Rehabilitation exercise posture evaluation method, device, medium and system
By acquiring and analyzing the user's rehabilitation project movement images and feedback information, and combining multi-dimensional and humanized judgment, the problems of subjectivity and high equipment cost in rehabilitation movement posture assessment in existing technologies are solved, and more accurate and efficient assessment results are achieved.
Patent Information
- Application Number
- CN202510800661.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-16
- Publication Date
- 2025-10-24
AI Technical Summary
Existing rehabilitation exercise posture assessment methods rely on manual observation and high-cost equipment. They are highly subjective, have poor repeatability, are inefficient, and lack automated and standardized data analysis. They are difficult to popularize in primary hospitals, and existing algorithms lack adaptability and stability in complex environments.
By continuously acquiring images of users performing rehabilitation programs, identifying key joint angles of the human body, judging whether the movements meet the standards based on preset standard conditions, recording changes in joint angles, and obtaining user feedback, an evaluation report is ultimately generated. Evaluation is conducted using a multi-dimensional and humanized judgment method, combined with voice prompts and feedback information.
It achieves more accurate rehabilitation movement posture assessment, improves user experience, reduces equipment costs, improves the automation and standardization of assessment, adapts to complex environments, and reduces subjective errors.
Smart Images
Figure CN120833879A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of limb rehabilitation assessment, and in particular to a rehabilitation exercise posture assessment method, device, medium and system. BACKGROUND
[0002] At present, the existing rehabilitation exercise posture assessment method mainly relies on the visual observation and manual measurement (such as using a protractor, a camera and a sensor, etc.) of a rehabilitation physician or a therapist and combines experience for assessment and guidance, but the following problems are prone to occur.
[0003] Firstly, in terms of measurement methods, manual observation, video playback, and scales, angle meters and other instruments are generally used for detection, and the detection accuracy depends on the experience of the physician, and there are problems of strong subjectivity, poor repeatability, and low efficiency; secondly, in terms of device cost and deployment, some high-end motion detection systems use high-cost special equipment and installation and debugging, such as wearable devices such as inertial measurement units (IMU), electromyographic sensors (SEMG), and flexible sensors, which are difficult to popularize to primary hospitals and sports training sites; thirdly, in terms of data processing, existing systems can only provide limited angle data or static images, and cannot comprehensively assess common rehabilitation assessment items such as straight leg raising, Thomas test, and body flexion, and professional reports rely on manual interpretation, lacking automated, standardized data analysis and comprehensive assessment.
[0004] Further, in order to improve the rehabilitation effect assessment efficiency and avoid the subjective judgment of the rehabilitation physician, a posture recognition scheme based on a deep learning model is proposed, however, this method generally faces the problems of insufficient training samples, large individual differences, and weak generalization ability, especially in actual rehabilitation scenarios, the environment is complex, and the device deployment is limited, which challenges the adaptability and stability of the algorithm.
[0005] Further, in order to avoid the problems of insufficient training samples and large individual differences, a module in the open source MediaPipe model is used to realize the recognition of the rehabilitation exercise posture, although 0-sample recognition can be realized, but using these modules alone to assess the rehabilitation exercise posture still has the following deficiencies, only angle data output can be realized, the assessment dimension is relatively single, and the user experience is poor.
[0006] Therefore, it is urgent to design a rehabilitation exercise posture assessment method, which can make the assessment result more accurate through multi-dimensional, dynamic and humanized judgment, and thus improve the user experience. SUMMARY
[0007] The main purpose of the embodiments of the present application is to at least solve one of the technical problems existing in the prior art, and propose a rehabilitation exercise posture evaluation method, device, medium and system, which can make the evaluation result more accurate through multi-dimensional, dynamic and personalized judgment, and thus improve the user experience.
[0008] To achieve the above-mentioned purpose, in a first aspect, the embodiments of the present application propose a rehabilitation exercise posture evaluation method, which comprises: continuously acquiring action images of a user performing a current rehabilitation project; identifying the action images to obtain human body key joint angles; judging whether the action meets the standard according to the human body key joint angles and a preset standard condition; if the action meets the standard, continuously recording the change value of the human body key joint angles; obtaining a target action image according to the change value of the human body key joint angles and a first preset value, and saving the human body key joint angles corresponding to the target action image as target human body key joint angles; acquiring feedback information of the user; obtaining an evaluation report of the current rehabilitation project according to the target action image, the target human body key joint angles and the feedback information, and prompting the user to perform a next rehabilitation project.
