Digital human model construction methods, electronic devices and computer-readable storage media
By registering human body temperature distribution data with the posture features of an initial 3D human body model, a human body thermal model is generated, solving the problem of difficulty in constructing an accurate digital human model in a natural state in existing technologies, and realizing accurate data acquisition without markers.
Patent Information
- Application Number
- CN202611021750.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-09
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies make it difficult to accurately construct digital human models for specific users in a natural state, and pasting markers can obscure acupoints, change temperature distribution, and affect physical condition.
By registering the acquired human body temperature distribution data with the posture features of the initial 3D human body model, a projection relationship is constructed to generate a human body thermal model, and then a digital human model is generated, thus avoiding the need to set marker points on the human body.
It enables accurate data collection of the human body in its natural state, and constructs a digital human model that does not require markers, thus avoiding any impact on the human body's condition.
Smart Images

Figure CN123118957A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical image processing, and more particularly to a method for constructing a digital human model, an electronic device, and a computer-readable storage medium. Background Technology
[0002] Currently, the digitization of TCM meridians mainly relies on standard human body models superimposed with static meridian line diagrams; however, this method cannot achieve model construction for specific users; in order to achieve model construction for specific users, existing technologies usually use the method of pasting markers on the human body surface and scanning the human body, and then determining the meridian direction of the human body based on the markers; however, pasting markers will block acupoints, change the human body temperature distribution, and even affect the user's physical condition, making it difficult to achieve accurate model construction under natural conditions. Summary of the Invention
[0003] The main objective of this invention is to propose a method for constructing a digital human model, an electronic device, and a computer-readable storage medium, aiming to solve the problem in the prior art of accurately constructing a model of the human body in a natural state.
[0004] To achieve the above objectives, the present invention provides a method for constructing a digital human model, the method comprising: Acquire human body temperature distribution data and an initial three-dimensional human body model of the target human body; Identify the first pose feature of the human body temperature distribution data and the second pose feature in the initial three-dimensional human body model; Based on the first posture features and the second posture features, the human body temperature distribution data is registered with the initial three-dimensional human body model to obtain the projection relationship between the human body temperature distribution data and the initial three-dimensional human body model. The temperature values in the human body temperature distribution data are mapped onto the initial three-dimensional human body model based on the projection relationship to obtain the human body thermal model; Human body part models are generated in the human body thermal model to obtain the digital human model corresponding to the target human body.
[0005] Optionally, the first posture feature includes a first joint point, and the second posture feature includes a second joint point; the step of registering the human body temperature distribution data with the initial three-dimensional human body model based on the first posture feature and the second posture feature to obtain the projection relationship between the human body temperature distribution data and the initial three-dimensional human body model includes: An initial projection matrix is constructed based on the first joint point and the second joint point, wherein the first joint point and the second joint point correspond one-to-one. An objective function is constructed based on the consistency difference between the human body temperature distribution data and the initial three-dimensional human body model; The projection relationship is obtained by iteratively optimizing the initial projection matrix using the objective function.
[0006] Optionally, constructing the initial projection matrix based on the first joint point and the second joint point includes: The intrinsic parameter matrix of the image acquisition device and the extrinsic parameter matrix of the image acquisition device and the 3D acquisition device are obtained. The human body temperature distribution data is acquired by the image acquisition device, and the initial 3D human body model is obtained by the point cloud data acquired by the 3D acquisition device. The image acquisition device and the 3D acquisition device simultaneously acquire data from the target human body. The device projection relationship is determined based on the intrinsic parameter matrix and the extrinsic parameter matrix; The initial projection matrix is obtained by matching the first joint point and the second joint point based on the device projection relationship.
[0007] Optionally, obtaining the projection relationship by iteratively optimizing the initial projection matrix using the objective function includes: In each iteration, the extrinsic parameter matrix is adjusted to generate a new initial projection matrix; After the iteration completion condition is met, the initial projection matrix is used as the projection relationship.
[0008] Optionally, the human body temperature distribution data is obtained from multiple infrared images; the objective function constructed based on the consistency difference between the human body temperature distribution data and the initial three-dimensional human body model includes: The initial three-dimensional human body model is projected onto a two-dimensional object based on the initial projection matrix to obtain a two-dimensional projection image, and the contour error between the two-dimensional projection image and the corresponding human body contour in the infrared image is determined. For the second joint point in the three-dimensional human body model, determine the projection temperature value corresponding to the projection of the second joint point on each infrared image; The temperature consistency error is obtained based on the differences between the projected temperature values. Determine the non-rigid deformation field corresponding to the three-dimensional human body model, and calculate the spatial gradient norm of the non-rigid deformation field; The objective function is constructed by combining the contour error, the temperature consistency error, and the spatial gradient norm.
