A three-dimensional image reconstruction method and system

Abstract

The present invention relates to the field of three-dimensional image processing technology, specifically to a three-dimensional image reconstruction method and system, comprising the following steps: based on the optical property data of the object, analyzing the light propagation behavior of various objects at differentiated angles, combining the type of light source, building a light propagation model, and generating light propagation behavior data. In the present invention, through the analysis of the optical properties of the object and the multi-angle light propagation behavior, the influence of the light source in the three-dimensional space is calculated, and the spatial positioning accuracy of the image is improved by combining the identification of the position and type of the light source, and the distortion correction ability of the image is optimized. When processing objects with complex optical properties, micro distortions and discontinuities are effectively analyzed and corrected, and the visual effect and data accuracy of the three-dimensional model of the object are enhanced. By analyzing the color, pattern and roughness of the surface of the object and adjusting the morphology of the non-rigid object, a rich and realistic three-dimensional visual performance is achieved, and the application scope of the three-dimensional image reconstruction technology is expanded.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional image processing, and in particular to a three-dimensional image reconstruction method and system. Background Art

[0002] The field of three-dimensional image processing technology aims to create and reconstruct spatial images using multi-dimensional data sources through multiple steps including image capture, three-dimensional data reconstruction, image rendering, and image display. It uses stereo vision, computer vision algorithms, image segmentation, object recognition, and three-dimensional data fusion to improve the spatial perception quality and depth information of images. It combines surface reconstruction and model-based rendering and is applied to medical imaging, industrial inspection, computer-aided design, virtual reality, and other fields to provide professionals with rich and intuitive image information.

[0003] Among them, the three-dimensional image reconstruction method focuses on converting two-dimensional images or scanned data captured from multiple angles and time points into three-dimensional models. It processes the input image data through a variety of algorithms to reconstruct the accurate surface and shape of objects in three-dimensional space and enhance the visual effects of the real environment. It includes image acquisition, feature point matching, three-dimensional geometric construction, and model rendering. It is applied to augmented reality, game development, three-dimensional city modeling, and creation of biological structure models in the medical field. It provides rich and intuitive visual effects and accurate data support, and improves the work efficiency of personnel.

[0004] Traditional 3D image reconstruction technology cannot accurately simulate the type of light source and light propagation behavior when dealing with complex lighting environments and changing viewing angles, resulting in inaccurate light source positioning, affecting the effect and accuracy of image distortion correction. It relies on traditional image analysis technology to identify and process the microstructure of objects, which leads to omissions and incorrect interpretation of key information when dealing with 3D reconstruction tasks with complex details and large amounts of data. It cannot capture subtle changes on the surface of objects in microstructure analysis and texture feature adjustment, resulting in the reconstructed 3D model lacking the necessary level of detail. It cannot simulate the actual physical properties of non-rigid objects, limiting the application scenarios and performance of the model, affecting its accuracy and practicality in professional applications, and resulting in visual and structural differences between the generated 3D model and the actual object. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a three-dimensional image reconstruction method and system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution, a three-dimensional image reconstruction method, comprising the following steps:

[0007] S1: Based on the optical property data of the object, analyze the light propagation behavior of various objects at different angles, combine the types of light sources, build a light propagation model, and generate light propagation behavior data;

[0008] S2: Based on the light propagation behavior data, by analyzing the brightness information of multiple regions in the input image data, identifying the type of light source, calculating the position of the light source in the target image in the three-dimensional space, and generating light source position positioning data;

[0009] S3: Based on the light source position positioning data, by evaluating the deviation between the image data and the light propagation model, identifying the image distortion features and calculating the distortion correction parameters, thereby generating an image data correction result;

[0010] S4: Based on the image data correction result, perform microstructure analysis on the corrected image data, identify and correct micro distortion and discontinuity of the object contour, construct three-dimensional models of multiple objects in the image, and generate a three-dimensional model of the target object;

[0011] S5: Based on the three-dimensional model of the target object, the color, pattern and roughness of the surface of the object are analyzed through the image data, the texture features are identified and the three-dimensional model is adjusted to generate surface texture mapping parameters;

[0012] S6: Based on the surface texture mapping parameters, the shape information of the three-dimensional model is analyzed to identify the type of the object, and the model parameters of the non-rigid object are adjusted to generate a three-dimensional image reconstruction result.

[0013] As a further solution of the present invention, the light propagation behavior data includes the scattering angle, reflection intensity information and refractive index data of the light, the light source position positioning data includes the spatial coordinates, irradiation direction and light intensity level information of the light source, the image data correction result includes the adjusted pixel position, color correction value and brightness adjustment parameter, the target object three-dimensional model includes the object's geometric structure data, surface feature information and volume attribute data, the surface texture mapping parameters include texture map alignment, map scaling ratio and texture depth information, and the three-dimensional image reconstruction result includes the elastic model parameters of non-rigid objects, object dynamic response characteristic parameters and object type recognition results.

