Training effect evaluation method and system based on limb freedom degree decomposition

Abstract

The invention provides a training effect evaluation method based on limb freedom degree decomposition. The training effect evaluation method comprises the following steps: carrying out region division on limbs; defining a mapping relation of attitude angles; collecting posture data of joint anatomical mark points in the limb area; establishing a coordinate system on the limb region; calculating a corresponding attitude angle according to the attitude data; collecting the complete cycle time of the training action; calculating a training index of each limb region; inputting the training indexes into the evaluation model; and outputting a comprehensive evaluation coefficient. The method has the beneficial effects that independent quantitative evaluation of the activity ability of each limb region is realized through limb partitioning, multi-coordinate system construction and training index dynamic calculation, and a trainer is helped to identify local action compensation or limitation problems; a dual redundancy mechanism of deviation correction and optical motion compensation is integrated, through cooperative calibration of the inertial measurement unit and the optical mark, measurement errors caused by equipment installation deviation or limb shaking are remarkably reduced, and the reliability of multi-source data fusion is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of training effect evaluation, and in particular relates to a training effect evaluation method and system based on limb degree of freedom decomposition. Background Art

[0002] Training effect evaluation is a key link to gain an in-depth understanding of whether the actual effect of sports training has achieved the expected goals, thereby optimizing training plans and improving sports performance.

[0003] Traditional assessments often rely on subjective manual observation or simple one-dimensional kinematic parameters (such as the amplitude of joint angle changes). This qualitative analysis model struggles to accurately capture the spatiotemporal characteristics of movements. In particular, in the assessment of complex movements, the multi-degree-of-freedom coupled motions of different body parts are simplified into a single metric, resulting in the loss of critical details. On the one hand, existing assessment methods focus data collection on a single plane or local limb, lacking three-dimensional motion coupling analysis across multiple joints throughout the body. On the other hand, the assessment system still utilizes traditional metrics and lacks a quantitative model deeply integrated with the principles of sports biomechanics. This results in assessment results that fail to fully reflect movement quality and make it difficult to support precise training decisions. Therefore, a method and system for assessing the effectiveness of sports training that addresses these issues is urgently needed. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a training effect evaluation method and system based on limb degree of freedom decomposition, which is particularly suitable for accurate evaluation of specific training movements in different scenarios.

[0005] The technical solution adopted by the present invention is as follows: First, a training effect evaluation method based on limb degree of freedom decomposition is provided, comprising the following steps:

[0006] Divide the limbs into zones;

[0007] Defining the mapping relationship between posture angles and the range of motion of the divided limb areas;

[0008] Collecting posture data of joint anatomical landmarks in the limb area;

[0009] Establish a global coordinate system and a local coordinate system on the limb area respectively;

[0010] Calculate the corresponding attitude angle according to the attitude data;

[0011] Collect the complete cycle time of training movements;

[0012] Calculate training indicators for each limb area;

[0013] Feed the training metrics into the evaluation model;

[0014] Output comprehensive evaluation coefficient.

[0015] Furthermore, calculating the corresponding attitude angle according to the attitude data includes the following steps:

[0016] Generate the corresponding rotation matrix according to the posture data;

[0017] Use homogeneous coordinate transformation to construct the transformation matrix from the local coordinate system to the global coordinate system;

[0018] The posture angle of each limb region is obtained from the rotation matrix.

[0019] Furthermore, the attitude angle is calculated by the following equation:

[0020]

[0021] in, is the roll angle, is the pitch angle, is the heading angle, It is the quaternary data of the joint anatomical landmark points of each limb area.

[0022] Furthermore, calculating the training index of each limb area includes the following steps:

[0023] Divide the complete cycle of training action into multiple sub-time series;

[0024] Within each sub-time series, the repeatability index and accuracy index of each limb region were calculated separately;

[0025] The repeatability index is given by Eq. Calculate, where 𝑚 is the number of times the same trainee completes the standard action, is the standard deviation of the roll angle, is the standard deviation of the pitch angle, is the standard deviation of the heading angle, λ is the normalization coefficient, and the constraint 𝑅 𝑗 ∈[0,1];

[0026] The time series of measured attitude angles {α(𝑡), β(𝑡), γ(𝑡)} are compared with the standard trajectory { , , After alignment, the accuracy index is calculated by equation ,in , , is the weight coefficient (𝜔 𝛼 +𝜔 𝛽 +𝜔 𝛾 =1); 、 is the angular range of each attitude, is the interval 𝑇 𝑗 The number of internal sampling points.