[0009] Optionally, generating human body part models in the human body thermal model to obtain a digital human model corresponding to the target human body includes: Obtain standard meridian coordinates and project the standard meridian coordinates onto the surface of the human thermal model to obtain the initial meridian path; Determine the temperature gradient of the initial meridian path on the human thermal model; Determine the positional differences between the initial meridian path and the standard anatomical atlas, and the direction of the temperature gradient; The initial meridian path is optimized based on the location differences to obtain the target path; The digital human model is obtained by generating the corresponding organs based on the target path.
[0010] Optionally, determining the positional differences between the initial meridian path and the standard anatomical atlas, and the direction of the temperature gradient, includes: Identify multiple key acupoints on the initial meridian path, and determine the corresponding landmarks in the standard anatomical atlas; Determine the point distance between the key acupoint and the marker point; Calculate the cosine of the angle between the tangent direction of the path and the direction of the temperature gradient; A path optimization function is constructed by combining the distance between the points and the cosine of the included angle, and the path optimization function is used as the position difference.
[0011] Optionally, generating the corresponding organs based on the target path to obtain the digital human model includes: A skin layer is generated on the outer side of the target path, wherein the target path includes meridian paths and collateral paths; The collateral pathway is projected onto the subcutaneous region to generate collaterals; The meridian pathway is projected onto the muscle fascia layer under the skin to generate a meridian; The digital human model is obtained by connecting the meridians with the corresponding organ models.
[0012] To achieve the above objectives, the present invention also provides an electronic device, the electronic device including a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the digital human model construction method as described above.
[0013] To achieve the above objectives, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the digital human model construction method described above.
[0014] This invention proposes a method, electronic device, and computer-readable storage medium for constructing a digital human model. The method involves acquiring human body temperature distribution data and an initial 3D human body model of a target human body; identifying a first posture feature of the human body temperature distribution data and a second posture feature in the initial 3D human body model; registering the human body temperature distribution data and the initial 3D human body model based on the first and second posture features to obtain a projection relationship between the two; mapping the temperature values in the human body temperature distribution data to the initial 3D human body model based on the projection relationship to obtain a human thermal model; and generating human body part models in the human thermal model to obtain a digital human model corresponding to the target human body. By registering the human body temperature distribution data and the initial 3D human body model based on posture features, accurate mapping of temperature values can be achieved without setting marker points on the human body, thus avoiding the influence of the human body's state during data acquisition. This allows for data acquisition of the human body in its natural state, thereby constructing an accurate digital human model. Attached Figure Description
[0015] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0016] 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, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 This is a flowchart illustrating the first embodiment of the digital human model construction method of the present invention; Figure 2 This is a schematic diagram of the module structure of the electronic device of the present invention. Detailed Implementation
[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.
[0019] This invention provides a method for constructing a digital human model, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the digital human model construction method of the present invention, the method comprising the following steps: Step S10: Obtain the body temperature distribution data and initial three-dimensional human body model of the target human body; The target human body is the user who needs to create a digital human model.
[0020] Human body temperature distribution data reflects the temperature status of the surface area of the target human body. The method of acquiring human body temperature distribution data can be set according to actual needs, such as acquiring the target human body through a high-resolution infrared thermal imager. Since the infrared thermal imager acquires planar images, multiple infrared thermal images can be acquired from the human body multiple times and from multiple perspectives, and then combined to obtain human body temperature distribution data.
[0021] The initial 3D human body model is a 3D model obtained by scanning the target human body. The specific method of obtaining the initial 3D human body model can be set according to actual needs. For example, a structured light scanner or laser scanner can be used to scan the target human body to obtain a high-density point cloud of the human body, and then a triangular mesh model can be constructed from the high-density point cloud of the human body to obtain the initial 3D human body model. It is understandable that there may be problems such as noise and missing data in the point cloud acquisition. Therefore, after acquiring the high-density point cloud of the human body, it can be denoised, smoothed, and hole-filled before constructing the initial 3D human body model.
[0022] Step S20: Identify the first pose features of the human body temperature distribution data and the second pose features in the initial three-dimensional human body model; The first posture feature is a posture-related feature extracted from human body temperature distribution data. The first posture feature reflects the human body posture as manifested by temperature. It can be understood that different parts of the human body have different temperatures, such as the location of meridians having obvious high or low temperature characteristics. Therefore, human body temperature distribution data can reflect the location of different parts of the human body structure. In this embodiment, the first posture feature is extracted to clarify this.
[0023] The second pose feature is the pose-related feature extracted from the initial three-dimensional human body model; the second pose feature reflects the human body pose embodied by the human body structure.
