Archery posture evaluation method, device, edge computing server and storage medium

Abstract

The embodiment of the present invention provides an archery posture evaluation method, device, edge computing server and storage medium, the method comprising: obtaining a first video through a selected image acquisition device, and judging whether a user enters a designated position area according to the first video; when the user enters the archery starting area, obtaining a second video through an image acquisition device, identifying the position of the skeletal joint points of the key parts of the user in the second image for each frame of the second image in the second video; determining the posture data corresponding to the second image according to the position of the skeletal joint points of the key parts of the user; selecting target posture data from the posture data of the user in the archery process, and comparing it with the preset standard posture data to evaluate the user's archery posture. In this way, the standard degree of the archery posture can be objectively and accurately determined, and then the user can be guided to correct the posture, which greatly facilitates the user's training.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of image processing technology, and in particular, to an archery posture evaluation method, device, edge computing server and storage medium. Background Art

[0002] During archery competitions, archery performance will be affected by many factors, such as the wind speed of the venue, the standardization of the archers' movements, the athletes' psychological state, etc.

[0003] The archery training process of athletes requires long-term one-on-one coaching and correction by coaches, and the teaching cost is very high. Although some auxiliary training tools have appeared in related technologies, such tools cannot give intuitive and vivid training evaluation and guidance results. For example, the method of installing sensors on the bow to collect the displacement parameters and jitter parameters of the bow through the sensors is used for the athletes or coaches to refer to after the athletes' training. This method cannot intuitively and quickly evaluate the posture of archery athletes during the movement process. Summary of the invention

[0004] In order to solve the technical problem that the archery motion evaluation method in the above-mentioned prior art cannot intuitively and quickly evaluate the motion process of archery athletes, the embodiments of the present invention provide an archery posture evaluation method, device, edge computing server and storage medium.

[0005] In a first aspect of an embodiment of the present invention, a method for evaluating an archery posture is first provided, the method comprising:

[0006] Acquire a first video through a selected image acquisition device, and determine whether the user enters a designated location area based on the first video;

[0007] If yes, triggering an archery posture detection event for the user;

[0008] Acquire a second video through the image acquisition device, and identify the positions of the skeletal joint points of the key parts of the user in each frame of the second image in the second video;

[0009] Determine the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user, and form the posture data of the user during the archery process by the posture data corresponding to each of the second images in the second video;

[0010] Target posture data is selected from the posture data of the user during the archery process, and compared with preset standard posture data to evaluate the archery posture of the user.

[0011] In an optional implementation, the candidate image acquisition devices are respectively deployed on the left and right sides with the bull's eye as the central axis, and the candidate image acquisition devices are located on the same horizontal line as the user;

[0012] The step of acquiring the first video through the selected image acquisition device includes:

[0013] Acquiring the archery habits of the user, and determining an image acquisition device from the candidate image acquisition devices according to the archery habits of the user;

[0014] The image acquisition device is started to perform video acquisition, and a first video is obtained through the image acquisition device.

[0015] In an optional implementation, acquiring the first video through an image acquisition device includes:

[0016] In the archery detection mode, acquiring a first video through an image acquisition device;

[0017] The determining, based on the first video, whether the user enters a designated location area includes:

[0018] For a first image in the first video, input the first image into a preset human position detection model to obtain a user position in the first image output by the human position detection model;

[0019] The user position is compared with a preset archery standing position area to determine whether the user enters the designated position area.

[0020] In an optional implementation, the human body position detection model is obtained in the following manner:

[0021] Acquire archery videos corresponding to multiple target users, and perform frame extraction processing on the archery videos to obtain a preset number of archery images;

[0022] Using a rectangular frame marking method to mark human targets on the preset number of archery images to generate human detection training samples;

[0023] Based on the human body detection training samples, supervised training is performed on the initial model for human body position detection to obtain a human body position detection model.

[0024] In an optional implementation, the identifying the positions of the skeletal joints of the key parts of the user in the second image includes:

[0025] The second image is input into a preset human posture assessment model to obtain the positions of the skeletal joint points of the key parts of the user in the second image output by the human posture assessment model.

[0026] In an optional embodiment, the human body posture assessment model is obtained specifically in the following manner:

[0027] Acquire archery videos corresponding to multiple target users, and perform frame extraction processing on the archery videos to obtain a preset number of archery images;

[0028] Using a skeleton joint point marking method to mark human joint points on the preset number of archery images to generate archery posture evaluation training samples;

[0029] Based on the archery posture evaluation training samples, supervised training is performed on the initial model for human posture evaluation to obtain a human posture evaluation model.

[0030] In an optional implementation, determining the posture data corresponding to the second image according to the skeletal joint point positions of the key parts of the user includes:

[0031] According to the positions of the skeletal joints of the key parts of the user, forming vectors corresponding to the key parts of the user;

[0032] The angle between the vectors corresponding to the key parts of the user is calculated, and the angle is determined to be the angle corresponding to the key parts of the user corresponding to the second image.

