CN113657237A - Weight lifting motion analysis system based on vision - Google Patents
Weight lifting motion analysis system based on vision Download PDFInfo
- Publication number
- CN113657237A CN113657237A CN202110916792.7A CN202110916792A CN113657237A CN 113657237 A CN113657237 A CN 113657237A CN 202110916792 A CN202110916792 A CN 202110916792A CN 113657237 A CN113657237 A CN 113657237A
- Authority
- CN
- China
- Prior art keywords
- data
- analysis
- motion
- vision
- weight lifting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 65
- 230000033001 locomotion Effects 0.000 title claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 claims abstract description 20
- 230000007246 mechanism Effects 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 8
- 238000007405 data analysis Methods 0.000 claims abstract description 7
- 238000013500 data storage Methods 0.000 claims abstract description 7
- 230000005540 biological transmission Effects 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 21
- 230000008569 process Effects 0.000 claims description 16
- 230000008859 change Effects 0.000 claims description 12
- 230000009471 action Effects 0.000 claims description 11
- 230000001133 acceleration Effects 0.000 claims description 10
- 210000003127 knee Anatomy 0.000 claims description 7
- 230000004927 fusion Effects 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 4
- 239000012141 concentrate Substances 0.000 claims description 2
- 238000012549 training Methods 0.000 description 9
- 230000009286 beneficial effect Effects 0.000 description 8
- 239000013598 vector Substances 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000005484 gravity Effects 0.000 description 4
- 238000007621 cluster analysis Methods 0.000 description 3
- 238000000354 decomposition reaction Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 210000001624 hip Anatomy 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000011480 coordinate descent method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000002650 habitual effect Effects 0.000 description 1
- 210000004394 hip joint Anatomy 0.000 description 1
- 210000000629 knee joint Anatomy 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B21/00—Exercising apparatus for developing or strengthening the muscles or joints of the body by working against a counterforce, with or without measuring devices
- A63B21/06—User-manipulated weights
- A63B21/072—Dumb-bells, bar-bells or the like, e.g. weight discs having an integral peripheral handle
- A63B21/075—Dumb-bells, bar-bells or the like, e.g. weight discs having an integral peripheral handle with variable weights, e.g. weight systems with weight selecting means for bar-bells or dumb-bells
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0062—Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0062—Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance
- A63B2024/0071—Distinction between different activities, movements, or kind of sports performed
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Physical Education & Sports Medicine (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Orthopedic Medicine & Surgery (AREA)
- Probability & Statistics with Applications (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention provides a weight lifting motion analysis system based on vision, which comprises the following components: the system comprises data acquisition equipment, an operation interface, an analysis module and a data storage module; the data acquisition equipment acquires the front and side video data of the athlete and transmits the data to an operation interface; the operation interface displays data and transmits the data to the analysis module through a transmission protocol; the analysis module preprocesses data through a space attention mechanism SAM, realizes athlete key point tracking and various motion data analysis by utilizing a deep neural network, and stores the data to the data storage module; and the analysis module forms an analysis report according to the key points and various motion data and transmits the analysis report to the operation interface. The invention greatly reduces the operation amount, fills the gap in similar application of motion analysis, and supports concurrent processing of simultaneous operation of multiple users by adopting containerization deployment.
Description
Technical Field
The invention relates to the field of motion analysis, in particular to a weight lifting motion analysis system based on vision.
Background
The sports data analysis industry is mainly focused on information statistics such as scores and tactics of multi-person projects at present, a small number of single-person motion analysis systems using computer vision are focused on a motion video editing function, and accurate data are not acquired and are not combined with kinematic analysis.
The existing motion analysis system achieves action decomposition through video unframing in a single project, and achieves track reappearance through timing sequence key point marking and smooth curve drawing.
