Biomimetic body manufacturing method, device, system, electronic equipment, storage medium and computer program product supporting user remote customization

Abstract

The application belongs to the field of intelligent manufacturing, and discloses a bionic body manufacturing method, device and system supporting remote customization of a user, electronic equipment, a storage medium and a computer program product, solves the problem that the finished product does not meet the expectation due to image transmission distortion in remote customization, and the method comprises the following steps: acquiring customized images (previously obtained by the user by shooting a reference object and a first calibration card by using a first camera) shot from multiple angles; the first calibration card comprises a color calibration card and a skin calibration card; a calibration body is arranged in the color calibration card and / or the skin calibration card, and is used for calibrating an illumination condition and a camera spatial frequency response; for each angle, imaging scene parameters on the user side are generated according to the customized images; cross-scene calibration parameters are determined according to the imaging scene parameters on the user side and the manufacturing side; after the reference object images in the customized images are corrected according to the cross-scene calibration parameters, bionic body 3D modeling is performed according to the reference object images of each angle to manufacture a bionic body.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing technology, specifically relating to a method, apparatus, system, electronic device, storage medium, and computer program product that supports remote customization of bionic bodies by users. Background Technology

[0002] Bionic bodies refer to artificial entities created using advanced materials and manufacturing technologies (such as 3D scanning, additive manufacturing, flexible electronics, etc.) based on 3D data of specific biological individuals (humans or animals). They possess highly biomimetic characteristics in terms of form, structure, and even function. These include prostheses for medical purposes, as well as lifelike models of people or animals for scientific research or artistic purposes.

[0003] Typically, when users customize bionic bodies, they need to provide a reference model to the bionic body manufacturer so that the manufacturer can create the bionic body based on that reference model. For example, when users customize prostheses for medical use, they usually need to come to the laboratory in person to undergo testing, such as testing data on the shape, color, and texture of the skin around the area where the prosthesis will be installed. This allows the manufacturer to manufacture the prosthesis based on the measured data, ensuring that the manufactured prosthesis matches the skin condition of the area where the prosthesis will be installed.

[0004] However, it is sometimes inconvenient to send reference subjects to the manufacturer's laboratory. For example, when customizing prostheses for patients, the patient's residence may be far from the laboratory (sometimes even across national borders) or their mobility may be limited, making it impossible for them to come to the laboratory in person for data measurement. Therefore, for remote users, they can only send a rendering of the reference subject to the manufacturer via the internet, and the manufacturer will manufacture the prosthesis based on this rendering. This creates a problem: when the manufacturer's technicians view the rendering on a monitor or print it out in the laboratory, the effect is often affected by lighting conditions, camera characteristics, and the characteristics of the monitor or photo paper. The effect presented in the laboratory often differs from the effect when the user observes the reference subject with their naked eye. This results in a significant difference in visual appearance between the final manufactured bionic body and the reference subject provided by the user. For example, a prosthesis manufactured by the manufacturer based on a photograph of the patient taken by the user may look significantly different from the patient's body and be unusable. Summary of the Invention

[0005] To address the aforementioned problems in the prior art, this invention provides a method, apparatus, system, electronic device, storage medium, and computer program product that supports remote customization of bionic bodies by users.

[0006] In a first aspect, the present invention provides a method for manufacturing biomimetic bodies that supports remote customization by users, comprising: Step 1: Acquire customized images taken from multiple angles; the customized images are images obtained by the user in advance using a first camera to photograph a reference object and a first calibration card; the reference object is a reference provided by the user for customizing the bionic body; the first calibration card includes a color calibration card and a skin calibration card; the skin calibration card contains multiple skin sample blocks; the multiple skin sample blocks visually present different states; the color calibration card contains multiple color blocks; the multiple color blocks have different colors; the color calibration card and / or the skin calibration card also have a calibration body; the calibration body is used to correct lighting conditions and camera spatial frequency response; Step 2: For each angle, generate imaging scene parameters on the user side based on the customized image; Step 3: For each angle, determine the cross-scene calibration parameters based on the imaging scene parameters of the user side and the manufacturing side; wherein, the imaging scene parameters of the manufacturing side are generated by the manufacturing side using the second camera to take pictures of the second calibration card in advance, and based on the pictures taken; the first calibration card and the second calibration card are the same calibration card or two identical calibration cards; Step 4: For each angle, correct the reference object image in the customized image according to the cross-scene calibration parameters; Step 5: Create a 3D model of the bionic object based on the corrected reference object images from various angles to manufacture the bionic object.

[0007] Secondly, the present invention provides a bionic body manufacturing device that supports remote customization by users. The device includes two sets of modules. The first set of modules includes: a calibration card image receiving module, a calibration card image acquisition module, an illumination parameter acquisition module, a camera spatial response acquisition module, an epidermal parameter acquisition module, a color parameter acquisition module, and a calibration factor calculation module. The second set of modules includes: a reference object image receiving module, a first image conversion module, a second image conversion module, a third image conversion module, a digital three-dimensional image reconstruction module, a three-dimensional prosthesis production CAD data generation module, a three-dimensional prosthesis production module, and a product evaluation and reconstruction module. The calibration card image receiving module is used to receive multi-angle calibration card images captured by the user side. These calibration card images are images obtained by the user in advance using a first camera to capture images of the first calibration card. The first calibration card includes a color calibration card and a skin calibration card. The skin calibration card contains multiple skin sample blocks. The multiple skin sample blocks visually present different states. The color calibration card contains multiple color blocks. The multiple color blocks have different colors. The color calibration card and / or the skin calibration card also include a calibration body. The calibration body is used to correct lighting conditions and camera spatial frequency response. The calibration card image acquisition module is used to acquire multi-angle calibration card images taken by the manufacturing side. These calibration card images are obtained by the manufacturing side using a second camera to capture images of the second calibration card. The first calibration card and the second calibration card are either the same calibration card or two identical calibration cards. The illumination parameter acquisition module is used to extract the calibration body shadow images of the user side and the manufacturing side from the calibration card images taken by the user side and the manufacturing side, and input them into the first neural network model to obtain the illumination parameters of the user side and the manufacturing side respectively. The camera spatial response acquisition module is used to extract calibration volume images from calibration card images captured by the user side and the manufacturing side, perform two-dimensional Fourier transform on them, and obtain the camera spatial frequency response of the user side and the manufacturing side. The epidermal parameter acquisition module is used to extract epidermal calibration card images from calibration card images taken from the user side and the manufacturing side, and input them into the second neural network model to obtain epidermal state parameters from the user side and the manufacturing side respectively. The color parameter acquisition module is used to extract color calibration card images from calibration card images taken from the user side and the manufacturing side, and input them into the third neural network model to obtain color parameters from the user side and the manufacturing side, respectively. The calibration factor calculation module is used to calculate the illumination condition calibration factor, color calibration factor, skin state calibration factor, and spatial frequency response calibration factor according to the illumination parameters, color parameters, skin state parameters, and camera spatial frequency response on the user side and the manufacturing side, respectively. The reference object image receiving module is used to receive multi-angle reference object images captured by the user; the reference object is a reference provided by the user for customizing the bionic body. The first image conversion module is used to convert the illumination parameters of the reference object image on the user side using the illumination condition calibration factor; The second image conversion module is used to perform camera spatial frequency response conversion on the reference object image on the user side using the spatial frequency response calibration factor; The third image conversion module is used to convert the color and skin state of the reference object image on the user side using the color calibration factor and the skin state calibration factor. The digital 3D image reconstruction module is used to perform bionic 3D modeling using multi-angle converted reference object images to form 3D structural data and surface data. The 3D prosthesis production CAD data generation module is used to input the 3D structural data and surface data into the fourth neural network model to obtain CAD data for manufacturing bionic bodies. The three-dimensional prosthesis production module is used to manufacture bionic bodies using the CAD data through CAD production methods; The product evaluation and reconstruction module is used to evaluate the quality of the bionic body. If the evaluation result does not meet the preset quality standard, the digital three-dimensional image reconstruction module is triggered to optimize the 3D modeling to achieve iterative improvement of the bionic body until the evaluation result meets the quality standard.