[0027] Furthermore, the evaluation model is based on the distribution of stage weights, degree of freedom weights and indicator weights, and the equation for calculating the comprehensive evaluation coefficient is: ,in is the stage weight, is the degree of freedom weight, is the indicator weight, and For the degrees of freedom in Repeatability and accuracy indicators of time periods, is the comprehensive evaluation coefficient.

[0028] In a second aspect, a training effect evaluation system based on limb degree of freedom decomposition is provided, comprising: a data acquisition module, including an inertial measurement unit deployed at joint anatomical landmarks in the limb region, for collecting quaternion posture data of the joint anatomical landmarks in each limb region and a complete cycle time of a training movement;

[0029] The deviation correction module is used to record the initial quaternion of the inertial measurement unit when the user is in the standard anatomical posture, and calculate the deviation with the preset theoretical reference posture quaternion to obtain the installation deviation quaternion, and correct the real-time quaternion posture data;

[0030] The motion compensation module is an array of optical reflective markers arranged around the inertial measurement unit. It is used to calculate the rigid body transformation matrix of the rotation matrix and translation vector through singular value decomposition based on the synchronized 3D coordinate data of the markers, and correct the position data of the inertial measurement unit.

[0031] An attitude angle calculation module is used to calculate the corresponding attitude angle through attitude data;

[0032] An index calculation module, used to calculate repeatability index and accuracy index through cycle time and posture angle;

[0033] Comprehensive evaluation module, used to calculate comprehensive evaluation coefficients of repeatability index and accuracy index.

[0034] Furthermore, the installation deviation quaternion is obtained by the equation Δq=q init ⊗q ref −1 Calculation, where ⊗ represents quaternion multiplication, q ref −1 is the inverse of the reference quaternion, q init is the initial quaternion; through the equation For the original quaternion q raw Correction is performed, where Δq is the installation deviation quaternion, qcorrected is the corrected quaternion.

[0035] In a third aspect, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so as to enable the at least one processor to execute the training effect evaluation method based on limb degree of freedom decomposition provided in the present disclosure.

[0036] In a fourth aspect, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the training effect evaluation method based on limb degree of freedom decomposition provided by the present disclosure.

[0037] In a fifth aspect, a computer program product is provided, comprising a computer program / instruction, which, when executed by a processor, implements the training effect evaluation method based on limb degree of freedom decomposition provided by the present disclosure.

[0038] The advantages and positive effects of the present invention are as follows: due to the adoption of the above-mentioned technical scheme, by decomposing the target limb into 8 independent analysis units, using three-axis posture angle dynamic modeling of spatial motion trajectory, combining full-cycle angle data collection and time period analysis, constructing an evaluation model that integrates the dual indicators of repeatability and accuracy, thereby realizing quantitative evaluation of the standardization of specific training movements, reducing the error rate of movement standardization evaluation, providing objective and comprehensive technical support for personalized training results, and significantly improving the accuracy and standardization of evaluation results; through limb partitioning, multi-coordinate system construction and dynamic calculation of training indicators, independent quantitative evaluation of the activity ability of each limb area is realized, helping trainers to identify local movement compensation or limitation problems, and improving the refinement of evaluation and the pertinence of movement correction; integrating the dual redundant mechanism of deviation correction and optical motion compensation, through the coordinated calibration of the inertial measurement unit and the optical marker, significantly reducing the measurement error caused by equipment installation offset or limb jitter, and ensuring the reliability of multi-source data fusion. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of a flow chart of a training effect evaluation method for limb freedom decomposition according to an embodiment of the present invention

[0040] Figure 2 This is a structural diagram of the limb area division according to an embodiment of the present invention.

