Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is only exemplary and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. In addition, in the following description, descriptions of well-known structures and techniques are omitted so as not to unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and/or the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It should be noted that the terms used herein should be construed to have meanings consistent with the context of the present specification and should not be construed in an idealized or overly formal manner.
Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a convention should be interpreted in accordance with the meaning of one of skill in the art having generally understood the convention (e.g., "a system having at least one of A, B and C" would include, but not be limited to, systems having a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).
In the process of implementing the inventive concept of the present disclosure, the inventor found that although the above index can reflect the physiological characteristics of the above personnel, it is difficult to accurately determine whether the actions of the above personnel conform to the standard action specification, and it is difficult to accurately determine whether the personnel have abnormal actions or compensatory conditions during the actions. Therefore, the method is difficult to provide targeted training guidance for the personnel to improve the training quality and reduce the damage risk. The standard action specification refers to correct action collar and skill to be followed by the person when the person performs the training action, and determines the training effect and safety of the person. Abnormal action or compensation conditions refer to muscle activation or exercise patterns which are not in accordance with the standard action specification and appear due to the influence of muscle functions or fatigue degrees when the personnel perform training actions, and can reduce the training effect and safety of the personnel and increase the risk of injury and accidents.
Taking firefighters as an example, muscle function and fatigue can directly affect firefighter athletic performance and safety. Muscle function refers to the characteristics of strength, speed, endurance, coordination, etc. of muscles during contraction and relaxation, which determines the athletic ability and efficiency of firefighters. Muscle fatigue refers to a phenomenon that muscle functions are reduced due to changes in energy metabolism, nerve conduction, muscle fiber structure, etc. after long-time or high-intensity exercise of muscles, which can reduce exercise performance and safety of firefighters and increase risks of injury and accidents.
Therefore, the method for monitoring and evaluating the muscle function and fatigue degree of the firefighter in the training process, as well as the standard action specification and abnormal action or compensation situation, has important significance for optimizing the training plan of the firefighter, improving the training effect and physical condition of the firefighter, and preventing and reducing risks and damages in training.
Based on this, the inventor found that, because there are some problems and disadvantages in the training monitoring and evaluation of firefighters, there is a need for a system that can comprehensively, accurately and timely reflect the muscle functions and fatigue of firefighters, and abnormal actions or compensation conditions, to guide and optimize the training plan of firefighters, improve the training effect of firefighters, and prevent and reduce risks and injuries in training. It should be understood that firemen are only examples in this disclosure and are not intended to limit the disclosure in any way.
In view of this, the present disclosure provides a motion estimation system based on multidimensional data analysis, comprising: the data acquisition device is used for acquiring I pieces of multidimensional data of the target object under the condition that the target object moves, wherein each piece of multidimensional data comprises an electromyographic signal and gesture data, the I pieces of multidimensional data correspond to the I moments, and I is an integer larger than 1. And the processor is connected with the data acquisition device and is used for: dividing the I multi-dimensional data based on the I gesture data to obtain N multi-dimensional data groups, wherein each multi-dimensional data group corresponds to different motion phases, and N is a positive integer smaller than or equal to I. And querying N groups of target template information corresponding to the N multidimensional data groups from the action template database. And analyzing the N multi-dimensional data sets based on the N groups of target template information to obtain a motion evaluation result corresponding to the target object.
Fig. 1 schematically illustrates an architecture diagram of a motion estimation system based on multidimensional data analysis according to a first embodiment of the present disclosure.
As shown in fig. 1, the motion estimation system based on multi-dimensional data analysis of this embodiment includes a data collector 110 and a processor 120.
According to an embodiment of the present disclosure, the data collector 110 may be worn on the body of the target object for collecting I pieces of multi-dimensional data of the target object in case the target object moves. Each piece of multidimensional data comprises an electromyographic signal and gesture data, I pieces of multidimensional data correspond to I times, and I is an integer greater than 1.
The target object may be an object to be detected. The movement may be a detection movement for detecting a condition of the object. The detected dimensions include, but are not limited to, at least one of physical state and athletic ability. The physical state includes, but is not limited to, at least one of a muscle state and a joint state. However, embodiments of the present disclosure are not limited thereto, and in other embodiments of the present disclosure, the above-described motion may also be a training motion for training the athletic ability of the subject.
The data collector 110 includes a surface myoelectricity sensing unit and an inertial measurement unit. The surface electromyographic sensor is used for acquiring electromyographic signals of the target object by using an electromyographic acquisition electrode worn by the target object. The inertial measurement unit is used for acquiring attitude data of the target object. Myoelectricity-harvesting electrodes may be located in the subject's lateral femoral muscle (VL), medial femoral muscle (VM), rectus Femoris (RF), biceps femoris (BL), semitendinosus (SE), anterior tibial muscle (TIA), medial gastrocnemius (gas.m), and lateral gastrocnemius (gas.l). An inertial measurement unit is located at the thigh, calf and foot of the target object. In embodiments of the present disclosure, the myoelectricity harvesting electrode may be affixed to the dominant lateral limb of the target subject. Similarly, the inertial measurement unit may also be tethered to the dominant lateral limb of the target object.
The electromyographic signal may be a surface electromyographic signal (surface electromyography, sEMG). Surface myoelectricity is a non-invasive, painless, and non-destructive technique that records the electrical signal generated when the muscle contracts by attaching electrodes to the surface of the skin. The electrical signal is used to reflect the electrophysiological activity of the muscle. The technique has the advantages that the muscle activity state can be measured continuously and dynamically in real time, the technique is not influenced by environment and movement modes, and the normal movement of firefighters is not disturbed. It should be noted that the present disclosure is presented by way of example only with firefighters, and it is to be understood that the athletic evaluation system of the present disclosure may also be used in other similar professions, including but not limited to athletes and the like. The sEMG signal has close relation with functional parameters such as muscle strength, fatigue and the like, and the extracted characteristic parameters can be used as a means for effectively monitoring and evaluating the muscle functions and fatigue degrees by processing the sEMG signal.
