Method for calculating rotation of golf ball moved by hit and rotation calculating device using the same

Abstract

A method for calculating rotation of a golf ball moved by being hit according to an embodiment of the present invention includes the steps of acquiring a first image and a second image, which are successive images, with respect to a golf ball moved by being hit, the golf ball being marked with a plurality of marks; calculating information about a reference mark and a plurality of peripheral marks around the reference mark with respect to the plurality of marks on a first ball image corresponding to a portion of the golf ball of the first image; specifying a combination of marks corresponding to the reference mark and the plurality of peripheral marks of the first ball image in a second ball image corresponding to a portion of the golf ball of the second image; and calculating a rotation axis and a rotation amount using respective center points of corresponding mark pairs on the first ball image and the second ball image.

Description

Technical Field

[0001] This invention relates to a method and a spin calculation device that can calculate the spin of a golf ball by analyzing markings on the surface of the golf ball in an image captured by the invention, and by analyzing the spin information of the golf ball as it moves rapidly as it is hit by a golf club. Background Technology

[0002] In ball-based sports, especially golf, there has been a long-standing effort to accurately sense the physical properties of a ball as it is struck by a golfer and to use the sensed values ​​for shot analysis, or to visualize them for use in simulated golf, such as so-called screen golf.

[0003] In particular, the spin of a ball that flies away after being hit is difficult to measure because it rotates at a very high speed around an axis in three-dimensional space. Furthermore, accurate measurement requires very expensive equipment, such as radar sensors.

[0004] However, such expensive sensing devices are not suitable for general-purpose sensing devices used in so-called screen golf or golf driving ranges to analyze ball play by sensing the ball hit by the user's golf swing, calculating the ball's trajectory, and thereby performing golf simulations on virtual golf courses. Furthermore, there is a current need to develop a technology that can quickly and accurately sense the ball's spin in relatively inexpensive and low-performance systems.

[0005] In addition to the radar sensor mentioned above, a sensing system using a camera is used as a device to sense the rotation of a golf ball.

[0006] In the aforementioned camera sensing system, when a golf ball is hit and moves, the camera captures an image of it and analyzes the portion corresponding to the golf ball from the captured image, thereby calculating the rotation of the golf ball as the subject.

[0007] Since the golf ball rotates at a very high speed when it is hit and moves, when calculating the rotation of the golf ball using the camera sensing system described above, it is preferable to use an ultra-high-speed camera with a high shooting speed.

[0008] There are two main ways to use camera sensing systems to detect the spin of a golf ball: one is to manually mark the golf ball with a specific shape, extract the mark from the image captured by the camera, and calculate the spin by analyzing how it moves; the other is to analyze the trademarks or dents already on the golf ball in the captured image without making any markings, and calculate the spin based on these.

[0009] In the latter case, since it is necessary to analyze unspecified markings on the golf ball, a high-specification camera system with high resolution and ultra-high speed is required. Furthermore, since it is necessary to find unspecified markings and determine similarity through images each time a sensor is detected, there is a problem that the accuracy of the rotation calculation results is not constant.

[0010] As prior art corresponding to the former situation mentioned above, there are Korean Patent No. 10-1386793, Korean Patent No. 10-1182393, Japanese Patent No. 3235987, and US Patent No. 7324663, etc.

[0011] The existing technology described above is based on the shape features of specific marks marked on a golf ball. It matches the marks to each other by the shape similarity of specific marks between two consecutive ball images and uses the matching result to calculate the rotation. Therefore, it has the advantage of being able to calculate the rotation to a certain extent.

[0012] However, in the case of calculating the rotation based on the shape features of specific marks on a golf ball as described above, there are problems. Even at a relatively fast rotation speed, it is necessary to ensure that the specific marks are on the captured image. Therefore, it is necessary to use expensive camera devices with very fast image acquisition speed. Furthermore, the computational workload of extracting specific marks by image processing of the acquired images and judging the shape similarity of specific marks in each of the two ball images is quite large, resulting in slow processing speed. Summary of the Invention

[0013] Technical issues

[0014] This invention provides a method for calculating the spin of a golf ball that has been struck and moved, and a spin calculation device using the same. The method marks multiple markers on the golf ball and quickly acquires images of a designated area using a camera. It then uses the relative relationships between the multiple markers in the acquired ball images to accurately match the markers between two consecutive ball images to calculate the spin, thereby increasing the calculation speed for spin calculation and enabling accurate spin calculation.

[0015] Technical solution

[0016] An embodiment of the present invention provides a method for calculating the rotation of a golf ball that has been struck and moved, comprising the steps of: acquiring consecutive images, namely a first image and a second image, for a golf ball marked with multiple marks that has been struck and moved; calculating information about a reference mark and multiple peripheral marks surrounding the reference mark for multiple marks on a first ball image corresponding to a portion of the golf ball in the first image; specifying a combination of marks corresponding to the reference mark and multiple peripheral marks of the first ball image in a second ball image corresponding to a portion of the golf ball in the second image; and calculating the rotation axis and the amount of rotation using the center points of corresponding mark pairs on the first ball image and the second ball image.

[0017] Furthermore, preferably, the method is characterized in that the step of combining the specific corresponding marks includes the step of combining marks on a second sphere image that have a specific relative relationship matching information about the relative relationship of each of the plurality of peripheral marks relative to the reference mark on the first sphere image.