[0033] In an optional embodiment, the selecting target posture data from the posture data of the user during the archery process and comparing the target posture data with preset standard posture data to evaluate the archery posture of the user includes:

[0034] Determine the similarity between any adjacent N frames of the second image in the second video;

[0035] Determine, from the second video, N adjacent frames of the second image whose similarity exceeds a preset similarity threshold; wherein N≥2;

[0036] Selecting a target second image from the adjacent N frames of the second image whose similarity exceeds a preset similarity threshold, and selecting an angle corresponding to the target second image as a target angle;

[0037] The target angle is compared with a preset standard angle to evaluate the user's archery posture.

[0038] In an optional embodiment, the motion sensor is mounted on the bow, and the method further comprises:

[0039] Acquire the motion data of the bow during the user's archery shooting currently collected by the motion sensor, and draw the current moving track of the bow according to the motion data;

[0040] Determine a collection time of the motion data, and search the second video for the second image corresponding to the collection time;

[0041] The current movement trajectory of the bow is marked on the second image corresponding to the acquisition moment and transmitted to the display screen for display.

[0042] In an optional embodiment, the motion sensor is mounted on the bow, and the method further comprises:

[0043] Acquire the complete motion data of the bow collected by the motion sensor during the user's archery process, and draw a complete movement trajectory of the bow according to the complete motion data;

[0044] The complete movement trajectory of the bow is marked on the second image of the target and transmitted to a display screen for display.

[0045] In a second aspect of an embodiment of the present invention, a device for evaluating an archery posture is provided, the device comprising:

[0046] A video acquisition module, used for acquiring a first video through a selected image acquisition device;

[0047] A user determination module, used to determine whether the user enters a designated location area according to the first video;

[0048] an event triggering module, configured to trigger an archery posture detection event for the user if yes;

[0049] A position recognition module, used to obtain a second video through the image acquisition device, and for each frame of a second image in the second video, identify the positions of the skeletal joints of the key parts of the user in the second image;

[0050] a data determination module, configured to determine the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user, and to form the posture data of the user during the archery shooting process from the posture data corresponding to each of the second images in the second video;

[0051] The posture evaluation module is used to select target posture data from the posture data of the user during the archery process, and compare it with preset standard posture data to evaluate the user's archery posture.

[0052] In a third aspect of an embodiment of the present invention, an edge computing server is further provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0053] Memory, used to store computer programs;

[0054] The processor is used to implement the archery posture evaluation method described in the first aspect when executing the program stored in the memory.

[0055] In a fourth aspect of the embodiments of the present invention, a storage medium is further provided, wherein instructions are stored in the storage medium, and when the instructions are executed on a computer, the computer executes the archery posture evaluation method described in the first aspect above.

[0056] In a fifth aspect of the embodiments of the present invention, a computer program product comprising instructions is also provided, which, when executed on a computer, enables the computer to execute the archery posture evaluation method described in the first aspect above.

[0057] The technical solution provided by the embodiment of the present invention obtains the first video through the selected image acquisition device, and determines whether the user enters the designated position area according to the first video. When the user enters the designated position area, the archery posture detection event of the user is triggered, and the second video is obtained through the image acquisition device. For each frame of the second image in the second video, the skeletal joint point position of the key part of the user in the second image is identified, and the posture data corresponding to the second image is determined according to the skeletal joint point position of the key part of the user. The posture data corresponding to each second image in the second video constitutes the posture data of the user in the archery process, and the target posture data is selected from the posture data of the user in the archery process, and compared with the preset standard posture data to evaluate the user's archery posture. In this way, video acquisition and analysis are performed during the archery process to realize the detection of the user's archery posture during the archery process, the user's archery posture is quantified through data, and the obtained target posture data is compared with the standard posture data, so that the standard degree of the archery posture can be determined intuitively and accurately, and then the user is guided to correct the posture, which greatly facilitates the user's training. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0060] Figure 1 It is a schematic diagram of an implementation flow of an archery posture evaluation method shown in an embodiment of the present invention;

[0061] Figure 2 A schematic diagram of a scenario for deploying a candidate image acquisition device shown in an embodiment of the present invention;

[0062] Figure 3 It is a schematic diagram of an implementation process of a training method for a human body position detection model shown in an embodiment of the present invention;

[0063] Figure 4 A schematic diagram of an implementation flow of a training method for a human posture assessment model shown in an embodiment of the present invention;

[0064] Figure 5 A schematic diagram of a bone joint point marking method shown in an embodiment of the present invention;

[0065] Figure 6 It is a schematic diagram of an implementation flow of a method for determining posture data corresponding to a second image shown in an embodiment of the present invention;

[0066] Figure 7 It is a schematic diagram of the architecture of a system for evaluating the archery posture of a user shown in an embodiment of the present invention;

[0067] Figure 8 It is a schematic diagram of an implementation flow of a method for evaluating a user's archery posture shown in an embodiment of the present invention;

[0068] Fig. 9 A second target image in a process of a user aiming at the bull's eye shown in an embodiment of the present invention;

[0069] Fig.10 It is a structural schematic diagram of an archery posture evaluation device shown in an embodiment of the present invention;

[0070] Fig.11 A schematic diagram of the structure of an edge computing server shown in an embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0073] like Figure 1 As shown, it is a schematic diagram of an implementation process of an archery posture evaluation method provided by an embodiment of the present invention. The method is applied to an edge computing server and may specifically include the following steps:

[0074] S101, obtaining a first video through a selected image acquisition device, and determining whether a user enters a designated location area based on the first video.

