CN114783001A - Swimming posture evaluation method, system, device and computer readable storage medium - Google Patents
Swimming posture evaluation method, system, device and computer readable storage medium Download PDFInfo
- Publication number
- CN114783001A CN114783001A CN202210708412.5A CN202210708412A CN114783001A CN 114783001 A CN114783001 A CN 114783001A CN 202210708412 A CN202210708412 A CN 202210708412A CN 114783001 A CN114783001 A CN 114783001A
- Authority
- CN
- China
- Prior art keywords
- swimming
- coordinates
- dimensional
- posture
- camera
- 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
- 230000009182 swimming Effects 0.000 title claims abstract description 64
- 238000011156 evaluation Methods 0.000 title description 16
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000001914 filtration Methods 0.000 claims abstract description 20
- 239000013598 vector Substances 0.000 claims abstract description 19
- 238000013528 artificial neural network Methods 0.000 claims abstract description 10
- 238000012937 correction Methods 0.000 claims abstract description 7
- 238000010586 diagram Methods 0.000 claims description 21
- 238000004422 calculation algorithm Methods 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- 210000001503 joint Anatomy 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 210000002310 elbow joint Anatomy 0.000 claims description 6
- 210000003127 knee Anatomy 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 210000004394 hip joint Anatomy 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 230000001629 suppression Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000033001 locomotion Effects 0.000 claims description 3
- 238000003702 image correction Methods 0.000 claims 2
- 238000012797 qualification Methods 0.000 claims 1
- 210000003423 ankle Anatomy 0.000 abstract description 3
- 238000013135 deep learning Methods 0.000 abstract description 3
- 230000007613 environmental effect Effects 0.000 abstract description 3
- 238000011160 research Methods 0.000 abstract description 3
- 230000006872 improvement Effects 0.000 abstract description 2
- 230000036544 posture Effects 0.000 description 58
- 230000009471 action Effects 0.000 description 10
- 238000004590 computer program Methods 0.000 description 7
- 230000009466 transformation Effects 0.000 description 6
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000012821 model calculation Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/08—Learning methods
-
- G06T3/08—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/20—Linear translation of a whole image or part thereof, e.g. panning
Abstract
The invention discloses a swimming posture assessment method, a system, a device and a computer readable storage medium, belonging to the technical field of swimming posture assessment, comprising the following steps: acquiring a swimming image, recognizing human body posture data and generating two groups of coordinates; establishing a three-dimensional attitude model; filtering anchor point coordinates generated in the same scene to convert two-dimensional coordinates into three-dimensional coordinates; generating three-dimensional vectors from the three-dimensional coordinates, and calculating included angles among the vectors; the invention sets the joint angle threshold of the swimmer, judges whether the swimming gesture is qualified or not, continuously and deeply researches gesture recognition along with the improvement of deep learning and operation speed, marks the joint parts of elbows, ankles, heads and the like by shooting the gesture of a human body, establishes a neural network to train the marked pictures after generating a marked picture, generates a model and provides a corresponding correction tool to improve the recognition efficiency under different environmental backgrounds.
Description
Technical Field
The invention belongs to the technical field of swimming posture assessment, and particularly relates to a swimming posture assessment method, a swimming posture assessment system, a swimming posture assessment device and a computer readable storage medium.
Background
In daily swimming exercise, the correct action posture can not only improve the movement efficiency, but also protect the abrasion of human joints and other parts, so that the correct guidance is particularly important, and in the actual swimming process, due to the refraction of water, a coach is difficult to quantitatively evaluate the action of a swimmer, and the swimming posture is difficult to reach a uniform standard;
for the judgment of swimming postures, the existing method is basically a new auxiliary belt mode with old belts, and each coach and referee has a teaching method, but in the actual teaching process, the situation that the actions of the trainees are not in accordance with the standard textbook standard can be caused, so that a set of standards is defined through the joint angle, and the problem that the actions of the swimmers cannot be quantitatively evaluated is solved.