[0018] Furthermore, preferably, the method is characterized in that the step of specifying the corresponding combination of markers includes: calculating feature information as relative angle information of the tilt of each of the plurality of peripheral markers relative to the reference marker on the first sphere image; calculating relative angle information of the tilt in each combination of any reference marker on the second sphere image and the plurality of peripheral markers relative to it; and matching the feature information of the markers on the first sphere image with the relative angle information of the markers on the second sphere image, and specifying the matched combination of markers on the second sphere image.

[0019] Furthermore, preferably, the method is characterized in that the steps of calculating the rotation axis and the amount of rotation include: plotting the center points of corresponding marker pairs on the first sphere image and the second sphere image on a sphere; and calculating the line connecting the center points of the corresponding marker pairs and the line intersecting two or more planes derived from the center point of the sphere as the rotation axis.

[0020] Furthermore, preferably, the method is characterized in that the steps of calculating the rotation axis and the amount of rotation include: calculating vectors perpendicular to the rotation axis from each center point of the corresponding pair of marks; projecting the calculated vectors onto a reference plane perpendicular to the rotation axis; and calculating the amount of rotation centered on the rotation axis using the vectors used in the reference plane.

[0021] Furthermore, preferably, the method further includes: calculating the amount of line-of-sight rotation at each position in the image based on the fixed line of sight of the camera between two consecutive images of the moving golf ball, and performing a correction step of subtracting the calculated amount of line-of-sight rotation from the amount of rotation calculated by the step of calculating the rotation axis and the amount of rotation.

[0022] On the other hand, an embodiment of the present invention provides a spin calculation device for a golf ball that has been hit and moved, comprising: a single camera separately configured from the camera system of a sensing device for calculating motion information of the golf ball that has been hit and moved, to acquire consecutive images of the golf ball marked with multiple marks as it has been hit and moved, namely a first image and a second image; a mark matching processing unit that calculates information about a reference mark and multiple peripheral marks around the reference mark for multiple marks on a first ball image corresponding to a portion of the golf ball in the first image, and specifically corresponds to a combination of marks of the reference mark and multiple peripheral marks of the first ball image in a second ball image corresponding to a portion of the golf ball in the second image; and a spin calculation unit that calculates the axis of rotation and the amount of rotation using the center points of each corresponding mark pair on the first ball image and the second ball image.

[0023] On the other hand, another embodiment of the present invention provides a spin calculation device for a golf ball that is moved by being hit, comprising: a mark matching processing unit that receives, from a camera of a sensing device for calculating motion information of a golf ball that is moved by being hit, consecutive images of a golf ball marked with multiple marks, namely a first image and a second image, and calculates information about a reference mark and multiple peripheral marks around the reference mark for multiple marks on a first ball image corresponding to a portion of the golf ball in the first image, and specifically corresponds to a combination of marks of the reference mark and multiple peripheral marks of the first ball image in a second ball image corresponding to a portion of the golf ball in the second image; and a spin calculation unit that calculates the rotation axis and the amount of rotation using the center points of each corresponding mark pair on the first ball image and the second ball image.

[0024] Furthermore, preferably, the device is characterized in that the marker matching processing unit is configured to: receive coordinate information of the placement of the golf ball from the sensing device, and based thereon set the area of ​​the image to be acquired by the single camera as a region of interest; and acquire the image of the set region of interest as the first image and the second image.

[0025] Furthermore, preferably, the device is characterized in that the mark matching processing unit is configured to: calculate feature information as relative angle information of the tilt of each of the plurality of peripheral marks relative to the reference mark on the first ball image; calculate relative angle information of the tilt in each combination of any reference mark on the second ball image and the plurality of peripheral marks relative to it; and match the feature information of the marks on the first ball image with the relative angle information of the marks on the second ball image, and specify the matched combination of marks on the second ball image.

[0026] Furthermore, preferably, the device is characterized in that the rotation calculation unit is configured to: display the center points of corresponding marker pairs on the first sphere image and the second sphere image on a sphere; calculate the line connecting the center points of the corresponding marker pairs and the line intersecting two or more planes derived from the center points of the sphere as the rotation axis; calculate the vectors perpendicular to the rotation axis from the center points of the corresponding marker pairs respectively; and project the calculated vectors onto a reference plane perpendicular to the rotation axis respectively, so as to calculate the rotation amount centered on the rotation axis using the vectors used in the reference plane.

[0027] The effects of the invention

[0028] The method for calculating the rotation of a golf ball that has been struck and moved, and the rotation calculation device thereof, of the present invention have the following effects: multiple marks are marked on the golf ball and images of a set area are quickly acquired by a camera, and the relative relationships between the multiple marks in the acquired ball images are used to accurately match the marks between two consecutive ball images to calculate the rotation, thereby increasing the calculation speed for calculating the rotation and enabling accurate rotation calculation. Attached Figure Description

[0029] Figure 1 (a) is a block diagram illustrating the structure of a rotational computing device according to an embodiment of the present invention. Figure 1 (b) is a block diagram illustrating the structure of a rotational computing device according to another embodiment of the present invention.

[0030] Figure 2 This is a flowchart illustrating a rotation calculation method according to an embodiment of the present invention.

[0031] Figure 3 This is an example of a ball image showing a golf ball with multiple markings on its surface and a diagram showing the markings in the ball image to specify the coordinates of their center points. This golf ball is used in a spin calculation device and spin calculation method according to an embodiment of the present invention.

[0032] Figure 4 It is shown as follows Figure 3 (b) shows the results of markings on a specific sphere image and calculating the tilt information for each mark.

[0033] Figure 5 and Figure 6 This diagram illustrates the process of matching markers on a first sphere image and a second sphere image using a rotation calculation apparatus and rotation calculation method based on an embodiment of the present invention.

