A gait analysis method and apparatus

Abstract

This invention discloses a gait analysis method and apparatus, relating to the field of data processing. It addresses the problems of existing gait analysis methods, such as high cost, complex operation and maintenance, and the interference of reflective markers / sensors with the test subject's natural gait. The method includes: acquiring temporal initial depth images simultaneously acquired by at least two depth image acquisition devices; processing a preset original skeleton model based on the temporal initial depth images to obtain the test skeleton of the subject and the first motion trajectory of each feature point of the test skeleton; matching the test skeleton with a preset original musculoskeletal model to obtain a test musculoskeletal model that conforms to the body shape characteristics of the subject; acquiring the second motion trajectory of each joint point of the test musculoskeletal model; and performing gait analysis calculations on the subject based on the second motion trajectory of each joint point of the test musculoskeletal model to obtain the average dynamic parameters, average kinematic parameters, and average spatiotemporal parameters of the subject's gait movement.

Description

Technical Field

[0001] This invention relates to the field of data processing, and in particular to a gait analysis method and apparatus. Background Technology

[0002] Human gait is a complex task that requires the coordination of multiple joints to achieve a precise movement trajectory. Gait analysis, a specialized branch of biomechanics, involves the kinematic observation and dynamic analysis of limb and joint movements during human walking, providing a series of gait parameters such as spatiotemporal parameters, joint angle parameters, dynamic parameters, and muscle parameters.

[0003] In existing technologies, typical gait analysis devices consist of reflective markers, an infrared camera, and a computer. The reflective markers are attached to the part of the body to be measured. The infrared camera receives the infrared signals emitted by the markers and reflected back, and the computer generates the body's motion coordinates. Finally, the relevant gait parameters are calculated based on these coordinates.

[0004] Existing gait analysis processes require specialized equipment such as infrared cameras and reflective markers / sensors, which are costly and complex to operate and maintain; moreover, wearing such equipment can interfere with the test subject's natural gait. Summary of the Invention

[0005] In view of this, the main objective of the present invention is to solve the problems of existing gait analysis methods being costly, complex to operate and maintain, and the fact that wearing markers / sensors and other devices can interfere with the test subject's natural gait.

[0006] On one hand, the present invention provides a gait analysis method, comprising: acquiring temporal initial depth images synchronously acquired by at least two depth image acquisition devices, wherein the temporal initial depth images are obtained by omnidirectional acquisition of the gait movement process of a person under test by the at least two depth image acquisition devices; processing a preset original skeleton model based on the temporal initial depth images to obtain a skeleton to be tested of the person under test and the first motion trajectory of each feature point of the skeleton to be tested; matching a preset original musculoskeletal model based on the skeleton to be tested to obtain a musculoskeletal model to be tested that conforms to the body shape characteristics of the person under test; and performing drive tracking optimization on the musculoskeletal model to be tested based on the first motion trajectory of each feature point of the skeleton to be tested to obtain each joint of the musculoskeletal model to be tested. The second motion trajectory of the points; based on the second motion trajectory of each joint point of the musculoskeletal model to be tested, gait analysis calculations are performed on the person under test to obtain the average dynamic parameters, average kinematic parameters, and average spatiotemporal parameters of the gait movement of the person under test; the average dynamic parameters include the average change curves of the triaxial ground reaction force and torque variation curves of each gait cycle, the average change curves of the joint contact force and torque variation curves of each gait cycle, and the average change curves of the muscle force and muscle activation variation curves of each gait cycle; the average kinematic parameters include the average change curves of the joint trajectory variation curves and the average change curves of the joint angle variation curves of each gait cycle; the average spatiotemporal parameters include the average stride length, average gait speed, and average gait cycle.

[0007] On the other hand, the present invention provides a gait analysis device, comprising:

[0008] The image acquisition module is used to acquire temporal initial depth images synchronously acquired by at least two depth image acquisition devices. The temporal initial depth images are obtained by omnidirectional acquisition of the gait movement process of the person under test by the at least two depth image acquisition devices.

[0009] The first trajectory acquisition module, connected to the image acquisition module, is used to process the preset original skeleton model according to the temporal initial depth image to obtain the skeleton of the person to be tested and the first motion trajectory of each feature point of the skeleton to be tested.

