A method for individualization of images based on real - scene three - dimension

Abstract

The present invention discloses a method for image monomerization based on real scene three-dimensional, including the steps: S1, obtaining oblique photography data, preprocessing it to obtain the preprocessed oblique photography data; S2, performing segmentation training on the preprocessed oblique photography data through AI image recognition technology to obtain a segmentation mask image; S3, performing geometric adjustment and topology repair on the segmented target objects in the segmentation mask image to establish a monomer model; S4, binding the spatial attributes and geometric features of the monomer model; S5, constructing a tree structure of the monomer model to adapt to the multi-scene business requirements, and completing the method for image monomerization based on real scene three-dimensional. The present invention is applicable to the requirements of multiple fields such as architectural design, urban planning, and urban management, has high practical value and economic benefits, solves the problems of insufficient data quality, insufficient segmentation accuracy, insufficient model optimization, and weak attribute organization ability in the prior art, and improves the applicability and practical value of the model.

Description

Technical Field

[0001] The present invention belongs to the technical field of real - scene three - dimensional modeling and data processing, and particularly relates to a method for image monomerization based on real - scene three - dimensional Background Art

[0002] With the rapid development of real - scene three - dimensional modeling technology, the oblique photography technology collected by devices such as unmanned aerial vehicles and LiDAR can generate high - precision three - dimensional models, which have been widely used in fields such as building management and urban planning. However, especially in terms of monomerization segmentation and model optimization, it is difficult to meet diverse business requirements, and there are the following main deficiencies:

[0003] (1) Insufficient data quality:

[0004] The three - dimensional models generated by oblique photography usually contain problems such as noise, blurred details, and inconsistent colors. These problems stem from limitations in shooting angles, light conditions, and device performance, resulting in a decrease in model clarity and difficulty in meeting the requirements of high - precision segmentation.

[0005] (2) Insufficient segmentation accuracy:

[0006] Traditional segmentation techniques have limited performance in dealing with complex scenes. Especially when extracting and separating multi - target objects, shape deviations or target overlaps are likely to occur. In addition, the existing algorithms have weak segmentation capabilities for details and are difficult to achieve the accuracy required for practical applications.

[0007] (3) Insufficient model optimization:

[0008] The segmented models often have problems in geometric shapes and topological structures, including irregular boundaries, non - manifold edges, isolated vertices, and redundant patches. These defects limit the integrity and usability of the models and cannot directly support subsequent analysis and application requirements.

[0009] (4) Weak attribute organization ability:

[0010] Lack of spatial attribute binding and hierarchical organization functions for monomer models, making it difficult to support multi - scene applications. Summary of the Invention

[0011] Aiming at the above deficiencies in the prior art, a method for image monomerization based on real - scene three - dimensional provided by the present invention solves the problems of insufficient data quality, insufficient segmentation accuracy, insufficient model optimization, and weak attribute organization ability in the prior art.

[0012] To achieve the above - mentioned invention purpose, the technical solution adopted by the present invention is: A method for image monomerization based on real - scene three - dimensional, comprising the following steps:

[0013] S1. Obtain the oblique photography data, preprocess it to obtain the preprocessed oblique photography data;

[0014] S2. Perform segmentation training on the preprocessed oblique photography data through AI image recognition technology to obtain a segmentation mask map;

[0015] S3. Perform geometric adjustment and topological repair on the segmented target objects in the segmentation mask map to establish a single model;

[0016] S4. Bind the spatial attributes and geometric features of the single model;

[0017] S5. Construct the tree structure of the single model to adapt to the multi-scenario business requirements and complete the image singleization method based on real scene three-dimensional.

[0018] Furthermore: In the above S1, the preprocessing method includes the following sub-steps:

[0019] S11. Use the Gaussian filtering algorithm to perform noise reduction processing on the oblique photography data to smooth the noise interference;

[0020] S12. Based on the image after noise reduction processing, extract the image edges through the Laplace operator to enhance the image details;

[0021] S13. Eliminate color distortion through a linear mapping function to obtain the preprocessed oblique photography data.