The disadvantages of the method are as follows: decomposing the motion according to frames increases the work load of the practical coach, and the coach needs to pay the same attention to each frame to observe the motion change; the real time sequence labeling of the key points to draw the track is that the user draws on the video instead of the real tracking and displaying of the key points, and the difference between the drawn curve and the real track is large.
Meanwhile, the conventional single-person motion analysis system often uses a traditional video analysis method, has the problems of being not accurate enough, being incapable of automatically positioning an analysis object and the like, only feeds back measurement data, does not perform data fusion analysis to generate a report, and is not intuitive and difficult to analyze for a user.
Disclosure of Invention
The invention belongs to an image detection technology based on a computer vision technology. The computer vision is used for simulating the function of human vision, and the image is acquired, processed, calculated, and finally actually detected, controlled and applied from a specific real object.
The invention provides a weight lifting motion analysis system based on vision, which comprises the following steps:
the system comprises data acquisition equipment, an operation interface, an analysis module and a data storage module;
the data acquisition equipment acquires the front and side video data of the athlete and transmits the data to an operation interface; the operation interface displays data and transmits the data to the analysis module through a transmission protocol;
the analysis module preprocesses data through a space attention mechanism SAM and realizes player key point tracking and various motion data analysis by utilizing a deep neural network;
and the analysis module forms an analysis report according to the key points and various motion data and transmits the analysis report to the operation interface.
Further, the transmission protocol comprises MQTT and FTP protocols.
Further, the pretreatment specifically comprises: the spatial attention mechanism SAM is used to select the subject of each frame of the data.
Further, the method utilizes a deep neural network to realize the tracking of the key points of the athlete, and specifically comprises the following steps: extracting features of each frame of picture main body by adopting VggNet, identifying 18 key points of a human body by combining context semantics, and determining the relative position of each frame of picture where the key points are located; and storing two-dimensional information data by adopting a key point thermodynamic diagram PCM representation.
Further, the items of motion data include: knee angle, arm angle, barbell acceleration, and center acceleration.
Further, each item of motion data is obtained through key points and preset acquisition equipment position information.
Further, the motion data analysis includes: the athlete can concentrate the strength of the whole body when getting out of arm, begin and end the action of the second part of lifting, tighten the body at the end of the lifting movement and begin and end the action of putting down the barbell.
The forming process of the analysis report specifically comprises the following steps: a model is built through time sequence change of various motion data, the model is combined with historical data to serve as input of a motion state analysis neural network, and multi-dimensional data fusion is achieved by utilizing the signal processing capacity of the artificial neural network.
The beneficial effects provided by the invention are as follows:
1. the main body selection preprocessing is carried out on the video through a space attention mechanism, so that the operation amount is reduced to the great extent;
2. the real track of key points in the motion process of the athlete is generated by the posture estimation technology, and the gap in similar application is filled;
3. combining the reverse dynamics and the weightlifting motion theory, and increasing the motion decomposition process to a unit of motion process instead of a unit of video frame;
4. carrying out multi-dimensional key information fusion of weightlifting motion through a neural network, and carrying out kinematic analysis by using data acquired from a video and historical data;
5. the system adopts containerization deployment and supports concurrent processing of simultaneous operation of multiple users.
Drawings
FIG. 1 is a block diagram of a vision-based weight lifting motion analysis system of the present invention;
FIG. 2 is a graph showing the trend of knee angle change with time at various stages of weight lifting according to the present invention;
fig. 3 is a graph showing the change of hip angle with time in each stage of weight lifting.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be further described with reference to the accompanying drawings.
Referring to fig. 1, a system for analyzing weight lifting movement based on vision includes the following:
the system comprises data acquisition equipment, an operation interface, an analysis module and a data storage module;
the data acquisition equipment acquires the front and side video data of the athlete and transmits the data to an operation interface;
the operation interface displays data and transmits the data to the analysis module through a transmission protocol;
the analysis module preprocesses data through a space attention mechanism SAM and realizes player key point tracking and various motion data analysis by utilizing a deep neural network;
and the analysis module forms an analysis report according to the key points and various motion data and transmits the analysis report to the operation interface.