The gesture data may include, but is not limited to, at least one of a displacement, a velocity, an angle, and an acceleration of the target object, which may describe a kinematic parameter of the motion feature. The kinematic analysis has the advantage of intuitively, objectively and quantitatively reflecting the gesture of the target object.
According to an embodiment of the present disclosure, the processor 120 is connected with the data collector 110. The processor 120 may be connected to the data collector 110 by communication, electrical, or the like. Thus, the processor 120 can receive and process the electromyographic signals and gesture data from the data collector 110.
For example, the processor 120 may preprocess the electromyographic signals and gesture data received from the data collector 110, resulting in preprocessed electromyographic signals and preprocessed gesture data.
Specifically, the surface myoelectricity sensing unit is used for: the initial myoelectric signal of the target object is acquired by using the myoelectric acquisition electrode worn by the target object. The processor 120 is configured to: and performing direct current filtering on the initial electromyographic signal to obtain a first filtered electromyographic signal. And carrying out power frequency notch filtering on the first filter electromyographic signal to obtain a second filter electromyographic signal. And 4-order 20-500 Hz Butterworth band-pass filtering is carried out on the second filtered electromyographic signals to obtain third filtered electromyographic signals, wherein the power frequency is 50Hz. And full-wave rectifying the third filtered electromyographic signal to obtain a rectified electromyographic signal. The rectified electromyographic signal is subjected to 6 Hz butterworth low pass filtering to obtain an electromyographic signal envelope. And normalizing the electromyographic signal envelope to obtain the electromyographic signal. For example, the electromyographic signal envelope may be normalized to a maximum value of the electromyographic signal envelope, resulting in a preprocessed electromyographic signal. Through the preprocessing process, noise in the initial electromyographic signal can be removed in a targeted manner, so that an accurate electromyographic signal is obtained, and accurate electromyographic signal characteristics can be extracted.
The inertia measurement unit is used for: initial pose data of the target object is acquired by using an inertial measurement unit worn by the target object. The processor 120 is configured to: and 4-order 5Hz zero-lag Butterworth low-pass filtering is carried out on the initial gesture data, so that filtered gesture data are obtained. And eliminating zero drift of the filtered attitude data to obtain preprocessed attitude data. Through the preprocessing process, high-frequency noise in the initial gesture data can be removed in a targeted manner, accurate gesture data can be obtained, and accurate gesture features can be extracted.
For example, the processor 120 is further configured to: dividing the I multi-dimensional data based on the I gesture data to obtain N multi-dimensional data groups, wherein each multi-dimensional data group corresponds to different motion phases, and N is a positive integer smaller than or equal to I. And querying N groups of target template information corresponding to the N multidimensional data groups from the action template database. And analyzing the N multi-dimensional data sets based on the N groups of target template information to obtain a motion evaluation result corresponding to the target object.
It is understood that the motion performed by the target object may include N motion phases. The I pose data may be used to identify the motion phase to which each multi-dimensional data belongs. For example, the extreme values in the I pose data may be used to identify the motion phase to which the multidimensional data pertains. Based on this, I multi-dimensional data can be divided by I pose data, resulting in a multi-dimensional data set corresponding to each motion phase.
For example, the processor 120 is configured to: and determining J action change moments of the target object from the I moments by matching the pre-stored action change posture data with the I posture data, wherein the action change posture data represents the limb posture of the object when the object changes actions, and J is an integer greater than or equal to 1 and less than I. And dividing the I multi-dimensional data based on J action change moments to obtain N multi-dimensional data groups. In this case, by matching the motion change posture data with the I posture data, the target posture data that matches the motion change posture data can be determined from the I posture data, and the time corresponding to the target posture data can be determined as the motion change time. On this basis, by dividing the I multi-dimensional data according to the motion change timing, the multi-dimensional data group corresponding to each motion phase can be accurately obtained.
In one embodiment of the present disclosure, the pre-stored motion-altering gesture data is determined manually from historical gesture data of the target object. In another embodiment of the present disclosure, the motion-altering gesture data may also be determined from historical gesture data acquired manually from other objects than the target object while performing the same motion as the target object.
In the embodiment of the present disclosure, the multi-dimensional data of the object that has been collected may be matched with the historical multi-dimensional data of each of M candidate movements stored in the database, to determine, from the M candidate movements, a target movement that matches the movement that is being performed by the object, where M is a positive integer greater than 1, and the database may be the above-mentioned action template database, or may be another database different from the above-mentioned action template database, and the present disclosure is not limited thereto. On this basis, the motion change posture data stored in association with the target motion may be determined as the motion change posture data for matching with the I posture data. Based on the above, by matching the multi-dimensional data of the target object with the historical multi-dimensional data, the motion of the target object can be accurately determined, so that the accurate motion change posture data can be determined, the motion change time can be accurately determined, and the multi-dimensional data set can be accurately divided according to the motion change time.
Specifically, the candidate motion information matching the N-th multi-dimensional data group may be determined from the candidate motion information matching the N-1-th multi-dimensional data group of the N multi-dimensional data groups based on the N-th multi-dimensional data group of the N multi-dimensional data groups, N being an integer of N or less and greater than 1. And determining candidate motion information matched with the Nth multidimensional data set as motion information. Then, the motion change posture data stored in association with the motion information is determined as the previously stored motion change posture data. Thus, by performing motion matching on the multi-dimensional data sets one by one, the motion change posture data can be accurately determined.
However, the embodiment of the present disclosure is not limited thereto, and in the embodiment of the present disclosure, a motion identifier input by a target object may also be acquired, and a motion performed by the target object may be determined based on the motion identifier, so as to query the database for motion change gesture data.