[0024] Step S30: Based on the first posture features and the second posture features, register the human body temperature distribution data with the initial three-dimensional human body model to obtain the projection relationship between the human body temperature distribution data and the initial three-dimensional human body model; The human body temperature distribution data and the initial three-dimensional human body model were collected from the same user; therefore, they are consistent in terms of human body structure. In this embodiment, the human body posture corresponding to the human body temperature distribution data and the initial three-dimensional human body model is determined by extracting the first posture features and the second posture features. Therefore, based on the consistency in human body structure, the correspondence between the first posture features and the second posture features can be determined, thereby determining the projection relationship between the human body temperature distribution data and the initial three-dimensional human body model, and achieving registration.
[0025] The projection relationship indicates the rules for transforming the human body temperature distribution data with the coordinates of the initial three-dimensional human body model; the specific form of the projection relationship can be set according to actual needs, such as a mapping matrix.
[0026] Step S40: Map the temperature values in the human body temperature distribution data to the initial three-dimensional human body model based on the projection relationship to obtain the human body thermal model; After obtaining the mapping matrix, the correspondence between specific points in the human body temperature distribution data and specific points in the initial three-dimensional human body model can be clearly defined. Then, the temperature values in the human body temperature distribution data can be mapped to the corresponding positions in the initial three-dimensional human body model, thereby determining the temperature distribution on the surface of the initial three-dimensional human body model and obtaining the human body thermal model.
[0027] In actual mapping, human body temperature distribution data may include multiple infrared images. In this case, for a specific vertex in the initial 3D human body model, the temperature value corresponding to the corresponding temperature value in multiple infrared images is weighted and averaged to obtain the temperature value of that vertex for mapping. The weight corresponding to the temperature value is determined based on the consistency between the initial 3D human body model and the human body contour in the corresponding infrared image. The consistency is positively correlated with the weight.
[0028] A human thermal model is a human body model that has been fitted with body surface temperature.
[0029] Step S50: Generate human body part models in the human body thermal model to obtain a digital human model corresponding to the target human body.
[0030] Human body part models are models corresponding to different parts of the human body; human body part models can include models of skin, collaterals, meridians, internal organs, etc.
[0031] It is understandable that there is a correspondence between the temperature distribution on the human body surface and the location of its organs. Therefore, after obtaining the human body thermal model, the location of the organs in the human body can be determined based on the temperature distribution of the human body thermal model, and then the final digital human model can be obtained by generating the corresponding human body part model.
[0032] This embodiment registers human body temperature distribution data with an initial 3D human body model based on posture features, enabling accurate mapping of temperature values without setting markers on the human body. This avoids the impact on the human body's state during data acquisition, allowing data to be collected from the human body in its natural state, thereby constructing an accurate digital human model.
[0033] Furthermore, in the second embodiment of the digital human model construction method of the present invention based on the first embodiment, the first posture feature includes a first joint point, and the second posture feature includes a second joint point; step S30 includes the following steps: Step S31: Construct an initial projection matrix based on the first joint point and the second joint point, wherein the first joint point and the second joint point correspond one-to-one; Step S32: Construct an objective function based on the consistency difference between the human body temperature distribution data and the initial three-dimensional human body model; Step S33: Iteratively optimize the initial projection matrix using the objective function to obtain the projection relationship.
[0034] The first joint is the joint contained in the first pose feature.
[0035] The second joint is the joint contained in the second pose feature.
[0036] The method for determining joints can be set based on actual needs, such as using a human posture estimation model to perform joint recognition on human body temperature distribution data and an initial 3D human body model respectively; the specific type of human posture estimation model can be set based on actual needs, such as OpenPose.
[0037] The initial projection matrix is a transformation matrix constructed based on the first and second joints.
[0038] Consistency discrepancies indicate the differences in parameters between human body temperature distribution data and the initial 3D human body model for the same location in the mapping relationship.
[0039] It is understandable that when the initial projection matrix accurately reflects the projection relationship between the human body temperature distribution data and the initial 3D human body model, the consistency difference is 0. However, in practical applications, there are errors in the registration based on the joints, which causes the obtained initial projection matrix to fail to accurately reflect the projection relationship between the human body temperature distribution data and the initial 3D human body model. Therefore, in this embodiment, an objective function is constructed based on the consistency difference. This objective function reflects the error in the initial projection matrix. Then, the initial projection matrix is iteratively optimized based on the objective function to continuously reduce the error of the initial projection matrix, thereby obtaining a projection matrix that can accurately reflect the projection relationship between the human body temperature distribution data and the initial 3D human body model, i.e., the projection relationship.
[0040] Further, step S31 includes the following steps: Step S311: Obtain the intrinsic parameter matrix of the image acquisition device and the extrinsic parameter matrix of the image acquisition device and the 3D acquisition device. The human body temperature distribution data is acquired by the image acquisition device, and the initial 3D human body model is obtained from the point cloud data acquired by the 3D acquisition device. The image acquisition device and the 3D acquisition device simultaneously acquire data from the target human body. Step S312: Determine the device projection relationship based on the intrinsic parameter matrix and the extrinsic parameter matrix; Step S313: Match the first joint point and the second joint point based on the device projection relationship to obtain the initial projection matrix.