[0022] The beneficial effect of the above further solution is: By preprocessing the oblique photography data, the image quality is improved through denoising, enhancement and detail repair, which is beneficial for the subsequent steps.

[0023] Furthermore: In the above S11, the expression of the Gaussian filtering algorithm is specifically:

[0024]

[0025] Wherein, is the pixel value of the original image after noise reduction processing, is the pixel value of the original image, is the filtering window, is the Gaussian weight;

[0026] In the above S12, the expression of the image after enhancing the details is specifically:

[0027]

[0028] In the formula, is the pixel value of the image after enhancing the details, λ is the enhancement coefficient, used to control the degree of edge protrusion, is the edge gradient of the image, and its expression is specifically:

[0029]

[0030] In the formula, is the second derivative of the image in the horizontal axis direction, reflecting the brightness change rate of the image in this direction, is the second derivative of the image in the vertical axis direction, reflecting the brightness change rate of the image in this direction;

[0031] In the above S13, the preprocessed oblique photography data The expression of is specifically:

[0032]

[0033] In the formula, is the gain coefficient, is the color deviation correction value.

[0034] Furthermore: The above S2 includes the following sub-steps:

[0035] S21. Input the preprocessed oblique photography data and the corresponding label data into the deep learning network, and optimize the deep learning network through multiple rounds of training;

[0036] S22. Correct the output of the deep learning network through manual annotation to obtain a trained deep learning network;

[0037] S23. Output a segmentation mask image through the trained deep learning network, and each pixel in the segmentation mask image is assigned a corresponding category label.

[0038] The beneficial effect of the above further solution is: Through the AI image recognition technology, the oblique photography data can be segmented and trained, and the segmentation accuracy can be optimized by combining user annotation.

[0039] Furthermore: In the above S21, the parameters of the deep learning network are optimized through the cross-entropy loss function, and the cross-entropy loss function The expression of is specifically:

[0040]

[0041] In the formula, is the true label of each pixel, is the model prediction value, N 1 is the total number of samples.

[0042] Furthermore: The above S3 includes the following sub-steps:

[0043] S31. Obtain the model of the segmented target object in the segmentation mask map, fit the geometric shape of the model by the least squares method, correct the segmentation error, calculate the deviation between the vertex and the fitting surface, and correct the vertices with a deviation greater than the deviation threshold;

[0044] S32. Use the Bezier curve interpolation method to repair the model boundary to ensure the boundary continuity of the model;

[0045] S33. Repair the non-manifold edges in the model through the topology optimization algorithm, remove the isolated vertices, and merge the redundant patches to ensure the integrity of the model topology and establish a single model.

[0046] The beneficial effects of the above further solution are: geometric adjustment and topology repair are performed on the segmented target object to ensure the integrity and accuracy of the monomerized model.

[0047] Further: In the above S31, the expression for fitting the geometric shape of the model by the least squares method is specifically:

[0048]

[0049] In the formula, is the point cloud vertex coordinate, A is the first plane equation parameter, B is the second plane equation parameter, C is the third plane equation parameter, D is the fourth plane equation parameter, x i is the horizontal axis coordinate of the point cloud vertex, y i is the vertical axis coordinate of the point cloud vertex, z i is the vertical axis coordinate of the point cloud vertex, i is the ordinal number of the point cloud vertex, N 2 is the total number of point cloud vertices;

[0050] In the above S32, the formula of the Bezier curve interpolation method is specifically:

[0051]

[0052] In the formula, P 0 is the first control point, P 1 is the second control point, P 2 is the third control point, t is the interpolation parameter, .

[0053] Further: The above S4 includes the following sub-steps:

[0054] S41. Calculate the geometric center of the monomer model;

[0055] S42. Calculate the geometric properties of the monomer model, including volume and surface area;

[0056] S43. Add spatial attributes and geometric features to the monomer model. The geometric features include the geometric center and geometric properties, and the spatial attributes include name, use, hierarchical relationship, and geographical coordinates.

