Video AI training physical test system

Abstract

The invention relates to the technical field of AI intelligent systems, in particular to a video AI training physical test system which comprises a video data acquisition module, a data classification module, a data storage module, an AI training module and a physical test monitoring module. The data classification module classifies the acquired body test video data, the data storage module stores the classified body test video data, and the AI training module performs pre-training based on an existing video AI training model and the acquired body test video to obtain a pre-trained body test monitoring model; the physical testing and monitoring module performs physical testing and monitoring on the basis of a pre-trained physical monitoring model, so that the problems that in an existing physical testing system, monitoring data of the physical testing and monitoring module cannot be trained and corrected, and only fixed physical testing and monitoring data can be applied are solved; therefore, errors of monitoring data can be caused in the physical fitness testing and monitoring process.

Description

Technical Field

[0001] The present invention relates to the technical field of AI intelligent systems, and in particular to a video AI training physical test system. Background Art

[0002] Physical fitness test, also known as physical fitness test, is the process of testing and evaluating the comprehensive ability of the human body in adapting to the environment. Physical fitness test includes the testing and evaluation of improving human health and basic activity ability, and also covers the testing and evaluation of competitive ability in sports competitions.

[0003] The prior art discloses a physical test system, including a physical test monitoring module, which includes a judgment module and an evaluation module. During the physical test of an athlete, the judgment module is used to judge whether the tester has any illegal behavior during the physical test, and the evaluation module is used to evaluate the specific physical test results to obtain physical test data.

[0004] However, by adopting the above method, the monitoring data of the physical fitness monitoring module cannot be corrected by training, and only fixed physical fitness monitoring data can be applied. Therefore, errors may occur in the monitoring data during the physical fitness test monitoring process. Summary of the invention

[0005] The purpose of the present invention is to provide a video AI training physical test system, which aims to solve the problem of existing physical test systems in which the monitoring data of the physical test monitoring module cannot be trained and corrected and can only apply fixed physical test monitoring data, which may cause errors in the monitoring data during the physical test monitoring process.

[0006] To achieve the above-mentioned object, the present invention provides a video AI training and physical measurement system, including a video data acquisition module, a data classification module, a data storage module, an AI training module and a physical measurement monitoring module;

[0007] The data classification module is connected to the video data acquisition module; the data storage module is connected to the data classification module; the AI ​​training module is connected to the data storage module; the physical measurement monitoring module is connected to the AI ​​training module;

[0008] The video data acquisition module is used to acquire physical measurement video data based on the big data platform;

[0009] The data classification module is used to classify the acquired physical measurement video data according to different physical measurement types for subsequent storage;

[0010] The data storage module is used to store the classified body measurement video data;

[0011] The AI ​​training module is used to obtain different physical measurement videos stored in the data storage module, and then perform pre-training based on the existing video AI training model to obtain a pre-trained physical measurement monitoring model;

[0012] The physical fitness monitoring module is used to monitor the physical fitness test based on the pre-trained physical fitness monitoring model.

[0013] The video data acquisition module includes a big data platform docking submodule, a video retrieval submodule and a video crawling submodule, wherein the video retrieval submodule is connected to the big data platform docking submodule; the video crawling submodule is connected to the video retrieval submodule;

[0014] The big data platform docking submodule is used to connect online with the physical test video search engine to obtain the physical test videos in the physical test database online in real time;

[0015] The video retrieval submodule is used to search for videos based on keywords of different physical test types through a search engine;

[0016] The video crawling submodule is used to crawl and search for corresponding physical examination videos.

[0017] Wherein, the data classification module includes a video recognition submodule and a video classification submodule, the video recognition submodule is connected to the video data acquisition module; the video classification submodule is connected to the video recognition submodule;

[0018] The video recognition submodule is used to recognize the decoded physical measurement video content and recognize different types of physical measurement videos;

[0019] The video classification submodule is used to classify and stack the physical examination videos identified as the same type for subsequent storage.