[0075] In an embodiment of the present invention, candidate image acquisition devices, such as cameras, etc., can be deployed in a (sports) venue to capture real-time video in the venue. Specifically, they can be deployed in an archery venue to capture real-time video in the venue; wherein the arrow shooting position area and the arrow shooting direction are marked in the background (i.e., in the screen of the image acquisition device).

[0076] It should be noted that, for the deployment of candidate image acquisition devices, the bull's eye can be used as the central axis, and the candidate image acquisition devices can be deployed on the left and right sides respectively, and the candidate image acquisition devices and the user are located at the same horizontal line, such as Figure 2 As shown, the viewing angle of the candidate image acquisition device may correspond to the left side or the right side when the user is shooting arrows, and this embodiment of the present invention is not limited to this.

[0077] Based on the deployment of the above-mentioned candidate image acquisition devices, for the edge computing server, the user's archery habits can be obtained, and then the image acquisition device can be determined from the above-mentioned candidate image acquisition devices according to the user's archery habits, and then the image acquisition device can be started to perform video acquisition, and the first video can be obtained through the image acquisition device.

[0078] It should be noted that the user's archery habits are usually understood as the user's dominant hand for pulling the bowstring. For example, if the user is left-handed, the user is accustomed to pulling the bowstring with the left hand. For example, if the user is right-handed, the user is accustomed to pulling the bowstring with the right hand. This means that in practice, different candidate image acquisition devices are started according to the user's different archery habits, and only one of the image acquisition devices is actually used.

[0079] For example, for an edge computing server, the user's archery habits are obtained (the user is left-handed and is accustomed to pulling the bowstring with his left hand). From the archery direction, the candidate image acquisition device on the left can be determined as the image acquisition device based on the user's archery habits, and then the image acquisition device can be started to perform video acquisition, and the first video can be obtained through the image acquisition device.

[0080] For example, for an edge computing server, the user's archery habits are obtained (the user is right-handed and is accustomed to pulling the bowstring with his right hand). From the archery direction, the candidate image acquisition device on the right can be determined as the image acquisition device based on the user's archery habits, and then the image acquisition device can be started to perform video acquisition, and the first video can be obtained through the image acquisition device.

[0081] After the image acquisition device is started to acquire video, in the archery detection mode, during the user's archery, the image acquisition device can capture the archery picture in real time, and transmit the captured picture to the edge computing server. Thus, in the archery detection mode, the edge computing server acquires the first video through the image acquisition device.

[0082] For the first video, it is possible to determine whether the user has entered the designated location area based on the first video. For example, in the archery detection mode, for the first video, it is possible to determine whether the student has entered the designated location area based on the first video. The designated location area here can be as follows: Figure 2 shown.

[0083] Among them, for the first image in the first video, the first image is input into a preset human position detection model to obtain the user position in the first image output by the human position detection model, and the user position is compared with the preset archery standing position area to determine whether the user enters the designated position area.

[0084] It should be noted that the human body position detection model may specifically be a PeopleNet model, and of course may also be other models, which is not limited in the embodiment of the present invention.

[0085] like Figure 3 FIG. 1 is a schematic diagram of an implementation flow of a training method for a human position detection model provided by an embodiment of the present invention. The method may specifically include the following steps:

[0086] S301, obtaining archery videos corresponding to a plurality of target users, and performing frame extraction processing on the archery videos to obtain a preset number of archery images.

[0087] In the embodiment of the present invention, during the model training phase, multiple target users are gathered, and archery videos are collected for each target user, so that the archery videos corresponding to the multiple target users can be obtained.

[0088] It should be noted that, in the process of collecting archery videos for each target user, an image acquisition device is determined from the above-mentioned candidate image acquisition devices according to the archery habits of each target user, and then the image acquisition device is started to perform video acquisition, and the archery video of each target user is obtained through the image acquisition device.

[0089] For example, during the model training phase, 4,000 students are recruited, including 2,000 males and 2,000 females, to obtain the archery habits of each student. An image acquisition device is determined from the candidate image acquisition devices mentioned above based on the archery habits of each student, and then the image acquisition device is started to perform video acquisition. The archery video of each student is obtained through the image acquisition device. In this way, the archery videos of 4,000 students can be obtained.

[0090] For the acquired archery videos corresponding to the multiple target users, a frame extraction process is performed in the embodiment of the present invention, so that a preset number of archery images can be obtained. The frame extraction process can be random, which is not limited in the embodiment of the present invention.

[0091] For example, for the acquired archery videos corresponding to 4,000 students, frame extraction is performed in the embodiment of the present invention, wherein a script may be used to randomly extract frames, so that approximately 50,000 archery images may be obtained.

[0092] S302, annotating the preset number of archery images with human targets by using a rectangular frame annotation method to generate human detection training samples.

[0093] For a preset number of archery images, in an embodiment of the present invention, a rectangular frame annotation method can be used to annotate human targets, so that human detection training samples can be generated, which means that for each archery image, a rectangular frame is used to frame the person in the archery image.

[0094] S303: Based on the human body detection training samples, supervised training is performed on the initial model for human body position detection to obtain a human body position detection model.