Disclosure of Invention
The invention aims to provide a swimming posture assessment method, a swimming posture assessment system, a swimming posture assessment device and a computer readable storage medium, which solve the problem that the action of a swimmer cannot be quantitatively assessed.
In order to achieve the purpose, the invention provides the following technical scheme: a swimming stroke assessment method includes:
acquiring a swimming image, identifying human body posture data, and generating two groups of coordinates;
establishing a three-dimensional attitude model;
filtering anchor point coordinates generated in the same scene to convert two-dimensional coordinates into three-dimensional coordinates;
generating three-dimensional vectors from the three-dimensional coordinates, and calculating included angles among the vectors;
and setting a joint angle threshold of the swimmer, and judging whether the swimming posture is qualified.
Preferably, use wide-angle camera image rectification algorithm to correct the image that positive side camera was afferent after acquireing swimming image, wide-angle camera image rectification algorithm includes:
setting radial distortion parameters
Tangential distortion parameter
Camera parameters
In the radial direction according to the formula
，
Tangentially according to the formula
，
And calculating a corrected image coordinate, wherein the radial distortion parameter, the tangential distortion parameter and the camera parameter are obtained through a correction function.
Preferably, the recognizing the human body gesture comprises:
generating a joint thermodynamic diagram from the corrected image, calculating the offset and the relative offset position of the joint relative to the coordinate, establishing a neural network framework, receiving the thermodynamic diagram and the joint offset as a training set by the network framework, supervising the training process and marking anchor point diagram positions of 33 joints after identification; the neural network is a residual error network, all layers are connected by using a jump mode and a regression method, and the training set uses a more shielded graph position.
Preferably, the image-generating joint thermodynamic diagram comprises:
the recognition sequence starts from the face of a human body, the face is recognized through a non-maximum suppression algorithm, the framed area is cut, and then a trained model is called to generate a coordinate result.
Preferably, the anchor coordinate filtering includes:
coordinate filtering by a filtering algorithm comprising
Combining the generated anchor point information of 33 joints to generate a 66 x N matrix, wherein N represents the frame number of the motion, and performing noise reduction processing on the combined data, wherein the formula is as follows:
in the formula
、
Is the pixel coordinates of the first camera head,
、
、
is a projection matrix of the camera on three coordinate planes,
the three-dimensional plane projection is carried out, and the analogy is carried out to obtain the pixel coordinate of the second camera
、
A is a coordinate matrix, and the coordinate matrix is calculated by the following formula;
preferably, the method of converting two-dimensional coordinates into three-dimensional coordinates includes converting by direct linear transformation:
the generated two-dimensional plane coordinates are directly and linearly transformed, the formula is shown as follows,
is the coordinates of the control points of each point,
coordinates on the image corresponding to the image points of the three-dimensional control point:
wherein
、
、
、
、
、
、
、
、
、
、
And determining parameters according to the parameters of the camera and the underwater environment parameters.
Preferably, the data for determining the qualified swimming posture comprises:
obtaining three-dimensional coordinates of each joint point
，
Then, the included angle of each joint is obtained in the three-dimensional space, and a three-dimensional vector is set according to the starting point of each joint
And calculating an included angle:
and obtaining eight joint included angles which are respectively included angles of a left elbow joint, a right elbow joint, a left arm, a right arm, a trunk, a left knee, a right knee and a left hip joint, defining four grades of excellence, good, qualified and unqualified through preset judgment conditions, and evaluating the posture of the swimmer.
