures images of the front, side, and top surfaces of the teacher's three-dimensional movements projected in a three-dimensional space onto a two-dimensional imaging plane, thereby providing two-dimensional representation of the teacher's posture and ideal sample position taken in each direction. Compare with. Here, the standard posture means an ideal posture obtained by collecting postures of a professional golfer and applying a statistical method thereto. In addition, comparing the teacher's posture with the standard posture means that the posture of the instructor is displayed on the image in which the posture of the instructor is displayed and the standard posture is synthesized and displayed to guide the instructor to correct the wrong part by himself.

The imaging apparatuses C10, C20, and C30 are arranged in three directions independent of each other based on the instructor (or scaffold pad 120), the first camera C10 disposed at the front and the second disposed at the side. It is preferable to basically include the camera C20 and the 3rd camera C30 arrange | positioned upwards. For example, three images captured from the front, the side, and the upper surface are divided and displayed in divided regions on one screen on the display panel 150 at the same posture. Can be checked at the same time and induce self-correction.

The host computer (PC) has a built-in control logic program for controlling all the sensors (LD, PD), imaging devices (C10, C20, C30), and storage devices, etc., so as to analyze the posture of the teacher. There may be. In particular, in the present invention, since the posture of the instructor is recognized by applying a series of image processing techniques to the captured images, the host computer PC preferably includes image processing logic. On the other hand, the display panel 150 induces posture correction by finally returning the result of analyzing the posture of the tutor as graphic information, for example, supporting the display panel 150 at a predetermined height in accordance with the eye level of the instructor. To be placed on the cradle 140. On the other hand, the posture analysis system may be further provided with a user control unit 130 for receiving a variety of commands from the teacher to convert the data into a readable form. For example, the user manipulation unit 130 may be provided at the end of the cable extending from the host computer (PC) in charge of the overall storage and processing of data, or a wireless communication module mounted on the host computer (PC) It may be provided in the form of a remote control for transmitting and receiving data through. The user operation unit 130 may include a start button and an end button for starting and ending a posture analysis procedure. In addition, an address posture, a back swing posture, a hitting posture, etc., for confirming the contents of the lesson regarding a specific classification operation It may include selection buttons for the classification operation of the.

FIG. 2 is a view schematically showing a general golf swing motion. In order to clearly show a posture, the body node is schematically illustrated as a link section. Referring to the drawings, the overall full swing motion of a golf may consist of a series of segmentation operations shown. Here, (a) of Figure 2 is a preliminary operation to start the swing motion to determine the position of the trainer's feet and standing the club head against the ground, Figure 2 (c) is to lift the golf club as far as possible The raised back swing posture, FIG. 2E shows a striking posture at the moment of hitting the golf ball with a golf club, and FIG. 2G shows a finish posture, which is a finishing step after the striking. The series of pre-stroke operations from the address posture (FIG. 2 (a)) to the strike posture (FIG. 2 (e)) can be understood as a swinging movement with the left shoulder as the center of rotation (the hinge axis). The rotation angle θ1 of the left shoulder shows the largest change. That is, the rotation angle θ1 of the left shoulder shows the maximum variation in the movement of the body simplified by the link structure. In the present invention, the posture of the instructor and the standard posture are compared with each other, and it is necessary to compare the division motions of the same step in the overall swing motion. For example, it is necessary to match the teacher's address posture with the standard address posture to be compared with each other, and if the teacher's address posture and the standard hitting posture are compared with each other, accurate correction is impossible. In addition, it is not easy to extract a comparison object (standard posture) that exactly corresponds to the posture of the instructor even with the same address posture in view of successive swing motions. In the present invention, the instructor and the standard posture are matched with each other by using the rotation angle θ1 of the left shoulder which is instantaneously changed in the swing motion. That is, when the rotational angle θ1 = α of the teacher's left shoulder is measured, the standard position corresponding to the rotational angle θ1 = α of the left shoulder is recommended from the previously inputted standard position.