[0057] The beneficial effect of the above further solution is: Bind spatial attributes to the model to facilitate subsequent storage, management, and application.

[0058] Further: In the above S41, when calculating the geometric center of the monomer model E The specific expression is:

[0059]

[0060] In the formula, is the j th vertex coordinate, is the total number of vertices;

[0061] In the above S42, for the volume V The specific expression is:

[0062]

[0063] In the formula, is the th triangle patch area, is the th perpendicular height from the triangle patch to the reference plane, M is the total number of triangle patches;

[0064] For the surface area S The specific expression is:

[0065] .

[0066] Further: The above S5 includes the following sub-steps:

[0067] S51. Organize the monomer models using a tree structure. The tree structure includes several root nodes and the sub-nodes subordinate to the several root nodes. Each sub-node stores the spatial attributes and geometric features of the monomer model;

[0068] S52. The tree structure adopts a way that supports dynamic expansion and adjustment, allowing the addition, deletion, or modification of root nodes and sub-nodes to meet the requirements of different business scenarios;

[0069] S53. The tree structure is output in a standardized format to meet the business requirements of multiple scenarios and complete the method for image monomerization based on real scene three-dimensional.

[0070] The beneficial effect of the above further solution is: constructing the tree structure of the monomer model can meet the need for the monomer model to be organized in a hierarchical tree structure for dynamic recombination.

[0071] The beneficial effects of the present invention are as follows:

[0072] (1) The present invention provides a method for image monomerization based on real scene three-dimensional. By preprocessing data to improve image quality, using AI image recognition technology combined with user annotation to optimize segmentation accuracy, and performing geometric adjustment and topological repair on the model to ensure model integrity and accuracy; at the same time, binding spatial attributes to the model, supporting the compilation and dynamic recombination of the tree structure, significantly enhancing the model's organization ability and application flexibility, applicable to the needs of multiple fields such as architectural design, urban planning, and urban management, with high practical value and economic benefits, solving the problems of insufficient data quality, insufficient segmentation accuracy, insufficient model optimization, and weak attribute organization ability in the prior art, and improving the applicability and practical value of the model.

[0073] (2) The monomerization method of the present invention based on real scene three-dimensional images can efficiently and accurately generate and manage monomer models, providing data support for further spatial analysis and applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a flowchart of a method for image monomerization based on real scene three-dimensional of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0075] The following describes the specific embodiments of the present invention to facilitate those skilled in the art of the present technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art of the present technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0076] As Figure 1 shown, in an embodiment of the present invention, a method for image monomerization based on real scene three-dimensional includes the following steps:

[0077] S1. Obtain the oblique photography data, preprocess it to obtain the preprocessed oblique photography data;

[0078] S2. Perform segmentation training on the preprocessed oblique photography data through AI image recognition technology to obtain a segmentation mask image;

[0079] S3. Geometric adjustment and topological repair are performed on the segmented target objects in the segmentation mask image to establish a single model;

[0080] S4. Bind the spatial attributes and geometric features of the single model;

[0081] S5. Construct the tree structure of the single model to meet the requirements of multi-scenario services, and complete the method for image singleization based on real-scene 3D.

[0082] In the above S1, the method for preprocessing includes the following sub-steps:

[0083] S11. Use the Gaussian filtering algorithm to perform noise reduction processing on the oblique photography data to smooth the noise interference;

[0084] S12. Based on the image after noise reduction processing, extract the image edges through the Laplace operator to enhance the image details;

[0085] S13. Eliminate color distortion through a linear mapping function to obtain the preprocessed oblique photography data.

[0086] In this embodiment, the method for preprocessing is specifically to repair and improve the quality of the oblique photography data by means of noise reduction processing, detail enhancement, and color correction in sequence.

