A method, system, device and medium for establishing a sports video course library

Abstract

The present invention discloses a method, system, device and medium for establishing a sports video course library. The method comprises the following steps: performing frame processing on a sports video; identifying the following parameters for a continuous multi-frame sports image: the number of times a complete action is repeated N; the time taken to complete a single complete action T; the vertical total displacement V of a single moving part i in a complete action; i ; The total horizontal displacement H of a single moving part i in a complete movement i ; The total displacement time t of a single moving part i in a complete movement i ; The total number of moving parts i n; Get the reference mass M of moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ; total energy consumption D of the action; based on the total energy consumption D of the action, the total number n of the moving parts i, the total time NT and the moving frequency #imgabs0#, the course difficulty C of the sports video is generated; the sports video is marked with the course difficulty C, and then stored. The present invention can quickly establish a database with a large number of sports video courses marked with difficulty.

Description

Technical Field

[0001] The present invention belongs to the field of video processing technology, and in particular relates to a method, system, device and medium for establishing a sports video course library. Background Art

[0002] With the advancement of sports and national fitness, more and more people are using smart fitness equipment and sports equipment (such as smart dumbbells, bicycles or treadmills), or obtaining amateur fitness courses and even professional sports training content on the Internet through tools such as smart phones and computers.

[0003] At present, due to differences in users' physical fitness, they have different tolerance for different sports training contents. If the sports training content is too difficult, there is a high possibility of danger such as sports injuries. Based on this, some technologies for personalized matching of users have emerged to tailor exercise plans for users. However, the difficulty in this technology is that it is difficult to measure the difficulty of sports video courses, resulting in insufficient materials when matching exercise plans for users.

[0004] At present, the difficulty measurement of sports video courses is still mainly based on human judgment, which leads to a lack of scientific judgment on the difficulty measurement of sports video courses. In addition, human judgment requires a lot of professional knowledge, as well as a lot of time and energy. This makes it difficult to measure the difficulty of a large number of sports video courses, and it is impossible to establish a database with a large number of annotated sports video courses. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method, system, device and medium for establishing a sports video course library in view of the deficiencies in the above-mentioned prior art, which can quickly establish a database with a large number of sports video courses marked with difficulty.

[0006] The first aspect of the present invention discloses a method for establishing a sports video course library, comprising the following steps:

[0007] Step 1: Obtain a motion video and perform frame processing on the motion video to obtain a continuous multi-frame motion image;

[0008] Step 2: Identify the continuous multi-frame motion images and identify the following parameters:

[0009] Repeat the complete action number of times N;

[0010] The time it takes to complete a single complete action is T;

[0011] The total vertical displacement V of a single moving part i in a complete movement i ;

[0012] The total horizontal displacement H of a single moving part i in a complete movement i ;

[0013] The total displacement time t of a single moving part i in a complete movement i ;

[0014] The total number of moving parts i is n;

[0015] Step 3: Get the reference mass M of the moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ;

[0016] Step 4: Calculate the total energy consumption D of the motion video:

[0017]

[0018] Step 5: Based on the total energy consumption D of the action, the total number n of the movement parts i, the total time NT and the movement frequency Course difficulty C for generating motion videos;

[0019] Step 6: Label the course difficulty C for the sports video and then store it in the sports video course library.

[0020] The above-mentioned method for establishing a sports video course library, when identifying the number of times N of repeating a complete action in step 2, includes the following steps:

[0021] Recognize multiple frames of motion images, identify multiple joints of the human body in each frame of motion images, and construct a simplified line diagram of the human body based on the multiple joints of the human body. Each simplified line diagram is a decomposed action diagram, and continuous multiple frames of decomposed action diagrams are obtained;

[0022] The number of decomposed action graphs corresponding to a decomposed action in the multi-frame decomposed action graphs is counted as the number N of repeating the complete action.

[0023] In the above-mentioned method for establishing a sports video course library, the time T required to complete a single complete action in step 2 is identified, and the following steps are included:

[0024] The time between two decomposition action diagrams that correspond to a decomposition action and are adjacent in time sequence is counted as the time consumption T for completing a single complete action.

[0025] In the above method for establishing a sports video course library, in step 2, the vertical total displacement V of a single moving part i in a complete action is identified. i The following steps are included:

[0026] Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps:

[0027] Step a1, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A;

[0028] Step b1, obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and calculate the vertical displacement v of the moving part i by comparing the current frame decomposition action graph A with the next frame decomposition action graph A. i ;

[0029] Step c1, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b1;

[0030] Step b1 and step c1 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple displacements v are obtained. i , based on multiple displacements v i Calculate the total vertical displacement V of a single moving part i i .