As an embodiment, the system is based on a body segmentation technology of an attitude estimation technology, inverse dynamics, a weightlifting motion analysis theory, an artificial neural network and a space attention mechanism.
The data acquisition equipment collects videos of the front and the side of the weight lifting athlete, and a user uploads the videos to the analysis module through an MQTT and FTP protocol by using an operation interface. The system supports the data interaction of external network users, namely the data interaction does not need to be carried out under the same local area network.
The analysis module obtains the video and then preprocesses the video, and selects a subject to be analyzed in each frame of picture-a region range of the action of the weightlifting athlete through a Spatial Attention Mechanism (SAM).
Extracting the characteristics of each frame of main body Part through a deep neural network VggNet, identifying 18 key points of the human body by combining context semantics, determining the relative positions of the key points of each Part of the human body in each frame, representing and storing two-dimensional information data by using a PCM (Part consistency map, key point thermodynamic diagram), and realizing the tracking of the key points of athletes; a line drawing function is embedded in an analysis algorithm, the track is drawn on a user-defined mask by taking a frame as a unit, and the user-defined tracking display of the track of the key points of the athlete is supported on an operation interface.
According to the key point data and the position information of the collection equipment preset by the system, the motion data of the athlete such as the knee included angle, the arm included angle, the barbell acceleration, the center acceleration and the like in the whole weight lifting process are calculated by combining reverse dynamics, various data information are calibrated by the preset focal length of the collection equipment and the relative position of the equipment and the athlete, and the data information is stored. In the embodiment of the invention, the coordinate descent method is adopted to acquire various motion data.
Based on the kinematic characteristics of weight lifting, the data of the included angle of the arm and the included angle of the knee are combined and analyzed to judge the beginning and the end of the first part of the lifting action, and the concentration degree of the whole body strength of the athlete in lifting is judged by comparing the independent change and the relative change of the time sequence of the two and the time coincidence degree of the maximum angle; through the data combination analysis of the zero acceleration and the gravity center acceleration of the bar, the strength and the tension action time of the exerciser in the process of lifting the barbell are deduced, and the beginning and the end of the action of the second part of the weight lifting are judged; and analyzing the height change of the barbell independently to judge the start and end of actions of tightening the body at the end of the lifting movement and putting down the barbell.
Specifically, the first part of the jack-up starts from the beginning to the end of the athlete squat to the lowest point (point 3), for a total of 6 points:
the time difference between the point 1 (the point where the gravity center coincides with the height of the barbell) and the point 2 (the highest point of the barbell in the first stage) is in negative correlation with the continuous pulling force in the process (the lifting speed of the barbell in the time difference needs to be counted, so that the force is reversely exerted); the included angles of the big arm and the small arm are in negative correlation with the concentration degree of the whole body force at the point 3; point 4 is the barbell and body center of gravity stabilization point in the knee-leading and bell-lifting stage, and the first stage is finished (from preparation to knee-leading and bell-lifting); point 5 is the key point of the second part of action, the starting point of the force stage; point 6 is the end of the force phase, representing the beginning of the inertia up phase; and the time difference between the two points is in negative correlation with the continuous thrust in the process;
referring to FIGS. 2-3, FIG. 2 is a graph showing the trend of knee angle change with time at various stages of weight lifting according to the present invention; fig. 3 is a graph showing the change of hip angle with time in each stage of weight lifting.
And constructing a model by using the calculation results of the data of each part (the height of the barbell, the acceleration of the knee, the included angle of the arm and the central acceleration) in time sequence change, combining the model with historical data to be used as the input of a motion state analysis neural network, and realizing multi-dimensional data fusion by using the signal processing capability and the automatic reasoning function of an artificial neural network to generate a feedback report for a user.