In this case, the motion change time of the I times can be identified by matching the motion change posture data that matches the motion being performed by the target object with the posture data. For example, the posture data includes an acceleration change value and a joint angle value. The action change gesture data comprises a pre-stored acceleration change value and a pre-stored joint angle value. The processor 120 is further configured to: and identifying E target acceleration change values matched with the pre-stored acceleration change values from the I acceleration change values, wherein E is a positive integer smaller than I. From the I joint angle values, Q target joint angle values matched with the pre-stored joint angle values are identified, and Q is a positive integer smaller than I. J motion change times are determined based on times corresponding to the E target acceleration change values and times corresponding to the Q target joint angle values. Thus, the I multi-dimensional data can be divided according to the J operation change times, and N multi-dimensional data groups can be obtained. Accordingly, by matching the acceleration change value and the joint angle value in the posture data, the target joint angle value and the target acceleration change value are determined, and the time corresponding to at least one of the target joint angle value and the target acceleration change value is determined as the operation change time, so that the accuracy of determining the operation change time is improved.
The partitioning of the multi-dimensional data of the present disclosure is specifically described below with reference to fig. 2, taking vertical jump landing motion as an example. It should be understood that this is by way of example only and is not intended to limit the movement of the target object in embodiments of the present disclosure.
Vertical jump landing is a lower limb strength training paradigm that involves a target object jumping off a fixed platform, landing on both feet, immediately followed by a best effort vertical jump, and landing four parts again. As shown in fig. 2, in the embodiment of the present disclosure, the vertical jump landing motion is divided into two stages in advance: a braking phase and a propulsion phase. The braking stage is defined as a stage from the contact of feet with the ground to the maximum buckling angle of the knee joint after the target object jumps down from a high place; the propulsion phase is defined as the phase from the time when the target object reaches the maximum flexion angle of the knee joint to the time when the feet leave the ground again.
When the electromyographic signals of the vertical jump landing task are analyzed, the vertical jump landing motion needs to be automatically divided into stages. The method for dividing the motion phase comprises the following steps: ① Providing original data by the IMU on the thigh and shank body section of the target object for analysis to obtain knee joint angle data, so that the moment corresponding to the maximum knee joint buckling angle can be determined according to the knee joint angle data; ② In the first jump of the target object, the downward speed is suddenly changed to 0 at the moment of contacting the ground, the acceleration also can be greatly changed, the characteristic of the change can be searched as a mark point of a dividing stage, the mark point corresponds to the action change moment, and other mark points in the embodiment of the disclosure are the same; ③ At the moment of the second jump of the target object when the feet are separated from the ground, the speed of the feet is rapidly increased from 0, the acceleration is also changed drastically, and the characteristic of the change can be searched as a mark point of the dividing stage. The motion stage can be divided by searching characteristic mark points by utilizing acceleration data measured by IMUs arranged on feet, thighs and shank body sections and comparing the acceleration data with kinematic data obtained by a motion capture system and ground reaction force data provided by a three-dimensional force measuring platform after the acceleration data is subjected to smoothing treatment by a 4-order 5Hz zero-lag Butterworth low-pass filter. The kinematic data obtained by the motion capture system may include marker (identification) point spatial position data. The above ground reaction force data may be used as the core reference data for comparison.
Fig. 3 schematically illustrates a flowchart of a landmark determining method according to an embodiment of the present disclosure.
As shown in FIG. 3, the method for determining the marker point of the embodiment includes operations S301 to S307
In operation S301, first acceleration data corresponding to the foot of the target object is acquired by using the IMU at the moment the foot is in contact with the ground after the target object jumps down from the box.
In operation S302, the first acceleration data and the kinematic data of the motion capture system are analyzed to obtain a first pre-stored acceleration variation value.
In operation S303, for the time of maximum knee flexion of the target object, joint angle data of the target object at the time is obtained by analyzing the raw data acquired by the IMU at the time, and the joint angle data is used to characterize the angle between the thigh and the calf of the target object.
In operation S304, the joint angle data and the kinematic data of the motion capture system are analyzed to obtain a pre-stored joint angle value.
In operation S305, at the moment when the feet are in contact with the ground at the moment when the target object is lifted up again, second acceleration data corresponding to the feet of the target object are acquired by using the IMU.
In operation S306, the second acceleration data and the kinematic data of the motion capture system are analyzed to obtain a second pre-stored acceleration variation value.
In operation S307, pre-stored motion change gesture data corresponding to the vertical jump landing is obtained based on the first pre-stored acceleration change value, the pre-stored joint angle value, and the second pre-stored acceleration change value.
The first pre-stored acceleration variation value and the second pre-stored acceleration variation value may correspond to a speed abrupt point of the target object. Other pre-stored acceleration change values in the embodiments of the present disclosure are the same, and are not described herein. The pre-stored joint angle values correspond to the maximum angle of the joint buckling of the target object, and other pre-stored joint angle values in the embodiment of the present disclosure are the same and are not described herein.
It should be noted that the above example only shows the process of dividing the I electromyographic signals. It should be appreciated that in the disclosed embodiments, a process of partitioning the I pose data is also included. The process of dividing the I pose data is similar to the above process, and will not be described in detail herein. Based on this, it is possible to divide into a plurality of data sets corresponding to each movement stage according to the movement stage.
In the embodiment of the disclosure, the target template information corresponding to the movement phases can be queried from the movement template database for the multidimensional data group corresponding to each movement phase.
For example, the processor 120 is configured to: and determining candidate motion information matched with the nth multi-dimensional data group from the candidate motion information matched with the nth multi-dimensional data group in the N multi-dimensional data groups based on the nth multi-dimensional data group in the N multi-dimensional data groups, wherein N is an integer less than or equal to N and greater than 1. Candidate motion information matching the nth multi-dimensional data set is determined as motion information. And inquiring to obtain N groups of target template information from the action template database based on the motion information.