In one embodiment of the attitude analysis system, several designated momentary motions are captured in the full swing motion consisting of continuous motions, and are compared with the standard postures of the same viewpoint stored in the system. Here, in the designated momentary motion, the address posture (FIG. 2A), the back swing posture in which the golf club is lifted as far as possible (FIG. 2C), and the striking attitude at the moment of hitting the golf ball (FIG. 2). It is preferable to include (e)) basically, and in addition to this, it may also include a finish posture (FIG. 2 (g)) which is a finishing step after the hitting.

At this time, the posture analysis system applies an image processing technique to the captured images in order to automatically capture the moment when self-designated actions are taken without relying on the instructor, and confirms that the unique characteristics of each motion appear. . For example, since the club head is rotated as far as possible in the back swing posture (FIG. 2C), the club head rotation angle is maximized by analyzing image signals continuously output from the upper camera C30. The detected time may be determined as the back swing time, and an image frame of the back swing time may be accurately selected from a plurality of image frames stored in the memory based on the time axis.

And, the moment hitting the golf ball ((e) of Figure 2) can be detected by grasping the time when the golf ball is separated from the stop position (corresponding to the position display 110), the golf ball sensor (LD, PD) The hitting moment t1 can be detected based on the output signal, and the hitting posture of the hitting moment t1 detected from the continuous flow of images obtained from the first to third cameras C10, C20, and C30. Can be judged by the captured image. Alternatively, the moment of hitting the golf ball may be detected by applying an image recognition algorithm to capture the moment when the club head and the golf ball contact each other. In this case, it is possible to detect the moment of hitting in a sensorless manner that does not require a separate sensor, and there is an advantage in that the cost of constructing the entire system can be reduced by using software processing.

On the other hand, in the address operation (Fig. 2 (a)), it is possible to capture the time when the club head close to the ground starts to rise and define it as the address moment. The tutor generally requires some preparation time for aiming the golf ball along with aligning both feet on the scaffold pad 120 and obtaining a correct body angle. Therefore, it would be desirable to define the starting point of the club head with sufficient preparation time to enter the full swing motion as the address moment. An additional algorithm is required to capture the address moment, which will be described with reference to FIG. 3. Figure 3 shows how the rotation angle θ1 of the left shoulder changes from address to striking. Referring to the drawing, it may be divided into a fine movement section (P1) showing the minute change in the rotation angle (θ1) and the rising section (P2) in which the club head is raised in earnest from the ground in the beginning The instant when the first line L1 and the second line L2 interpolate the fine movement section P1 and the rising section P2 linearly may be detected as an address instant t0.

On the other hand, from a kinematic point of view, the present invention simplifies and recognizes the body structure of the instructor as a link structure, and more specifically, regards the link structure of a plurality of sections rotatably connected to each other using a pin joint. That is, as shown in FIG. 4, each body node of the teacher is modeled as a link clause, and each joint part is modeled as a pin joint. By applying an image recognition algorithm to capture each body node of the teacher, the body structure can be shaped into a link structure. And the intersection where these body nodes intersect with each other can be detected as a joint part. As the body node, a shoulder node connecting the left shoulder and the right shoulder to each other, and an upper body upper part and a lower node which shape the spine skeleton across the upper body can be considered. The upper and lower nodes can be considered for each of the limbs of the left arm, right arm, left leg, and right leg. On the other hand, although not the body node of the teacher, the golf club adjusted by the rotation of the wrist of the teacher provides important information in the posture analysis, it can be treated like the body node. Joint parts of the body are parts that can be rotated relative to each other, and are present in upper torso parts such as shoulders, elbows and wrists, and lower torso parts such as waists and knees.

In the present invention, each body node is displayed as a photosensitive identification mark so that image recognition based on the distribution of luminance signals can be efficiently performed. For example, by adopting a light sensitive identification mark captured in a high luminance pattern or a low luminance pattern in response to ambient background light, a series of preprocessing steps for enhancing the recognition rate can be omitted in the image processing process. It can dramatically reduce the processing burden on the host computer and eliminate the fear of recognition failure.