[0014] As a further solution of the present invention, based on the optical property data of the object, the light propagation behavior of various objects at different angles is analyzed, and a light propagation model is constructed in combination with the type of light source. The steps of generating light propagation behavior data are specifically as follows:

[0015] S101: Based on the optical property data of the objects, surface material data of various objects are collected, light scattering and absorption ratios are analyzed and recorded, and object material property information is generated;

[0016] S102: Analyze the light reflection paths of various objects at different angles based on the material characteristic information of the objects, record the deviation and bending of light propagation, and generate light path simulation results;

[0017] S103: Based on the light path simulation result, a Monte Carlo ray tracing algorithm is used to calculate the light propagation path and effect in three-dimensional space, consider the influence of light source type and propagation medium, build a light propagation model, and generate light propagation behavior data.

[0018] As a further solution of the present invention, the Monte Carlo ray tracing algorithm is according to the formula:

[0019] ;

[0020] Calculate the transmission intensity of light in the medium, where is the light intensity after transmission, is the initial light intensity, is the base of natural logarithms, is the absorption coefficient of the medium, is the distance that light travels in the medium, is the scattering coefficient of the medium, is the angle adjustment coefficient when light interacts with the medium, is the incident angle of the light on the medium surface.

[0021] As a further solution of the present invention, based on the light propagation behavior data, by analyzing the brightness information of multiple regions in the input image data, identifying the type of light source, calculating the position of the light source in the target image in three-dimensional space, and generating the light source position positioning data, the specific steps are:

[0022] S201: Analyze the brightness and color information in the input image data based on the light propagation behavior data, evaluate the brightness distribution and color temperature characteristics of multiple regions, and generate brightness and color distribution information;

[0023] S202: Based on the brightness and color distribution information, by analyzing the light reflection characteristics of various light sources, identifying the light source type of the target image, including sunlight, fluorescent light, and natural light source, and generating light source type identification data;

[0024] S203: Based on the light source type identification data, using a light propagation model, considering the influence of scattering and reflection behavior of light on position calculation, analyzing the position of the target light source in three-dimensional space, and generating light source position positioning data.

[0025] As a further solution of the present invention, based on the light source position positioning data, by evaluating the deviation between the image data and the light propagation model, identifying the image distortion features and calculating the distortion correction parameters, the steps of generating the image data correction result are specifically as follows:

[0026] S301: using the light source position positioning data, analyzing the brightness and color of multiple image regions and comparing them with a light propagation model, calculating light distortion values ​​of multiple regions, and generating regional distortion data;

[0027] S302: Calculating distortion correction parameters of multiple image regions based on the regional distortion data, including pixel position adjustment and light intensity correction, to generate a correction parameter data set;

[0028] S303: Apply the correction parameter data set to adjust the input image data to eliminate optical distortion, optimize the visual consistency of the image, and generate an image data correction result.

[0029] As a further solution of the present invention, based on the image data correction result, the corrected image data is subjected to microstructure analysis, the microscopic distortion and discontinuity of the object contour are identified and corrected, and a three-dimensional model of multiple objects in the image is constructed. The steps of generating the three-dimensional model of the target object are specifically as follows:

[0030] S401: Analyze the microstructure of the object in the input image data based on the image data correction result, identify the micro distortion and discontinuity of the object contour, and generate microstructure analysis data;

[0031] S402: Based on the microstructure analysis data, correct the object contour and eliminate distortion and discontinuity, optimize the three-dimensional shape of the object, and generate object contour correction data;

[0032] S403: Analyze the geometric shapes of multiple objects using the object contour correction data, construct three-dimensional models of multiple objects in the image, and generate a three-dimensional model of the target object.

[0033] As a further solution of the present invention, based on the three-dimensional model of the target object, the color, pattern and roughness of the object surface are analyzed through image data, texture features are identified and the three-dimensional model is adjusted, and the steps of generating surface texture mapping parameters are specifically as follows:

[0034] S501: Based on the three-dimensional model of the target object, collect color and pattern data of the surface of the target object, perform color saturation and contrast analysis on each surface area, evaluate color characteristics of multiple areas, and generate color characteristic data;

[0035] S502: Based on the color characteristic data, by analyzing the surface gloss and reflection characteristics of the target object, evaluating texture characteristics, including concavity and convexity, texture depth, and generating texture characteristic information;

[0036] S503: According to the texture feature information, adjust the texture parameters of the surface of the target object model, including color, pattern, and roughness, match the input image data, and generate surface texture mapping parameters.

[0037] As a further solution of the present invention, based on the surface texture mapping parameters, the shape information of the three-dimensional model is analyzed, the type of the object is identified, and the model parameters of the non-rigid object are adjusted to generate the three-dimensional image reconstruction result in the following steps:

[0038] S601: Analyzing the geometric shape and surface texture information of the three-dimensional model using the surface texture mapping parameters, identifying visual features of multiple parts of the target object, and generating shape texture feature information;

[0039] S602: Based on the shape and texture feature information, analyze the types of multiple objects in the image and identify non-rigid objects, including cloth, metal, and plants, and generate object type recognition results;

[0040] S603: Based on the object type recognition result, adjust the three-dimensional model parameters of the target object, simulate the physical response behavior of the non-rigid object, and generate a three-dimensional image reconstruction result.