The candidate motion information may be information to be determined as a motion being performed by the target object. By matching the N multi-dimensional data sets and the candidate motion information layer by layer, it is possible to efficiently and accurately determine information of the motion being performed by the target object from the candidate motion information without previously setting what motion the target object is performing, thereby obtaining target template information corresponding to the motion being performed by the target object.
The action template database can be a database under standard actions established by standardizing and normalizing actions through kinematic analysis. For example, the database stores target template information corresponding to standard actions. The target template information can be representative action characteristic parameters obtained by collecting, processing, analyzing and other operations on surface electromyographic signals and kinematic data in the movement process of the object. The target template information may include template electromyographic signals corresponding to the surface electromyographic signals described above, template posture data corresponding to the kinematic data described above, and the like. Based on the motion template database can be used for providing standard data and target data of the motion, so that the motion of the target object can be evaluated and guided, and the motion effect and safety of the target object are improved.
In the embodiment of the disclosure, the initial electromyographic signals acquired in the history period may be subjected to direct current filtering, power frequency notch filtering, 4-order 20-500 Hz butterworth band-pass filtering, full-wave rectification, 6 Hz butterworth low-pass filtering, normalization and the like, so as to obtain template electromyographic signals, and the processing process is similar to the preprocessing process of the electromyographic signals, and will not be repeated herein.
In the embodiment of the disclosure, 4-order 5Hz zero-lag butterworth low-pass filtering processing may be performed on initial gesture data acquired in a history period to obtain filtered gesture data. Thereby, high-frequency noise in the initial attitude data can be removed. And then, by detecting the resting bias of the resting state, calibrating to eliminate zero drift, and then, establishing a lower limb movement model based on the static calibration to obtain knee-ankle joint angle data. Template posture data can thus be obtained. The processing procedure is similar to the preprocessing procedure of the gesture data, and is not described herein.
Taking the vertical jump landing motion as an example, the action template database comprises a set of accurate, flexible and easy-to-use template electromyographic signals capable of being supplemented for the vertical jump landing motion, wherein the template electromyographic signals are unique. The template electromyographic signal has high accuracy and stability due to the accurate acquisition and processing of a plurality of strictly selected test data. And, the exclusive nature of the template electromyographic signal means that all subsequent operations are performed based on the template electromyographic signal. A specially designed algorithm or system can be adopted to realize specific functions or purposes by comparing with a unique template electromyographic signal of vertical jump landing motion.
Based on the foregoing, the template gesture data stored in the action template database may include template gesture features. The template electromyographic signals stored in the action template database may include template electromyographic signal characteristics. Based on the above, in the embodiment of the present disclosure, feature extraction may be performed on the posture data and the electromyographic signals of the target object, so as to obtain the posture feature and the electromyographic signal feature of the target object. For example, the start-stop time point of the motion may be determined according to the kinematic data, and the electromyographic signal in the motion execution period may be intercepted, where the start-stop time point corresponds to the motion change time. Thus, the electromyographic signal characteristics can be extracted for the electromyographic signal within the time period, and the electromyographic signal characteristics are used as template electromyographic signals. The electromyographic signal characteristics may include a root mean square value of the electromyographic signal, an integrated electromyographic value, an average power frequency value, a muscle co-contraction index, a muscle co-mode, and the like.
In the embodiment of the disclosure, in order to evaluate the muscle contraction strategy of the target object, that is, how the target object coordinates and controls the contraction of different muscle groups when performing training actions, the present invention extracts the following characteristic parameters for the myoelectric signal:
The muscular Co-contraction index (Co-contraction Index, CCI) is used to reflect the coordination between antagonistic and active muscles, which manifests itself in the stability and stiffness of the muscles acting oppositely around each joint, and thus in response to factors that maintain joint stability during exercise when influenced by the external environment (e.g., severe impact forces).
（1）
Wherein CCI is the muscle co-contraction index. EMGANT is the myoelectric root mean square value (RMS) of antagonistic muscles; EMGAG is the electromyographic Root Mean Square (RMS) value of the active muscle. Antagonistic muscle is defined as a muscle or group of muscles that have lower activity during the action phase, i.e., a muscle or group of muscles that have less RMS. An active muscle is defined as a muscle or group of muscles that have a higher activity during the action phase, i.e. a muscle or group of muscles with a greater RMS.
Muscle Synergy (MS) refers to how a group of muscles work together to produce effective movement during a particular exercise task. This may involve adjustments in the order, magnitude, and timing of activation of particular muscles.
Vm*n=Wm*s * Hs*n （2）
Wherein Vm*n is the electromyographic signal matrix. Wm*s is a co-coefficient matrix. Hs*n is an activation coefficient matrix. Wherein m is the number of muscles. n is the number of data points of the electromyographic signal. s is the number of synergy.
Electromyogram amplitude-related parameters are used to measure the strength of muscle contraction, reflecting the strength and fatigue of the muscle, such as Root Mean Square (RMS), integral myoelectricity (Integral Electromyography, iEMG), median Frequency (MF), etc.
Illustratively, the processor 120 of embodiments of the present disclosure may include a data processing module and a standard action comparison module. The data processing module transmits the electromyographic signals and the characteristic parameters of the gesture data to the action analysis module, and the action analysis module compares and analyzes the characteristic parameters with corresponding standard templates in the action template database to generate a movement evaluation result, so that whether the action is standard and normative or whether abnormal muscle activation or compensation conditions exist in the action process can be judged based on the movement evaluation result, wherein the standard templates correspond to the target template information. The method for analyzing by the action analysis module comprises at least one of the following steps:
(1) And (3) based on a similarity method, namely calculating the similarity between the characteristic parameters and the target template information, such as Euclidean distance, cosine similarity, correlation coefficient and the like, judging whether the motion is standard or not according to the similarity, or judging whether abnormal muscle activation or compensation exists in the motion process, and giving feedback and evaluation of the motion, wherein if the similarity is higher, the motion is more standard, and otherwise, the motion is abnormal or compensation.