[0008] Thirdly, the present invention provides a bionic manufacturing system that supports remote customization by users, comprising: at least one calibration card and a computer program product; The calibration card includes a color calibration card and an epidermal calibration card; the epidermal calibration card contains multiple epidermal sample blocks; the multiple epidermal sample blocks visually present different states; the color calibration card contains multiple color blocks; the multiple color blocks have different colors; the color calibration card and / or the epidermal calibration card also include a calibration body; the calibration body is used to correct lighting conditions and camera spatial frequency response; when the computer program product is run on a computer, the computer executes the steps of the above-described bionic body manufacturing method that supports remote customization by the user.

[0009] Fourthly, the present invention provides an electronic device including an input / output unit, a processing unit, a storage unit, a control unit, and a communication unit; the input / output unit, the processing unit, the storage unit, the control unit, and the communication unit are connected to an address bus, a communication bus, and a data / control bus; when a computer program in the storage unit is executed, the steps of the above-mentioned bionic body manufacturing method supporting remote customization by the processor in the processing unit are implemented.

[0010] Fifthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for manufacturing bionic bodies that supports remote customization by a user.

[0011] In a sixth aspect, the present invention provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the steps of the above-described bionic body manufacturing method that supports remote customization by the user.

[0012] This invention provides a method, apparatus, system, electronic device, storage medium, and computer program product for supporting remote customization of bionic bodies. By introducing a calibration card containing color, skin samples, and a calibration body, and photographing it together with a user reference object, and generating imaging scene parameters for both the user and manufacturing sides based on the captured images, and then calculating cross-scene calibration parameters to correct the user-side image, a traceable visual transmission chain is constructed from the user's uncontrolled environment to the standard environment of the manufacturing end. This fundamentally eliminates cross-scene imaging distortion caused by differences in lighting, cameras, and display devices, enabling the final manufactured bionic body to faithfully reproduce the true visual appearance of the user reference object under standard observation conditions. This solves the problem of serious discrepancies between the finished product and expectations caused by image transmission distortion in remote customization.

[0013] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0014] Figure 1 This is a flowchart of a bionic body manufacturing method that supports remote customization by users, provided by the present invention. Figure 2 The image illustrates a scenario where a user takes a custom image using a first camera. Figure 3 A color calibration card is schematically shown in the image. Figure 4 The image schematically illustrates a skin calibration card; Figure 5 A calibration body is schematically shown in the figure; Figure 6 yes Figure 5 Top view of the calibration body in the diagram; Figure 7 The diagram schematically illustrates the overall flow of the bionic body manufacturing method supporting remote customization provided by the present invention; Figure 8 The diagram illustrates Figure 7 The flowchart for correcting the epidermal state during the process of correcting the reference object image; Figure 9 The diagram illustrates Figure 7 The flowchart for color correction during the correction of the reference object image; Figure 10 This is a schematic diagram of the structure of a bionic body manufacturing device that supports remote customization by users, provided by the present invention. Figure 11 This is a schematic diagram of the structure of an electronic device provided by the present invention. Detailed Implementation

[0015] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0016] To address the problem of significant discrepancies between the final product and the expected result caused by image transmission distortion during remote customization of bionic bodies, this invention provides a bionic body manufacturing method that supports remote customization by users. (See [link to relevant documentation]). Figure 1 The method includes the following steps: Step 1: Acquire customized images taken from multiple angles; these customized images are images obtained by the user in advance using a first camera to photograph a reference object and a first calibration card; the reference object is a reference provided by the user for the customized bionic body; the first calibration card includes a color calibration card and a skin calibration card; the skin calibration card contains multiple skin sample blocks; the multiple skin sample blocks visually present different states; the color calibration card contains multiple color blocks; the multiple color blocks have different colors; the color calibration card and / or the skin calibration card also have a calibration body; this calibration body is used to correct lighting conditions and camera spatial frequency response.

[0017] Specifically, before proceeding to step 1, the manufacturer will mail the first calibration card to the user so that the user can take a customized image. Then, the user will return the customized image to the manufacturer via email or other means. The manufacturer will then import the customized image sent by the user into a computer or other electronic device capable of running computer programs, thereby executing step 1 and subsequent steps.

[0018] Alternatively, when taking custom images, users can photograph the reference object, color calibration card, and skin calibration card together, placing all three in the same custom image. Or, users can photograph the reference object, color calibration card, and skin calibration card separately, ensuring consistent shooting conditions; this will create three separate images that together constitute the user's custom image. Alternatively, users can photograph the reference object and color calibration card (or skin calibration card) together, and then photograph the remaining calibration card; this will create two images for the custom image, again requiring consistent shooting conditions each time.

[0019] In this invention, the user-customized bionic body can be a prosthesis for parts such as the face, ear, nose, and fingers. The corresponding reference object is provided by the user. For example, when the prosthesis is used to replace a damaged left ear, the user's right ear can be used as a reference object. Of course, the bionic body in this invention is not limited to the above-mentioned prostheses for medical purposes; it can also be a customized humanoid robot, the skin / appearance covering of an animal robot, or an animal toy, etc.

[0020] In this invention, the color calibration card contains multiple color blocks. Figure 3The example shown is a color calibration card containing eight color blocks, including five standard colors (red, green, blue, white, and black) and three skin colors (skin color1-skin color3). These three skin colors can be, for example, brown, yellow, and pink, but are not limited to these. The central area surrounded by the eight color blocks is used to fix the calibration object. Of course, this invention does not limit the number of color blocks or the specific colors included in the color calibration card.

[0021] In this invention, the epidermal calibration card contains multiple epidermal sample blocks that visually present different states. Specifically, these epidermal sample blocks exhibit visual differences in various aspects such as skin color (dark or light), diffuse reflectance (dry or oily), and / or texture (smooth or wrinkled). Figure 4 The example illustrates an epidermal calibration card containing eight epidermal sample blocks, representing eight different skin conditions such as oiliness, wrinkles, and dullness. The central area surrounded by the eight epidermal sample blocks is used to fix the calibration body. Of course, this invention does not limit the number or specific colors of the color blocks included in the epidermal calibration card.

[0022] In this invention, the calibration body can exist in various forms. For example, in one implementation, the calibration body can be as follows: Figure 5 and Figure 6 The structure shown includes: a base 1, a hollow cube 2 located above the base 1, and a translucent sphere 3 located above the cube 2. The surface of the cube 2 is a polarization surface composed of gratings, with different grating spacings for different polarization surfaces. The 12 edges of the cube 2 are composed of rods of varying thicknesses.

[0023] It should be noted that the calibration body used in this invention is primarily for using its shadow to correct lighting conditions. Additionally, it can be used to correct the camera's spatial frequency response. Subsequent sections will discuss the use of this calibration body to calibrate lighting conditions and the camera's spatial frequency response.

[0024] In another implementation, the cube 2 described above can be replaced with a pyramid or cone, and gratings with different properties can be arranged on its inclined sides. This type of calibration body can also be used to calibrate lighting conditions and camera spatial frequency response.