[0010] The musculoskeletal matching module, connected to the first trajectory acquisition module, is used to match the preset original musculoskeletal model with the skeleton to be tested to obtain a musculoskeletal model that conforms to the body shape characteristics of the person to be tested.

[0011] The trajectory optimization module is connected to the first trajectory acquisition module and the musculoskeletal matching module respectively, and is used to drive and track the musculoskeletal model under test according to the first motion trajectory of each feature point of the skeleton under test, and to obtain the second motion trajectory of each joint of the musculoskeletal model under test.

[0012] The gait analysis module is connected to the image acquisition module and the trajectory optimization module respectively. It is used to perform gait analysis calculations on the test subject based on the second motion trajectory of each joint of the musculoskeletal model under test, and to obtain the average dynamic parameters, average kinematic parameters and average spatiotemporal parameters of the test subject's gait motion.

[0013] The average dynamic parameters include the average variation curves of triaxial ground reaction force and torque variation curves for each gait cycle, the average variation curves of joint contact force and torque variation curves for each gait cycle, and the average variation curves of muscle force and muscle activation variation curves for each gait cycle.

[0014] The average kinematic parameters include the average curves of joint trajectory changes and the average curves of joint angle changes for each gait cycle.

[0015] The average spatiotemporal parameters include average stride length, average gait speed, and average gait cycle.

[0016] In summary, the technical solution provided by this invention involves simultaneously acquiring temporal initial depth images of the gait movement of a person under test using at least two depth image acquisition devices. After processing and driving tracking optimization using the original skeleton model and original musculoskeletal model, gait analysis is performed to obtain the gait analysis results of the person under test. The technical solution provided by this invention reduces costs by requiring only at least two depth image acquisition devices and eliminating the need for reflective markers, simplifying operation and maintenance. It allows for convenient measurement and analysis anytime, anywhere. Furthermore, it eliminates the need for markers / sensors and other equipment, allowing for the acquisition of the test subject's natural gait. It also provides gait parameters such as average dynamic parameters, average kinematic parameters, and average spatiotemporal parameters, enabling complete gait analysis. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart of the gait analysis method provided in Embodiment 1 of the present invention;

[0019] Figure 2 for Figure 1 A schematic diagram of the original skeleton model in the gait analysis method shown;

[0020] Figure 3 This is a schematic diagram of the gait analysis device provided in Embodiment 2 of the present invention;

[0021] Figure 4 for Figure 3 A schematic diagram of the structure of the first trajectory acquisition module in the gait analysis device shown;

[0022] Figure 5 for Figure 3 The diagram shows the structure of the gait analysis module in the gait analysis device. Detailed Implementation

[0023] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Example 1

[0025] like Figure 1 As shown, the present invention provides a gait analysis method, comprising:

[0026] Step 101: Obtain the temporal initial depth images synchronously acquired by at least two depth image acquisition devices.

[0027] In this embodiment, the initial temporal depth image in step 101 is obtained by acquiring the gait movement process of the person under test from all directions through at least two depth image acquisition devices. At least two depth image acquisition devices can capture the gait movement process of the person under test in 360° without blind spots.

[0028] For example, when there are 2 depth image acquisition devices, the two depth image acquisition devices can be set up facing each other; when there are 3 depth image acquisition devices, the three depth image acquisition devices can be set up at the three vertices of an equilateral triangle; other methods can also be used to set up depth image acquisition devices, as long as they can capture the gait movement process of the person being tested in 360° without blind spots, which will not be elaborated here.

[0029] At least two image acquisition devices can capture the gait movement of the subject within a specific range at a certain frequency, such as 30 Hz, without limitation.

[0030] Step 102: Process the preset original skeleton model based on the initial depth image of the time series to obtain the skeleton of the person to be tested and the first motion trajectory of each feature point of the skeleton.

[0031] In this embodiment, the specific processing procedure of step 102 may include: fusing at least two synchronized initial temporal depth images to obtain a temporal fused depth image; extracting the temporal gait depth image of the person under test from the temporal fused depth image; and processing the preset original skeleton model according to the temporal gait depth image of the person under test to obtain the skeleton of the person under test and the first motion trajectory of each feature point of the skeleton of the skeleton under test.