[0041] Figure 3 This is a flow chart of calculating the corresponding attitude angle according to an embodiment of the present invention.

[0042] Figure 4 This is a structural diagram of a training effect evaluation system for limb freedom decomposition according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The present disclosure is described more fully below with reference to the accompanying drawings, which illustrate exemplary embodiments of the present disclosure. The technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present disclosure.

[0044] like Figure 1 and Figure 2 As shown, the present invention provides a training effect evaluation method based on limb degree of freedom decomposition, comprising the following steps:

[0045] S100, dividing the limbs into regions;

[0046] Based on the biomechanics of human motion, the target limb area is divided into eight DOF segments: head and neck, trunk, left / right upper arm, forearm, and hand. Figure 2 The posture angle of each limb segment is defined according to the physiological range of joint motion.

[0047] S200, defining a mapping relationship between posture angles and the activity ranges of the divided limb areas;

[0048] Attitude describes an object's orientation in space. It typically consists of three Euler angles: yaw, pitch, and roll. Roll α represents coronal motion, rotating around the sagittal axis (X-axis); pitch β represents sagittal motion, rotating around the coronal axis (Y-axis); and yaw γ represents horizontal motion, rotating around the vertical axis (Z-axis).

[0049] Based on the principle of anatomical degree-of-freedom decomposition, the assessment area covers key kinematic chain units in the upper limbs and trunk. The mapping relationship between physiological activities and posture angles for each degree of freedom is as follows: In the head and neck, flexion / extension movement: the pitch angle (β) represents the amplitude of the head's up and down swing, corresponding to nodding and tilting the head; lateral flexion movement: the roll angle (α) represents the left and right tilt angle of the head, such as the ear-to-shoulder movement; axial rotation movement: the heading angle (γ) represents the ability to turn the head, such as the left and right gaze movement. In the upper arm (shoulder joint complex), sagittal plane flexion and extension: the pitch angle (β) is used to monitor the forward and backward swing of the arm, such as the arm swing and press; frontal plane abduction / adduction: the roll angle (α) is used to assess the lateral expansion and contraction of the arm, such as the lateral raise; horizontal plane internal / external rotation: the heading angle (γ) is used to quantify the rotation of the upper arm around the longitudinal axis, such as the throwing power and swimming stroke. In the forearm (elbow-proximal radioulnar joint complex), sagittal plane flexion and extension: elbow flexion and extension are recorded using the pitch angle (β), such as in dumbbell curls. Horizontal plane pronation / supination: forearm rotation around the radial and ulnar axes is characterized by the heading angle (γ), such as in screwdriving or a tennis backhand. In the wrist (wrist complex), sagittal plane palmar flexion / dorsiflexion: hand flexion (palm toward the arm) and dorsiflexion (palm away from the arm) are measured using the pitch angle (β), such as in wrist swing. Coronal plane ulnar / radial deviation: hand tilt toward the ulnar (medial) or radial (lateral) side is quantified using the roll angle (α), such as in wrist swing. In the trunk (thoracolumbar complex), sagittal flexion / extension: the pitch angle (β) is used to monitor the forward and backward tilt of the trunk, such as bowing and sit-ups; coronal plane tilt: the roll angle (α) is used to evaluate the left and right bending of the trunk, such as side flexion and yoga triangle pose; horizontal plane axial rotation: the heading angle (γ) is used to reflect the left and right twisting of the trunk, such as golf swing and boxing rotation.

[0050] S300, collecting posture data of joint anatomical landmarks in the limb area;

[0051] S400, establishing a global coordinate system and a local coordinate system in the limb region;

[0052] The global coordinate system is defined as fixed to the trunk reference point (T1 thoracic spinous process); the origin of the local coordinate system is set at the joint rotation center (such as the center of the humeral head of the shoulder joint).

[0053] S500, calculating the corresponding attitude angle according to the attitude data;

[0054] S600: Collect the complete cycle time of the training action;

[0055] S700, calculating the training index of each limb area;

[0056] S800, input the training indicators into the evaluation model;

[0057] S900: Output comprehensive evaluation coefficient.