[0020] Among them, the AI ​​training module includes a data receiving submodule, a model application submodule, a model training submodule and a model generation submodule, the data receiving submodule is connected to the data storage module; the model application submodule is connected to the data receiving submodule; the model training submodule is connected to the model application submodule; the model generation submodule is connected to the model training submodule;

[0021] The data receiving submodule is used to obtain different types of body measurement video data stored in the data storage module;

[0022] The model application submodule is used to fill different types of body measurement video data into the model library based on the existing video AI training model;

[0023] The model training submodule is used for training on different types of body measurement videos based on filling;

[0024] The model generation submodule is used to generate an accurate physical measurement monitoring model under the continuous training of the model training submodule, and has the function of real-time update and improvement to ensure the accuracy of monitoring.

[0025] Wherein, the video AI training physical test system further includes a physical fitness assessment module, and the physical fitness assessment module is connected to the physical test monitoring module;

[0026] The physical fitness assessment module is used to evaluate the physical fitness value based on the physical fitness test data monitored by the physical fitness monitoring module, so as to facilitate physical fitness assessment for different physical conditions.

[0027] The physical fitness evaluation module includes a physical fitness data acquisition submodule, an evaluation and judgment submodule and a physical fitness data generation submodule, wherein the physical fitness data acquisition submodule is connected to the physical fitness monitoring module; the evaluation and judgment submodule is connected to the physical fitness data acquisition submodule; and the physical fitness data generation submodule is connected to the evaluation and judgment submodule;

[0028] The physical measurement data acquisition submodule is used to acquire the physical measurement monitoring data of the physical measurement monitoring module;

[0029] The evaluation and judgment submodule is used to judge the acquired physical test monitoring data, and analyze the physical test monitoring data in combination with factors such as environmental weather, climate and the physical condition of the athlete to obtain an analysis result;

[0030] The physical fitness data generation submodule is used to generate specific physical fitness data of the athlete based on the analysis results.

[0031] Among them, the video AI training physical test system also includes an evaluation data push module, and the evaluation data push module is connected to the physical fitness evaluation module;

[0032] The evaluation data push module is used to push the evaluated athlete's physical test data to the mobile terminal for subsequent reference by the athlete, so as to facilitate special physical training.

[0033] The present invention provides a video AI training physical measurement system, wherein the video data acquisition module acquires physical measurement video data based on a big data platform, and the data classification module classifies the acquired physical measurement video data according to different physical measurement types, and the data storage module stores the classified physical measurement video data, and the AI ​​training module acquires different physical measurement videos stored in the data storage module, and then pre-trains based on the existing video AI training model to obtain a pre-trained physical measurement monitoring model, and then the physical measurement monitoring module monitors the physical measurement based on the pre-trained physical measurement monitoring model, thereby solving the problem of the existing physical measurement system, in which the monitoring data of the physical measurement monitoring module cannot be trained and corrected and can only apply fixed physical measurement monitoring data, and thus may cause errors in the monitoring data during the physical measurement monitoring process. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0035] Figure 1 It is a schematic diagram of the overall structure of the first embodiment of the present invention.

[0036] Figure 2 It is a schematic diagram of the structure of a video data acquisition module according to the first embodiment of the present invention.

[0037] Figure 3 It is a schematic diagram of the structure of the AI ​​training module of the first embodiment of the present invention.

[0038] Figure 4 It is a schematic diagram of the overall structure of the second embodiment of the present invention.

[0039] Figure 5 It is a schematic diagram of the structure of a physical fitness assessment module according to the second embodiment of the present invention.

[0040] Figure 6 It is a schematic diagram of the overall structure of the third embodiment of the present invention.

[0041] 101- video data acquisition module, 102- data classification module, 103- data storage module, 104- AI training module, 105- physical measurement monitoring module, 106- big data platform docking submodule, 107- video retrieval submodule, 108- video crawling submodule, 109- video recognition submodule, 110- video classification submodule, 111- data receiving submodule, 112- model application submodule, 113- model training submodule, 114- model generation submodule, 115- storage judgment submodule, 116- video compression submodule, 117- mesh storage submodule, 118- tree storage submodule, 119- video decoding submodule, 201- physical fitness assessment module, 202- physical measurement data acquisition submodule, 203- assessment judgment submodule, 204- physical fitness data generation submodule, 301- assessment data push module. DETAILED DESCRIPTION

[0042] The first embodiment of the present application is:

[0043] See also Figure 1-Figure 3 ,in, Figure 1 is a schematic diagram of the overall structure of the first embodiment of the present invention, Figure 2 is a schematic diagram of the structure of a video data acquisition module according to a first embodiment of the present invention, Figure 3 It is a schematic diagram of the structure of the AI ​​training module of the first embodiment of the present invention.