[0032] The pre-defined original skeleton model includes at least the feature points necessary for gait feature analysis, such as the pelvic position, and points at the hip, knee, ankle, and toes on both sides of the lower limbs. The structure of the original skeleton model can be referenced from [reference needed]. Figure 2 , Figure 2 The joint points, skeleton length, etc. are for reference only and can be adjusted as needed. They will not be elaborated on here.

[0033] When performing image fusion on at least two synchronized initial depth images, the image fusion can be performed based on the relative positions of at least two depth image acquisition devices.

[0034] The temporal gait depth image of the subject can be extracted using the background segmentation extraction method. The specific process may include: acquiring temporal background depth images pre-acquired by at least two depth image acquisition devices; extracting the temporal gait depth image of the subject from the temporal fused depth image based on the temporal background depth image using the background segmentation extraction method; the background segmentation extraction method includes: background difference method, optical flow method or adjacent frame difference method.

[0035] The specific process of processing the preset original skeleton model based on the temporal gait depth image of the subject may include: unifying the coordinate system of the temporal gait depth image of the subject with the coordinate system of the preset original skeleton model to obtain a temporal unified depth image and a unified skeleton model; matching and adjusting the unified skeleton model according to the initial frame image in the temporal unified depth image to obtain an initial skeleton model that matches the subject; iteratively optimizing the initial skeleton model and all temporal unified depth images to obtain the subject skeleton; and obtaining the first motion trajectory of each feature point of the subject skeleton. The method for obtaining the first motion trajectory of each feature point of the subject skeleton can be: iteratively optimizing the subject skeleton and all temporal unified depth images to obtain the first motion trajectory of each feature point of the subject skeleton; or, obtaining the first motion trajectory of each feature point of the subject skeleton based on all temporal unified depth images and pose images in a preset pose standard training set.

[0036] Matching and adjusting the unified skeleton model based on the initial frame image in the temporal unified depth image can be achieved by rotating, translating, aligning, and fitting the unified skeleton model according to the initial frame image. The process of obtaining the first motion trajectory of each feature point of the skeleton under test based on all temporal unified depth images and pose maps in the preset pose standard training set can be as follows: match the skeleton under test with the temporal unified depth images respectively, determine the best matching image, and then perform interpolation retrieval on the best matching image to obtain the first motion trajectory of each feature point of the skeleton under test. To save matching time, this matching can be a bottom-up matching.

[0037] Step 103: Match the preset original musculoskeletal model with the skeleton to be tested to obtain a musculoskeletal model that matches the body shape characteristics of the person to be tested.

[0038] In this embodiment, the musculoskeletal model in step 103 is a model simulated based on the muscle and skeletal structure of the human body.

[0039] Step 104: Based on the first motion trajectory of each feature point of the skeleton under test, drive tracking optimization is performed on the musculoskeletal model under test to obtain the second motion trajectory of each joint point of the musculoskeletal model under test.

[0040] In this embodiment, the method of obtaining the second motion trajectory through step 104 can be as follows: during the driving tracking process, the first motion trajectory of each feature point of the skeleton under test is matched with the motion trajectory of each joint point of the musculoskeletal model under test, and the motion trajectory of each joint point of the musculoskeletal model under test is optimized by minimizing the labeling error method to obtain the second motion trajectory of each joint point of the musculoskeletal model under test.

[0041] During the optimization process, the maximum range of motion of each joint and the maximum force exerted by each muscle can be referenced to make the second motion trajectory of each joint of the musculoskeletal model under test more closely match the actual situation of human movement.

[0042] Step 105: Based on the second motion trajectory of each joint of the musculoskeletal model to be tested, perform gait analysis and calculation on the test subject to obtain the average dynamic parameters, average kinematic parameters and average spatiotemporal parameters of the test subject's gait movement.

[0043] In this embodiment, the average kinematic parameters in step 105 include the average curves of the triaxial ground reaction force and torque variation curves for each gait cycle, the average curves of the joint contact force and torque variation curves for each gait cycle, and the average curves of the muscle force and muscle activation variation curves for each gait cycle. The average kinematic parameters include the average curves of the joint trajectory variation curves and the average curves of the joint angle variation curves for each gait cycle. The average spatiotemporal parameters include the average stride length, average gait speed, and average gait cycle.