(2) And (3) setting a threshold value between the characteristic parameter and the target template information, such as a maximum value, a minimum value, an average value, a standard deviation and the like, judging whether the action is standard according to whether the characteristic parameter is in a threshold range or whether abnormal muscle activation or compensation conditions exist in the action process, giving feedback and evaluation of the action, and if the characteristic parameter is in the threshold range, indicating the standard action specification, otherwise, indicating the action abnormal or compensation conditions.
(3) The method based on classification, i.e. a machine learning or deep learning method, such as a support vector machine, a decision tree, a neural network, etc., is used to train and learn the target template information, so as to obtain a classifier. Under the condition that the characteristic parameters are input into the classifier, whether the action is standard or not can be judged according to the output of the classifier, or whether abnormal muscle activation or compensation conditions exist in the action process or not is judged, feedback and evaluation of the action are given, if the output of the classifier is positive, standard action specification is indicated, and otherwise, abnormal action or compensation conditions exist in the action.
Based on the above method, embodiments of the present disclosure may implement at least one of the following analysis of the above-described feature parameters:
(1) And (3) comparing the characteristic parameters of the target object with the target template information, finding out the difference and the deviation, evaluating whether the muscle contraction strategy of the target object is reasonable and effective, and giving feedback and advice.
(2) And (3) trend analysis, namely comparing characteristic parameters of the target object in different time periods or different training phases, observing the change and the trend, evaluating whether the muscle contraction strategy of the target object is stable and consistent, and giving feedback and advice.
(3) And (3) performing correlation analysis, namely performing correlation analysis on characteristic parameters of the target object and other indexes such as movement characteristics, performance, effects, risks and the like, exploring causal relations, evaluating the influence of a muscle contraction strategy of the target object on training, and giving feedback and advice.
Based on this, the motion estimation result of the embodiment of the present disclosure may include at least one of the following information:
(1) The merits of the target subject's muscle contraction strategy, i.e., whether the target subject is able to effectively coordinate and control the contraction of different muscle groups while performing the training action, to achieve optimal exercise effects and minimal risk.
(2) The improvement of the target subject's muscle contraction strategy, i.e. how the target subject should adjust and optimize the contraction of different muscle groups when performing the training actions, to improve the quality and safety of the movement.
(3) The evaluation of the muscle contraction strategy of the target object, namely, the target object gives the score and the grade of the muscle contraction strategy of the target object according to different evaluation standards when the target object performs the training action so as to reflect the training level and the training capacity of the target object.
Based on this, by analyzing the multidimensional data in accordance with the target template data, a motion estimation result can be obtained. For example, the collected gesture data may be analyzed according to the template gesture data, to obtain a first motion estimation result corresponding to the gesture data. And analyzing the collected electromyographic signals according to the template electromyographic signals to obtain a second motion evaluation result corresponding to the electromyographic signals. And generating the motion estimation result based on the first motion estimation result and the second motion estimation result. It should be noted that, the motion estimation result of the embodiment of the present disclosure is only used as reference data, and is not directly used for treating the target object or diagnosing the disease of the target object.
In the disclosed embodiment, the processor 120 is further configured to: the N multi-dimensional data sets are divided into N myoelectric signal sets and N gesture data sets. Generating nth first action error information based on an nth one of the N sets of module electromyographic signals and an nth one of the N sets of electromyographic signals. And generating first motion trend information of the target object based on the N pieces of first motion error information. Thus, a first motion estimation result can be generated based on the first motion trend information. The analysis of the gesture data is similar to the analysis method of the electromyographic signals, and will not be described here.
According to an embodiment of the present disclosure, the target object is an object subjected to a target training motion. The pose data includes joint angle values. The target template information comprises template posture data and template electromyographic signals, and the template posture data comprises a template joint angle change rate. The processor 120 is further configured to: and determining an nth joint angle change rate of the target object based on the joint angle value in the nth multi-dimensional data set and the movement period corresponding to the nth multi-dimensional data set for the nth multi-dimensional data set, wherein N is a positive integer less than or equal to N. Generating an nth joint angle change score based on the nth joint angle change rate and the nth template joint angle change rate, wherein the joint angle change score is used for representing the joint buckling condition of the target object in the movement process. An nth muscle activation score for the target subject is generated based on the electromyographic signals in the nth multidimensional dataset and the nth set of template electromyographic signals, wherein the muscle activation score is used to characterize the degree of muscle activation of the target subject. Based on the N joint angle change scores and the N muscle activation scores, a reaction index of the target subject is calculated, wherein the reaction index is used for representing the reaction condition of muscles and joints of the target subject in the movement process. And obtaining a pre-stored reaction index of the target object before the target training exercise, wherein the exercise corresponding to the pre-stored reaction index is the same as the exercise corresponding to the reaction index. And generating a motion estimation result corresponding to the target training motion based on the reaction index and a pre-stored reaction index of the target object, wherein the pre-stored reaction index is obtained under the condition that the target object which does not perform the target training motion moves, and the motion estimation result is used for representing the activation capability of the target training motion to joints and muscles of the target object.
In the embodiment of the disclosure, the response index of the target object may be calculated, and then the pre-stored response index of the target object when the target object is not subjected to the target training exercise may be compared, so as to generate an exercise evaluation result for representing the influence capability of the target training exercise on the target object.
Based on the method, the response index of the target object is determined based on the joint angle change score and the muscle activation score of the target object, and then the response index is compared with a pre-stored response index before the target object performs target training exercise, so that an exercise evaluation result used for representing the activation capability of the target training exercise on the joints and muscles of the target object is obtained, the accuracy of determining the activation capability of the target training exercise on the joints and muscles of the target object is improved, and therefore the training plan of the target object can be accurately adjusted.