[0044] A video AI training physical test system of the present invention comprises a video data acquisition module 101, a data classification module 102, a data storage module 103, an AI training module 104 and a physical test monitoring module 105, wherein the video data acquisition module 101 comprises a big data platform docking submodule 106, a video retrieval submodule 107 and a video crawling submodule 108, the data classification module 102 comprises a video recognition submodule 109 and a video classification submodule 110, the AI ​​training module 104 comprises a data receiving submodule 111, a model application submodule 112, a model training submodule 113 and a model generation submodule 114, the data storage module 103 comprises a storage judgment submodule 115, a video compression submodule 116, a mesh storage submodule 117, a tree storage submodule 118 and a video decoding submodule 119, and the above-mentioned scheme solves the problem of the existing physical test system, in which the monitoring data of the physical test monitoring module 105 cannot be trained and corrected and can only apply fixed physical test monitoring data, so the monitoring data may cause errors in the physical test monitoring process.

[0045] Among them, the data classification module 102 is connected to the video data acquisition module 101; the data storage module 103 is connected to the data classification module 102; the AI ​​training module 104 is connected to the data storage module 103; the physical measurement monitoring module 105 is connected to the AI ​​training module 104, the video data acquisition module 101 acquires physical measurement video data based on the big data platform, the data classification module 102 is used to classify the acquired physical measurement video data according to different physical measurement types for subsequent storage, the data storage module 103 is used to store the classified physical measurement video data, the AI ​​training module 104 acquires different physical measurement videos stored in the data storage module 103, and then pre-trains based on the existing video AI training model to obtain a pre-trained physical measurement monitoring model, the physical measurement monitoring module 105 is based on the pre-trained The trained physical fitness monitoring model is used to monitor the physical fitness test. The video data acquisition module 101 acquires the physical fitness video data based on the big data platform, and the acquired physical fitness video data is classified according to different physical fitness types through the data classification module 102. The data storage module 103 stores the classified physical fitness video data. The AI ​​training module 104 acquires different physical fitness videos stored in the data storage module 103, and then pre-trains based on the existing video AI training model to obtain a pre-trained physical fitness monitoring model. Then, the physical fitness monitoring module 105 monitors the physical fitness test based on the pre-trained physical fitness monitoring model, which solves the problem of the existing physical fitness system, in which the monitoring data of the physical fitness monitoring module 105 cannot be trained and corrected and can only apply fixed physical fitness monitoring data. Therefore, the monitoring data may have errors during the physical fitness test monitoring process.

[0046] Secondly, the video retrieval submodule 107 is connected to the big data platform docking submodule 106; the video crawling submodule 108 is connected to the video retrieval submodule 107; the big data platform docking submodule 106 is used to connect online with the physical test video search engine, and obtain the physical test videos of the physical test database online in real time. The video retrieval submodule 107 searches for videos based on keywords of different physical test types through the search engine, and the video crawling submodule 108 is used to crawl and search for the corresponding physical test videos. The big data platform docking submodule 106 is connected online with the physical test video search engine to obtain the physical test videos of the physical test database, and then the video retrieval submodule 107 searches for keywords of different physical test types through the search engine to obtain a large number of physical test videos, and then the video crawling submodule 108 crawls and searches for the corresponding physical test videos.

[0047] Again, the video recognition submodule 109 is connected to the video data acquisition module 101; the video classification submodule 110 is connected to the video recognition submodule 109; the video recognition submodule 109 is used to identify the decoded physical measurement video content and identify different types of physical measurement videos, and the video classification submodule 110 is used to classify and stack the physical measurement videos identified as the same type for subsequent storage. The video recognition submodule 109 identifies the decoded physical measurement video content and identifies different types of physical measurement videos, such as physical measurement videos such as rope skipping, short-distance running, and long-distance running. Then the video classification submodule 110 classifies and stacks the physical measurement videos identified as the same type for subsequent storage.

