CN118179002A - A fitness equipment linkage control method and system - Google Patents

Abstract

The present invention relates to the field of equipment linkage control technology, and in particular to a fitness equipment linkage control method and system, including a data acquisition sensor module, a camera module and a music playing module, wherein the data acquisition sensor module includes a face recognition unit, a fingerprint recognition unit and a motion data sensor unit, wherein the face recognition unit and the fingerprint recognition unit establish a user fitness database based on user identity information, and the camera module recognizes and detects the posture information of the key points of the human skeleton after preprocessing the captured image or video, and counts the motion posture after determining the linkage change range of the key points of the skeleton. The fitness equipment linkage control method and system clusters the frequency of the user's fitness action per unit time, determines the optimal frequency collection per unit time when the user is exercising, and counts once if it is within the optimal frequency collection, which can effectively improve the accuracy of the user's fitness action, thereby improving the fitness effect.

Description

A fitness equipment linkage control method and system

Technical Field

The present invention relates to the technical field of equipment linkage control, and in particular to a fitness equipment linkage control method and system.

Background technique

With the progress of social civilization, people hope to be healthy physically and mentally, prolong their life, pursue spiritual satisfaction and enjoy the fun of life. Sports are one of the important ways to achieve this goal. It has many functions such as fitness, entertainment, and thought. It can enrich human cultural life, improve cultural standards and quality of life, improve people's living habits, cultivate noble character, and have a positive regulatory effect on the healthy development of society. Therefore, going to the gym has become the preferred way for people to exercise.

As indoor fitness exercises become more and more popular, people have taken fitness exercises as a daily program to improve their physical condition. However, existing fitness equipment and systems are large in size, heavy, inconvenient to move, and cannot form an immersive fitness scene, let alone adjust the functions of fitness equipment with each other.

A data interaction fitness system disclosed in the patent document with publication number CN107158661A includes an intelligent gateway, an intelligent fitness device, a mobile app, and a PC application system. The intelligent gateway and the intelligent fitness device, the mobile app, and the PC application system are connected and communicated through a wireless network; the intelligent gateway is equipped with a private cloud, a wireless router, and a fitness device control hub; the intelligent fitness device needs to be equipped with a human body sensor, an NFC control chip, and a ZigBee module; the mobile app is used to obtain user fitness health index information through NFC communication; the PC application system is used to obtain user fitness information uploaded to the private cloud by the intelligent gateway through the wireless network, and analyze and process data for user-specified health goals and user fitness information. This invention provides a highly intelligent and professional fitness system to help people perform scientific exercises based on their personal characteristics and achieve the best results.

However, this data interactive fitness system cannot simulate a wider variety of fitness equipment using fewer components, nor can it build an immersive fitness scene for users, and cannot control the accuracy of users' fitness movements.

Summary of the invention

The purpose of the present invention is to provide a fitness equipment linkage control method and system to solve the problems raised in the above background technology.

To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a method and system for linkage control of fitness equipment, comprising a data acquisition sensor module, a camera module, a music playing module and a control module, wherein the data acquisition sensor module comprises a face recognition unit, a fingerprint recognition unit and a motion data sensing unit, wherein the face recognition unit and the fingerprint recognition unit establish a user fitness database based on user identity information, and the motion data sensing unit establishes a user fitness database after obtaining optimal motion data based on a clustering algorithm.

The camera module recognizes and detects the posture information of the key points of the human skeleton after preprocessing the captured image or video, and counts the movement posture after determining the linkage change range of the key points of the skeleton.

The music playing module includes importing the user's listening log into the user fitness database and playing the songs that the user is interested in during the user's fitness process.

Preferably, the face recognition unit and the fingerprint recognition unit are used to obtain the user's identity information and generate a corresponding user ID serial number, and the user information in the fitness database is bound and matched with the collected user fitness information through the user ID serial number.

Preferably, the face recognition unit and fingerprint recognition unit include face recognition devices and fingerprint collection devices at the entrance of the gym and on fitness equipment, and the face recognition devices and fingerprint collection devices are both connected to the terminal device via network signals.

Preferably, the motion data sensing unit includes a counting sensor and a lighting device. The counting sensor clusters the frequencies of the user's fitness movements after transmitting the data to the terminal device to determine the optimal frequency set for the user's fitness.

Preferably, the control module determines whether the frequency of the user's fitness is in the optimal frequency collection, generates corresponding control instructions according to the received real-time frequency value of the fitness equipment, and controls the real-time color state of the lighting equipment through the control instructions.

