Methods, systems, computer equipment, and storage media for inverse calculation of dynamic pressure distribution

Abstract

This invention relates to the field of digital biomechanical assessment technology, providing a method for inverse calculation of dynamic pressure distribution, including: S1: offline initialization, performing finite element discretization on a transparent flexible elastomer, constructing and preprocessing a global stiffness matrix; S2: online scanning and displacement boundary mapping, acquiring the initial height field of the elastomer under no-load conditions and the real-time height field of the bottom of the test part after the elastomer is pressed under load; S3: multi-layer medium refraction dynamic compensation, performing layered refraction correction on the laser beam path based on the surface normal vector of the elastomer after compression; S4: inverse finite element analysis (iFEA) pressure inverse calculation, inputting the corrected displacement field as boundary conditions into a preset mechanical model; S5: geometric-dynamic data fusion output, mapping high-precision three-dimensional geometric mesh and high-resolution pressure texture pixel by pixel. This simultaneously acquires high-precision three-dimensional foot morphology and sub-millimeter-level pressure data, solving the problems of scanning distortion and disconnection between shape and pressure data.

Description

Technical Field

[0001] This invention relates to the technical field of digital biomechanical assessment, and more particularly to an integrated three-dimensional plantar scanning and dynamic pressure distribution inverse calculation method and system. Background Technology

[0002] This invention belongs to the field of foot biomechanical testing and digital orthotics technology, specifically involving technical solutions related to three-dimensional plantar scanning and plantar pressure distribution detection. This technical field is a core technology supporting applications such as clinical diagnosis and treatment of foot and ankle diseases, customized orthopedic insoles, and early prevention and control of diabetic foot. Its technological development directly affects the design accuracy of foot orthotics products, the treatment effect of foot and ankle diseases, and the prevention and control level of chronic foot diseases such as diabetic foot. In the practical application of digital foot assessment and customized orthotics, achieving a deep integration of "realistic morphological acquisition" and "precise stress perception" of the human foot has long been a core technical challenge that needs to be overcome in this field. Existing mainstream solutions for this need all have insurmountable technical defects and cannot meet the actual application needs of clinical diagnosis and treatment and industrial-grade customized orthotics. The specific technical defects are described in detail below: 1. Foot geometric distortion caused by rigid support surfaces The closest existing technology to the present invention is a laser 3D scanning device based on a reinforced glass plane, such as the iQube series and Voxelcare series laser 3D scanners. Although this type of scanning solution can achieve the scanning and acquisition of the three-dimensional shape of the foot, it suffers from fundamental biomechanical and scanning accuracy problems due to design defects in the support surface, specifically in the following two aspects: 1.1 Significant deviations in biomechanical characteristics When the subject stands on the rigid glass support surface of this type of equipment, the boundary conditions between the sole of the foot and the support surface exhibit completely rigid characteristics, which is completely inconsistent with the support environment of the human foot in a natural shoe-wearing state. For subjects with special physiological characteristics of the foot, such as high arches and flexible flat feet, the arch of the foot is prone to compensatory structural collapse under its own weight due to the lack of close physical support. At the same time, the soft tissue of the foot will over-expand in all directions due to the reaction force of the rigid support, resulting in a significant deviation between the physiological morphology of the foot during the scanning process and its actual physiological state.

[0003] 1.2 Distortion model leads to failure of subsequent orthopedic design The rigid glass support plane cannot simulate the enveloping environment of a real shoe cavity, nor can it reproduce the fit, support, and enveloping feel of a custom insole. The 3D foot model obtained by scanning with such equipment is a "collapsed" model after the arch has collapsed. This model directly loses the real anatomical curves and physiological structural features of the arch, which is a severely morphologically distorted model. If personalized orthotic insoles are designed and manufactured based on such distorted 3D foot models, the resulting orthotic insoles will generally have problems such as poor fit to the human foot, misalignment of the arch support position, and incorrect matching of force points. They cannot provide precise orthotic support to the foot, ultimately rendering the orthotic insole completely ineffective, and may even cause additional biomechanical damage to the ankle due to support deviations.

[0004] 2. The decoupling problem between foot geometry and dynamic pressure data In existing technologies, another widely used foot detection solution involves using pressure pads that integrate resistive or capacitive sensors. This type of solution is mainly used for detecting plantar pressure distribution. However, when combined with digital foot assessment and customized orthotics, it has significant technical limitations. The core issue is the decoupling between foot geometry and dynamic pressure data. Specific problems are as follows: 2.1 "Seeing the force but not the form," lacking three-dimensional spatial morphological information These pressure pads can only acquire two-dimensional pressure distribution data on the sole contact surface through their integrated physical sensor array. They cannot capture and collect three-dimensional spatial morphology data of the lateral wall, dorsum, and complex arch of the foot. However, in practical applications such as personalized orthotic insole customization and foot and ankle disease diagnosis and treatment, the morphological characteristics of the lateral wall and dorsum of the foot and the three-dimensional spatial structure of the arch are the key basis for designing a fit-fitting support scheme and matching the stress characteristics of the foot. Relying solely on two-dimensional pressure data cannot meet the needs of high-precision digital assessment and orthotic design.

[0005] 2.2 Multiple factors combined cause alignment errors, making pixel-level stress-geometry mapping impossible. In practical applications, attempting to fuse independent pressure plate data obtained from pressure pads with 3D foot scans obtained from other devices to achieve a joint assessment of "morphology + pressure" results in severe interpolation errors due to inconsistencies in several key factors: First, inconsistent sampling frequencies—the inherent differences in signal acquisition frequencies between pressure sensors and laser scanning equipment lead to inaccurate matching of the time dimension of the data; second, inconsistent spatial coordinate systems—the independent reference coordinate systems of the pressure pad and the 3D scanning equipment lack a unified calibration benchmark; and third, inconsistent foot positions under pressure—the foot's posture and pressure level are difficult to completely unify in two separate testing processes. The combined effect of these three factors makes it impossible to eliminate interpolation errors after data fusion, ultimately preventing pixel-level stress-geometric precision mapping of the sole and hindering effective technical synergy between morphological and pressure data.

[0006] 3. The problem of physical sensor arrays obstructing the optical path of laser scanning To overcome the aforementioned problem of "decoupling of morphology and pressure data," some improvements have been attempted in existing technologies. Specifically, a pressure sensor film is placed over the scanning area of ​​a laser 3D scanner, attempting to integrate foot 3D morphology scanning and plantar pressure detection functions into the same device. However, this improvement not only failed to solve the original problem but also triggered new and more serious technical contradictions, directly leading to a significant decrease in the device's detection performance. The specific technical contradictions are as follows: 3.1 Severe optical interference blocks the laser vision pathway. Traditional flexible pressure sensors, such as FSR arrays, have a non-transparent structure. When such a sensor film is placed over a scanner, it directly blocks the optical vision path of the laser scanning sensor below, preventing the laser from penetrating the sensor film to perform three-dimensional scanning of the foot. In this case, the integrated device can only make a choice between two detection functions: either abandon the acquisition of the foot's three-dimensional shape and retain only the pressure detection function, or remove the sensor film and retain only the three-dimensional scanning function. It is impossible to achieve effective integration of the two functions.

[0007] 3.2 The detection resolution has a significant bottleneck, which can easily lead to missed detection of local stress concentration points. Physical pressure sensors have inherent technical limitations in contact density, typically ranging from 5 to 10 mm, while laser 3D scanning equipment can achieve sub-millimeter level point cloud detection resolution, representing an order-of-magnitude difference. Using such low-resolution physical sensors for plantar pressure detection cannot accurately capture subtle stress changes in the plantar surface, easily leading to missed detections of stress concentration points. Key detection points, such as early diabetic foot ulcers, are precisely located in areas of localized stress concentration. Missing these points directly results in the failure of early prevention and control of diabetic foot, posing serious clinical health risks to patients and failing to meet the technical requirements for refined diagnosis and treatment of foot and ankle diseases.