[0041] A three-dimensional image reconstruction system, the three-dimensional image reconstruction system is used to perform the above three-dimensional image reconstruction method, the system comprises:

[0042] The light analysis module builds a light propagation model and generates light path analysis data based on the optical property data of the object by comparing and analyzing the light propagation behaviors of various object materials at different angles.

[0043] The light source positioning module analyzes the brightness information of multiple areas in the input image data based on the light path analysis data, identifies the type of light source and calculates the position in the three-dimensional space, and generates the light source space coordinate information;

[0044] The image correction module evaluates the deviation between the image and the light propagation model based on the light source spatial coordinate information, calculates the image correction parameters, adjusts the image to eliminate optical distortion, and generates an image adjustment record;

[0045] The structural analysis module uses the image adjustment record to perform microscopic structural analysis on the corrected image, identify microscopic distortions and discontinuities of the object contour, construct three-dimensional models of multiple objects in the image, and generate object model construction results;

[0046] The texture adjustment module analyzes the color, pattern and roughness of the object surface based on the object model construction result, adjusts the shape information of the three-dimensional model, identifies the object type and adjusts the model parameters of the non-rigid object, and generates a three-dimensional image reconstruction result.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are:

[0048] In the present invention, the influence of the light source in three-dimensional space is calculated by analyzing the optical characteristics of the object and the multi-angle light propagation behavior, and the spatial positioning accuracy of the image is improved by combining the identification of the position and type of the light source, and the distortion correction ability of the image is optimized. When processing objects with complex optical properties, micro distortions and discontinuities are effectively analyzed and corrected, and the visual effect and data accuracy of the three-dimensional model of the object are enhanced. By analyzing the color, pattern and roughness of the object surface and adjusting the morphology of non-rigid objects, rich and realistic three-dimensional visual performance is achieved, and the application scope of three-dimensional image reconstruction technology is expanded. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0050] Figure 2 This is a detailed flow chart of S1 of the present invention;

[0051] Figure 3 This is a detailed flow chart of S2 of the present invention;

[0052] Figure 4 This is a detailed flow chart of S3 of the present invention;

[0053] Figure 5 This is a detailed flow chart of S4 of the present invention;

[0054] Figure 6 This is a detailed flow chart of S5 of the present invention;

[0055] Figure 7 This is a detailed flow chart of S6 of the present invention;

[0056] Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0058] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0059] See also Figure 1 The present invention provides a technical solution, a three-dimensional image reconstruction method, comprising the following steps:

[0060] S1: Based on the optical property data of the object, analyze the light propagation behavior of various objects at different angles, combine the types of light sources, build a light propagation model, and generate light propagation behavior data;

[0061] S2: Based on the light propagation behavior data, by analyzing the brightness information of multiple areas in the input image data, identifying the type of light source, calculating the position of the light source in the target image in three-dimensional space, and generating light source position positioning data;

[0062] S3: Based on the light source position positioning data, by evaluating the deviation between the image data and the light propagation model, identifying the image distortion features and calculating the distortion correction parameters, the image data correction result is generated;

[0063] S4: Based on the image data correction result, perform microstructure analysis on the corrected image data, identify and correct micro distortion and discontinuity of the object contour, construct three-dimensional models of multiple objects in the image, and generate a three-dimensional model of the target object;

[0064] S5: Based on the 3D model of the target object, the color, pattern and roughness of the object surface are analyzed through image data, the texture features are identified and the 3D model is adjusted to generate surface texture mapping parameters;

[0065] S6: Based on the surface texture mapping parameters, the shape information of the three-dimensional model is analyzed, the type of the object is identified, and the model parameters of the non-rigid object are adjusted to generate a three-dimensional image reconstruction result.

[0066] The light propagation behavior data includes the scattering angle, reflection intensity information and refractive index data of the light. The light source position positioning data includes the spatial coordinates, irradiation direction and light intensity level information of the light source. The image data correction results include the adjusted pixel position, color correction value and brightness adjustment parameters. The three-dimensional model of the target object includes the object's geometric structure data, surface feature information and volume attribute data. The surface texture mapping parameters include the texture map alignment, map scaling and texture depth information. The three-dimensional image reconstruction results include the elastic model parameters of non-rigid objects, the object's dynamic response characteristic parameters and the object type recognition results.

[0067] See also Figure 2 Based on the optical property data of the object, the light propagation behavior of various objects at different angles is analyzed, and the light propagation model is constructed in combination with the type of light source. The specific steps for generating light propagation behavior data are as follows:

[0068] S101: Based on the optical property data of the object, the surface material data of various objects are collected, and the light scattering and absorption ratios are analyzed and recorded to generate the material property information of the object. Specifically, the process is as follows;

[0069] In sub-step S101, based on the optical property data of the objects, the surface material data of various objects are collected, and the spectral analysis is performed on each material to obtain the light scattering and absorption ratio, and the reaction data of each material under the light of different wavelengths is recorded. The transmittance and reflectivity of the light are measured by a spectrophotometer, and the light scattering coefficient and absorption coefficient of each material are obtained by analysis. Using the light scattering absorption model, the formula is: ,in, is the total light scattering absorbance, is the light scattering coefficient, is the light absorption coefficient, is the adjustment factor of the scattering coefficient, It is the adjustment factor of the absorption coefficient, predicts the effect of different materials on light, and generates information on the material characteristics of the object.