[0031] The above-mentioned method for establishing a sports video course library, in step 2, identifies the total horizontal displacement H of a single moving part i in a complete action i The following steps are included:

[0032] Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps:

[0033] Step a2, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A;

[0034] Step b2: Obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and calculate the horizontal displacement h of the motion part i when comparing the current frame decomposition action graph A with the next frame decomposition action graph A. i ;

[0035] Step c2, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b2;

[0036] Step b2 and step c2 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple displacements h are obtained. i , based on multiple displacements h i Calculate the total horizontal displacement H of a single moving part i i .

[0037] The above-mentioned method for establishing a sports video course library, in step 2, identifies the total displacement time t of a single sports part i i The following steps are included:

[0038] Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps:

[0039] Step a3, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A;

[0040] Step b3, obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and compare the current frame decomposition action graph A with the next frame decomposition action graph A to determine whether the motion part i is displaced. If so, record the time t once. t ;

[0041] Step c3, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b3;

[0042] Step b3 and step c3 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple time t t , based on multiple time t t The total displacement time t of moving part i is calculated by the sum of i .

[0043] The above-mentioned method for establishing a sports video course library, when identifying the total number n of sports parts i in step 2, includes the following steps:

[0044] Filter out a frame of decomposed action graph B with the largest number of joint points among the multiple frame decomposed action graphs;

[0045] Based on the number of joints in the decomposed motion graph B, the total number n of motion parts i is calculated.

[0046] The second aspect of the present invention discloses a sports video course library establishment system, including a frame processing module, an identification processing module, a parameter acquisition module, an energy consumption calculation module, a difficulty calculation module and a storage module;

[0047] A frame processing module is used to acquire motion video and perform frame processing on the motion video to obtain continuous multi-frame motion images;

[0048] The recognition processing module is used to recognize continuous multi-frame motion images and identify the following parameters:

[0049] Repeat the complete action number of times N;

[0050] The time it takes to complete a single complete action is T;

[0051] The total vertical displacement V of a single moving part i in a complete movement i ;

[0052] The total horizontal displacement H of a single moving part i in a complete movementi ;

[0053] The total displacement time t of a single moving part i in a complete movement i ;

[0054] The total number of moving parts i is n;

[0055] Parameter acquisition module, used to obtain the reference mass M of the moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ;

[0056] The energy consumption calculation module is used to calculate the total energy consumption D of the motion video:

[0057]

[0058] Difficulty calculation module, based on the total energy consumption D of the action, the total number n of movement parts i, the total time NT and the movement frequency Course difficulty C for generating motion videos;

[0059] The storage module is used to mark the course difficulty C of the sports video and then store it in the sports video course library.

[0060] A third aspect of the present invention discloses an electronic device, comprising:

[0061] processor;

[0062] A storage medium having a computer program stored thereon, wherein the computer program is executed by the processor to perform the method according to the first aspect.

[0063] A fourth aspect of the present invention discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method according to the first aspect is executed.

[0064] Compared with the prior art, the present invention has the following advantages: the present invention performs frame processing on the motion video, then identifies multiple motion parameters contained in the motion video, and based on the idea of ​​mathematical modeling, makes a calculation formula for measuring the total energy consumption D of the action, combines the energy consumption of various parts of the human body during human motion, and finally gives the difficulty of the motion video. The method is scientific and effective, and can automatically mark the difficulty of a large number of motion videos, thereby realizing the rapid establishment of a database with a large number of sports video courses marked with difficulty.

[0065] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flow chart of the method of Example 1.

[0067] Figure 2 This is a system block diagram of Example 2. DETAILED DESCRIPTION

[0068] like Figure 1 As shown, a method for establishing a sports video course library includes the following steps:

[0069] Step 1: Obtain a motion video and perform frame processing on the motion video to obtain a continuous multi-frame motion image;

[0070] Step 2: Identify the continuous multi-frame motion images and identify the following parameters:

[0071] Repeat the complete action number of times N;

[0072] The time it takes to complete a single complete action is T;

[0073] The total vertical displacement V of a single moving part i in a complete movement i ;

[0074] The total horizontal displacement H of a single moving part i in a complete movement i ;

[0075] The total displacement time t of a single moving part i in a complete movement i ;

[0076] The total number of moving parts i is n;

[0077] Step 3: Get the reference mass M of the moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ;

[0078] It should be noted that the reference mass M i and static maintenance energy consumption E i Based on the physiological experience parameters, it can be selected according to the specific situation of the person in the sports video;

[0079] Step 4: Calculate the total energy consumption D of the motion video:

[0080]

[0081] Step 5: Based on the total energy consumption D of the action, the total number n of the movement parts i, the total time NT and the movement frequency Course difficulty C for generating motion videos;

[0082] It should be noted that the total energy consumption D, the total number of movement parts i n, the total time NT and the movement frequency Represents the four dimensions of measuring the difficulty of sports. The course difficulty C can be obtained by adding different coefficients to the four dimensions and then summing them up. The specific choice can be made according to the actual situation;

[0083] Step 6: Label the course difficulty C for the sports video and then store it in the sports video course library.