[0009] In some embodiments, the identification of the action images to obtain human body key joint angles comprises: identifying the action images to obtain human body key points, the human body key points comprising a left shoulder, a left hip, a left knee, a left ankle, a left big toe, a right shoulder, a right hip, a right knee, a left palm root and a finger root; acquiring coordinates of the human body key points; obtaining human body key joint angles according to the coordinates of the human body key points.
[0010] In some embodiments, the human body key joint angles comprise hip joint angles, knee joint angles and ankle joint angles; the preset standard condition comprises a first preset condition, a second preset condition and a third preset condition; and the judgment of whether the action meets the standard according to the human body key joint angles and the preset standard condition comprises: if the hip joint angles meet the first preset condition, the knee joint angles meet the second preset condition and the ankle joint angles meet the third preset condition, it is judged that the action meets the standard; if the hip joint angles do not meet the first preset condition, or the knee joint angles do not meet the second preset condition, or the ankle joint angles do not meet the third preset condition, it is judged that the action does not meet the standard.
[0011] In some embodiments, the method further comprises: if the action is not up to standard, prompting the user to adjust the human key joint angle through voice.
[0012] In some embodiments, the target action image is obtained according to the change value of the human key joint angle and a first preset value, comprising: if the change value of the human key joint angle is lower than the first preset value within a first preset time length, the current action image is intercepted as the target action image.
[0013] In some embodiments, the evaluation report of the current rehabilitation project is obtained according to the target action image, the target human key joint angle and the feedback information, comprising: an evaluation report of the current rehabilitation project is obtained according to the target action image, the target human key joint angle, the feedback information and preset medical guideline information; or, the target action image, the target human key joint angle and the feedback information are transmitted to a first preset report generation model to obtain an evaluation report of the current rehabilitation project.
[0014] In some embodiments, the method further comprises: if it is detected that the user has completed all the rehabilitation projects, transmitting all the evaluation reports corresponding to the rehabilitation projects to a second preset report generation model to obtain a comprehensive evaluation report.
[0015] In a second aspect, an embodiment of the present application provides a running control device, comprising a memory and a processor, the memory stores a computer program, and the processor implements the evaluation method of the rehabilitation exercise posture when executing the computer program.
[0016] In a third aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program implements the evaluation method of the rehabilitation exercise posture when executed by a processor.
[0017] In a fourth aspect, an embodiment of the present application provides an evaluation system of a rehabilitation exercise posture, which comprises: an image acquisition module, which is used to continuously acquire action images of a user performing a current rehabilitation project; a posture recognition module, which is used to recognize the action images; an angle analysis module, configured to obtain a human key joint angle according to the action image, determine whether the action meets a standard according to the human key joint angle and a preset standard condition, continuously record a change value of the human key joint angle, obtain a target action image according to the change value of the human key joint angle and a first preset value, and save a human key joint angle corresponding to the target action image as a target human key joint angle; a voice recognition module, configured to obtain feedback information of a user; an evaluation module, configured to obtain an evaluation report of a current rehabilitation project according to the target action image, the target human key joint angle and the feedback information; a voice broadcast module, configured to prompt the user to perform a next rehabilitation project.
[0018] According to the rehabilitation movement posture evaluation method, device, medium and system provided by the embodiment of the present application, in the process of evaluating a single rehabilitation project, first, the action image of the user performing the current rehabilitation project is continuously obtained, the action image is recognized to obtain the human key joint angle, whether the action meets the standard is determined according to the human key joint angle and the preset standard condition, if the action meets the standard, in order to obtain the action image and the human key joint angle in the stable state, it is further determined whether the stable state is reached, specifically, the change value of the human key joint angle is continuously recorded, the target action image is obtained according to the change value of the human key joint angle and the first preset value, and the human key joint angle corresponding to the target action image is saved as the target human key joint angle, further, after obtaining the action image and the human key joint angle in the stable state, the user has performed the current rehabilitation project, the feedback information of the user is continuously obtained, finally, the evaluation report of the current rehabilitation project is obtained according to the target action image, the target human key joint angle and the feedback information, and the user is prompted to perform the next rehabilitation project, compared with the evaluation scheme of the prior art which can only realize the output of the angle data, on the basis of the target action image and the target human key joint angle, the feedback information of the user is further combined to make a deeper evaluation, so that the evaluation result is more accurate, and the use experience of the user is improved. BRIEF DESCRIPTION OF DRAWINGS
[0019] The accompanying drawings are used to provide a further understanding of the technical scheme of the present application, and constitute a part of the specification, and are used to explain the technical scheme of the present application together with the embodiments of the present application, and do not constitute a limitation on the technical scheme of the present application.