[0070] S102: Based on the material characteristic information of the object, the light reflection paths of various objects at different angles are analyzed, the offset and bending of the light propagation are recorded, and the process of generating the light path simulation result is specifically as follows;

[0071] In sub-step S102, based on the material characteristic information of the object, the scattering and absorption data of the object surface are used to analyze the light reflection path at different angles. By setting different incident angles and observation angles, the reflection and refraction of the light on the object surface are recorded, and the offset and bending path of the light propagation are analyzed. The reflection path analysis model is used, and the formula is: ,in, represents the light path, is the light reflectivity, is the incident angle, is the reflectivity adjustment factor, It is the adjustment coefficient of the incident angle, simulating the behavior of light under different materials and structures, and generating light path simulation results.

[0072] S103: Based on the light path simulation results, the Monte Carlo ray tracing algorithm is used to consider the influence of the light source type and the propagation medium, calculate the propagation path and effect of the light in the three-dimensional space, build a light propagation model, and generate the light propagation behavior data. The specific process is as follows;

[0073] In sub-step S103, based on the light path simulation results, taking into account the type of light source and the characteristics of the propagation medium, the Monte Carlo method is used to simulate the random propagation path of light in three-dimensional space, record the intensity and wavelength of the light source and the refractive index and absorption coefficient of the medium, build a light propagation model, and apply the Monte Carlo ray tracing model to reflect the influence of the light source and medium on light propagation, and generate light propagation behavior data.

[0074] Monte Carlo ray tracing algorithm, according to the formula:

[0075] ;

[0076] Calculate the transmission intensity of light in the medium, where The intensity of light after transmission, which is used to indicate the intensity of light after passing through a medium. is the initial light intensity, which is used to indicate the original intensity of light before it enters the medium. It is the base of the natural logarithm, used to represent the basis of the exponential function, and is key to simulating the nonlinear effect of light attenuation. is the absorption coefficient of the medium, which is used to measure the efficiency of the medium in absorbing light energy. is the propagation distance of light in the medium, which is used to calculate the attenuation of light during propagation. is the scattering coefficient of the medium, which is used to describe the scattering intensity of the medium to the light and affects the deviation and diffusion of the light propagation path. It is the angle adjustment coefficient when light interacts with the medium, which is used to adjust the effect of the light source angle on the light intensity. It is the incident angle between the light and the medium surface. It is used to calculate the specific effect of the angle on the light intensity regulation. The cosine value of the angle is used to adjust the light intensity attenuation.

[0077] The specific execution process of the formula is as follows:

[0078] Measuring the initial light intensity by optical equipment and the distance light travels in the medium , using a spectrum analyzer to determine the absorption coefficient of the medium and scattering coefficient , the incident angle of light is measured by an angle sensor And calculate the cosine value, the angle adjustment coefficient obtained through experimental data , substitute the parameters into the formula to calculate the light intensity after passing through the medium , which is used to analyze the propagation effects of light under various conditions and optimize the practicality and prediction accuracy of the model.

[0079] See also Figure 3 Based on the light propagation behavior data, by analyzing the brightness information of multiple areas in the input image data, identifying the type of light source, calculating the position of the light source in the target image in three-dimensional space, and generating the light source position positioning data, the specific steps are:

[0080] S201: Based on the light propagation behavior data, analyzing the brightness and color information in the input image data, evaluating the brightness distribution and color temperature characteristics of multiple regions, and generating the brightness and color distribution information is specifically as follows;

[0081] In sub-step S201, based on the light propagation behavior data, the input image data is analyzed to extract the brightness and color information of multiple regions in the image. Through color space conversion and brightness correction, the brightness distribution and color temperature characteristics of different regions in the image are evaluated. Using the brightness color analysis model, the formula is: ,in, Represents brightness and color distribution information, is the brightness value of the image, is the color value, is the adjustment coefficient of the brightness value, It is the adjustment coefficient of the color value, analyzes the light effect in the image, and generates brightness and color distribution information.

[0082] S202: Based on the brightness and color distribution information, by analyzing the light reflection characteristics of various light sources, the light source type of the target image is identified, including sunlight, fluorescent light, and natural light source. The specific process of generating light source type identification data is as follows;

[0083] In sub-step S202, based on the brightness and color distribution information, the light source characteristics reflected in the image are analyzed to identify the type of light source in the image. By comparing the spectral characteristics and reflection behaviors of differentiated light sources, including daylight, fluorescent lamps, and natural light sources, a light source type recognition model is used, and the formula is: ,in, Represents the light source type identification data, is the brightness data, is the color temperature data, and They are adjustment coefficients for brightness and color temperature data, respectively, distinguish and identify the various light source types in the image, and generate light source type identification data.