[0095] For human body detection training samples, in an embodiment of the present invention, supervised training can be performed on an initial model for human body position detection based on the human body detection training samples to obtain a human body position detection model.

[0096] It should be noted that, in the embodiment of the present invention, when the loss function converges or the number of iterations reaches a threshold, the model training can be regarded as terminated, and the embodiment of the present invention does not limit this.

[0097] S102: If yes, triggering an archery posture detection event for the user.

[0098] S103, acquiring a second video through the image acquisition device, and identifying positions of skeletal joints of key parts of the user in each frame of a second image in the second video.

[0099] S104, determining the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user, and forming the posture data of the user during the archery process from the posture data corresponding to each of the second images in the second video.

[0100] In an embodiment of the present invention, the image acquisition device can be a real-time video capture device of the entire archery process of the user, and transmit the video to the edge computing server for processing. When the user enters the archery starting area, the edge computing server triggers an archery posture detection event for the user, which means that the detection of the user's archery posture is started at this time. At this time, the user can be tracked by the target tracking algorithm, and the image acquisition device captures the motion picture in real time, and the edge computing server obtains the second video.

[0101] It should be noted that, in the embodiment of the present invention, the first video and the second video may be sub-segments in the complete video code stream of archery posture evaluation. The complete video code stream of archery posture evaluation is constituted by the first video and the second video. Here, the archery posture evaluation code stream refers to the video code stream collected from the entire process of the user entering the (camera) screen, then entering the archery standing position area, and then starting to shoot arrows from the archery standing position area and ending the archery, and then exiting the (camera) screen.

[0102] In an embodiment of the present invention, for each frame of the second image in the second video, the positions of the skeletal joints of the key parts of the user in the second image are identified, and the posture data corresponding to the second image is determined based on the positions of the skeletal joints of the key parts of the user. Here, the posture data corresponding to each second image in the second video constitutes the posture data of the user during archery.

[0103] For example, in an embodiment of the present invention, for the i-th (i=1, 2, 3, 4...) second image in the second video, the positions of the skeletal joints of the key parts of the user in the i-th second image are identified, and the posture data corresponding to the i-th second image is determined based on the positions of the skeletal joints of the key parts of the user, as shown in Table 1 below. The posture data corresponding to each second image in the second video thus constitutes the posture data of the user during archery.

[0104] The second image of the i-th frame Posture data Frame 1 Second Image Posture data 1 Frame 2 Second Image Posture data 2 …… ……

[0105] Table 1

[0106] It should be noted that the key parts in the embodiment of the present invention specifically refer to the shoulder joint, elbow joint, chin joint and / or ankle joint, etc., which can be selected according to actual needs, and the embodiment of the present invention does not limit this. In addition, the skeletal joint point positions of the key parts of the user can be as follows:

[0107] {P i (x i ,y i , c i )}, i∈0,...,21; where (x i ,y i ) represents the horizontal and vertical coordinates of the bone joint points in the image, c i It represents the confidence level, which represents the credibility of the position of this bone joint point. If it is low, it can be discarded and the position of the bone joint point with higher confidence level can be retained.

[0108] Among them, in an embodiment of the present invention, the positions of the skeletal joint points of the key parts of the user in the second image can be identified through the human posture evaluation model. Specifically, the second image can be input into a preset human posture evaluation model to obtain the positions of the skeletal joint points of the key parts of the user in the second image output by the human posture evaluation model.

[0109] It should be noted that the human body posture assessment model may specifically be a YOLO model, and of course may also be other models, which is not limited in the embodiment of the present invention.

[0110] like Figure 4 FIG. 1 is a schematic diagram of an implementation flow of a training method for a human posture assessment model provided by an embodiment of the present invention. The method may specifically include the following steps:

[0111] S401, obtaining archery videos corresponding to a plurality of target users, and performing frame extraction processing on the archery videos to obtain a preset number of archery images.

[0112] In the embodiment of the present invention, during the model training phase, multiple target users are gathered, and archery videos are collected for each target user, so that the archery videos corresponding to the multiple target users can be obtained.

[0113] It should be noted that, in the process of collecting archery videos for each target user, an image acquisition device is determined from the above-mentioned candidate image acquisition devices according to the archery habits of each target user, and then the image acquisition device is started to perform video acquisition, and the archery video of each target user is obtained through the image acquisition device.

[0114] For example, during the model training phase, 4,000 students are recruited, including 2,000 males and 2,000 females, to obtain the archery habits of each student. An image acquisition device is determined from the candidate image acquisition devices mentioned above based on the archery habits of each student, and then the image acquisition device is started to perform video acquisition. The archery video of each student is obtained through the image acquisition device. In this way, the archery videos of 4,000 students can be obtained.

[0115] For the acquired archery videos corresponding to the multiple target users, a frame extraction process is performed in the embodiment of the present invention, so that a preset number of archery images can be obtained. The frame extraction process can be random, which is not limited in the embodiment of the present invention.

[0116] For example, for the acquired archery videos corresponding to 4,000 students, frame extraction is performed in the embodiment of the present invention, wherein a script may be used to randomly extract frames, and roughly about 50,000 archery images may be obtained.

[0117] S402, annotating the human body joints of the preset number of archery images by using a skeleton joint annotation method to generate archery posture evaluation training samples.

