Golf swing motion analysis method, analysis device and electronic equipment

Abstract

The invention is suitable for the technical field of motion analysis, and provides a golf swing action analysis method and device and electronic equipment. The method comprises the following steps: acquiring a golf swing video; performing key point detection on image frames in the golf swing video by using the key point detection model, determining a key point coordinate set corresponding to each image frame, and determining a key point frame sequence corresponding to the golf swing video; the key point coordinate set comprises coordinates of human body key parts in the image frame; the key point frame sequence is a sequence corresponding to the key point coordinate set; based on the key point frame sequence and a preset judgment condition, determining a key action frame sequence corresponding to the golf swing action; and according to the key point frame sequence and the key action frame sequence, determining a golf swing action score. According to the technical scheme provided by the invention, the accuracy of golf swing action analysis is improved.

Description

Technical Field

[0001] The present application relates to the technical field of motion analysis, and in particular to a golf swing motion analysis method, analysis device and electronic equipment. Background Art

[0002] For golf, the posture and movement of the player have a very important impact on the player's performance when playing golf. Sometimes a slight change in posture will affect the direction and strength of the ball. Many researchers have tried to study the swing of players when playing golf through human posture, because the naked eye cannot accurately observe whether the athlete's movements are standard in an instant.

[0003] Existing golf swing analysis methods include: using the time series network MLPFormer to process the image feature sequence of the golf swing video to obtain key frame images; using the human posture estimation method to obtain 2D / 3D human skeleton point coordinates for the key frame images, and comparing them with the action images, 3D skeleton models, and joint angles of professional players; using the Euclidean distance calculation with the 3D human skeleton coordinates of professional players to obtain the action standard score. However, the accuracy of the key frame obtained by processing the image feature sequence using MLPFormer is low, which makes the accuracy of the 2D / 3D human skeleton point coordinates obtained by processing the key frame using the posture estimation model low, resulting in low accuracy of golf swing analysis.

[0004] Therefore, how to improve the accuracy of golf swing analysis has become an urgent problem to be solved. Summary of the invention

[0005] The present application provides a golf swing analysis method, an analysis device, an electronic device, a computer-readable storage medium, and a computer program product, aiming to solve the problem of low accuracy in golf swing analysis.

[0006] In a first aspect, an embodiment of the present application provides a method for analyzing a golf swing, the method comprising: obtaining a golf swing video; performing key point detection on image frames in the golf swing video using a key point detection model, determining a set of key point coordinates corresponding to each of the image frames, and determining a key point frame sequence corresponding to the golf swing video; the key point coordinate set comprises coordinates of key parts of a human body in the image frames; the key point frame sequence is a sequence corresponding to the key point coordinate set; determining a key action frame sequence corresponding to the golf swing based on the key point frame sequence and preset judgment conditions; and determining a golf swing score based on the key point frame sequence and the key action frame sequence.

[0007] In a possible implementation, determining a key action frame sequence corresponding to a golf swing action based on the key point frame sequence and a preset judgment condition includes: calculating key parameters of each of the image frames in the golf swing video based on the key point frame sequence to obtain a key parameter sequence set; the key parameter sequence set includes key parameter sequences corresponding to a plurality of key parameters; the key parameters represent parameters related to the golf swing action; determining a swing action frame sequence in the golf swing video according to the key parameter sequence set; and determining a key action frame sequence corresponding to the golf swing action from the swing action frame sequence according to the preset judgment condition.

[0008] In a possible implementation, determining the swing action frame sequence in the golf swing video according to the key parameter sequence set includes: determining a first key part parameter sequence corresponding to a calculation window in a first key parameter sequence in the key parameter sequence set; determining a second key parameter sequence corresponding to the first key parameter sequence in the key parameter sequence set; determining a second key part parameter sequence corresponding to the calculation window in the second key parameter sequence; calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence; and determining the swing action frame sequence in the golf swing video according to the correlation and the similarity.

[0009] In a possible implementation, the key action frame sequence includes a preparatory posture frame sequence, a start swing action frame, a ball contact frame and a swing to apex frame; determining the key action frame sequence corresponding to the golf swing action from the swing action frame sequence according to the preset judgment condition includes: determining the preparatory posture frame sequence based on the key parameter frame sequence, the swing action frame sequence and a preset preparatory posture judgment condition in the preset judgment condition; determining the start swing action frame based on the key parameter frame sequence, the swing action frame sequence and a preset start swing judgment condition in the preset judgment condition; determining the ball contact frame based on the key parameter frame sequence, the swing action frame sequence and a preset ball contact moment judgment condition in the preset judgment condition; determining the swing to apex frame based on the key parameter frame sequence, the swing action frame sequence and a preset swing to apex judgment condition in the preset judgment condition.

[0010] In a possible implementation, after determining the golf swing action score based on the key point frame sequence and the key action frame sequence, the method further includes: determining a swing posture and / or a swing action based on the key parameter sequence set, and determining a hitting suggestion corresponding to the swing posture and / or the swing action.

[0011] In a possible implementation, the calculation of the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence includes: using the Pearson correlation coefficient calculation method to calculate the correlation between the first key part parameter sequence and the second key part parameter sequence; using a dynamic time warping algorithm to calculate the similarity between the first key part parameter sequence and the second key part parameter sequence.

[0012] In a possible implementation, the key action frame sequence includes a preparation posture frame sequence, and the golf swing action score includes a preparation posture score and a swing rhythm score; determining the golf swing action score according to the key point frame sequence and the key action frame sequence includes: determining the swing rhythm and the left elbow joint angle and the right elbow joint angle corresponding to the preparation posture frame sequence according to the key point frame sequence and the key action frame sequence; calculating the preparation posture score according to the left elbow joint angle and the right elbow joint angle corresponding to the preparation posture frame sequence; and calculating the swing rhythm score according to the swing rhythm.

