A rehabilitation training evaluation method, device, electronic device and storage medium

Abstract

The present application provides a rehabilitation training evaluation method, device, electronic device and storage medium, which relates to the field of rehabilitation equipment technology. The method includes responding to a rehabilitation training evaluation request, determining the target rehabilitation training indicated by the rehabilitation training evaluation request; determining the rehabilitation action execution stage detection rule to be executed based on the type of the target rehabilitation training; determining the rehabilitation action execution stage of the practitioner according to the determined rehabilitation action execution stage detection rule for the acquired rehabilitation action frame image of the practitioner; when the rehabilitation action execution stage of the practitioner meets the evaluation conditions, determining the real-time action evaluation value corresponding to the rehabilitation action frame image and feeding it back to the practitioner. It is used to improve the accuracy of the action evaluation of the practitioner's rehabilitation training.

Description

Technical Field

[0001] The present application relates to the technical field of rehabilitation equipment, and in particular to a method, device, electronic device and storage medium for evaluating rehabilitation training. Background Art

[0002] This technology is used in rehabilitation therapy related technologies. The evaluation and correction of patient movements is achieved based on image human key point detection technology.

[0003] Since during rehabilitation training, the practitioner performs an action over a period of time rather than in an instant, the evaluation of the practitioner's action standard should be based on a comprehensive assessment of all action states within the execution time of the action rather than just considering a certain key frame. Therefore, the determination of the evaluation time period is more important.

[0004] In addition, requiring the practitioner to perform a certain rehabilitation action within a relatively fixed time period will reduce their user experience and affect their rehabilitation enthusiasm. And if the patient completes the required rehabilitation action ahead of time or fails to complete the required rehabilitation action within the time period, it will have a greater impact on the evaluation results. Therefore, it is more important to accurately determine the rehabilitation action evaluation time period based on the user's insensitivity. Summary of the invention

[0005] The purpose of the embodiments of the present application is to provide a rehabilitation training evaluation method, device, electronic device and storage medium, so as to improve the accurate recognition of the movements of the practitioner at different stages of rehabilitation training, so as to improve the accuracy of rehabilitation training evaluation.

[0006] In a first aspect, the present invention provides a method for evaluating rehabilitation training, the method comprising responding to a rehabilitation training evaluation request, determining a target rehabilitation training indicated by the rehabilitation training evaluation request; determining a rehabilitation action execution phase detection rule to be executed based on the type of the target rehabilitation training; determining the rehabilitation action execution phase of the practitioner according to the determined rehabilitation action execution phase detection rule for an acquired rehabilitation action frame image of the practitioner; when the rehabilitation action execution phase of the practitioner meets the evaluation conditions, determining a real-time action evaluation value corresponding to the rehabilitation action frame image and feeding it back to the practitioner.

[0007] In an optional embodiment, when the type of rehabilitation training is two-state rehabilitation training, the rehabilitation action execution stage includes an evaluation stage and a non-evaluation stage. For each acquired rehabilitation action frame image, the rehabilitation action execution stage of the practitioner is determined in the following manner: for the rehabilitation action frame image, the real-time coordinates of the key points of the human body are extracted to calculate the first calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed; determine whether the first calculated similarity is less than a first preset similarity; if so, determine that the rehabilitation action execution stage of the practitioner is the non-evaluation stage; if not, determine that the rehabilitation action execution stage of the practitioner is the evaluation stage.

[0008] In an optional embodiment, when the type of rehabilitation training is three-state rehabilitation training, the rehabilitation action execution stage includes a non-evaluation stage, an evaluation preparatory stage and an evaluation stage. For each frame of rehabilitation action image acquired, the rehabilitation action execution stage of the practitioner is determined in the following manner: for the rehabilitation action frame image, the real-time coordinates of the key points of the human body are extracted to calculate the second calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the preparatory action of the rehabilitation action to be performed; determine whether the second calculated similarity is greater than the second preset similarity; if it is less than the second preset similarity, determine that the rehabilitation action execution stage of the practitioner is the non-evaluation stage; if it is greater than the second preset similarity, determine that the rehabilitation action execution stage of the practitioner is the evaluation preparatory stage, and calculate the third calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed. If the third calculated similarity is less than the third preset similarity, return to the step of calculating the second calculated similarity. If the third calculated similarity is greater than the third preset similarity, determine that the rehabilitation action execution stage of the practitioner is the evaluation stage.

