Method and apparatus for spine generation based on 3D body scan

Abstract

The present application belongs to the technical field of human body image digital information processing, and provides a spine generation method and device based on 3D human body scanning. The method comprises the following steps: fitting a three-dimensional reconstruction model according to a basic digital human model to obtain a fitted digital human model; calculating a body shape parameter used in the fitting of the fitted digital human model; performing regression processing on a preset basic spine model using the body shape parameter to obtain a preliminary spine model matched with the size of the skeleton of the three-dimensional reconstruction model; performing deformation processing on the basic digital human model using the body shape parameter to obtain a digital human model deformed using the body shape parameter; matching the position of the skeleton of the preliminary spine model with the digital human model deformed using the body shape parameter to generate a spine model matched with the spine of the three-dimensional reconstruction model. Thus, the accuracy, efficiency and cost of spine judgment are improved, and the popularization and application of spine judgment are promoted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of human body image digital information processing, and more particularly, to a spine generation method and device based on 3D human body scanning. BACKGROUND

[0002] The spine bears important functions of supporting the human body, conducting load, moving, maintaining stability, protecting the spinal cord and flexible movement of six degrees of freedom. Modern people often cause spine health problems due to long-term desk work, bending over, and not paying attention to standing and sitting posture. However, the damage and abnormal growth of the spine are concealed and often easily ignored. At present, the methods for judging the health of the human spine mainly include spine CT scanning, manual bone palpation and traditional 3D scanning. Spine CT scanning is a commonly used method for checking spine diseases at present. It can truly reflect the morphological structure of the spine in a specific posture by using a penetrating type of tomographic scanning. However, CT radiation has adverse effects on health, and CT scanning is high in cost, low in efficiency, and highly professional, which is not conducive to popularization and application. Manual bone palpation needs to be performed in a specific posture or with the help of related equipment, and then combined with the experience of doctors for diagnosis, which highly depends on the experience and guidance of doctors and is difficult to popularize on a large scale. At the same time, contact diagnosis and treatment may also have some awkward situations in actual operation. The traditional 3D scanning is to register the human body model in the three-dimensional image with the standard spine model established in advance, and then detect the path of the back hollow, so as to infer the coordinates of each vertebra and obtain the spine model. However, the traditional 3D scanning fails to provide an effective key point positioning method, and the deviation of key point positioning may cause abnormal stretching of the spine in the longitudinal direction, abnormal displacement of the spine in the transverse direction, etc. At the same time, the back hollow reflects the position of the muscle rather than the real position of each vertebra, so that the obtained spine model has a large error.

[0003] In summary, the existing spine judgment technology has the technical problems of high cost, low efficiency, low precision and being not conducive to popularization and application. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a spine generation method and device based on 3D human body scanning to improve the precision of spine judgment, improve the efficiency of spine judgment, reduce the cost of spine judgment, and thereby promote the popularization and application of spine judgment, in view of the deficiencies in the above technical solutions.

[0005] In a first aspect, the present application provides a spine generation method based on 3D human body scanning, which comprises the following steps:

[0006] The preset basic digital human model is fitted according to a three-dimensional reconstruction model generated after a target human body is scanned by a 3D human body scanning device, to obtain a fitted digital human model; the basic digital human model has a point cloud topological relationship representing a human body composition structure;

[0007] The body shape parameters used when the fitted digital human model is obtained by fitting the basic digital human model and the three-dimensional reconstruction model are calculated;

[0008] The preset basic spine model is subjected to regression processing using the body shape parameters, to obtain a preliminary spine model matched with the bone size of the three-dimensional reconstruction model, and the basic digital human model is subjected to deformation processing using the body shape parameters, to obtain a digital human model deformed using the body shape parameters;

[0009] The preliminary spine model is matched with the position of the bone of the digital human model deformed using the body shape parameters, to generate a spine model matched with the spine of the three-dimensional reconstruction model.

[0010] In a second aspect, the present application provides a spine generation device based on 3D human body scanning, which comprises:

[0011] The fitted digital human model generation module is configured to preset a basic digital human model, and fit a three-dimensional reconstruction model generated after a target human body is scanned by a 3D human body scanning device according to the basic digital human model, to obtain a fitted digital human model; the basic digital human model has a point cloud topological relationship representing a human body composition structure;

[0012] The body shape parameter calculation module is configured to calculate body shape parameters used when the fitted digital human model is obtained by fitting the basic digital human model and the three-dimensional reconstruction model;

[0013] The body shape parameter use module is configured to subject a preset basic spine model to regression processing using the body shape parameters, to obtain a preliminary spine model matched with the bone size of the three-dimensional reconstruction model, and subject the basic digital human model to deformation processing using the body shape parameters, to obtain a digital human model deformed using the body shape parameters;

[0014] The spine model generation module is configured to match the preliminary spine model with the position of the bone of the digital human model deformed using the body shape parameters, to generate a spine model matched with the spine of the three-dimensional reconstruction model.