[0013] In a second aspect, an embodiment of the present application provides a golf swing analysis device, the device comprising:

[0014] A video acquisition module, used to acquire golf swing videos;

[0015] A key point detection module, configured to perform key point detection on the image frames in the golf swing video using a key point detection model, determine a key point coordinate set corresponding to each of the image frames, and determine a key point frame sequence corresponding to the golf swing video; the key point coordinate set includes coordinates of key parts of the human body in the image frames; and the key point frame sequence is a sequence corresponding to the key point coordinate set;

[0016] A key action frame sequence determination module, used to determine a key action frame sequence corresponding to a golf swing action based on the key point frame sequence and a preset judgment condition;

[0017] The score determination module is used to determine the score of the golf swing action according to the key point frame sequence and the key action frame sequence.

[0018] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0020] In a fifth aspect, an embodiment of the present application provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the method described in the first aspect or any one of the implementation methods thereof.

[0021] Compared with the prior art, the embodiments of the present application have the following beneficial effects: using a key point detection model to perform key point detection on image frames in a golf swing video, obtaining a set of key point coordinates corresponding to each image frame, the key point coordinate set including coordinates of key parts of the human body in the corresponding image frame, determining a key point frame sequence corresponding to the golf swing video according to the key point coordinate set corresponding to each image frame; and then determining a key action frame sequence corresponding to the golf swing action based on the key point frame sequence and preset judgment conditions; compared with using a temporal network MLPFormer to directly process the image feature sequence of a golf swing video to obtain a key frame, due to the key point frame sequence corresponding to the golf swing video, the key point frame sequence corresponding to the golf swing action is more convenient than using a temporal network MLPFormer to directly process the image feature sequence of a golf swing video to obtain a key frame. The key point frame sequence includes a set of key point coordinates in each image frame of the golf swing video, and the key point coordinate set in each image frame includes the coordinates of key parts of the human body in each image frame. Therefore, the key action frame sequence is obtained by analyzing the key point frame sequence, which is equivalent to analyzing the coordinates of key parts of the human body in each image frame of the golf swing video to determine the key action frame sequence, so that the analysis process of determining the key action frame sequence is more accurate and detailed, which effectively avoids missing the detection of key action frames, improves the accuracy of determining the key action frame sequence, and further improves the accuracy of golf swing action analysis when performing golf swing action analysis based on the key action frame sequence.

[0022] It can be understood that the golf swing motion analysis device, electronic device, computer-readable storage medium and computer program product provided in the embodiments of the present application have the same beneficial effects as the above-mentioned golf swing motion analysis method, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1A flowchart of a golf swing analysis method provided in one embodiment of the present application;

[0025] Figure 2 A schematic diagram of a flow chart of determining a key action frame sequence corresponding to a golf swing action based on a key point frame sequence and a preset judgment condition provided in an embodiment of the present application;

[0026] Figure 3 A schematic diagram of a key parameter sequence set provided in an embodiment of the present application;

[0027] Figure 4 A schematic diagram of a key parameter sequence set for dividing key actions provided in an embodiment of the present application;

[0028] Figure 5 A schematic diagram of changes in joint angles during a golf swing provided by an embodiment of the present application;

[0029] Figure 6 A structural block diagram of a golf swing analysis device provided in one embodiment of the present application;

[0030] Figure 7 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0031] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0032] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0033] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0034] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0035] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0036] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0037] For ease of understanding, the technical solution of the present application will be described in detail below with reference to the accompanying drawings.

[0038] Figure 1 This is a flowchart of a golf swing analysis method provided by an embodiment of the present application. For ease of description, only the part related to the present embodiment is shown. The method provided by the present embodiment includes the following steps:

[0039] S110, obtaining a golf swing video.

[0040] Specifically, a video recording tool such as a camera is used to record a golf swing video, and the recorded golf swing video includes the complete process of the golfer from the preparation posture to the hitting of the ball.

[0041] In a specific implementation, a video recording tool such as a camera is used to record a golf swing video in the direction in which the golfer's body is facing. The golf swing video obtained from this shooting angle can clearly record every posture and movement of the golfer during the hitting process, ensuring the accuracy of subsequent calculation steps.

[0042] S120, use the key point detection model to perform key point detection on the image frames in the golf swing video, determine the key point coordinate set corresponding to each image frame, and determine the key point frame sequence corresponding to the golf swing video; the key point coordinate set includes the coordinates of the key parts of the human body in the image frame; the key point frame sequence is the sequence corresponding to the key point coordinate set.

[0043] Specifically, the key point detection model outputs the 3-dimensional (3D) key point coordinates of the human body through the input 2-dimensional (2D) image. The 3D key point coordinates refer to the coordinates of the key points on the X, Y, and Z axes, respectively. X and Y refer to the pixel coordinates of the key points on the 2D image, and Z refers to the depth information of the key points. The key part categories of the human body output by the key point detection model include at least the left shoulder, right shoulder, left elbow, right elbow, left wrist, right wrist, left hip joint, right hip joint, left knee, right knee, left ankle and right ankle, and may also include other key parts of the human body, which are not limited in this embodiment.

[0044] In a specific implementation, each image frame of the golf swing video is input into a key point detection model, and key point detection is performed on each image frame through the key point detection model to obtain the three-dimensional coordinates of key parts of the human body such as the left shoulder, right shoulder, left elbow, right elbow, left wrist, right wrist, left hip joint, right hip joint, left knee, right knee, left ankle, and right ankle in each image frame, and determine the key point coordinate set corresponding to each image frame. After key point detection is performed on each image frame of the golf swing video, a key point frame sequence consisting of the key point coordinate set of each image frame is obtained.