[0009] In an optional implementation, if the evaluation phase is exited, the process returns to the step of calculating the second similarity.

[0010] In an optional implementation, whether the rehabilitation action execution stage of the practitioner meets the evaluation condition is determined in the following manner: when the rehabilitation action execution stage of the practitioner is the evaluation stage, it is determined that the rehabilitation action execution stage of the practitioner meets the evaluation condition.

[0011] In an optional embodiment, it also includes determining a final score of the rehabilitation training of the rehabilitation action to be performed by the practitioner based on the real-time action evaluation value corresponding to all the rehabilitation action frame images in the rehabilitation stage.

[0012] In an optional implementation, it also includes determining and displaying rehabilitation training guidance information based on the scores of the rehabilitation movements to be performed by the practitioner.

[0013] In a second aspect, the present invention provides a rehabilitation training evaluation device, the device comprising:

[0014] A response module, used to respond to the rehabilitation training evaluation request and determine the target rehabilitation training indicated by the rehabilitation training evaluation request;

[0015] A rule analysis module, used to determine the detection rules of the rehabilitation action execution phase to be performed based on the type of target rehabilitation training;

[0016] A state analysis module is used to determine the rehabilitation action execution stage of the practitioner according to the determined rehabilitation action execution stage detection rule based on the acquired rehabilitation action frame image of the practitioner;

[0017] The evaluation module is used to determine the real-time action evaluation value corresponding to the rehabilitation action frame image and feed it back to the practitioner when the rehabilitation action execution stage of the practitioner meets the evaluation conditions.

[0018] In a third aspect, the present invention provides an electronic device comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and the processor executes the machine-readable instructions to perform the steps of the rehabilitation training evaluation method as described in any of the aforementioned implementations.

[0019] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the rehabilitation training evaluation method in any of the aforementioned embodiments are executed.

[0020] The present application provides a method, device, electronic device and storage medium for evaluating rehabilitation training, wherein the method includes responding to a rehabilitation training evaluation request, determining the target rehabilitation training indicated by the rehabilitation training evaluation request; determining the rehabilitation action execution phase detection rule to be performed based on the type of the target rehabilitation training; determining the rehabilitation action execution phase of the practitioner according to the determined rehabilitation action execution phase detection rule for the acquired rehabilitation action frame image of the practitioner; when the rehabilitation action execution phase of the practitioner meets the evaluation conditions, determining the real-time action evaluation value corresponding to the rehabilitation action frame image and feeding it back to the practitioner. By classifying rehabilitation training and designing different action state detection rules for different types of rehabilitation training in combination with the state machine architecture, the state of the rehabilitation action performed by the practitioner can be better captured more accurately, giving a more accurate evaluation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A flowchart of a rehabilitation training evaluation method provided in an embodiment of the present application;

[0023] Figure 2 A flowchart of steps for determining the rehabilitation action execution stage of a practitioner provided in an embodiment of the present application;

[0024] Figure 3 A flowchart of another step of determining the rehabilitation action execution stage of a practitioner provided in an embodiment of the present application;

[0025] FIG4 (a) is a schematic diagram of a non-evaluation stage of a “standing straight leg lift” provided in an embodiment of the present application;

[0026] FIG4( b ) is a schematic diagram of an evaluation stage of “standing straight leg lift” provided in an embodiment of the present application;

[0027] FIG. 5 ( a ) is a non-evaluation stage intention of “assisted knee flexion in sitting position” provided in an embodiment of the present application;

[0028] FIG5( b ) is a schematic diagram of a preparatory stage of an assessment of “assisted knee flexion in a sitting position” provided in an embodiment of the present application;

[0029] FIG5( c ) is a schematic diagram of an evaluation stage of “assisted knee flexion in sitting position” provided in an embodiment of the present application;

[0030] Figure 6 A schematic diagram of the structure of a rehabilitation training evaluation device provided in an embodiment of the present application;

[0031] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] First, the application scenario of the present application is described. The technical solution of the present application can be applied to the rehabilitation training evaluation and guidance of the exerciser by the rehabilitation machine.