[0020] The present application is further illustrated below in combination with the drawings and embodiments; Figure 1It is a whole flow schematic diagram of a rehabilitation exercise posture evaluation method provided by the application; Figure 2 It is a detailed flow schematic diagram of step S200 in the rehabilitation exercise posture evaluation method provided by the application; Figure 3 It is a detailed flow schematic diagram of step S300 in the rehabilitation exercise posture evaluation method provided by the application; Figure 4 It is a detailed flow schematic diagram of step S700 in the rehabilitation exercise posture evaluation method provided by the application; Figure 5 It is a schematic diagram of a first target action image provided by the application; Figure 6 It is a schematic diagram of a second target action image provided by the application; Figure 7 It is a schematic diagram of a third target action image provided by the application. DETAILED DESCRIPTION
[0021] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.
[0022] It should be noted that the terms "first", "second" and the like in the description and claims and the above-described drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence.
[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.
[0024] In the prior art, in order to avoid the problems of insufficient training samples and large individual differences, a module in the open source MediaPipe model is proposed to realize the recognition of rehabilitation exercise postures. Although 0-sample recognition can be realized, there are still the following deficiencies in evaluating rehabilitation exercise postures by using these modules alone. Only angle data output can be realized, the evaluation dimension is relatively single, and the user experience is poor.
[0025] Based on this, the embodiments of the present application provide a rehabilitation exercise posture evaluation method, device, medium and system, which make the evaluation result more accurate through multi-dimensional, dynamic and humanized judgment, and further improve the user experience.
[0026] The embodiments of the present application will be further described below in combination with the drawings.
[0027] In a first aspect, referring to Figure 1 The application provides a rehabilitation exercise posture evaluation method and system. The method comprises the following steps: S100: continuously acquiring action images of a user performing a current rehabilitation project; S200: identifying the action images to obtain human body key joint angles; S300: determining whether the action meets a standard according to the human body key joint angles and preset standard conditions; S410: if the action meets the standard, continuously recording a change value of the human body key joint angles; S500: obtaining a target action image according to the change value of the human body key joint angles and a first preset value, and saving a human body key joint angle corresponding to the target action image as a target human body key joint angle; S600: acquiring feedback information of the user; S700: obtaining an evaluation report of the current rehabilitation project according to the target action image, the target human body key joint angle and the feedback information, and prompting the user to perform a next rehabilitation project.
[0028] It should be noted that before the step S100, the method further comprises prompting the user to start performing the current rehabilitation project through voice, wherein the rehabilitation project comprises a straight leg raise test, a Thomas test and a sit-and-reach test.
[0029] It should be noted that before the step S600, the method further comprises prompting the user to select subjective feeling information in the test process. Specifically, for example, the method provides options such as whether the user feels pain, numbness or stretching, and displays the options on an operation interface of the evaluation system or plays the options through voice to prompt the user to select, and correspondingly, the user can manually select to provide the feedback information on the interface or output the option information through voice to obtain and identify the feedback information by the evaluation system. It should be noted that the step S410 of continuously recording the change value of the human body key joint angles comprises continuously recording an angle change curve of the human body key joint angles changing over time, and obtaining the change value of the human body key joint angles according to the angle change curve.
[0030] In some embodiments, referring to Figure 2 The step S200 of identifying the action images to obtain the human body key joint angles comprises the following steps: S210: identifying the action images to obtain human body key points, wherein the human body key points comprise a left shoulder, a left hip, a left knee, a left ankle, a left big toe, a right shoulder, a right hip, a right knee, a left palm root and a finger root; S220: Obtain the coordinates of the human body key points. S230: Obtain the human body key joint angles according to the coordinates of the human body key points.
[0031] It should be noted that based on the coordinates of the human body key points, the human body key joint angles are obtained by the three-point angle method. Specifically, first, based on the coordinates of the three human body key points of the shoulder, the hip and the knee, the hip joint angle can be obtained by the three-point angle method; second, based on the coordinates of the three human body key points of the hip, the knee and the ankle, the knee joint angle can be obtained by the three-point angle method; and third, based on the coordinates of the three human body key points of the knee, the ankle and the foot, the ankle joint angle can be obtained by the three-point angle method.
[0032] It should be noted that if the current rehabilitation project is the straight leg raising test, the user needs to complete the corresponding action of the straight leg raising test, and the MediaPipe Pose model is used to identify the human body key points such as the shoulder, the hip, the knee and the ankle in the action image. Further, the human body key points are connected to calculate the coordinates of the human body key points. Then, based on the coordinates of the human body key points, the hip joint angle, the knee joint angle and the ankle joint angle are obtained by the three-point angle method, realizing the automatic recognition of the straight leg raising test.