[0045] As an example, the coordinates of key parts of the human body are shown in Table 1:

[0046] Table 1

[0047] Left shoulder coordinates <![CDATA[left_shoulder=(x l_shoulder ,y l_shoulder ,z l_shoulder )]]> Right shoulder coordinates <![CDATA[righ_shoulder=(x r_shoulder ,y r_shoulder ,z r_shoulder )]]> Left elbow coordinates <![CDATA[left_elbow=(x l_elbow ,y l_elbow ,z l_elbow )]]> Right elbow coordinates <![CDATA[right_elbow=(x r_elbow ,y r_elbow ,z r_elbow )]]> Left wrist coordinates <![CDATA[left_wrist=(x l_wrist ,y l_wrist ,z l_wrist )]]> Right wrist coordinates <![CDATA[right_wrist=(x r_wrist ,y r_wrist ,z r_wrist )]]> Left hip joint coordinates <![CDATA[left_hip=(x l_hip ,y l_hip ,z l_hip )]]> Right hip joint coordinates <![CDATA[right_hip=(x r_hip ,y r_hip ,z r_hip )]]> Left knee coordinates <![CDATA[left_knee=(x l_knee ,y l_knee ,z l_knee )]]> Right knee coordinates <![CDATA[right_knee=(x r_knee ,y r_knee ,z r_knee )]]> Left ankle coordinates <![CDATA[left_ankle=(x l_ankle ,y l_ankle ,z l_ankle )]]> Right ankle coordinates <![CDATA[right_ankle=(x r_ankle ,y r_ankle ,z r_ankle )]]>

[0048] Exemplarily, based on the coordinates of the key parts of the human body in Table 1, the key point coordinate set corresponding to the i-th image frame in the image frame of the golf swing video is defined as:

[0049] frame i ={left_shoulder i ,right_shoulder i ,left_elbow i ,…,right_ankle i};

[0050] The key point frame sequence consisting of the key point set corresponding to each image frame in the golf swing video is defined as:

[0051] golf_video i ={frame1,frame2,…,frame i ,…,frame T};

[0052] Wherein, T represents the total number of image frames contained in the golf swing video.

[0053] S130, determining a key action frame sequence corresponding to a golf swing action based on the key point frame sequence and a preset judgment condition.

[0054] Specifically, the key action frame sequence corresponding to the golf swing action is a sequence of frames corresponding to the key actions in the golf swing action in the golf swing video. The key actions in the golf swing action include the backswing action and the downswing action. The backswing action includes the preparation posture, the middle of the backswing, the backswing to 3 / 4, and the backswing to the top; the downswing action includes the upswing and downswing transition and the moment of contact with the ball. The preset judgment condition includes the conditions that the key point coordinate set in the image frame corresponding to the key action must meet.

[0055] In a specific implementation, the key point frame sequence corresponding to the golf swing video is analyzed, the image frames corresponding to the key point coordinate sets in the key point frame sequence that meet the preset judgment conditions are determined as key action frames, and all key action frames are combined to determine a key action frame sequence corresponding to the golf swing action.

[0056] As an example, a deep learning model is used to determine a key action frame sequence corresponding to a golf swing action based on a key point frame sequence and preset judgment conditions. The deep learning model includes Long Short-Term Memory (LSTM), Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and Feed-forward Neural Networks (FNNs).

[0057] S140, determining a golf swing action score according to the key point frame sequence and the key action frame sequence.

[0058] Specifically, the golf swing action score is used to evaluate the standardization of the golf swing action. The higher the golf swing action score is, the more standard the golf swing action is.

[0059] As an example, according to the key point coordinate set of the image frame corresponding to any key action in the key action frame sequence, the score of the key action is calculated, and thus the score of the golf swing action is obtained.

[0060] The technical solution provided by this embodiment uses a key point detection model to perform key point detection on image frames in a golf swing video, obtains a set of key point coordinates corresponding to each image frame, and the key point coordinate set includes coordinates of key parts of the human body in the corresponding image frame. A key point frame sequence corresponding to the golf swing video is determined according to the key point coordinate set corresponding to each image frame; then a key action frame sequence corresponding to the golf swing action is determined based on the key point frame sequence and preset judgment conditions; compared with using a temporal network MLPFormer to directly process the image feature sequence of the golf swing video to obtain a key frame, since the key point frame sequence corresponding to the golf swing video The key point coordinate set in each image frame of the golf swing video is included, and the key point coordinate set in each image frame includes the coordinates of the key parts of the human body in each image frame. Therefore, the key action frame sequence is obtained by analyzing the key point frame sequence, which is equivalent to analyzing the coordinates of the key parts of the human body in each image frame of the golf swing video to determine the key action frame sequence, so that the analysis process of determining the key action frame sequence is more accurate and detailed, which effectively avoids missing the detection of key action frames, improves the accuracy of determining the key action frame sequence, and further improves the accuracy of golf swing action analysis when performing golf swing action analysis based on the key action frame sequence.

[0061] Figure 2 A flowchart of determining a key action frame sequence corresponding to a golf swing action based on a key point frame sequence and a preset judgment condition is provided in accordance with an embodiment of the present application. Figure 2 As shown, based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, based on the key point frame sequence and the preset judgment condition, the key action frame sequence corresponding to the golf swing action is determined, including:

[0062] S210, calculating key parameters of each image frame in the golf swing video based on the key point frame sequence to obtain a key parameter sequence set; the key parameter sequence set includes key parameter sequences corresponding to multiple key parameters respectively; the key parameters represent parameters related to the golf swing action.

[0063] Specifically, the key parameters include the left elbow joint angle, the right elbow joint angle, the left elbow Y-axis coordinate, the right elbow Y-axis coordinate, the left wrist Y-axis coordinate, the right wrist Y-axis coordinate, the left wrist and torso geometry values, the right wrist and torso geometry values, etc. This embodiment does not limit the specific types of key parameters.

[0064] In a specific implementation, the key parameters in each image frame of the golf swing video are first calculated according to the key point frame sequence, and then the corresponding key parameter sequence is determined for each key parameter, that is, the left elbow joint angle sequence, the right elbow joint angle sequence, the left elbow Y-axis coordinate sequence, the right elbow Y-axis coordinate sequence, the left wrist Y-axis coordinate sequence, the right wrist Y-axis coordinate sequence, the left wrist and torso geometry value sequence, and the right wrist and torso geometry value sequence are obtained; a key parameter sequence set is determined according to each key parameter sequence.