The present invention specifically explains the process as follows:
the invention uses the correlation algorithm to obtain various indexes which have influence on the performance of the athletes, and the indexes are called as 'beneficial indexes'.
In addition, the data acquisition system of the invention can obtain beneficial indexes of each training, the indexes can be used as a multi-dimensional vector, and each training video can obtain the vector.
The invention uses a cluster analysis algorithm and a neural network to predict the value of the beneficial index when the athlete reaches the best performance, and the value can be used as a training target to assist the athlete to improve.
The principle is that the value of the beneficial index of each training of the athlete is not completely unchanged, and the fluctuation of some values can cause the change of the performance. We believe that similar floating trends may yield similar results (better or worse). Therefore, beneficial indexes obtained by multiple times of training can be classified, a class which can make the performance become good is taken out, neural network fitting is carried out, a multidimensional vector is extracted from the class of data, and the vector can be considered as a due index when the athlete can achieve the best performance.
The invention acquires data such as the gravity center height, the barbell speed, the included angle between the knee joint and the hip joint and the like of historical training videos of athletes to obtain a plurality of multidimensional vectors, and each index of each vector represents each acquired data in a training process. We can then use the K-means cluster analysis algorithm to calculate the distance between vectors using Euclidean distances, and classify all data into K classes. The value of K may be selected in conjunction with the classification. The K types have similar beneficial indexes and represent a weight lifting mode.
The invention counts the average weight-lifting achievement of each type, and takes out the type with the best average achievement for the following continuous analysis. Finally, the mode represented by the index is considered to be easier to achieve.
As the training of athletes possibly has a certain habitual defect, the invention introduces the first three match data in each big-sports event final competition, each match data is regarded as a multi-dimensional vector and becomes the champion data, the champion data and the class with the optimal average performance obtained in the previous step are subjected to K-mean cluster analysis again, the average weight lifting performance of each class is counted again, and the class with the optimal average performance is taken out and is called as the optimal class.
And then, discrete data serialization is carried out in the optimal class by utilizing a spline interpolation method. Because the density of points in a class is high, the error of the continuously obtained data is small. And finally, well fitting the continuous key data distribution condition in the optimal class and the corresponding prediction achievement thereof. The continuous class becomes a continuous optimal class.
In the invention, the artificial Neural Network performs optimization calculation in a continuous optimal class by utilizing a LeNet-AlexNet-VGG Network in an RNN (Current Neural Network). The weighted gradient is decreased by utilizing a RMSProp (root mean square transfer) algorithm, and the point with the highest performance in the optimal class is obtained in the process, so that the optimal performance which can be achieved by the athlete can be predicted to reach each key data index of the performance.
After the processing is finished, the analysis module informs a user that the system is finished through an MQTT protocol, and sends a data processing result and the line drawing mask information to an operation interface, so that the user can display the analysis result through the operation interface and support to custom check each item of data and each key point track.
In the invention:
1. data acquisition equipment for acquiring video information
2. The operation interface is used for uploading and playing back videos and presenting analysis reports
3. The analysis module is used for acquiring data information of the athlete from the video, analyzing the training process of the athlete by combining a kinematics theory and generating a report;
4. the data storage module is used for storing historical data and historical analysis results of the athletes.
The invention has the beneficial effects that:
1. the main body selection preprocessing is carried out on the video through a space attention mechanism, so that the operation amount is reduced to the great extent;
2. the real track of key points in the motion process of the athlete is generated by the posture estimation technology, and the gap in similar application is filled;
3. combining the reverse dynamics and the weightlifting motion theory, and increasing the motion decomposition process to a unit of motion process instead of a unit of video frame;
4. carrying out multi-dimensional key information fusion of weightlifting motion through a neural network, and carrying out kinematic analysis by using data acquired from a video and historical data;
5. the system adopts containerization deployment and supports concurrent processing of simultaneous operation of multiple users.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (8)
1. A vision-based weight lifting motion analysis system, characterized by: the method comprises the following steps:
the system comprises data acquisition equipment, an operation interface, an analysis module and a data storage module;
the data acquisition equipment acquires the front and side video data of the athlete and transmits the data to an operation interface;
the operation interface displays data and transmits the data to the analysis module through a transmission protocol;
the analysis module preprocesses data through a space attention mechanism SAM, realizes athlete key point tracking and various motion data analysis by utilizing a deep neural network, and stores the data to the data storage module;
and the analysis module forms an analysis report according to the key points and various motion data and transmits the analysis report to the operation interface.