[0033] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0034] Embodiment 1

[0035] Figure 1A flowchart of a rehabilitation training evaluation method provided in an embodiment of the present application. Figure 1 As shown, an embodiment of the present application provides a method that can be performed through a rehabilitation application. The rehabilitation application here can be in the form of an APP, etc., installed in the practitioner's electronic device such as a mobile phone or a tablet. Among them, the evaluation method of rehabilitation training includes:

[0036] S10: respond to the rehabilitation training evaluation request and determine the target rehabilitation training indicated by the rehabilitation training evaluation request.

[0037] The practitioner can select the rehabilitation training that he / she wants to perform in the graphical user interface provided by the rehabilitation application, and determine the exercise to start the rehabilitation training. At this time, the rehabilitation application can determine that the rehabilitation training evaluation request sent by the practitioner is received, and determine the target rehabilitation training indicated by the rehabilitation training evaluation request.

[0038] The rehabilitation training here may include "standing straight leg lift", "assisted knee flexion in sitting position", "left and right transfer", "single-leg standing balance", etc.

[0039] The types of rehabilitation training here may include two-state rehabilitation training and three-state rehabilitation training. Among them, the two-state rehabilitation training may be a situation where rehabilitation preparatory movements are not required, and the three-state rehabilitation training may be a situation where rehabilitation preparatory movements are required.

[0040] Among them, "standing straight leg lift" and "single-leg standing balance" are two-state rehabilitation training, while "sitting assisted knee flexion" and "left-right transfer" are three-state rehabilitation training.

[0041] It is understandable that in the rehabilitation application, the type of each rehabilitation training may be pre-set.

[0042] S20: Determine a rehabilitation action execution phase detection rule to be performed based on the type of target rehabilitation training.

[0043] In step S20, after the type of target rehabilitation training to be performed by the practitioner is determined, the corresponding rehabilitation action execution stage detection rule is determined. The rehabilitation action execution stage detection rule is used to judge the rehabilitation action execution stage of the practitioner's body.

[0044] S30 , determining the rehabilitation action execution stage of the practitioner according to the determined rehabilitation action execution stage detection rule for the acquired rehabilitation action frame image of the practitioner.

[0045] Here, the practitioner can use the camera of the mobile phone to record his or her own practice process. For the video stream captured by the camera, the rehabilitation application can capture the rehabilitation action frame image at a certain sampling frequency.

[0046] Each frame of rehabilitation action image can be input into a pre-trained machine learning model (such as a convolutional neural network) to obtain the real-time coordinates of the key points of the human body output by the machine learning model. The key points of the human body here can include the shoulders, elbows, wrists, hips, knees, ankles and other positions.

[0047] Based on the real-time coordinates of key points of the human body, the execution stage of the rehabilitation action performed by the practitioner can be further determined.

[0048] When the type of rehabilitation training is two-state rehabilitation training, the rehabilitation action execution phase includes two phase states: an evaluation phase and a non-evaluation phase.

[0049] When the type of rehabilitation training is three-state rehabilitation training, the rehabilitation action execution phase includes three phase states: non-assessment phase, assessment preparation phase and assessment phase.

[0050] S40: When the rehabilitation action execution stage of the practitioner meets the evaluation conditions, the real-time action evaluation value corresponding to the rehabilitation action frame image is determined and fed back to the practitioner.

[0051] In step S40, whether the exerciser's rehabilitation action execution stage meets the evaluation conditions can be determined in the following manner:

[0052] When the rehabilitation action execution stage of the practitioner is the evaluation stage, it is determined that the rehabilitation action execution stage of the practitioner meets the evaluation condition.

[0053] That is, for two-state rehabilitation training or three-state rehabilitation training, when the rehabilitation movements of the practitioner reach the evaluation stage, the real-time movement evaluation value of the practitioner's rehabilitation movements begins to be calculated. After the evaluation is completed, it can be displayed on the screen or fed back to the practitioner through voice broadcasting and other methods.

[0054] The real-time action evaluation value here corresponds one-to-one to the rehabilitation action frame image, and is used to indicate the degree of completion of the practitioner's real-time action relative to the rehabilitation training.