Preferably, the real-time color status of the lighting device is divided into three colors: red, green and blue. When the frequency of the user during fitness is lower than the optimal frequency collection, the instructions issued by the control module control the real-time color of the lighting device to blue. When the frequency of the user during fitness is higher than the optimal frequency collection, the instructions issued by the control module control the real-time color of the lighting device to red. When the frequency of the user during fitness is within the optimal frequency collection, the instructions issued by the control module control the real-time color of the lighting device to green.

Preferably, the key points of the human skeleton refer to 16 specific positions of the human body, such as the head, shoulders, knees and ankles, and the range of change is determined by observing the angle change of the key points in a certain fitness movement.

Preferably, after the camera module determines the identity of the user based on the face detector, the captured images and videos are transmitted to the processor for image preprocessing, objects are detected and located from the processed images or videos, and the detected human body information is used to identify the human skeleton key points and posture information through a posture estimation model, and the changes in the skeleton key points are observed according to different fitness movements, and they are counted using a counting algorithm.

Preferably, the control module includes a host terminal of the gym, the host terminal is connected to the face recognition device and fingerprint collection device in the data collection sensor module through a network signal, and the host terminal is electrically connected to the counting sensor and lighting equipment in the motion data sensing unit.

A method for controlling linkage of a fitness device comprises the following steps:

S1: When a user enters a gym, the face recognition device and fingerprint recognition device at the entrance of the gym will identify the user, obtain the user's identity information and generate a corresponding user ID serial number, use the serial number to establish a user fitness database, and use the user ID serial number to bind and match the user information in the fitness database with the collected user fitness information;

S2: Users can import their own music listening logs into the music playing module of the fitness equipment and play music according to their own preferences during fitness;

S3: After the camera module determines the user's identity based on the face detector, it transmits the captured images and videos to the processor for image preprocessing, detects and locates objects from the processed images or videos, identifies the key points of the human skeleton and posture information through the posture estimation model, observes the changes of the key points of the skeleton according to different fitness movements, and counts them using the counting algorithm;

S4: During the user's fitness process, the counting sensor on the fitness equipment calculates the frequency of the user's fitness movements, and the control module clusters the frequency of the user's fitness movements per unit time to determine the optimal frequency set per unit time during the user's fitness. If it is within the optimal frequency set, a count is performed and displayed on the fitness equipment display screen, otherwise the count is not displayed.

Compared with the prior art, the present invention has the following beneficial effects:

The fitness equipment linkage control method and system establishes a user fitness database after acquiring optimal motion data based on user identity information and a clustering algorithm, detects and locates objects from the processed images or videos after preprocessing the captured images or videos, identifies human skeleton key points and posture information through a posture estimation model based on the detected human body information, observes changes in skeleton key points according to different fitness movements, clusters the frequencies of user fitness movements per unit time, determines the optimal frequency collection per unit time when the user is exercising, and counts once if the frequency is within the optimal frequency collection, which can effectively improve the accuracy of the user's movements during fitness, thereby improving the fitness effect.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

FIG. 1 is a flow chart of a linkage control system for fitness equipment according to the present invention.

Detailed ways

The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

Please refer to Figure 1, the present invention provides a technical solution:

A linkage control system for fitness equipment, comprising a data acquisition sensor module, a camera module, a music player module and a control module, wherein the data acquisition sensor module comprises a face recognition unit, a fingerprint recognition unit and a motion data sensor unit, wherein the face recognition unit and the fingerprint recognition unit establish a user fitness database based on user identity information, and the motion data sensor unit establishes a user fitness database after acquiring optimal motion data based on a clustering algorithm;

The camera module recognizes and detects the posture information of the key points of the human skeleton after preprocessing the captured image or video, and counts the motion posture after determining the linkage change range of the key points of the skeleton;

The music playing module includes importing the user's listening log into the user fitness database and playing the songs that the user is interested in during the user's fitness process.

The face recognition unit and the fingerprint recognition unit are used to obtain the user's identity information and generate a corresponding user ID serial number. The user ID serial number is used to bind and match the user information in the fitness database with the collected user fitness information. The face recognition unit and the fingerprint recognition unit include face recognition devices and fingerprint collection devices at the entrance of the gym and on fitness equipment. The face recognition device and the fingerprint collection device are connected to the terminal device through network signals.

The motion data sensing unit includes a counting sensor, a lighting device and a control module electrically connected thereto. The control module clusters the frequencies of the user's fitness movements and determines the optimal frequency set for the user during fitness.