[0008] In summary, in the research and application of existing technologies, an effective solution that takes into account multiple technical requirements has yet to be formed to address the core needs of digital foot assessment and customized orthotics. Currently, there is a lack of an integrated technical solution that can simultaneously provide a shoe insole-like fit and physical support environment for the foot, non-destructively reproduce the full-field pressure distribution of the sole through purely optical means, and has the advantage of low cost. This has become a key technical bottleneck restricting the further development of foot biomechanical testing and digital orthotics technology. Summary of the Invention

[0009] To address the aforementioned problems, the present invention aims to provide a method and system for inverse calculation of dynamic pressure distribution. This invention can simultaneously acquire high-precision three-dimensional foot morphology and sub-millimeter-level pressure data, solving the problems of scanning distortion and disconnect between shape and pressure data. It offers low cost and real-time output, providing complete shape-force integrated data support for foot orthotics and clinical assessment.

[0010] The above-mentioned objective of this invention is achieved through the following technical solutions: The method for inverse calculation of dynamic pressure distribution includes the following steps: S1: Perform offline initialization, discretize the transparent flexible elastomer using finite element method, construct and preprocess the global stiffness matrix, and generate a pre-computation kernel to accelerate online calculation; S2: Perform online scanning and displacement boundary mapping, collect the initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed in under load conditions, and construct displacement boundary conditions. The test part is the part of the human body, including the feet, where pressure distribution data needs to be acquired. S3: Perform multi-layer medium refraction dynamic compensation. Based on the surface normal vector of the elastic body after compression, perform layered refraction correction on the laser beam path to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. S4: Perform inverse finite element analysis (iFEA) pressure back calculation, input the corrected displacement field as the boundary condition into the preset mechanical model, and use a fast solution algorithm to convert the three-dimensional deformation of the elastic body into pressure distribution data at the bottom of the part to be measured. S5: Performs geometric-dynamic data fusion output, mapping the high-precision 3D bottom geometric mesh of the test site with the high-resolution pressure texture pixel by pixel, generating an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis.

[0011] Furthermore, the hardware component of the 3D scanning device upon which the method relies includes: Base module: As the core of the system's underlying hardware, it integrates a wired laser emitter and a multi-camera vision sensor array, and is equipped with a precision motion mechanism. It is capable of performing high-precision rotational scanning or linear motion scanning tasks, and is responsible for transmitting structured light upwards and capturing the refracted optical path signal. The base module performs rotational scanning or linear motion scanning through the precision motion mechanism to adapt to three-dimensional data acquisition under different measurement ranges and spatial resolution requirements. Support plate: Located above the base, it is made of high-transmittance tempered glass and serves as the load-bearing reference surface for the entire system. It bears the entire weight of the human body while ensuring the non-destructive penetration of laser and visual signals. The thickness of the support plate is 12-15mm to achieve a balance between load-bearing stability and optical transmission performance. Transparent flexible elastomer layer: Tightly adhered and fixed to the surface of the support plate, serving as the core mechanical conversion interface; this layer uses a transparent addition-cured flexible elastic material with optimized material parameters, including an optimized Young's modulus of 0.78-2.50 MPa and a Shore hardness of 20-40 A, designed to accurately simulate the combined mechanical response of the plantar fat pad and typical shoe midsole materials, providing a soft support environment for the foot; the thickness of the transparent flexible elastomer layer is 5-20 mm; the transparent flexible elastomer material is transparent addition-cured silicone or PDMS, to balance compressive deformation response, light transmittance, and scanning reconstruction accuracy; Data processing system: Connected to the base module via a high-speed data interface, it has a built-in high-performance computing unit and integrates three core algorithm modules: a refractive index dynamic compensation algorithm for correcting optical path distortion, a high-precision point cloud reconstruction algorithm for restoring geometric morphology, and an inverse finite element analysis algorithm for calculating pressure from deformation; the data processing system is also used to register and fuse the scanned point cloud and pressure texture in a unified coordinate system to generate a fused model that simultaneously contains three-dimensional geometric morphology information and plantar pressure distribution information.

[0012] Further, in step S1, offline initialization is performed, the transparent flexible elastomer is discretized using finite element methods, a global stiffness matrix is ​​constructed and preprocessed, and a pre-computation kernel is generated to accelerate online calculations, specifically: Mesh generation is performed, and the transparent flexible elastomer region is discretized using finite element method, with the mesh size set to 1mm×1mm×1mm; Global stiffness matrix assembly is performed by constructing the material's elasticity matrix D based on the constitutive relation of linear elasticity, calculating the stiffness matrix of standard elements, and generating the global stiffness matrix through a vectorized assembly method. , where N is the total number of degrees of freedom of the system; Perform matrix preprocessing and storage, for Symmetry processing is performed, and the matrix is ​​pre-divided into blocks or Cholesky decomposition based on the fixed bottom surface constraint to generate and store pre-computation files, thereby significantly reducing the computing power consumption during online computation. In this system, the bottom nodes of the transparent flexible elastomer are set as fixed constraints to characterize the displacement restriction of the bottom boundary of the transparent flexible elastomer by the supporting plate; the global stiffness matrix is ​​divided into blocks after symmetry processing to pre-separate the matrix sub-blocks corresponding to the known boundary degrees of freedom and the internal degrees of freedom to be calculated; the global stiffness matrix is ​​decomposed by Cholesky and the decomposition factors are stored for direct use in the online stage for fast substitution and solution; the pre-computation kernel includes the stiffness matrix block results and / or Cholesky decomposition results.

[0013] Further, in step S2, online scanning and displacement boundary mapping are performed. The initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed in under load conditions are collected respectively to construct displacement boundary conditions. The test part includes the parts of the human body, including the feet, where pressure distribution data needs to be acquired. Specifically: No-load baseline data acquisition: Before the user stands up, in a no-load state, the laser passes through the glass and elastomer, and the acquisition terminal performs a full-field scan of the upper surface of the transparent flexible elastomer layer to obtain the initial height field of the elastomer surface. This data serves as the zero-point reference for subsequent displacement calculations; Weight-bearing state scan: The patient stands with both feet on the elastomer layer. The elastomer undergoes compression deformation, locally filling the arch area and providing upward support. The system simultaneously performs the following two acquisitions using laser technology: Geometric contour of foot skin: capturing 3D point clouds of the dorsum and lateral walls of the foot; Pressure surface displacement: Capturing the real-time surface height field formed after the sole of the foot is pressed into a transparent flexible elastomer layer. ; The initial height field serves as the zero-point reference for displacement calculation; the difference between the real-time surface height field and the initial height field constitutes the displacement boundary condition of the upper surface of the transparent flexible elastomer; during the scanning process under load, the foot outer contour point cloud and the height field of the pressure surface, which are collected simultaneously, establish a spatial correspondence based on the same scanning coordinate system; the displacement boundary condition is mapped to the nodes on the upper surface of the finite element model after coordinate registration, so as to serve as the boundary input for subsequent inverse finite element pressure back calculation.

[0014] Furthermore, in step S3, multi-layer medium refraction dynamic compensation is performed. Based on the surface normal vector of the elastic body after compression, the laser beam path is corrected by layered refraction to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. Specifically: As the surface of the elastomer is no longer smooth after being compressed, the deflection angle of light at the air-elastomer-glass interface changes with the spatial position. System Extraction The local gradient is calculated, and the normal vector at each scan point is calculated. Based on the surface normal vector of the elastic body obtained by real-time scanning, the spatial repositioning of millions of point cloud coordinates is performed point by point using Snell's law. The layered refraction correction formula is applied to perform real-time repositioning of point cloud coordinates, eliminating geometric ghosting caused by medium deformation and avoiding severe radial distortion of the scanned foot model due to refraction. The normal vector is calculated based on the local gradient of the surface of the transparent flexible elastomer under pressure to obtain the interface normal direction corresponding to each scanning point; the layered refraction correction is performed on the interface of the multi-layer medium composed of air, transparent flexible elastomer and supporting plate; the point cloud coordinates are repositioned point by point based on Snell's law to correct the optical path deflection error caused by the change of interface normal at different spatial positions; the correction result is used to eliminate geometric ghosting and radial distortion caused by the deformation of the flexible medium under pressure, and improve the spatial position accuracy of the point cloud and the accuracy of subsequent geometric reconstruction.