2. A vision-based weight lifting motion analysis system as claimed in claim 1 wherein: the transmission protocol comprises MQTT and FTP protocols.
3. A vision-based weight lifting motion analysis system as claimed in claim 1 wherein: the pretreatment specifically comprises the following steps: the spatial attention mechanism SAM is used to select the subject of each frame of the data.
4. A vision-based weight lifting motion analysis system as claimed in claim 3 wherein: the method for realizing the tracking of the key points of the athletes by utilizing the deep neural network specifically comprises the following steps: extracting features of each frame of picture main body by adopting VggNet, identifying 18 key points of a human body by combining context semantics, and determining the relative position of each frame of picture where the key points are located; and storing two-dimensional information data by adopting a key point thermodynamic diagram PCM representation.
5. A vision-based weight lifting motion analysis system as claimed in claim 1 wherein: the items of motion data include: knee angle, arm angle, barbell acceleration, and center acceleration.
6. A vision-based weight lifting motion analysis system as claimed in claim 5 wherein: each item of motion data is obtained through key points and preset acquisition equipment position information.
7. A vision-based weight lifting motion analysis system as claimed in claim 1 wherein: the motion data analysis includes: the athlete can concentrate the strength of the whole body when getting out of arm, begin and end the action of the second part of lifting, tighten the body at the end of the lifting movement and begin and end the action of putting down the barbell.
8. A vision-based weight lifting motion analysis system as claimed in claim 1 wherein: the forming process of the analysis report specifically comprises the following steps: a model is built through time sequence change of various motion data, the model is combined with historical data to serve as input of a motion state analysis neural network, and multi-dimensional data fusion is achieved by utilizing the signal processing capacity of the artificial neural network.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110916792.7A CN113657237A (en) | 2021-08-11 | 2021-08-11 | Weight lifting motion analysis system based on vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110916792.7A CN113657237A (en) | 2021-08-11 | 2021-08-11 | Weight lifting motion analysis system based on vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113657237A true CN113657237A (en) | 2021-11-16 |
Family
ID=78479464
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110916792.7A Pending CN113657237A (en) | 2021-08-11 | 2021-08-11 | Weight lifting motion analysis system based on vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113657237A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140066201A1 (en) * | 2012-08-31 | 2014-03-06 | Blue Goji Corp. | Mobile and Adaptable Fitness System |
CN106709453A (en) * | 2016-12-24 | 2017-05-24 | 北京工业大学 | Sports video key posture extraction method based on deep learning |
RU2658255C1 (en) * | 2017-07-21 | 2018-06-19 | Анатолий Александрович Шалманов | Device for biomechanical control of technical preparedness and physical fitness of weightlifters |
CN111680608A (en) * | 2020-06-03 | 2020-09-18 | 长春博立电子科技有限公司 | Intelligent sports auxiliary training system and training method based on video analysis |
CN113177455A (en) * | 2021-04-23 | 2021-07-27 | 中国科学院计算技术研究所 | Method and system for identifying exercise intensity |
-
2021
- 2021-08-11 CN CN202110916792.7A patent/CN113657237A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140066201A1 (en) * | 2012-08-31 | 2014-03-06 | Blue Goji Corp. | Mobile and Adaptable Fitness System |