[0044] The average muscle activation can be obtained by the ratio of the average muscle force of a muscle to the maximum force exerted by that muscle.

[0045] The specific process of gait analysis calculation in step 105 may include: dividing the gait of the person under test according to the second motion trajectory of each joint of the musculoskeletal model under test, and obtaining the average spatiotemporal parameters; driving the musculoskeletal model under test to perform inverse dynamics calculation according to the second motion trajectory of each joint of the musculoskeletal model under test, and obtaining the corresponding average values, to obtain the average dynamic parameters and average kinematic parameters.

[0046] The process of obtaining the average spatiotemporal parameters involves dividing the second motion trajectory of each joint of the musculoskeletal model under test into gait segments to determine the motion trajectory of each joint within a complete gait cycle; then determining the spatiotemporal parameters corresponding to each complete gait cycle based on the motion trajectory of each joint within each complete gait cycle; and finally averaging the spatiotemporal parameters corresponding to all complete gait cycles to obtain the average spatiotemporal parameters.

[0047] It can obtain the curve of the displacement difference between the heel and the center of the pelvis in the forward direction over time, and take the time period corresponding to the maximum displacement difference of two adjacent maxima as a gait cycle.

[0048] The feedback from the preset contact points on the sole of the foot of the musculoskeletal model under test identifies the moment the foot touches the ground, and then divides the gait cycle; that is, the time interval between two adjacent heel touches of the same foot is taken as a gait cycle.

[0049] The process of obtaining average dynamic parameters and average kinematic parameters may include: obtaining the triaxial ground reaction force and torque variation curves for each gait cycle based on the feedback results of the preset contact points on the sole of the foot of the musculoskeletal model under test during the inverse dynamics calculation process; establishing the second motion trajectory of each joint of the musculoskeletal model under test and the mechanical balance equation of the average triaxial ground reaction force and torque, and obtaining the joint contact force and torque variation curves for each gait cycle; establishing the mechanical balance equation between the average joint contact force and torque and the muscle force according to the muscle recruitment principle, and obtaining the muscle force variation curves for each part of the muscles in each gait cycle; obtaining the muscle activation variation curves for each gait cycle based on the muscle force variation curves for each part of the muscles and the preset maximum force value; and obtaining the average variation curves corresponding to the triaxial ground reaction force and torque variation curves, the joint contact force and torque variation curves, and the muscle force and muscle activation variation curves for each part of the muscles in each gait cycle, respectively, to obtain the average dynamic parameters.

[0050] The maximum range of motion of each joint and the maximum force exerted by each muscle can be set in advance according to human physiological characteristics; the mechanical balance equation mainly includes the laws of translation and rotation, and the sum of all external forces is equal to the sum of the products of the weight and acceleration of all individual body segments; muscle recruitment refers to which muscles and muscle activation levels are selected when performing inverse dynamics calculations in order to minimize the total sum of muscle activation levels; the average change curve of each parameter can be obtained by interpolating all the change curves of each parameter to a specified number and then averaging them, which will not be elaborated here.

[0051] Furthermore, based on the variation curves of each parameter in each gait cycle, the maximum and minimum values ​​of each parameter in each gait cycle can be obtained, thereby determining the variation range of each parameter; the parameter values ​​and corresponding percentages of each parameter can also be obtained, which will not be elaborated here.

[0052] In summary, the technical solution provided by this invention involves simultaneously acquiring temporal initial depth images of the gait movement of a person under test using at least two depth image acquisition devices. After processing and driving tracking optimization using the original skeleton model and original musculoskeletal model, gait analysis is performed to obtain the gait analysis results of the person under test. The technical solution provided by this invention reduces costs by requiring only at least two depth image acquisition devices and eliminating the need for reflective markers, simplifying operation and maintenance. It allows for convenient measurement and analysis anytime, anywhere. Furthermore, it eliminates the need for markers / sensors and other equipment, allowing for the acquisition of the test subject's natural gait. It also provides gait parameters such as average dynamic parameters, average kinematic parameters, and average spatiotemporal parameters, enabling complete gait analysis.