[0033] It should be noted that if the current rehabilitation project is the Thomas test, the user needs to complete the corresponding action of the Thomas test, and the MediaPipe Pose model is used to identify the human body key points such as the left shoulder, the left hip, the left knee, the right shoulder, the right hip, the right knee and the left ankle in the action image. Further, the human body key points are connected to calculate the coordinates of the human body key points. Then, based on the coordinates of the human body key points, the left hip joint angle, the left knee joint angle and the right hip joint angle are obtained by the three-point angle method.
[0034] It should be noted that if the current rehabilitation project is the forward bending test, the user needs to complete the corresponding action of the forward bending test. First, the MediaPipe Pose model is used to identify the human body key points such as the shoulder, the hip, the knee and the ankle in the action image. Second, the MediaPipe Hands model is used to identify the human body key points such as the left thumb and the left finger tip, the left finger root and the left palm root in the action image. Further, the human body key points are connected to calculate the coordinates of the human body key points. Then, based on the coordinates of the human body key points, the hip joint angle, the knee joint angle and the ankle joint angle are obtained by the three-point angle method. In addition, based on the coordinates of the human body key points, the distances between the left thumb and the left finger tip, the left finger root and the left palm root are obtained. Further, the flexibility data is obtained according to the distances between the left thumb and the left finger tip, the left finger root and the left palm root.
[0035] In some embodiments, the human key joint angles include a hip joint angle, a knee joint angle, and an ankle joint angle; the preset standard conditions include a first preset condition, a second preset condition, and a third preset condition; the reference Figure 3 In step S300, determining whether the movement meets the standard according to the human key joint angles and the preset standard conditions includes step S310 or step S320: Step S310: If the hip joint angle meets the first preset condition, the knee joint angle meets the second preset condition, and the ankle joint angle meets the third preset condition, it is determined that the movement meets the standard. It should be noted that if the current rehabilitation project is a straight leg raising test, the first preset condition corresponds to a hip joint angle less than 150°, indicating effective leg lifting; the second preset condition corresponds to a knee joint angle greater than 170°, indicating that the knee joint is straight; and the third preset condition corresponds to an ankle joint angle between 90° and negative 120°, indicating that the foot is hooked up; therefore, if the hip joint angle meets the first preset condition, the knee joint angle meets the second preset condition, and the ankle joint angle meets the third preset condition, it is determined that the movement meets the standard.
[0036] Step S320: If the hip joint angle does not meet the first preset condition, or the knee joint angle does not meet the second preset condition, or the ankle joint angle does not meet the third preset condition, it is determined that the movement does not meet the standard.
[0037] In some embodiments, if the current rehabilitation project is the Thomas test, the first preset condition corresponds to a left hip joint angle between 80° and 180° and a right hip joint angle less than 80°, and the second preset condition corresponds to a left knee joint angle between 55° and negative 55°; therefore, if the left hip joint angle and the right hip joint angle both meet the first preset condition, and the left knee joint angle meets the second preset condition, it is determined that the movement meets the standard.
[0038] In some embodiments, if the current rehabilitation project is the trunk forward bending test, the first preset condition corresponds to a hip joint angle between 80° and 100°, the second preset condition corresponds to a knee joint angle greater than 170°, and the third preset condition corresponds to an ankle joint angle between 100° and 120°; therefore, if the hip joint angle meets the first preset condition, the knee joint angle meets the second preset condition, and the ankle joint angle meets the third preset condition, it is determined that the movement meets the standard.
[0039] In some embodiments, with reference to Figure 1 The method further includes step S420: if the movement does not meet the standard, prompting the user to adjust the human key joint angles through voice.
[0040] It should be noted that if the current rehabilitation project is the body forward bending test, the knee joint angle in the human key joint angle does not meet the standard, and the user is prompted by voice "knee joint is not straightened", thereby prompting the user to adjust the knee joint angle; It should be noted that if the current rehabilitation project is the body forward bending test, the hip joint angle in the human key joint angle does not meet the standard, and the user is prompted by voice "ankle is not hooked", thereby prompting the user to adjust the hip joint angle; It can be understood that until the human key joint angle in the current rehabilitation project meets the standard, the voice broadcast is stopped, which is helpful to help the patient correct the error action in time, guide the user to complete the standard action, and avoid invalid training or secondary injury due to angle deviation.
[0041] In some embodiments, the target action image is obtained according to the change value of the human key joint angle and the first preset value in step S500, comprising: if the change value of the human key joint angle is lower than the first preset value within the first preset time length, the current action image is intercepted as the target action image.
[0042] Preferably, the first preset time length is 5s, and the first preset value is 5°. If the change value of the human key joint angle is lower than 5° within 5s, it indicates that the current action image is the action image in the stable state, which is intercepted and saved as the target action image, and is displayed in the subsequent generated evaluation report. Further, the target action image can be marked as a standard sample to improve the accuracy of subsequent detection and evaluation.