[0058] By adopting the above method, through limb partitioning, multi-coordinate system construction and dynamic calculation of training indicators, an independent quantitative assessment of the activity ability of each limb area can be achieved, helping trainers to identify local movement compensation or limitation problems, and improving the refinement of the assessment and the targetedness of movement correction.

[0059] In order to solve the problem that the posture angle calculation deviation of the local posture data of the limbs is caused by the inconsistent reference system, which affects the credibility of the action evaluation, an implementation method is provided in this embodiment.

[0060] like Figure 3 As shown, in one embodiment, calculating the corresponding attitude angle according to the attitude data includes the following steps:

[0061] S501, generating a corresponding rotation matrix according to the posture data;

[0062] S502, constructing a transformation matrix from the local coordinate system to the global coordinate system using homogeneous coordinate transformation;

[0063] S503: Obtain the posture angle of each limb region from the rotation matrix.

[0064] By adopting the above method, homogeneous coordinate transformation is used to unify the data conversion in the global and local coordinate systems, eliminating the attitude angle calculation error caused by the difference in sensor installation position, ensuring the comparability of motion data in different areas, and enhancing the consistency of evaluation results.

[0065] In order to solve the problem that posture angle calculation relies on empirical formulas or simplified models, which may lead to inaccurate angle mapping due to the coupling of joint degrees of freedom, an implementation method is provided in this embodiment.

[0066] In one embodiment, the attitude angle is calculated using the following equation:

[0067]

[0068] in, is the roll angle, is the pitch angle, is the heading angle, It is the quaternary data of the joint anatomical landmark points of each limb area.

[0069] Using the above method, the standardized posture angle calculation formula based on quaternions can fully restore the three-dimensional motion freedom of the joint, avoid the one-sidedness of single angle calculation, and support accurate analysis of complex movements (such as shoulder external rotation and flexion).

[0070] In order to solve the problem that traditional indicator calculation ignores the stage differences within the action cycle and cannot dynamically track the changes in action stability and accuracy, an implementation method is provided in this embodiment.

[0071] In one embodiment, calculating the training index of each limb area includes the following steps:

[0072] Divide the complete cycle of training action into multiple sub-time series;

[0073] For a specific training action (such as shooting action), the complete cycle time T from the start to the end of the action is collected, and the action cycle T is divided into n sub-time series T1, T2...T n .

[0074] Within each sub-time series, the repeatability index and accuracy index of each limb region were calculated separately;

[0075] For each limb unit (such as the right forearm), the three-axis posture angle time series is synchronously collected: α sequence: 𝛼(T)={𝛼𝑇1,𝛼𝑇2,...,𝛼𝑇 𝑛}, characterizing the lateral stability of the limb segment; β sequence: 𝛽(T)={𝛽𝑇1,𝛽𝑇2,...,𝛽𝑇 𝑛}, reflecting the amplitude of the forward and backward movement; γ sequence: 𝛾(T)={𝛾𝑇1,𝛾𝑇2,...,𝛾𝑇 𝑛}, describes the steering accuracy in the horizontal plane.

[0076] The repeatability index is given by Eq. Calculate, where 𝑚 is the number of times the same trainee completes the standard action, is the standard deviation of the roll angle, is the standard deviation of the pitch angle, is the standard deviation of the heading angle, λ is the normalization coefficient, and the constraint 𝑅 𝑗 ∈[0,1];

[0077] The same trainee completes 𝑚 standard movements (𝑚≥10), and obtains m training data. The 𝑘th movement of the target limb is calculated at T 𝑗 Standard deviation of attitude angle within the time period 、 、 :

[0078]

[0079] Where:

[0080] is the rolling angle value of the k-th action at time point t;

[0081] : The kth action is in 𝑇 𝑗 The average roll angle within

[0082] = sampling rate × time interval length, which is the interval 𝑇 𝑗 The number of internal sampling points is calculated similarly (T j ), (T j ).