[0041] Image acquisition equipment is used to collect data on the distribution of human body temperature. Examples of image acquisition equipment include infrared thermal imagers.
[0042] 3D acquisition equipment is used to acquire source data, i.e. point cloud data, of the initial 3D human body model; 3D acquisition equipment can be a structured light scanner or a laser scanner.
[0043] The intrinsic parameter matrix is the inherent hardware parameter matrix of the image acquisition device; the intrinsic parameter matrix may include focal length, image principal point, lens distortion coefficient, etc.; the intrinsic parameter matrix is determined by the hardware parameters of the image acquisition device and remains unchanged.
[0044] The extrinsic parameter matrix is the rotation and translation transformation matrix of the image acquisition device relative to the 3D acquisition device; the extrinsic parameter matrix is determined by the relative pose between the image acquisition device and the 3D acquisition device; the image acquisition device and the 3D acquisition device simultaneously acquire data from the target human body, determine the relative position of the image acquisition device and the 3D acquisition device at the acquisition time, and then obtain the corresponding extrinsic parameter matrix.
[0045] The device projection relationship indicates the rigid projection relationship between the image acquisition device and the 3D acquisition device, and is entirely based on the intrinsic parameter matrix and the poses of both. The device projection relationship determines the positional transformation between the 2D coordinate system of the image acquisition device and the 3D coordinate system of the 3D acquisition device. Based on this, the first and second joint points are matched using the device projection relationship, thereby coarsely aligning the target human body between the 2D and 3D coordinate systems to obtain an approximate mapping relationship, i.e., the initial projection matrix.
[0046] Furthermore, the human body temperature distribution data is obtained from multiple infrared images; step S32 includes the following steps: Step S321: Project the initial three-dimensional human body model onto a two-dimensional object based on the initial projection matrix to obtain a two-dimensional projection image, and determine the contour error between the two-dimensional projection image and the corresponding human body contour in the infrared image. Step S322: For the second joint point in the three-dimensional human body model, determine the projection temperature value corresponding to the projection of the second joint point on each of the infrared images; Step S323: Obtain the temperature consistency error based on the differences between the projected temperature values; Step S324: Determine the non-rigid deformation field corresponding to the three-dimensional human body model, and calculate the spatial gradient norm of the non-rigid deformation field; Step S325: Construct the objective function by combining the contour error, the temperature consistency error, and the spatial gradient norm.
[0047] Infrared images can be obtained by capturing images of a target human body from infrared thermal imagers set up at different angles at the same time.
[0048] The two-dimensional projection image is the virtual human outline obtained by projecting the initial three-dimensional human body model onto a two-dimensional coordinate system.
[0049] The human body contour in the infrared image is the human body surface contour obtained by edge detection of the infrared image. Contour error indicates the difference in contour between the 3D human body model and the infrared image; the specific calculation method of contour error can be set according to actual needs, such as obtaining the contour error by calculating the chamfer distance between the corresponding human body contour in the 2D projection image and the infrared image.
[0050] The projected temperature value is the temperature value of the corresponding position of the vertex in the 3D human body model projected onto the infrared image; the backprojection is the transformation of the vertex in the 3D human body model onto each infrared image based on the initial projection matrix.
[0051] It is understandable that the initial projection matrix has errors. Therefore, when back-projecting vertices based on the initial projection matrix, the corresponding human body surface positions are different in different infrared images, and thus the corresponding projection temperature values are different. Therefore, the difference in projection temperature values in each infrared image can reflect the error of the initial projection matrix, i.e., the temperature consistency error.
[0052] The specific quantification method for temperature consistency error can be set based on actual needs, such as using the variance between each projected temperature value as the temperature consistency error.
[0053] The non-rigid deformation field is a three-dimensional displacement field that is performed vertex-by-vertex in a three-dimensional human body model.
[0054] The spatial gradient norm is the value obtained by solving for the spatial gradient of a non-rigid deformation field and taking the norm; the spatial gradient norm characterizes the magnitude of the change in deformation displacement between adjacent vertices.
[0055] Contour error provides shape constraints for the registration process, temperature consistency error provides temperature consistency constraints from multiple perspectives for the registration process, and spatial gradient norm provides smoothness constraints for the registration process. By combining the three to construct an objective function, the objective function can optimize the initial projection matrix by incorporating the three key constraints, thereby obtaining an accurate projection relationship.