[0043] Further, the skeleton line formed by superimposing the human key point connection on the target action image, the extrapolation for estimating the distance between the toe position and the hand key point, and the human key point mark are added. In the process of rehabilitation training, by comparing the past target action images, the changes of the skeleton line and the extrapolation circle are intuitively understood, the rehabilitation training regulations are continuously optimized according to the changes of the skeleton line and the extrapolation circle, the scientificity of the rehabilitation training process is improved, and the rehabilitation process of the user is accelerated. In addition, the human key joint angle data is attached to the target action image, so that the user can not only understand the action posture in the rehabilitation process from the target action image, but also more intuitively understand the human key joint angle data corresponding to the action posture in the rehabilitation process, without manual measurement, thereby improving the user experience.
[0044] In some embodiments, with reference to Figure 4 , the evaluation report of the current rehabilitation project is obtained according to the target action image, the target human key joint angle and the feedback information in step S700, comprising step S710 or S720: Step S710: obtaining the evaluation report of the current rehabilitation project according to the target action image, the target human key joint angle, the feedback information and the preset medical guideline information; Or, step S720: transmitting the target action image, the target human body key joint angle and the feedback information to the first preset report generation model to obtain an evaluation report of the current rehabilitation project.
[0045] It can be understood that by fusing and analyzing the feedback information such as "numbness", "pulling sensation", "pain" of the user, the target action image and the target human body key joint angle, a more professional evaluation report can be generated for detailed evaluation of the current rehabilitation project.
[0046] It should be noted that the evaluation report of the current rehabilitation project includes the target action image, the posture execution condition evaluation, the improvement suggestion, the recommended rehabilitation action and the possible muscle or skeletal problem speculation, wherein, with reference to Figure 5 、 Figure 6 、 Figure 7 , the target action image is additionally provided with the data display of the connection line of the human body key points and the human body key joint angle, and the muscle or skeletal problem includes muscle tension, nerve pulling and insufficient knee joint control.
[0047] Preferably, by setting the preset medical guideline information, the target human body key joint angle, the feedback information and the target action image are processed through the preset medical guideline information to generate the evaluation report of the current rehabilitation project, thereby avoiding the problems of low efficiency and strong subjectivity in traditional manual analysis.
[0048] Preferably, the first preset report generation model is a large language model (such as Qwen2.5-Instruct), which automatically generates the evaluation report of the current rehabilitation project by means of semantic understanding and medical knowledge transfer of the target human body key joint angle and the feedback information, and gives targeted suggestions in the evaluation report, thereby significantly improving the analysis efficiency and content quality.
[0049] In some embodiments, the method further includes step S800: if it is detected that the user has completed all the rehabilitation projects, transmitting the evaluation reports corresponding to all the rehabilitation projects to a second preset report generation model to obtain a comprehensive evaluation report.
[0050] It should be noted that the rehabilitation projects include the straight leg raise test, the Thomas test and the body forward bending test, and if it is detected that the user has completed the three rehabilitation projects of the straight leg raise test, the Thomas test and the body forward bending test, the three evaluation reports corresponding to the straight leg raise test, the Thomas test and the body forward bending test are transmitted to the second preset report generation model to generate a more comprehensive comprehensive evaluation report.
[0051] It can be understood that when the evaluation report is output for a single rehabilitation project, the data between the straight leg raising test, the Thomas test and the body flexion test projects are fragmented and fail to form a complete rehabilitation evaluation system. By transmitting all the evaluation reports corresponding to the rehabilitation projects to the second preset report generation model to obtain a comprehensive evaluation report, the user's lower limb symptoms can be comprehensively evaluated and more reasonable rehabilitation suggestions can be given, which makes up for the dependence on artificial expertise in traditional evaluation. In addition, by transmitting all the evaluation reports corresponding to the rehabilitation projects to the second preset report generation model to obtain a comprehensive evaluation report, the comprehensive evaluation report contains the detection time and all the target motion images of the three rehabilitation projects of the leg raising test, the Thomas test and the body flexion test, as well as comprehensive suggestions, which is convenient for the user to archive and subsequent tracking.
[0052] Preferably, the second preset report generation model can be the same as the first preset report generation model, both being a large language model (such as Qwen2.5-Instruct).
[0053] Further, the system automatically calls a document interface (python-docx) to output and save the comprehensive evaluation report in the form of a document, which is convenient for doctors to remotely analyze, rehabilitate, track and manage individual cases or for users to self-evaluate.