[0015] Compared with the prior art, the present application has the following beneficial effects:

[0016] The application provides a spine generation method and device based on 3D human body scanning, which comprises the following steps: presetting a basic digital human model, fitting a three-dimensional reconstruction model generated after a 3D human body scanning device scans a target human body according to the basic digital human model, obtaining a fitted digital human model, the basic digital human model having a point cloud topological relationship representing the composition structure of a human body, calculating a body shape parameter used when fitting the fitted digital human model from the basic digital human model and the three-dimensional reconstruction model, performing regression processing on a preset basic spine model using the body shape parameter to obtain a preliminary spine model matched with the bone size of the three-dimensional reconstruction model, and performing deformation processing on the basic digital human model using the body shape parameter to obtain a digital human model deformed using the body shape parameter, matching the position of the bones of the preliminary spine model and the digital human model deformed using the body shape parameter to generate a spine model matched with the spine of the three-dimensional reconstruction model, thereby improving the accuracy, efficiency and cost of spine judgment and promoting the popularization and application of spine judgment. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a flowchart of the spine generation method based on 3D human body scanning of the application;

[0018] Figure 2 is another flowchart of the spine generation method based on 3D human body scanning of the application;

[0019] Figure 3 is a schematic diagram of the corresponding point positions of the spine model and the digital human model of the application;

[0020] Figure 4 is a schematic diagram of the predicted bone point coordinates of the application;

[0021] Figure 5 is a schematic diagram of the model calculated bone points and the corresponding bone points in the spine model of the application;

[0022] Figure 6 is a schematic diagram of the spine generation device based on 3D human body scanning of the application;

[0023] Figure 7 is a schematic diagram of the server of the application. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical scheme and advantages of the application clearer, the application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the application and do not limit the application.

[0025] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and in the above description are used for distinguishing between similar objects, and do not necessarily have to follow a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances so that the embodiments of the present application described herein can be implemented in other than the order illustrated or described herein.

[0026] Embodiment one

[0027] Referring to Figures 1 to 7 , embodiment one provides a spine generation method based on 3D human body scanning, comprising steps S1, S2, S3 and S4, the embodiment provides a spine generation method based on 3D human body scanning, by presetting a basic digital human model, and according to the basic digital human model, fitting the three-dimensional reconstruction model generated after the 3D human body scanning device scans the target human body, to obtain the fitted digital human model, the basic digital human model has point cloud topological relationship representing the composition structure of the human body, calculating the body shape parameters used when fitting the basic digital human model and the three-dimensional reconstruction model to obtain the fitted digital human model, using the body shape parameters to perform regression processing on the preset basic spine model, to obtain a preliminary spine model matched with the bone size of the three-dimensional reconstruction model, and using the body shape parameters to perform deformation processing on the basic digital human model, to obtain a digital human model deformed using body shape parameters, matching the position of the bones of the preliminary spine model and the digital human model deformed using body shape parameters, to generate a spine model matched with the spine of the three-dimensional reconstruction model, thereby improving the accuracy of spine judgment, improving the efficiency of spine judgment, reducing the cost of spine judgment, and promoting the popularization and application of spine judgment.

[0028] It should be noted that the method for generating a spine based on a 3D human body scan provided in this embodiment can be run on a server, which is the execution subject of all or part of the steps in the method for generating a spine based on a 3D human body scan. In addition to being able to execute steps S1, S2, S3 and S4 in this embodiment, it can also run part or all of the steps of the methods referred to below. The server includes a memory, a processor and a network interface that are communicatively connected to each other through a system bus. It should be noted that only part of the components of the server are shown in the figure, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Those skilled in the art in the technical field can understand that the server here is a device that can automatically perform numerical calculation and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc. The server can be a smart phone, a smart wearable device or other computing device. The server can interact with the user through a keyboard, a mouse, a remote control, a touchpad or a voice control device, etc. The memory includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory can be an internal storage unit of the server, such as the hard disk or the memory of the server. In other embodiments, the memory can also be an external storage device of the server, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the server. Of course, the memory can include both the internal storage unit and the external storage device of the server.