[0015] Further, in step S4, inverse finite element analysis (iFEA) pressure back calculation is performed. The corrected displacement field is used as the boundary condition input to the preset mechanical model. The three-dimensional deformation of the elastic body is converted into pressure distribution data at the bottom of the measured part through a fast solution algorithm. Specifically: Displacement field construction: Read the point cloud data corrected in step S3, perform coordinate centering calibration to align it with the coordinate system of the finite element model, map the unstructured scanned point cloud height data onto the upper surface nodes of the finite element mesh, and calculate the full-field displacement. The point cloud data is used as the boundary condition input parameter and imported into the preset finite element mechanical model to calculate the normal pressure intensity at each coordinate point and construct the displacement boundary condition vector U. After the point cloud data is corrected in step S3 and aligned with the coordinate system of the finite element model, a one-to-one correspondence is established between the scanning space and the mechanical solution space. The unstructured scanning point cloud height data is mapped to the nodes on the upper surface of the finite element mesh to form a discrete displacement boundary input that can be used for inverse finite element analysis. Perform a fast inverse finite element method: divide the system's degrees of freedom into boundary degrees of freedom with known displacements. and the internal degrees of freedom to be determined Based on the stiffness matrix pre-calculated in step S1, the blocks are divided. , Solve the equilibrium equations: ; Since K11 has been decomposed in step S1, the full-field displacement can be solved in milliseconds using the vectorized substitution method. ; Divide the upper surface of the stiffness matrix into blocks The calculated pressure is the pressure exerted by the sole of the foot on the elastic body. The system's degrees of freedom are divided into boundary degrees of freedom and internal degrees of freedom, corresponding to the region with known displacement and the region with unknown displacement, respectively. In the online stage, a fast substitution solution is performed based on the pre-decomposed matrix obtained in step S1 to reduce the computational load in the real-time back-calculation process. The normal pressure distribution at the bottom of the test part is obtained through the reaction force of the upper surface node to obtain the pressure texture information corresponding to the contact area of ​​the test part.

[0016] Further, in step S5, geometric-dynamic data fusion is performed, mapping the high-precision 3D bottom geometric mesh of the test site to the high-resolution pressure texture pixel by pixel to generate an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis. Specifically: The high-precision three-dimensional geometric mesh, including the dorsum, lateral wall, and sole of the foot, corrected in step S3, is mapped pixel-by-pixel with the high-resolution pressure texture calculated in step S4 to generate a fusion model that not only contains high-precision three-dimensional geometric information but also pixel-level pressure texture mapping. This model can be used for 3D printing or clinical diagnosis. The pressure texture and the three-dimensional geometric mesh are mapped based on a unified coordinate system to ensure a consistent correspondence between the pressure information and the surface spatial position of the tested area; the fusion model contains three-dimensional geometric information of the dorsum, lateral wall and sole of the foot; the sole texture pixels in the fusion model carry the corresponding pressure values ​​to form a visual representation that combines geometric shape and force distribution; the fusion model is used for 3D printing, orthopedic insole design or clinical evaluation.

[0017] A dynamic pressure distribution inverse calculation system for performing the dynamic pressure distribution inverse calculation method as described above, comprising: The offline initialization module is used to perform offline initialization, discretize the transparent flexible elastomer using finite element method, construct and preprocess the global stiffness matrix, and generate a pre-computation kernel for accelerating online calculations. The online scanning and mapping module is used to perform online scanning and displacement boundary mapping. It collects the initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed into it under load conditions, and constructs displacement boundary conditions. The test part is the part of the human body, including the feet, where pressure distribution data needs to be acquired. The refraction compensation module is used for dynamic compensation of refraction in multi-layer media. Based on the surface normal vector of the elastic body after compression, it performs layered refraction correction on the laser beam path to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. The iFEA pressure back calculation module is used to perform inverse finite element analysis. The iFEA pressure back calculation takes the corrected displacement field as the boundary condition input to the preset mechanical model, and converts the three-dimensional deformation of the elastic body into pressure distribution data at the bottom of the part to be measured through a fast solution algorithm. The geometric-dynamic fusion module is used to perform geometric-dynamic data fusion output, mapping the high-precision 3D bottom geometric mesh of the test site with the high-resolution pressure texture pixel by pixel, generating an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis.

[0018] A computer device includes a memory and one or more processors, the memory storing computer code that, when executed by the one or more processors, causes the one or more processors to perform the dynamic pressure distribution inverse calculation method as described above.

[0019] A computer-readable storage medium storing computer code that, when executed, performs an action such as the aforementioned inverse calculation method for dynamic pressure distribution.

[0020] Compared with existing technologies, this invention achieves efficient integration of three-dimensional plantar scanning and dynamic pressure distribution inverse calculation through an innovative integrated design of the physical layer and algorithm layer. This effectively solves technical problems in traditional foot detection technologies, such as weight-bearing geometric distortion, disconnect between morphology and pressure data, insufficient detection resolution, and high equipment costs. Simultaneously, it addresses the real-time requirements of clinical applications and possesses significant multi-dimensional technical benefits, as detailed below: (1) Accurately recreate the foot's support state within the shoe and obtain a non-collapsed foot morphology in a biomechanical sense. This invention designs a transparent and flexible composite interface that simulates the insole environment, precisely setting the Young's modulus of the elastomer layer to 0.78-2.50 MPa and the Shore hardness to 20-40 A. These parameters match the coupling stiffness of the plantar fat pad and typical shoe midsole materials (EVA / PU), allowing the elastomer layer (such as PDMS / addition-cured silicone) to provide a fit to special foot structures such as high arches. This makes the foot's weight-bearing scanning state closer to the physiological state of the actual support state within the shoe. The foot morphology obtained in this way is free from collapse distortion. When designing and customizing orthotic insoles based on this morphology, the amount of subsequent grinding correction can be greatly reduced.

[0021] (2) Breaking through the limitations of physical sensors, achieving sub-millimeter-level high-resolution pressure sensing, and realizing precise synchronous acquisition of foot geometry and pressure distribution. This invention constructs an inverse finite element pressure model based on displacement boundary conditions. The three-dimensional displacement vector ΔZ of the elastic body measured by laser is used as the displacement boundary condition of the finite element model. Foot pressure distribution data is generated by calculating the nodal reaction force or stress tensor of the mesh model through the material constitutive relation. This achieves a technological breakthrough of "visual sensor", making the resolution of pressure distribution completely aligned with the point cloud density of laser scanning, reaching the sub-millimeter level. It gets rid of the discrete limitations of traditional physical sensor arrays and realizes the essential leap from "seeing the shape" to "measuring pressure".

[0022] (3) Eliminate optical distortion introduced by flexible media and ensure the geometric accuracy of foot 3D scanning. In view of the problem of dynamic deflection of laser optical path caused by non-uniform indentation after the elastic body is compressed, this invention designs a real-time correction algorithm for laser refraction path driven by dynamic deformation field. Based on the surface normal vector of the elastic body obtained by real-time scanning, the spatial repositioning of millions of point cloud coordinates is performed point by point through Snell's Law, which effectively solves the problem of "geometric ghosting" caused by flexible media, avoids serious radial distortion of the scanned foot model, and ensures the spatial position accuracy of laser point cloud.

[0023] (4) Achieve sub-second real-time solution of scan data and realize the clinical application experience of "scan and get". This invention designs a linearization acceleration scheme for real-time output. Taking advantage of the invariant geometric structure of the elastic body, the global stiffness matrix K is pre-constructed in the offline stage. The inverse operator or decomposition factor of the stiffness matrix is ​​pre-calculated and stored through matrix decomposition techniques such as Cholesky decomposition or LU decomposition. In actual online scanning, the traditional iterative solution or complete matrix inversion operation is abandoned. Only the real-time displacement boundary conditions are substituted into the pre-decomposed matrix equation. The nodal reaction force is quickly solved by forward / backward substitution method. The calculation process that originally required several minutes of high-performance workstation is compressed to sub-second response, which meets the actual needs of clinical or store real-time model taking.