[0065] Correspondingly, such as Figure 3 As shown, the key parameter sequence set includes sequences corresponding to the left elbow joint angle, right elbow joint angle, left elbow Y-axis coordinate, right elbow Y-axis coordinate, left wrist Y-axis coordinate, right wrist Y-axis coordinate, left wrist and torso geometry values, and right wrist and torso geometry values ​​in each image frame of the golf swing video.

[0066] As an example, the left elbow joint angle refers to the angle formed by the left elbow as the vertex, and the left elbow to the left wrist and the left elbow to the left shoulder as the edges. The left elbow joint angle is calculated according to the left elbow joint angle calculation formula based on the key point frame sequence.

[0067] Exemplarily, the left elbow joint angle calculation formula includes:

[0068] dec_x_lwe=x l_wrist -x l_elbow ;

[0069] dec_x_lse=x l_shoulder -x l_elbow ;

[0070] dec_y_lwe=y l_wrist -y l_esbow ;

[0071] dec_y_lse=y s_shoulder -y l_elbow ;

[0072]

[0073] Among them, θ l_elbow_joint represents the left elbow joint angle; cos -1 Represents the arccosine function; dec_x_lwe represents the difference between the left wrist X-axis coordinate and the left elbow X-axis coordinate; dec_x_lse represents the difference between the left shoulder X-axis coordinate and the left elbow X-axis coordinate; dec_y_lwe represents the difference between the left wrist Y-axis coordinate and the left elbow Y-axis coordinate; dec_x_lse represents the difference between the left shoulder Y-axis coordinate and the left elbow Y-axis coordinate.

[0074] As an example, the right elbow joint angle refers to the angle formed by the right elbow as the vertex, and the right elbow to the right wrist and the right elbow to the right shoulder as the edges. The right hand joint angle is calculated according to the key point frame sequence using the right hand joint angle calculation formula.

[0075] Exemplarily, the right hand joint angle calculation formula includes:

[0076] dec_x_rwe=x r_wrist -x r_elbow ;

[0077] dec_x_rse=x r_shoulder -x r_elbow ;

[0078] dec_y_rwe=y r_wrist -y r_elbow ;

[0079] dec_y_rse=y r_shoulder -y r_elbow ;

[0080]

[0081] Among them, θ r_elbow_joint represents the right elbow joint angle; cos -1 Represents the arccosine function; dec_x_rwe represents the difference between the X-axis coordinate of the right wrist and the X-axis coordinate of the right elbow; dec_x_rse represents the difference between the X-axis coordinate of the right shoulder and the X-axis coordinate of the right elbow; dec_y_rwe represents the difference between the Y-axis coordinate of the right wrist and the Y-axis coordinate of the right elbow; dec_x_rse represents the difference between the Y-axis coordinate of the right shoulder and the Y-axis coordinate of the right elbow.

[0082] As an example, the torso polygon is composed of the left shoulder, left hip joint, left ankle, right ankle, right hip joint, and right shoulder. trunk , based on the torso polygon polygon trunk , use the left wrist and torso geometry calculation formula and the right wrist and torso geometry calculation formula to calculate the left wrist and torso geometry value and the right wrist and torso geometry value respectively.

[0083] Exemplarily, the left wrist and torso geometry calculation formula is:

[0084]

[0085] Exemplarily, the right wrist and torso geometry calculation formula is:

[0086]

[0087] S220: Determine a swing action frame sequence in the golf swing video according to the key parameter sequence set.

[0088] The swing action frame sequence refers to the frame sequence corresponding to the whole process of the golfer's swing action. In this step, the swing action frame sequence in the golf swing video is located according to the key parameter sequence set and the swing action feature.

[0089] S230, determining a key action frame sequence corresponding to the golf swing action from the swing action frame sequence according to a preset judgment condition.

[0090] The key action frame sequence refers to a sequence of frames corresponding to key actions in a golf swing action. Specifically, after determining the swing action frame sequence, the key action frame sequence is determined from the swing action frame sequence in combination with features corresponding to each key action in the golf swing action; specifically, image frames in the swing action frame sequence that meet preset judgment conditions corresponding to key actions are determined as key action frames, thereby obtaining the key action frame sequence.

[0091] The technical solution provided in this embodiment first determines a key parameter sequence set based on a key point frame sequence, then determines a swing action frame sequence in a golf swing video based on the key parameter sequence set, and finally determines a key action frame sequence corresponding to the golf swing action from the swing action frame sequence based on preset judgment conditions, thereby improving the accuracy of determining the key action frame sequence.

[0092] Based on the above embodiments, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, a swing action frame sequence in a golf swing video is determined according to a key parameter sequence set, including:

[0093] Determine a first key part parameter sequence corresponding to the calculation window in a first key parameter sequence of the key parameter sequence set;

[0094] Determine a second key parameter sequence corresponding to the first key parameter sequence in the key parameter sequence set;

[0095] Determining a second key part parameter sequence corresponding to the calculation window in the second key parameter sequence;

[0096] Calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence;

[0097] Determine the swing action frame sequence in the golf swing video based on correlation and similarity.

[0098] Specifically, the first key parameter sequence is any key parameter sequence in the key parameter sequence set, the first key part parameter sequence is a sequence segment in any calculation window in the first key parameter sequence; the key parameters corresponding to the second key parameter sequence and the key parameters corresponding to the first key parameter sequence are key parameters corresponding to the left and right sides of the same type of human body parts, for example, the first key part parameter sequence is the Y-axis coordinate sequence of the left wrist, and the corresponding second key parameter sequence is the Y-axis coordinate sequence of the right wrist; the second key part parameter sequence is a sequence segment in the second key parameter sequence in the same calculation window as the calculation window corresponding to the first key part parameter sequence. The swing action frame sequence is a sequence composed of image frames corresponding to the golf swing action.

[0099] In a specific implementation, a sliding window method is used to process the key parameter sequence. In each calculation window, the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence are calculated, and it is determined whether the calculated correlation and similarity are greater than a preset threshold. If both the correlation and the similarity are greater than the preset threshold, the first key part parameter sequence and the second key part parameter sequence corresponding to the calculation window are determined as the swing action sequence.