[0048] In addition, the data receiving submodule 111 is connected to the data storage module 103; the model application submodule 112 is connected to the data receiving submodule 111; the model training submodule 113 is connected to the model application submodule 112; the model generation submodule 114 is connected to the model training submodule 113, the data receiving submodule 111 is used to obtain different types of physical measurement video data stored in the data storage module 103, the model application submodule 112 is used to fill different types of physical measurement video data into the model library based on the existing video AI training model, the model training submodule 113 is trained based on the filled different types of physical measurement videos, and the model generation submodule 114 generates a model under the continuous training of the model training submodule 113 An accurate physical measurement monitoring model is provided, and has the function of real-time updating and improvement to ensure the accuracy of monitoring. The data receiving submodule 111 obtains different types of physical measurement video data stored in the data storage module 103. The model application submodule 112 has an existing video AI training model built in. Different types of physical measurement video data are filled into the model library based on the existing video AI training model. Then, the model training submodule 113 is trained based on the different types of physical measurement videos filled in the video AI training model. Then, the model generation submodule 114 generates an accurate physical measurement monitoring model under the continuous training of the model training submodule 113. Moreover, the monitoring data of the physical measurement monitoring model can be updated and improved in real time as the database is continuously updated to ensure the accuracy of monitoring.

[0049] Furthermore, the data storage module 103 includes a storage judgment submodule 115, a video compression submodule 116, a mesh storage submodule 117 and a tree storage submodule 118. The storage judgment submodule 115 is connected to the data classification submodule; the video compression submodule 116 is connected to the storage judgment submodule 115; the mesh storage submodule 117 is connected to the video compression submodule 116; the tree storage submodule 118 is connected to the video compression submodule 116. The storage judgment submodule 115 is used to obtain the classified body measurement video data and then judge which storage mode is suitable for it. The video compression submodule 116 is used to compress the judged video data. The mesh storage submodule 117 is used to receive the compressed body measurement video data suitable for mesh storage and store it in a mesh manner. The tree storage submodule 118 is used to receive the compressed body measurement video data suitable for tree storage and store it in a tree manner. The storage judgment submodule 115 obtains the classified body measurement video data and judges which storage method is suitable for it. Then, the body measurement video data is compressed by the video compression submodule 116. When the video data is suitable for mesh storage, the mesh storage submodule 117 performs mesh storage on it. When the video data is suitable for tree storage, the tree storage submodule 118 performs tree storage on it, so as to summarize and store the body measurement video data for subsequent video reading.

[0050] Finally, the data storage module 103 also includes a video decoding submodule 119, which is respectively connected to the mesh storage submodule 117 and the tree storage submodule 118. The video decoding submodule 119 decodes the stored compressed volume measurement video based on the video decoder to achieve the original clarity of the video. The video decoding submodule 119 decodes the acquired compressed volume measurement video based on the video decoder, decodes the compressed video into the original pixel value, and then plays it in combination with the audio signal to improve the clarity of the video, so as to facilitate subsequent AI model training.

[0051] When using a video AI training physical fitness system described in the present invention, the video data acquisition module 101 acquires physical fitness video data based on a big data platform, and the data classification module 102 classifies the acquired physical fitness video data according to different physical fitness types, and the data storage module 103 stores the classified physical fitness video data. The AI ​​training module 104 acquires different physical fitness videos stored in the data storage module 103, and then pre-trains based on the existing video AI training model to obtain a pre-trained physical fitness monitoring model. Then, the physical fitness monitoring module 105 monitors the physical fitness test based on the pre-trained physical fitness monitoring model, thereby solving the problem of the existing physical fitness system, in which the monitoring data of the physical fitness monitoring module 105 cannot be trained and corrected and can only apply fixed physical fitness monitoring data, thereby causing errors in the monitoring data during the physical fitness test monitoring process.