The invention provides a swimming posture evaluation system based on a multi-angle camera, which comprises:
the acquisition module is used for acquiring a swimming image;
the recognition module is used for recognizing human body gestures;
the generating module is used for generating two groups of coordinates according to the recognized human body posture data;
the three-dimensional posture model is used for training the recognized human body posture data;
filtering anchor coordinates, namely filtering the anchor coordinates generated in the same scene;
the conversion module is used for converting the two-dimensional coordinates into three-dimensional coordinates;
the generating module is used for generating a three-dimensional vector from the three-dimensional coordinate;
the included angle calculation module is used for calculating the included angle between the vectors;
the judgment threshold setting module is used for setting a judgment threshold;
and the judging module is used for converting the angle and the posture into good and bad after the swimming posture is set to be multi-terminal, and judging the qualified swimming posture data.
The invention provides a swimming posture evaluation device based on a multi-angle camera, which comprises:
the at least two cameras are used for acquiring a front image and a side image of a human body;
a memory for storing non-transitory computer readable instructions; and
and the processor is used for operating the computer readable instructions, so that the computer readable instructions can realize the swimming posture evaluation method based on the multi-angle camera when being executed by the processor.
The present invention provides a computer-readable storage medium for storing non-transitory computer-readable instructions, which, when executed by a computer, cause the computer to perform the above-described multi-angle camera-based swimming stroke assessment method.
The invention has the technical effects and advantages that: the swimming posture evaluation method, the system, the device and the computer readable storage medium deeply research posture recognition through deep learning and operation rate, mark joint parts such as elbows, ankles, heads and the like by shooting human body postures, build a neural network to train the marked pictures after generating a marked picture, generate a model and provide corresponding correction tools to improve recognition efficiency under different environmental backgrounds;
acquiring an underwater swimmer image through a front camera and a side camera, correcting the acquired wide-angle image, calling an algorithm model to realize gesture recognition, generating two groups of coordinates, synchronizing the two groups of coordinates through direct linear transformation and a filtering algorithm, thereby obtaining human body joint coordinates in a three-dimensional plane, acquiring joint angles, and realizing underwater action (such as swimming) gesture evaluation through setting a set threshold; the integral identification method has the characteristics of high accuracy and moderate model calculation amount;
a set of standards are defined through the joint angles, so that the posture of the swimmer can be judged, the swimmer can be better helped to master action requirements, and more excellent technical breakthrough is realized.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of the present invention;
FIG. 2 is an anchor point diagram for human gesture recognition in accordance with the present invention;
FIG. 3 is a schematic view of the camera head of the present invention mounted thereon;
FIG. 4 is a three-dimensional synchronized human body posture diagram of the present invention;
FIG. 5 is a block flow diagram of an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
The invention provides a swimming posture assessment method based on a multi-angle camera, which is shown in figures 1 and 5 and comprises the following steps:
s10, obtaining swimming images, recognizing human body posture data and generating two groups of coordinates; in the embodiment, the images of the front side camera and the side camera are read through the USB cable to carry out wide-angle correction so as to identify the posture coordinates of the human body in the images, two groups of coordinates are respectively generated, and the images transmitted by the front side camera and the side camera are corrected by defining radial distortion parameters and tangential distortion parameters, so that the posture identification coordinate is prevented from being distorted;
setting radial distortion parameters
Tangential distortion parameter
Camera parameters
In the radial direction according to the formula
，
Tangentially according to the formula
，
And calculating and correcting image coordinates, wherein the radial distortion parameter, the tangential distortion parameter and the camera parameter are obtained through a correction function.
It should be noted that in the present embodiment, as shown in fig. 3, acquiring a swimming image includes two cameras taken from the side and the front, wherein one camera takes the swimmer posture from the side position and one camera takes the swimmer posture from the head; the camera is connected with the recognition system through a USB cable to complete image transmission, and coordinates output by the camera are converted from two-dimensional coordinates to three-dimensional coordinates through direct linear transformation; in the embodiment, the model of the camera is H65V1, PIXEL SIZE is 720P/2.23MM, or 720P/AHD;
in the process of recognizing human body posture data, an introduced image training residual error neural network frame is used, a maximum value suppression algorithm is used, firstly, characteristic regions such as human faces are found out in pictures, a model is called after the whole human body is extracted, posture recognition is achieved, and human body anchor point coordinates are generated; aiming at human body posture recognition, a joint thermodynamic diagram is generated, offset of each joint relative to coordinates is calculated, relative offset positions are calculated, a neural network framework is established, the network framework receives the thermodynamic diagram and the joint offset as a training set, the training process is supervised, and anchor point positions of 33 joints are marked after recognition, as shown in fig. 2.