The control module includes a host terminal of the gym, which is connected to the face recognition device and fingerprint collection device in the data collection sensor module through a network signal, and is electrically connected to the counting sensor and lighting device in the motion data sensor unit.

The control module determines whether the frequency of the user's fitness is in the optimal frequency collection. The control module generates corresponding control instructions according to the real-time frequency value of the fitness equipment received, and controls the real-time color state of the lighting equipment through the control instructions.

The real-time color status of the lighting equipment is divided into three colors: red, green and blue. When the frequency of the user's fitness is lower than the optimal frequency collection, the control module issues an instruction to control the real-time color of the lighting equipment to blue. When the frequency of the user's fitness is higher than the optimal frequency collection, the control module issues an instruction to control the real-time color of the lighting equipment to red. When the frequency of the user's fitness is within the optimal frequency collection, the control module issues an instruction to control the real-time color of the lighting equipment to green.

The fitness database includes a data storage and processing device that only stores fitness data information of a single user 30 days before the login time.

After the camera module determines the user's identity based on the face detector, it transmits the captured images and videos to the processor for image preprocessing, detects and locates objects from the processed images or videos, identifies the key points of the human skeleton and posture information through the posture estimation model, and observes the changes of the key points of the skeleton according to different fitness movements. The key points of the human skeleton refer to 16 specific positions of the human body, such as the head, shoulders, knees and ankles. The range of change is determined by observing the angle changes of the key points in a certain fitness movement; the counting algorithm is used to count them, and the camera module is electrically connected to the fitness equipment display module through the counting sensor, and counts when the subsequent fitness movements are within the range of change.

A method for controlling linkage of a fitness device comprises the following steps:

S1: When a user enters a gym, the face recognition device and fingerprint recognition device at the entrance of the gym will identify the user, obtain the user's identity information and generate a corresponding user ID serial number, use the serial number to establish a user fitness database, and use the user ID serial number to bind and match the user information in the fitness database with the collected user fitness information;

S2: Users can import their own listening logs into the music playback module of the fitness equipment and play music according to their own preferences during fitness. They can also use the equipment to recommend music based on the user's listening logs.

S3: After the camera module determines the identity of the user based on the face detector, it transmits the captured images and videos to the processor for image preprocessing.

The captured images are transmitted to the image processor, which performs VLAD fusion on the image data to obtain vectors. First, the features of each frame of a video are clustered to obtain multiple cluster centers, and all features are assigned to the specified cluster centers. The feature vectors in each cluster area are averaged, and finally the feature vectors of all cluster areas are merged as the feature vector of the entire video. The feature vectors of all frames are clustered to one center point, and VLAD clusters the feature vectors of all frames to multiple center points. By stacking all feature vectors, richer features can be obtained with less information loss. The obtained vectors are used to model the image through the low-rank approximation model in 3D convolution to obtain feature results.

A 3×3×3 convolution kernel is used in all images, and the convolution layer stride is set to 1. The input of VGG is set to an RGB image of size 224×244. The RGB mean is calculated for all images in the training set, and then the image is passed as input to the VGG convolutional network.

Assuming that n recognition points are extracted from an image, the visual features of the image can be expressed as χ = {x ₁ , ..., x _j , ... x _n }, and the corresponding position features can be expressed as P = P = {P ₁ , ..., P _j , ..., P _n }, where x _j and P _j represent the visual features and position features of the jth recognition point. In order to obtain a more efficient recognition representation, it is necessary to encode the underlying image features to obtain a vector-level representation of the image.

First, k-means clustering is used to cluster visual features to generate a visual dictionary. Assuming that the size of the visual dictionary VD ₁ obtained by clustering is k ₁ , VD ₁ is expressed as VD ₁ = {C ₁ , ..., C _i , ..., C _k1 }, where C _i represents the i-th cluster center in the visual dictionary. The original VLAD encoding is to calculate the difference between each cluster center and its elements, that is, the encoding vector of the i-th cluster center is expressed as:

Where _xj represents the jth visual feature contained in cluster center _Ci , and _Ni represents the number of visual features contained in cluster center _Ci . In order to enhance the role of cluster center in visual features, the calculation of the residual between the cluster center and its most similar element is added. The VLAD calculation formula of the i-th cluster center is: Where _xt represents the visual feature most similar to the visual feature contained in the cluster center _Ci . In summary, when the size of the visual dictionary is k ₁ , the image encoding vector of the image can be expressed as:

The image data is enhanced and the amplitude is randomly enhanced. After the effective image data is extracted, data x _valid is obtained, and the number of sampling points is N _valid . With N _clip as the number of sampling points of each image, x _valid is sliced to obtain multiple slice data x _clip (λ _c ), where λ _c is the slice number.

Perform random amplitude enhancement or attenuation on each slice data x _clip (λ _c ). First, randomly select a gain value rn_s from [rn_l, rn_h] (e.g. [-10, 10]), then select a second random value in [rn_s-rn_o, rn_h+rn_o] as the gain of a slice, and directly multiply it with the slice data to obtain the slice data after gain processing. The gain values of all slices are randomly obtained in the range of [rn_s-rn_o, rn_h+rn_o] and gain processed in the same way. Among them, rn_l and rn_h are the lower and upper limits of the first-level random gain, respectively, and rn_o is the offset of the second-level gain. The data slices after random gain processing are resynthesized into complete image data.

Detect and locate objects from processed images or videos, identify human skeleton key points and posture information through the posture estimation model, observe the changes of skeleton key points according to different fitness movements, and use counting algorithms to count them;

Extract global features, extract Rx and Ry nodes, and upsample the extracted feature maps. The upsampling module consists of 3 consecutive deconvolution layers, batch normalization, and ReLU activation functions. Then, a 1×1 convolution is performed to generate a 2D heat map of the root node. Softargmax is used to extract Rx and Ry nodes from the 2D heat map. To extract the RZ node, the feature map extracted by the Backbone layer is processed by global average pooling, and then a 1×1 convolution is performed to output the scalar value γ, which is multiplied by the k value to obtain the final absolute depth value RZ.

Different fitness movements have different fitness counting algorithms. The difference lies in the linkage changes of the key points of the human skeleton, by observing the angle changes of a key point. Assuming that the coordinates of the three key points of the skeleton are (1, yxA), (2, yxB) and (3, yxC), the lengths of the sides corresponding to each angle are a, b and c, respectively, and then the arc cosine formula is used to determine whether the angle B is within the selectable angle range.

Taking the high pull-down as an example, to define the rules for counting the high pull-down fitness exercise, it is necessary to strictly abide by the requirements of the high pull-down fitness exercise, that is, to ensure that the upper torso is in a straight line, and counting is only performed when the two elbows are within a specific angle range and the arm state changes during the pull-down process.

S4: During the user's fitness process, the counting sensor on the fitness equipment calculates the frequency of the user's fitness movements, and the control module clusters the frequency of the user's fitness movements per unit time to determine the optimal frequency set per unit time during the user's fitness. If it is within the optimal frequency set, a count is performed and displayed on the fitness equipment display screen, otherwise the count is not displayed.

The specific steps of counting are as follows:

First, the following definitions are given:

The average distance between any two frequency data points is:

Where: d( _xi , _xj ) is the Euclidean distance between any two points, and n represents the number of frequency data points.

Define the area radius as R, Reler is the adjustment coefficient. When Reler is 0.13, the clustering effect is the best.

Definition of frequency point aggregation:

The average cluster distance is defined as:

G( _xi ) is determined by comparing the degree of aggregation Dp( _xi ), which is used to measure the differences between different clusters. Among all the frequency points, when the degree of aggregation of _xi is the largest, G( _xi ) is the maximum distance between _xi and all the remaining points, otherwise it is the minimum distance between _xi and all the remaining points.

The central aggregation parameter is defined as:

Calculate the central aggregation parameter ω(X _i ) of each point;

Removal of free points. Free points are those points that are far away from all other points. Their existence will cause the deviation of the center point in the class to which they belong, thus affecting the classification effect.

Select the point _Xi that makes ω( _Xi ) the largest, and use it as the first initial cluster center. Calculate the distance between _Xi and the remaining points, and compare the obtained distance value with the neighborhood radius R. If the distance is less than R, it means that it can be classified into the same category as _Xi , so these points will be removed from the data points. If the distance is greater than R, it means that the distance to _Xi is too far and it is not suitable to be classified into the same category as _Xi , so these points will be retained and proceed to the next step.

Select the point with the largest ω(X _i ) from the remaining points as the second cluster center and repeat the above steps;

The above steps are repeated until all points X ₁ , X ₂ , ..., X _n in the data set are removed, and k optimal initial centers M _i (i=1, 2, ..., k) are output.

After counting, it will be displayed on the display screen of the fitness equipment. It will only be displayed when the count reaches the fitness requirement, so as to control the accuracy of the user's fitness movements.

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

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims (10)

1. The utility model provides a health and fitness facilities coordinated control system, includes data acquisition sensing module, camera module, music play module and control module, its characterized in that: the data acquisition sensing module comprises a face recognition unit, a fingerprint identification unit and a motion data sensing unit, wherein the face recognition unit and the fingerprint identification unit establish a user body-building database based on user identity information, and the motion data sensing unit establishes the user body-building database after acquiring optimal motion data based on a clustering algorithm;

The camera module is used for identifying and detecting gesture information of key points of human bones after preprocessing a shot image or video, and counting the moving gestures after determining the linkage change range of the key points of the bones;

the music playing module plays songs interested by the user in the user body building process after the user listening log is imported into the user body building database.

2. The exercise machine coordinated control system of claim 1, wherein: the face recognition unit and the fingerprint recognition unit are used for acquiring identity information of a user and generating corresponding user ID serial numbers, and binding and matching of the user information and the acquired user body-building information in the body-building database are achieved through the user ID serial numbers.

3. The exercise machine coordinated control system of claim 1, wherein: the face recognition unit and the fingerprint recognition unit comprise face recognition equipment and fingerprint acquisition equipment at the entrance of the gymnasium and on the gymnasium, and the face recognition equipment and the fingerprint acquisition equipment are connected with the terminal equipment through network signals.

4. The exercise machine coordinated control system of claim 1, wherein: the motion data sensing unit comprises a counting sensor and light equipment, wherein the counting sensor clusters the frequency of the user body-building action after transmitting data to the terminal equipment, and determines the optimal frequency aggregate during the user body-building.

5. The exercise machine coordinated control system of claim 4, wherein: the control module generates corresponding control instructions according to the received real-time frequency values of the body-building equipment by judging whether the frequency of the user during body-building is in the optimal frequency aggregate, and controls the real-time color state of the light equipment through the control instructions.

6. The exercise machine coordinated control system of claim 5, wherein: the real-time color state of the light equipment is divided into three colors of red, green and blue, when the frequency of the user during exercise is lower than the optimal frequency set, the real-time color of the light equipment is controlled to be blue by the instruction sent by the control module, when the frequency of the user during exercise is higher than the optimal frequency set, the real-time color of the light equipment is controlled to be red by the instruction sent by the control module, and when the frequency of the user during exercise is in the optimal frequency set, the real-time color of the light equipment is controlled to be green by the instruction sent by the control module.

7. The exercise machine coordinated control system of claim 1, wherein: after the camera module determines the identity of the user based on the face detector, the camera module transmits the shot image and video to the processor for image preprocessing, detects and positions objects from the processed image or video, recognizes key points and gesture information of human bones through the gesture estimation model according to the detected human body information, observes the change of the key points of the bones according to different body-building actions, and counts the changes by using a counting algorithm.

8. The exercise machine coordinated control system of claim 7, wherein: the key points of the human bones refer to a plurality of specific positions of the head, the shoulders, the knees and the ankles of the human body, and the change range is determined by observing the angle change of the key points in a certain body-building action.

9. The exercise machine coordinated control system of claim 8, wherein: the control module comprises a host terminal of the gymnasium, the host terminal is connected with face recognition equipment and fingerprint acquisition equipment in the data acquisition sensing module through network signals, and the host terminal is electrically connected with a counting sensor and lamplight equipment in the motion data sensing unit.

10. The exercise machine coordinated control method according to any one of claims 1 to 9, comprising the steps of:

S1: when a user enters a gymnasium, face recognition equipment and fingerprint recognition equipment at the entrance of the gymnasium can recognize the user, acquire identity information of the user and generate corresponding user ID serial numbers, a user gymnasium database is built by using the serial numbers, and binding and matching of user information and acquired user gymnasium information in the gymnasium database are realized through the user ID serial numbers;

S2: the user can import own song listening logs into the music playing module of the exercise equipment, and can play music according to own preference in the exercise process;

S3: after the camera module determines the identity of the user based on the face detector, the camera module transmits the shot image and video to the processor for image preprocessing, detects and positions objects from the processed image or video, recognizes key points and gesture information of human bones through the gesture estimation model according to the detected human body information, observes the change of the key points of the bones according to different body-building actions, and counts the changes by using a counting algorithm;

s4: in the user body building process, a counting sensor on the body building equipment calculates the action frequency of the user body building, a control module clusters the frequency in the unit time of the user body building action, an optimal frequency set in the unit time of the user body building is determined, if the optimal frequency set is in the optimal frequency set, the counting is carried out once and is displayed on a display screen of the body building equipment, and otherwise, the counting is not displayed.