[0056] The objective function Ereg has:
[0057] Among them, E geo For contour error; E temp For temperature consistency error; E smooth is the spatial gradient norm; w1~w3 are the corresponding weights, and the specific values of the weights can be set according to actual needs, such as w1=0.5, w2=0.3, w3=0.2.
[0058] Further, step S33 includes the following steps: Step S331: In each iteration, the extrinsic parameter matrix is adjusted to generate a new initial projection matrix; Step S332: After the iteration completion condition is met, the initial projection matrix is used as the projection relationship.
[0059] In each iteration, the extrinsic parameter matrix is adjusted to update the initial projection matrix. After the initial projection matrix is updated, the objective function is calculated; this process is repeated. The specific iterative algorithm can be set according to actual needs, such as LM (Levenberg-Marquardt nonlinear least squares optimization algorithm).
[0060] The iteration completion condition is the condition for the optimization to be completed; the iteration completion condition can be set according to actual needs, such as the number of iterations or the objective function parameter index.
[0061] When the iteration completion condition is the number of iterations, a target number of iterations can be set. When the number of iterations reaches the target number of iterations, the iteration completion condition is met. The specific value of the target number of iterations can be set based on actual needs, such as 50 iterations.
[0062] When the iteration completion condition is the objective function parameter index, a threshold can be set for the value of the objective function or for a specific parameter in the objective function. For example, the iteration completion condition is met when the objective function reflects that the contour error is less than the contour error threshold or the temperature consistency error is less than the temperature consistency error threshold.
[0063] You can also set the number of iterations and the objective function parameter index in the iteration completion condition.
[0064] Furthermore, in the third embodiment of the digital human model construction method of the present invention based on the first embodiment of the present invention, step S50 includes the following steps: Step S51: Obtain standard meridian coordinates and project the standard meridian coordinates onto the surface of the human body thermal model to obtain the initial meridian path; Step S52: Determine the temperature gradient of the initial meridian path on the human thermal model; Step S53: Determine the positional differences between the initial meridian path and the standard anatomical atlas, and the direction of the temperature gradient; Step S54: Optimize the initial meridian path based on the location difference to obtain the target path; Step S55: Generate the corresponding organs based on the target path to obtain the digital human model.
[0065] Standard meridian coordinates are the industry-standard spatial coordinates of meridians and acupoints; for example, standard meridian coordinates can be obtained from the GB / T12346 standard meridian and acupoint database.
[0066] Standard meridian coordinates are projected onto the surface of a human thermal model; the specific projection method can be set according to actual needs, such as UV mapping projection.
[0067] The initial meridian path is the path formed by the corresponding coordinates after the standard meridian coordinates are mapped onto the surface of the human body thermal model.
[0068] The temperature gradient is the rate of temperature change calculated along the normal direction of the initial meridian path; based on the temperature gradient, the high-temperature ridge line and low-temperature valley line of the human body can be determined; the high-temperature ridge line represents the area with vigorous Qi and blood; the low-temperature valley line represents the area with stagnation and deficiency.
[0069] Standard anatomical atlases are standard 3D anatomical models that include the spatial locations of human bones, muscles, and organs; they can be obtained from standard sources in the field of anatomy.
[0070] It is understandable that the meridian pathways of the human body have a certain degree of uniformity, and there is a strong correlation between the meridian pathways and temperature distribution. Therefore, in this embodiment, by determining the positional differences between the initial meridian path and the standard anatomical atlas and the direction of the temperature gradient, the positional differences can reflect the differences between the initial meridian path and the theoretically correct meridian distribution. Then, by reducing or eliminating the positional differences, the initial meridian path can be optimized to obtain a target path that conforms to human anatomy and fits the distribution of Qi and blood on the body surface.
[0071] Further, step S53 includes the following steps: Step S531: Determine multiple key acupoints on the initial meridian path and identify the corresponding landmarks in the standard anatomical atlas; Step S532: Determine the point distance between the key acupoint and the marker point; Step S533: Calculate the cosine of the angle between the tangent direction of the path and the direction of the temperature gradient; Step S534: Construct a path optimization function by combining the distance between the points and the cosine of the included angle, and use the path optimization function as the position difference.
[0072] Key acupoints are acupoints that correspond to the standard meridian coordinates in the initial meridian path.
[0073] Landmarks are feature points in standard anatomical atlases that have a standard relative position to key acupoints, such as feature points of bones or muscles near specific acupoints.
[0074] The point distance indicates the distance between the key acupoint and the corresponding landmark point; the specific calculation method of the point distance can be set according to actual needs, such as the point distance being the Euclidean distance between the key acupoint and the corresponding landmark point.
[0075] It is understandable that the point distance indicates the distance between the key acupoint and the corresponding landmark. The larger the point distance, the more the key acupoint deviates from the landmark corresponding to the anatomical standard. Therefore, in this embodiment, the point distance is combined to construct the path optimization function.