[0034] Figure 7 It shows that based on such Figure 6 The matching of markers between the first and second sphere images described above will plot the center point of each matching marker on a sphere.

[0035] Figure 8 It shows the use of Figure 7 A diagram illustrating an example of a method for calculating the axis of rotation from the center points of the marks on the sphere shown.

[0036] Figure 9 It shows the use of Figure 8 The diagram shows an example of a method for calculating rotation using the axis of rotation and the center points of each marked point in the sphere.

[0037] Figure 10 This diagram illustrates the amount of rotation that occurs during a non-rotational movement from a first sphere image to a second sphere image while the camera's line of sight remains fixed in a rotational calculation device according to an embodiment of the present invention. Detailed Implementation

[0038] The following description, with reference to the accompanying drawings, details the method for calculating the spin of a golf ball that has been struck and moved according to the present invention, and the spin calculation device utilizing it.

[0039] In this invention, the rotation calculation device is basically implemented by using a predetermined camera to capture a user hitting a golf ball with a golf club and analyzing the captured image to calculate the rotation of the hit ball. The rotation calculation device of this invention can be implemented as a function of the sensing device by using the camera of the sensing device that senses the motion of the golf ball, or it can be implemented independently of the aforementioned sensing device as a device that uses a single camera for calculating rotation and uses the image acquired by the single camera to calculate rotation.

[0040] For example, regarding the former above. Figure 1 As shown in (a), regarding the latter above, for example Figure 1 As shown in (b).

[0041] First, regarding the present invention... Figure 1The embodiment shown in (a) will be described. One embodiment of the rotation calculation device of the present invention is a device that calculates rotation using the camera of the sensing device SD, and may be configured to include a mark matching processing unit 510 and a rotation calculation unit 520.

[0042] Here, the sensing device SD can be a device that calculates the three-dimensional coordinate information of the golf ball moving in space by taking a picture of the golf ball being hit by a golf club by user P and analyzing the captured image, and calculates information about the initial velocity, direction angle, altitude angle and other motion characteristics of the golf ball based on the three-dimensional coordinate information.

[0043] The sensing device SD can be applied to various fields such as golf swing analysis based on the user or virtual golf simulation based on virtual reality.

[0044] The sensing device SD may be configured to include cameras 110 and 120 and a sensing processing unit 200.

[0045] The camera of the sensing device SD is configured to continuously acquire images from the perspective of observing a moving golf ball. To calculate the three-dimensional spatial position information of the moving golf ball, it is preferable to use multiple cameras that acquire images of the same object from different perspectives, for example... Figure 1 (a) and Figure 1 The first camera 110 and the second camera 120 shown in (b) are synchronized with each other and configured in a stereoscopic manner.

[0046] As described above, by configuring the cameras of the sensing device in a stereo method with multiple cameras 110 and 120 synchronized with each other, the two-dimensional information of the golf ball extracted from each of the images obtained by the first camera 110 and the images obtained by the second camera 120 for the same object (golf ball) can be converted into three-dimensional information.

[0047] like Figure 1 (a) and Figure 1 As shown in (b), the sensing processing unit 200 of the sensing device SD can be configured to include an image processing unit 210 that collects images from multiple cameras 110 and 120 respectively and performs predetermined image processing to extract corresponding golf balls, and an information calculation unit 220 that calculates three-dimensional position information from the two-dimensional position information of the golf balls extracted from the images.

[0048] The sensing processing unit 200 extracts the moving golf ball from each of the images collected by each of the cameras 110 and 120, calculates the position information of the golf ball, and transmits the calculated information to the client 300 so that the client 300 can perform inherent functions such as calculating new information or calculating and analyzing information using the received position information of the golf ball.

[0049] For example, when the client 300 is implemented as a simulator used in a screen golf system, the position information of the golf ball and golf club can be received from the sensing processing unit 200 and used to realize a simulated image of the trajectory of the virtual golf ball flying on a virtual golf course.

[0050] Furthermore, when the client 300 is implemented as a golf swing analysis device, it can receive position information of the golf ball and golf club from the sensing and processing unit 200 and use it to provide analysis information on the user's golf swing, diagnosis of swing problems, and experience information for solving them.

[0051] The image processing unit 210 described above can be configured to perform image processing to extract a differential operation image relative to a reference image of each of the images successively acquired by the cameras 110 and 120, and the information calculation unit 220 can be configured to calculate the position information of the moving golf ball from each of the differential operation images extracted by the image processing unit described above.

[0052] As a method for extracting the motion of a golf ball from an acquired image, in addition to the method of using difference operations on the image as described above, a method can also be used that prepares a template image of a golf ball in advance and extracts the part corresponding to the golf ball from each acquired image by the similarity with the template image of the golf ball.

[0053] like Figure 1 The rotation calculation device of an embodiment of the present invention shown in (a) is a component that receives a ball image from one of the multiple cameras of the sensing device SD, calculates the rotation of the moving golf ball through the mark matching processing unit 510 and the rotation calculation unit 520, and transmits it to the client 300.

[0054] In contrast, Figure 1 The rotation calculation device of another embodiment of the present invention shown in (b) may be independently equipped with a single camera 600 for calculating the rotation of a golf ball, and may be configured to include a marker matching processing unit 510 and a rotation calculation unit 520 for calculating the rotation using ball images captured by the single camera 600.

[0055] In the above Figure 1In the embodiment shown in (a), one of the multiple cameras 110, 120 of the sensing device SD or Figure 1 In the embodiment shown in (b), the single camera 600 can be configured to move in response to the hitting of a golf ball marked with multiple markers, and acquire consecutive images, namely a first image and a second image.