[0024] (5) Low-cost integration of vision and force perception without relying on expensive hardware. The invention realizes the inverse calculation of plantar pressure and the processing of scanning data through software algorithms throughout the process. It does not require the configuration of expensive resistive / capacitive pressure sensing chips. The detection function of "vision and force perception integration" is realized by simply using optical scanning combined with algorithms such as inverse finite element analysis and refraction correction. This significantly reduces the manufacturing cost of high-end foot biomechanical assessment equipment. At the same time, it avoids the technical contradiction of traditional physical sensor arrays blocking the optical path and takes into account both the stability of optical scanning and the accuracy of pressure detection. Attached Figure Description

[0025] Figure 1 This is a flowchart of the integrated three-dimensional plantar scanning and dynamic pressure distribution inverse calculation method of the present invention; Figure 2 This is a schematic diagram of the hardware system structure of the present invention; Figure 3 A finite element mesh model of an elastic body with displacement boundary conditions applied to it according to the present invention; Figure 4 This is a diagram of plantar pressure distribution based on inverse finite element analysis, as presented in this invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0028] This invention discloses an integrated three-dimensional plantar scanning and dynamic pressure distribution inverse calculation method and system. This technical solution is a cross-disciplinary digital biomechanical assessment platform. Its purpose is to simultaneously acquire high-precision three-dimensional anatomical morphology of the human foot and high-resolution plantar pressure distribution field through a single optical scanning process.

[0029] This solution achieves integration in the following three dimensions at the product architecture and core capability levels: (1) Physical layer: Adaptive flexible sensing interface Unlike traditional scanners that use rigid, reinforced glass support surfaces, this adaptive flexible sensing interface introduces a layer of transparent, flexible elastomer (such as PDMS or specialized silicone) into the optical path. This component functions as both a load-bearing interface and a physical "simulator." Through its own elastic deformation (Young's modulus ranging from 0.78 to 2.50 MPa), it simulates the physical support environment of a real insole for the arch of the foot, allowing the foot to be in an anatomical position closer to its natural physiological and shoe-wearing state during weight-bearing scanning.

[0030] (2) Algorithm layer: Visual-mechanical integrated inverse calculation engine The core of this vision-force integrated inverse calculation engine system is a calculation model based on inverse finite element analysis (iFEA). This model abandons traditional resistive or capacitive physical sensors and directly uses the three-dimensional displacement field of the elastic body obtained by laser scanning as the driving input. Combined with the linear elasticity model, it converts the deformation captured by vision into a precise plantar pressure distribution in real time, achieving sub-millimeter pressure sensing resolution.

[0031] (3) Data layer: Geometry-dynamics deep fusion model The final output of this system, a deep fusion model of geometry and dynamics, is a multi-dimensional digital image of the foot. This model not only includes a geometric mesh of the foot composed of point clouds, but also encompasses the dorsum, lateral walls, and plantar curves. Furthermore, the plantar texture pixels in this model carry physically meaningful pressure values. This "form-force integration" data structure provides complete underlying data support for subsequent automated CAD design of orthotic insoles and deep tissue stress risk assessment of diabetic foot.

[0032] Overall, this invention is an intelligent scanning device that achieves tactile perception through purely optical means. By introducing the technical means of "flexible medium + reverse simulation", it effectively solves the two major technical pain points that have long existed in the industry: "weight-bearing geometric distortion" and "disconnect between morphology / pressure data".

[0033] Furthermore, to verify the progressiveness of this invention, the performance parameters of this invention are compared with those of existing mainstream technologies as follows: (1) In terms of dimensional capabilities, traditional flatbed laser scanners can only acquire 3D morphological data and cannot provide pressure information; traditional pressure pads (FSR arrays) can only acquire 2D pressure data and lack morphological information; while the present invention (elastomer iFEA) can achieve integrated acquisition of 3D morphology and 3D pressure, forming a multi-dimensional data output of "form and force in one".

[0034] (2) Regarding the arch support mode, the traditional flatbed laser scanner uses rigid contact support, which is prone to arch collapse; the traditional pressure pad (FSR array) support mode is flexible / rigid selectable; the present invention uses adaptive flexible support, which can simulate the real shoe internal environment and avoid arch shape distortion.

[0035] (3) In terms of sampling resolution, the resolution of traditional flatbed laser scanners is 0.2-0.5mm, and the resolution of traditional pressure pads (FSR arrays) is only 5.0-10.0mm. This invention can maintain a sub-millimeter level sampling resolution of 0.2-0.5mm consistent with laser scanning, and achieve high-precision detection.

[0036] (4) In terms of optical stability, traditional flatbed laser scanners have extremely high optical stability; traditional pressure pads (FSR arrays) have poor optical stability due to sensor obstruction; the present invention can achieve high optical stability through refraction compensation algorithm.

[0037] (5) In terms of equipment cost, traditional flatbed laser scanners have moderate costs and traditional pressure pads (FSR arrays) have high costs; this invention relies on pure algorithm empowerment and does not need to rely on high-cost physical sensors, so the equipment cost is low.

[0038] (6) In terms of clinical data consistency, traditional flatbed laser scanners have low data consistency due to foot arch collapse, especially for patients with high arches; traditional pressure pads (FSR arrays) have moderate data consistency; this invention can achieve extremely high clinical data consistency by simulating the microenvironment inside the shoe.

[0039] The following is an illustration through specific examples: First Embodiment like Figure 1 As shown, this embodiment provides an integrated three-dimensional plantar scanning and dynamic pressure distribution inverse calculation method. First, let's introduce the hardware components used in this embodiment: like Figure 2 As shown, the hardware consists of four parts: a base module 101, a support plate 102, a transparent flexible elastomer layer 103, and a data processing system 104. Specifically, it includes: Base module 101: As the core of the system's underlying hardware, it integrates a wired laser emitter and a multi-camera vision sensor array, and is equipped with a precision motion mechanism. It is capable of performing high-precision rotational scanning or linear motion scanning tasks, and is responsible for transmitting structured light upwards and capturing the refracted optical path signal. The base module performs rotational scanning or linear motion scanning through the precision motion mechanism to adapt to three-dimensional data acquisition under different measurement ranges and spatial resolution requirements. Support plate 102: Located above the base, it is made of high-transmittance tempered glass with a thickness of 12-15mm. It serves as the load-bearing reference surface for the entire system, bearing the entire weight of the human body while ensuring the non-destructive penetration of laser and visual signals. The thickness of the support plate is 12-15mm to achieve a balance between load-bearing stability and optical transmission performance. Transparent flexible elastomer layer 103: Tightly adhered and fixed to the surface of the support plate, serving as the core mechanical conversion interface; this layer uses a transparent addition-cured flexible elastic material, such as transparent addition-cured silicone PDMS, with a thickness set at 5-20mm (preferably 10mm in this embodiment). The material parameters have been optimized, with the Young's modulus optimized to 0.78-2.50MPa and the Shore hardness to 20-40A, aiming to accurately simulate the combined mechanical response of the plantar fat pad and typical shoe midsole materials, providing a soft support environment for the foot; the thickness of the transparent flexible elastomer layer is 5-20mm; the transparent flexible elastomer material is transparent addition-cured silicone or PDMS, to balance compressive deformation response, light transmittance, and scanning reconstruction accuracy; Data processing system 104: Connects to the base module via a high-speed data interface, has a built-in high-performance computing unit, and integrates three core algorithm modules: a refractive index dynamic compensation algorithm for correcting optical path distortion, a high-precision point cloud reconstruction algorithm for restoring geometric morphology, and an inverse finite element analysis algorithm for calculating pressure from deformation; the data processing system is also used to register and fuse the scanned point cloud and pressure texture in a unified coordinate system to generate a fused model that simultaneously contains three-dimensional geometric morphology information and plantar pressure distribution information.

[0040] The method specifically includes the following steps: S1: Perform offline initialization, discretize the transparent flexible elastomer using finite element method, construct and preprocess the global stiffness matrix, and generate a pre-computation kernel for accelerating online calculation.