[0100] As an example, set the calculation window size to window_width frames and the overlap window size to overlap frames. The calculation method for the calculation window and the overlap window is as follows:

[0101] window_width=3×frame_rate;

[0102] overlap=[0.05×frame_rate];

[0103] Where frame_rate represents the frame rate of the golf swing video.

[0104] As an example, in any calculation window, the correlation and similarity between the left wrist and torso geometry value sequence and the right wrist and torso geometry value sequence, the correlation and similarity between the left wrist Y-axis coordinate sequence and the right wrist Y-axis coordinate sequence, and the correlation and similarity between the left elbow Y-axis coordinate sequence and the right elbow Y-axis coordinate sequence are calculated respectively. When all the correlations and similarities calculated above are greater than a preset threshold, the sequence corresponding to the calculation window among the left wrist and torso geometry value sequence, the right wrist and torso geometry value sequence, the left wrist Y-axis coordinate sequence, the right wrist Y-axis coordinate sequence, the left elbow Y-axis coordinate sequence, and the right elbow Y-axis coordinate sequence is determined as a swing action sequence. Exemplarily, the preset threshold is 0.95.

[0105] The technical solution provided in this embodiment determines the swing action frame sequence in the golf swing video by calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence, thereby improving the accuracy of determining the swing action frame sequence.

[0106] On the basis of the above-mentioned embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, the key action frame sequence includes a preparation posture frame sequence, a swing start frame, a ball contact frame, and a swing to top frame; according to a preset judgment condition, a key action frame sequence corresponding to a golf swing action is determined from the swing action frame sequence, including:

[0107] The preparatory posture frame sequence is determined based on the key parameter frame sequence, the swing action frame sequence and the preset preparatory posture judgment condition in the preset judgment condition.

[0108] Determine a swing start frame based on the key parameter frame sequence, the swing action frame sequence and a preset swing start judgment condition in the preset judgment condition;

[0109] Determine the ball touching frame based on the key parameter frame sequence, the swing action frame sequence and the preset ball touching instant judgment condition in the preset judgment condition;

[0110] Based on the key parameter frame sequence, the swing action frame sequence and the preset swing to top judgment condition in the preset judgment condition, the swing to top frame is determined.

[0111] Specifically, the preset judgment conditions include a preset preparation posture judgment condition, a preset backswing start judgment condition, a preset ball contact moment judgment condition, and a preset backswing to top judgment condition.

[0112] The preset preparatory posture judgment conditions include that the average angle of the left elbow joint angle and the average angle of the right hand joint angle are greater than the preset angle, the left wrist and trunk geometry value and the right wrist and trunk geometry value are equal to the first preset value; the variance of the normalized left elbow Y-axis coordinate sequence, the variance of the normalized right elbow Y-axis coordinate sequence, the variance of the normalized left wrist Y-axis coordinate sequence, and the variance of the normalized right wrist Y-axis coordinate sequence are less than the second preset value. Exemplarily, the preset angle is 170 degrees, the first preset value is 1, and the second preset value is 0.1.

[0113] The preset conditions for judging the start of the swing include: the preset time period to the left of the moment when the left wrist and trunk geometry value changes from 1 to 0, and the corresponding frame sequence is defined as left_wrist_trunk_left_0.5; the preset time period to the left of the moment when the right wrist and trunk geometry value changes from 1 to 0, and the corresponding frame sequence is defined as left_wrist_trunk_right_0.5; the frame corresponding to the maximum mutation point of the left elbow joint angle corresponding to the union of the sequence left_wrist_trunk_left_0.5 and the sequence left_wrist_trunk_right_0.5 is the start swing action frame. Exemplarily, the preset time period is 0.5 seconds.

[0114] The preset conditions for judging the moment of touching the ball include being located in the interval where the geometric values ​​of the left wrist and torso change from 0 to 1 and then to 0, and the corresponding frame sequence is defined as left_wrist_trunk_010; being located in the interval where the geometric values ​​of the right wrist and torso change from 0 to 1 and then to 0, and the corresponding frame sequence is defined as right_wrist_trunk_010; the frame corresponding to the maximum value point of the Y-axis coordinate of the right wrist corresponding to the union of the sequence left_wrist_trunk_010 and the sequence right_wrist_trunk_010 is the ball touching frame.

[0115] The preset swing-to-apex judgment conditions include being between the swing-start frame and the ball-touching frame, and the corresponding frame sequence is defined as address_to_impact; the right wrist Y-axis coordinate sequence corresponding to address_to_impact is between two minimum minimum points, and the corresponding frame sequence is defined as inter_min_extremal; the frame corresponding to the maximum value point of the right wrist Y-axis coordinate sequence corresponding to inter_min_extremal is the swing-to-apex frame.

[0116] In a specific implementation, the preparation posture frame sequence, the start swing action frame, the ball touch frame and the swing to the top frame are determined in the swing action frame sequence according to the preset preparation posture judgment condition, the preset start swing judgment condition, the preset ball touch moment judgment condition and the preset swing to the top judgment condition, and the swing action and the downswing action are divided in the swing action frame sequence through the determined key action frames. Figure 4 As shown in FIG. 1 , the three vertical dotted lines represent the beginning of the swing, the swing to the top and the moment of contact with the ball. The frame between the beginning of the swing and the frame between the swing to the top is the swing action, and the frame between the swing to the top and the frame between the swing to the top is the downswing action.

[0117] The technical solution provided in this embodiment is based on the key parameter frame sequence. By presetting the preparatory posture judgment condition, the presetting start swing judgment condition, the presetting ball contact moment judgment condition and the presetting swing to apex judgment condition, the presetting preparatory posture frame sequence, the start swing action frame, the ball contact frame and the swing to apex frame are determined in the swing action frame sequence, thereby improving the accuracy of determining the key action frame sequence.

[0118] Based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, after determining the golf swing action score according to the key point frame sequence and the key action frame sequence, it also includes:

[0119] A swing posture and / or a swing action is determined according to the set of key parameter sequences, and a shot suggestion corresponding to the swing posture and / or the swing action is determined.