[0076] The tangent direction of the path is the tangent vector of the curve at any point on the initial meridian path, representing the direction of meridian extension.
[0077] The direction of the temperature gradient indicates the direction of body surface temperature.
[0078] The closer the cosine value of the included angle is to 1, the closer the meridian path is to the trend of Qi and blood circulation. Conversely, the smaller the cosine value of the included angle, the more the meridian deviates from the trend of Qi and blood circulation. Therefore, in this embodiment, the cosine value of the included angle is used to construct the path optimization function.
[0079] Path optimization function L * have:
[0080] Among them, D geo L is the distance between points; L is the initial meridian path in the current iteration; L std Standard meridian coordinates; D therm The value is the cosine of the included angle; T represents the temperature gradient; K(L) represents the path curvature smoothing term; α, β, and γ are the corresponding weights, and the specific values of the weights can be set according to actual needs, such as α=0.4, β=0.4, and γ=0.2.
[0081] The initial meridian path is optimized using a path optimization function to obtain the target path. The specific method can be set according to actual needs, such as using a dynamic programming algorithm.
[0082] Further, step S55 includes the following steps: Step S551: Generate a skin layer on the outer side of the target path, wherein the target path includes a meridian path and a collateral path; Step S552: Project the collateral pathway onto the subcutaneous tissue to generate collaterals; Step S553: Project the meridian path onto the muscle fascia layer under the skin to generate meridians; Step S554: Connect the meridians with the corresponding organ models to obtain the digital human model.
[0083] The skin region is a sheet-like area that extends outward to both sides of the curved surface of the human body thermal model with the target path as the center. The skin region corresponds to the twelve skin regions and is the operation layer for moxibustion, scraping, and skin examination. Its anatomical location is the epidermis and superficial dermis.
[0084] The meridian pathway is the baseline of the twelve primary meridians in the target pathway. It is the main meridian baseline that runs longitudinally through the whole body and is the main channel for the circulation of Qi and blood.
[0085] The collateral pathway is the branch baseline extending from the main branch in the target pathway, which is distributed laterally on the body surface and corresponds to the superficial subcutaneous blood and qi branches.
[0086] The projection of the meridian path onto the subcutaneous tissue determines the meridian. Since the meridian is distributed in the muscle fascia layer, the meridian path is projected onto the subcutaneous muscle fascia layer to generate the meridian.
[0087] The projection of the collateral vessels under the skin determines the collateral vessels, and the specific depth can be set according to actual needs, such as 1.5mm.
[0088] The viscera model is a three-dimensional anatomical model of the internal organs corresponding to the functional affiliation of each meridian in the standard anatomical atlas; it is a standard model.
[0089] Extending from the main meridians to the organ models, connecting channels are generated to establish a three-dimensional topological connection between the surface meridians and the deep internal organs.
[0090] In this embodiment, a digital human model is constructed using a four-layer three-dimensional structure of skin-network-meridian-organ, which can clearly show the spatial relationship between meridians and deep anatomical structures, providing an intuitive internal perspective for acupuncture depth control and massage technique guidance.
[0091] After completing the construction of the individualized four-layer three-dimensional meridian structure, the system enters the multimodal fusion, noise reduction, and model output processing flow. First, topological consistency verification is carried out, traversing the complete meridian path and detecting its spatial intersection and overlap with the anatomical models of blood vessels and nerves. The meridian points with interference are fine-tuned in strict accordance with the human anatomy avoidance rules to avoid the distortion problem of conflict between meridians and important soft tissue anatomical structures. After verification, the overall three-dimensional model is subjected to Laplacian global smoothing to uniformly eliminate various noises such as surface burrs and vertex mutations left over from the previous 2D-3D registration, meridian path optimization, and layered modeling process, ensuring the continuous smoothness of the human body surface and meridian pipelines. Finally, the processed geometric mesh data, vertex-by-vertex body surface temperature and physiological attributes, the four-layer three-dimensional meridian hierarchy structure of skin-network-meridian-organ, and the anatomical association information of meridians and bones and organs are integrated to export GLTF / OBJ three-dimensional model files with JSON attribute description files to obtain a complete and usable digital human model.
[0092] The overall implementation principle of this application is explained below based on practical examples: I. System Structure Composition: Markless acquisition module: integrates an infrared thermal imager array and a 3D structured light scanner; Feature registration engine: performs multimodal feature extraction, pose estimation, and nonlinear iterative optimization registration; Meridian reconstruction algorithm module: includes built-in thermal gradient calculation, path dynamic programming solver, and hierarchical extension logic; Anatomy Fusion Module: Responsible for non-rigid deformation adaptation and topological conflict detection of standard anatomical models; Visualization terminal: Provides 3D rendering, layered display, heat overlay and data export functions.