[0056] For example, the image of the nth frame of a camera or a single camera can be called the first image, and the image of the (n+1)th frame can be called the second image.

[0057] The portion corresponding to the golf ball included in the first image is extracted as the first ball image, and the portion corresponding to the golf ball included in the second image is extracted as the second ball image, so that the label matching processing unit 510 and the rotation calculation unit 520 can calculate the rotation by analyzing the first ball image and the second ball image respectively.

[0058] The marker matching processing unit 510 can perform the following functions: find and specify multiple markers on each ball image, calculate information about a reference marker and multiple surrounding markers around the reference marker for the multiple markers on the first ball image, and find and specify a combination of markers corresponding to the reference marker and multiple surrounding markers of the first ball image in the second ball image.

[0059] The rotation calculation unit 520 can calculate the rotation axis and rotation amount by using the center points of each corresponding mark pair on the first sphere image and the second sphere image through geometric calculation.

[0060] The specific functions of the aforementioned mark matching processing unit and rotation calculation unit will be described later.

[0061] On the other hand, refer to Figure 2 The flowchart shown describes a method for calculating the spin of a golf ball that is struck and moved according to an embodiment of the present invention.

[0062] First, a series of images, including a first image of a golf ball marked with multiple marks and a second image following it, are acquired by the camera or a single camera of the sensing device (S110).

[0063] Using the images obtained as described above, the marker matching processing unit can extract and analyze multiple markers marked on the golf ball for rotation calculation from the images to determine specific combinations of corresponding markers between the golf ball in the first image and the golf ball in the second image. This can be achieved through... Figure 2 The steps S120 to S160 on the flowchart are executed.

[0064] After acquiring the first image and the second image in step S110, the portion of the first image corresponding to the golf ball is extracted as the first ball image, and the portion of the second image corresponding to the golf ball is extracted as the second ball image (S120).

[0065] The marker matching processing unit can define a reference marker and a plurality of peripheral markers for a plurality of markers on the first ball image (S130), and calculate information about the relative relationship between each of the plurality of peripheral markers and the reference marker on the first ball image.

[0066] The relative relationship of each of the plurality of peripheral marks relative to the reference mark can be defined according to the shape or phase of the plurality of marks marked on the golf ball. For example, when the plurality of marks marked on the golf ball have different phases from each other, that is, when they are tilted at different angles, the relative relationship of each of the plurality of peripheral marks relative to the reference mark can be calculated as the relative angle information of the tilt of each of the plurality of peripheral marks relative to the reference mark.

[0067] That is, the marker matching processing unit can define a reference marker and a plurality of peripheral markers for a plurality of markers on the first ball image (S130), and calculate the relative angle information of the tilt of each of the plurality of peripheral markers relative to the reference marker on the first ball image, namely "feature information" (S140).

[0068] The marker matching processing unit can calculate the relative angle information of the tilt in each combination of any reference marker on the second sphere image and a plurality of surrounding markers relative to it (S150), and specify the matched combination of markers on the second sphere image by matching the feature information of the markers on the first sphere image and the relative angle information of the markers on the second sphere image (S160).

[0069] Figures 3 to 6 This shows a specific example of matching the tags as described above.

[0070] Figure 3 This is an example of a ball image showing a golf ball with multiple markings on its surface and a diagram showing the markings in the ball image to specify the coordinates of their center points. This golf ball is used in a spin calculation device and spin calculation method according to an embodiment of the present invention.

[0071] Figure 4 It is shown as follows Figure 3 (b) shows the results of markings on a specific sphere image and calculating the tilt information for each mark.

[0072] Figure 5 and Figure 6This is a diagram used to illustrate the process of matching markers on the first ball image and markers on the second ball image.

[0073] First, refer to Figure 3 , Figure 3 (a) shows an image of a golf ball that has been photographed with multiple markings. Figure 3 (b) shows that... Figure 3 The results of identifying each marker on the sphere image shown in (a) by marking the region and calculating the coordinates of the center point of each marker are as follows. Figure 3 (c) and Figure 3 (d) shows Figure 3 The marks shown in (b) and the marked areas shown in an enlarged view.

[0074] like Figure 3 As shown in (a), the golf ball used in this invention has multiple markings on its surface, and multiple markings mk can be identified on the ball image BI.

[0075] The multiple markings on a golf ball can be, for example, Figure 3 (a) shows multiple marks with the same shape, or multiple marks with different shapes from each other.

[0076] In either case, it is preferable to use multiple markers as a reference to calculate relative information in relation to other markers in the surrounding area.

[0077] For example, such as Figure 3 As shown in (a), when multiple markers mk with the same shape are labeled as having different phases, i.e. tilts, the relative angle information of the tilt of each of the other markers can be calculated with a reference marker as a reference, and this can become the feature information for matching markers.

[0078] In such Figure 3 In the sphere image BI shown in (a), the marker matching processing unit can be as follows: Figure 3 As shown in (b), each mark mk is specified, and a mark region R is specified for each mark mk, i.e., the region of each mark mk, and the center point C of each mark mk is specified through each mark region R.

[0079] like Figure 3 As shown in (c), a marker region R1 can be generated in a manner that includes a tilted marker mk1, and the coordinates (cx1, cy1) of the center point of the marker region R1 can be calculated using the center point C1 of the marker region R1 as the center point of the marker.

[0080] exist Figure 3In case (d), the marker mk2 is shown horizontally, preferably with the marker region R2 having a predetermined size to include a background to some extent with the marker mk2, rather than setting the marker region R2 to be flush with the marker mk2.