In relation to labeling body nodes with a photosensitive identification mark, for example, a luminous band is produced in the joint portion of the body that generates high brightness in response to ambient light, or conversely absorbs ambient light to produce low brightness. A way of fitting the dark band to the joint part of the body can be illustrated. Feature points may be detected by capturing a high luminance pattern or a low luminance pattern from an image output from each of the first to third cameras C10, C20, and C30 in comparison with a peripheral region. However, based on the applicant's experimental results and considering light reflecting elements scattered in the surrounding environment, it is judged that capturing a low luminance pattern rather than capturing a high luminance pattern in the collected image data is advantageous in accuracy of position detection. do.

In order to reduce the trouble of individually marking the identification marks for each body segment, a group of identification marks can be garmented. That is, by wearing a garment with an identification mark attached to each required position, it is possible to simultaneously mark a plurality of feature points with one detachment. 5 shows an embodiment of the garment with the identification mark (M) attached. The garment attached with the identification mark (M) can be provided in the top / bottom for the convenience of wearing, and the attachment position of the identification mark (M), two for each limb of the body, one on each shoulder, and waist It can be provided one by one at the chest and waist to detect the baby's bow. Looking more specifically, the upper left arm, the lower left arm, the upper right arm, the lower right arm, and the left shoulder and right shoulder, the identification mark (M) may be attached to each of the chest and waist. And, the bottom of the upper left leg and lower left leg, and the upper and lower right leg can be attached to the identification mark (M), respectively. By capturing the identification mark M in the image of the teacher, the corresponding body node can be detected. For example, the shoulder node can be detected by capturing the identification marks (M) marked on both shoulders of the instructor and connected to each other, and the rotation state of the shoulder node can be known.

On the other hand, since the teacher's head rotation angle, that is, the eyeline direction, is also an important factor for evaluating the swing posture, the identification mark M for displaying the head rotation angle is also separated from the identification mark M for each body node position. It is preferred to be provided together. For example, as shown in Figure 5, the identification mark (M) is attached to the hat worn by the teacher, more specifically, may be provided in a stripe pattern extending along the front direction of the teacher. In addition, it is preferable that the identification mark M also basically has an optical characteristic captured in high or low luminance in response to ambient light. At this time, it is possible to easily recognize the identification mark having a clear luminance difference from the surrounding background from the image data obtained from the third camera C30 in the upper position.

On the other hand, Figure 6 is a view showing the address posture from the front by shaping the body of the teacher in a link structure. In the following, the linked body structure is considered and from the kinematic point of view, the swing position of the instructor is described, and the angle variables to be analyzed / corrected will be described. Here, describing the swing posture of the instructor means that the posture taken by the instructor can be largely described using the selected angular parameters, and other kinematic features observed in the instructor's posture are linked depending on the angular variables. . However, the angular variables introduced below are merely exemplary, and it is possible to increase the number of angular variables for more detailed and detailed analysis.

As described above, the teacher's posture and the ideal sample posture are matched with each other for comparison by the rotation angle θ1 of the left shoulder. Here, the ideal sample position is stored in the form of a data table in which the arbitrary rotation angle θ1 and the set of remaining angle variables (θ2, θ3, .., θn) 1: 1 correspond to each other, so that the remainder for a specific rotation angle θ1 = α. The values of the angle variable θ2 = β, θ3 = δ, .., θn = Ω may be provided. Therefore, the values θ2 = β, θ3 = δ,., Θn = Ω of the other angle variables provided from the data table are recommended as the teaching attitude to the teacher. However, rather than providing the value of such an angle variable as a quantitative value itself, it is possible to give a visual effect by displaying the recommended angle variable as a wedge mark having an angle of inclination. In addition to the data table, the sample position is stored as a mathematical function relationship that takes the rotation angle θ1 of the left shoulder as an argument and returns other angle variables θ2, θ3, .., θn as a function value. There may be. Accordingly, the rotation angle extracted from the teacher's posture, θ1 = α, is input, and values of other angle variables, θ2 = β, θ3 = δ, .., θn = Ω, are provided from the functional relationship stored in the system. In other words, the so-called self-teaching can be performed by the instructor to identify and correct the wrong part of his posture.