According to an embodiment of the present disclosure, the processor 120 is further configured to: and determining an nth first target time and an nth second target time from a motion period corresponding to the nth multi-dimensional data group based on the joint angle value in the nth multi-dimensional data group, wherein the first target time represents the time at which the joint bending angle of the target object is minimum, and the second target time represents the time at which the joint bending angle of the target object is maximum. An nth joint bending duration of the target object is calculated based on the nth first target time and the nth second target time. The nth joint angle variation amount of the target object is calculated based on the joint angle value corresponding to the nth first target time and the joint angle value corresponding to the nth second target time. The nth joint angle change rate of the target object is calculated based on the nth joint angle change amount and the nth joint bending time period. An nth joint angle change score may thus be generated based on the nth joint angle change rate and the nth template joint angle change rate. Based on the above, by determining the accurate joint bending period according to the first target time and the second target time, the joint angle change rate can be accurately determined based on the length of the joint bending period and the joint angle change amount, thereby improving the accuracy of determining the joint angle change score.
In the embodiment of the present disclosure, taking a vertical jump landing motion as an example, the rate of change from the time when the target object initially contacts the ground to the time when the target object reaches the maximum joint angle in the motion process may be calculated, and the Z score normalization may be performed, so as to obtain an nth joint angle change score. The following formula is shown:
（3）
Where ACR represents the joint angle change score. Δθ is the angle change amount. RT is the Reaction Time (RT), the duration of the above-mentioned joint bending, for recording the duration of the joint from the initial pose to the Time when the maximum angle is reached. MuACR is the mean of the rate of change of joint angle in the template pose data. σACR is the standard deviation of the rate of change of joint angle in the template pose data. The larger the value of the joint angle change score, the more rapid the joint flexion of the target subject during this phase of motion, and the lower the risk of injury.
According to an embodiment of the present disclosure, the processor 120 is further configured to: an nth electromyographic root mean square value is calculated based on the electromyographic signals in the nth multidimensional data set. And calculating the myoelectric root mean square value of the nth template based on the myoelectric signals of the nth template. An nth muscle activation score is calculated based on the nth myoelectric root mean square value and the nth template myoelectric root mean square value. Based on the above, by calculating the muscle activation score according to the myoelectric root mean square value, the accuracy of the calculated muscle activation score is improved.
In the embodiment of the present disclosure, taking a vertical jump landing motion as an example, a surface myoelectricity root mean square value from a time point when a target object initially contacts the ground to a time point when a maximum joint angle is reached in a motion process may be calculated, and Z fraction normalization may be performed, so as to obtain a muscle activation score, where the following formula is shown:
（4）
Wherein N is the number of the collected surface electromyographic signal data points. xi is the ith data in the superficial electromyographic signal data sequence, which includes the acquired superficial electromyographic signals of the popliteal muscle (semitendinosus and biceps femoris). MuRMSH is the mean value of the surface myoelectric root mean square values of the popliteal muscles (semitendinous and biceps femoris) in the template myoelectric signals. σRMSH is the standard deviation of the surface myoelectric root mean square values of the popliteal muscles (semitendinous and biceps femoris) in the template myoelectric signals.
By calculating the surface myoelectric root mean square value of the popliteal muscle (semitendinous muscle and biceps femoris), and determining the muscle activation score based on this value, the greater the muscle activation score, the higher the degree of activation of the corresponding muscle during the course of action, indicating that preferential activation of the popliteal muscle helps to reduce the risk of injury in skip-like actions.
Based on this, a reaction index of the target subject, which is dimensionless, can be calculated based on the joint angle change score and the muscle activation score. The following formula is shown:
（5）
Wherein MJCRI is the above reaction index. n is the number of actions of the target object and corresponds to the number of the action phases. ACRj is the joint angle change score described above. RMSjH is the muscle activation score described above.
The reaction index may be a muscle-joint reaction index (Muscle Joint Combined Response Index, MJCRI) which can be used to characterize the state and risk of injury of the knee and ankle joint of a human during vertical jump landing movements, and can also be used to evaluate the ability of the joint and muscle to move during dynamic movements and the activation of the muscle to maintain a stable effect. By means of the muscle-joint response index parameters, the movements and activation of joints and muscles in dynamic actions can be quantified, thus helping to assess the risk of injury. The muscle-joint response index (MJCRI) combines factors such as the angular change of the joint, the response time, and the surface myoelectricity root mean square value. The larger the parameter value is, the better the joint movement ability is, the related muscles are correctly activated, and the risk of movement injury is lower.
Based on this, the above-described reaction index and the pre-stored reaction index may be combined to generate a motion estimation result. The above-described reaction index may be included in the exercise evaluation result.
In the embodiment of the present disclosure, the target objects may be B. The motion estimation results of the B target objects can be obtained. The system further includes a human-machine interface coupled to the processor 120. Thus, in one embodiment of the present disclosure, the processor 120 may control the human-machine interaction interface to present the motion estimation results of the B target objects. In another embodiment of the present disclosure, the processor 120 is further configured to: and determining a risk object from the B target objects based on the motion evaluation results corresponding to the B target objects, wherein the B target objects are different from each other, the risk object represents an object with the highest risk of motion injury in the B target objects, and B is an integer greater than 1. Training advice information is generated based on the motion estimation result corresponding to the risk object and the multi-dimensional data corresponding to the risk object, wherein the training advice information includes training motions for advice of the target object. And controlling the human-computer interaction interface to display training suggestion information.
In embodiments of the present disclosure, a risk subject that is more susceptible to injury during exercise may be determined by comparing the response index in the exercise assessment of the B target subjects. And, training advice information matching at least one of the motion estimation result, the electromyographic signal, and the posture data may be queried from the advice database based on the motion estimation result, the electromyographic signal, and the posture data. However, embodiments of the present disclosure are not limited thereto, and in another embodiment of the present disclosure, training advice information may be generated based on the exercise assessment results and a predetermined joint exercise injury risk assessment model.