[0083] The time series of measured attitude angles {α(𝑡), β(𝑡), γ(𝑡)} are compared with the standard trajectory { , , After alignment, the accuracy index is calculated by equation ,in , , is the weight coefficient (𝜔 𝛼 +𝜔 𝛽 +𝜔 𝛾 =1); 、 is the angular range of each attitude, is the interval 𝑇 𝑗 The number of internal sampling points.

[0084] By adopting the above method, through the decomposition of action cycles and the alignment design of time series, the repeatability (action consistency) and accuracy (benchmarking standard trajectory) indicators are output synchronously, and the fluctuation characteristics of the trainee's action control ability are intuitively fed back, assisting in the formulation of segmented intensive training plans.

[0085] In order to solve the problem that comprehensive evaluation relies on manual experience weighting, is highly subjective and difficult to adapt to the core objectives of different training stages, an implementation method is provided in this embodiment.

[0086] In one embodiment, the evaluation model is based on the distribution of stage weights, degree of freedom weights and indicator weights, and the equation for calculating the comprehensive evaluation coefficient is: ,in is the stage weight, is the degree of freedom weight, is the indicator weight, and For the degrees of freedom in Repeatability and accuracy indicators of time periods, is the comprehensive evaluation coefficient.

[0087] : Stage weight, which is allocated according to the importance of motion mechanics to each time series (e.g., the weight of T1 is 0.5); : Degree of freedom weight, set according to the importance of each limb segment function (e.g., the weight of the upper arm β angle is 0.3); 𝜆: Index weight (𝜆∈[0,1]), balancing repeatability and accuracy (default 0.5); , : The dth degree of freedom (each limb segment) is in 𝑇 𝑗 Repeatability and accuracy scores for time periods.

[0088] By using the above method and combining the multi-dimensional dynamic allocation mechanism of stage weights, degree of freedom weights and indicator weights, we can automatically generate evaluation results that adapt to individual needs (such as focusing on safety in the early stages of postoperative rehabilitation and focusing on explosive power in competitive training), thereby improving the versatility of the model.

[0089] like Figure 4 As shown, in order to facilitate the use of the training effect evaluation method based on limb degree of freedom decomposition provided by the present disclosure, the present disclosure also provides a training effect evaluation system based on limb degree of freedom decomposition, including:

[0090] The data acquisition module 10 includes an inertial measurement unit 11 deployed at the joint anatomical landmarks in the limb region, and is used to collect quaternion posture data of the joint anatomical landmarks in each limb region and the complete cycle time of the training movement;

[0091] The deviation correction module 20 is used to record the initial quaternion of the inertial measurement unit 11 when the user is in the standard anatomical posture, and calculate the deviation with the preset theoretical reference posture quaternion to obtain the installation deviation quaternion, and correct the real-time quaternion posture data;

[0092] When the subject is in a standard anatomical posture (such as upright position, arms hanging naturally, palms facing inward), record the initial quaternion q output by each IMU init At the same time, set the quaternion q of the standard reference attitude ref , the theoretical pose under this posture is usually used as a reference (for example, the q ref (corresponding to the head facing forward horizontally without deflection).

[0093] The motion compensation module 30 is an array of optical reflective markers arranged around the inertial measurement unit 11, which is used to calculate the rigid body transformation matrix of the rotation matrix and translation vector through singular value decomposition based on the synchronized three-dimensional coordinate data of the markers, and correct the position data of the inertial measurement unit 11;

[0094] Paste reflective markers near each IMU installation location and simultaneously collect the IMU original position data P IMU-rawThe three-dimensional coordinates of the marker points. It is recommended to arrange at least three non-collinear marker points around each IMU to build a stable rigid body model. Based on the three-dimensional coordinates of the marker points, the rigid body transformation matrix T is calculated by least squares method or singular value decomposition (SVD). This matrix contains rotation (Rotation) and translation (Translation) information, which is used to describe the displacement change of the sensor relative to the bone when the soft tissue moves: T=[ Among them, the rotation matrix R describes the posture change, and the translation vector t describes the position offset. The rigid body transformation matrix is ​​used to dynamically compensate the original position coordinates of the IMU: P IMU-corrected = P IMU-raw Corrected coordinate P IMU-corrected It reflects the true position of the sensor after removing the interference of soft tissue movement, improving motion trajectory analysis.