[0054] The following is a complete embodiment of the process of the user performing multiple rehabilitation projects to obtain a comprehensive evaluation report: First, referring to Figure 5 , continuously acquire the motion images of the user performing the straight leg raising test project; identify the motion images to obtain the hip joint angle, the knee joint angle and the ankle joint angle; if the hip joint angle is less than 150°, the knee joint angle is greater than 170° and the ankle joint angle is between 90° and negative 120°, the motion is up to standard; if the motion is up to standard, the user is prompted to maintain the motion through voice, and the change values of the hip joint angle, the knee joint angle and the ankle joint angle are continuously recorded; if the change values of the hip joint angle, the knee joint angle and the ankle joint angle are less than 5° within 5 seconds, it indicates that the current motion image is a motion image in a stable state, which is intercepted and saved as a first target motion image and the hip joint angle, the knee joint angle and the ankle joint angle corresponding to the first target motion image are saved; acquire the first feedback information of the user; obtain the evaluation report of the straight leg raising test project according to the first target motion image, the hip joint angle, the knee joint angle and the ankle joint angle corresponding to the first target motion image and the first feedback information, and prompt the user to enter the Thomas test project through voice.
[0055] Secondly, referring to Figure 6, continuously acquire action images of the user performing the Thomas test project; identify the action images to obtain a left knee joint angle, a left hip joint angle, and a right hip joint angle; if the left hip joint angle is between 80° and 180°, the right hip joint angle is less than 80°, and the left knee joint angle is between 55° and -55°, the action is qualified; if the action is qualified, prompt the user to maintain the action through voice, and continuously record the change values of the left knee joint angle, the left hip joint angle, and the right hip joint angle; if the change values of the left knee joint angle, the left hip joint angle, and the right hip joint angle are less than 5° within 5 seconds, it indicates that the current action image is an action image in a stable state, which is intercepted and saved as a second target action image and the left knee joint angle, the left hip joint angle, and the right hip joint angle corresponding to the second target action image are saved; acquire second feedback information of the user; obtain an evaluation report of the Thomas test project according to the second target action image, the left knee joint angle, the left hip joint angle, and the right hip joint angle corresponding to the second target action image, and the second feedback information, and prompt the user to enter the sit-and-reach test project through voice.
[0056] Next, with reference to Figure 7 , continuously acquire action images of the user performing the sit-and-reach test project; identify the action images to obtain a hip joint angle, a knee joint angle, and an ankle joint angle; if the knee joint angle is greater than 170° and the ankle joint angle is between 100° and 120°, broadcast “please sit straight and flatten your hands” through voice, and continue to the subsequent judgment; if the hip joint angle is between 80° and 100°, broadcast “please stretch your hands forward” through voice; continuously record the change values of the left knee joint angle, the left hip joint angle, and the right hip joint angle; if the hip joint angle is between 80° and 100° within 5 seconds, it indicates that the current action image is an action image in a stable state, which is intercepted and saved as a third target action image and the hip joint angle, the knee joint angle, and the ankle joint angle corresponding to the third target action image are saved; acquire third feedback information of the user; obtain an evaluation report of the sit-and-reach test project according to the third target action image, the hip joint angle, the knee joint angle, and the ankle joint angle corresponding to the third target action image, and the third feedback information; in addition, the evaluation report of the sit-and-reach test project also includes flexibility degree information, specifically, by judging the distance between the left foot big toe and the finger tip, the finger root and the palm root, confirming one state that meets from the five states of “not touching the toe”, “touching the toe with the finger tip”, “touching the toe with the finger root”, and “touching the toe with the palm root” is displayed in the evaluation report of the sit-and-reach test project.
[0057] Finally, the evaluation report of the straight leg raise test project, the evaluation report of the Thomas test project, and the evaluation report of the sit-and-reach test project are all transmitted to the second preset report generation model to obtain a comprehensive evaluation report.
[0058] In a second aspect, an embodiment of the present invention proposes an operation control device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements a method for evaluating rehabilitation exercise posture as described in any one of the embodiments of the first aspect.
[0059] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer executable programs. Furthermore, memory can include high-speed random access memory and non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device.
[0060] In some embodiments, the memory may optionally include a memory remotely located relative to the processor, and the remote memory may be connected to the processor via a network. Examples of the aforementioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0061] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a method for evaluating rehabilitation exercise posture as described in any one of the embodiments of the first aspect.