[0118] For a preset number of archery images, in an embodiment of the present invention, a skeleton joint point labeling method can be used to label the human body joint points, so that archery posture evaluation training samples can be generated, which means that for each archery image, the skeleton joint points therein are labeled.

[0119] For example, for 50,000 standing long jump images, the human body joints are labeled using the skeleton joints labeling method, and a total of 22 skeleton joints are labeled, such as Figure 5 As shown, the meaning of each bone joint point is shown in Table 2 below.

[0120]

[0121]

[0122] Table 2

[0123] In an embodiment of the present invention, in order to improve the speed of computing and processing while meeting the detection requirements, during the above-mentioned training process, the training sample images are annotated with human body joints, or only the 16 skeletal joints including the above-mentioned upper torso and lower torso are annotated, or only the 8 skeletal joints in the upper torso are annotated, and this is not limited to the embodiment of the present invention.

[0124] S403, based on the archery posture evaluation training samples, supervised training is performed on the human body posture evaluation initial model to obtain a human body posture evaluation model.

[0125] For the archery posture evaluation training samples, in the embodiment of the present invention, supervised training can be performed on the human body posture evaluation initial model based on the archery posture evaluation training samples to obtain the human body posture evaluation model.

[0126] It should be noted that, in the embodiment of the present invention, when the loss function converges or the number of iterations reaches a threshold, the model training can be regarded as terminated, and the embodiment of the present invention does not limit this.

[0127] In addition, in the embodiment of the present invention, if Figure 6 FIG. 1 is a schematic diagram of an implementation flow of a method for determining posture data corresponding to a second image provided by an embodiment of the present invention, which may specifically include the following steps:

[0128] S601, composing vectors corresponding to the key parts of the user according to the positions of the skeletal joints of the key parts of the user.

[0129] S602, calculating the angle between the vectors corresponding to the key parts of the user, and determining the angle as the angle corresponding to the key parts of the user corresponding to the second image.

[0130] Regarding the positions of the skeletal joint points of the key parts of the user in the second image, in an embodiment of the present invention, vectors corresponding to the key parts of the user can be composed based on the positions of the skeletal joint points of the key parts of the user, and the angle between the vectors corresponding to the key parts of the user can be calculated, and the angle can be determined to be the angle corresponding to the key parts of the user corresponding to the second image.

[0131] For example, for the skeletal joint points P7, P9, P 13 , which constitutes the vectors corresponding to the key parts of the user, that is, the two vectors corresponding to the left shoulder joint of the user, namely refer to Figure 5 As well as the meaning shown in the above Table 2, the angle between the two vectors is calculated, and the angle is determined to be the angle corresponding to the left shoulder joint of the user corresponding to the second image, which is recorded as angle_shoulder.

[0132] Similarly, for the positions of the remaining skeletal joint points of the key parts of the user in the second image, two vectors corresponding to the user's left waist joint, two vectors corresponding to the left knee joint, and two vectors corresponding to the left ankle joint are formed, and then the angle between the two vectors is calculated to determine the angle corresponding to the left waist joint, the angle corresponding to the left knee joint, and the angle corresponding to the left ankle joint of the user corresponding to the second image, which are recorded as angle_waist, angle_knee, and angle_ankle, respectively, as shown in Table 3 below.

[0133] In this way, the angles corresponding to the above joints of the user corresponding to the second image can be determined.

[0134]

[0135] Table 3

[0136] S105, selecting target posture data from the posture data of the user during the archery process, and comparing the target posture data with preset standard posture data to evaluate the archery posture of the user.

[0137] In an embodiment of the present invention, the posture data corresponding to each second image in the second video constitutes the posture data of the user during the archery process, and the target posture data is selected from the posture data of the user during the archery process, that is, the target posture data is selected from the posture data corresponding to each second image, and compared with the preset standard posture data to evaluate the user's archery posture, that is, the corresponding sports expert knowledge base, evaluates the archery posture problem, and gives training guidance opinions in the corresponding expert database.

[0138] Among them, for the target posture data (such as the right shoulder joint angle, the right knee joint angle, etc.), the target posture data is compared with the preset standard posture data (such as the standard right shoulder joint angle, the right knee joint angle, etc.), to detect whether the user's archery posture is standard, or to give the standard degree, and if the user's archery posture is not standard, the user's archery posture is played back on the display screen and the non-standard position is marked; Figure 7 As shown, the edge computing server is connected to the camera and the display screen respectively.

[0139] In addition, if Figure 8 As shown, it is a schematic diagram of an implementation flow of a method for evaluating a user's archery posture provided by an embodiment of the present invention. The method is applied to an edge computing server and may specifically include the following steps:

[0140] S801: Determine the similarity between any adjacent N frames of the second image in the second video.

[0141] S802, determining, from the second video, N adjacent frames of the second image whose similarity exceeds a preset similarity threshold; wherein N≥2.

[0142] S803: Select a target second image from the adjacent N frames of the second image whose similarity exceeds a preset similarity threshold, and select an angle corresponding to the target second image as a target angle.