[0084] In this embodiment, when identifying the number of times N of repeating the complete action in step 2, the following steps are included:

[0085] Recognize multiple frames of motion images, identify multiple joints of the human body in each frame of motion images, and construct a simplified line diagram of the human body based on the multiple joints of the human body. Each simplified line diagram is a decomposed action diagram, and continuous multiple frames of decomposed action diagrams are obtained;

[0086] The number of decomposed action graphs corresponding to a decomposed action in the multi-frame decomposed action graphs is counted as the number N of repeating the complete action.

[0087] It should be noted that the multiple joints are the fingertips of the left human body, the left wrist joints, the elbow joints of the left arm, the shoulder joints of the left arm, the fingertips of the right human body, the right wrist joints, the elbow joints of the right arm, the shoulder joints of the right arm, the top of the human head, the clavicle joints, the left hip joints of the human body, the right hip joints of the human body, the left knee joints of the human body, the right knee joints of the human body, the left ankle joints of the human body, the right ankle joints of the human body, the left toe joints of the human body, and the right toe joints of the human body.

[0088] When constructing a simplified line drawing of a human body:

[0089] Generate the line of the left arm of the human body based on the lines of the fingertips of the left human body, the joints of the left wrist of the human body, the elbow joints of the left arm of the human body, and the shoulder joints of the left arm of the human body;

[0090] Generate the line of the right arm of the human body based on the lines of the fingertips of the right human body, the joints of the right wrist of the human body, the elbow joints of the right arm of the human body, and the shoulder joints of the right arm of the human body;

[0091] Generate the human head lines based on the human head vertices and human clavicle points.

[0092] Generate a human body trunk line based on a line connecting a right arm shoulder joint point and a left arm shoulder joint point, and a line connecting a clavicle point and a midpoint of a line connecting a left hip joint point and a right hip joint point;

[0093] Generate the left leg line of the human body based on the left hip joint point, the left knee joint point, the left ankle joint point and the left toe joint point of the human body;

[0094] Based on the right hip joint point, the right knee joint point, the right ankle joint point and the right toe joint point, the right leg line of the human body is generated.

[0095] It should be noted that a complete action in a sports video is composed of multiple decomposed actions. For example, a complete kicking action is composed of three actions: leg lifting action, leg extending action and leg retracting action.

[0096] In this embodiment, the time taken to complete a single complete action in step 2 is identified as T, which includes the following steps:

[0097] The time between two decomposition action diagrams that correspond to a decomposition action and are adjacent in time sequence is counted as the time consumption T for completing a single complete action.

[0098] For example, if the complete action is a kicking action, the time taken T is the time between the first leg-lifting action and the next leg-lifting action.

[0099] In this embodiment, the total vertical displacement V of a single moving part i in a complete action is identified in step 2. i The following steps are included:

[0100] Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps:

[0101] Step a1, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A;

[0102] Step b1, obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and calculate the vertical displacement v of the moving part i by comparing the current frame decomposition action graph A with the next frame decomposition action graph A. i ;

[0103] Step c1, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b1;

[0104] Step b1 and step c1 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple displacements v are obtained. i , based on multiple displacements v i Calculate the total vertical displacement V of a single moving part i i .

[0105] It should be noted that, assuming that the moving part i is the left hand of the human body, the midpoint of the line connecting the fingertips of the left hand and the joints of the left wrist of the human body is set as the center of mass of the left hand of the human body, and the total vertical displacement of this point in a complete movement is the total vertical displacement V of the left hand of the human body. i; Of course, the motion direction of the point in the comparison of two consecutive frames of decomposed action graph A can also be used as a reference to calculate the single displacement v i Add positive and negative marks. For example, moving upward is positive displacement, moving downward is negative displacement, or both can be positive displacement.

[0106] Calculate a single displacement v i When the two consecutive frames of decomposed action graph A are placed in the same two-dimensional coordinate system, the displacement v is obtained by changing the coordinates of the points. i .