[0025] Figure 2 The diagram illustrates a user taking a customized image using the first camera. In actual customization, when the user takes pictures of the reference object and the first calibration card using the first camera, the pictures can be taken from angles such as 0°, ±30°, ±45°, and ±60° from the front of the reference object, but are not limited to these angles.

[0026] Step 2: For each angle, generate imaging scene parameters on the user side based on the customized image.

[0027] Here, imaging scene parameters include illumination parameters, camera spatial frequency response, color parameters, and / or skin state parameters. Illumination parameters characterize the inherent properties of the light source (color temperature, spectrum, polarization) and its illumination pattern in three-dimensional space (illumination direction, illumination range, uniformity). Skin state parameters quantify the macroscopic visual and perceptual characteristics of the skin (such as sheen, smoothness, and color). Skin state parameters can be defined as functions of physical parameters. These physical parameters include geometric and optical parameters; geometric parameters include surface roughness (arithmetic mean roughness or root mean square roughness) and autocorrelation length, etc.; optical parameters include diffuse reflectance, scattering coefficient, surface haze, gloss, absorption coefficient, brightness, and contrast, etc. These parameters are all clearly defined physical parameters in the prior art. Camera spatial frequency response is the same as that referred to in imaging optics and computer vision; color parameters, as the name suggests, are parameters characterizing color.

[0028] For example, when the imaging scene parameters include four types: illumination parameters, camera spatial frequency response, color parameters, and skin state parameters, step 2 specifically includes: For each angle, the calibration body image, calibration body shadow image, color calibration card image, skin calibration card image, and reference object image are extracted from the customized image. The illumination parameters on the user side are generated based on the calibration body shadow image, the color parameters on the user side are generated based on the color calibration card image, and the skin state parameters on the user side are generated based on the skin calibration card image. The spatial frequency domain distribution of the calibration body image is calculated as the camera spatial frequency response on the user side.

[0029] In step 2, generating user-side illumination parameters based on the calibration body shadow image includes: inputting the calibration body shadow image into the first neural network model so that the first neural network model outputs illumination parameters as user-side illumination parameters. For example, the first neural network model will output the following illumination parameters: illuminance, light source direction (latitude and longitude), light source diffusion type (such as point light source of lamps, line light source of fluorescent lamps or area light source of indirect lighting), light source color (such as daylight white, warm yellow of lamplight or cool white of fluorescent lamps), and polarization state (such as linearly polarized light under blue sky or unpolarized light of artificial lighting).

[0030] In this invention, the first neural network model is trained based on multiple first training samples; the first training samples include: calibration body shadow image samples (i.e., calibration body shadow images used as training samples) and the illumination parameters corresponding to the calibration body shadow image samples; the calibration body shadow image samples are calibration body shadow images taken under illumination conditions with known illumination parameters, and the illumination parameters are the illumination parameters corresponding to the calibration body shadow image samples; different calibration body shadow image samples correspond to different illumination conditions.

[0031] In practical applications, the first neural network model can be obtained by the manufacturer constructing and training the first training samples. Alternatively, when deploying the method of this invention in a computer imaging product, the first neural network model can also be pre-trained and deployed in the computer program product by the computer program product provider; both are acceptable.

[0032] In step 2, epidermal state parameters on the user side are generated based on the epidermal calibration card image, including: (1) Extract each first sub-block from the epidermal calibration card image; the first sub-block is the image of the epidermal sample block (that is, the image of the epidermal sample block in the epidermal calibration card image extracted from the customized image).

[0033] (2) Input each first sub-block into the second neural network model so that the second neural network model outputs the epidermal state parameters corresponding to each first sub-block.

[0034] Here, the second neural network model is trained based on multiple second training samples; the second training samples include: first sub-block samples (i.e., the first sub-block as training samples) and the epidermal state parameters corresponding to the first sub-block samples; the first sub-block samples in the multiple second training samples include: the original first sub-block obtained by photographing the epidermal sample block, and the synthetic first sub-block generated based on the original first sub-block using data augmentation.

[0035] In practical applications, the second neural network model can be obtained by the manufacturer constructing and training the second training samples. Alternatively, when deploying the method of this invention in a computer program product, the second neural network model can also be pre-trained by the provider of the computer program product and deployed in the computer imaging product; both are acceptable.

[0036] (3) Determine the skin state parameters on the user side based on the skin state parameters corresponding to each first sub-block.

[0037] Understandably, after step (2) is completed, for each angle, the second neural network model will output corresponding skin state parameters for each first sub-block at that angle. Therefore, it is necessary to comprehensively determine the user-side skin state parameters at the current angle based on the skin state parameters corresponding to each first sub-block. For example, in one implementation, the set of skin state parameters corresponding to each first sub-block can be directly used as the user-side skin state parameters at the current angle. In another implementation, the skin state parameters corresponding to each first sub-block can be further integrated and calculated to form the user-side skin state parameters.

[0038] For example, suppose the epidermal condition parameters include: the oiliness index corresponding to the diffuse reflectance of the epidermis, the wrinkle depth corresponding to the epidermal texture, and the skin darkness corresponding to the skin tone. These three parameters are defined as follows: Oiliness index = a × diffuse reflectance + b × gloss + c × scattering coefficient; Wrinkle depth = d × surface roughness + e × diffuse reflectance + f × autocorrelation length; Skin darkness = g × lightness + h × contrast + i × absorption coefficient.

[0039] Where a~i are the weighting coefficients of each physical parameter.

[0040] Accordingly, determining the user-side epidermal state parameters based on the epidermal state parameters corresponding to each first sub-block can include: identifying first sub-blocks related to epidermal shine from the epidermal sample blocks corresponding to each first sub-block, extracting an oiliness index from the epidermal state parameters corresponding to these first sub-blocks, and finally calculating a parameter value that can measure the diffuse reflectivity of the epidermis based on the extracted oiliness index; identifying first sub-blocks related to epidermal smoothness from the epidermal sample blocks corresponding to each first sub-block, extracting wrinkle depth from the epidermal state parameters corresponding to these first sub-blocks, and finally calculating a parameter value that can measure the texture of the epidermis based on the extracted wrinkle depth; and identifying first sub-blocks related to epidermal color from the epidermal sample blocks corresponding to each first sub-block, extracting skin darkness from the epidermal state parameters corresponding to these first sub-blocks, and finally calculating a parameter value that can measure skin color based on the extracted skin darkness.

[0041] by Figure 4The skin calibration card shown is an example. It corresponds to eight sub-blocks. The average oiliness index of the four sub-blocks with an epidermal shine of "dry" is taken to obtain the dryness value; the average oiliness index of the four sub-blocks with an epidermal shine of "oily" is taken to obtain the oiliness value; the average wrinkle depth of the four sub-blocks with an epidermal smoothness of "wrinkled" is taken to obtain the wrinkle depth; the average wrinkle depth of the four sub-blocks with an epidermal smoothness of "smooth" is taken to obtain the smoothness value; the average wrinkle depth of the four sub-blocks with an epidermal color of "dark skin" is taken to obtain the depth value; and the average wrinkle depth of the four sub-blocks with an epidermal color of "light skin" is taken to obtain the shallowness value. Thus, a set of epidermal state parameters is obtained to measure diffuse reflectivity (dryness value, oiliness value), texture (wrinkle depth, smoothness value), and skin color (depth value, shallowness value).

[0042] It is understandable that when the epidermal state parameters directly output by the second neural network model include the oiliness index, wrinkle depth, and skin darkness mentioned above, the epidermal state parameters corresponding to the first sub-block sample during the training of the second neural network model can be calculated using the definitions of these three parameters.