[0081] exist Figure 3 In the case of (d), similarly, when the marked region R2 is specified, the coordinates (cx2, cy2) can be calculated with the center point C2 as the center point of the marked mk2.

[0082] In addition to the specific center point coordinates of each marker, the tilt information of each marker can also be calculated.

[0083] For the tilt angle of each marker, the principal component of the directional gradient formed by the pixels within the marker region can be extracted and defined as the marker's angle. Singular Value Decomposition (SVD) can be used to extract the principal component of the pixel gradient.

[0084] Figure 4 The image shows an example of the results of calculating the tilt angle of each of a specific set of markers in a sphere image as described above.

[0085] The steps for calculating the tilt angle information of each mark appearing on the sphere image are applicable to both the first and second sphere images.

[0086] On the other hand, refer to Figure 5 In the attached figure, reference numeral 610 indicates the first ball image, reference numeral 620 indicates the second ball image, and the golf ball shown in the first ball image rotates and moves to become the golf ball shown in the second ball image. At this time, it is necessary to match the same marks 710 on the first ball image 610 and the same marks 720 on the second ball image 620.

[0087] Regarding the matching of the markers as described above, the conventional approach is to set multiple markers as graphics with different shapes from each other, and to match the same markers on the first and second sphere images based on the similarity of the shapes of the graphics.

[0088] However, the problem is that in order to match the same markers by the similarity of their shapes, the camera used to acquire the ball image needs to have a very high resolution. Even with high resolution, there are many cases where the shapes of the markers appearing in the image are not exactly the same. Therefore, when matching markers by shape similarity, the accuracy decreases and there are limitations.

[0089] In this invention, since the markers are matched using information about the tilt of each of the aforementioned markers, rather than matching the markers based on the shape of the graphic, the speed of the operation that can be used for marker matching is very fast. At the same time, it also has the advantage of being able to match markers very accurately.

[0090] When matching the graphic shape of a marker, the computation speed is very slow because two-dimensional data needs to be processed. In contrast, when matching is performed using the tilt information of the marker according to the present invention, the tilt information is calculated as a single value, so the computation speed can be very fast from the perspective of matching using an equivalent tilt value.

[0091] In this regard, if we refer to Figure 6 More specifically, one of the multiple marks that will be displayed on the first sphere image 610 can be defined as a reference mark, and the marks surrounding the reference mark can be defined as peripheral marks.

[0092] For example, such as Figure 6 As shown, the marker 711 closest to the center point of the first sphere image 610 can be defined as the "reference marker", and the markers around the reference marker 711 can be defined in a clockwise direction as the first peripheral marker 712, the second peripheral marker 713, the third peripheral marker 714, the fourth peripheral marker 715, etc.

[0093] Thus, a combination of reference markers and peripheral markers on a second sphere image 620 is found that matches the reference marker 711 and peripheral markers 712-715 defined on the first sphere image 610.

[0094] Therefore, for the reference mark 711 and the surrounding marks 712-715 on the first sphere image 610, the relative angle information can be calculated using each tilt angle information as follows, which is referred to here as "feature information".

[0095] [Table 1]

[0096] <Feature Information>

[0097] Reference mark First perimeter marker Second perimeter markings Third perimeter markings Fourth perimeter marker tilt angle 2° 47° 137° 2° 92° relative angle 0° 45° 135° 0° 90°

[0098] That is, by finding the combination of reference markers and surrounding markers with relative angle information corresponding to the above-mentioned feature information for multiple markers on the second sphere image, marker matching can be performed.

[0099] Since the second ball image 620 is in a state rotated from the first ball image 610, it is impossible to directly identify the combination of the reference mark - peripheral mark on the first ball image 610, and all the marks on the second ball image 620 may become reference marks. Therefore, for all cases, the relative angle information of the reference mark - peripheral mark is calculated by determining any reference mark on the second ball image 620 and determining the peripheral marks based on this reference mark.

[0100] For example, as Figure 6 shown, in the case of S1, the relative angle information can be calculated by determining the mark 723 on the second ball image 621 as the reference mark and defining the remaining peripheral marks respectively; in the case of S2, the relative angle information can be calculated by determining the mark 721 on the second ball image 622 as the reference mark and defining the remaining peripheral marks respectively; in the case of S3, the relative angle information can be calculated by determining the mark 722 on the second ball image 623 as the reference mark and defining the remaining peripheral marks respectively. In this way, the relative angle information for each combination of the reference mark - peripheral mark when each of all the marks is used as the reference mark can be calculated.

[0101] Figure 6 The relative angles in the case of the combination of the reference mark 723 and the peripheral marks on the second ball image 621 in the case of S1 and the relative angles in the case of the combination of the reference mark 721 and the peripheral marks on the second ball image 622 in the case of S2 can be expressed as follows.

[0102] [Table 2]

[0103] <Case of S1>

[0104] Reference mark First perimeter marker Second perimeter markings Third perimeter markings Fourth perimeter marker Tilt angle 49° 94° 4° 4° 139° relative angle 0° 45° 135° 135° 90°

[0105] [Table 3]

[0106] <Case of S2>

[0107] Reference mark First perimeter marker Second perimeter markings Third perimeter markings Fourth perimeter marker Tilt angle 4° 49° 139° 4° 94° relative angle 0° 45° 135° 0° 90°

[0108] Comparing the above characteristic information and the relative angle information in the case of S1 above, it can be seen that they are inconsistent with each other, so the case of S1 is excluded.