[0053] Example 2

[0054] like Figure 3 As shown, an embodiment of the present invention provides a gait analysis device, comprising:

[0055] The image acquisition module 301 is used to acquire the temporal initial depth image synchronously acquired by at least two depth image acquisition devices. The temporal initial depth image is obtained by omnidirectional acquisition of the gait movement process of the person under test by at least two depth image acquisition devices.

[0056] The first trajectory acquisition module 302 is connected to the image acquisition module and is used to process the preset original skeleton model according to the temporal initial depth image to obtain the skeleton of the person to be tested and the first motion trajectory of each feature point of the skeleton.

[0057] Musculoskeletal matching module 303 is connected to the first trajectory acquisition module and is used to match the preset original musculoskeletal model according to the skeleton to be tested, so as to obtain a musculoskeletal model to be tested that conforms to the body shape characteristics of the person to be tested.

[0058] The trajectory optimization module 304 is connected to the first trajectory acquisition module and the musculoskeletal matching module respectively. It is used to drive and track the musculoskeletal model under test according to the first motion trajectory of each feature point of the skeleton under test, and to obtain the second motion trajectory of each joint of the musculoskeletal model under test.

[0059] The gait analysis module 305 is connected to the image acquisition module and the trajectory optimization module respectively. It is used to perform gait analysis calculations on the test subject based on the second motion trajectory of each joint of the musculoskeletal model to be tested, and to obtain the average dynamic parameters, average kinematic parameters and average spatiotemporal parameters of the test subject's gait motion.

[0060] The average dynamic parameters include the average variation curves of triaxial ground reaction force and torque variation curves for each gait cycle, the average variation curves of joint contact force and torque variation curves for each gait cycle, and the average variation curves of muscle force and muscle activation variation curves for each gait cycle.

[0061] The average kinematic parameters include the average curves of joint trajectory changes and the average curves of joint angle changes for each gait cycle.

[0062] The average spatiotemporal parameters include average stride length, average gait speed, and average gait cycle.

[0063] In this embodiment, the process of gait analysis through the above modules is similar to that provided in Embodiment 1 of the present invention, and will not be described in detail here.

[0064] Furthermore, such as Figure 4 As shown, the first trajectory acquisition module 302 in the gait analysis device provided in this embodiment includes:

[0065] The image fusion submodule 3021 is used to fuse at least two synchronized temporal initial depth images to obtain a temporal fused depth image;

[0066] The gait image extraction submodule 3022 is connected to the image fusion submodule and is used to extract the temporal gait depth image of the person under test from the temporal fusion depth image;

[0067] The trajectory acquisition submodule 3023 is connected to the gait image extraction submodule. It is used to process the preset original skeleton model based on the temporal gait depth image of the person under test to obtain the skeleton of the person under test and the first motion trajectory of each feature point of the skeleton.

[0068] In this embodiment, the process of obtaining the first motion trajectory through the above sub-modules is similar to that provided in Embodiment 1 of the present invention, and will not be described in detail here.

[0069] Furthermore, such as Figure 5As shown, the gait analysis module 305 in the gait analysis device provided in this embodiment includes:

[0070] The spatiotemporal parameter acquisition submodule 3051 is used to perform gait segmentation on the person under test based on the second motion trajectory of each joint of the musculoskeletal model under test, and obtain the average spatiotemporal parameters.

[0071] The calculation submodule 3052 is used to drive the musculoskeletal model under test to perform inverse dynamics calculation based on the second motion trajectory of each joint of the musculoskeletal model under test, and then obtain the corresponding average values ​​to obtain the average dynamic parameters and average kinematic parameters.

[0072] In this embodiment, the gait analysis module implements the gait analysis process through the above-mentioned sub-modules, which is similar to that provided in Embodiment 1 of the present invention, and will not be described in detail here.