Based on the training advice information, the training advice can be pertinently carried out on the risk object by generating the training advice information based on the motion evaluation result and the multidimensional data of the risk object, and the accuracy of generating the training advice information is improved. On the basis, training suggestion information is displayed by controlling the human-computer interaction interface, so that user experience is improved.
Fig. 4 schematically illustrates a flowchart of a training advice information generation method according to an embodiment of the present disclosure.
As shown in FIG. 4, the training advice information generation method of the embodiment includes operations S401 to S404.
In operation S401, myoelectric signals and posture data are collected during a history period, and an action template database is constructed.
In operation S402, multi-dimensional data of a target object is acquired.
In operation S403, the multi-dimensional data of the target object is analyzed based on the target template information in the action template database, to obtain a motion estimation result.
In operation S404, training advice information is generated based on the exercise assessment result and a predetermined joint exercise injury risk assessment model.
According to the embodiment of the disclosure, the motion phases of the target object are divided according to the acquired gesture data, so that a multi-dimensional data set corresponding to each motion phase is obtained. On the basis, the multi-dimensional data sets of each movement stage of the target object are analyzed by using the target template information corresponding to each multi-dimensional data set in the action template database, so that the accuracy of judging whether the action of the target object in the movement process accords with the standard action specification is improved, and the accuracy of determining the muscle activation condition and compensation condition of the target object in the action process is improved.
For a better understanding of the present disclosure, the present disclosure is further described below with reference to fig. 5, taking firefighters as an example.
Fig. 5 schematically illustrates an architecture diagram of a motion estimation system based on multidimensional data analysis according to a second embodiment of the present disclosure.
As shown in fig. 5, the system of this embodiment includes a data collector 510 and a processor 520. The data collector 510 includes a surface myoelectricity sensing unit 511 and an inertial measurement unit 512. Processor 520 may include a data storage module 521, a human-machine interaction interface 522, a standard action comparison module 523, and a data processing module 524. The data storage module 521 may be constructed based on a solid state disk. The human-machine interface 522 may be coupled to a keyboard, speaker, etc. for presenting information and retrieving input information. The standard action comparison template 523 is used to compare and analyze the multidimensional data with the target template information. The data processing module 524 is used for executing all data processing operations except the data processing operations executed by the data storage module 521, the man-machine interaction interface 522, and the standard action comparison module 523.
In the embodiment of the disclosure, 3-5 firefighters (hereinafter referred to as "template subjects") with healthy body and rich experience can be selected as the collectors of the template signals. In the case of training a firefighter, the surface myoelectricity sensing unit 511 and the inertia measuring unit 512 are attached to the corresponding muscle parts of the firefighter, basic information and training parameters of the firefighter are input through the human-computer interaction module, and training is started. For example, after wiping the skin surface with alcohol, the electromyographic signal acquisition electrodes of the surface electromyographic sensor unit 511 are respectively adhered to the lateral femoral muscle, medial femoral muscle, rectus femoris, biceps femoris, semitendinosus, tibialis anterior, medial gastrocnemius, and lateral gastrocnemius of the dominant limb of the template subject. And binding the inertial measurement unit 512 to the thigh, calf and foot body sections of the dominant side limb of the template subject. And enabling the template subject to repeatedly complete the task of the vertical jump landing action for 10 times under the guidance of a professional trainer. And synchronously acquiring electromyographic signals and kinematic signals in the action process. The surface myoelectricity sensing unit 511 and the inertia measuring unit 512 collect data to the data processing module 524, and data preprocessing is performed in the data processing module 524.
The data processing module 524 performs preprocessing and analysis on the data, and calculates joint angles, muscle activation states, and fatigue degrees during execution of the actions. Meanwhile, the data processing module 524 transmits the characteristic parameters of the data to the standard action comparison module 523, the standard action comparison module 523 compares the characteristic parameters of the data with corresponding standard templates in the action template database, judges whether the action is standard and normative or whether abnormal muscle activation or compensation exists in the action process, gives feedback and evaluation of the action, and provides visual feedback and suggestion for firefighters and coaches through the man-machine interaction interface 522 in combination with the existing firefighter athletic injury risk assessment model. After training the firefighter, the data processing module 524 stores the raw data and processed data in the data storage module 521 for subsequent interrogation and analysis by the firefighter and trainer.
Based on the foregoing, the embodiments of the present disclosure provide a motion estimation system based on multidimensional data analysis, which monitors and estimates the muscle function and fatigue of a target object, and abnormal actions or compensation conditions by collecting surface electromyographic signals and kinematic data of the target object. Visual feedback and advice are provided for firefighters and coaches through a human-computer interaction interface, the training plan of the firefighters is guided and optimized, the training effect of the firefighters is improved, and risks and damages in training are prevented and reduced.
According to the embodiment of the disclosure, the muscle function and fatigue degree of the firefighter, and standard action specification and abnormal or compensatory conditions can be monitored and evaluated in real time by using a noninvasive, painless and lossless surface myoelectricity technology, whether the action is standard specification or not is judged, various feedback and advice are provided for firefighters and coaches by using a man-machine interaction module, the training plan of the firefighter is guided and optimized, the training effect and physical condition of the firefighter are improved, risks and injuries in training are prevented and reduced, and the firefighter has the advantages of accuracy, portability, instantaneity, dynamics, practicality, interactive friendliness and the like.
According to the embodiment of the disclosure, the system provided by the disclosure can solve the problem of insufficient training or overload caused by the fact that the focus of the daily scientific training of the firefighter is focused on the subjective experience judgment of a guide coach, the surface electromyographic signals in the training process of the firefighter can be processed in real time and compared with the target template information, whether the action is standard and normative or whether abnormal muscle activation or compensation conditions exist in the action process is judged, feedback and evaluation of the action are given, and the action quality and safety of the firefighter are improved.