[0120] Specifically, the swing posture represents a static action of a golfer in any image frame in a golf video, and the swing action represents a continuous swing action of the golfer corresponding to multiple image frames.

[0121] In a specific implementation, the backswing time, downswing time, swing rhythm, joint angles, shoulder rotation angles, hip rotation angles and other calculation parameters are first calculated based on the key parameter sequence set, and then the swing posture and / or swing action are determined based on the obtained calculation parameters. Combined with the batting suggestions given in different situations shown in Table 2, the batting suggestions corresponding to the swing posture and the batting suggestions corresponding to the swing action are determined.

[0122] Table 2

[0123]

[0124] As an example, the backswing time calculation formula is:

[0125]

[0126] Among them, backswing_time represents the backswing time; frame top Indicates the frame number from the swing up to the top frame; frame address Indicates the frame number of the start swing frame; frame__rate indicates the frame rate of the golf swing video.

[0127] As an example, the downswing time calculation formula is:

[0128]

[0129] Among them, downswing_time represents the downswing time; frame top Indicates the frame number from the swing up to the top frame; frameimpact Indicates the frame number of the ball-touching frame; frame__rate indicates the frame rate of the golf swing video.

[0130] As an example, the formula for calculating swing tempo is:

[0131]

[0132] Among them, swing_tempo represents the swing rhythm; backswing_time represents the backswing time; downswing_time represents the downswing time.

[0133] As an example, based on the above-mentioned left elbow joint angle calculation formula and the right elbow joint angle calculation formula, the left shoulder joint angle calculation formula, the left hip joint angle calculation formula, the left knee joint angle calculation formula, the right shoulder joint angle calculation formula, the right hip joint angle calculation formula, and the right knee joint angle calculation formula can be inferred, and the left shoulder joint angle, left hip joint angle, left knee joint angle, right shoulder joint angle, right hip joint angle, and right knee joint angle can be calculated.

[0134] Among them, the left shoulder joint angle refers to the angle with the left shoulder as the vertex, and the left shoulder to the left elbow, and the left shoulder to the right shoulder as the edges; the left hip joint angle refers to the angle with the left hip as the vertex, and the left hip to the left knee, and the left hip to the left shoulder as the edges; the left knee joint angle refers to the angle with the left knee as the vertex, and the left knee to the left hip, and the left knee to the left ankle as the edges; the right shoulder joint angle refers to the angle with the right shoulder as the vertex, and the right shoulder to the right elbow, and the right shoulder to the left shoulder as the edges; the right hip joint angle refers to the angle with the right hip as the vertex, and the right hip to the right knee, and the right hip to the right shoulder as the edges; the right knee joint angle refers to the angle with the right knee as the vertex, and the right knee to the right hip, and the right knee to the right ankle as the edges.

[0135] Exemplarily, the changes in the joint angles during the swing are as follows: Figure 5 shown.

[0136] As an example, the shoulder rotation angle calculation formula is:

[0137]

[0138] Among them, shoulder_turn represents the shoulder rotation angle, z r_shoulder Indicates the Z-axis coordinate of the right shoulder; l_shoulder Indicates the Z-axis coordinate of the left shoulder; x r_shoilder Indicates the X-axis coordinate of the right shoulder; l_shoulder Indicates the X-axis coordinate of the left shoulder.

[0139] As an example, the hip rotation angle calculation formula is:

[0140]

[0141] Where hip_turn represents the hip rotation angle; z r_hip Represents the Z-axis coordinate of the right hip joint; z l_hip Indicates the Z-axis coordinate of the left hip joint; x r_hip Represents the X-axis coordinate of the right hip joint; x l_hip Represents the X-axis coordinate of the left hip joint.

[0142] The technical solution provided in this embodiment can provide corresponding batting suggestions based on the swing posture and / or swing action determined by the key parameter sequence set, thereby improving the user experience.

[0143] On the basis of the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence includes:

[0144] The correlation between the first key part parameter sequence and the second key part parameter sequence is calculated using the Pearson correlation coefficient calculation method;

[0145] A dynamic time warping algorithm is used to calculate the similarity between the first key part parameter sequence and the second key part parameter sequence.

[0146] Among them, the Pearson correlation coefficient calculation method is a statistic used to measure the degree of linear correlation between two variables. The value range of the Pearson correlation coefficient is between -1 and 1, where 1 represents a perfect positive correlation, that is, when one variable increases, the other variable also increases; -1 represents a perfect negative correlation, that is, when one variable increases, the other variable decreases; 0 represents no linear correlation.

[0147] The technical solution provided in this embodiment uses the Pearson correlation coefficient calculation method to perform correlation calculation, which can accurately and conveniently determine the correlation between the first key part parameter sequence and the second key part parameter sequence.

[0148] Among them, Dynamic Time Warping (DTW) is an algorithm used to measure the similarity of two time series, especially for those sequences of different lengths or that may differ in time. The implementation of the dynamic time warping algorithm usually involves constructing a cost matrix and finding the best matching path between the two sequences through dynamic programming. This path minimizes the cumulative distance between the two sequences, thus providing an effective way to compare their similarity.

[0149] The technical solution provided in this embodiment uses a dynamic time warping algorithm to calculate the similarity, which can accurately and conveniently determine the similarity between the first key part parameter sequence and the second key part parameter sequence.

[0150] Based on the above embodiments, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, the key action frame sequence includes a preparation posture frame sequence, and the golf swing action score includes a preparation posture score and a swing rhythm score; determining the golf swing action score according to the key point frame sequence and the key action frame sequence includes:

[0151] According to the key point frame sequence and the key action frame sequence, determine the left elbow joint angle and the right elbow joint angle corresponding to the swing rhythm and the preparation posture frame sequence;

[0152] Calculate the preparatory posture score according to the left elbow joint angle and the right elbow joint angle corresponding to the preparatory posture frame sequence;

[0153] Based on the swing tempo, a swing tempo score is calculated.