[0029] Specifically, in step S1, the server pre-sets a basic digital human model, and according to the basic digital human model, fits the three-dimensional reconstruction model generated after the 3D human body scanning device scans the target human body, to obtain a fitted digital human model. The basic digital human model has a point cloud topological relationship representing the composition structure of the human body.

[0030] It should be noted that the 3D human body scanning device can include a depth camera, wherein the depth camera can be a monocular depth camera. The depth camera can include an infrared camera, a dot matrix projector, and a depth calculation processor. It can be understood that the depth camera scans the human body, is safe and radiation-free, and only collects depth information of the human body without collecting color information, thereby protecting the privacy of the user.

[0031] It should also be noted that the three-dimensional reconstruction model belongs to an unordered data model and can generally reflect the body model of the target human body. However, the specific human body parts and sizes cannot be identified from the three-dimensional reconstruction model alone, and the human body structure of the three-dimensional reconstruction model cannot be obtained. Meanwhile, the specific human body parts and sizes of the preset basic digital human model are known data, although the preset basic digital human model is not from the target human body. The specific human body parts and sizes of the preset basic digital human model are represented by the point cloud topological relationship of the basic digital human model, so that the human body composition structure can be obtained. In this embodiment, the three-dimensional reconstruction model is fitted according to the basic digital human model to obtain a fitted digital human model, which can reflect the specific human body parts and sizes and obtain the human body structure of the three-dimensional reconstruction model. When fitting the three-dimensional reconstruction model, a basic digital human model close to the three-dimensional reconstruction model can be selected, so that the fitted digital human model obtained by fitting is more realistic.

[0032] It should also be noted that fitting the three-dimensional reconstruction model generated by the 3D human body scanning device after scanning the target human body according to the basic digital human model to obtain a fitted digital human model can include the following steps: dividing the basic digital human model into spine-related parts, and aligning the spine-related parts with the three-dimensional reconstruction model generated by the 3D human body scanning device after scanning the target human body; fusing the point clouds of the spine-related parts by point cloud fusion technology to generate a fusion model; performing posture correction on the arms and legs of the fusion model so that the arms and legs of the fusion model are in the same position as the arms and legs of the basic digital human model; and performing a fitting operation on the point cloud of the posture-corrected fusion model so that the fusion model is close to the three-dimensional reconstruction model, thereby obtaining a fitted digital human model. Further, the step of dividing the basic digital human model into spine-related parts and aligning the spine-related parts with the three-dimensional reconstruction model generated by the 3D human body scanning device after scanning the target human body can include the following steps: dividing the basic digital human model into shoulder-chest, waist, and hip-leg parts; and aligning the shoulder-chest, waist, and hip-leg parts with the three-dimensional reconstruction model generated by the 3D human body scanning device after scanning the target human body by point cloud registration technology.

[0033] Specifically, in step S2, the server calculates the body shape parameters used when fitting the fitted digital human model from the basic digital human model and the three-dimensional reconstruction model.

[0034] It should be noted that the calculation of body shape parameters used when fitting the fitted digital human model from the basic digital human model and the 3D reconstruction model may include the following steps: setting an energy equation, and minimizing the energy equation to obtain a set of body shape parameters β1, β2, ... β 10 The body shape parameters β1, β2, ... β 10 The body shape parameters are those required when fitting the fitted digital human model from the basic digital human model and the 3D reconstruction model; the energy equation is E. Vdiff =M(β1, β2, ... β) 10 )-m; where M represents the body shape parameters β1, β2, ... β 10 A function to obtain the coordinate values ​​of the model after deformation from the basic digital human model, where m represents the coordinate values ​​of the fitted digital human model.

[0035] Furthermore, the energy equation is minimized to obtain a set of body size parameters β1, β2, ... β 10 This may include the following steps: minimizing the energy equation using gradient descent to obtain a set of body size parameters β1, β2, ... β 10 Furthermore, the energy equation is minimized using the gradient descent method to obtain a set of body size parameters β1, β2, ... β 10 This may include the following steps:

[0036] Given a continuously differentiable function J(β1, β2, ... β) to be optimized 10 ), where β1, β2, ... β 10 Given the body size parameter, a learning rate or step size 'a', and a set of initial values ​​β1, β2, ... β 10 =0;

[0037] Calculate the gradient of the continuously differentiable function J to be optimized.