On the other hand, some angle variables, including the rotation angle (θ1) of the left shoulder described above are measured based on the horizontal axis and the vertical axis, the image input from the outside is required for the horizontal axis and the vertical axis. For example, a stripe pattern is artificially formed along a vertical axis on a fixed structure, such as a cradle 140 for fixing the display panel 150, and the image of the stripe pattern is recognized as a vertical axis, and a right angle relationship is obtained. It can be recognized as a horizontal axis. In addition, with respect to the upper image acquired from the upper position of the instructor, the first and second axes are formed in a cross shape around the scaffold pad 120 where the instructor is located, and the reference direction can be set when the images are recognized. will be.

Hereinafter, various angle variables defined from the front image, the side image, and the top image of the instructor will be described. However, the angular variables listed below are exemplary, and various other angular variables may be further defined. Referring to FIG. 6, from the front image, a rotation angle θ1 of the left shoulder is defined, which is an angle formed by the upper axis of the left arm and the vertical axis, or an angle formed by the vertical axis of the upper and lower nodes of the left arm (θ1, the rotation angle of the left shoulder). It can be defined as. Here, the upper and lower nodes of the left arm mean a virtual body node consisting of a straight line interpolating the upper and lower nodes of the left arm. Also, from the front image, an angle θ2 formed between the upper and lower nodes of the left arm may be defined, and an angle θ3 formed between the lower arm and the golf club may be defined. In addition, an angle θ4 at which the shoulder node forms the horizontal axis may be defined. For example, the shoulder node may be defined as a line connecting the right shoulder and the left shoulder to which the identification mark M is attached. .

FIG. 7 is a side view of the instructor in an address posture, in which the body is shaped into a link structure. Referring to the drawings, an angle θ5 formed between the upper and lower nodes of the right arm and the golf club may be defined. The upper and lower nodes of the right arm cover both the upper and lower nodes of the right arm, and the angle θ5 is roughly shown. By interpolating the upper and lower nodes to form a new upper and lower nodes, it can be defined as the angle between the upper and lower nodes and the golf club. Alternatively, the angle θ5 may be defined as the angle between the upper or lower node and the golf club. On the other hand, in the upper body, the lower back lean angle θ6 and the back unfolding angle θ7 may be defined. The back spread angle may be defined as an angle between the upper and lower nodes of the upper body. In the lower body, an angle θ8 formed between the upper and lower nodes of the right leg may be defined.

FIG. 8 is an upper image of a teacher in an address posture, in which a body is formed into a link structure. Referring to the drawings, an angle θ9 formed between the upper and lower nodes of the left arm and the first axis may be defined. The upper and lower nodes of the left arm cover both the upper and lower nodes of the left arm, and the angle θ9 roughly interpolates the upper and lower nodes to form a new upper and lower nodes, at an angle formed by the first axis. Can be defined. Alternatively, the angle θ9 may be defined as the angle between the upper or lower node of the left arm and the first axis.

In addition, an angle θ10 between the upper and lower nodes of the left arm and the club head may be defined, and between the shoulder node and the second axis connecting the right shoulder and the left shoulder with the identification mark M in a straight line. The angle θ11 of may be defined. In addition, an angle θ12 between the teacher's eyeline direction S and the first axis may be defined from the identification mark M attached to the hat worn by the teacher.

The angle variables θ1 to θ12 defined above may be defined in all division motions of the full swing motion, including the address posture and subsequent back swing posture, hitting posture, and finish posture. For these angle variables, the recommendation values returned from a data table or a function relation previously stored in the system are presented by inputting the rotation angle θ1 of the left shoulder taken by the teacher. However, when performing an automatic lesson by capturing any momentary motions among the full swing motions of the instructor, all the recommended values for the above-described angular variables may be presented, but in some cases, selected according to the classification operation step. Attitude analysis may be performed based on operating variables.