[0143] In an embodiment of the present invention, for the second video, the similarity between any adjacent N frames of second images in the second video is determined, so as to determine from the second video adjacent N frames of second images whose similarity exceeds a preset similarity threshold, where N≥2, and a target second image is selected from the adjacent N frames of second images whose similarity exceeds the preset similarity threshold, and the angle corresponding to the target second image is selected as the target angle.

[0144] In an embodiment of the present invention, any two adjacent second image frames in the second video are taken as an example for explanation. For any two adjacent second image frames in the second video, the similarity between the any two adjacent second image frames is determined, so that the target angle can be selected from the angles corresponding to each second image in the second video based on the similarity.

[0145] For example, for the edge computing server, for any two adjacent frames of the second image in the second video, the similarity between any two adjacent frames of the second image is determined, as shown in Table 4 below, so that the target angle can be selected from the angles corresponding to each second image in the second video based on the similarity.

[0146] Any two frames of the second image Similarity The first frame second image and the second frame second image 85％ The second image of the second frame and the second image of the third frame 80％ …… ……

[0147] Table 4

[0148] It should be noted that the similarity between any two frames of the second image may be calculated by referring to a relatively mature algorithm on the market, such as a background difference method, and the embodiment of the present invention does not limit this.

[0149] Among them, during the archery process, the user usually needs to combine his own breathing frequency and natural factors, aim at the center of the bull's eye for a period of time, and then shoot the arrow at the center of the bull's eye. During this period, the process of the user aiming at the center of the bull's eye is more important, and the archery posture at this time can be focused on.

[0150] Based on this, in an embodiment of the present invention, the edge computing server can determine the two adjacent frames of second images whose similarity exceeds a preset similarity threshold from the second video, select the target second image from the two adjacent frames of second images whose similarity exceeds the preset similarity threshold, and select the angle corresponding to the target second image as the target angle.

[0151] For example, the edge computing server determines from the second video that the two frames of the second image have a similarity of more than 95%, which means that the user is now in a state of aiming at the bull's eye. At this time, the user needs to maintain aiming at the bull's eye for a period of time (for example, 1 to 2 seconds). Correspondingly, there are many adjacent frames of the second image, and their similarities are all more than 95%.

[0152] At this time, a second image can be randomly selected from a large number of adjacent second images with a similarity of more than 95% as the target second image, such as Fig. 9 As shown, the second image here corresponds to the process of the user aiming at the bull's eye, and is a frame image in the process of the user aiming at the bull's eye. At this time, the angle corresponding to the target second image is selected as the target angle.

[0153] It should be noted that, given that it takes a certain amount of time for the user to aim at the bull's eye, for example 1 to 2 seconds, the process of the user aiming at the bull's eye is a continuous process, and the number of the two adjacent frames of second images whose corresponding similarity exceeds the preset similarity threshold is relatively large, then a second image is randomly selected from the two adjacent frames of second images whose similarity exceeds the preset similarity threshold as the target second image, and the embodiments of the present invention are not limited to this.

[0154] S804: Compare the target angle with a preset standard angle to evaluate the user's archery posture.

[0155] In the embodiment of the present invention, the angle corresponding to the target second image is used as the target angle, and the target angle is compared with the preset standard angle to evaluate the user's archery posture. This means that the angle corresponding to the key part of the user corresponding to a certain frame image (i.e., the target second image) during the user's aiming at the bull's eye is compared with the preset standard angle to evaluate the user's archery posture.

[0156] In an embodiment of the present invention, the edge computing server may, after obtaining the posture data of the user during archery, mark the posture data on the second image and present the marked second image on a display screen.

[0157] In an embodiment of the present invention, the above-mentioned edge computing server can also mark the user's posture data on the target second image. For example, the angle corresponding to the target second image is the target angle. In an embodiment of the present invention, the target angle can be marked on the target second image, so that it can be displayed on the display screen after marking.

[0158] In addition, in an embodiment of the present invention, a motion sensor (such as an accelerometer, a gyroscope, etc.) is installed on the bow, specifically, it can be installed on the anti-vibration rod of the bow. The motion sensor is used to collect the motion data of the bow during archery (posture direction, acceleration, etc.) in real time, and send the motion data to the edge computing server.

[0159] Based on this, the edge computing server can obtain the motion data of the bow currently collected by the motion sensor during the user's archery process, and draw the current movement trajectory of the bow based on the motion data. The current movement trajectory of the bow can be marked in a specific image and transmitted to the display screen for display for the user to watch.

[0160] Specifically, the collection time of the motion data is determined, and the second image corresponding to the collection time is searched from the second video, which means that the collection time of the second image here is consistent with the collection time of the motion data, indicating that the motion data and the corresponding second image are collected at the same time, and the current moving trajectory of the bow is marked on the second image corresponding to the collection time, and transmitted to the display screen for display.

[0161] For example, for the edge computing server, the motion data of the bow collected by the motion sensor at the current time (12:00) during the user's archery process is obtained, and the current moving trajectory of the bow is drawn according to the motion data, the collection time of the motion data (12:00) is determined, and the second image corresponding to the collection time is searched from the second video, which means that at the moment of 12:00, both the motion data and the corresponding second image are collected. At this time, the current moving trajectory of the bow can be marked on the second image corresponding to the collection time and transmitted to the display screen for display.