[0041] In this embodiment, step S1 specifically includes: Mesh generation is performed, and the transparent flexible elastomer region is discretized using finite element method, with the mesh size set to 1mm×1mm×1mm; Global stiffness matrix assembly is performed by constructing the material's elasticity matrix D based on the constitutive relation of linear elasticity, calculating the stiffness matrix of standard elements, and generating the global stiffness matrix through a vectorized assembly method. , where N is the total number of degrees of freedom of the system; Perform matrix preprocessing and storage, for Symmetry processing is performed, and the matrix is ​​pre-divided into blocks or Cholesky decomposition based on the fixed bottom surface constraint to generate and store pre-computation files, thereby significantly reducing the computing power consumption during online computation. In this system, the bottom nodes of the transparent flexible elastomer are set as fixed constraints to characterize the displacement restriction of the bottom boundary of the transparent flexible elastomer by the supporting plate; the global stiffness matrix is ​​divided into blocks after symmetry processing to pre-separate the matrix sub-blocks corresponding to the known boundary degrees of freedom and the internal degrees of freedom to be calculated; the global stiffness matrix is ​​decomposed by Cholesky and the decomposition factors are stored for direct use in the online stage for fast substitution and solution; the pre-computation kernel includes the stiffness matrix block results and / or Cholesky decomposition results.

[0042] Offline initialization is a "pre-calculation preparation" performed by this system before the formal scanning. Its core is to transform the physical properties and structural information of the transparent flexible elastomer into a quickly callable computational model in advance. Specifically, the elastomer region is first divided into a large number of 1mm×1mm×1mm micro-grid elements, turning the complex elastomer structure into a set of computable discrete elements. Then, based on the elastic mechanical properties of the material, the force relationships of these small elements are integrated to form a global stiffness matrix describing the stiffness characteristics of the entire elastomer. Finally, this matrix is ​​symmetricized and decomposed, completing and saving the complex matrix operations in advance. This allows the system to directly call these pre-calculated results during subsequent actual scanning, significantly reducing the workload of real-time calculations and making the pressure back-calculation process more efficient.

[0043] S2: Perform online scanning and displacement boundary mapping, and collect the initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed in under load conditions to construct displacement boundary conditions. The test part includes the parts of the human body, including the feet, where pressure distribution data needs to be acquired.

[0044] In this embodiment, step S2 specifically includes: No-load baseline data acquisition: Before the user stands up, in a no-load state, the laser passes through the glass and elastomer, and the acquisition terminal performs a full-field scan of the upper surface of the transparent flexible elastomer layer to obtain the initial height field of the elastomer surface. This data serves as the zero-point reference for subsequent displacement calculations; Weight-bearing state scan: The patient stands with both feet on the elastomer layer. The elastomer undergoes compression deformation, locally filling the arch area and providing upward support. The system simultaneously performs the following two acquisitions using laser technology: Geometric contour of foot skin: capturing 3D point clouds of the dorsum and lateral walls of the foot; Pressure surface displacement: Capturing the real-time surface height field formed after the sole of the foot is pressed into a transparent flexible elastomer layer. ; The initial height field serves as the zero-point reference for displacement calculation; the difference between the real-time surface height field and the initial height field constitutes the displacement boundary condition of the upper surface of the transparent flexible elastomer; during the scanning process under load, the foot outer contour point cloud and the height field of the pressure surface, which are collected simultaneously, establish a spatial correspondence based on the same scanning coordinate system; the displacement boundary condition is mapped to the nodes on the upper surface of the finite element model after coordinate registration, so as to serve as the boundary input for subsequent inverse finite element pressure back calculation.

[0045] Online scanning and displacement boundary mapping are crucial steps in preparing core displacement data for subsequent pressure inverse calculation. First, with the user neither standing nor under load, a full-field scan of the upper surface of the elastomer layer is performed to obtain the initial height of the elastomer surface; this data serves as the zero-point reference for subsequent displacement calculations. When the patient stands on the elastomer layer, the elastomer deforms under pressure, and the system simultaneously completes two acquisitions: first, capturing the three-dimensional geometric contours of the foot, including the dorsum and lateral walls; and second, recording the real-time surface height formed after the sole of the foot is pressed into the elastomer. By comparing the initial height with the real-time height after compression, the displacement change of the elastomer can be obtained, thereby constructing the displacement boundary conditions required for pressure inverse calculation.

[0046] S3: Perform multi-layer medium refraction dynamic compensation. Based on the surface normal vector of the elastic body after compression, perform layered refraction correction on the laser beam path to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud.

[0047] In this embodiment, step S3 specifically includes: As the surface of the elastomer is no longer smooth after being compressed, the deflection angle of light at the air-elastomer-glass interface changes with the spatial position. System Extraction The local gradient is calculated, and the normal vector at each scan point is calculated. Based on the surface normal vector of the elastic body obtained by real-time scanning, the spatial repositioning of millions of point cloud coordinates is performed point by point using Snell's law. The layered refraction correction formula is applied to perform real-time repositioning of point cloud coordinates, eliminating geometric ghosting caused by medium deformation and avoiding severe radial distortion of the scanned foot model due to refraction. The normal vector is calculated based on the local gradient of the surface of the transparent flexible elastomer under pressure to obtain the interface normal direction corresponding to each scanning point; the layered refraction correction is performed on the interface of the multi-layer medium composed of air, transparent flexible elastomer and supporting plate; the point cloud coordinates are repositioned point by point based on Snell's law to correct the optical path deflection error caused by the change of interface normal at different spatial positions; the correction result is used to eliminate geometric ghosting and radial distortion caused by the deformation of the flexible medium under pressure, and improve the spatial position accuracy of the point cloud and the accuracy of subsequent geometric reconstruction.

[0048] Multilayer media refraction dynamic compensation is used to correct scanning errors caused by light refraction after the elastomer is compressed. When the elastomer is compressed and deformed, the surface is no longer flat, and the refraction angle of light between air, the elastomer, and glass changes with position, which can easily cause "geometric ghosting" or even radial distortion in the scanned foot model. This step first extracts the local gradient of the surface of the compressed elastomer, calculates the normal vector of each scanning point, and then repositions all point cloud coordinates point by point based on these normal vectors. By correcting the refraction through layers, the positional deviation caused by the deformation of the medium is eliminated, ensuring that the spatial position of the three-dimensional point cloud of the foot obtained by scanning is accurate and reliable.

[0049] S4: Perform inverse finite element analysis (iFEA) pressure back calculation, input the corrected displacement field as the boundary condition into the preset mechanical model, and use a fast solution algorithm to convert the three-dimensional deformation of the elastic body into pressure distribution data at the bottom of the part to be measured.

[0050] In this embodiment, step S4 specifically includes: Displacement field construction: Read the point cloud data corrected in step S3, perform coordinate centering calibration to align it with the coordinate system of the finite element model, map the unstructured scanned point cloud height data onto the upper surface nodes of the finite element mesh, and calculate the full-field displacement. The point cloud data is used as the boundary condition input parameter and imported into the preset finite element mechanical model to calculate the normal pressure intensity at each coordinate point and construct the displacement boundary condition vector U. After the point cloud data is corrected in step S3 and aligned with the coordinate system of the finite element model, a one-to-one correspondence is established between the scanning space and the mechanical solution space. The unstructured scanning point cloud height data is mapped to the nodes on the upper surface of the finite element mesh to form a discrete displacement boundary input that can be used for inverse finite element analysis. like Figure 3 As shown in the figure, the system algorithm accurately maps the plantar depth data (Z-axis displacement) obtained from 3D scanning onto the nodes on the upper surface of the finite element mesh of the elastic body. The concave regions of the mesh intuitively reflect the displacement field distribution during plantar contact.