[0038] Towards the gradient Update β1, β2, ... β in the direction of fastest descent. 10 ;

[0039] Recalculate the new gradient of the continuously differentiable function J to be optimized. Calculate the gradient The model is used to determine whether the loop needs to be terminated, in order to obtain a set of body shape parameters β1, β2, ... β 10 .

[0040] Specifically, in step S3, the server uses the body shape parameters to regress the preset basic spine model to obtain a preliminary spine model matching the size of the bones of the three-dimensional reconstruction model, and uses the body shape parameters to deform the basic digital human model to obtain a digital human model deformed using the body shape parameters.

[0041] It should be noted that step S3 can include the following steps: using the body shape parameters β1, β2,... β 10 regressing the preset basic spine model to obtain a preliminary spine model matching the size of the bones of the three-dimensional reconstruction model; using the body shape parameters β1, β2,... β 10 deforming the basic digital human model to obtain a digital human model deformed using the body shape parameters.

[0042] It can be understood that people with different body shape parameters have different skeletal characteristics, such as different sizes of the pelvis, different sizes of the scapula, different lengths of the spine, and the like. The basic spine model cannot be directly applied to people with different body shapes, and different sizes and lengths of the skeleton need to be set according to different body shape characteristics. In this embodiment, by using 10-dimensional body shape parameters β1, β2,... β 10 The preset basic spine model is regressed to obtain a skeleton matching the size of the bones of the three-dimensional reconstruction model.

[0043] Specifically, in step S4, the server matches the preliminary spine model with the positions of the bones of the digital human model deformed using the body shape parameters to generate a spine model matching the spine of the three-dimensional reconstruction model.

[0044] It should be noted that step S4 can include the following steps: using the fixed topological structure of the preliminary spine model and the digital human model deformed using the body shape parameters, configuring key point information of the digital human model deformed using the body shape parameters, the key point information being used to determine the positions of the bones of the digital human model deformed using the body shape parameters; and according to the key point information, rotating and translating the bones of different parts of the preliminary spine model that have been determined in size to appropriate positions to generate a spine model matching the spine of the three-dimensional reconstruction model. Further, the configuration of the key point information of the digital human model deformed using the body shape parameters includes: finding corresponding points of the digital human model deformed using the body shape parameters and the bones at positions where the skin is thin; and after deriving ideal spine model positioning points from the digital human model deformed using the body shape parameters, minimizing the sum of coordinate differences between the ideal spine model positioning points and corresponding positioning points of the preliminary spine model.

[0045] It should be noted that, in order to make the spinal model and use 10-dimensional body shape parameters β1, β2, ... β 10 The deformed digital human model exhibits good adaptability. It requires rotating and translating the bones of different parts, whose lengths and sizes are already determined, to appropriate positions. Since both the basic spinal model and the digital human model deformed using body parameters have fixed topological structures, key point information can be configured first, selecting points that play a crucial role in determining the positions of each bone. These crucial points can be divided into two categories: one type is the corresponding points between the spinal model and the digital human model. These corresponding points are found in areas with thinner skin, such as the scapula, clavicle, and lumbar vertebrae. The corresponding points between the spinal model and the digital human model are as follows: Figure 3 As shown, the sum of the coordinate differences of the corresponding points can be used as a loss function. Another key point is the point obtained by minimizing the sum of the coordinate differences between the ideal spine model positioning points and the corresponding positioning points of the basic spine model after deriving the ideal spine model positioning points from the digital human model deformed using body shape parameters.

[0046] In some improved embodiments, the ideal skeletal positioning points are derived from the digital human model deformed using body shape parameters. This can be done using a linear derivation method, combining the coordinates of multiple digital human models deformed using body shape parameters. For example... Figure 4 As shown, assuming Figure 4 In the model, the first type point 100 and the second type point 101 represent a ring of horizontal points near the neck of the digital human model. The third type point 102 represents the predicted cervical skeleton of the digital human model. In the second type point 101, P1, P2, P3, and P4 are reference points used in the coordinate calculation. Therefore, the coordinates V of the cervical skeleton points can be calculated using the following formula: V = P1 × 0.45 + P2 × 0.45 + P3 × 0.05 + P4 × 0.05. Similarly, other skeletal positioning points can be calculated from the coordinates of the digital human model deformed using body shape parameters. Figure 5 The model shows the skeletal points calculated by the model and their corresponding skeletal points in the spine model.

[0047] It should be noted that matching skeletons and human models can be achieved by minimizing the sum of coordinate differences between the two types of skeleton pairing points mentioned above, thus determining the optimal positions of each bone that makes up the spine model.