[0162] For the edge computing server, the complete movement trajectory of the bow can be obtained in the end. Specifically, the complete motion data of the bow collected by the motion sensor during the user's archery process is obtained, and the complete movement trajectory of the bow is drawn according to the complete motion data. The complete movement trajectory of the bow is marked on the second image of the target and transmitted to the display screen for display for users to watch.

[0163] In the embodiment of the present invention, a camera is used to capture the athlete's sports video in real time, and the athlete's key body parts are identified in the video frame image, and the posture data of the key parts are calculated. Then, the user's sports movements are quantified, and the posture data and the movement curve of the bow are simultaneously presented on the video image, which is convenient for intuitively evaluating the standardization of the athlete's movements.

[0164] It should be noted that, for the motion data or the complete motion data, reference may be made to the existing motion data processing method, so as to obtain the current motion trajectory or the complete motion trajectory, which is not limited in the embodiment of the present invention.

[0165] Through the above description of the technical solution provided by the embodiment of the present invention, a first video is acquired through a selected image acquisition device, and based on the first video, it is determined whether the user enters a designated position area, and when the user enters the position area where he stands when shooting arrows, an archery posture detection event for the user is triggered, and a second video is acquired through the image acquisition device, and for each frame of the second image in the second video, the positions of the skeletal joints of the key parts of the user in the second image are identified, and based on the positions of the skeletal joints of the key parts of the user, the posture data corresponding to the second image is determined, and the posture data of the user during the archery process is composed of the posture data corresponding to each second image in the second video, and the target posture data is selected from the posture data of the user during the archery process, and compared with the preset standard posture data to evaluate the user's archery posture.

[0166] At the same time, the motion curve of the bow is drawn on the video frame image, and the image is sent to the display screen for presentation. The video frame image containing the complete motion curve of the bow and the video frame image containing key actions can be extracted and stored locally or uploaded to the cloud for subsequent reference.

[0167] In this way, video collection and analysis are performed during the archery process to detect the user's archery posture during the archery process, and the user's archery posture is quantified through data. The obtained target posture data is compared with the standard posture data, which can objectively and accurately determine the standard degree of the archery posture, and then guide the user to correct the posture, which greatly facilitates the user's training.

[0168] An archery posture evaluation method provided in an embodiment of the present invention is applied to an edge computing server. In a specific application scenario, the edge computing server and the camera and the display screen can be independently set, and the edge computing server communicates with the camera and the display screen respectively; or the edge computing server (processor) and the camera and the display screen are set as an integrated machine, which is not limited in the embodiment of the present invention.

[0169] Corresponding to the above method embodiment, the embodiment of the present invention also provides an archery posture evaluation device, such as Fig.10 As shown, the device is applied to an edge computing server and may include: a video acquisition module 1010, a user judgment module 1020, an event triggering module 1030, a location identification module 1040, a data determination module 1050, and a posture evaluation module 1060.

[0170] The video acquisition module 1010 is used to acquire a first video through a selected image acquisition device;

[0171] A user determination module 1020 is used to determine whether the user enters a designated location area according to the first video;

[0172] An event triggering module 1030 is used for triggering an archery posture detection event for the user if yes;

[0173] The position recognition module 1040 is used to obtain a second video through the image acquisition device, and for each frame of the second image in the second video, recognize the position of the skeletal joint points of the key parts of the user in the second image;

[0174] A data determination module 1050 is used to determine the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user, and the posture data corresponding to each of the second images in the second video is used to form the posture data of the user during the archery process;

[0175] The posture evaluation module 1060 is used to select target posture data from the posture data of the user during the archery process, and compare it with preset standard posture data to evaluate the archery posture of the user.

[0176] The embodiment of the present invention also provides an edge computing server, such as Fig.11 As shown, it includes a processor 111, a communication interface 112, a memory 113 and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114.

[0177] Memory 113, used for storing computer programs;

[0178] The processor 111 is used to execute the program stored in the memory 113 to implement the following steps:

[0179] A first video is acquired through a selected image acquisition device, and based on the first video, it is determined whether a user enters a designated location area; if so, an archery posture detection event for the user is triggered; a second video is acquired through the image acquisition device, and for each frame of a second image in the second video, the positions of the skeletal joints of the key parts of the user in the second image are identified; based on the positions of the skeletal joints of the key parts of the user, the posture data corresponding to the second image is determined, and the posture data of the user during the archery process is composed of the posture data corresponding to each of the second images in the second video; target posture data is selected from the posture data of the user during the archery process, and compared with preset standard posture data to evaluate the user's archery posture.

[0180] The communication bus mentioned in the edge computing server can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0181] The communication interface is used for communication between the above-mentioned edge computing server and other devices.

[0182] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0183] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0184] In another embodiment provided by the present invention, a storage medium is also provided, in which instructions are stored, and when the instructions are executed on a computer, the computer executes the archery posture evaluation method described in any one of the above embodiments.

[0185] In another embodiment provided by the present invention, a computer program product comprising instructions is also provided, which, when executed on a computer, enables the computer to execute the archery posture evaluation method described in any one of the above embodiments.

[0186] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a storage medium, or transmitted from one storage medium to another storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk Solid State Disk (SSD)), etc.