The neural network is a residual error network, and all layers are connected by adopting a jump mode and a regression method so as to improve the feature extraction efficiency of different levels. And the training set adopts the picture that shelters from more picture position, like the arm alternately embraces in the front of the chest, whole people squat down, and then increases network identification efficiency, improves the degree of accuracy.
In the identification process, firstly, a human body is identified, because the face is relatively easy to detect, the face is identified by adopting a non-maximum suppression algorithm, and a trained model is called to generate a coordinate result after a selected area is cut;
s20, establishing a three-dimensional attitude model;
s30, filtering anchor point coordinates generated in the same scene to convert the two-dimensional coordinates into three-dimensional coordinates;
coordinate filtering is carried out on anchor point coordinates generated by cameras with different angles on the same scene through a singular value decomposition algorithm, and a two-dimensional coordinate is converted into a three-dimensional coordinate through a direct linear transformation algorithm;
s40, generating three-dimensional vectors from the three-dimensional coordinates, and calculating included angles among the vectors; generating three-dimensional vectors through three-dimensional coordinates, calculating included angles among the vectors, and calculating angles of human joints in a three-dimensional space to generate a result; as shown in fig. 4;
when filtering anchor coordinates, because of having two cameras of different angles, the anchor information of 33 joints that will produce is made up, produces a 66X N matrix, and N stands for the frame number of action, carries on the noise reduction to the combined data, the formula is as follows:
in the formula
、
Is the pixel coordinates of the first camera,
、
、
is a projection matrix of the camera on three coordinate planes,
the three-dimensional plane projection is carried out, and the analogy is carried out to obtain the pixel coordinate of the second camera
、
A is a coordinate matrix, and the coordinate matrix is calculated by the following formula;
direct linear transformation:
after the noise reduction processing is carried out, the generated two-dimensional plane coordinate is directly and linearly transformed, the formula is shown as follows,
is the coordinates of the control points of each point,
coordinates on the image corresponding to the image points of the three-dimensional control point:
wherein
、
、
、
、
、
、
、
、
、
、
And determining parameters according to the parameters of the camera and the underwater environment parameters.
S50, obtaining three-dimensional coordinates of each joint point
，
Then, the included angle of each joint is obtained in the three-dimensional space, and a three-dimensional vector is set according to the starting point of each joint
And calculating an included angle:
after the included angles of all joints are obtained, the included angles are eight included angles which are respectively the included angles of the left elbow joint, the right elbow joint, the left arm and the trunk, the included angles of the left knee and the right knee, and the included angles of the left hip joint and the right hip joint, four grades of excellence, good, qualified and unqualified are defined through preset judgment conditions, and the posture of the swimmer is evaluated. The swimming posture is divided into a plurality of ends by setting a judgment threshold value, the conversion of the angle and the posture is realized according to the set threshold value, the posture evaluation result can be stored, and after the evaluation is finished, the posture of the swimmer is displayed and stored.
The invention provides a swimming posture evaluation system based on a multi-angle camera, which comprises:
the acquisition module is used for acquiring a swimming image;
the recognition module is used for recognizing human body gestures;
the generating module is used for generating two groups of coordinates from the recognized human body posture data;
the three-dimensional posture model is used for training the recognized human body posture data;
filtering anchor coordinates, namely filtering the anchor coordinates generated in the same scene;
the conversion module is used for converting the two-dimensional coordinates into three-dimensional coordinates;
the generating module is used for generating a three-dimensional vector from the three-dimensional coordinate;
the included angle calculation module is used for calculating the included angle between the vectors;
the judgment threshold setting module is used for setting a judgment threshold;
and the judging module is used for converting the angle and the posture into good and bad after the swimming posture is set to be multi-terminal, and judging the qualified swimming posture data.