[0084] S203: Based on the light source type identification data, using the light propagation model, considering the influence of the scattering and reflection behavior of light on the position calculation, analyzing the position of the target light source in the three-dimensional space, and generating the light source position positioning data, the specific process is as follows;

[0085] In sub-step S203, based on the light source type identification data, a light propagation model is applied to calculate the position of the target light source in three-dimensional space. By considering the scattering and reflection behavior of light, a light source position analysis model is adopted, and the formula is: ,in, Represents the light source position positioning data, is the light source type data, is the light direction data, is the adjustment factor of the light source type, It is the adjustment coefficient of the light direction, locates the position of the light source in space, and generates the light source position positioning data.

[0086] See also Figure 4 Based on the light source position data, the steps of generating the image data correction result are as follows:

[0087] S301: using the light source position positioning data, analyzing the brightness and color of multiple image regions and comparing them with the light propagation model, calculating the light distortion values ​​of multiple regions, and generating the regional distortion data in detail;

[0088] In sub-step S301, based on the light source position positioning data, the brightness and color of multiple image regions are analyzed, and the deviation between the actual image and the theoretical model is analyzed by comparison and analysis with the light propagation model to determine the light distortion value of the region. The regional light deviation model is used, and the formula is: ,in, represents the regional distortion data, is the observed brightness or color data, is the theoretical ray propagation data, The difference between the actual light and the predicted model is identified and quantified for the distortion sensitivity coefficient, generating regional distortion data.

[0089] S302: Based on the regional distortion data, the distortion correction parameters of multiple image regions are calculated, including pixel position adjustment and light intensity correction. The specific process of generating the correction parameter data set is as follows;

[0090] In sub-step S302, based on the regional distortion data, the distortion correction parameters of multiple image regions are calculated, including pixel position adjustment and light intensity correction. The pixel adjustment algorithm is optimized according to the regional distortion degree. The pixel correction coefficient model is used, and the formula is: ,in, represents the calibration parameter data, is the regional distortion data, is the initial measurement value of the area to be corrected in the image, To correct the intensity coefficient, the image data is adjusted according to the severity of the distortion and the regional characteristics to generate a correction parameter data set.

[0091] S303: Apply the correction parameter data set to adjust the input image data to eliminate optical distortion, optimize the visual consistency of the image, and generate the image data correction result. The specific process is as follows;

[0092] In sub-step S303, the correction parameter data set is applied to adjust the input image data to eliminate optical distortion and enhance the visual consistency of the image. By implementing the image processing algorithm, considering the degree of distortion and the correction parameters, a visual consistency optimization model is adopted, and the formula is: ,in, Represents data correction information, To calibrate the parameter data set, is the image data before adjustment, For the overall visual optimization coefficient, element multiplication is used to represent the impact of the correction parameters on each pixel, optimize image quality and reduce visual distortion, and generate image data correction results.

[0093] See also Figure 5 Based on the image data correction results, the corrected image data is subjected to microstructure analysis, the microscopic distortion and discontinuity of the object contour are identified and corrected, and the three-dimensional models of multiple objects in the image are constructed. The specific steps of generating the three-dimensional model of the target object are as follows:

[0094] S401: Based on the image data correction result, the microstructure of the object in the input image data is analyzed, the microscopic distortion and discontinuity of the object contour are identified, and the process of generating microstructure analysis data is specifically as follows;

[0095] In sub-step S401, based on the image data correction result, the microstructure of the object in the input image data is analyzed by high-resolution imaging technology to identify the microscopic distortion and discontinuity of the object contour, and the microstructure analysis model is used, and the formula is: ,in, Represents structural analysis information, is the microscopic distortion data identified in the image, is the analytical sensitivity coefficient, which represents the sensitivity of distortion recognition at the micro level. The integral symbol It means that the overall integral calculation of the microscopic distortion of the image data is performed, covering the entire image area, and generating microstructure analysis data.

[0096] S402: Based on the microstructure analysis data, the object contour is corrected and distortion and discontinuity are eliminated, the three-dimensional shape of the object is optimized, and the process of generating object contour correction data is specifically as follows;

[0097] In sub-step S402, based on the microstructure analysis data, the object contour is corrected to eliminate distortion and discontinuity, thereby optimizing the three-dimensional shape of the object. The object contour correction model is adopted, and the formula is: ,in, Represents contour correction information, For microstructural analysis data, For the preset object contour reference data, To correct the intensity coefficient, which indicates the degree of contour adjustment, object contour correction data is generated.

[0098] S403: Analyze the geometrical forms of multiple objects by using the object contour correction data, construct three-dimensional models of multiple objects in the image, and generate the three-dimensional model of the target object.