According to the embodiment of the disclosure, the system provided by the disclosure acquires the surface electromyographic signals of the firefighter in the process of executing the specific training action, compares the surface electromyographic signals with the template surface electromyographic signals which are more reasonable in kinematic analysis, can obtain detection and reminding of the firefighter that the firefighter uses muscles or compensates for errors in the process of completing the action, gives out the damage risk degree by combining with the existing firefighter athletic injury risk assessment model, reminds the firefighter of consciously improving the action requirement by utilizing the athletic injury assessment result, further completes the target action better, improves the training quality and reduces the damage risk.
Fig. 6 schematically illustrates a schematic diagram of a motion estimation method based on multidimensional data analysis according to an embodiment of the present disclosure.
As shown in fig. 6, the motion estimation method based on multi-dimensional data analysis of this embodiment includes operations S610 to S640.
In operation S610, in the case where the target object moves, I pieces of multidimensional data of the target object are acquired, wherein each piece of multidimensional data includes an electromyographic signal and posture data, the I pieces of multidimensional data correspond to I times, and I is an integer greater than 1.
In operation S620, the I multi-dimensional data are divided based on the I pose data, to obtain N multi-dimensional data sets, where each multi-dimensional data set corresponds to a different motion phase, and N is a positive integer less than or equal to I.
In operation S630, N sets of target template information corresponding to the N multi-dimensional data sets are queried from the active template database.
In operation S640, the N multi-dimensional data sets are analyzed based on the N sets of target template information, to obtain a motion estimation result corresponding to the target object.
It should be noted that, the motion estimation method based on multi-dimensional data analysis in the embodiments of the present disclosure corresponds to the data processing process of the motion estimation system based on multi-dimensional data analysis, which is not described herein.
It will be appreciated that in embodiments of the present disclosure, the processor may be integrated with the data collector or may be integrated into a separate electronic device.
Fig. 7 schematically illustrates a block diagram of a processor-based implemented electronic device, according to an embodiment of the disclosure.
As shown in fig. 7, an electronic device 700 according to an embodiment of the present disclosure includes a processor 701 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 702 or a program loaded from a storage section 708 into a Random Access Memory (RAM) 703. The processor 701 may include, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or an associated chipset and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), or the like. The processor 701 may also include on-board memory for caching purposes. The processor 701 may comprise a single processing unit or a plurality of processing units for performing different actions of the method flows according to embodiments of the disclosure.
In the RAM 703, various programs and data necessary for the operation of the electronic apparatus 700 are stored. The processor 701, the ROM 702, and the RAM 703 are connected to each other through a bus 704. The processor 701 performs various operations of the method flow according to the embodiments of the present disclosure by executing programs in the ROM 702 and/or the RAM 703. Note that the program may be stored in one or more memories other than the ROM 702 and the RAM 703. The processor 701 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in the one or more memories.
According to an embodiment of the present disclosure, the electronic device 700 may further include an input/output (I/O) interface 705, the input/output (I/O) interface 705 also being connected to the bus 704. The electronic device 700 may also include one or more of the following components connected to an input/output (I/O) interface 705: an input section 706 including a keyboard, a mouse, and the like; an output portion 707 including a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, a speaker, and the like; a storage section 708 including a hard disk or the like; and a communication section 709 including a network interface card such as a LAN card, a modem, or the like. The communication section 709 performs communication processing via a network such as the internet. The drive 710 is also connected to an input/output (I/O) interface 705 as needed. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 710 as necessary, so that a computer program read therefrom is mounted into the storage section 708 as necessary.
The present disclosure also provides a computer-readable storage medium that may be embodied in the apparatus/device/system described in the above embodiments; or may exist alone without being assembled into the apparatus/device/system. The computer-readable storage medium carries one or more programs which, when executed, implement methods in accordance with embodiments of the present disclosure.
According to embodiments of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, which may include, for example, but is not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this disclosure, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to embodiments of the present disclosure, the computer-readable storage medium may include ROM 702 and/or RAM 703 and/or one or more memories other than ROM 702 and RAM 703 described above.
Embodiments of the present disclosure also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowcharts. The program code, when executed in a computer system, causes the computer system to perform the methods provided by embodiments of the present disclosure.
The above-described functions defined in the system/apparatus of the embodiments of the present disclosure are performed when the computer program is executed by the processor 701. The systems, apparatus, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the disclosure.
In one embodiment, the computer program may be based on a tangible storage medium such as an optical storage device, a magnetic storage device, or the like. In another embodiment, the computer program may also be transmitted, distributed over a network medium in the form of signals, downloaded and installed via the communication section 709, and/or installed from the removable medium 711. The computer program may include program code that may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.
In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion 709, and/or installed from the removable medium 711. The above-described functions defined in the system of the embodiments of the present disclosure are performed when the computer program is executed by the processor 701. The systems, devices, apparatus, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the disclosure.
According to embodiments of the present disclosure, program code for performing computer programs provided by embodiments of the present disclosure may be written in any combination of one or more programming languages, and in particular, such computer programs may be implemented in high-level procedural and/or object-oriented programming languages, and/or assembly/machine languages. Programming languages include, but are not limited to, such as Java, c++, python, "C" or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., connected via the Internet using an Internet service provider).
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Those skilled in the art will appreciate that the features recited in the various embodiments of the disclosure may be combined and/or combined in various combinations, even if such combinations or combinations are not explicitly recited in the disclosure. In particular, features recited in various embodiments of the present disclosure may be combined and/or combined in various ways without departing from the spirit and teachings of the present disclosure. All such combinations and/or combinations fall within the scope of the present disclosure.
The embodiments of the present disclosure are described above. These examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described above separately, this does not mean that the measures in the embodiments cannot be used advantageously in combination. Various alternatives and modifications can be made by those skilled in the art without departing from the scope of the disclosure, and such alternatives and modifications are intended to fall within the scope of the disclosure.