The invention provides a swimming posture evaluation device based on a multi-angle camera, which comprises:
the at least two cameras are used for acquiring a front image and a side image of a human body;
a memory for storing non-transitory computer readable instructions; and
and the processor is used for operating the computer readable instructions, so that the computer readable instructions can realize the swimming posture evaluation method based on the multi-angle camera when being executed by the processor.
The present invention provides a computer-readable storage medium for storing non-transitory computer-readable instructions, which, when executed by a computer, cause the computer to perform the above-described multi-angle camera-based swimming stroke assessment method.
With the improvement of deep learning and operation rate, the swimming posture evaluation method, the swimming posture evaluation system, the swimming posture evaluation device and the computer readable storage medium continuously and deeply research on posture recognition, marks joint parts such as elbows, ankles, heads and the like by shooting human postures, establishes a neural network to train the marked pictures after generating a marked picture, generates a model, and provides corresponding correction tools to improve recognition efficiency under different environmental backgrounds.
The underwater swimmer image can be obtained through the front camera and the side camera, the obtained wide-angle image is corrected, the gesture recognition is realized by calling an algorithm model, two groups of coordinates are generated, the two groups of coordinates are synchronized through direct linear transformation and singular value decomposition, so that the human body joint coordinates in a three-dimensional plane are obtained, the joint angle is obtained, the underwater action such as swimming gesture evaluation is realized by setting a set threshold value, and the overall recognition method has the characteristics of high accuracy and moderate model calculation amount;
a set of standards are defined through the joint angles, so that the posture of the swimmer can be judged, the swimmer can be better helped to master action requirements, and more excellent technical breakthrough is realized.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments or portions thereof without departing from the spirit and scope of the invention.
Claims (10)
1. A swimming posture assessment method based on a multi-angle camera is characterized by comprising the following steps: the method comprises the following steps:
acquiring a swimming image for recognizing human body posture data to generate two groups of coordinates;
establishing a three-dimensional attitude model;
filtering anchor point coordinates generated in the same scene to convert the two-dimensional coordinates into three-dimensional coordinates;
generating three-dimensional vectors from the three-dimensional coordinates, and calculating included angles among the vectors;
and setting a joint angle threshold of the swimmer, and judging whether the swimming posture is qualified.
2. The swimming posture assessment method based on the multi-angle camera as claimed in claim 1, wherein: after obtaining the swimming image, use wide-angle camera image correction algorithm to correct the image that positive side camera was afferent, wide-angle camera image correction algorithm includes:
setting radial distortion parameters
Tangential distortion parameter
Camera parameters
In the radial direction according to the formula
，
Tangentially according to the formula
，
And calculating and correcting image coordinates, wherein the radial distortion parameter, the tangential distortion parameter and the camera parameter are obtained through a correction function.
3. The swimming posture assessment method based on the multi-angle camera as claimed in claim 1, wherein: the recognizing the human body gesture comprises the following steps:
generating a joint thermodynamic diagram from the corrected image, calculating the offset and the relative offset position of the joint relative to the coordinate, establishing a neural network framework, receiving the thermodynamic diagram and the joint offset as a training set by the network framework, supervising the training process and marking anchor point diagram positions of 33 joints after identification; the neural network is a residual error network, all layers are connected by using a jump mode and a regression method, and the training set uses the graph position with more shelters.
4. The swimming stroke assessment method based on the multi-angle camera as claimed in claim 3, wherein: the image-generated joint thermodynamic diagram includes:
the recognition sequence starts from the face of the human body, the face is recognized through a non-maximum suppression algorithm, the selected area is cut, and then the trained model is called to generate a coordinate result.