[0107] In this embodiment, the total horizontal displacement H of a single moving part i in a complete action is identified in step 2. i The following steps are included:

[0108] Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps:

[0109] Step a2, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A;

[0110] Step b2: Obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and calculate the horizontal displacement h of the motion part i when comparing the current frame decomposition action graph A with the next frame decomposition action graph A. i ;

[0111] Step c2, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b2;

[0112] Step b2 and step c2 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple displacements h are obtained. i , based on multiple displacements h i Calculate the total horizontal displacement H of a single moving part i i Here you can refer to the total vertical displacement V i Calculation.

[0113] In this embodiment, the total displacement time t of a single moving part i identified in step 2 is i The following steps are included:

[0114] Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps:

[0115] Step a3, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A;

[0116] Step b3, obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and compare the current frame decomposition action graph A with the next frame decomposition action graph A to determine whether the motion part i is displaced. If so, record the time t once. t ;

[0117] Step c3, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b3;

[0118] Step b3 and step c3 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple time t t , based on multiple time t t The total displacement time t of moving part i is calculated by the sum of i .

[0119] It should be noted that the total displacement time t of any moving part i is i Are both less than or equal to the time T required to complete a single complete action.

[0120] In this embodiment, when identifying the total number n of motion parts i in step 2, the following steps are included:

[0121] Filter out a frame of decomposed action graph B with the largest number of joint points among the multiple frame decomposed action graphs;

[0122] Based on the number of joints in the decomposed motion graph B, the total number n of motion parts i is calculated.

[0123] It should be noted that, in the present embodiment, the total number of the moving parts i is n=14, which are respectively the left hand part of the human body, the left forearm part of the human body, the left upper arm part of the human body, the left foot part of the human body, the left calf part of the human body, the left thigh part of the human body, the right hand part of the human body, the right forearm part of the human body, the right upper arm part of the human body, the right foot part of the human body, the right calf part of the human body, the right thigh part of the human body, the head part of the human body, and the trunk part of the human body; (limbs * 3 + trunk + head and neck). The more joints there are, the more parts of the human body that can be divided, and the specific selection can be based on actual conditions. The purpose of selecting a frame of decomposed action graph B with the largest number of joints is to avoid missing identifiable human body parts i.

[0124] Example 2

[0125] like Figure 2 As shown, a sports video course library establishment system includes a frame processing module 100, a recognition processing module 200, a parameter acquisition module 300, an energy consumption calculation module 400, a difficulty calculation module 500 and a storage module 600;

[0126] The frame processing module 100 is used to obtain a motion video and perform frame processing on the motion video to obtain a continuous multi-frame motion image;

[0127] The recognition processing module 200 is used to recognize the continuous multi-frame motion images and identify the following parameters:

[0128] Repeat the complete action number of times N;

[0129] The time it takes to complete a single complete action is T;

[0130] The total vertical displacement V of a single moving part i in a complete movement i ;

[0131] The total horizontal displacement H of a single moving part i in a complete movement i ;

[0132] The total displacement time t of a single moving part i in a complete movement i ;

[0133] The total number of moving parts i is n;

[0134] Parameter acquisition module 300, used to obtain the reference mass M of the moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ;

[0135] The energy consumption calculation module 400 is used to calculate the total energy consumption D of the motion video:

[0136]

[0137] The difficulty calculation module 500 is based on the total energy consumption D of the action, the total number n of the movement parts i, the total time NT and the movement frequency Course difficulty C for generating motion videos;

[0138] The storage module 600 is used to mark the course difficulty C of the sports video and then store it in the sports video course library.

[0139] The technical solution of this embodiment is substantially the same as that of Embodiment 1. If there are any unclear points, please refer to Embodiment 1.

[0140] Example 3

[0141] An electronic device, comprising:

[0142] processor;

[0143] A storage medium having a computer program stored thereon, wherein the computer program executes the method described in embodiment 1 when executed by the processor.

[0144] Example 4

[0145] A computer-readable storage medium stores a computer program, and the computer program executes the method described in Example 1 when executed by a processor.

[0146] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for establishing a sports video course library, characterized in that: The following steps are involved: Step 1: Obtain a motion video and perform frame processing on the motion video to obtain a continuous multi-frame motion image; Step 2: Identify the continuous multi-frame motion images and identify the following parameters: Repeat the complete action number of times N; The time it takes to complete a single complete action is T; The total vertical displacement V of a single moving part i in a complete movement i ; The total horizontal displacement H of a single moving part i in a complete movement i ; The total displacement time t of a single moving part i in a complete movement i ; The total number of moving parts i is n; Step 3: Get the reference mass M of the moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ; Step 4: Calculate the total energy consumption D of the motion video: Step 5: Based on the total energy consumption D of the action, the total number n of the movement parts i, the total time NT and the movement frequency Course difficulty C for generating motion videos; Step 6: Label the course difficulty C for the sports video and then store it in the sports video course library.