[0043] In step 2, user-side color parameters are generated based on the color calibration card image, including: (a) Extract each second sub-block from the color calibration card image; the second sub-block is an image of the color block (i.e., an image of the color block in the color calibration card image extracted from the custom image).

[0044] (b) Input each second sub-block image into the third neural network model so that the third neural network model outputs the color parameters corresponding to each second sub-block.

[0045] Here, the color parameter is not necessarily the original RGB value, L*a*b* value or HVC value of the pixel in the second sub-block image, but an equivalent color value after being understood and normalized by the third neural network model.

[0046] The third neural network model is trained based on multiple third training samples. The third training samples include: second sub-block samples (i.e., second sub-blocks used as training samples) and color parameters corresponding to the second sub-block samples. The second sub-block samples in the multiple third training samples include: the original second sub-blocks obtained by photographing the color blocks, and the synthetic second sub-blocks generated by data augmentation based on the original second sub-blocks.

[0047] In practical applications, the third neural network model can be obtained by the manufacturer constructing and training the third training samples. Alternatively, when deploying the method of this invention in a computer program product, the third neural network model can also be pre-trained by the provider of the computer program product and deployed in the computer imaging product; both are acceptable.

[0048] (c) Determine the color parameters on the user side based on the color parameters corresponding to each second sub-block.

[0049] Understandably, after step (b) is completed, for each angle, the third neural network model will output the corresponding color parameters for each second sub-block at that angle. Therefore, it is necessary to comprehensively determine the user-side color parameters at the current angle based on the color parameters corresponding to each second sub-block. Here, the set of color parameters corresponding to each second sub-block can be directly used as the user-side color parameters at the current angle.

[0050] In step 2, the spatial frequency domain distribution of the calibration volume image is calculated as the spatial frequency response of the camera on the user side. Specifically, the calibration volume image can be transformed into the spatial frequency domain through a two-dimensional Fourier transform, and the resulting spectrum distribution can be used as the spatial frequency response of the camera on the user side.

[0051] Understandably, after performing step 2, each angle will obtain its corresponding lighting parameters, skin condition parameters, and color parameters.

[0052] Step 3: For each angle, determine the cross-scene calibration parameters based on the imaging scene parameters of the user side and the manufacturing side; wherein, the imaging scene parameters of the manufacturing side are generated by the manufacturing side using the second camera to take pictures of the second calibration card in advance, and based on the pictures taken; the first calibration card and the second calibration card are the same calibration card or two identical calibration cards.

[0053] It is understandable that when the first and second calibration cards are the same, the user and the manufacturer can exchange and use the calibration card via mail. When the first and second calibration cards are two identical calibration cards, these two calibration cards are usually from the same production batch.

[0054] In step 3, specifically for each angle, the illumination condition calibration factor, color calibration factor, skin state calibration factor, and spatial frequency response calibration factor are calculated based on the illumination parameters, color parameters, skin state parameters, and camera spatial frequency response of the user side and the manufacturing side, respectively, and used as cross-scene calibration parameters.

[0055] Specifically, for each angle, the ratio of illumination parameters on the user side to those on the manufacturing side is calculated. This ratio is the illumination condition calibration factor. When there are multiple illumination parameters, the ratios between parameters of the same type are calculated, resulting in a series of ratios that constitute the illumination condition calibration factor. Similarly, the ratio of color parameters on the user side to those on the manufacturing side is calculated to obtain the color calibration factor. This ratio is calculated between color parameters of corresponding color patches on both the user side and the manufacturing side. The ratio of epidermal condition parameters on the user side to those on the manufacturing side is calculated to obtain the epidermal condition calibration factor. For example, when the illumination parameters include oiliness index, wrinkle depth, and skin darkness, the ratios of oiliness index, wrinkle depth, and skin darkness on the user side and the manufacturing side are calculated. These three ratios constitute the epidermal condition calibration factor. The ratio of the camera spatial frequency response (spectral distribution) on the user side and the manufacturing side is calculated to obtain the spatial frequency response calibration factor, wherein the ratio between the same frequencies in the spectral distribution on the user side and the manufacturing side is calculated accordingly.

[0056] Step 4: For each angle, correct the reference object image in the customized image according to the cross-scene calibration parameters.

[0057] Specifically, in step 4, steps 4-1 to 4-4 are performed for each angle: Step 4-1: Correct the lighting conditions of the reference object image according to the lighting condition calibration factor.

[0058] Specifically, the reference object image is segmented into several first sub-regions; each first sub-region is input into a first neural network model so that the first neural network model outputs the illumination parameters corresponding to the first sub-region, and the illumination conditions corresponding to the first sub-region are corrected to the illumination conditions on the manufacturing side according to the illumination condition calibration factor.

[0059] Here, the lighting conditions corresponding to the first sub-region are corrected to the manufacturing-side lighting conditions based on the lighting condition calibration factor. Specifically, this can be achieved by dividing or multiplying the lighting parameters corresponding to the first sub-region by the lighting condition calibration factor. If the lighting parameters include multiple parameters, each parameter is divided or multiplied by a different lighting condition calibration factor of the same type. For example, the color temperature in the lighting parameters corresponds to the lighting condition calibration factor formed by dividing or multiplying by the ratio of the color temperatures on the user side and the manufacturing side. Then, the reference object image is adjusted in image processing software (such as Microsoft Photo, Microsoft Paint, Adobe Photoshop, or other common general-purpose image processing tools) to ensure that its lighting parameters meet the corrected lighting parameters, thus achieving the correction of the lighting conditions of the reference object image.

[0060] Understandably, if the lighting condition calibration factor is obtained by dividing the lighting parameters on the user side by the lighting parameters on the manufacturing side, then in step 4-1, the illumination parameters corresponding to the first sub-region are divided by the lighting condition calibration factor. Otherwise, if the lighting condition calibration factor is obtained by dividing the lighting parameters on the manufacturing side by the lighting parameters on the user side, then in step 4-1, the illumination parameters corresponding to the first sub-region are multiplied by the lighting condition calibration factor.

[0061] Step 4-2: Perform camera spatial frequency response correction on the reference object image based on the spatial frequency response calibration factor.

[0062] Specifically, in the spatial frequency domain, the camera spatial frequency response (spectral distribution) of the reference object image is divided or multiplied by a spatial frequency response calibration factor. Then, the corrected spectral distribution is subjected to an inverse Fourier transform, thus achieving the correction of the camera spatial frequency response of the reference object image.

[0063] Step 4-3: Correct the epidermal condition of the reference object image according to the epidermal condition calibration factor.

[0064] Specifically, the reference object image is segmented into several second sub-regions; each second sub-region is input into a second neural network model so that the second neural network model outputs the epidermal state parameters corresponding to the second sub-region, and the epidermal state of the second sub-region is corrected to the display conditions on the manufacturing side according to the epidermal state calibration factor.

[0065] Here, the skin state of the second sub-region is corrected to the display conditions on the manufacturing side according to the skin state calibration factor. Specifically, this can be similar to step 4-1, where the skin state parameters of the second sub-region are divided or multiplied by the skin state calibration factor. Then, the reference image is adjusted in image processing software (such as Microsoft Photo, Microsoft Paint, Adobe Photoshop, or other common general-purpose image processing tools) so that its skin state parameters meet the corrected skin state parameters, thus achieving the correction of the skin state of the reference image.

[0066] Step 4-4: Perform color correction on the reference object image according to the color calibration factor.