Hereinafter, the attitude analysis method of the golf swing will be described. The host computer PC, which is supplied with the operating power and waits for the instruction of the teacher, may start a series of processes for posture analysis by detecting a start signal from the user operator 130. However, the posture analysis may be started only by the start signal, but the posture analysis may be started with sufficient conditions that the weight of the instructor is detected from the scaffold pad 120 on the platform 100 together with the start signal. More specifically, the host computer PC operates the imaging apparatuses C10, C20, and C30 to capture the swing motion of the teacher as continuous images (S11). For example, image data output at 60 frames per second from the imaging devices C10, C20, and C30 are accumulated in the storage device in a continuous flow about the time axis. After the trainee's full swing motion is captured, when the end signal of the trainee is detected from the user manipulation unit 130, the host computer PC stops shooting operations of the imaging devices C10, C20, and C30 according to the detected signal. Let's do it. Then, a full-fledged posture analysis is started by applying an image recognition algorithm to the image data accumulated in the storage device (S12). That is, the host computer PC detects from the luminance distribution of the image data a high luminance having a relatively high luminance or a low luminance region having a relatively low luminance. In this regard, the light sensitive identification mark M attached to the garment and hat worn by the instructor will be captured in a relatively high or low luminance area relative to the surroundings. In addition, the golf club may be easily captured in a high luminance region compared to the surroundings even if it is not labeled with a distinctive metallic luster.

Next, by using the high and low luminance regions that are recognized in the image, an image obtained by capturing respective postures, for example, an address posture, a back swing posture, and a hitting posture, is extracted from the teacher's full swing motion (S13). ). In this case, a plurality of image data continuously stored on the time axis may be stored in the storage device, and the host computer PC may select images of respective postures. For example, the moment of striking can be accurately detected based on the output signals of the golf ball sensors LD and PD, and the front, side, and top images of the striking posture can be selected. In addition, as shown in FIG. 3, the address posture defined as the point in time at which the club head starts to rise is monitored by the change in the rotation angle θ1 of the left shoulder, at which point in time t0 at which the ascending section P2 starts. Can be determined as an address moment to extract images of the front, side, and top of which the address posture is captured. In the back swing posture, the club head is rotated as far as possible, and the image of the camera C30 positioned upward is analyzed to detect the back swing time point when the club head rotation angle is maximized. Images of the front, side, and top of which the back swing posture is captured can be extracted.

After the images of the address posture, the back swing posture, and the hitting posture are captured, the rotation angle of the left shoulder taken by the instructor is extracted from the selected images (S14). Then, the database inputted to the system is inquired by using the extracted left shoulder rotation angle (θ1 = α), or another angle is derived from the functional relationship input to the system by inputting the rotation angle (θ1 = α). Variables are obtained, for example, the recommended values of the various angular variables [theta] 2, ..., [theta] 12 shown in FIGS.

Next, the teacher's posture is captured and the recommended values of the angle variables θ2,..., Θ12 are synthesized (overlayed) and displayed. The recommended values of the angle variables θ2,..., Θ12 are preferably provided in a graphical representation in a wedge shape branched at an angle between the angles so that the instructor can easily understand the text. something to do. On the other hand, it is preferable to allocate one screen for each division operation such as an address posture, a back swing posture, and a hitting posture, and to display the front, side, and top images of each division operation by dividing the assigned one screen. . For example, front, side, and top images taken from an address posture are provided on one screen along with the contents of the lesson (corresponding to the recommendation for the angular variable), and then the back swing posture and As the screen is switched according to the selection of the hitting posture, the front, side, and top images corresponding to the selected posture are displayed on one screen.

Meanwhile, the present invention divides the golf swing into an embodiment in which the entire course of the golf swing from start to completion is continuously photographed, and then the teaching content is added to the posture taken by the instructor, and the golf swing is divided into several division motion steps. Consider all embodiments that provide intensive instruction for any selected step. In the former case, the form of post-checking one's posture and teaching contents for each stage after the full swing is performed. In the latter case, the posture correction can be performed while checking his posture and teaching contents in real time. to be. For example, if you want to focus on address posture, you can use the mode selection function to select an address mode. At this time, the instructor can perform intensive training on a specific step by correcting his posture to approach the posture while checking the actual posture synthesized in the image of the posture he is currently taking. Preferably, a separate mode may be supported for each partitioning operation, for example an address mode, a back swing mode, and a strike mode may be provided by default.

Meanwhile, FIGS. 10A to 10D are photographs taken from the front of the swing motion, and more specifically, photographs of the address posture, the back swing posture, the hitting posture and the finish posture, respectively. 11A to 11D are photographs taken from the side of the swing motion, and photographs photographed from the posture of the finish posture, respectively. 12A to 12D are photographs taken from above of swing motion, respectively, and photographs photographing a finish posture from an address posture.