[0095] An attitude angle calculation module 40 is used to calculate the corresponding attitude angle through the attitude data;

[0096] An index calculation module 50 is used to calculate a repeatability index and an accuracy index through cycle time and posture angle;

[0097] The comprehensive evaluation module 60 is used to calculate the comprehensive evaluation coefficient based on the repeatability index and the accuracy index.

[0098] The above-mentioned device integrates a dual redundant mechanism of deviation correction and optical motion compensation. Through the coordinated calibration of the inertial measurement unit and the optical marker, it significantly reduces the measurement error caused by equipment installation offset or limb shaking, and ensures the reliability of multi-source data fusion.

[0099] In order to solve the problem that sensor installation deviation requires repeated manual calibration, which is cumbersome and has poor real-time performance, an implementation method is provided in this embodiment.

[0100] In one embodiment, the installation deviation quaternion is given by the equation Δq=q init ⊗q ref −1 Calculation, where ⊗ represents quaternion multiplication, q ref −1 is the inverse of the reference quaternion, q init is the initial quaternion; through the equation For the original quaternion q raw Correction is performed, where Δq is the installation deviation quaternion, q corrected is the corrected quaternion.

[0101] The above device can realize automatic compensation of installation deviation through quaternion inverse operation, without relying on manual calibration process, simplifying the equipment deployment steps, supporting real-time correction in dynamic training scenarios, and improving the system usability and response speed.

[0102] The following describes the contents involved in the above embodiment in conjunction with a preferred embodiment.

[0103] In the assessment of shooting athletes' gun-holding stability, the right arm data is analyzed first in this example because the movement characteristics and training goals focus on the right arm (gun-holding side).

[0104] According to the biomechanical characteristics of the shooting action, the target limb is decomposed into the following independent analysis units:

[0105] Head and neck: pitch angle (β), heading angle (γ), monitoring the head fine-tuning amplitude;

[0106] Right upper arm: pitch angle (β), heading angle (γ), quantitative gun stability and rotation control;

[0107] Right forearm: pitch angle (β), heading angle (γ), to assess elbow flexion and extension and forearm rotation;

[0108] Right wrist joint: pitch angle (β), roll angle (α), detecting wrist dorsiflexion / lateral tilt;

[0109] Torso: Heading angle (γ), reflecting the accuracy of the torso axial alignment.

[0110] The data acquisition module utilizes a hybrid inertial-optical capture solution. Nine-axis inertial measurement units (IMUs) (model BMI270, 200Hz sampling rate) are deployed at key anatomical locations on the athlete to simultaneously capture motion data from the head, neck, torso, shoulder, elbow, and wrist, collecting quaternion data. Optical reflective markers are also installed on the front end of the barrel, allowing for simultaneous recording of the gun's trajectory using a Qualisys infrared capture system (0.1mm accuracy). The head and neck IMU is secured to the mastoid bone behind the right ear. This anatomical location is close to the center of head rotation, enabling precise capture of the pitch angle caused by chin tuck during aiming, as well as the heading angle changes caused by correcting the aiming baseline. The IMU is secured in an X-shaped pattern using breathable medical tape. The right upper arm IMU is positioned at the center of shoulder rotation (the greater tuberosity of the humerus, 5cm below the acromion), while the right forearm (right elbow) IMU is mounted 2cm distal to the lateral epicondyle of the humerus. Both IMUs are secured with medical-grade elastic bands, with the band tension adjusted to allow for insertion of a finger to prevent sensor slippage and blood flow restriction that could affect performance. The right wrist IMU was mounted 1.5 cm distal to the radial styloid process, using an elastic wrist brace with a 0.5 mm silicone cushion to minimize skin deformation. The torso IMU was located at the T1 thoracic vertebra, adhered to the skin with double-sided medical tape and reinforced with a breathable elastic strap. The global coordinate system had its origin at the T1 thoracic vertebra; the local coordinate system had its origin at the shoulder joint's center of rotation, with the Z axis pointing toward the muzzle.