[0109] Comparing the above characteristic information and the relative angle information in the case of S2, it can be seen that they are consistent with each other. From this, it can be known that the combination of marks on the second ball image that matches the characteristic information on the first ball image is the case of S2.

[0110] That is, in Figure 6In the case of S2, the change from the first ball image 610 to the second ball image 622 can be specifically due to the change caused by rotation, and the combination of reference mark and peripheral mark on the first ball image 610 can be matched with the combination of reference mark and peripheral mark on the second ball image 622.

[0111] As mentioned above Figures 3 to 6 The process shown can be achieved through Figure 2 The process is carried out according to steps S120 to S160 of the flowchart shown.

[0112] On the other hand, such as Figure 2 As shown, after the aforementioned mark matching is completed, the rotation calculation unit can use the matching results to calculate the rotation axis and rotation amount.

[0113] In order to calculate the rotation axis and the amount of rotation, preferably, firstly, the center points of the corresponding marker pairs on the first sphere image and the second sphere image are plotted on a sphere (S210).

[0114] Then, the lines connecting the center points of the corresponding marker pairs and the lines intersecting two or more planes derived from the center point of the sphere can be calculated as the axis of rotation (S220).

[0115] After calculating the rotation axis as described above, the vector perpendicular to the rotation axis from the center point of each corresponding mark pair is calculated (S230), and the vectors calculated as described above are projected onto the reference plane perpendicular to the rotation axis (S240).

[0116] The amount of rotation centered on the rotation axis is calculated using the vector projected onto the reference plane as described above (S250).

[0117] By calculating the rotation axis and rotation amount as described above, the rotation information from the first sphere image to the second sphere image can be calculated.

[0118] However, the change caused by the rotation from the first ball image to the second ball image is not solely due to the actual rotation of the golf ball. Therefore, the line of sight of the cameras capturing the first and second ball images is fixed, and they have moved from the position of the golf ball when capturing the first ball image to the position when capturing the second ball image, resulting in a rotational amount that occurs despite the non-rotational movement from the first ball image to the second ball image.

[0119] Thus, since the rotation amount corresponding to the non-rotational movement from the first ball image to the second ball image occurs due to the fixed line of sight of the camera, after calculating the line of sight rotation amount at each position in the image corresponding to the fixed line of sight of the camera and calculating the rotation amount as described above, the accurate rotation amount can be calculated by subtracting the calculated line of sight rotation amount from the calculated rotation amount (S260).

[0120] Figures 7 to 10 The diagram shows a specific example of the calculation of the rotation as described above.

[0121] Figure 7 It is shown as follows Figure 6 As described above, the matching of markers between the first and second sphere images will plot the center point of each matched marker on a sphere. Figure 8 It shows the use of Figure 7 A diagram illustrating an example of a method for calculating the axis of rotation from the center points of the marks on the sphere shown.

[0122] Figure 9 It shows the use of Figure 8 The diagram shows an example of a method for calculating rotation using the axis of rotation and the center points of each marked point in a sphere. Figure 10 It is a diagram used to illustrate the amount of rotation that occurs when the camera's line of sight moves without rotation from the first sphere image to the second sphere image while the camera's line of sight remains fixed.

[0123] First, refer to Figure 7 ,like Figure 7 As shown in (a), the markers 712 and 714 of the first sphere image 610 are matched with the markers 722 and 724 of the second sphere image 620, respectively, and can be as follows: Figure 7 As shown in (b), the center points (Ca1-Cb1, Ca2-Cb2) of the respective matching marks are represented on a sphere Ob.

[0124] See Figure 7 As shown in (b), the sphere Ob rotates from Ca1 to Cb1 and from Ca2 to Cb2. Here, the sphere Ob can be a virtually assumed sphere based on the contour of the sphere image.

[0125] Here, the rotation axis and rotation amount corresponding to the rotation from Ca1 to Cb1 may not be consistent with the rotation axis and rotation amount corresponding to the rotation amount and the rotation from Ca2 to Cb2.

[0126] Such inconsistencies may be caused by errors that occur during image analysis, such as the loss of some pixels representing the markers during pixel analysis.

[0127] Although Figure 7The image shows two pairs of marker center points, but it is not limited to this. Obviously, more pairs of marker center points can appear depending on the number of markers, such as three pairs, four pairs, etc.

[0128] like Figure 7 As shown in (b), the rotation axis can be calculated using each center point, given a pair of center points that match the markers on the sphere Ob. Figure 8 This has been illustrated.

[0129] exist Figure 8 (a) to Figure 8 In (c), the center point of sphere Ob is Co, and Ca1 and Cb1, as well as Ca2 and Cb2, are pairs of matching marker center points.

[0130] Assuming Figure 8 As shown in (a), the plane PL1 is perpendicular to the line L1 connecting the center points Ca1 and Cb1 of the markers and passes through the center point Co of the sphere Ob, and it is assumed that... Figure 8 As shown in (b), the plane PL2 is perpendicular to the straight line L2 connecting the center points Ca2 and Cb2 of the sphere Ob and passes through the center point Co of the sphere Ob.

[0131] Then, the intersection of planes PL1 and PL2 is determined, and it can be calculated as the axis of rotation SA.

[0132] In the same manner as above, for combinations of three or more pairs of marker center points, with each pair as a unit, the rotation axis is calculated for all combinations, and the results of multiple rotation axis calculations are used, for example, by averaging the values.

[0133] On the other hand, when the rotation axis is calculated as described above, vectors perpendicular to the rotation axis can be calculated from the center points of the corresponding mark pairs, and the calculated vectors can be projected onto a reference plane perpendicular to the rotation axis. The amount of rotation centered on the rotation axis can be calculated using the vectors projected onto the reference plane.