[0073] In summary, the technical solution provided by this invention involves simultaneously acquiring temporal initial depth images of the gait movement of a person under test using at least two depth image acquisition devices. After processing and driving tracking optimization using the original skeleton model and original musculoskeletal model, gait analysis is performed to obtain the gait analysis results of the person under test. The technical solution provided by this invention reduces costs by requiring only at least two depth image acquisition devices and eliminating the need for reflective markers, simplifying operation and maintenance. It allows for convenient measurement and analysis anytime, anywhere. Furthermore, it eliminates the need for markers / sensors and other equipment, allowing for the acquisition of the test subject's natural gait. It also provides gait parameters such as average dynamic parameters, average kinematic parameters, and average spatiotemporal parameters, enabling complete gait analysis.

[0074] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A gait analysis method, characterized in that, include: Acquire temporal initial depth images simultaneously acquired by at least two depth image acquisition devices, wherein the temporal initial depth images are obtained by omnidirectional acquisition of the gait movement process of the person under test by the at least two depth image acquisition devices; The preset original skeleton model is processed based on the time-series initial depth image to obtain the skeleton of the person to be tested and the first motion trajectory of each feature point of the skeleton to be tested. The preset original musculoskeletal model is matched with the skeleton to be tested to obtain a musculoskeletal model that conforms to the body shape characteristics of the person to be tested. Based on the first motion trajectory of each feature point of the skeleton under test, the driving tracking optimization of the musculoskeletal model under test is performed to obtain the second motion trajectory of each joint point of the musculoskeletal model under test. Based on the second motion trajectory of each joint of the musculoskeletal model under test, gait analysis and calculation are performed on the person under test to obtain the average dynamic parameters, average kinematic parameters and average spatiotemporal parameters of the gait motion of the person under test; The average dynamic parameters include the average variation curves of triaxial ground reaction force and torque variation curves for each gait cycle, the average variation curves of joint contact force and torque variation curves for each gait cycle, and the average variation curves of muscle force and muscle activation variation curves for each gait cycle. The average kinematic parameters include the average curves of joint trajectory changes and the average curves of joint angle changes for each gait cycle. The average spatiotemporal parameters include average stride length, average gait speed, and average gait cycle. The step of processing the preset original skeleton model based on the temporal initial depth image includes: At least two synchronized initial temporal depth images are fused to obtain a temporally fused depth image; Extract the temporal gait depth image of the person under test from the temporal fusion depth image; The preset original skeleton model is processed based on the temporal gait depth image of the person under test to obtain the skeleton of the person under test and the first motion trajectory of each feature point of the skeleton of the person under test. The step of processing the preset original skeleton model based on the temporal gait depth image of the person under test includes: Unify the coordinate system of the temporal gait depth image of the person under test with the coordinate system of the preset original skeleton model to obtain a temporal unified depth image and a unified skeleton model; The unified skeleton model is matched and adjusted based on the initial frame image in the time-series unified depth image to obtain an initial skeleton model that matches the person to be tested. Based on the initial skeleton model and all temporal unified depth images, iterative optimization is performed to obtain the skeleton of the person to be tested; Obtain the first motion trajectory of each feature point of the skeleton under test; The step of performing gait analysis and calculation on the subject based on the second motion trajectory of each joint of the musculoskeletal model to be tested includes: The gait of the person under test is divided according to the second motion trajectory of each joint of the musculoskeletal model under test, and the average spatiotemporal parameters are obtained. After driving the musculoskeletal model under test to perform inverse dynamics calculations based on the second motion trajectory of each joint, the corresponding average values ​​are obtained to obtain the average dynamic parameters and average kinematic parameters.

2. The gait analysis method according to claim 1, characterized in that, The step of acquiring the first motion trajectory of each feature point of the skeleton under test includes: Iterative optimization is performed based on the skeleton to be tested and all temporally unified depth images to obtain the first motion trajectory of each feature point of the skeleton to be tested; or, Based on all temporally unified depth images and pose maps in the preset pose standard training set, the first motion trajectory of each feature point of the skeleton under test is obtained.

3. The gait analysis method according to claim 1, characterized in that, Extracting the temporal gait depth image of the person under test from the temporal fusion depth image includes: Acquire temporal background depth images pre-acquired by the at least two depth image acquisition devices; The temporal gait depth image of the person under test is extracted from the temporal fused depth image based on the temporal background depth image using the background segmentation extraction method. The background segmentation and extraction methods include: background difference method, optical flow method, or adjacent frame difference method.