[0067] Specifically, the reference object image is segmented into several third sub-regions; each third sub-region is input into a third neural network model so that the third neural network model outputs the color parameters corresponding to the third sub-region, and the color of the third sub-region is corrected to the display conditions on the manufacturing side according to the color calibration factor.

[0068] Here, the color of the third sub-region is corrected to the display conditions on the manufacturing side according to the color calibration factor. Specifically, this can be similar to step 4-1, where the color parameters of the third sub-region are divided or multiplied by the skin state calibration factor. Then, the reference image is adjusted in image processing software (such as Microsoft Photo, Microsoft Paint, Adobe Photoshop, or other common general-purpose image processing tools) so that its color parameters meet the corrected color parameters, thus achieving the color correction of the reference image.

[0069] Step 5: Create a 3D model of the bionic object based on the corrected reference object images from various angles to manufacture the bionic object.

[0070] Step 5 specifically includes: Step 5-1: Construct a 3D structure based on the reference object images from various angles, and generate surface data for the 3D structure to obtain 3D structure data and surface data.

[0071] The surface data includes the color and skin state of the application area corresponding to the 3D structure. It can be understood that the color of the application area can be formed based on the color parameter distribution of the corrected reference object image, and the skin state of the application area can be formed based on the skin state parameter distribution of the corrected reference object image.

[0072] It is understandable that the reference object image contains the outline of the reference object, so the 3D structure of the biomimetic body can be constructed based on the outline information contained in the reference object images from various angles. Furthermore, since the lighting conditions, camera spatial frequency response, skin condition, and color of the reference object images from various angles have been corrected to the manufacturing conditions, the surface data for the 3D structure can be directly formed based on the corrected reference object images.

[0073] Step 5-2: Input the 3D structural data and surface data into the fourth neural network model so that the fourth neural network model outputs CAD data for manufacturing the bionic body.

[0074] The fourth neural network is trained based on historical data, which includes 3D structural data, surface data, and corresponding CAD data used in previous bionic body manufacturing processes. Furthermore, the historical data preferably consists of 3D structural data, surface data, and corresponding CAD data of non-remotely customized bionic bodies.

[0075] After using the fourth neural network model to output CAD data for manufacturing bionics, bionics can be manufactured according to the CAD data.

[0076] The present invention provides a bionic manufacturing method that supports remote customization by introducing a calibration card containing color, skin samples, and calibration bodies, which is photographed together with a user reference object. Based on the captured images, imaging scene parameters on the user side and the manufacturing side are generated, and cross-scene calibration parameters are calculated to correct the user-side image. This constructs a traceable visual transmission chain from the user's uncontrolled environment to the standard environment of the manufacturing end, fundamentally eliminating cross-scene imaging distortion caused by differences in lighting, cameras, and display devices. This allows the final manufactured bionic object to reproduce the true visual appearance of the user reference object with high fidelity under standard observation conditions, solving the problem of serious discrepancies between the finished product and expectations caused by image transmission distortion in remote customization.

[0077] In one embodiment, before the final bionic body is manufactured, the bionic body manufacturing method supporting remote customization provided by the present invention further includes the following steps: i. Acquire biomimetic images at least once from multiple angles (the same multiple angles used when taking custom images); the biomimetic images are obtained by the manufacturing side using a second camera to capture images of the biomimetic being manufactured.

[0078] ii. For each angle, compare the differences in shape between the bionic object and the reference object in the manufacturing process, based on the bionic image and the corrected reference object image.

[0079] Understandably, by comparing the outlines of the bionic object and the reference object in the bionic image and the corrected reference object image, it is possible to determine whether there are any differences in shape between the two.

[0080] iii. If the shape difference does not meet the shape error requirements, then optimize the 3D modeling of the biomimetic body based on the shape difference.

[0081] Specifically, if the shape difference does not meet the shape error requirements, the corresponding displacement field of pixels between the two images can be calculated using an optical flow algorithm to obtain a difference map. This difference map is then applied inversely to the image of the bionic body in the bionic body 3D modeling software. Image distortion technology can be used to adjust the contour of the image so that the shape difference between it and the contour of the reference object meets the error requirements.

[0082] iv. If the shape difference meets the shape error requirements, then further compare the differences in color and surface condition between the bionic body and the reference object during manufacturing based on the bionic body image and the corrected reference object image.

[0083] Here, based on the bionic image and the corrected reference object image, the differences in color and skin condition between the bionic image and the reference object during manufacturing can be compared. This can be achieved by segmenting the bionic image and the corrected reference object image into small blocks, and then inputting the segmented blocks into the second neural network model and the third neural network model. Based on the color parameters and skin condition parameters output by the models, it can be determined whether there are differences in color and skin condition.

[0084] vi. If the differences in color and epidermal condition do not meet the requirements for color and skin error, then optimize the 3D modeling of the bionic body based on the differences in color and epidermal condition.

[0085] Specifically, if the differences in color and skin condition do not meet the requirements for color and skin error, the ratio or difference between the color parameters and skin condition parameters of each block is given to the bionic 3D modeling software, so that the software can re-render the color and skin condition of the bionic body based on this ratio or difference, thereby optimizing the bionic 3D modeling.

[0086] In this embodiment, during the manufacturing process of the bionic body, the images of the actual bionic body taken are compared with the reference images multiple times, and the 3D model is dynamically corrected based on the quantitative differences in shape, color and skin condition. This ensures that the final manufactured bionic body can highly match the user's expectations in terms of shape and appearance, significantly improving the customization accuracy and success rate.

[0087] In a specific example, the bionic body manufacturing method provided by this invention, which supports remote customization by the user, is used to manufacture a prosthesis for a patient. The process is as follows: Figure 7 As shown. In Figure 7 In this context, the "reference object image" refers to the image of the reference object provided by the patient for the customized prosthesis. For example, the reference object could be the patient's normal limb. The process of "correcting the reference object image at each angle" can be found in steps 2-4 above. Figure 8 , Figure 9 .

[0088] The method provided in this invention can be applied to electronic devices. Specifically, the electronic device can be a desktop computer, a portable computer, a server, etc. No limitation is made herein; any electronic device that can implement this invention falls within the protection scope of this invention.

[0089] Based on the same inventive concept as the aforementioned biomimetic manufacturing method supporting remote user customization, this invention also provides a biomimetic manufacturing apparatus supporting remote user customization. See also... Figure 10The device comprises two sets of modules. The first set of modules includes: a calibration card image receiving module 10, a calibration card image acquisition module 20, an illumination parameter acquisition module 30, a camera spatial response acquisition module 40, a skin parameter acquisition module 50, a color parameter acquisition module 60, and a calibration factor calculation module 70. The second set of modules includes: a reference object image receiving module 11, a first image conversion module 21 for realizing illumination parameter conversion, a second image conversion module 31 for realizing camera spatial frequency response conversion, a third image conversion module 41 for realizing color and skin state conversion, a digital 3D image reconstruction module 51, a 3D prosthesis production CAD data generation module 61, a 3D prosthesis production module 71, and a product evaluation and reconstruction module 81 for realizing bionic body iterative improvement.

[0090] In the first group of modules mentioned above, the calibration card image receiving module 10 receives calibration card images taken from multiple angles by the user, such as multi-angle calibration card images taken by the user using a smartphone or other ordinary camera under home lighting conditions. These calibration card images are images obtained by the user beforehand by taking pictures of the first calibration card using a first camera. The first calibration card includes a color calibration card and a skin calibration card. The skin calibration card contains multiple skin sample blocks. The multiple skin sample blocks present different visual states. The color calibration card contains multiple color blocks. The multiple color blocks have different colors. The color calibration card and / or the skin calibration card are also provided with a calibration body. The calibration body is used to correct lighting conditions and camera spatial frequency response.