[0055] Furthermore, the number of action trainings can be selected to determine the timing of exiting the rehabilitation training detection and evaluation service, for example, to determine whether the practitioner has completed a single exercise or multiple exercises of the target rehabilitation training, or to determine whether the number of exercises corresponding to the target rehabilitation training currently performed by the practitioner has ended.

[0056] The embodiment of the present application provides a method for evaluating rehabilitation training. By classifying rehabilitation training and combining it with a state machine architecture, different action state detection rules are designed for different types of rehabilitation training. This can more accurately capture the state of the rehabilitation movements performed by the practitioner and provide more accurate evaluation results.

[0057] Embodiment 2

[0058] like Figure 2 As shown, in one embodiment of the present application, when the type of rehabilitation training is two-state rehabilitation training, taking the "standing straight leg raising" training as an example, for each acquired rehabilitation action frame image, the rehabilitation action execution stage of the practitioner can be determined in the following manner:

[0059] S3001. Extracting real-time coordinates of key points of the human body from the rehabilitation action frame image to calculate a first calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed.

[0060] Here, taking the "standing straight leg raise" training as an example, the rehabilitation action that the practitioner needs to perform may actually be raising one leg to 60°, but in the rehabilitation action frame image, the practitioner's rehabilitation action is only raising one leg to 30°.

[0061] In step S3001, the image can be converted into the format required by the machine learning model, such as resizing, normalization, etc.

[0062] Key points of the human body here may include the hips, knees and ankles.

[0063] It is understandable that for each rehabilitation training and each different action state of the rehabilitation training, the standard value coordinates of each key point of the human body can be preset.

[0064] The calculated similarity here may be the similarity between the key point coordinates, for example, it may be calculated by Euclidean distance, cosine similarity, etc.

[0065] S3003: Determine whether the first calculated similarity is less than a first preset similarity.

[0066] In a specific embodiment, in addition to similarity calculation, other additional conditions may be combined for judgment, and the design may be performed through specific rehabilitation training.

[0067] S3005: If yes, determine that the rehabilitation action execution phase of the practitioner is a non-evaluation phase.

[0068] S3007: If not, determine that the rehabilitation action execution phase of the practitioner is the evaluation phase.

[0069] When performing the "standing straight leg lift" movement, as shown in Figure 4 (a), the non-assessment stage can be when the practitioner stands straight and lifts one leg forward to a certain height. As shown in Figure 4 (b), other situations can be regarded as the practitioner's rehabilitation movement as the assessment stage.

[0070] When the practitioner stands up straight and lifts one leg forward to a certain height, the calculation of the real-time action evaluation value can begin.

[0071] Some rehabilitation movements consist of two movements that need to be performed continuously. After completing the previous movement, there is usually a recovery movement, and then the next movement is performed. At this time, it is easy to distinguish one rehabilitation training as two, thus causing unreasonable score evaluation and motion capture errors. When distinguishing through the above embodiment, when the practitioner performs the rehabilitation movement, the continuity of the rehabilitation training is used to statistically evaluate the scores of all movements in the state. In this way, for some complex rehabilitation training, we can more accurately determine the time period for statistical movement scores and give scientific and reasonable scores.

[0072] Embodiment 3

[0073] like Figure 3 As shown, in one embodiment of the present application, when the type of rehabilitation training is three-state rehabilitation training, the rehabilitation action execution stage includes a non-evaluation stage, an evaluation preparation stage, and an evaluation stage. Taking the "assisted knee flexion in sitting position" training as an example, for each acquired rehabilitation action frame image, the rehabilitation action execution stage of the practitioner can be determined in the following manner:

[0074] S3000: extracting real-time coordinates of key points of the human body from the rehabilitation action frame image to calculate a second calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the preparatory action of the rehabilitation action to be performed.

[0075] The similarity calculation here can refer to step S3001, and the similarity is calculated based on the coordinates. Alternatively, a body part vector is formed based on the coordinates of the key points of the human body. Then, the angle feature and the distance feature are calculated based on the body part vector, and the similarity is calculated through the angle feature and the distance feature.

[0076] For example, two adjacent human key points A and B can be randomly selected or selected from the extracted human key points to form a vector :

[0077] ;

[0078] in, and are the real-time coordinates of A and B respectively.

[0079] The angle between two vectors It can be calculated by the following formula:

[0080] ;

[0081] in,( , ) is the first vector, ( , ) is the second vector. The first vector and the second vector here are not limited, or can be selected according to the corresponding rehabilitation action.