4. The gait analysis method according to claim 1, characterized in that, The method involves driving the musculoskeletal model under test to perform inverse dynamics calculations based on the second motion trajectory of each joint point, and then obtaining the corresponding average values ​​to obtain the average dynamic parameters and average kinematic parameters, including: Based on the feedback results of the preset contact points on the sole of the foot of the musculoskeletal model under test during the inverse dynamics calculation process, the triaxial ground reaction force and torque variation curves of each gait cycle are obtained. Establish the second motion trajectory of each joint of the musculoskeletal model under test and the mechanical equilibrium equation of the average triaxial ground reaction force and torque, and obtain the joint contact force and torque change curves of each gait cycle. Based on the muscle recruitment principle, a mechanical balance equation is established between the average joint contact force and torque and the muscle force, and the muscle force change curves of each muscle group in each gait cycle are obtained. Based on the muscle force change curves of each muscle group in each gait cycle and the preset maximum force value, obtain the muscle activation change curves for each gait cycle. The average dynamic parameters are obtained by acquiring the average curves corresponding to the triaxial ground reaction force and torque variation curves, the joint contact force and torque variation curves, and the muscle force and muscle activation variation curves of each muscle group in each gait cycle.

5. A gait analysis device, characterized in that, To implement the gait analysis method according to any one of claims 1-4, comprising: The image acquisition module is used to acquire temporal initial depth images synchronously acquired by at least two depth image acquisition devices. The temporal initial depth images are obtained by omnidirectional acquisition of the gait movement process of the person under test by the at least two depth image acquisition devices. The first trajectory acquisition module, connected to the image acquisition module, is used to process the preset original skeleton model according to the temporal initial depth image to obtain the skeleton of the person to be tested and the first motion trajectory of each feature point of the skeleton to be tested. The musculoskeletal matching module, connected to the first trajectory acquisition module, is used to match the preset original musculoskeletal model with the skeleton to be tested to obtain a musculoskeletal model that conforms to the body shape characteristics of the person to be tested. The trajectory optimization module is connected to the first trajectory acquisition module and the musculoskeletal matching module respectively, and is used to drive and track the musculoskeletal model under test according to the first motion trajectory of each feature point of the skeleton under test, and to obtain the second motion trajectory of each joint of the musculoskeletal model under test. The gait analysis module is connected to the image acquisition module and the trajectory optimization module respectively. It is used to perform gait analysis calculations on the test subject based on the second motion trajectory of each joint of the musculoskeletal model under test, and to obtain the average dynamic parameters, average kinematic parameters and average spatiotemporal parameters of the test subject's gait motion. The average dynamic parameters include the average variation curves of triaxial ground reaction force and torque variation curves for each gait cycle, the average variation curves of joint contact force and torque variation curves for each gait cycle, and the average variation curves of muscle force and muscle activation variation curves for each gait cycle. The average kinematic parameters include the average curves of joint trajectory changes and the average curves of joint angle changes for each gait cycle. The average spatiotemporal parameters include average stride length, average gait speed, and average gait cycle.

6. The gait analysis device according to claim 5, characterized in that, The first trajectory acquisition module includes: The image fusion submodule is used to fuse at least two synchronized temporal initial depth images to obtain a temporal fused depth image; A gait image extraction submodule, connected to the image fusion submodule, is used to extract the temporal gait depth image of the person under test from the temporal fusion depth image; The trajectory acquisition submodule, connected to the gait image extraction submodule, is used to process the preset original skeleton model based on the temporal gait depth image of the person under test to obtain the skeleton of the person under test and the first motion trajectory of each feature point of the skeleton.

7. The gait analysis device according to claim 5, characterized in that, The gait analysis module includes: The spatiotemporal parameter acquisition submodule is used to perform gait segmentation on the person under test based on the second motion trajectory of each joint of the musculoskeletal model under test, and obtain the average spatiotemporal parameters. The calculation submodule is used to drive the musculoskeletal model under test to perform inverse dynamics calculations based on the second motion trajectory of each joint of the musculoskeletal model under test, and then obtain the corresponding average values ​​to obtain the average dynamic parameters and average kinematic parameters.

Images