[0062] In a fourth aspect, an embodiment of the present invention provides a rehabilitation exercise posture evaluation system, which is applied to a rehabilitation exercise posture evaluation device. The system includes: An image acquisition module, which is used to continuously acquire action images of the user performing the current rehabilitation program; A gesture recognition module is used to recognize action images; An angle analysis module is used to obtain the key joint angles of the human body according to the action image; determine whether the action meets the standard according to the key joint angles of the human body and the preset standard conditions; continuously record the change value of the key joint angles of the human body; obtain the target action image according to the change value of the key joint angles of the human body and the first preset value, and save the key joint angles of the human body corresponding to the target action image as the target key joint angles of the human body; Speech recognition module, which is used to obtain user feedback information; Evaluation module, which is used to obtain an evaluation report of the current rehabilitation project based on the target action image, target key joint angles of the human body, and feedback information; The voice broadcast module is used to prompt the user to perform the next rehabilitation project.
[0063] Preferably, the image acquisition module can be a camera, the posture recognition module can be MediaPipe Pose and MediaPipe Hands, which can detect human key points in real time, automatically extract and calculate key joint angles such as hips, knees and ankles in real time, replace manual measurement, and improve the accuracy and real-time performance of evaluation; in addition, high-precision posture detection can be achieved using only a common camera and open-source software modules, reducing the overall cost of the system.
[0064] It should be noted that the voice recognition module transmits the voice information of the user to the voice recognition library (for example, SpeechRecognition) to obtain the feedback information of the user, without the need for the user to manually select after approaching the device interface, thereby improving the user experience.
[0065] It should be noted that the voice broadcast module is also used to prompt the user to adjust the human key joint angle in a timely manner in the case that the action does not meet the standard, for example, in the body forward bending test, if the knee joint is less than 170 degrees, an error “knee joint not straightened” is prompted to enable the user to adjust the knee joint angle in a timely manner, and if the ankle joint is greater than 115 degrees, an error “ankle not hooked up” is prompted to enable the user to adjust the ankle joint angle in a timely manner; in addition, the voice broadcast module is also used to prompt the user to perform feedback, after obtaining the target action image, the user performs the current rehabilitation project, and the voice broadcast module is used to prompt the user to perform feedback to obtain the real-time subjective feeling of the user in a timely manner, so that the subsequent evaluation result is more accurate.
[0066] Preferably, the evaluation system of the rehabilitation movement posture is encapsulated using the PyQt5 framework, and a graphical interface is designed, the user can start the test through the way of clicking or voice instruction, the interface displays real-time video, angle information and detection result image at the same time, and has the function of automatic size adjustment.
[0067] According to the rehabilitation movement posture evaluation method, device, medium and system provided by the embodiment of the application, in the process of evaluating a single rehabilitation project, the action image of the user performing the current rehabilitation project is continuously acquired, the action image is identified to obtain the human body key joint angle, and whether the action meets the standard is determined according to the human body key joint angle and the preset standard condition. If the action meets the standard, in order to obtain the action image and the human body key joint angle in the stable state, it is necessary to further determine whether the stable state is reached. Specifically, the change value of the human body key joint angle is continuously recorded, the target action image and the human body key joint angle corresponding to the target action image are obtained according to the change value of the human body key joint angle and the first preset value, and the target human body key joint angle is saved. Further, after obtaining the action image and the human body key joint angle in the stable state, the user has completed the current rehabilitation project, and the feedback information of the user is continuously acquired. Finally, the evaluation report of the current rehabilitation project is obtained according to the target action image, the target human body key joint angle and the feedback information, and the user is prompted to perform the next rehabilitation project. Compared with the evaluation scheme in the prior art which can only realize the output of angle data, on the basis of obtaining the target action image and the target human body key joint angle, the feedback information of the user is further combined to evaluate more deeply, so that the evaluation result is more accurate, and the use experience of the user is improved.
[0068] The embodiments described in the embodiments of the application are used to more clearly illustrate the technical solutions of the embodiments of the application, and do not constitute a limitation on the technical solutions provided by the embodiments of the application. Those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the application are also applicable to similar technical problems.
[0069] Those skilled in the art can understand that the technical solutions shown in the drawings do not constitute a limitation on the embodiments of the application, and can include more or fewer steps than the drawings, or combine certain steps or different steps.
[0070] The above-described running control device embodiments are only schematic, and the units described as separate components can or can not be physically separated, that is, can be located in one place or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments.
[0071] Those skilled in the art can understand that all or some of the steps in the above method, the functions of the modules / units in the system and the device can be implemented as software, firmware, hardware and appropriate combinations thereof.
[0072] The terms "first", "second", "third", "fourth", and the like in the description of the application and in the claims hereof, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of the terms so termed herein is solely for the purpose of distinctability and that the embodiments of the application described herein can be carried out in other sequences than the one described herein without departing from the scope of the application. Further, the terms "comprise" and "comprising" and any variations thereof, as used in the specification and in the claims hereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.