[0048] Example 2

[0049] See Figures 1 to 7 Based on the above embodiments, this embodiment provides a spine generation device based on 3D human body scanning, comprising:

[0050] The fitted digital human model generation module is configured to preset a basic digital human model, and fit a three-dimensional reconstruction model generated after a target human body is scanned by a 3D human body scanning device according to the basic digital human model, to obtain a fitted digital human model; the basic digital human model has a point cloud topological relationship representing a human body composition structure.

[0051] The body shape parameter calculation module is configured to calculate a body shape parameter used when the fitted digital human model is obtained by fitting the basic digital human model and the three-dimensional reconstruction model.

[0052] The body shape parameter use module is configured to use the body shape parameter to perform regression processing on a preset basic spine model, to obtain a preliminary spine model matched with a bone size of the three-dimensional reconstruction model, and use the body shape parameter to perform deformation processing on the basic digital human model, to obtain a digital human model deformed using the body shape parameter.

[0053] The spine model generation module is configured to match the preliminary spine model with a position of a bone of the digital human model deformed using the body shape parameter, to generate a spine model matched with a spine of the three-dimensional reconstruction model.

[0054] It should be noted that in the embodiment, a basic digital human model is preset, and a three-dimensional reconstruction model generated after a target human body is scanned by a 3D human body scanning device is fitted according to the basic digital human model, to obtain a fitted digital human model; the basic digital human model has a point cloud topological relationship representing a human body composition structure; a body shape parameter used when the fitted digital human model is obtained by fitting the basic digital human model and the three-dimensional reconstruction model is calculated; the body shape parameter is used to perform regression processing on a preset basic spine model, to obtain a preliminary spine model matched with a bone size of the three-dimensional reconstruction model; and the body shape parameter is used to perform deformation processing on the basic digital human model, to obtain a digital human model deformed using the body shape parameter; the preliminary spine model is matched with a position of a bone of the digital human model deformed using the body shape parameter, to generate a spine model matched with a spine of the three-dimensional reconstruction model, thereby improving the accuracy of spine judgment, improving the efficiency of spine judgment, reducing the cost of spine judgment, and promoting the popularization and application of spine judgment.

[0055] It should be noted that the above embodiments are only preferred specific embodiments of the present application, and the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application, and the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A spine generation method based on 3D body scan, characterized in that, The spine generation method based on the 3D human body scanning comprises the following steps: A basic digital human model is preset, and a three-dimensional reconstruction model generated after a target human body is scanned by a 3D human body scanning device is fitted according to the basic digital human model to obtain a fitted digital human model; the basic digital human model has a point cloud topological relationship representing the composition structure of a human body; Body shape parameters used in fitting the fitted digital human model from the basic digital human model and the three-dimensional reconstruction model are calculated; The basic spine model is subjected to regression processing using the body shape parameters to obtain a preliminary spine model matched with the bone size of the three-dimensional reconstruction model, and the basic digital human model is subjected to deformation processing using the body shape parameters to obtain a digital human model deformed using the body shape parameters; The preliminary spine model is matched with the position of the bone of the digital human model deformed using the body shape parameters to generate a spine model matched with the spine of the three-dimensional reconstruction model; The body shape parameters used in fitting the fitted digital human model from the basic digital human model and the three-dimensional reconstruction model are calculated, comprising: setting up an energy equation, minimizing the energy equation to obtain a set of body shape parameters , the body shape parameters being used in fitting the fitted digital human model from the base digital human model and the three-dimensional reconstructed model; The energy equation is ; wherein, represents the body shape parameters obtaining a function of the model coordinate values after deformation from the base digital human model, represents the coordinate values of the fitted digital human model; the set of body shape parameters is obtained by minimizing the energy equation through the gradient descent method .

2. The 3D body scan based spine generation method of claim 1, wherein, The three-dimensional reconstruction model generated after the target human body is scanned by the 3D human body scanning device is fitted according to the basic digital human model to obtain the fitted digital human model, comprising: The basic digital human model is divided into spine-related parts, and the spine-related parts are aligned with the three-dimensional reconstruction model generated after the target human body is scanned by the 3D human body scanning device; Point cloud fusion technology is used to fuse the point clouds of the spine-related parts to generate a fusion model; The arms and legs of the fusion model are subjected to posture correction so that the arms and legs of the fusion model are in the same position as the arms and legs of the basic digital human model; The point clouds of the posture-corrected fusion model are subjected to fitting operation so that the fusion model approaches the three-dimensional reconstruction model to obtain the fitted digital human model.