[0051] Perform a fast inverse finite element method: divide the system's degrees of freedom into boundary degrees of freedom with known displacements. and the internal degrees of freedom to be determined Based on the stiffness matrix pre-calculated in step S1, the blocks are divided. , Solve the equilibrium equations: ; Since K11 has been decomposed in step S1, the full-field displacement can be solved in milliseconds using the vectorized substitution method. ; Divide the upper surface of the stiffness matrix into blocks The calculated pressure is the pressure exerted by the sole of the foot on the elastic body. The system's degrees of freedom are divided into boundary degrees of freedom and internal degrees of freedom, corresponding to the region with known displacement and the region with unknown displacement, respectively. In the online stage, a fast substitution solution is performed based on the pre-decomposed matrix obtained in step S1 to reduce the computational load in the real-time back-calculation process. The normal pressure distribution at the bottom of the test part is obtained through the reaction force of the upper surface node to obtain the pressure texture information corresponding to the contact area of ​​the test part.

[0052] Inverse finite element analysis (iFEA) is the core step in converting elastic body deformation into plantar pressure data. First, the point cloud data, after refraction correction, is calibrated to align with the coordinate system of the finite element model. Then, the height changes obtained from the scan are mapped onto the upper surface nodes of the finite element mesh to calculate the displacement field of the elastic body, which serves as the boundary condition for the pressure calculation. Next, using pre-calculated stiffness matrix blocks from the offline stage, the system's degrees of freedom are divided into boundary parts with known displacements and internal parts to be determined. A fast solution method is used to solve for the internal displacements within milliseconds. Finally, based on the upper surface blocks of the stiffness matrix, the pressure exerted by the foot on the elastic body is calculated, thus obtaining accurate plantar pressure distribution data.

[0053] S5: Performs geometric-dynamic data fusion output, mapping the high-precision 3D bottom geometric mesh of the test site with the high-resolution pressure texture pixel by pixel, generating an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis.

[0054] In this embodiment, step S5 specifically includes: The high-precision three-dimensional geometric mesh, including the dorsum, lateral wall, and sole of the foot, corrected in step S3, is mapped pixel-by-pixel with the high-resolution pressure texture calculated in step S4 to generate a fusion model that not only contains high-precision three-dimensional geometric information but also pixel-level pressure texture mapping. This model can be used for 3D printing or clinical diagnosis. The pressure texture and the three-dimensional geometric mesh are mapped based on a unified coordinate system to ensure a consistent correspondence between the pressure information and the surface spatial position of the tested area; the fusion model contains three-dimensional geometric information of the dorsum, lateral wall and sole of the foot; the sole texture pixels in the fusion model carry the corresponding pressure values ​​to form a visual representation that combines geometric shape and force distribution; the fusion model is used for 3D printing, orthopedic insole design or clinical evaluation.

[0055] like Figure 4 As shown in the figure, the system's calculation output is displayed. On the XY plane, the normal pressure intensity of different anatomical regions of the foot is visualized using grayscale mapping. The right-hand scale indicates the correspondence between pressure values ​​and grayscale levels.

[0056] The geometry-dynamics data fusion output is the final step in integrating the results of all previous steps into a usable final model. This step involves mapping the high-precision 3D geometric mesh, which includes the dorsum, lateral walls, and sole of the foot and has undergone refraction correction, to a pixel-by-pixel high-resolution pressure texture obtained from inverse finite element analysis. This ensures that each geometric point carries precise pressure numerical information, generating a fused model that integrates form and force. This model preserves the true 3D morphology of the foot and clearly presents the pressure distribution in different areas of the sole. It can be directly used for 3D printing custom orthotic insoles and can also provide intuitive and accurate biomechanical data references for clinical diagnosis.

[0057] Second Embodiment To further illustrate the practical application scenarios and technical effects of the present invention, this embodiment takes the customized production of high-arch orthotic insoles as an example to explain in detail the specific application process of this integrated three-dimensional plantar scanning and dynamic pressure distribution inverse calculation system, as follows: Physical environment simulation: A transparent flexible elastomer with a thickness of 10mm and a Shore hardness of 20A was selected as the scanning support medium. When a patient with high arches stands on the surface of the elastomer layer, the arch area of ​​the foot, which is suspended on the rigid support surface, will press into the elastomer layer. The elastomer layer undergoes elastic deformation under pressure and forms an upward support force, keeping the patient's arch in a "subtalar neutral position", thus simulating the real support environment inside the shoe.

[0058] Submillimeter-level deformation and pressure sensing: The scanning system of this invention captures the full-field compression deformation of the elastomer layer with a spatial resolution of 0.2mm. At the same time, it uses the inverse finite element analysis (iFEA) algorithm to back-calculate the plantar pressure distribution and accurately identify the stress concentration point at the first metatarsal head of the patient's foot. The pressure value at this stress concentration point is 350kPa.

[0059] Automated design of orthopedic insoles: The foot geometry-dynamics fusion model scanned and generated by this system is directly imported into the CAD design software for orthopedic insoles. Based on the plantar pressure cloud map in the model, the software automatically performs the "pressure relief cavity" structure design at the insole design position corresponding to the first metatarsal head. At the same time, in the arch support area, the corresponding filling stiffness is matched according to the plantar support force data calculated by the system, and the automated design of the support structure is completed.

[0060] Application effect verification: The customized high arch orthotic insoles based on the technical solution of this invention have a fit to the patient's foot that is more than 30% better than traditional customization methods. This significantly reduces the manual grinding process in the later stages of orthotic insole production and greatly improves the customization efficiency and fitting accuracy of orthotic insoles.

[0061] Third Embodiment The core of this invention is to achieve three-dimensional foot scanning and inverse calculation of dynamic pressure distribution based on "transparent flexible medium + laser scanning + inverse finite element analysis". Based on the same inventive concept, two types of positive alternative implementation schemes can be derived. At the same time, there are potential bypass schemes in the field that target the core technology of this invention. Related entities usually try to break the technical combination of "transparent elastomer + laser scanning" and bypass the inverse finite element analysis (iFEA) algorithm of this invention through physical or chemical means. This embodiment will describe the positive alternative schemes and potential technical bypass schemes in detail.

[0062] I. Positive Alternative Solution to the Invention The solutions in this section are all based on the core inventive concept of this invention and are different implementations of the same technical idea, specifically including the following two: (a) Back-calculation scheme for structured feature points (marked points) Specific implementation: Within the transparent elastomer layer of the present invention, tiny nano-fluorescent dots or black micro-grid dots that do not affect the penetration of lasers are uniformly embedded. Implementation principle: The scanning device no longer directly calculates the continuous displacement of the surface of the elastomer layer, but tracks the spatial displacement vector of the above-mentioned preset "marker points" through a high-speed camera, and uses the discrete point displacement as the core calculation basis; Technical advantages: It significantly reduces the dependence on the accuracy of laser scanning point cloud reconstruction. By acquiring the displacement of discrete marker points and combining it with the Kriging interpolation algorithm, the full-field strain of the elastic layer can be quickly calculated, simplifying the calculation logic of displacement detection.

[0063] (II) Pneumatic / Hydraulic Compensated Transparent Interface Solution Specific implementation: The solid transparent flexible elastomer layer in this invention is replaced with a sealed transparent soft capsule, and the soft capsule is filled with transparent silicone oil or high-pressure gas. Implementation principle: By monitoring the overall pressure change inside the soft bladder and combining the volume data of the foot pressed into the soft bladder obtained by laser optical scanning, the local contact pressure at the contact point between the sole of the foot and the soft bladder is calculated using the hydrostatic equation. Technical advantages: The soft bladder can provide more uniform support to the foot, and the refractive index of the fluid inside is relatively constant, eliminating the need for complex dynamic refraction correction and greatly reducing the difficulty of designing optical path correction algorithms.