[0154] In a specific implementation, according to the key point frame sequence, based on the above-mentioned left elbow joint angle calculation formula and right elbow joint angle calculation formula, the left elbow joint angle and the right elbow joint angle corresponding to the preparatory posture frame sequence are calculated, and then the preparatory posture score is calculated according to the preparatory posture score calculation formula; based on the swing rhythm calculated in the above embodiment, the swing rhythm score is calculated according to the swing rhythm score.

[0155] As an example, the formula for calculating the ready posture score is:

[0156]

[0157] Among them, address_score represents the score of the preparation posture; left_elbow_anger represents the left elbow joint angle corresponding to the preparation posture frame sequence; right_elbow_anger represents the right elbow joint angle corresponding to the preparation posture frame sequence.

[0158] As an example, the formula for calculating the swing tempo score is:

[0159] swing_tempo_score=40×swing_tempo-20;

[0160] Among them, swing_tempo_score represents the swing tempo score; swing_tempo represents the swing tempo.

[0161] The technical solution provided in this embodiment determines the left elbow joint angle and the right elbow joint angle corresponding to the swing rhythm and the preparation posture frame sequence according to the key point frame sequence and the key action frame sequence; and further calculates the preparation posture score and the swing rhythm score based on this, thereby improving the accuracy of calculating the golf swing action score.

[0162] Figure 6 This is a structural block diagram of a golf swing analysis device provided by an embodiment of the present application. For the sake of convenience, only the parts related to the embodiment of the present application are shown. Figure 6 The golf swing action analysis device 600 may include a video acquisition module 601 , a key point detection module 602 , a key action frame sequence determination module 603 and a score determination module 604 .

[0163] The video acquisition module 601 is used to acquire a golf swing video.

[0164] The key point detection module 602 is used to use the key point detection model to perform key point detection on the image frames in the golf swing video, determine the key point coordinate set corresponding to each image frame, and determine the key point frame sequence corresponding to the golf swing video; the key point coordinate set includes the coordinates of the key parts of the human body in the image frame; the key point frame sequence is the sequence corresponding to the key point coordinate set.

[0165] The key action frame sequence determination module 603 is used to determine the key action frame sequence corresponding to the golf swing action based on the key point frame sequence and the preset judgment condition.

[0166] The score determination module 604 is used to determine the score of the golf swing action according to the key point frame sequence and the key action frame sequence.

[0167] A golf swing motion analysis device provided in an embodiment of the present application has the same beneficial effects as the above-mentioned golf swing motion analysis method.

[0168] In one embodiment, the key action frame sequence determination module 603 includes:

[0169] The key parameter calculation submodule is used to calculate the key parameters of each image frame in the golf swing video based on the key point frame sequence to obtain a key parameter sequence set; the key parameter sequence set includes key parameter sequences corresponding to multiple key parameters respectively; the key parameters represent parameters related to the golf swing action;

[0170] A swing action frame sequence determination submodule, used to determine a swing action frame sequence in a golf swing video according to a key parameter sequence set;

[0171] The key action frame sequence determination submodule is used to determine the key action frame sequence corresponding to the golf swing action from the swing action frame sequence according to a preset judgment condition.

[0172] In one embodiment, the swing action frame sequence determination submodule includes:

[0173] A first key part parameter sequence determining unit, used to determine a first key part parameter sequence corresponding to the calculation window in a first key parameter sequence of the key parameter sequence set;

[0174] A second key parameter sequence determining unit, configured to determine a second key parameter sequence corresponding to the first key parameter sequence in the key parameter sequence set;

[0175] A second key part parameter sequence determining unit, used to determine a second key part parameter sequence corresponding to the calculation window in the second key parameter sequence;

[0176] A calculation unit, used for calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence;

[0177] The swing action frame sequence determination unit is used to determine the swing action frame sequence in the golf swing video according to the correlation and similarity.

[0178] In one embodiment, the key action frame sequence determination submodule includes:

[0179] A first determining unit is used to determine a preparatory posture frame sequence based on the key parameter frame sequence, the swing action frame sequence and a preset preparatory posture judgment condition in the preset judgment condition;

[0180] A second determination unit is used to determine a swing start action frame based on the key parameter frame sequence, the swing action frame sequence and a preset swing start judgment condition in the preset judgment condition;

[0181] A third determination unit is used to determine the ball touching frame based on the key parameter frame sequence, the swing action frame sequence and the preset ball touching instant judgment condition in the preset judgment condition;

[0182] The fourth determination unit is used to determine the backswing to top frame based on the key parameter frame sequence, the swing action frame sequence and the preset backswing to top judgment condition in the preset judgment condition.

[0183] In one embodiment, the golf swing analysis device 600 further includes:

[0184] The batting suggestion determination module is used to determine the swing posture and / or the swing action according to the key parameter sequence set, and determine the batting suggestion corresponding to the swing posture and / or the swing action.

[0185] In one embodiment, the computing unit includes:

[0186] A first calculation subunit is used to calculate the correlation between the first key part parameter sequence and the second key part parameter sequence by using a Pearson correlation coefficient calculation method;

[0187] The second calculation subunit is used to calculate the similarity between the first key part parameter sequence and the second key part parameter sequence by using a dynamic time warping algorithm.

[0188] In one embodiment, the score determination module 604 includes:

[0189] A calculation parameter determination submodule is used to determine the left elbow joint angle and the right elbow joint angle corresponding to the swing rhythm and the preparation posture frame sequence according to the key point frame sequence and the key action frame sequence;

[0190] A preparatory posture score calculation submodule, used to calculate the preparatory posture score according to the left elbow joint angle and the right elbow joint angle corresponding to the preparatory posture frame sequence;

[0191] The swing rhythm score calculation submodule is used to calculate the swing rhythm score according to the swing rhythm.

[0192] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0193] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0194] Figure 7 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 7 As shown, the electronic device 7 of this embodiment includes: at least one processor 70 ( Figure 7 Only one is shown in the figure), a memory 71, and a computer program 72 stored in the memory 71 and executable on at least one processor 70, the processor 70 executes the computer program 72 to implement the above Figure 1 or Figure 2 The steps in the method embodiment, or the implementation of the above Figure 6 Functions of each module / unit in the device embodiment.