[0052] The second embodiment of the present application is:

[0053] See also Figure 4-Figure 5 ,in, Figure 4 is a schematic diagram of the overall structure of the second embodiment of the present invention, Figure 5 It is a schematic diagram of the structure of a physical fitness assessment module according to the second embodiment of the present invention.

[0054] On the basis of the first embodiment, the present invention provides a video AI training physical measurement system, which also includes a physical fitness assessment module 201, and the physical fitness assessment module 201 includes a physical measurement data acquisition submodule 202, an assessment and judgment submodule 203 and a physical fitness data generation submodule 204.

[0055] Among them, the physical fitness assessment module 201 is connected to the physical measurement monitoring module 105, and the physical fitness assessment module 201 is used to evaluate the physical fitness value based on the physical fitness test data monitored by the physical measurement monitoring module 105, so as to facilitate physical fitness assessment of different physical conditions. The physical fitness assessment module 201 is used to evaluate the physical fitness value based on the physical fitness test data monitored by the physical measurement monitoring module 105, so as to facilitate the specific physical measurement data for physical fitness assessment of different physical conditions.

[0056] Secondly, the physical measurement data acquisition submodule 202 is connected to the physical measurement monitoring module 105; the evaluation and judgment submodule 203 is connected to the physical measurement data acquisition submodule 202; the physical fitness data generation submodule 204 is connected to the evaluation and judgment submodule 203, the physical measurement data acquisition submodule 202 is used to obtain the physical measurement monitoring data of the physical measurement monitoring module 105, the evaluation and judgment submodule 203 is used to judge the acquired physical measurement monitoring data, and analyze the physical measurement monitoring data in combination with factors such as environmental weather, climate and physical condition of the athlete to obtain an analysis result, and the physical fitness data generation submodule 204 is used to generate specific physical measurement data of the athlete based on the analysis result, the physical measurement data acquisition submodule 202 acquires the physical measurement monitoring data of the physical measurement monitoring module 105, and then the evaluation and judgment submodule 203 judges the acquired physical measurement monitoring data, and analyzes the physical measurement monitoring data in combination with factors such as environmental weather, climate and physical condition of the athlete to obtain an analysis result, and then generates specific physical measurement data of the athlete based on the analysis result through the physical fitness data generation submodule 204.

[0057] In the video AI training physical fitness system described in this embodiment, the physical fitness data acquisition submodule 202 acquires the physical fitness monitoring data of the physical fitness monitoring module 105, and then the evaluation and judgment submodule 203 judges the acquired physical fitness monitoring data, analyzes the physical fitness monitoring data in combination with factors such as environmental weather, climate and the physical condition of the athlete to obtain an analysis result, and then generates the athlete's specific physical fitness data based on the analysis result through the physical fitness data generation submodule 204.

[0058] The third embodiment of the present application is:

[0059] See also Figure 6 ,in, Figure 6 It is a schematic diagram of the overall structure of the third embodiment of the present invention.

[0060] On the basis of the first embodiment, the present invention provides a video AI training physical test system, wherein the video AI training physical test system further includes an evaluation data push module 301,

[0061] Among them, the evaluation data push module 301 is connected to the physical fitness evaluation module 201. The evaluation data push module 301 is used to push the evaluated athlete's physical test data to the mobile terminal for the athlete's subsequent reference, so as to facilitate special physical fitness training. The evaluation data push module 301 sends the evaluated physical test data to the mobile terminal based on the Internet, so that the athlete can check the physical test situation after the physical test, and then carry out special physical fitness strengthening training.

[0062] In the video AI training physical fitness test system described in this embodiment, the evaluation data push module 301 sends the evaluated physical fitness test data to the mobile terminal based on the Internet, so that the athletes can check the physical fitness test results after the physical fitness test and then carry out special physical fitness training.