CN106709453A (en) * | 2016-12-24 | 2017-05-24 | 北京工业大学 | Sports video key posture extraction method based on deep learning |
RU2658255C1 (en) * | 2017-07-21 | 2018-06-19 | Анатолий Александрович Шалманов | Device for biomechanical control of technical preparedness and physical fitness of weightlifters |
CN111680608A (en) * | 2020-06-03 | 2020-09-18 | 长春博立电子科技有限公司 | Intelligent sports auxiliary training system and training method based on video analysis |
CN113177455A (en) * | 2021-04-23 | 2021-07-27 | 中国科学院计算技术研究所 | Method and system for identifying exercise intensity |
Non-Patent Citations (5)
Title |
---|
KRISTOF KIPP: "Predicting net joint moments during a weightlifting exercise with a neural network model", JOURNAL OF BIOMECHANIC, 30 June 2018 (2018-06-30) * |
吴昊;王修信;郝艳;梁冬冬;蒋金艳;刘海克;于秋萍;蒋炳德;: "举重序列图像杠铃动力学参数的提取与分析", 广西物理, no. 01, 15 March 2008 (2008-03-15) * |
白雪岭;王洪生;张希安;季文婷;魏高峰;王成焘;: "三维运动学仿真男子举重运动员抓举技术及膝关节运动分析", 中国组织工程研究与临床康复, no. 35, 27 August 2009 (2009-08-27) * |
邓宇;刘国翌;李华;: "基于视频的杠铃轨迹跟踪与分析系统", 中国图象图形学报, no. 12, 30 December 2006 (2006-12-30) * |
韩美林;: "基于BP神经网络的体育视频关键姿态检测", 商洛学院学报, no. 06, 20 December 2019 (2019-12-20) * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11544928B2 (en) | Athlete style recognition system and method | |
Theagarajan et al. | Soccer: Who has the ball? Generating visual analytics and player statistics | |
US10186041B2 (en) | Apparatus and method for analyzing golf motion | |
JP2014522035A (en) | Object posture search apparatus and method | |
Edwards et al. | From pose to activity: Surveying datasets and introducing CONVERSE | |
Vantigodi et al. | Real-time human action recognition from motion capture data | |
CN101826155B (en) | Method for identifying act of shooting based on Haar characteristic and dynamic time sequence matching | |
CN106846372B (en) | Human motion quality visual analysis and evaluation system and method thereof | |
Hsu et al. | Coachai: A project for microscopic badminton match data collection and tactical analysis | |
CN110705418B (en) | Taekwondo kicking motion video capture and scoring system based on deep LabCut | |
Yang et al. | Research on face recognition sports intelligence training platform based on artificial intelligence | |
Doughty et al. | Who’s better, who’s best: Skill determination in video using deep ranking | |
CN113975775A (en) | Wearable inertial body feeling ping-pong exercise training system and working method thereof | |
Yan et al. | A review of basketball shooting analysis based on artificial intelligence | |
EP4176798A1 (en) | Monitoring exercise activity in a gym environment | |
Mavroudi et al. | Deep moving poselets for video based action recognition | |
CN113657237A (en) | Weight lifting motion analysis system based on vision | |
Zeng et al. | Deep learning approach to automated data collection and processing of video surveillance in sports activity prediction | |
CN114882580A (en) | Measuring method for motion action consistency based on deep learning | |
Wu | Image Recognition of Standard Actions in Sports Videos Based on Feature Fusion. | |
Klempous et al. | Neural networks classification for training of five German longsword mastercuts-a novel application of motion capture: analysis of performance of sword fencing in the historical European martial arts (HEMA) domain | |
He et al. | Recognition and prediction of badminton attitude based on video image analysis | |
Malawski et al. | Automatic analysis of techniques and body motion patterns in sport | |
CN115273243B (en) | Fall detection method, device, electronic equipment and computer readable storage medium | |
Neskorodieva et al. | Real-time Classification, Localization and Tracking System (Based on Rhythmic Gymnastics) |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |