Method and device for automatically evaluating athletic ability of old people, medium and product

Abstract

The invention discloses an automatic assessment method, device, medium and product for the athletic ability of old people, and relates to the field of computer vision, and the method comprises the steps: obtaining a to-be-tested video of a subject; the to-be-tested video comprises videos that the two hands of the subject are placed behind the brain, squat and sit down; extracting skeleton key point data from the to-be-detected video by using a key point detection algorithm; according to the skeleton key point data, a feature extraction network model obtained based on graph convolutional network training is adopted to determine skeleton features; according to the skeleton features, a multi-modal classification model obtained through training based on a long-short-term memory network and a full convolutional network is adopted, and an athletic ability evaluation result is determined. According to the invention, the method can achieve the automatic assessment of the exercise ability of the old people, which is simple to implement and has high precision and high efficiency, in a home environment.

Description

Technical Field

[0001] The present application relates to the field of computer vision, and in particular to a method, device, medium and product for automatically evaluating the motor ability of the elderly. Background Art

[0002] As the global aging trend intensifies, the health of the elderly has received more and more attention. Exercise capacity is an important indicator of the health status of the elderly. Reasonable exercise can effectively prevent a variety of chronic diseases and improve the quality of life of the elderly. However, due to the gradual deterioration of the physical functions of the elderly, scientific, objective and continuous exercise capacity assessment is particularly important.

[0003] At present, traditional motor ability assessment mostly relies on manual measurement and observation by professional medical staff. Although these methods are accurate, they have certain limitations, such as complex assessment process, reliance on professional equipment, long time consumption and high cost. In addition, the elderly generally have problems such as decreased motor ability and weakened cognitive ability, and are unable to freely participate in complex assessment processes. Therefore, the existing assessment methods are difficult to achieve daily motor ability monitoring and management, and are difficult to adapt to the personalized needs of the elderly population.

[0004] With the development of artificial intelligence and smart wearable devices, automated assessment methods based on sensor technology have gradually been applied to the medical and health field. By collecting motion data through sensors and using machine learning and deep learning algorithms for data analysis, it is possible to automatically assess the motor ability of the elderly. However, most of the existing automated assessment algorithms based on sensor technology are complex and difficult to promote and use in ordinary households. At the same time, these methods usually fail to fully consider the movement characteristics and personalized needs of the elderly in real life scenarios, resulting in certain limitations on the accuracy and practicality of the assessment results.

[0005] Therefore, there is an urgent need for an automated assessment method that can be easily implemented in a home environment and has both high accuracy and efficiency. Summary of the invention

[0006] The purpose of this application is to provide a method, device, medium and product for automatically evaluating the motor ability of the elderly, which can realize simple, high-precision and efficient automated evaluation of the motor ability of the elderly in a home environment.

[0007] To achieve the above objectives, this application provides the following solutions:

[0008] In a first aspect, the present application provides a method for automatically evaluating the motor ability of the elderly, the method comprising:

[0009] Obtain a video of the subject to be tested; the video to be tested includes: a video of the subject placing his hands behind his head, squatting, and sitting;

[0010] Use key point detection algorithm to extract skeleton key point data from the video to be tested;

[0011] According to the skeleton key point data, the feature extraction network model trained based on the graph convolutional network is used to determine the skeleton features;

[0012] According to the skeleton features, a multimodal classification model trained based on long short-term memory network and fully convolutional network is used to determine the motor ability assessment results.

[0013] Optionally, obtaining the video to be tested of the subject specifically includes:

[0014] Using video equipment to film the subjects;

[0015] Use OpenCV to obtain the video to be tested.

[0016] Optionally, the key point detection algorithm is an HrNet algorithm.

[0017] Optionally, the loss function of the multimodal classification model adopts a cross-entropy loss function.

[0018] Optionally, obtaining a video of the subject to be tested, and then further comprising:

[0019] Perform data cleaning and data filling operations on the video to be tested.

[0020] In a second aspect, the present application provides an automatic assessment device for the motor ability of the elderly, the automatic assessment device for the motor ability of the elderly comprising:

[0021] The test acquisition and input module is used to obtain the test video of the subject; the test video includes: the video of the subject putting his hands behind his head, squatting and sitting;

[0022] The motion ability feature extraction module is used to extract the skeleton key point data from the video to be tested by using the key point detection algorithm; according to the skeleton key point data, the skeleton features are determined by using the feature extraction network model obtained by training based on the graph convolution network;

[0023] The physical motor ability assessment module is used to determine the motor ability assessment results based on skeleton features using a multimodal classification model trained based on long short-term memory networks and fully convolutional networks.

[0024] Optionally, the automatic assessment device for the motor ability of the elderly further comprises: a body motor ability assessment report module;

[0025] The body movement ability assessment report module is used to generate a movement ability assessment report from the movement ability assessment result.

[0026] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for automatically assessing the motor ability of the elderly.

[0027] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for automatically assessing the motor ability of the elderly.

[0028] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method for automatically assessing the motor ability of the elderly.

[0029] According to the specific embodiments provided in this application, this application has the following technical effects:

[0030] The present application provides a method, device, medium and product for automatically evaluating the motor ability of the elderly. It adopts computer vision technology combined with graph convolutional network (Graph Convolutional Network, GCN) and multimodal network trained based on long short-term memory network (Long Short-Term Memory, LSTM) and full convolutional network (Fully Convolutional Networks for Semantic Segmentation, FCN) to propose a new method for evaluating motor ability. Through the recorded video to be tested, the motor ability evaluation results of different actions can be obtained. The elderly can use it independently at home, and the scoring is fully automatic. The evaluation can be objective and quantitative, which is helpful for the elderly to carry out motor ability training at home; and the motor ability evaluation results obtained by this application have a certain encouraging effect on patients; at the same time, the scores recorded by date and the collected videos help doctors to analyze and make corresponding training suggestions. The application of this application not only improves the scientificity and personalization of treatment, but also effectively improves the quality of life of the elderly. Through continuous training and scientific evaluation, the elderly can gradually improve their motor function and improve their quality of life. At the same time, this application also provides medical personnel with a more convenient and efficient evaluation tool, which promotes the further development of motor ability evaluation for the elderly. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 This is a diagram of an application environment of a method for automatically evaluating the motor ability of the elderly in one embodiment of the present application;

[0033] Figure 2 A schematic diagram of a flow chart of a method for automatically evaluating the motor ability of the elderly provided in one embodiment of the present application;

[0034] Figure 3 Set up schematics for the action;

[0035] Figure 4 This is a schematic diagram of the key points of the skeleton;

[0036] Figure 5 This is a schematic diagram of the structure of the convolutional network of MALSTM-MFCN;

[0037] Figure 6 Schematic diagram of the MALSTM-MFCN model evaluation algorithm flow. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0039] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0040] The automatic assessment method for the motor ability of the elderly provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the video to be tested to the server 104. After the server 104 receives the video to be tested, for the video to be tested, the server 104 extracts the skeleton key point data of the video to be tested based on the key point detection algorithm; according to the skeleton key point data, the feature extraction network model obtained by training the graph convolution network is used to determine the skeleton features; according to the skeleton features, the multimodal classification model obtained by training the long short-term memory network and the full convolution network is used to determine the motor ability evaluation result. The server 104 can feedback the obtained motor ability evaluation result to the terminal 102. In addition, in some embodiments, the automatic evaluation method for the motor ability of the elderly can also be implemented by the server 104 or the terminal 102 alone, such as the terminal 102 can directly process the video to be tested, or the server 104 can obtain the video to be tested from the data storage system and process the video to be tested.

[0041] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0042] In an exemplary embodiment, Figure 2 As shown, a method for automatically evaluating the motor ability of the elderly is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or can be executed by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following S201 to S204. Among them:

[0043] S201, obtaining a video of the subject to be tested; the video to be tested includes: a video of the subject placing his hands behind his head, squatting, and sitting down; wherein the three actions in the video to be tested are determined according to the standards of the American College of Sports Medicine, and are used to judge the motor ability of the elderly; the action of placing both hands behind the head can include elbow flexion and extension and shoulder extension; in the process of obtaining the video to be tested of the subject, the subject with reduced motor ability will have problems such as insufficient flexibility of the shoulder or elbow joints leading to inability to bend, inability to bend the knee joints normally, dizziness when sitting down and standing up, etc., and the action settings are as follows Figure 3 shown.

[0044] S201 specifically includes:

[0045] S11, photographing the subject using a camera; when the subject puts both hands behind the head, the subject faces the camera; when the subject squats or sits, the subject faces the camera sideways.

[0046] S12, using OpenCV to obtain the video to be tested. OpenCV processes and analyzes the video to be tested at the frame level. OpenCV provides a wealth of video processing functions, including video stream acquisition, storage, and playback. It has good cross-platform and compatibility, and can easily implement video operations in various environments.

[0047] In order to maximize the use of data and automatically clean up and delete irregular videos in the data set in batches, reduce the interference of irregular data on the overall analysis, and save time and resources for viewing irregular videos, after S201, the following steps are also included:

[0048] Perform data cleaning and data filling operations on the video to be tested.

[0049] There are two main reasons for data non-standardization. One is that the patient makes corresponding movements, and the key points cannot be fully detected, resulting in errors in the extraction of feature data. The other is the wrong shooting angle. For example, for action one, the subject needs to face the camera head-on, but the subject faces the camera sideways, which makes it impossible to segment the action through rules, and thus cannot be accurately evaluated. For these two main reasons, the rules are used to limit them. If the rules cannot be passed, the next evaluation operation will not be performed. For undetected skeletons, the data filling method is used. Since the movement changes of the elderly between two frames will not be too large, the amount of data is increased by filling forward or backward. Among them, the rule settings are shown in Table 1:

[0050] Table 1

[0051]

[0052] S202, extracting skeleton key point data from the video to be tested using a key point detection algorithm;

[0053] The existing key point detection algorithm can detect 17 key points of the human body in photos or videos. Among them, the upper limbs can identify 6 key points including the shoulder joint, elbow joint, and wrist joint, and the lower limbs can identify 7 key points including the hip joint, knee joint, and ankle joint. The information obtained for each key point includes: coordinates and confidence; the x and y coordinates in the coordinates represent the horizontal and vertical coordinates respectively. The x coordinate and y coordinate represent the horizontal and vertical position information of the skeletal key point in the picture respectively.

[0054] In an exemplary embodiment, the present application uses the HrNet algorithm to extract the skeletal key point data in the video to be tested. The HrNet algorithm can realize the real-time detection and tracking of the skeletal key points of the elderly in motion. The HrNet algorithm uses deep learning technology to identify the skeletal key points of the human body from the video stream in real time; the skeletal key points include: skeletal key points of the arms, hands and other parts, and track their movement trajectories.

[0055] Each bone key point represents some specific points in the human body. For example, the key points of the lower limbs include: the left hip joint, right hip joint, left knee joint, right knee joint, left ankle joint, and right ankle joint. Figure 4 shown.

[0056] S203, determining skeleton features based on the skeleton key point data using a feature extraction network model trained based on a graph convolutional network;

[0057] S203 specifically includes:

[0058] The data of each skeleton key point in each frame of the video to be tested are respectively represented as nodes in the graph, and the connections between the nodes are represented as edges in the graph, thereby forming a graph;

[0059] Due to the sparsity and connectivity of joints in motor ability assessment, the use of graph convolutional neural networks for skeleton feature extraction can effectively extract spatial features during the movement process.

[0060] GCN is a model dedicated to processing graph data and is widely used to process the dependencies between nodes and edges in graph structures. GCN learns the local and global features of nodes in the graph through graph convolution operations and captures the association information between nodes and neighboring nodes in the graph structure. By stacking multiple graph convolution layers, more complex graph features are gradually extracted.

[0061] Skeleton key point data is usually composed of multiple time frames, each time frame contains several skeleton key points, and each skeleton key point has corresponding spatial coordinates, i.e., each skeleton key point data in each frame in the video to be tested. The embodiment of the application extracts features from human skeleton data through a graph convolutional neural network. For each frame of image, each skeleton key point in the human skeleton data is represented as a node in the graph, and the connection between the skeleton key points is represented as an edge in the graph. After extracting each frame of image data through a graph convolutional network, the graphs of different time frames are connected to form a new time series.

[0062] S204, according to the skeleton features, a multimodal classification model obtained by training a long short-term memory network and a fully convolutional network is used to determine the athletic ability evaluation result. The loss function of the multimodal classification model is a cross-entropy loss function.

[0063] Among them, the evaluation algorithm of the multimodal classification model is MLSTM-MFCN or MALSTM-MFCN. The multimodal classification model is divided into multiple long short-term memory networks and multi-scale fully convolutional networks, which together constitute the sports ability evaluation network.

[0064] The input layer of the multimodal classification model receives the skeleton features extracted by GCN, among which the gate-based LSTM layer can naturally extract the temporal dependency between features, while the FCN with multi-scale filter groups can perceive spatial features of different ranges from the time series curve. Adding the attention mechanism to the LSTM layer can better allow the model to focus on the more important features in time. In FCN, the dilated convolution filter expands the range of the receptive field without compressing information or increasing the number of parameters, which is conducive to learning long-term features from high-dimensional and continuous time series, but too high a dilation rate corresponds to extensive skipping of sequence nodes, which may lead to misperception of patterns. Therefore, the dilation rate of the multimodal classification model is less than or equal to 4. Finally, the outputs of LSTM and FCN are spliced ​​together to realize the fully connected layer. Through linear transformation, the features are mapped to the rating level using the SoftMax activation function, and the multimodal classification model is trained. The final output is the probability distribution of each sample belonging to each rating level, and the classification result is obtained, which is the athletic ability evaluation result.

[0065] The structure of the convolutional network of MALSTM-MFCN in this application can be found in Figure 5 As shown in the figure, the MLSTM-MFCN network is obtained by extracting features from several sample videos through the GCN algorithm and training the corresponding labels, where the labels represent the level of the elderly in the current exercise process.

[0066] For the loss function used in the training process, since the data set adapted by this application does not have the problem of data similarity between categories, and the number between the three levels of 0, 1, and 2 is relatively balanced, the loss function uses the cross-entropy loss function Cross-Entropy Loss. If subsequent data have problems such as data imbalance, it is modified to weighted cross-entropy loss. The weighted cross-entropy loss can make the model pay more attention to these minority categories during the training process, thereby improving the classification performance.

[0067] The MALSTM-MFCN model was trained using labeled motion data, and the motion results were divided into three levels: 0, 1, and 2. Through supervised learning, the MALSTM-MFCN model can learn the characteristic representations of normal and abnormal motion patterns. During the MALSTM-MFCN model training process, the appropriate learning rate and number of iterations were set according to the specific performance of the MALSTM-MFCN model and the distribution of the data set, and ultimately better results were achieved in the classification problem.

[0068] After training the MALSTM-MFCN model (multimodal classification model) with the best performance indicators, this application extracts the coordinate information of the key points of the human skeleton in the video stream, and extracts the features of all frames through GCN to obtain new time series data, which is input into the trained MALSTM-MFCN model. The motor ability evaluation result corresponding to the input video stream can be obtained to achieve the effect of automatic scoring. The evaluation algorithm process corresponding to this application can be found in Figure 6 .

[0069] This application provides an automatic assessment method for the motor ability of the elderly based on the GCN-MALSTM-MFCN model. According to the assessment results, the kinetic energy state of the upper and lower limbs of the subject can be judged, including the performance of muscle strength, coordination, flexibility, etc., to provide reference and basis for medical personnel. Based on the existing samples, this application uses GCN to extract skeleton features and train the MALSTM-MFCN model. The trained model can be used directly to automatically evaluate new videos in the future, making the work of motor ability assessment more efficient. At the same time, the obtained assessment can also provide help for medical personnel.

[0070] This application uses computer vision to detect the motor function of the upper and lower limbs of the elderly, and can automatically evaluate the motor ability of the elderly, aiming to solve the following problems:

[0071] 1. Automatically detect the upper and lower limb movement assessment of the elderly. To address the issue of motor ability assessment for the elderly, high-precision computer vision technology is used to capture the coordinates of key points of the patient's body during movement, analyze the patient's movement status, and achieve real-time monitoring and analysis of the elderly's motor ability.

[0072] 2. Realize the objectivity and standardization of automatic evaluation of motor ability. Traditional motor ability evaluation usually relies on subjective observation and manual testing by medical personnel, so there are problems of subjectivity and inaccuracy. However, this application can realize objective and automated detection of the motor ability of the elderly by using computer vision technology, thereby improving the objectivity and standardization of the evaluation.

[0073] 3. Realize the real-time automatic assessment of the motor ability of the elderly. Traditional motor ability assessment often requires a lot of time and human resources. The manual screening of large amounts of data and the instability of data collection may be a burden for doctors. This application can achieve real-time monitoring and analysis of motor ability, greatly improving the efficiency of the assessment, helping to timely detect motor changes in the elderly and conduct targeted rehabilitation interventions. By analyzing the results of the elderly's motor assessment, appropriate rehabilitation training suggestions can be provided for the elderly.

[0074] 4. Realize the continuity and tracking of the motor ability assessment of the elderly. The decline in motor ability is usually a continuous process that requires long-term tracking and evaluation. This application can realize the continuous monitoring and long-term tracking of the motor ability of the elderly, and can record the movement status and change trends of the subjects in real time, provide data support for doctors, and help to develop personalized rehabilitation plans and treatment plans for the elderly with reduced motor ability. The real-time feedback results can evaluate the treatment effect so that doctors can adjust the rehabilitation plan in time.

[0075] 5. Realize data collection and collation in the field of motor ability assessment. Sorting out a large amount of data on the elderly through the database will help further study the algorithm of the elderly's motor ability; by accumulating a large amount of data from subjects and combining it with deep learning algorithms, we can continuously optimize the multimodal classification model and improve the accuracy of diagnosis and treatment.

[0076] Based on the same inventive concept, the embodiment of the present application also provides an automatic assessment device for the motor ability of the elderly for implementing the automatic assessment method for the motor ability of the elderly involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the automatic assessment device for the motor ability of the elderly provided below can refer to the limitations of the automatic assessment method for the motor ability of the elderly above, and will not be repeated here.

[0077] In an exemplary embodiment, a device for automatically assessing the motor ability of the elderly is provided, comprising:

[0078] The test acquisition and input module is used to obtain the test video of the subject; the test video includes: the video of the subject putting his hands behind his head, squatting and sitting;

[0079] The motion ability feature extraction module is used to extract the skeleton key point data from the video to be tested by using the key point detection algorithm; according to the skeleton key point data, the skeleton features are determined by using the feature extraction network model obtained by training based on the graph convolution network;

[0080] The physical motor ability assessment module is used to determine the motor ability assessment results based on skeleton features using a multimodal classification model trained based on long short-term memory networks and fully convolutional networks.

[0081] In another exemplary embodiment, the automatic assessment device for the motor ability of the elderly further includes: a body motor ability assessment report module;

[0082] The body movement ability assessment report module is used to generate a movement ability assessment report from the movement ability assessment result.

[0083] In an exemplary embodiment, the test collection and entry module also includes: a motion video recording device and a motion data collection device; the information collected by the motion video recording device and the motion data collection device is entered and used as the basic data and basic model of the physical movement ability evaluation module.

[0084] The automatic assessment device for the exercise ability of the elderly provided by this application has the following technical effects:

[0085] 1. The module is simple and clear;

[0086] 2. The functions of each module are clear and the design is reasonable;

[0087] 3. The movement is classified according to the computer program, which avoids the subjectivity and error of human classification and makes the test more accurate;

[0088] 4. Separately assessing the upper and lower limb movements of the subject can help you better understand the subject's motor abilities and provide targeted rehabilitation training suggestions for the upper or lower limbs;

[0089] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for automatically evaluating the motor ability of the elderly is implemented.

[0090] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0091] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0092] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0093] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0094] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0095] In this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0096] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0097] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for automatically evaluating the motor ability of the elderly, characterized in that: The automatic assessment method for the motor ability of the elderly comprises: Obtain a video of the subject to be tested; the video to be tested includes: a video of the subject placing his hands behind his head, squatting, and sitting; Use key point detection algorithm to extract skeleton key point data from the video to be tested; According to the skeleton key point data, the feature extraction network model trained based on the graph convolutional network is used to determine the skeleton features; According to the skeleton features, a multimodal classification model trained based on long short-term memory network and fully convolutional network is used to determine the motor ability assessment results.

2. The method for automatically evaluating the motor ability of the elderly according to claim 1, characterized in that: The obtaining of the test video of the subject specifically includes: Using video equipment to film the subjects; Use OpenCV to obtain the video to be tested.

3. The method for automatically evaluating the motor ability of the elderly according to claim 1, characterized in that: The key point detection algorithm is the HrNet algorithm.

4. The method for automatically evaluating the motor ability of the elderly according to claim 1, characterized in that: The loss function of the multimodal classification model adopts the cross-entropy loss function.

5. The method for automatically evaluating the motor ability of the elderly according to claim 1, characterized in that: Get the video of the subject to be tested, and then include: Perform data cleaning and data filling operations on the video to be tested.

6. An automatic assessment device for the motor ability of the elderly, characterized in that: The automatic assessment device for the elderly's motor ability comprises: The test acquisition and input module is used to obtain the test video of the subject; the test video includes: the video of the subject putting his hands behind his head, squatting and sitting; The motion ability feature extraction module is used to extract the skeleton key point data from the video to be tested by using the key point detection algorithm; according to the skeleton key point data, the skeleton features are determined by using the feature extraction network model obtained by training based on the graph convolution network; The physical motor ability assessment module is used to determine the motor ability assessment results based on skeleton features using a multimodal classification model trained based on long short-term memory networks and fully convolutional networks.

7. The automatic assessment device for the motor ability of the elderly according to claim 6 is characterized in that: The automatic assessment device for the motor ability of the elderly also includes: a body motor ability assessment report module; The body movement ability assessment report module is used to generate a movement ability assessment report from the movement ability assessment result.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for automatically assessing the motor ability of the elderly as described in any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for automatically evaluating the motor ability of the elderly described in any one of claims 1 to 5 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for automatically evaluating the motor ability of the elderly described in any one of claims 1 to 5 is implemented.

Images