[0134] like Figure 9 As shown, a reference plane PO perpendicular to the rotation axis SA can be defined, and vectors perpendicular to the rotation axis SA can be calculated from the center points of the matching marker pairs marked on the sphere Ob, and the calculated vectors can be projected onto the reference plane PO.

[0135] like Figure 9As shown, the center point of the rotation axis SA projected onto the reference plane PO is point Cp. The vector perpendicular to the rotation axis SA at point Ca1 can be projected as vector v1 from point Pa1 on the reference plane PO to point Cp. The vector perpendicular to the rotation axis SA at point Cb1 can be projected as vector v2 from point Pb1 on the reference plane PO to point Cp. The vector perpendicular to the rotation axis SA at point Ca2 can be projected as vector v3 from point Pa2 on the reference plane PO to point Cp. The vector perpendicular to the rotation axis SA at point Cb2 can be projected as vector v4 from point Pb2 on the reference plane PO to point Cp.

[0136] Therefore, the rotation amount Q1 centered on the rotation axis Cp can be calculated using vectors v1 and v2 projected onto the reference plane PO, and the rotation amount Q2 centered on the rotation axis Cp can be calculated using vectors v3 and v4.

[0137] The value of rotation Q1 can be determined using the angle formed by vectors v1 and v2 centered at the center point Cp on the reference plane PO and the circumference of the circle, or by using the central angle of the triangle connecting Pa1-Pb1-Cp as the length of the arc from Pa1 to Pb1. Similarly, the value of rotation Q2 can be determined using the angle formed by vectors v3 and v4 centered at the center point Cp on the reference plane PO and the circumference of the circle, or by using the central angle of the triangle connecting Pa2-Pb2-Cp as the length of the arc from Pa2 to Pb2.

[0138] Thus, for multiple rotation values ​​Q1 and Q2, the average value can be taken as the final rotation value. When more rotation values ​​are calculated, statistical analysis of each rotation value, such as average and dispersion, standard deviation, etc., can be used to calculate the statistically accurate value of the rotation value.

[0139] However, as mentioned above, the rotation amount from the first sphere image to the second sphere image includes the rotation amount corresponding to the non-rotational movement of the first sphere image to the second sphere image as the camera's line of sight is fixed, i.e., the line of sight rotation amount that varies depending on the position within the image. Therefore, it is necessary to subtract this line of sight rotation amount from the previously calculated rotation amount.

[0140] It can be done Figure 10 Confirm the amount of rotation (line-of-sight rotation) that occurs as the object moves without rotation from the first sphere image to the second sphere image.

[0141] like Figure 10 As shown in (a), when a golf ball is fixed to one side of the clamp ZG, it will be as follows: Figure 10 As shown in (b), the golf ball, fixed to the other side of the clamp ZG, moves in a straight line, and an image of the ball can be acquired for each case to confirm this, although from... Figure 10(a) shows B1 direction Figure 10 As shown in (b), the second sphere image B2 moves without rotation, yet a rotation still occurs. That is, it can be known that a line-of-sight rotation occurs at each position within the image.

[0142] Therefore, by measuring and pre-setting the amount of rotation that occurs during a non-rotational movement from the first sphere image B1 to the second sphere image B2, and performing a rotation from the line of sight rotation, as shown in... Figures 7 to 9 The rotation amount calculated in the manner described above is subtracted from the correction of the pre-set line-of-sight rotation amount, and the final accurate rotation amount can be calculated.

[0143] This eye rotation amount can be pre-measured and pre-set to be taken into account during camera calibration, or the eye rotation amount described above can be calculated each time the user makes a golf shot to use it to correct the rotation amount.

[0144] As described above, the present invention marks multiple markers on a golf ball and quickly acquires images of a designated area using a camera. It then uses the relative relationships between the multiple markers in the acquired ball images to accurately match the markers between two consecutive ball images to calculate the spin, thus having the advantage of being able to calculate the spin faster and more accurately than existing technologies.

[0145] Industrial utilization potential

[0146] The method for calculating the spin of a golf ball that is struck and moved, and the spin calculation device thereof, disclosed in this application, can be applied to the fields of golf analysis or golf simulation systems, which are based on the analysis of a ball struck by a golf club during a golf swing.

Claims

1. A method for calculating the spin of a golf ball that has been struck and moved, characterized in that, include: The step of acquiring consecutive images, namely the first image and the second image, as a golf ball marked with multiple marks moves after being hit; The step of calculating information about a reference mark and a plurality of surrounding marks around the reference mark for a plurality of marks on a first ball image corresponding to a portion of the first image of the golf ball; The step of specifically combining a reference mark and multiple peripheral marks corresponding to the reference mark of the first ball image in the second ball image corresponding to the golf ball portion of the second image; as well as The steps of calculating the rotation axis and rotation amount using the center points of corresponding marker pairs on the first and second sphere images. The steps for combining the specific corresponding tags include: The step of calculating feature information as relative angular information of the tilt of each of the plurality of peripheral marks relative to the reference mark on the first sphere image; The steps of calculating relative angular information of the tilt in each combination of an arbitrary reference marker on the second sphere image and a plurality of surrounding markers relative to it; and The steps include matching the feature information of the markers on the first sphere image with the relative angle information of the markers on the second sphere image, and specifying the combination of the matched markers on the second sphere image.