[0093] II. Constructing a markerless, individualized 3D digital human model of the meridians of an adult female volunteer; 1. Data collection; Environment: Room temperature 22±0.5℃, humidity 50%, volunteers sat quietly for 15 minutes until their body temperature stabilized; Infrared acquisition: Using a FLIRA700 thermal imager, thermal images were acquired from four perspectives: front, back, left, and right, with a resolution of 640×480 and no markers. 3D scanning: Using the Shining3DEinScanPro2XPlus scanner, full-body point cloud data is acquired simultaneously to reconstruct the initial 3D human body model; Data loading: Load the GB / T12346 standard meridian database and the female standard anatomical model.
[0094] 2. Markerless natural registration; Feature extraction: High curvature points such as acromion, elbow tip, anterior superior iliac spine, and knee eye are extracted at the 3D end; Canny edge detection is performed on the thermal image at the infrared end to extract body contours and temperature change edge points such as the neck and armpit. Initial registration: 18 key points in the 2D heatmap are identified using OpenPose, and an initial mapping is established with the corresponding vertices of the 3D model. The initial projection matrix is then calculated. Iterative optimization: Construct an energy function, setting w1=0.5, w2=0.3, and w3=0.2. Run the LM algorithm iteratively 50 times to achieve a 3D model projection contour with an infrared image contour overlap of >95% and a temperature distribution error of <0.2℃. Complete texture mapping to generate a human thermal model.
[0095] 3. Meridian reconstruction; Taking the Lung Meridian of Hand-Taiyin as an example: Project the standard acupoints LU1-LU11 onto the human thermal model; Thermal analysis revealed a faint high-temperature band along the lung meridian on the anterior medial border of the volunteer's upper limb (ΔT≈0.5℃, indicating relatively vigorous Qi and blood). Path correction: Set α=0.4, β=0.4, γ=0.2. The dynamic programming algorithm shifts the path towards the center of the high-temperature zone, with the maximum shift controlled within 3mm (within the allowable range for dissection), generating the target path; Layered modeling: Extend outward by 4mm to generate the skin; project inward to the subcutaneous layer by 1.5mm to generate the vascular network; penetrate into the intermuscular space of the pectoralis major / biceps brachii to generate the main meridian; connect to the lung model.
[0096] 4. Output and Verification; Topology check: Confirm that the path does not pass through major blood vessels (such as the brachial artery); Output: Generates a GLTF model. After expert review, the individualized path overlaps with the "sensitive points" found by traditional palpation by more than 85%, which is better than the standard model (overlap ~60%).
[0097] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.
[0098] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.
[0099] Reference Figure 2 In terms of hardware structure, the electronic device may include components such as a communication module 10, a memory 20, and a processor 30. In the electronic device, the processor 30 is connected to both the memory 20 and the communication module 10. The memory 20 stores a computer program, which is executed by the processor 30. When the computer program is executed, it implements the steps of the above-described method embodiments.
[0100] The communication module 10 can connect to external communication devices via a network. The communication module 10 can receive requests from the external communication devices and can also send requests, instructions, and information to the external communication devices. The external communication devices can be other electronic devices, servers, or IoT devices, such as televisions, etc.
[0101] The memory 20 can be used to store software programs and various data. The memory 20 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as acquiring human body temperature distribution data and an initial three-dimensional human body model of the target human body), etc.; the data storage area may include a database, and may store data or information created based on system usage. Furthermore, the memory 20 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.
[0102] The processor 30 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in the memory 20, and by calling data stored in the memory 20, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. The processor 30 may include one or more processing units; optionally, the processor 30 may integrate an application processor and a modem processor. The application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 30.
[0103] although Figure 2 Not shown, but the above-described electronic device may further include a circuit control module for connecting to a power supply to ensure the normal operation of other components. Those skilled in the art will understand that... Figure 2 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.
[0104] The present invention also proposes a computer-readable storage medium having a computer program stored thereon. The computer-readable storage medium may be... Figure 2 The memory 20 in the electronic device may also be at least one of ROM (Read-Only Memory) / RAM (Random Access Memory), magnetic disk, optical disk, etc. The computer-readable storage medium includes a number of instructions to cause a terminal device with a processor (which may be a television, automobile, mobile phone, computer, server, terminal, or network device, etc.) to execute the methods described in the various embodiments of the present invention.
[0105] In this invention, the terms "first," "second," "third," "fourth," and "fifth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.