5. The swimming posture assessment method based on the multi-angle camera as claimed in claim 3, wherein: the anchor coordinate filtering includes:
coordinate filtering by a filtering algorithm, the filtering algorithm comprising:
combining the generated anchor point information of 33 joints to generate a 66 x N matrix, wherein N represents the frame number of the motion, and performing noise reduction processing on the combined data, wherein the formula is as follows:
in the formula
、
Is the pixel coordinates of the first camera head,
、
、
is a projection matrix of the camera in three coordinate planes,
the three-dimensional plane projection is carried out, and the analogy is carried out to obtain the pixel coordinate of the second camera
、
A is a coordinate matrix, and the coordinate matrix is calculated by the following formula;
6. the swimming posture assessment method based on the multi-angle camera as claimed in claim 1, wherein: the method for converting the two-dimensional coordinates into the three-dimensional coordinates comprises the following steps of:
the generated two-dimensional plane coordinates are directly and linearly transformed, the formula is shown as follows,
is the coordinates of the control points of each point,
coordinates on the image corresponding to the image points of the three-dimensional control point:
wherein
、
、
、
、
、
、
、
、
、
、
The parameters are determined according to the parameters of the camera and the underwater environment parameters for parameter conversion.
7. The swimming stroke assessment method based on the multi-angle camera as claimed in claim 1, wherein: the swimming posture qualification judging data comprises:
obtaining three-dimensional coordinates of each joint point
，
Then, the included angle of each joint is obtained in the three-dimensional space, and a three-dimensional vector is set according to the starting point of each joint
And calculating an included angle:
and obtaining eight joint included angles which are respectively included angles of a left elbow joint, a right elbow joint, a left arm, a right arm, a trunk, a left knee, a right knee and a left hip joint, defining four grades of excellence, good, qualified and unqualified through preset judgment conditions, and evaluating the posture of the swimmer.
8. A multi-angle camera-based swimming stroke assessment system, the system comprising:
the acquisition module is used for acquiring a swimming image;
the recognition module is used for recognizing human body gestures;
the generating module is used for generating two groups of coordinates according to the recognized human body posture data;
the three-dimensional posture model is used for training the recognized human body posture data;
filtering anchor coordinates, namely filtering the anchor coordinates generated in the same scene;
the conversion module is used for converting the two-dimensional coordinates into three-dimensional coordinates;
the generating module is used for generating a three-dimensional vector from the three-dimensional coordinate;
the included angle calculation module is used for calculating the included angle between the vectors;
the judgment threshold setting module is used for setting a judgment threshold;
and the judging module is used for setting the swimming posture into a multi-end mode, converting the angle and the posture to be good or bad and judging the qualified swimming posture data.
9. A swimming posture assessment device based on a multi-angle camera comprises:
the at least two cameras are used for acquiring front images and side images of a human body;
a memory for storing non-transitory computer readable instructions; and
a processor for executing the computer readable instructions, such that the computer readable instructions, when executed by the processor, implement the multi-angle camera-based swimming stroke assessment method according to any one of claims 1 to 7.