2. A method for calculating the spin of a golf ball that has been struck and moved, characterized in that, include: The step of acquiring consecutive images, namely the first image and the second image, as a golf ball marked with multiple marks moves after being hit; The step of calculating information about a reference mark and a plurality of surrounding marks around the reference mark for a plurality of marks on a first ball image corresponding to a portion of the first image of the golf ball; The step of specifically combining a reference mark and multiple peripheral marks corresponding to the reference mark of the first ball image in the second ball image corresponding to the golf ball portion of the second image; as well as The steps of calculating the rotation axis and rotation amount using the center points of corresponding marker pairs on the first and second sphere images. The steps for calculating the rotation axis and the amount of rotation include: The step of plotting the center points of corresponding marker pairs on the first sphere image and the second sphere image onto a sphere; as well as The step of calculating the line connecting the center points of the corresponding pair of marks and the line intersecting two or more planes derived from the center point of the sphere as the axis of rotation.

3. The method for calculating the spin of a golf ball that has been struck and moved according to claim 2, characterized in that, The steps for calculating the rotation axis and the amount of rotation include: The steps include calculating the vector perpendicular to the rotation axis from each center point of the corresponding mark pair; The steps of projecting the calculated vectors onto a reference plane perpendicular to the rotation axis; and The step of calculating the amount of rotation centered on the rotation axis using the vectors used in the reference plane.

4. A method for calculating the spin of a golf ball that has been struck and moved, characterized in that, include: The step of acquiring consecutive images, namely the first image and the second image, as a golf ball marked with multiple marks moves after being hit; The step of calculating information about a reference mark and a plurality of surrounding marks around the reference mark for a plurality of marks on a first ball image corresponding to a portion of the first image of the golf ball; The step of specifically combining a reference mark and multiple peripheral marks corresponding to the reference mark of the first ball image in the second ball image corresponding to the golf ball portion of the second image; as well as The steps of calculating the rotation axis and rotation amount using the center points of corresponding marker pairs on the first and second sphere images. The method for calculating the spin of the golf ball that moves after being struck also includes: Based on the fixed line of sight of the camera between two consecutive images of the moving golf ball, calculate the amount of line-of-sight rotation at each position in the image. And a correction step is performed by subtracting the calculated line-of-sight rotation from the rotation amount calculated by the step of calculating the rotation axis and the rotation amount.

5. A device for calculating the spin of a golf ball that moves after being struck, characterized in that, include: A single camera is separately configured from the camera system of the sensing device that calculates the motion information of the golf ball that is hit and moves, so as to acquire consecutive images of the golf ball marked with multiple marks being hit and moving, namely the first image and the second image. The marker matching processing unit calculates information about a reference marker and multiple surrounding markers around the reference marker for multiple markers on a first ball image corresponding to the golf ball portion of the first image, and specifically matches the combination of markers corresponding to the reference marker and multiple surrounding markers of the first ball image in a second ball image corresponding to the golf ball portion of the second image; as well as The rotation calculation unit uses the center points of corresponding marker pairs on the first and second sphere images to calculate the rotation axis and rotation amount through geometric calculations. The marker matching processing unit is configured to: Calculate feature information as relative angular information of the tilt of each of the plurality of peripheral marks relative to the reference mark on the first sphere image; For each combination of an arbitrary reference marker on the second sphere image and a plurality of surrounding markers relative to it, calculate the relative angle information of the tilt in each combination; and The feature information of the markers on the first sphere image and the relative angle information of the markers on the second sphere image are matched, and the combination of the matched markers on the second sphere image is specified.

6. The spin calculation device for a golf ball that moves after being struck, as described in claim 5, is characterized in that, The marker matching processing unit is configured to: The sensor receives coordinate information of the placement of the golf ball, and based on this, sets the area of ​​the image to be acquired by the single camera as the region of interest; and The image of the defined region of interest is acquired as the first image and the second image.

7. A device for calculating the spin of a golf ball that moves upon being struck, characterized in that, include: The marker matching processing unit receives consecutive images of a golf ball being hit and moving, marked with multiple markers, from the camera of a sensing device that calculates motion information of a golf ball that has been hit and moved, namely a first image and a second image. For the multiple markers on the first ball image corresponding to the golf ball portion of the first image, it calculates information about a reference marker and multiple surrounding markers around the reference marker. In the second ball image corresponding to the golf ball portion of the second image, it specifies a combination of markers that correspond to the reference marker and multiple surrounding markers of the first ball image. as well as The rotation calculation unit uses the center points of corresponding marker pairs on the first and second sphere images to calculate the rotation axis and rotation amount through geometric calculations. The marker matching processing unit is configured to: Calculate feature information as relative angular information of the tilt of each of the plurality of peripheral marks relative to the reference mark on the first sphere image; For each combination of an arbitrary reference marker on the second sphere image and a plurality of surrounding markers relative to it, calculate the relative angle information of the tilt in each combination; and The feature information of the markers on the first sphere image and the relative angle information of the markers on the second sphere image are matched, and the combination of the matched markers on the second sphere image is specified.

8. The spin calculation device for a golf ball that moves upon being struck, as described in claim 5 or 7, is characterized in that... The rotation calculation unit is configured as follows: The center points of corresponding marker pairs on the first sphere image and the second sphere image are represented on a sphere; The line connecting the center points of the corresponding pair of marks and the line intersecting two or more planes derived from the center point of the sphere are calculated as the axis of rotation; Calculate the vector perpendicular to the rotation axis from each center point of the corresponding mark pair; and The calculated vectors are projected onto a reference plane perpendicular to the axis of rotation, so as to calculate the amount of rotation centered on the axis of rotation using the vectors used in the reference plane.

Images