[0187] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0188] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0189] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A method for evaluating archery posture, characterized in that: The method comprises: Acquire a first video through a selected image acquisition device, and determine whether the user enters a designated location area based on the first video; If yes, triggering an archery posture detection event for the user; Acquire a second video through the image acquisition device, and identify the positions of the skeletal joint points of the key parts of the user in each frame of the second image in the second video; Determine the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user, and form the posture data of the user during the archery process by the posture data corresponding to each of the second images in the second video; Selecting target posture data from the posture data of the user in the process of archery, and comparing with preset standard posture data to evaluate the archery posture of the user, including: determining the similarity between any adjacent N frames of the second image in the second video; determining the adjacent N frames of the second image whose similarity exceeds a preset similarity threshold from the second video; wherein N≥2; selecting a target second image from the adjacent N frames of the second image whose similarity exceeds a preset similarity threshold, and selecting an angle corresponding to the target second image as a target angle; comparing the target angle with a preset standard angle to evaluate the archery posture of the user.

2. The method according to claim 1, characterized in that: Deploy candidate image acquisition devices on the left and right sides respectively with the bull's eye as the central axis, and the candidate image acquisition devices are located on the same horizontal line as the user; The step of acquiring the first video through the selected image acquisition device includes: Acquiring the archery habits of the user, and determining an image acquisition device from the candidate image acquisition devices according to the archery habits of the user; The image acquisition device is started to perform video acquisition, and a first video is obtained through the image acquisition device.

3. The method according to claim 1, characterized in that The determining, based on the first video, whether the user enters a designated location area includes: For a first image in the first video, input the first image into a preset human position detection model to obtain a user position in the first image output by the human position detection model; The user position is compared with a preset archery standing position area to determine whether the user enters the designated position area.

4. The method according to claim 3, characterized in that: The human body position detection model is specifically obtained in the following manner: Acquire archery videos corresponding to multiple target users, and perform frame extraction processing on the archery videos to obtain a preset number of archery images; Using a rectangular frame marking method to mark human targets on the preset number of archery images to generate human detection training samples; Based on the human body detection training samples, supervised training is performed on the initial model for human body position detection to obtain a human body position detection model.

5. The method according to claim 1, characterized in that: The identifying the positions of the skeletal joints of the key parts of the user in the second image includes: The second image is input into a preset human posture assessment model to obtain the positions of the skeletal joint points of the key parts of the user in the second image output by the human posture assessment model.

6. The method according to claim 5, characterized in that The human body posture assessment model is specifically obtained by the following method: Acquire archery videos corresponding to multiple target users, and perform frame extraction processing on the archery videos to obtain a preset number of archery images; Using a skeleton joint point marking method to mark human joint points on the preset number of archery images to generate archery posture evaluation training samples; Based on the archery posture evaluation training samples, supervised training is performed on the initial model for human posture evaluation to obtain a human posture evaluation model.

7. The method according to claim 1, characterized in that The step of determining the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user comprises: According to the positions of the skeletal joints of the key parts of the user, forming vectors corresponding to the key parts of the user; The angle between the vectors corresponding to the key parts of the user is calculated, and the angle is determined to be the angle corresponding to the key parts of the user corresponding to the second image.

8. The method according to any one of claims 1 to 7, characterized in that: The motion sensor is mounted on the bow, and the method further comprises: Acquire the motion data of the bow during the user's archery shooting currently collected by the motion sensor, and draw the current moving track of the bow according to the motion data; Determine a collection time of the motion data, and search the second video for the second image corresponding to the collection time; The current movement trajectory of the bow is marked on the second image corresponding to the acquisition moment and transmitted to the display screen for display.

9. The method according to claim 1, characterized in that: The motion sensor is mounted on the bow, and the method further comprises: Acquire the complete motion data of the bow collected by the motion sensor during the user's archery process, and draw a complete movement trajectory of the bow according to the complete motion data; The complete movement trajectory of the bow is marked on the second image of the target and transmitted to a display screen for display.

10. An archery posture evaluation device, characterized in that: The device comprises: A video acquisition module, used for acquiring a first video through a selected image acquisition device; A user determination module, used to determine whether the user enters a designated location area according to the first video; an event triggering module, configured to trigger an archery posture detection event for the user if yes; A position recognition module, used to obtain a second video through the image acquisition device, and for each frame of a second image in the second video, identify the positions of the skeletal joints of the key parts of the user in the second image; a data determination module, configured to determine the posture data corresponding to the second image according to the positions of the skeletal joints of the key parts of the user, and to form the posture data of the user during the archery shooting process from the posture data corresponding to each of the second images in the second video; A posture evaluation module is used to select target posture data from the posture data of the user in the process of archery, and compare it with preset standard posture data to evaluate the archery posture of the user, including: determining the similarity between any adjacent N frames of the second image in the second video; determining the adjacent N frames of the second image whose similarity exceeds a preset similarity threshold from the second video; wherein N≥2; selecting a target second image from the adjacent N frames of the second image whose similarity exceeds the preset similarity threshold, and selecting the angle corresponding to the target second image as the target angle; comparing the target angle with the preset standard angle to evaluate the archery posture of the user.

11. An edge computing server, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, for implementing the method steps described in any one of claims 1 to 9 when executing a program stored in a memory.

12. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images