[0063] What is disclosed above is only one or more preferred embodiments of the present application, and cannot be used to limit the scope of rights of the present application. Ordinary technicians in this field can understand that all or part of the processes of implementing the above embodiments and equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A video AI training physical test system, characterized in that: It includes video data acquisition module, data classification module, data storage module, AI training module and physical test monitoring module; The data classification module is connected to the video data acquisition module; the data storage module is connected to the data classification module; the AI ​​training module is connected to the data storage module; the physical measurement monitoring module is connected to the AI ​​training module; The video data acquisition module is used to acquire physical measurement video data based on the big data platform; The data classification module is used to classify the acquired physical measurement video data according to different physical measurement types for subsequent storage; The data storage module is used to store the classified body measurement video data; The AI ​​training module is used to obtain different physical measurement videos stored in the data storage module, and then perform pre-training based on the existing video AI training model to obtain a pre-trained physical measurement monitoring model; The physical fitness monitoring module is used to monitor the physical fitness test based on a pre-trained physical fitness monitoring model.

2. A video AI training physical test system as claimed in claim 1, characterized in that: The video data acquisition module includes a big data platform docking submodule, a video retrieval submodule and a video crawling submodule, wherein the video retrieval submodule is connected to the big data platform docking submodule; the video crawling submodule is connected to the video retrieval submodule; and the video decoding submodule is connected to the video crawling submodule. The big data platform docking submodule is used to connect online with the physical test video search engine to obtain the physical test videos in the physical test database online in real time; The video retrieval submodule is used to search for videos based on keywords of different physical test types through a search engine; The video crawling submodule is used to crawl and search for corresponding physical examination videos.

3. A video AI training physical test system as claimed in claim 2, characterized in that: The data classification module includes a video recognition submodule and a video classification submodule, wherein the video recognition submodule is connected to the video data acquisition module; the video classification submodule is connected to the video recognition submodule; The video recognition submodule is used to recognize the decoded physical measurement video content and recognize different types of physical measurement videos; The video classification submodule is used to classify and stack physical examination videos identified as the same type for subsequent storage.

4. A video AI training physical test system as claimed in claim 3, characterized in that: The AI ​​training module includes a data receiving submodule, a model application submodule, a model training submodule and a model generation submodule, wherein the data receiving submodule is connected to the data storage module; the model application submodule is connected to the data receiving submodule; the model training submodule is connected to the model application submodule; and the model generation submodule is connected to the model training submodule; The data receiving submodule is used to obtain different types of body measurement video data stored in the data storage module; The model application submodule is used to fill different types of body measurement video data into the model library based on the existing video AI training model; The model training submodule is used for training on different types of body measurement videos based on filling; The model generation submodule is used to generate an accurate physical measurement monitoring model under the continuous training of the model training submodule, and has the function of real-time update and improvement to ensure the accuracy of monitoring.

5. The video AI training physical test system according to claim 1, characterized in that: The video AI training physical test system also includes a physical fitness assessment module, which is connected to the physical test monitoring module; The physical fitness assessment module is used to evaluate the physical fitness value based on the physical fitness test data monitored by the physical fitness monitoring module, so as to facilitate physical fitness assessment for different physical conditions.

6. A video AI training physical test system as claimed in claim 5, characterized in that: The physical fitness evaluation module includes a physical fitness data acquisition submodule, an evaluation and judgment submodule and a physical fitness data generation submodule, wherein the physical fitness data acquisition submodule is connected to the physical fitness monitoring module; the evaluation and judgment submodule is connected to the physical fitness data acquisition submodule; and the physical fitness data generation submodule is connected to the evaluation and judgment submodule; The physical measurement data acquisition submodule is used to acquire the physical measurement monitoring data of the physical measurement monitoring module; The evaluation and judgment submodule is used to judge the acquired physical test monitoring data, and analyze the physical test monitoring data in combination with factors such as environmental weather, climate and the physical condition of the athlete to obtain an analysis result; The physical fitness data generation submodule is used to generate specific physical fitness data of the athlete based on the analysis results.

7. A video AI training physical test system as claimed in claim 1, characterized in that: The video AI training physical test system also includes an evaluation data push module, which is connected to the physical fitness evaluation module; The evaluation data push module is used to push the evaluated athlete's physical test data to the mobile terminal for subsequent reference by the athlete, so as to facilitate special physical training.

Images