[0073] In several embodiments provided in the present application, it should be understood that the disclosed operation control device and method can be implemented in other ways. For example, the above-described operation control device embodiments are only illustrative, for example, the division of the above-described units is only a logical function division, and actual implementation can have another division manner, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed units can be through some interfaces, indirect coupling or communication connection of the operation control device or units, which can be electrical, mechanical or other forms.
[0074] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or in the form of a software functional unit.
[0075] If the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes multiple instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program storage media.
[0076] The preferred embodiments of the present application are described above with reference to the accompanying drawings, and are not intended to limit the scope of the present application. Any modification, equivalent replacement and improvement made by those skilled in the art without departing from the scope and spirit of the present application shall fall within the scope of the present application.
Claims
1. A method of evaluating a rehabilitation movement posture, characterized by, The method comprises: continuously acquiring action images of a user performing a current rehabilitation project; identifying the action images to obtain human key joint angles; judging whether the action meets a standard according to the human key joint angles and a preset standard condition; if the action meets the standard, continuously recording a change value of the human key joint angles; obtaining a target action image according to the change value of the human key joint angles and a first preset value, and saving a human key joint angle corresponding to the target action image as a target human key joint angle; acquiring feedback information of the user; obtaining an evaluation report of the current rehabilitation project according to the target action image, the target human key joint angle and the feedback information, and prompting the user to perform a next rehabilitation project.
2. The evaluation method according to claim 1, characterized in that The identification of the action images to obtain human key joint angles comprises: identifying the action images to obtain human key points, the human key points comprising a left shoulder, a left hip, a left knee, a left ankle, a left big toe, a right shoulder, a right hip, a right knee, a left palm root and a finger root; acquiring coordinates of the human key points; obtaining human key joint angles according to the coordinates of the human key points.
3. The evaluation method according to claim 1, characterized in that The human key joint angles comprise hip joint angles, knee joint angles and ankle joint angles; the preset standard condition comprises a first preset condition, a second preset condition and a third preset condition; the judgment of whether the action meets the standard according to the human key joint angles and the preset standard condition comprises: if the hip joint angles meet the first preset condition, the knee joint angles meet the second preset condition and the ankle joint angles meet the third preset condition, it is judged that the action meets the standard; if the hip joint angles do not meet the first preset condition, or the knee joint angles do not meet the second preset condition, or the ankle joint angles do not meet the third preset condition, it is judged that the action does not meet the standard.
4. The evaluation method according to claim 1, characterized in that The method further comprises: if the action does not meet the standard, prompting the user to adjust the human key joint angles through voice.
5. The evaluation method according to claim 1, characterized in that The obtaining of the target action image according to the change value of the human key joint angles and the first preset value comprises: if the change value of the human key joint angles is lower than the first preset value within a first preset time length, a current action image is intercepted as the target action image.
6. The evaluation method according to claim 1, characterized in that The obtaining of the evaluation report of the current rehabilitation project according to the target action image, the target human key joint angle and the feedback information comprises: obtaining an evaluation report of a current rehabilitation project according to the target action image, the target human key joint angle, the feedback information and preset medical guideline information; or, transmitting the target action image, the target human key joint angle and the feedback information to a first preset report generation model to obtain an evaluation report of a current rehabilitation project.
7. The evaluation method according to claim 1, characterized in that The method further comprises: if it is detected that the user has performed all the rehabilitation projects, transmitting evaluation reports corresponding to all the rehabilitation projects to a second preset report generation model to obtain a comprehensive evaluation report.
8. A running control device characterized by comprising: The system comprises a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the rehabilitation exercise posture evaluation method of any one of claims 1 to 7.
9. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 8. The computer program is executed by the processor to implement the rehabilitation exercise posture evaluation method of any one of claims 1 to 7.
10. A system for evaluating a rehabilitation movement posture, characterized by, The system comprises: An image acquisition module, which is used to continuously acquire action images of a user performing a current rehabilitation project; A posture recognition module, which is used to recognize the action images; An angle analysis module, which is used to obtain human key joint angles according to the action images, determine whether the action meets the standard according to the human key joint angles and a preset standard condition, continuously record the change values of the human key joint angles, obtain a target action image according to the change values of the human key joint angles and a first preset value, and save the human key joint angles corresponding to the target action image as target human key joint angles; A voice recognition module, which is used to acquire feedback information of the user; An evaluation module, which is used to obtain an evaluation report of the current rehabilitation project according to the target action image, the target human key joint angles, and the feedback information; A voice broadcast module, which is used to prompt the user to perform a next rehabilitation project.