[0099] In sub-step S403, the geometric morphology analysis of multiple objects in the image is performed using the object contour correction data to construct a three-dimensional model of the object. The three-dimensional modeling model is applied, and the formula is: ,in, Represents the object model, Correct the data for the contour of a single object, is the weight factor of the corresponding object, indicating the relative importance of different objects in 3D modeling. For the overall modeling efficiency coefficient, the 3D model is constructed by considering the correction data and weight of the object to generate the 3D model of the target object.

[0100] See also Figure 6 Based on the three-dimensional model of the target object, the color, pattern and roughness of the object surface are analyzed through image data, the texture features are identified and the three-dimensional model is adjusted. The specific steps of generating surface texture mapping parameters are as follows:

[0101] S501: Based on the three-dimensional model of the target object, the color and pattern data of the surface of the target object are collected, the color saturation and contrast analysis is performed on each surface area, the color characteristics of multiple areas are evaluated, and the specific process of generating color characteristic data is as follows;

[0102] In sub-step S501, based on the three-dimensional model of the target object, the color and pattern data of the surface of the target object are collected, including using high dynamic range imaging technology to capture the color details of each surface area, and analyzing the color saturation and contrast of each area through image analysis software. The color saturation measurement uses color space conversion to quantitatively evaluate the purity of the color. The contrast analysis is evaluated by calculating the standard deviation of the pixel brightness in the area, and the color characteristic data is calculated. The formula is: ,in, Represents color characteristic data, is the surface area analyzed, is the color data in the area, is the standard contrast color data for the region, is the color contrast sensitivity coefficient, It represents the integration over a tiny portion of the surface area to generate color characteristic data.

[0103] S502: Based on the color characteristic data, by analyzing the surface gloss and reflection characteristics of the target object, the texture characteristics are evaluated, including concavity and convexity, and texture depth, and the specific process of generating texture characteristic information is as follows;

[0104] In sub-step S502, based on the color characteristic data, the surface gloss and reflection characteristics of the target object are analyzed, the gloss is evaluated by measuring the intensity of the reflected light, the microstructure of the surface is analyzed to determine the reflection characteristics, and the reflectivity under the differential wavelength is captured by a spectrometer. Combined with the color characteristic data, the following formula is used for processing: ,in, Represents texture feature information, is the glossiness data of a single area extracted from the color characteristic data. is the reflection characteristic data of the corresponding area, is the texture feature weight coefficient, The total number of regions analyzed to generate texture feature information.

[0105] S503: According to the texture feature information, the texture parameters of the surface of the target object model, including color, pattern, and roughness, are adjusted to match the input image data, and the process of generating the surface texture mapping parameters is specifically as follows;

[0106] In sub-step S503, the texture parameters of the target object model surface are adjusted according to the texture feature information, including the adjustment of color, pattern and roughness. The alignment and scaling of the pattern are adjusted using an image processing algorithm. The visual effect is optimized by simulating the effect of surface roughness on light scattering. The calculation process of texture parameter adjustment uses the following formula: ,in, Represents the surface texture mapping parameters, is the texture feature information item, is the texture fitness coefficient, is the overall texture adjustment factor, Generates surface texture mapping parameters for the number of texture feature items.

[0107] See also Figure 7 , based on the surface texture mapping parameters, the shape information of the 3D model is analyzed, the type of the object is identified, and the model parameters of the non-rigid object are adjusted. The specific steps for generating the 3D image reconstruction result are:

[0108] S601: Analyze the geometric shape and surface texture information of the three-dimensional model using the surface texture mapping parameters, identify the visual features of multiple parts of the target object, and generate the shape texture feature information.

[0109] In sub-step S601, the geometric shape and surface texture information of the three-dimensional model are analyzed by image analysis software using the surface texture mapping parameters, and the visual features of multiple parts of the model are identified, including the granularity, color variation and gloss difference of the texture. The formula is: ,in, Represents shape and texture feature information, is the total number of parts of the model being analyzed, Representative Some of the surface texture mapping parameters, is the geometric shape data of the corresponding part, It is a weighting coefficient that adjusts the degree of integration of texture and shape data, captures the surface characteristics of the model, and generates shape and texture feature information.

[0110] S602: Based on shape and texture feature information, analyze the types of multiple objects in the image and identify non-rigid objects, including cloth, metal, and plants. The specific process of generating object type identification results is as follows;

[0111] In sub-step S602, based on the shape and texture feature information, multiple types of objects in the image are identified, including the identification of non-rigid objects such as cloth, metal and plants. The commonalities and differences of texture and shape data are evaluated using pattern recognition technology. The formula is: ,in, Represents the object type recognition result, is the number of objects analyzed, It is extracted from the shape and texture feature information. The characteristic data of an object, is the feature importance coefficient, In order to analyze the sensitivity coefficient as a whole, the feature data of different objects are accumulated and weighted to identify various types of non-rigid objects and generate object type recognition results.