[0106] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0107] Although embodiments of the present invention have been shown and described above, the scope of protection of the present invention is not limited thereto. It is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, and substitutions to the above embodiments within the scope of the present invention, and such changes, modifications, and substitutions should all be covered 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 method for constructing a digital human model, characterized in that, The digital human model construction method includes: Acquire human body temperature distribution data and an initial three-dimensional human body model of the target human body; Identify the first pose feature of the human body temperature distribution data and the second pose feature in the initial three-dimensional human body model; Based on the first posture features and the second posture features, the human body temperature distribution data is registered with the initial three-dimensional human body model to obtain the projection relationship between the human body temperature distribution data and the initial three-dimensional human body model. The temperature values in the human body temperature distribution data are mapped onto the initial three-dimensional human body model based on the projection relationship to obtain the human body thermal model; Human body part models are generated in the human body thermal model to obtain the digital human model corresponding to the target human body.
2. The digital human model construction method as described in claim 1, characterized in that, The first posture feature includes a first joint point, and the second posture feature includes a second joint point; the step of registering the human body temperature distribution data with the initial three-dimensional human body model based on the first posture feature and the second posture feature to obtain the projection relationship between the human body temperature distribution data and the initial three-dimensional human body model includes: An initial projection matrix is constructed based on the first joint point and the second joint point, wherein the first joint point and the second joint point correspond one-to-one. An objective function is constructed based on the consistency difference between the human body temperature distribution data and the initial three-dimensional human body model; The projection relationship is obtained by iteratively optimizing the initial projection matrix using the objective function.
3. The digital human model construction method as described in claim 2, characterized in that, The step of constructing the initial projection matrix based on the first joint point and the second joint point includes: The intrinsic parameter matrix of the image acquisition device and the extrinsic parameter matrix of the image acquisition device and the 3D acquisition device are obtained. The human body temperature distribution data is acquired by the image acquisition device, and the initial 3D human body model is obtained by the point cloud data acquired by the 3D acquisition device. The image acquisition device and the 3D acquisition device simultaneously acquire data from the target human body. The device projection relationship is determined based on the intrinsic parameter matrix and the extrinsic parameter matrix; The initial projection matrix is obtained by matching the first joint point and the second joint point based on the device projection relationship.
4. The digital human model construction method as described in claim 3, characterized in that, The step of iteratively optimizing the initial projection matrix using the objective function to obtain the projection relationship includes: In each iteration, the extrinsic parameter matrix is adjusted to generate a new initial projection matrix; After the iteration completion condition is met, the initial projection matrix is used as the projection relationship.
5. The digital human model construction method as described in claim 2, characterized in that, The human body temperature distribution data is obtained from multiple infrared images; the objective function constructed based on the consistency difference between the human body temperature distribution data and the initial three-dimensional human body model includes: The initial three-dimensional human body model is projected onto a two-dimensional object based on the initial projection matrix to obtain a two-dimensional projection image, and the contour error between the two-dimensional projection image and the corresponding human body contour in the infrared image is determined. For the second joint point in the three-dimensional human body model, determine the projection temperature value corresponding to the projection of the second joint point on each infrared image; The temperature consistency error is obtained based on the differences between the projected temperature values. Determine the non-rigid deformation field corresponding to the three-dimensional human body model, and calculate the spatial gradient norm of the non-rigid deformation field; The objective function is constructed by combining the contour error, the temperature consistency error, and the spatial gradient norm.
6. The digital human model construction method as described in claim 1, characterized in that, The step of generating human body part models in the human body thermal model to obtain the digital human model corresponding to the target human body includes: Obtain standard meridian coordinates and project the standard meridian coordinates onto the surface of the human thermal model to obtain the initial meridian path; Determine the temperature gradient of the initial meridian path on the human thermal model; Determine the positional differences between the initial meridian path and the standard anatomical atlas, and the direction of the temperature gradient; The initial meridian path is optimized based on the location differences to obtain the target path; The digital human model is obtained by generating the corresponding organs based on the target path.
7. The digital human model construction method as described in claim 6, characterized in that, The determination of the positional differences between the initial meridian path and the standard anatomical atlas, and the direction of the temperature gradient, includes: Identify multiple key acupoints on the initial meridian path, and determine the corresponding landmarks in the standard anatomical atlas; Determine the point distance between the key acupoint and the marker point; Calculate the cosine of the angle between the tangent direction of the path and the direction of the temperature gradient; A path optimization function is constructed by combining the distance between the points and the cosine of the included angle, and the path optimization function is used as the position difference.
8. The digital human model construction method as described in claim 6, characterized in that, The process of generating the corresponding organs based on the target path to obtain the digital human model includes: A skin layer is generated on the outer side of the target path, wherein the target path includes meridian paths and collateral paths; The collateral pathway is projected onto the subcutaneous region to generate collaterals; The meridian pathway is projected onto the muscle fascia layer under the skin to generate a meridian; The digital human model is obtained by connecting the meridians with the corresponding organ models.
9. An electronic device, characterized in that, The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the digital human model construction method as described in any one of claims 1 to 8.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the digital human model construction method as described in any one of claims 1 to 8.