10. A computer-readable storage medium storing non-transitory computer-readable instructions which, when executed by a computer, cause the computer to perform the multi-angle camera-based swimming stroke assessment method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210708412.5A CN114783001A (en) | 2022-06-22 | 2022-06-22 | Swimming posture evaluation method, system, device and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210708412.5A CN114783001A (en) | 2022-06-22 | 2022-06-22 | Swimming posture evaluation method, system, device and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114783001A true CN114783001A (en) | 2022-07-22 |
Family
ID=82422343
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210708412.5A Pending CN114783001A (en) | 2022-06-22 | 2022-06-22 | Swimming posture evaluation method, system, device and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114783001A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115170911A (en) * | 2022-09-06 | 2022-10-11 | 浙江大学湖州研究院 | Human body key part positioning system and method based on image recognition |
| CN115240856A (en) * | 2022-08-29 | 2022-10-25 | 成都体育学院 | Exercise health assessment method, system and equipment based on exercise posture |
| CN117115925A (en) * | 2023-10-24 | 2023-11-24 | 广州华夏汇海科技有限公司 | Intelligent pool wall for swimming examination room |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8330611B1 (en) * | 2009-01-15 | 2012-12-11 | AvidaSports, LLC | Positional locating system and method |
| CN108288294A (en) * | 2018-01-17 | 2018-07-17 | 视缘(上海)智能科技有限公司 | A kind of outer ginseng scaling method of a 3D phases group of planes |
| CN113850865A (en) * | 2021-09-26 | 2021-12-28 | 北京欧比邻科技有限公司 | Human body posture positioning method and system based on binocular vision and storage medium |
-
2022
- 2022-06-22 CN CN202210708412.5A patent/CN114783001A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8330611B1 (en) * | 2009-01-15 | 2012-12-11 | AvidaSports, LLC | Positional locating system and method |
| CN108288294A (en) * | 2018-01-17 | 2018-07-17 | 视缘(上海)智能科技有限公司 | A kind of outer ginseng scaling method of a 3D phases group of planes |
| CN113850865A (en) * | 2021-09-26 | 2021-12-28 | 北京欧比邻科技有限公司 | Human body posture positioning method and system based on binocular vision and storage medium |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115240856A (en) * | 2022-08-29 | 2022-10-25 | 成都体育学院 | Exercise health assessment method, system and equipment based on exercise posture |
| CN115170911A (en) * | 2022-09-06 | 2022-10-11 | 浙江大学湖州研究院 | Human body key part positioning system and method based on image recognition |
| CN117115925A (en) * | 2023-10-24 | 2023-11-24 | 广州华夏汇海科技有限公司 | Intelligent pool wall for swimming examination room |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114783001A (en) | Swimming posture evaluation method, system, device and computer readable storage medium | |
| Qiao et al. | Real-time human gesture grading based on OpenPose | |
| CN106650687B (en) | Posture correction method based on depth information and skeleton information | |
| JP6124308B2 (en) | Operation evaluation apparatus and program thereof | |
| CN110321754B (en) | Human motion posture correction method and system based on computer vision | |
| CN110544301A (en) | Three-dimensional human body action reconstruction system, method and action training system | |
| CN111862296B (en) | Three-dimensional reconstruction method, three-dimensional reconstruction device, three-dimensional reconstruction system, model training method and storage medium | |
| CN109117753B (en) | Part recognition method, device, terminal and storage medium | |
| CN109274883B (en) | Posture correction method, device, terminal and storage medium | |
| CN106355147A (en) | Acquiring method and detecting method of live face head pose detection regression apparatus | |
| CN111402290A (en) | Action restoration method and device based on skeleton key points | |
| CN110544302A (en) | Human body action reconstruction system and method based on multi-view vision and action training system | |
| JPWO2012077286A1 (en) | Object detection apparatus and object detection method | |
| JP2009157767A (en) | Face image recognition apparatus, face image recognition method, face image recognition program, and recording medium recording this program | |
| CN113850865A (en) | Human body posture positioning method and system based on binocular vision and storage medium | |
| CN112001859A (en) | Method and system for repairing face image | |
| CN109858433B (en) | Method and device for identifying two-dimensional face picture based on three-dimensional face model | |
| CN114187656A (en) | Action detection method, device, equipment and storage medium | |
| CN111709365A (en) | Automatic human motion posture detection method based on convolutional neural network | |
| CN115035546B (en) | Three-dimensional human body posture detection method and device and electronic equipment | |
| Yang et al. | Human exercise posture analysis based on pose estimation | |
| CN115205737B (en) | Motion real-time counting method and system based on transducer model | |
| JP2021128592A5 (en) | ||
| CN108694348B (en) | Tracking registration method and device based on natural features | |
| Lin | 3D object detection and 6D pose estimation using RGB-D images and mask R-CNN |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220722 |