[0064] II. Potential technical bypass solutions for this invention Such solutions are not a positive replacement for the core concept of this invention, but rather bypass methods designed by relevant parties to circumvent the patent protection of this invention, specifically targeting the technical combination of "transparent elastomer + laser scanning" and the iFEA algorithm. These methods include the following three types: (a) Vision-tactile film (pressure-sensitive color-changing material) bypass solution Bypassing the approach: Instead of the transparent flexible elastomer layer design of this invention, a special transparent pressure-sensitive film, such as a film made of mechanical coloring material, is directly laid above the glass support surface of the scanning device; Technical details: When the transparent pressure-sensitive film is subjected to foot pressure, it will undergo a chemical-level color or polarization state change. The camera below the scanning device directly obtains the foot pressure distribution data by capturing the color distribution state of the film, rather than detecting the three-dimensional displacement of the elastomer. Purpose of bypassing: By this method, the complex inverse finite element method (iFEA) calculation in this invention is bypassed, and the mechanical inverse calculation process of plantar pressure is simplified into a simple image processing process, thereby circumventing the patent protection related to the iFEA algorithm.

[0065] (II) Hybrid "hollowed-out" array sensor bypass scheme Bypassing the path: A thin-film pressure array is fabricated using extremely fine transparent conductive materials, typically silver nanowires or graphene meshes, to replace the sensorless optical detection design of this invention; Technical details: The thin-film pressure array itself has a light transmittance of over 90%, and can be directly attached to the surface of the glass support. The laser can directly pass through the sensor array to complete the three-dimensional morphological scanning of the foot, and at the same time, the sensor array collects foot pressure data. Bypassing the core concept of the "visual sensor" of this invention and returning to the pressure measurement method of traditional physical sensors, although the equipment manufacturing cost of this solution is higher, it can effectively circumvent the algorithm-related patents of this invention.

[0066] (III) Multimodal Deep Learning End-to-End Prediction (Finite Element Model-Free) Bypass Solution Bypassing the path: Instead of establishing the physical finite element equation based on linear elasticity in this invention, a data-driven pressure prediction model is constructed using deep neural networks, with CNN-Transformer being a typical neural network model; Technical details: A large amount of paired data of foot pressure depth maps and actual plantar pressure maps are collected in advance. The deep neural network is trained using this paired data. After training, the system only needs to input a foot pressure depth map to directly predict and output the corresponding plantar pressure cloud map. Bypassing the objective: If this invention specifically defines the "finite element physical model" or "constitutive equation", this solution can completely circumvent the patent protection scope related to the physical formula through a purely data-driven AI training method.

[0067] Fourth embodiment This embodiment provides a dynamic pressure distribution inverse calculation system for performing the dynamic pressure distribution inverse calculation method as described in the first embodiment, including: The offline initialization module is used to perform offline initialization, discretize the transparent flexible elastomer using finite element method, construct and preprocess the global stiffness matrix, and generate a pre-computation kernel for accelerating online calculations. The online scanning and mapping module is used to perform online scanning and displacement boundary mapping. It collects the initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed into it under load conditions, and constructs displacement boundary conditions. The test part is the part of the human body, including the feet, where pressure distribution data needs to be acquired. The refraction compensation module is used for dynamic compensation of refraction in multi-layer media. Based on the surface normal vector of the elastic body after compression, it performs layered refraction correction on the laser beam path to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. The iFEA pressure back calculation module is used to perform pressure back calculation in inverse finite element analysis. It takes the corrected displacement field as the boundary condition input to the preset mechanical model and uses a fast solution algorithm to convert the three-dimensional deformation of the elastic body into pressure distribution data at the bottom of the part to be measured. The geometric-dynamic fusion module is used to perform geometric-dynamic data fusion output, mapping the high-precision 3D bottom geometric mesh of the test site with the high-resolution pressure texture pixel by pixel, generating an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis.

[0068] A computer-readable storage medium stores computer code that, when executed, performs the methods described above. Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. This program can be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0069] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

[0070] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0071] It should be noted that the above embodiments can be freely combined as needed. The above description is only a preferred embodiment of the present invention. It should be pointed out that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for inverse calculation of dynamic pressure distribution, characterized in that, Includes the following steps: S1: Perform offline initialization, discretize the transparent flexible elastomer using finite element method, construct and preprocess the global stiffness matrix, and generate a pre-computation kernel to accelerate online calculation; S2: Perform online scanning and displacement boundary mapping, collect the initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed in under load conditions, and construct displacement boundary conditions. The test part is the part of the human body, including the feet, where pressure distribution data needs to be acquired. S3: Perform multi-layer medium refraction dynamic compensation. Based on the surface normal vector of the elastic body after compression, perform layered refraction correction on the laser beam path to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. S4: Perform inverse finite element analysis (iFEA) pressure back calculation, input the corrected displacement field as the boundary condition into the preset mechanical model, and use a fast solution algorithm to convert the three-dimensional deformation of the elastic body into pressure distribution data at the bottom of the part to be measured. S5: Perform geometric-dynamic data fusion output, and perform pixel-by-pixel mapping between the high-precision 3D bottom geometric mesh of the test site and the high-resolution pressure texture to generate an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis. The hardware components of the 3D scanning equipment upon which the method relies include: Base module: As the core of the system's underlying hardware, it integrates a wired laser emitter and a multi-camera vision sensor array, and is equipped with a precision motion mechanism. It is capable of performing high-precision rotational scanning or linear motion scanning tasks, and is responsible for transmitting structured light upwards and capturing the refracted optical path signal. The base module performs rotational scanning or linear motion scanning through the precision motion mechanism to adapt to three-dimensional data acquisition under different measurement ranges and spatial resolution requirements. Support plate: Located above the base, it is made of high-transmittance tempered glass and serves as the load-bearing reference surface for the entire system. It bears the entire weight of the human body while ensuring the non-destructive penetration of laser and visual signals. The thickness of the support plate is 12-15mm to achieve a balance between load-bearing stability and optical transmission performance. Transparent flexible elastomer layer: Tightly adhered and fixed to the surface of the support plate, serving as the core mechanical conversion interface; this layer uses a transparent addition-cured flexible elastic material with optimized material parameters, including an optimized Young's modulus of 0.78-2.50 MPa and a Shore hardness of 20-40 A, designed to accurately simulate the combined mechanical response of the plantar fat pad and typical shoe midsole materials, providing a soft support environment for the foot; the thickness of the transparent flexible elastomer layer is 5-20 mm; the transparent flexible elastomer material is transparent addition-cured silicone or PDMS, to balance compressive deformation response, light transmittance, and scanning reconstruction accuracy; Data processing system: Connected to the base module via a high-speed data interface, it has a built-in high-performance computing unit and integrates three core algorithm modules: a dynamic refractive index compensation algorithm for correcting optical path distortion, a high-precision point cloud reconstruction algorithm for restoring geometric morphology, and an inverse finite element analysis algorithm for calculating pressure from deformation; the data processing system is also used to register and fuse the scanned point cloud and pressure texture in a unified coordinate system to generate a fused model that simultaneously contains three-dimensional geometric morphology information and plantar pressure distribution information; In step S4, inverse finite element analysis (iFEA) pressure back calculation is performed. The corrected displacement field is used as the boundary condition input to the preset mechanical model. The three-dimensional deformation of the elastic body is converted into pressure distribution data at the bottom of the measured part through a fast solution algorithm. Specifically: Displacement field construction: Read the point cloud data corrected in step S3, perform coordinate centering calibration to align it with the coordinate system of the finite element model, map the unstructured scanned point cloud height data onto the upper surface nodes of the finite element mesh, and calculate the full-field displacement. , Let be the initial height field of the elastic body surface. The real-time surface height field formed after the transparent flexible elastomer layer is pressed into the sole of the foot is used as the boundary condition input parameter and imported into the preset finite element mechanical model to calculate the normal pressure intensity at each coordinate point and construct the displacement boundary condition vector U. Among them, after the point cloud data corrected in step S3 is aligned with the coordinate system of the finite element model, a one-to-one correspondence between the scanning space and the mechanical solution space is established. The unstructured scanning point cloud height data is mapped to the upper surface nodes of the finite element mesh to form the discrete displacement boundary input for inverse finite element analysis. Perform a fast inverse finite element method: divide the system's degrees of freedom into boundary degrees of freedom with known displacements. and the internal degrees of freedom to be determined Based on the stiffness matrix pre-calculated in step S1, the blocks are divided. , Solve the equilibrium equations: ; Since K11 has been decomposed in step S1, the full-field displacement can be solved in milliseconds using the vectorized substitution method. ; Divide the upper surface of the stiffness matrix into blocks The calculated pressure is the pressure exerted by the sole of the foot on the elastic body. The system's degrees of freedom are divided into boundary degrees of freedom and internal degrees of freedom, corresponding to the region with known displacement and the region with unknown displacement, respectively. In the online stage, a fast substitution solution is performed based on the pre-decomposed matrix obtained in step S1 to reduce the computational load in the real-time back-calculation process. The normal pressure distribution at the bottom of the test part is obtained through the reaction force of the upper surface node to obtain the pressure texture information corresponding to the contact area of ​​the test part.