[0082] Distance feature between vectors It can be calculated by the following formula:

[0083] ;

[0084] in,( , ) is the corresponding vector.

[0085] It can be understood that when calculating the similarity, the similarity between the distance feature and the angle feature and the specified feature can be calculated.

[0086] S3002: Determine whether the second calculated similarity is greater than a second preset similarity.

[0087] S3004: If the similarity is less than a second preset similarity, determining that the rehabilitation action execution stage of the practitioner is a non-evaluation stage.

[0088] Taking the “assisted knee flexion in sitting position” training as an example, the movement can be shown in Figure 5 (a).

[0089] S3006: If the similarity is greater than the second preset similarity, determine that the rehabilitation action execution stage of the practitioner is the evaluation preparation stage, and calculate a third calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed.

[0090] As shown in Figure 5(b), the practitioner enters the rehabilitation preparation stage (sitting and lifting one leg to a certain height). It is then determined whether the practitioner remains in the rehabilitation preparation stage. If not, the similarity is calculated with the rehabilitation action to be performed (sitting and bending one leg at the knee, as shown in Figure 5(c)).

[0091] If the third calculated similarity is less than the third preset similarity, the process returns to the step of calculating the second calculated similarity.

[0092] S3008: If the third calculated similarity is greater than the third preset similarity, the rehabilitation action execution stage of the practitioner is determined to be the evaluation stage. If it is determined not to exit the rehabilitation training evaluation, the process returns to the step of calculating the second calculated similarity.

[0093] It is understandable that some rehabilitation movements require a preparatory movement that is quite different from the daily human movement. In the process from the daily human movement to the execution of the preparatory movement, it is easy to cause the movement to be misjudged as the rehabilitation training movement to be performed. Therefore, an evaluation preparation stage is introduced. The practitioner's movements are not evaluated in the evaluation preparation stage, so it will not cause any impact.

[0094] Embodiment 4

[0095] In one embodiment of the present application, for all rehabilitation action frame images in each evaluation stage, the corresponding real-time action evaluation value can be calculated in the following manner:

[0096] First determine the evaluation value of a body part in a single frame image :

[0097]

[0098] in, is the order of the body parts ( 1, 2, ..., n), For body parts The angle eigenvalue or distance eigenvalue calculated in real time between the corresponding vectors, For body parts The corresponding target eigenvalue, For body parts The corresponding limiting characteristic value, the target characteristic value here is determined based on relatively standard rehabilitation movements, and the limiting characteristic value is determined based on the situation that the human body can achieve with the farthest data distance from the target characteristic value.

[0099] If both angle and distance features are involved in the score evaluation of a body part, the action evaluation value of the target part in a single frame image is calculated as follows:

[0100] =

[0101] in, is the evaluation value calculated based on the angle feature value of the target part, is the evaluation value calculated based on the distance feature of the target part. is the corresponding normalization coefficient.

[0102] Next, the real-time action evaluation value of a single frame image can be calculated:

[0103] ;

[0104] in, For body parts The corresponding weights, for The real-time action evaluation value of the moment-by-moment rehabilitation action frame image. The distribution of weights may be normal distribution, average distribution, Poisson distribution, exponential distribution, Cauchy distribution, gamma distribution, chi-square distribution, t distribution, etc. It may also be customized, depending on the specific situation.

[0105] Finally, the total action evaluation value S in the evaluation phase can be calculated:

[0106]

[0107] in, for Real-time motion estimation value of moment-to-moment recovery motion frame images.

[0108] Considering that in some cases, the final comprehensive evaluation score of the evaluation stage cannot equally consider the image frames corresponding to each moment, the following formula can be used to calculate the total action evaluation value S in the evaluation stage:

[0109]

[0110] in, , The weights corresponding to the rehabilitation action frame images at different moments. The weight distribution may be normal distribution, Poisson distribution, exponential distribution, Cauchy distribution, gamma distribution, chi-square distribution, t distribution, etc. It may also be customized, depending on the specific situation.

[0111] Furthermore, rehabilitation training guidance information may be determined and displayed according to the total action evaluation value and / or real-time action evaluation value of a certain rehabilitation training action of the practitioner.