[0091] The calibration card image acquisition module 20 acquires multi-angle images of the calibration card taken from the manufacturing side (such as when laboratory experts take images using a high-fidelity camera under laboratory lighting conditions).

[0092] The illumination parameter acquisition module 30 extracts the calibration body shadow images of the user side and the manufacturing side from the calibration card images taken by the user and the manufacturing side, and inputs them into the first neural network model to obtain the illumination parameters of the user side and the manufacturing side (laboratory) respectively.

[0093] The camera spatial response acquisition module 40 extracts calibration volume images from calibration card images taken by the user and manufacturing sides, performs two-dimensional Fourier transform on them, and obtains the camera spatial frequency response of the user and manufacturing sides.

[0094] The epidermal parameter acquisition module 50 extracts epidermal calibration card images from calibration card images taken by the user and manufacturing sides, and inputs them into the second neural network model to obtain epidermal state parameters from the user and manufacturing sides respectively.

[0095] The color parameter acquisition module 60 extracts the color calibration card images of the user side and the manufacturing side from the calibration card images taken by the user and the manufacturing side, and inputs them into the third neural network model to obtain the color parameters of the user side and the manufacturing side respectively.

[0096] The calibration factor calculation module 70 calculates the illumination condition calibration factor, color calibration factor, skin state calibration factor, and spatial frequency response calibration factor based on the illumination parameters, color parameters, skin state parameters, and camera spatial frequency response from the user side and the manufacturing side, respectively.

[0097] In the second group of modules mentioned above, the reference object image receiving module 11 receives multi-angle reference object images captured by the user, such as multi-angle reference object images captured by the user in advance (under home lighting conditions) using a smartphone or other ordinary camera.

[0098] The first image conversion module 21 uses the illumination condition calibration factor obtained by the calibration factor calculation module 70 to convert the illumination parameters of the reference object image on the user side into an image taken under laboratory conditions.

[0099] Then, the second image conversion module 31 uses the spatial frequency response calibration factor obtained by the calibration factor calculation module 70 to perform camera spatial frequency response conversion on the current reference object image, converting it into an image taken under laboratory conditions.

[0100] Then, the third image conversion module 41 uses the color calibration factor and epidermal state calibration factor obtained by the calibration factor calculation module 70 to perform color and epidermal state conversion on the reference object image on the user side, converting it into an image taken under laboratory conditions.

[0101] This allows for the acquisition of reference object images of laboratory photographic quality from multiple perspectives. Then, the digital 3D image reconstruction module 51 uses the corrected reference object images from multiple angles to perform biomimetic 3D modeling, generating 3D structural data and surface data. Due to the correction processing, the reconstructed 3D structural surface can highly reproduce the morphology of a high-quality laboratory-grade reference object (such as a patient's body surface morphology).

[0102] Then, the 3D prosthesis production CAD data generation module 61 inputs the 3D structural data and surface data into the fourth neural network model to obtain CAD data for manufacturing the bionic body; then, the 3D prosthesis production module 71 uses the CAD data generated by the 3D prosthesis production CAD data generation module 61 to manufacture the bionic body through CAD production.

[0103] Then, the product evaluation and reconstruction module 81 evaluates the quality of the manufactured bionic body. If the evaluation result does not meet the preset quality standard, the digital 3D image reconstruction module is triggered to optimize the 3D modeling to achieve iterative improvement of the bionic body until the evaluation result meets the quality standard. Here, the method by which the product evaluation and reconstruction module 81 evaluates the quality of the manufactured bionic body can be referred to the method embodiment described above for evaluating the differences between the bionic body and the reference object in terms of shape, color, and skin condition, and will not be repeated here.

[0104] It should be noted that, for the embodiments of the device / electronic device / storage medium / computer program product, since each module is basically similar to the steps of the method embodiment, the description is relatively simple, and relevant parts can be referred to in the description of the method embodiment.

[0105] It is understandable that the aforementioned bionic body manufacturing device that supports remote customization by users can actually be a computer program within a bionic body manufacturing software that supports remote customization by users, and each of its modules is a program module within that computer program.

[0106] Based on the same inventive concept as the aforementioned bionic body manufacturing method supporting remote user customization, this invention also provides a bionic body manufacturing system supporting remote user customization, comprising: at least one calibration card and a computer program product. The calibration card includes a color calibration card and a skin calibration card; the skin calibration card contains multiple skin sample blocks; the multiple skin sample blocks visually present different states; the color calibration card contains multiple color blocks; the multiple color blocks have different colors; the color calibration card and / or the skin calibration card also include a calibration body; the calibration body is used to correct lighting conditions and camera spatial frequency response; when the computer program product is run on a computer, the computer executes the steps of the aforementioned bionic body manufacturing method supporting remote user customization.

[0107] Optionally, the system further includes: a lighting device; the lighting device is used to provide illumination to the manufacturing side so that all imaging activities related to the manufacturing of the bionic body are carried out under uniform lighting conditions from the initial shooting of the calibration card to form the imaging scene parameters of the manufacturing side until the completion of the bionic body manufacturing process.

[0108] Optionally, the system further includes: a photographing device and a 3D printing device. The photographing device is used to perform photographing tasks on the manufacturing side, such as photographing the second calibration card, photographing the bionic body, etc. The 3D printing device is used to manufacture the bionic body based on CAD data.

[0109] This invention also provides an electronic device, such as... Figure 11As shown, it includes: an input / output unit 101, a processing unit 102, a storage unit 103, a control unit 104, and a communication unit 105. These are connected to an address bus 201, a communication bus 202, and a data / control bus 203. When the program in the storage unit 103 is executed, any step in the above method embodiment is implemented by the processor in the processing unit 102, enabling the user to remotely customize the bionic body.

[0110] The input / output unit 101 may include an image acquisition device, a two-dimensional or three-dimensional display, etc. The processing unit 102 may include a central processing unit, a graphics processing unit, a network processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), etc. The storage unit 103 may include random access memory (RAM), read-only memory (ROM), non-volatile memory (NROM), and optical storage media such as CDs, DVDs, and Blu-ray discs. The control unit 104 may include an electronic control unit, a microcontroller unit, an image / audio control unit, a motor control unit, an input / output control unit, etc. The communication unit 105 may include an Ethernet communication unit, an LTE communication unit, a Wi-Fi communication unit, a Bluetooth communication unit, etc.

[0111] Address bus 201 is used to specify the physical address of required units (such as input / output units or memory units). Communication bus 202 is used to exchange signals between units within the system and to communicate with external devices. Data bus 203 is used to transmit numerical data and instruction data between units. Control bus 203 is also used to exchange control signals, transmit operation instructions and their timing information to each unit, and is responsible for interacting with external devices using control signals.

[0112] The present invention also provides a computer-readable storage medium. A computer program is stored in the computer-readable storage medium, and when executed by a processor, the computer program implements the steps of any of the above-described methods for manufacturing bionic bodies that support remote user customization.

[0113] Optionally, the computer-readable storage medium may be non-volatile memory (NVM), such as at least one disk storage device.

[0114] Optionally, the computer-readable storage medium may also be at least one storage device located remotely from the aforementioned processor.

[0115] In another embodiment of the present invention, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to perform the steps of any of the above-described bionic body manufacturing methods that support remote customization by the user.

[0116] It should be noted that, for the product embodiments of the system / electronic device / storage medium / computer program, since they are basically similar to the method embodiments, the description is relatively simple, and relevant parts can be referred to the description of the method embodiments.