[0112] S603: Based on the object type recognition result, the three-dimensional model parameters of the target object are adjusted to simulate the physical response behavior of the non-rigid object to generate the three-dimensional image reconstruction result. The specific process is as follows;

[0113] In sub-step S603, based on the object type recognition result, the three-dimensional model parameters of the target object are adjusted to simulate the physical response behavior of the non-rigid object, including adjusting the elasticity, compression resistance and texture of the model to match the requirements of differentiated physical properties. The formula is: ,in, represents the three-dimensional image reconstruction result, is the number of object types analyzed, For the Identification data of class objects, is the object's physical property adjustment coefficient, In order to adjust the global sensitivity coefficient, the physical parameters of the three-dimensional model are adjusted by nonlinear weighted accumulation of each object recognition result to generate a three-dimensional image reconstruction result.

[0114] See also Figure 8 , a three-dimensional image reconstruction system, the three-dimensional image reconstruction system is used to perform the above three-dimensional image reconstruction method, the system comprises:

[0115] The light analysis module builds a light propagation model and generates light path analysis data based on the optical property data of the object by comparing and analyzing the light propagation behaviors of various object materials at different angles.

[0116] The light source positioning module analyzes the brightness information of multiple areas in the input image data based on the light path analysis data, identifies the type of light source, calculates the position in three-dimensional space, and generates the spatial coordinate information of the light source;

[0117] The image correction module evaluates the deviation between the image and the light propagation model based on the spatial coordinate information of the light source, calculates the image correction parameters, adjusts the image to eliminate optical distortion, and generates an image adjustment record;

[0118] The structural analysis module uses the image adjustment record to perform microstructural analysis on the corrected image, identify microscopic distortions and discontinuities in the object contour, construct three-dimensional models of multiple objects in the image, and generate object model construction results;

[0119] Based on the object model construction results, the texture adjustment module analyzes the color, pattern and roughness of the object surface, adjusts the shape information of the three-dimensional model, identifies the object type and adjusts the model parameters of non-rigid objects to generate three-dimensional image reconstruction results.

[0120] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A three-dimensional image reconstruction method, characterized in that: The following steps are involved: Based on the optical property data of objects, analyze the light propagation behavior of various objects at different angles, build a light propagation model based on the types of light sources, and generate light propagation behavior data; Based on the light propagation behavior data, by analyzing the brightness information of multiple areas in the input image data, identifying the type of light source, calculating the position of the light source in the target image in the three-dimensional space, and generating light source position positioning data; Based on the light source position positioning data, by evaluating the deviation between the image data and the light propagation model, identifying the image distortion features and calculating the distortion correction parameters, thereby generating an image data correction result; Based on the image data correction result, performing microstructure analysis on the corrected image data, identifying and correcting microscopic distortions and discontinuities of the object contour, constructing three-dimensional models of multiple objects in the image, and generating a three-dimensional model of the target object; Based on the three-dimensional model of the target object, the color, pattern and roughness of the surface of the object are analyzed through image data, texture features are identified and the three-dimensional model is adjusted to generate surface texture mapping parameters; Based on the surface texture mapping parameters, the shape information of the three-dimensional model is analyzed to identify the type of the object, and the model parameters of the non-rigid object are adjusted to generate a three-dimensional image reconstruction result.

2. The three-dimensional image reconstruction method according to claim 1, characterized in that: The light propagation behavior data includes the scattering angle, reflection intensity information and refractive index data of the light, the light source position positioning data includes the spatial coordinates, irradiation direction and light intensity level information of the light source, the image data correction result includes the adjusted pixel position, color correction value and brightness adjustment parameter, the target object three-dimensional model includes the object's geometric structure data, surface feature information and volume attribute data, the surface texture mapping parameters include the texture map alignment, map scaling ratio and texture depth information, and the three-dimensional image reconstruction result includes the elastic model parameters of the non-rigid object, the object dynamic response characteristic parameters and the object type recognition result.

3. The three-dimensional image reconstruction method according to claim 1, characterized in that: Based on the optical property data of the object, the light propagation behavior of various objects at different angles is analyzed, and the light propagation model is constructed in combination with the type of light source. The specific steps for generating light propagation behavior data are as follows: Based on the optical property data of objects, the surface material data of various objects are collected, the light scattering and absorption ratios are analyzed and recorded, and the material property information of the objects is generated; Based on the material characteristic information of the object, analyzing the light reflection paths of various objects at different angles, recording the deviation and bending of light propagation, and generating light path simulation results; Based on the light path simulation results, a Monte Carlo ray tracing algorithm is used to calculate the propagation path and effect of light in three-dimensional space, consider the influence of light source type and propagation medium, build a light propagation model, and generate light propagation behavior data.

4. The three-dimensional image reconstruction method according to claim 3, characterized in that: The Monte Carlo ray tracing algorithm is based on the formula: ； Calculate the transmission intensity of light in the medium, where is the light intensity after transmission, is the initial light intensity, is the base of natural logarithms, is the absorption coefficient of the medium, is the distance that light travels in the medium, is the scattering coefficient of the medium, is the angle adjustment coefficient when light interacts with the medium, is the incident angle of the light on the medium surface.