Although the present invention has been described with reference to the embodiments illustrated in the accompanying drawings, it is merely exemplary, and various modifications and equivalent other embodiments are possible from those skilled in the art to which the present invention pertains. You will understand the point. Therefore, the true scope of protection of the present invention should be defined by the appended claims.

1 is a general configuration of a golf swing attitude analysis system.

FIG. 2 is a diagram sequentially illustrating a continuous sorting operation of a golf swing.

3 is a diagram for explaining a technique of capturing an address instant.

4 shows a body structure shaped as a link clause.

FIG. 5 is a diagram illustrating a correspondence between the garment identification mark and each body node recognized as the identification mark.

6 illustrates a front address posture and shows various angle variables defined in the front posture.

FIG. 7 illustrates a side address posture and shows various angle variables defined in the side posture.

8 is a diagram illustrating an upper address posture and showing various angle variables defined in an upper posture.

Figure 9 is a flow chart showing a series of steps performed in the attitude analysis method of the golf swing according to another aspect of the present invention.

10A to 10D are photographs taken from the front of the address, the back swing, the hitting, and the finish posture.

11A to 11D are photographs taken from the side of an address, a back swing, a strike, and a finish posture.

12A to 12D are photographs taken from above of an address, a back swing, a strike, and a finish posture.

<Description of the symbols for the main parts of the drawings>

100: platform 110: location mark

120: scaffold pad 130: the user control panel

140: cradle 150: display panel

C10, C20, C30: Image pickup device PC: Host computer

LD, PD: Golf ball sensor M: Identification mark

θ1, θ2, θ3, .., θ12: Angle variable

Claims (6)

Acquiring continuous images capturing the swing motion of the teacher from an imaging apparatus provided on the front, side, and top of the tutor wearing the garment having the photosensitive identification mark on the garment; Recognizing a high luminance or low luminance pattern including the identification mark by applying an image recognition algorithm to the acquired continuous images; Extracting an image in which an address posture, a back swing posture, and a hitting posture are taken using the recognized luminance pattern; Detecting a rotation angle of the left shoulder from the extracted image; Acquiring data on an ideal sample position from the detected rotational angle; And Displaying the sample position; including; The angle formed by the rotational angle of the head and the body node, and when defining the angle variables θ1, θ2, θ3, .., θn, The data relating to the sample position includes the values of the angle variables θ2, θ3, .., θn outputted from a data table or a functional relationship with the rotation angle θ1 of the left shoulder as an input. Posture analysis method. The method of claim 1, In the step of displaying the sample position, Posture analysis method of the golf swing, characterized in that the synthesis of the sample posture in addition to the captured image of the trainer. delete The method of claim 1, In the step of extracting the image of the posture, The posture taken at the moment when the rotational angle of the left shoulder starts to rise is determined as the address posture, The posture taken at the moment when the rotation angle of the left shoulder is maximized is determined as the back swing posture, A posture analysis method of a golf swing, characterized in that the position determined at the moment of contact of the club head and the golf ball as the hitting position. The method of claim 1, The identification mark is a posture analysis method of the golf swing, characterized in that attached to the top, bottom and hat worn by the instructor. Clothing having a photosensitive identification mark attached to each part of the body; An imaging device for capturing images of the front, side, and top of the tutor wearing the garment; Recognizing a high luminance or low luminance pattern including the identification mark from the image of the teacher, By using the recognized luminance pattern, the image of the address posture, back swing posture, and hitting posture is extracted. A host computer for detecting a rotational angle of the left shoulder from the extracted image, and acquiring data on an ideal sample posture from the detected rotational angle; And And a display device for displaying the sample position under the control of the host computer. The angle formed by the rotational angle of the head and the body node, and when defining the angle variables θ1, θ2, θ3, .., θn, The data relating to the sample position includes the values of the angle variables θ2, θ3, .., θn outputted from a data table or a functional relationship with the rotation angle θ1 of the left shoulder as an input. Posture Analysis System.

Images