2. A method for establishing a sports video course library according to claim 1, characterized in that: When identifying the number of times N to repeat the complete action in step 2, the following steps are included: Recognize multiple frames of motion images, identify multiple joints of the human body in each frame of motion images, and construct a simplified line diagram of the human body based on the multiple joints of the human body. Each simplified line diagram is a decomposed action diagram, and continuous multiple frames of decomposed action diagrams are obtained; The number of decomposed action graphs corresponding to a decomposed action in the multi-frame decomposed action graphs is counted as the number N of repeating the complete action.

3. A method for establishing a sports video course library according to claim 2, characterized in that: In step 2, the time T required to complete a single complete action is identified, which includes the following steps: The time between two decomposition action diagrams that correspond to a decomposition action and are adjacent in time sequence is counted as the time consumption T for completing a single complete action.

4. A method for establishing a sports video course library according to claim 2, characterized in that: In step 2, we identify the total vertical displacement V of a single moving part i in a complete movement. i The following steps are included: Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps: Step a1, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A; Step b1, obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and calculate the vertical displacement v of the moving part i by comparing the current frame decomposition action graph A with the next frame decomposition action graph A. i ; Step c1, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b1; Step b1 and step c1 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple displacements v are obtained. i , based on multiple displacements v i Calculate the total vertical displacement V of a single moving part i i .

5. A method for establishing a sports video course library according to claim 2, characterized in that: In step 2, the total horizontal displacement H of a single moving part i in a complete action is identified i The following steps are included: Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps: Step a2, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A; Step b2: Obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and calculate the horizontal displacement h of the motion part i when comparing the current frame decomposition action graph A with the next frame decomposition action graph A. i ; Step c2, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b2; Step b2 and step c2 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple displacements h are obtained. i , based on multiple displacements h i Calculate the total horizontal displacement H of a single moving part i i .

6. A method for establishing a sports video course library according to claim 2, characterized in that: The total displacement time t of a single moving part i identified in step 2 i The following steps are included: Through the continuous multi-frame decomposition action graph, obtain the continuous multi-frame decomposition action graph A corresponding to a complete action, and then perform the following steps: Step a3, obtaining the first frame decomposition action graph A among the continuous multiple frame decomposition action graphs A as the current frame decomposition action graph A; Step b3, obtain the next frame decomposition action graph A of the current frame decomposition action graph A, and compare the current frame decomposition action graph A with the next frame decomposition action graph A to determine whether the motion part i is displaced. If so, record the time t once. t ; Step c3, taking the next frame decomposition action graph A as the new current frame decomposition action graph A, and then proceeding to step b3; Step b3 and step c3 are executed repeatedly until the continuous multi-frame decomposition action graph A is executed and multiple time t t , based on multiple time t t The total displacement time t of moving part i is calculated by the sum of i .

7. A method for establishing a sports video course library according to claim 2, characterized in that: When the total number n of the moving parts i is identified in step 2, the following steps are included: Filter out a frame of decomposed action graph B with the largest number of joint points among the multiple frame decomposed action graphs; Based on the number of joints in the decomposed motion graph B, the total number n of motion parts i is calculated.

8. A sports video course library establishment system, characterized in that: It includes a frame processing module, an identification processing module, a parameter acquisition module, an energy consumption calculation module, a difficulty calculation module and a storage module; A frame processing module is used to obtain motion video and perform frame processing on the motion video to obtain continuous multi-frame motion images; The recognition processing module is used to recognize continuous multi-frame motion images and identify the following parameters: Repeat the complete action number of times N; The time it takes to complete a single complete action is T; The total vertical displacement V of a single moving part i in a complete movement i ; The total horizontal displacement H of a single moving part i in a complete movement i ; The total displacement time t of a single moving part i in a complete movement i ; The total number of moving parts i is n; Parameter acquisition module, used to obtain the reference mass M of the moving part i i , gravitational acceleration g, static energy consumption of a single moving part in the starting posture E i ; The energy consumption calculation module is used to calculate the total energy consumption D of the motion video: Difficulty calculation module, based on the total energy consumption D of the action, the total number n of movement parts i, the total time NT and the movement frequency Course difficulty C for generating motion videos; The storage module is used to mark the course difficulty C of the sports video and then store it in the sports video course library.

9. An electronic device, characterized in that: The electronic device comprises: processor; A storage medium having a computer program stored thereon, wherein the computer program, when executed by the processor, performs the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is executed.

Images