5. The three-dimensional image reconstruction method according to claim 1, characterized in that: Based on the light propagation behavior data, by analyzing the brightness information of multiple areas in the input image data, identifying the type of light source, calculating the position of the light source in the target image in three-dimensional space, and generating the light source position positioning data, the specific steps are: Based on the light propagation behavior data, analyzing the brightness and color information in the input image data, evaluating the brightness distribution and color temperature characteristics of multiple regions, and generating brightness and color distribution information; Based on the brightness and color distribution information, by analyzing the light reflection characteristics of various light sources, the light source type of the target image is identified, including fluorescent light and natural light source, and light source type identification data is generated; Based on the light source type identification data, using the light propagation model, considering the influence of scattering and reflection behavior of light on position calculation, the position of the target light source in three-dimensional space is analyzed to generate light source position positioning data.

6. The three-dimensional image reconstruction method according to claim 1, characterized in that: Based on the light source position positioning data, the steps of generating the image data correction result by evaluating the deviation between the image data and the light propagation model, identifying the image distortion features and calculating the distortion correction parameters are specifically as follows: Using the light source position positioning data, analyzing the brightness and color of multiple image regions and comparing them with the light propagation model, calculating the light distortion values ​​of multiple regions, and generating regional distortion data; Based on the regional distortion data, calculating distortion correction parameters of multiple image regions, including pixel position adjustment and light intensity correction, to generate a correction parameter data set; The correction parameter data set is applied to adjust the input image data to eliminate optical distortion, optimize the visual consistency of the image, and generate an image data correction result.

7. The three-dimensional image reconstruction method according to claim 1, characterized in that: Based on the image data correction result, the corrected image data is subjected to microstructure analysis, the microscopic distortion and discontinuity of the object contour are identified and corrected, and a three-dimensional model of multiple objects in the image is constructed. The steps of generating the three-dimensional model of the target object are specifically as follows: Based on the image data correction result, analyzing the microstructure of the object in the input image data, identifying the microscopic distortion and discontinuity of the object contour, and generating microstructure analysis data; Based on the microstructure analysis data, correct the object contour and eliminate distortion and discontinuity, optimize the three-dimensional shape of the object, and generate object contour correction data; The object contour correction data is used to analyze the geometric shapes of multiple objects, construct three-dimensional models of multiple objects in the image, and generate a three-dimensional model of the target object.

8. The three-dimensional image reconstruction method according to claim 1, characterized in that: Based on the target object three-dimensional model, the color, pattern and roughness of the object surface are analyzed through image data, texture features are identified and the three-dimensional model is adjusted, and the steps of generating surface texture mapping parameters are specifically as follows: Based on the three-dimensional model of the target object, collecting color and pattern data on the surface of the target object, performing color saturation and contrast analysis on each surface area, evaluating color characteristics of multiple areas, and generating color characteristic data; Based on the color characteristic data, by analyzing the surface gloss and reflection characteristics of the target object, evaluating texture characteristics, including concavity and convexity, texture depth, and generating texture characteristic information; According to the texture feature information, the texture parameters of the surface of the target object model, including color, pattern, and roughness, are adjusted to match the input image data and generate surface texture mapping parameters.

9. The three-dimensional image reconstruction method according to claim 1, characterized in that: Based on the surface texture mapping parameters, the shape information of the three-dimensional model is analyzed, the type of the object is identified, and the model parameters of the non-rigid object are adjusted to generate the three-dimensional image reconstruction result in the following steps: Analyzing the geometric shape and surface texture information of the three-dimensional model using the surface texture mapping parameters, identifying visual features of multiple parts of the target object, and generating shape texture feature information; Based on the shape and texture feature information, analyzing the types of multiple objects in the image and identifying non-rigid objects, including cloth and plants, to generate object type identification results; Based on the object type recognition result, the three-dimensional model parameters of the target object are adjusted to simulate the physical response behavior of the non-rigid object and generate a three-dimensional image reconstruction result.

10. A three-dimensional image reconstruction system, characterized in that: According to any one of claims 1 to 9, the three-dimensional image reconstruction method, the system comprising: The light analysis module builds a light propagation model and generates light path analysis data based on the optical property data of the object by comparing and analyzing the light propagation behaviors of various object materials at different angles. The light source positioning module analyzes the brightness information of multiple areas in the input image data based on the light path analysis data, identifies the type of light source and calculates the position in the three-dimensional space, and generates the light source space coordinate information; The image correction module evaluates the deviation between the image and the light propagation model based on the light source spatial coordinate information, calculates the image correction parameters, adjusts the image to eliminate optical distortion, and generates an image adjustment record; The structural analysis module uses the image adjustment record to perform microscopic structural analysis on the corrected image, identify microscopic distortions and discontinuities of the object contour, construct three-dimensional models of multiple objects in the image, and generate object model construction results; The texture adjustment module analyzes the color, pattern and roughness of the object surface based on the object model construction result, adjusts the shape information of the three-dimensional model, identifies the object type and adjusts the model parameters of the non-rigid object, and generates a three-dimensional image reconstruction result.

Images