[0195] The electronic device 7 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device 7 may include but is not limited to a processor 70 and a memory 71. Those skilled in the art will appreciate that Figure 7 It is only an example of the electronic device 7 and does not constitute a limitation on the electronic device 7. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0196] The processor 70 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0197] In some embodiments, the memory 71 may be an internal storage unit of the electronic device 7, such as a hard disk or memory of the electronic device 7. In other embodiments, the memory 71 may also be an external storage device of the electronic device 7, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 7. Further, the memory 71 may also include both an internal storage unit and an external storage device of the electronic device 7. The memory 71 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as program codes of a computer program. The memory 71 may also be used to temporarily store data that has been output or is to be output.

[0198] The embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0199] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include at least: any entity or device capable of carrying the computer program code to an electronic device, a recording medium, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), an electric carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0200] A computer-readable storage medium provided in an embodiment of the present application has the same beneficial effects as the above-mentioned golf swing action analysis method.

[0201] An embodiment of the present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0202] A computer program product provided in an embodiment of the present application has the same beneficial effects as the above-mentioned golf swing action analysis method.

[0203] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0204] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0205] In the embodiments provided in the present application, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0206] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0207] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A golf swing analysis method, characterized in that: The method comprises: Get golf swing videos; Using a key point detection model to perform key point detection on the image frames in the golf swing video, determine a key point coordinate set corresponding to each of the image frames, and determine a key point frame sequence corresponding to the golf swing video; the key point coordinate set includes coordinates of key parts of the human body in the image frames; the key point frame sequence is a sequence corresponding to the key point coordinate set; Determining a key action frame sequence corresponding to a golf swing action based on the key point frame sequence and a preset judgment condition; A golf swing action score is determined according to the key point frame sequence and the key action frame sequence.

2. The method according to claim 1, characterized in that: The determining of a key action frame sequence corresponding to a golf swing action based on the key point frame sequence and a preset judgment condition comprises: Calculating key parameters of each image frame in the golf swing video based on the key point frame sequence to obtain a key parameter sequence set; the key parameter sequence set includes key parameter sequences corresponding to a plurality of key parameters; the key parameters represent parameters related to the golf swing action; Determine a swing action frame sequence in the golf swing video according to the key parameter sequence set; According to the preset judgment condition, a key action frame sequence corresponding to the golf swing action is determined from the swing action frame sequence.

3. The method according to claim 2, characterized in that The step of determining a swing action frame sequence in the golf swing video according to the key parameter sequence set includes: Determining a first key part parameter sequence corresponding to the calculation window in a first key parameter sequence of the key parameter sequence set; Determining a second key parameter sequence corresponding to the first key parameter sequence in the key parameter sequence set; Determining a second key part parameter sequence corresponding to the calculation window in the second key parameter sequence; Calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence; A swing action frame sequence in the golf swing video is determined according to the correlation and the similarity.

4. The method according to claim 2, characterized in that: The key action frame sequence includes a preparation posture frame sequence, a swing start action frame, a ball touch frame, and a swing to top frame; The step of determining a key action frame sequence corresponding to the golf swing action from the swing action frame sequence according to the preset judgment condition comprises: Determining the preparatory posture frame sequence based on the key parameter frame sequence, the swing action frame sequence and a preset preparatory posture judgment condition in the preset judgment conditions; Determining the swing start frame based on the key parameter frame sequence, the swing action frame sequence and a preset swing start judgment condition in the preset judgment conditions; Determining the ball-touching frame based on the key parameter frame sequence, the swing action frame sequence and a preset ball-touching instant judgment condition in the preset judgment conditions; The swing-to-apex frame is determined based on the key parameter frame sequence, the swing action frame sequence and a preset swing-to-apex judgment condition in the preset judgment condition.

5. The method according to claim 2, characterized in that: After determining the golf swing action score according to the key point frame sequence and the key action frame sequence, the method further includes: A swing posture and / or a swing action is determined according to the key parameter sequence set, and a hitting suggestion corresponding to the swing posture and / or the swing action is determined.

6. The method according to claim 3, characterized in that The calculating the correlation and similarity between the first key part parameter sequence and the second key part parameter sequence comprises: The correlation between the first key part parameter sequence and the second key part parameter sequence is calculated using a Pearson correlation coefficient calculation method; A dynamic time warping algorithm is used to calculate the similarity between the first key part parameter sequence and the second key part parameter sequence.

7. The method according to any one of claims 1 to 6, characterized in that: The key action frame sequence includes a preparation posture frame sequence, and the golf swing action score includes a preparation posture score and a swing rhythm score; Determining a golf swing action score according to the key point frame sequence and the key action frame sequence includes: Determine the left elbow joint angle and the right elbow joint angle corresponding to the swing rhythm and the preparatory posture frame sequence according to the key point frame sequence and the key action frame sequence; Calculating the preparatory posture score according to the left elbow joint angle and the right elbow joint angle corresponding to the preparatory posture frame sequence; The swing rhythm score is calculated according to the swing rhythm.

8. A golf swing analysis device, characterized in that: The device comprises: A video acquisition module, used to acquire golf swing videos; A key point detection module, configured to perform key point detection on the image frames in the golf swing video using a key point detection model, determine a key point coordinate set corresponding to each of the image frames, and determine a key point frame sequence corresponding to the golf swing video; the key point coordinate set includes coordinates of key parts of the human body in the image frames; and the key point frame sequence is a sequence corresponding to the key point coordinate set; A key action frame sequence determination module, used to determine a key action frame sequence corresponding to a golf swing action based on the key point frame sequence and a preset judgment condition; The score determination module is used to determine the score of the golf swing action according to the key point frame sequence and the key action frame sequence.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer program product, characterized in that The invention comprises a computer program which, when executed, causes the method according to any one of claims 1 to 7 to be performed.

Images