2. The method for inverse calculation of dynamic pressure distribution according to claim 1, characterized in that, In step S1, offline initialization is performed, the transparent flexible elastomer is discretized using the finite element method, the global stiffness matrix is ​​constructed and preprocessed, and a pre-computation kernel is generated to accelerate online calculations. Specifically: Mesh generation is performed, and the transparent flexible elastomer region is discretized using finite element method, with the mesh size set to 1mm×1mm×1mm; Global stiffness matrix assembly is performed by constructing the material's elasticity matrix D based on the constitutive relation of linear elasticity, calculating the stiffness matrix of standard elements, and generating the global stiffness matrix through a vectorized assembly method. , where N is the total number of degrees of freedom of the system; Perform matrix preprocessing and storage, for Symmetry processing is performed, and the matrix is ​​pre-divided into blocks or Cholesky decomposition based on the fixed bottom surface constraint to generate and store pre-computation files, thereby significantly reducing the computing power consumption during online computation. In this system, the bottom nodes of the transparent flexible elastomer are set as fixed constraints to characterize the displacement restriction of the bottom boundary of the transparent flexible elastomer by the supporting plate; the global stiffness matrix is ​​divided into blocks after symmetry processing to pre-separate the matrix sub-blocks corresponding to the known boundary degrees of freedom and the internal degrees of freedom to be calculated; the global stiffness matrix is ​​decomposed by Cholesky and the decomposition factors are stored for direct use in the online stage for fast substitution and solution; the pre-computation kernel includes the stiffness matrix block results and / or Cholesky decomposition results.

3. The method for inverse calculation of dynamic pressure distribution according to claim 1, characterized in that, In step S2, online scanning and displacement boundary mapping are performed. The initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed into it under load conditions are collected respectively to construct displacement boundary conditions. The test part includes the parts of the human body, including the feet, where pressure distribution data needs to be acquired. Specifically: No-load baseline data acquisition: Before the user stands up, in a no-load state, the laser passes through the glass and elastomer, and the acquisition terminal performs a full-field scan of the upper surface of the transparent flexible elastomer layer to obtain the initial height field of the elastomer surface. This data serves as the zero-point reference for subsequent displacement calculations; Weight-bearing state scan: The patient stands with both feet on the elastomer layer. The elastomer undergoes compression deformation, locally filling the arch area and providing upward support. The system simultaneously performs the following two acquisitions using laser technology: Geometric contour of foot skin: capturing 3D point clouds of the dorsum and lateral walls of the foot; Pressure surface displacement: Capturing the real-time surface height field formed after the sole of the foot is pressed into a transparent flexible elastomer layer. ; The initial height field serves as the zero-point reference for displacement calculation; the difference between the real-time surface height field and the initial height field constitutes the displacement boundary condition of the upper surface of the transparent flexible elastomer; during the scanning process under load, the foot outer contour point cloud and the height field of the pressure surface, which are collected simultaneously, establish a spatial correspondence based on the same scanning coordinate system; the displacement boundary condition is mapped to the nodes on the upper surface of the finite element model after coordinate registration, so as to serve as the boundary input for subsequent inverse finite element pressure back calculation.

4. The method for inverse calculation of dynamic pressure distribution according to claim 1, characterized in that, In step S3, multi-layer medium refraction dynamic compensation is performed. Based on the surface normal vector of the elastic body after compression, the laser beam path is corrected by layered refraction to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. Specifically: As the surface of the elastomer is no longer smooth after being compressed, the deflection angle of light at the air-elastomer-glass interface changes with the spatial position. System Extraction The local gradient is calculated, and the normal vector at each scan point is calculated. Based on the surface normal vector of the elastic body obtained by real-time scanning, the spatial repositioning of millions of point cloud coordinates is performed point by point using Snell's law. The layered refraction correction formula is applied to perform real-time repositioning of point cloud coordinates, eliminating geometric ghosting caused by medium deformation and avoiding severe radial distortion of the scanned foot model due to refraction. The normal vector is calculated based on the local gradient of the surface of the transparent flexible elastomer under pressure to obtain the interface normal direction corresponding to each scanning point; the layered refraction correction is performed on the interface of the multi-layer medium composed of air, transparent flexible elastomer and supporting plate; the point cloud coordinates are repositioned point by point based on Snell's law to correct the optical path deflection error caused by the change of interface normal at different spatial positions; the correction result is used to eliminate geometric ghosting and radial distortion caused by the deformation of the flexible medium under pressure, and improve the spatial position accuracy of the point cloud and the accuracy of subsequent geometric reconstruction.

5. The method for inverse calculation of dynamic pressure distribution according to claim 1, characterized in that, In step S5, geometric-dynamic data fusion is performed, mapping the high-precision 3D bottom geometric mesh of the test site to the high-resolution pressure texture pixel by pixel to generate an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis. Specifically: The high-precision three-dimensional geometric mesh, including the dorsum, lateral wall, and sole of the foot, corrected in step S3, is mapped pixel-by-pixel with the high-resolution pressure texture calculated in step S4 to generate a fusion model that not only contains high-precision three-dimensional geometric information but also pixel-level pressure texture mapping. This model can be used for 3D printing or clinical diagnosis. The pressure texture and the three-dimensional geometric mesh are mapped based on a unified coordinate system to ensure a consistent correspondence between the pressure information and the surface spatial position of the tested area; the fusion model contains three-dimensional geometric information of the dorsum, lateral wall and sole of the foot; the sole texture pixels in the fusion model carry the corresponding pressure values ​​to form a visual representation that combines geometric shape and force distribution; the fusion model is used for 3D printing, orthopedic insole design or clinical evaluation.

6. A dynamic pressure distribution inverse calculation system for performing the dynamic pressure distribution inverse calculation method as described in any one of claims 1-5, characterized in that, include: The offline initialization module is used to perform offline initialization, discretize the transparent flexible elastomer using finite element method, construct and preprocess the global stiffness matrix, and generate a pre-computation kernel for accelerating online calculations. The online scanning and mapping module is used to perform online scanning and displacement boundary mapping. It collects the initial height field of the elastic body under no-load conditions and the real-time height field of the bottom of the test part after the elastic body is pressed into it under load conditions, and constructs displacement boundary conditions. The test part is the part of the human body, including the feet, where pressure distribution data needs to be acquired. The refraction compensation module is used for dynamic compensation of refraction in multi-layer media. Based on the surface normal vector of the elastic body after compression, it performs layered refraction correction on the laser beam path to eliminate the geometric ghosting introduced by the deformation of the flexible medium and ensure the spatial position accuracy of the point cloud. The iFEA pressure back calculation module is used to perform inverse finite element analysis. The iFEA pressure back calculation takes the corrected displacement field as the boundary condition input to the preset mechanical model, and converts the three-dimensional deformation of the elastic body into pressure distribution data at the bottom of the part to be measured through a fast solution algorithm. The geometric-dynamic fusion module is used to perform geometric-dynamic data fusion output, mapping the high-precision 3D bottom geometric mesh of the test site with the high-resolution pressure texture pixel by pixel to generate an integrated bottom model of the test site that can be directly used for 3D printing or clinical diagnosis.

7. A computer device, characterized in that, It includes a memory and one or more processors, the memory storing computer code that, when executed by the one or more processors, causes the one or more processors to perform the dynamic pressure distribution inverse calculation method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer code, and when the computer code is executed, the dynamic pressure distribution inverse calculation method as described in any one of claims 1 to 5 is executed.

Images