[0112] In this way, the movements can be scored and counted more scientifically and accurately, and reasonable corrections or movement guidance can be given.

[0113] Embodiment 5

[0114] like Figure 6 As shown, based on the same inventive concept, an embodiment of the present application further provides a rehabilitation training evaluation device 60, the device comprising:

[0115] A response module 610 is used to respond to the rehabilitation training evaluation request and determine the target rehabilitation training indicated by the rehabilitation training evaluation request;

[0116] A rule analysis module 620 is used to determine the rehabilitation action execution phase detection rules to be performed based on the type of target rehabilitation training;

[0117] The state analysis module 630 is used to determine the rehabilitation action execution stage of the practitioner according to the determined rehabilitation action execution stage detection rule based on the acquired rehabilitation action frame image of the practitioner;

[0118] The evaluation module 640 is used to determine the real-time action evaluation value corresponding to the rehabilitation action image and feed it back to the practitioner when the rehabilitation action execution stage of the practitioner meets the evaluation conditions.

[0119] In a preferred embodiment, when the type of rehabilitation training is two-state rehabilitation training, the rehabilitation action execution stage includes a non-evaluation stage and an evaluation stage. For each acquired rehabilitation action frame image, the state analysis module 630 determines the rehabilitation action execution stage of the practitioner in the following manner: for the rehabilitation action frame image, the real-time coordinates of the key points of the human body are extracted to calculate the first calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed; determine whether the first calculated similarity is less than the first preset similarity; if so, determine that the rehabilitation action execution stage of the practitioner is the non-evaluation stage; if not, determine that the rehabilitation action execution stage of the practitioner is the evaluation stage.

[0120] In a preferred embodiment, when the type of rehabilitation training is three-state rehabilitation training, the rehabilitation action execution stage includes a non-evaluation stage, an evaluation preparatory stage and an evaluation stage. For each acquired rehabilitation action frame image, the state analysis module 630 determines the rehabilitation action execution stage of the practitioner in the following manner: for the rehabilitation action frame image, the real-time coordinates of the key points of the human body are extracted to calculate the second calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the preparatory action of the rehabilitation action to be performed; determine whether the second calculated similarity is greater than the second preset similarity; if it is less than the second preset similarity, determine that the rehabilitation action execution stage of the practitioner is the non-evaluation stage; if it is greater than the second preset similarity, determine that the rehabilitation action execution stage of the practitioner is the evaluation preparatory stage, and calculate the third calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed. If the third calculated similarity is less than the third preset similarity, return to the step of calculating the second calculated similarity. If the third calculated similarity is greater than the third preset similarity, determine that the rehabilitation action execution stage of the practitioner is the evaluation stage.

[0121] In a preferred embodiment, the state analysis module 630 is further configured to return to the step of calculating the second similarity if it is determined to exit the evaluation phase.

[0122] In a preferred embodiment, the evaluation module 640 determines whether the rehabilitation action execution stage of the practitioner meets the evaluation condition in the following manner: when the rehabilitation action execution stage of the practitioner is the evaluation stage, it is determined that the rehabilitation action execution stage of the practitioner meets the evaluation condition.

[0123] In a preferred embodiment, the evaluation module 640 is further used to determine the final score of a certain rehabilitation action of the practitioner based on the real-time action evaluation values ​​corresponding to all the rehabilitation action frame images in the evaluation stage.

[0124] In a preferred embodiment, the evaluation module 640 is also used to determine and display rehabilitation training guidance information according to the final score of a certain rehabilitation training action of the practitioner.

[0125] See also Figure 7 , Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 7 As shown in , the electronic device 700 includes a processor 710 , a memory 720 and a bus 730 .