3. The spine generation method based on 3D body scan of claim 2, wherein, The basic digital human model is divided into spine-related parts, and the spine-related parts are aligned with the three-dimensional reconstruction model generated after the target human body is scanned by the 3D human body scanning device, comprising: The basic digital human model is divided into shoulder chest, waist, and hip leg parts; Point cloud registration technology is used to align the shoulder chest, waist, and hip leg parts with the three-dimensional reconstruction model generated after the target human body is scanned by the 3D human body scanning device.

4. The 3D body scan based spine generation method of claim 1, wherein, minimizing the energy equation by a gradient descent method to obtain a set of body shape parameters comprising: Given a continuously differentiable function J to be optimized ),in For the body size parameter, provide a learning rate or step size 'a', and provide a set of initial values. =0; computing the gradient of the function J to be optimized ; towards the gradient fastest direction of descent update ; calculating a new gradient of the function J to be optimized ; calculating the gradient of the model to determine whether the cycle needs to be terminated to obtain a set of body shape parameters .

5. The 3D body scan based spine generation method of claim 1, wherein, The basic spine model is subjected to regression processing using the body shape parameters to obtain a preliminary spine model matched with the bone size of the three-dimensional reconstruction model, and the basic digital human model is subjected to deformation processing using the body shape parameters to obtain a digital human model deformed using the body shape parameters, comprising: using the body size parameter performing a regression on a pre-set base spine model to obtain a preliminary spine model matching the bone size of the three-dimensional reconstructed model; using the body shape parameters deforming the base digital human model to obtain a digital human model deformed using the body shape parameters.

6. The 3D body scan based spine generation method according to any of claims 1 to 5, wherein, The preliminary spine model is matched with the position of the bone of the digital human model deformed using the body shape parameters to generate a spine model matched with the spine of the three-dimensional reconstruction model, comprising: The key point information of the digital human model deformed using the body shape parameters is configured by using the preliminary spine model and the fixed topology of the digital human model deformed using the body shape parameters, and the key point information is used to determine the positions of the bones of the digital human model deformed using the body shape parameters; According to the key point information, the bones of different parts of the preliminary spine model which have been determined in size are rotated and translated to appropriate positions to generate a spine model matching the spine of the three-dimensional reconstruction model.

7. The 3D body scan based spine generation method of claim 6, wherein, The configuration of the key point information of the digital human model deformed using the body shape parameters includes: finding the corresponding points of the digital human model deformed using the body shape parameters and the bones at the positions where the skin is thin; and after the ideal spine model positioning points are derived from the digital human model deformed using the body shape parameters, the sum of the coordinate differences between the ideal spine model positioning points and the corresponding positioning points of the preliminary spine model is minimized.

8. A spine generation apparatus based on 3D body scan, characterized by, The spine generation device based on the 3D human body scanning includes: The fitting digital human model generation module is configured to preset a basic digital human model, and fit a three-dimensional reconstruction model generated by a 3D human body scanning device after scanning a target human body, to obtain a fitting digital human model according to the basic digital human model; the basic digital human model has a point cloud topology representing the structure of a human body; The body shape parameter calculation module is configured to calculate a body shape parameter used when the fitting digital human model is fitted from the basic digital human model and the three-dimensional reconstruction model; The body shape parameter using module is configured to use the body shape parameter to perform regression processing on a preset basic spine model, to obtain a preliminary spine model matching the size of the bones of the three-dimensional reconstruction model, and use the body shape parameter to deform the basic digital human model, to obtain a digital human model deformed using the body shape parameters; The spine model generation module is configured to match the positions of the bones of the preliminary spine model and the digital human model deformed using the body shape parameters, to generate a spine model matching the spine of the three-dimensional reconstruction model; The calculation of the body shape parameter used when the fitting digital human model is fitted from the basic digital human model and the three-dimensional reconstruction model includes: setting up an energy equation, minimizing the energy equation to obtain a set of body shape parameters , the body shape parameters being the body shape parameters needed to be used when fitting the fitted digital human model from the base digital human model and the three-dimensional reconstructed model; The energy equation is ; wherein, represents the body shape parameters obtaining a function of the model coordinate values after deformation from the base digital human model, represents the coordinate values of the fitted digital human model; the set of body shape parameters is obtained by minimizing the energy equation through the gradient descent method .

Images