[0117] In this invention, the terms "first," "second," etc., are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention.

[0118] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature or characteristic described in connection with that embodiment or example is 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. Moreover, the specific features or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0119] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings and the disclosure in carrying out the claimed invention. In the description of the invention, the word "comprising" does not exclude other components or steps, "a" or "an" does not exclude a plurality, and "a plurality" means two or more, unless otherwise explicitly specified. Furthermore, while different embodiments may describe certain measures, this does not mean that these measures cannot be combined to produce good results.

[0120] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus (devices), or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects, all of which are collectively referred to herein as "modules" or "systems." Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The computer program may be stored / distributed in a suitable medium, provided with or as part of other hardware, or may take other distribution forms, such as via the Internet or other wired or wireless telecommunications systems.

[0121] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0122] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0123] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0124] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for manufacturing bionic bodies that supports remote customization by users, characterized in that, include: Step 1: Acquire customized images taken from multiple angles; the customized images are images obtained by the user in advance using a first camera to photograph a reference object and a first calibration card; the reference object is a reference provided by the user for customizing the bionic body; the first calibration card includes a color calibration card and a skin calibration card; the skin calibration card contains multiple skin sample blocks; the multiple skin sample blocks visually present different states; the color calibration card contains multiple color blocks; the multiple color blocks have different colors; the color calibration card and / or the skin calibration card also have a calibration body; the calibration body is used to correct lighting conditions and camera spatial frequency response; Step 2: For each angle, generate imaging scene parameters on the user side based on the customized image; Step 3: For each angle, determine the cross-scene calibration parameters based on the imaging scene parameters of the user side and the manufacturing side; wherein, the imaging scene parameters of the manufacturing side are generated by the manufacturing side using the second camera to take pictures of the second calibration card in advance, and based on the pictures taken; the first calibration card and the second calibration card are the same calibration card or two identical calibration cards; Step 4: For each angle, correct the reference object image in the customized image according to the cross-scene calibration parameters; Step 5: Create a 3D model of the bionic object based on the corrected reference object images from various angles to manufacture the bionic object.

2. The bionic body manufacturing method supporting remote user customization according to claim 1, characterized in that, The imaging scene parameters include illumination parameters, camera spatial frequency response, color parameters, and skin state parameters in the imaging scene.

3. The bionic body manufacturing method supporting remote user customization according to claim 2, characterized in that, Step 2 specifically includes: For each angle, a calibration body image, a calibration body shadow image, a color calibration card image, a skin calibration card image, and the reference object image are extracted from the customized image. Illumination parameters on the user side are generated based on the calibration body shadow image, color parameters on the user side are generated based on the color calibration card image, and skin state parameters on the user side are generated based on the skin calibration card image. The spatial frequency domain distribution of the calibration body image is calculated as the camera spatial frequency response on the user side.

4. The bionic body manufacturing method supporting remote user customization according to claim 3, characterized in that, Step 3 specifically includes: For each angle, based on the illumination parameters, color parameters, skin state parameters, and camera spatial frequency response from both the user side and the manufacturing side, corresponding illumination condition calibration factors, color calibration factors, skin state calibration factors, and spatial frequency response calibration factors are calculated as cross-scene calibration parameters.

5. The bionic body manufacturing method supporting remote user customization according to claim 1, characterized in that, Before the final bionic body is manufactured, the method further includes: At least once, images of the bionic body taken from the multiple angles are acquired; the bionic body images are images obtained by the manufacturing side using a second camera to capture images of the bionic body during manufacturing; For each angle, the differences in shape between the bionic object under manufacturing and the reference object are compared based on the bionic image and the corrected reference object image; If the difference in shape does not meet the shape error requirements, then the 3D modeling of the bionic body is optimized based on the difference in shape. If the difference in shape meets the shape error requirement, then the differences in color and skin condition between the bionic body in manufacturing and the reference object are further compared based on the bionic body image and the corrected reference object image. If the differences in color and epidermal condition do not meet the requirements for color and skin error, then the 3D modeling of the bionic body is optimized based on the differences in color and epidermal condition.

6. The bionic body manufacturing method supporting remote user customization according to claim 4, characterized in that, The step of generating user-side illumination parameters based on the calibration body shadow image includes: inputting the calibration body shadow image into a first neural network model, so that the first neural network model outputs illumination parameters as user-side illumination parameters; The first neural network model is trained based on multiple first training samples; the first training samples include: calibration body shadow image samples and the illumination parameters corresponding to the calibration body shadow image samples; the calibration body shadow image samples are calibration body shadow images taken under illumination conditions with known illumination parameters, and the illumination parameters are the illumination parameters corresponding to the calibration body shadow image samples; different calibration body shadow image samples correspond to different illumination conditions.

7. The bionic body manufacturing method supporting remote user customization according to claim 4, characterized in that, The step of generating user-side epidermal state parameters based on the epidermal calibration card image includes: Extract each first sub-block from the epidermal calibration card image; the first sub-block is the image of the epidermal sample block; Each first sub-block is input into the second neural network model so that the second neural network model outputs the epidermal state parameters corresponding to each first sub-block. The skin state parameters on the user side are determined based on the skin state parameters corresponding to each of the first sub-blocks. The second neural network model is trained based on multiple second training samples; the second training samples include: first sub-block samples and epidermal state parameters corresponding to the first sub-block samples; the first sub-block samples in the multiple second training samples include: the original first sub-block obtained by photographing the epidermal sample block, and the synthetic first sub-block generated based on the original first sub-block using data augmentation.

8. The bionic body manufacturing method supporting remote user customization according to claim 4, characterized in that, The step of generating user-side color parameters based on the color calibration card image includes: Extract each second sub-block from the color calibration card image; the second sub-block is an image of the color block; Each second sub-block image is input into the third neural network model, so that the third neural network model outputs the color parameters corresponding to each second sub-block; The color parameters on the user side are determined based on the color parameters corresponding to each of the second sub-blocks; The third neural network model is trained based on multiple third training samples; the third training samples include: second sub-block samples and color parameters corresponding to the second sub-block samples; the second sub-block samples in the multiple third training samples include: original second sub-blocks obtained by photographing the color blocks, and synthetic second sub-blocks generated based on the original second sub-blocks using data augmentation.

9. The bionic body manufacturing method supporting remote user customization according to claim 4, characterized in that, In step 4, the following operations are performed for each angle: Step 4-1: Correct the lighting conditions of the reference object image according to the lighting condition calibration factor; Step 4-2: Perform camera spatial frequency response correction on the reference object image according to the spatial frequency response calibration factor; Step 4-3: Correct the epidermal condition of the reference object image according to the epidermal condition calibration factor; Step 4-4: Perform color correction on the reference object image according to the color calibration factor.

10. The bionic body manufacturing method supporting remote user customization according to claim 9, characterized in that, Step 4-1 specifically includes: The reference object image is divided into several first sub-regions; Each first sub-region is input into the first neural network model so that the first neural network model outputs the illumination parameters corresponding to the first sub-region, and the illumination conditions corresponding to the first sub-region are corrected to the illumination conditions on the manufacturing side according to the illumination condition calibration factor. The first neural network model is trained based on multiple first training samples. The first training samples include: calibration body shadow image samples and the illumination parameters corresponding to the calibration body shadow image samples. The calibration body shadow image samples are calibration body shadow images taken under illumination conditions with known illumination parameters, and the illumination parameters are the illumination parameters corresponding to the calibration body shadow image samples. Different calibration body shadow image samples correspond to different illumination conditions.