[0126] The memory 720 stores machine-readable instructions executable by the processor 710. When the electronic device 700 is running, the processor 710 communicates with the memory 720 through the bus 730. When the machine-readable instructions are executed by the processor 710, the steps of a rehabilitation training evaluation method in the above-mentioned method embodiment can be executed. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0127] An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a rehabilitation training evaluation method in the above method embodiment can be executed. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0128] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0129] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0130] In addition, the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0131] Furthermore, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0132] It should be noted that if the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0133] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0134] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for evaluating rehabilitation training, characterized in that: The method comprises: In response to the rehabilitation training evaluation request, determine the target rehabilitation training indicated by the rehabilitation training evaluation request; Based on the type of the target rehabilitation training, determining a rehabilitation action execution phase detection rule to be performed; For the acquired rehabilitation action frame image of the practitioner, the rehabilitation action execution stage of the practitioner is determined according to the determined rehabilitation action execution stage detection rule; When the rehabilitation action execution stage of the practitioner is the evaluation stage, it is determined that the rehabilitation action execution stage of the practitioner meets the evaluation condition, and the real-time action evaluation value corresponding to the rehabilitation action frame image is determined and fed back to the practitioner; Determine the final score of the rehabilitation training of the rehabilitation action to be performed by the practitioner based on the real-time action evaluation values ​​corresponding to all the rehabilitation action frame images in the evaluation stage; Among them, when the type of rehabilitation training is two-state rehabilitation training, the rehabilitation action execution stage includes an evaluation stage and a non-evaluation stage; When the type of rehabilitation training is three-state rehabilitation training, the rehabilitation action execution phase includes a non-assessment phase, an assessment preparation phase, and an assessment phase.

2. The method according to claim 1, characterized in that When the type of rehabilitation training is two-state rehabilitation training, for each acquired rehabilitation action frame image, the rehabilitation action execution stage of the practitioner is determined by the following method: Extracting the real-time coordinates of the key points of the human body from the rehabilitation action frame image to calculate a first calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed; Determining whether the first calculated similarity is less than a first preset similarity; If so, it is determined that the practitioner's rehabilitation action execution phase is a non-assessment phase; If not, it is determined that the practitioner's rehabilitation action execution phase is the evaluation phase.

3. The method according to claim 1, characterized in that When the type of rehabilitation training is three-state rehabilitation training, for each acquired rehabilitation action frame image, the rehabilitation action execution stage of the practitioner is determined by the following method: Extracting the real-time coordinates of the key points of the human body from the rehabilitation action frame image to calculate the second calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the preparatory action of the rehabilitation action to be performed; determining whether the second calculated similarity is greater than a second preset similarity; If it is less than the second preset similarity, then determining that the rehabilitation action execution stage of the practitioner is a non-evaluation stage; If it is greater than the second preset similarity, the rehabilitation action execution stage of the practitioner is determined to be the evaluation preparation stage, and a third calculated similarity between the rehabilitation action of the practitioner in the rehabilitation action frame image and the rehabilitation action to be performed is calculated. If the third calculated similarity is less than the third preset similarity, returning to the step of calculating the second calculated similarity, If the third calculated similarity is greater than the third preset similarity, it is determined that the rehabilitation action execution stage of the practitioner is the evaluation stage.

4. The method according to claim 3, characterized in that If the evaluation phase is exited, the process returns to the step of calculating the second similarity.

5. The method according to claim 1, characterized in that Also includes: According to the final score of the rehabilitation training of the rehabilitation action to be performed by the practitioner, the rehabilitation training guidance information is determined and displayed.

6. A rehabilitation training evaluation device, characterized in that: The device comprises: A response module, used to respond to the rehabilitation training evaluation request and determine the target rehabilitation training indicated by the rehabilitation training evaluation request; A rule analysis module, used to determine the rehabilitation action execution phase detection rules to be performed based on the type of the target rehabilitation training; A state analysis module, for determining the rehabilitation action execution stage of the practitioner according to the determined rehabilitation action execution stage detection rule based on the acquired rehabilitation action frame image of the practitioner; An evaluation module, for determining, when the rehabilitation action execution stage of the practitioner is the evaluation stage, that the rehabilitation action execution stage of the practitioner meets the evaluation conditions, and then determining the real-time action evaluation value corresponding to the rehabilitation action frame image and feeding it back to the practitioner; Among them, when the type of rehabilitation training is two-state rehabilitation training, the rehabilitation action execution stage includes an evaluation stage and a non-evaluation stage; When the type of rehabilitation training is three-state rehabilitation training, the rehabilitation action execution phase includes a non-assessment phase, an assessment preparation phase, and an assessment phase.

7. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the rehabilitation training evaluation method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the rehabilitation training evaluation method according to any one of claims 1 to 5 are executed.

Images