[0111] According to the biomechanics research of shooting sports, the shooting action cycle T=4s is divided into four key stages: T1 preparation period (0-1.0s): stance adjustment and neutral position calibration; T2 gun raising period (1.0-2.0s): raising the gun arm to the aiming position; T3 aiming period (2.0-3.5s): muzzle fine-tuning and stabilization; T4 firing period (3.5-4.0s): pulling the trigger and maintaining the posture. The standard action data of the same trainee was collected continuously for 10 times, and the original data of the three-axis attitude angle of each time series was extracted from the IMU and optical system with a sampling frequency of 200Hz; based on the 3σ principle, the angle values ​​that exceeded the physiological activity range were eliminated (for example, the head β angle movement range is 15° flexion to 20° extension); taking the right upper arm as an example, the repeatability index and accuracy index calculation were performed respectively. Repeatability index 𝑅 𝑗 :The right arm only considers β (pitch angle) and γ (heading angle), so the standard deviation of the angle 𝛼 Set to zero. For 10 training data, calculate the right upper arm's 𝑘th action at T 𝑗 Standard deviation of attitude angle within the time period , :

[0112]

[0113]

[0114] Calculate the repeatability index score:

[0115]

[0116] Where m=10, λ is the normalization coefficient (constraint 𝑅 𝑗 ∈[0,1]).

[0117] Accuracy index 𝐴 𝑗 :Use the dynamic time warping algorithm to compare the right arm measured attitude angle time series {β(𝑡),γ(𝑡)} with the preset standard trajectory { , }After alignment, the weighted root mean square error is calculated.

[0118]

[0119] in, , is the weight coefficient ( + ), the right upper arm flexion and extension range is 0°~90°, and the rotation range is -30°~20°, so 𝐿 𝛽 =90°L γ =50°, n jis the number of sampling points in the time period (e.g., in time period T1, n1=1.0s×200Hz=200 points).

[0120] The evaluation model adopts a dynamic weight allocation mechanism, and all weight coefficients must meet the normalization condition (∑ =1,∑ =1) The specific settings are as follows: Stage weight : Weights are assigned based on the degree of influence of each stage of the shooting action on stability. T3 and T4 each have a weight of 0.4, T1 and T2 each have a weight of 0.1; Degree of freedom weight Based on the anatomical contribution distribution: right upper arm β accounts for 0.35; right upper arm γ accounts for 0.25; head and neck β accounts for 0.15; and the remaining degrees of freedom account for a total of 0.25. Indicator weight (𝜆): Because the shooting training goal focuses on stability, the repeatability weight 𝜆 is set to 0.7.

[0121] The comprehensive evaluation coefficient 𝑆 is calculated by the following formula:

[0122]

[0123] The value of the comprehensive evaluation coefficient S is positively correlated with the training effect, and its quantitative result can directly represent the standardization of the training results.

[0124] Based on the embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0125] An electronic device includes at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the training effect evaluation method based on limb degree of freedom decomposition provided by the present disclosure.

[0126] Electronic device is intended to refer to various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device may also refer to various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are intended to be examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0127] A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the training effect evaluation method based on limb degree of freedom decomposition provided by the present disclosure.

[0128] Various embodiments of the present disclosure may be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system comprising at least one programmable processor, which may be a special-purpose or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0129] A computer program product includes a computer program / instruction, which, when executed by a processor, provides a training effect evaluation method based on limb degree of freedom decomposition provided by the present disclosure.

[0130] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0131] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0132] The embodiments of the present invention are described in detail above, but the contents described are only preferred embodiments of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A training effect evaluation method based on limb freedom decomposition, characterized in that: The following steps are involved: Divide the limbs into zones; Defining the mapping relationship between posture angles and the range of motion of the divided limb areas; Collecting posture data of joint anatomical landmarks in the limb area; Establish a global coordinate system and a local coordinate system on the limb area respectively; Calculate the corresponding attitude angle according to the attitude data; Collect the complete cycle time of training movements; Calculate training indicators for each limb area; Feed the training metrics into the evaluation model; Output comprehensive evaluation coefficient.