11. The bionic body manufacturing method supporting remote user customization according to claim 9, characterized in that, Step 4-3 specifically includes: The reference object image is divided into several second sub-regions; Each second sub-region is input into the second neural network model so that the second neural network model outputs the epidermal state parameters corresponding to the second sub-region, and the epidermal state of the second sub-region is corrected to the display conditions on the manufacturing side according to the epidermal state calibration factor; The second neural network model is trained based on multiple second training samples. The second training samples include: a first sub-block sample and the epidermal state parameters corresponding to the first sub-block sample. The first sub-block is an image of the epidermal sample block. The first sub-block sample in the multiple second training samples includes: an original first sub-block obtained by taking a picture of the epidermal sample block, and a synthetic first sub-block generated based on the original first sub-block using data augmentation.

12. The bionic body manufacturing method supporting remote user customization according to claim 9, characterized in that, Step 4-4 specifically includes: The reference object image is divided into several third sub-regions; Each third sub-region is input into the third neural network model so that the third neural network model outputs the color parameters corresponding to the third sub-region, and the color of the third sub-region is corrected to the display conditions on the manufacturing side according to the color calibration factor; The third neural network model is trained based on multiple third training samples. The third training samples include: second sub-block samples and color parameters corresponding to the second sub-block samples. The second sub-block is an image of the color block. The second sub-block samples in the multiple third training samples include: the original second sub-block obtained by taking a picture of the color block, and the synthesized second sub-block generated based on the original second sub-block using data augmentation.

13. The bionic body manufacturing method supporting remote user customization according to claim 9, characterized in that, The process of performing biomimetic 3D modeling based on the corrected reference object images from various angles includes: A 3D structure is constructed based on reference object images from various angles, and surface data is generated for the 3D structure to obtain 3D structure data and surface data; wherein, the surface data includes the color and skin state corresponding to the application area in the 3D structure; the color of the application area is formed based on the color parameter distribution of the corrected reference object image, and the skin state of the application area is formed based on the skin state parameter distribution of the corrected reference object image; The 3D structural data and surface data are input into the fourth neural network model so that the fourth neural network model outputs CAD data for manufacturing bionic bodies. The fourth neural network is trained based on historical data, which includes 3D structural data, surface data, and corresponding CAD data used in previous bionic body manufacturing processes.

14. A bionic body manufacturing device that supports remote customization by users, characterized in that, The device comprises two sets of modules. The first set of modules includes: a calibration card image receiving module, a calibration card image acquisition module, an illumination parameter acquisition module, a camera spatial response acquisition module, an epidermal parameter acquisition module, a color parameter acquisition module, and a calibration factor calculation module. The second set of modules includes: a reference object image receiving module, a first image conversion module, a second image conversion module, a third image conversion module, a digital 3D image reconstruction module, a 3D prosthesis production CAD data generation module, a 3D prosthesis production module, and a product evaluation and reconstruction module. The calibration card image receiving module is used to receive multi-angle calibration card images captured by the user side. These calibration card images are images obtained by the user in advance using a first camera to capture images of the first calibration card. The first calibration card includes a color calibration card and a skin calibration card. The skin calibration card contains multiple skin sample blocks. The multiple skin sample blocks visually present different states. The color calibration card contains multiple color blocks. The multiple color blocks have different colors. The color calibration card and / or the skin calibration card also include a calibration body. The calibration body is used to correct lighting conditions and camera spatial frequency response. The calibration card image acquisition module is used to acquire multi-angle calibration card images taken by the manufacturing side. These calibration card images are obtained by the manufacturing side using a second camera to capture images of the second calibration card. The first calibration card and the second calibration card are either the same calibration card or two identical calibration cards. The illumination parameter acquisition module is used to extract the calibration body shadow images of the user side and the manufacturing side from the calibration card images taken by the user side and the manufacturing side, and input them into the first neural network model to obtain the illumination parameters of the user side and the manufacturing side respectively. The camera spatial response acquisition module is used to extract calibration volume images from calibration card images captured by the user side and the manufacturing side, perform two-dimensional Fourier transform on them, and obtain the camera spatial frequency response of the user side and the manufacturing side. The epidermal parameter acquisition module is used to extract epidermal calibration card images from calibration card images taken from the user side and the manufacturing side, and input them into the second neural network model to obtain epidermal state parameters from the user side and the manufacturing side respectively. The color parameter acquisition module is used to extract color calibration card images from calibration card images taken from the user side and the manufacturing side, and input them into the third neural network model to obtain color parameters from the user side and the manufacturing side, respectively. The calibration factor calculation module is used to calculate the illumination condition calibration factor, color calibration factor, skin state calibration factor, and spatial frequency response calibration factor according to the illumination parameters, color parameters, skin state parameters, and camera spatial frequency response on the user side and the manufacturing side, respectively. The reference object image receiving module is used to receive multi-angle reference object images captured by the user; the reference object is a reference provided by the user for customizing the bionic body. The first image conversion module is used to convert the illumination parameters of the reference object image on the user side using the illumination condition calibration factor; The second image conversion module is used to perform camera spatial frequency response conversion on the reference object image on the user side using the spatial frequency response calibration factor; The third image conversion module is used to convert the color and skin state of the reference object image on the user side using the color calibration factor and the skin state calibration factor. The digital 3D image reconstruction module is used to perform bionic 3D modeling using multi-angle converted reference object images to form 3D structural data and surface data. The 3D prosthesis production CAD data generation module is used to input the 3D structural data and surface data into the fourth neural network model to obtain CAD data for manufacturing bionic bodies. The three-dimensional prosthesis production module is used to manufacture bionic bodies using the CAD data through CAD production methods; The product evaluation and reconstruction module is used to evaluate the quality of the bionic body. If the evaluation result does not meet the preset quality standard, the digital three-dimensional image reconstruction module is triggered to optimize the 3D modeling to achieve iterative improvement of the bionic body until the evaluation result meets the quality standard.

15. A bionic body manufacturing system supporting remote user customization, characterized in that, include: At least one calibration card and computer program product; The calibration card includes a color calibration card and an epidermal calibration card; the epidermal calibration card contains multiple epidermal sample blocks; the multiple epidermal sample blocks visually present different states; the color calibration card contains multiple color blocks; the multiple color blocks have different colors; the color calibration card and / or the epidermal calibration card are further provided with a calibration body; the calibration body is used to correct lighting conditions and camera spatial frequency response; when the computer program product is run on a computer, the computer performs the steps of the bionic body manufacturing method supporting remote customization by any one of claims 1 to 13.

16. The system according to claim 15, characterized in that, The system also includes: lighting equipment; The lighting equipment is used to provide illumination to the manufacturing side so that all imaging activities related to the manufacturing of the bionic body are carried out under uniform lighting conditions from the initial imaging of the calibration card to form the imaging scene parameters of the manufacturing side until the completion of the bionic body manufacturing process.

17. The system according to claim 16, characterized in that, The system also includes: a camera and a 3D printing device.

18. An electronic device, characterized in that, It includes an input / output unit, a processing unit, a storage unit, a control unit, and a communication unit; the input / output unit, the processing unit, the storage unit, the control unit, and the communication unit are connected to an address bus, a communication bus, and a data / control bus; When the computer program in the storage unit is executed, the steps of the bionic body manufacturing method supporting remote customization by any one of claims 1 to 13 are implemented by the processor in the processing unit.

19. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the bionic body manufacturing method supporting remote customization as described in any one of claims 1 to 13.

20. A computer program product containing instructions, characterized in that, When the computer program product is run on a computer, it causes the computer to perform the steps of the bionic body manufacturing method supporting remote customization by any one of claims 1 to 13.

Images