2. The training effect evaluation method based on limb degree of freedom decomposition according to claim 1 is characterized in that: Calculating the corresponding attitude angle based on attitude data includes the following steps: Generate the corresponding rotation matrix according to the posture data; Use homogeneous coordinate transformation to construct the transformation matrix from the local coordinate system to the global coordinate system; The posture angle of each limb region is obtained from the rotation matrix.

3. The training effect evaluation method based on limb degree of freedom decomposition according to claim 1 or 2, characterized in that: The attitude angle is calculated using the following equation: in, is the roll angle, is the pitch angle, is the heading angle, ( , , , ) is the quaternary data of the joint anatomical landmark points in each limb area.

4. The training effect evaluation method based on limb degree of freedom decomposition according to claim 1 is characterized in that: Calculating the training index for each limb area includes the following steps: Divide the complete cycle of training action into multiple sub-time series; Within each sub-time series, the repeatability index and accuracy index of each limb region were calculated separately; The repeatability index is given by Eq. Calculate, where 𝑚 is the number of times the same trainee completes the standard action, is the standard deviation of the roll angle, is the standard deviation of the pitch angle, is the standard deviation of the heading angle, λ is the normalization coefficient, and the constraint 𝑅 𝑗 ∈[0,1]; The time series of measured attitude angles {α(𝑡), β(𝑡), γ(𝑡)} are compared with the standard trajectory { , , After alignment, the accuracy index is calculated by equation ,in , , is the weight coefficient (𝜔 𝛼 +𝜔 𝛽 +𝜔 𝛾 =1); 、 is the angular range of each attitude, is the interval 𝑇 𝑗 The number of internal sampling points.

5. The training effect evaluation method based on limb degree of freedom decomposition according to claim 1 is characterized in that: The evaluation model is based on the distribution of stage weights, degree of freedom weights and indicator weights. The equation for calculating the comprehensive evaluation coefficient is: ,in is the stage weight, is the degree of freedom weight, is the indicator weight, and For the degrees of freedom in Repeatability and accuracy indicators of time periods, is the comprehensive evaluation coefficient.

6. A training effect evaluation system based on limb freedom decomposition, characterized in that: include: A data acquisition module, including an inertial measurement unit deployed at joint anatomical landmarks in the limb region, for collecting quaternion posture data of the joint anatomical landmarks in each limb region and the complete cycle time of the training movement; The deviation correction module is used to record the initial quaternion of the inertial measurement unit when the user is in the standard anatomical posture, and calculate the deviation with the preset theoretical reference posture quaternion to obtain the installation deviation quaternion, and correct the real-time quaternion posture data; The motion compensation module is an array of optical reflective markers arranged around the inertial measurement unit. It is used to calculate the rigid body transformation matrix of the rotation matrix and translation vector through singular value decomposition based on the synchronized 3D coordinate data of the markers, and correct the position data of the inertial measurement unit. An attitude angle calculation module is used to calculate the corresponding attitude angle through attitude data; An index calculation module, used to calculate repeatability index and accuracy index through cycle time and posture angle; Comprehensive evaluation module, used to calculate comprehensive evaluation coefficients of repeatability index and accuracy index.

7. The training effect evaluation system based on limb degree of freedom decomposition according to claim 6, characterized in that: The installation deviation quaternion is obtained by the equation Δq=q init ⊗q ref −1 Calculation, where ⊗ represents quaternion multiplication, q ref −1 is the inverse of the reference quaternion, q init is the initial quaternion; through the equation For the original quaternion q raw Correction is performed, where Δq is the installation deviation quaternion, q corrected is the corrected quaternion.

8. An electronic device comprising: at least one processor; as well as a memory communicatively connected to at least one processor; wherein, The memory stores instructions that can be executed by at least one processor. The instructions are executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 5.

9. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to make a computer execute the method according to any one of claims 1 to 5.

10. A computer program product comprising a computer program / instructions, which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 5.

Images