[0087] In the above S11, the expression of the Gaussian filtering algorithm is specifically:

[0088]

[0089] Among them, is the pixel value of the original image after noise reduction processing, is the pixel value of the original image, is the filtering window, is the Gaussian weight;

[0090] Noise in the image will affect subsequent image analysis and object recognition. Noise reduction is the basis for improving image quality. The Gaussian filtering algorithm removes noise by smoothing the image. In this process, the calculation of each pixel value depends on the weighted average of its surrounding neighboring pixels. Specifically, when using Gaussian filtering, the value of each pixel is weighted-averaged with other pixels in its neighborhood, and the weight is determined by the Gaussian function. The shape of the Gaussian function determines the influence range of each pixel's neighborhood.

[0091] In the above S12, the expression of the image after enhancing details is specifically:

[0092]

[0093] In the formula, is the pixel value of the image after enhancing details, λis the enhancement coefficient, used to control the degree of edge prominence, is the edge gradient of the image, and its specific expression is:

[0094]

[0095] In the formula, is the second derivative of the image in the horizontal axis direction, reflecting the brightness change rate of the image in this direction, is the second derivative of the image in the vertical axis direction, reflecting the brightness change rate of the image in this direction;

[0096] In this embodiment, the detail enhancement aims to make the object contour clearer by strengthening the edge information of the image. The Laplace operator is widely used for image edge extraction. It obtains the degree of change of the image by calculating the image gradient, thereby enhancing the image details. Through detail enhancement, the edges and subtle changes in the image can be captured more clearly, which helps the subsequent segmentation of objects.

[0097] In S13, the preprocessed oblique photography data The specific expression is:

[0098]

[0099] In the formula, is the gain coefficient, used to adjust the brightness, is the color deviation correction value, used to adjust the color tone deviation.

[0100] In this embodiment, due to the inconsistent shooting conditions and lighting environment of oblique photography, color distortion may occur, affecting the realism of the image. The present invention can eliminate or reduce this color difference and restore the true color of the object through color correction. Color correction is usually carried out by the linear mapping method, aiming to adjust the brightness and color tone of the image to make it closer to the true value.

[0101] S2 includes the following sub-steps:

[0102] S21. Input the preprocessed oblique photography data and the corresponding label data into the deep learning network, and optimize the deep learning network through multiple rounds of training;

[0103] S22. Correct the output of the deep learning network through manual annotation to obtain the trained deep learning network;

[0104] S23. Output the segmentation mask image through the trained deep learning network, and each pixel in the segmentation mask image is assigned a corresponding category label.

[0105] In S21, the parameters of the deep learning network are optimized through the cross-entropy loss function, and the cross-entropy loss function The specific expression is as follows:

[0106]

[0107] In the formula, is the true label of each pixel, is the model prediction value, N 1 is the total number of samples.

[0108] In this embodiment, the model training is to learn by inputting the labeled data, automatically extract features from the image using a deep learning network, and learn to associate these features with object categories. During the training process, the present invention uses a cross-entropy loss function to optimize the parameters of the model to ensure that the predicted output is as close as possible to the true label.

[0109] In S22, the user can optimize through manual annotation. Since in a complex environment, the deep learning network may misjudge some boundaries or objects, the manually corrected annotation can help the model correct errors, especially for difficult scenarios or objects with unclear boundaries. The annotation after manual optimization will further improve the accuracy of the model.

[0110] After training and optimization, the deep learning network can generate a segmentation mask image, where each pixel is assigned a class label. The segmentation mask image provides a very important input for subsequent steps and can be used to build a 3D model or perform further geometric analysis.

[0111] S3 includes the following sub-steps:

[0112] S31. Obtain the model of the segmented target object in the segmentation mask image, fit the geometric shape of the model by the least squares method, correct the segmentation error, calculate the deviation between the vertex and the fitting surface, and correct the vertices with a deviation greater than the deviation threshold;

[0113] S32. Use the Bessel curve interpolation method to repair the model boundary to ensure the boundary continuity of the model;

[0114] S33. Repair the non-manifold edges in the model through a topology optimization algorithm, remove isolated vertices, and merge redundant patches to ensure the integrity of the model topology and establish a single model.

[0115] In S31, the specific expression for fitting the geometric shape of the model by the least squares method is as follows:

[0116]

[0117] In the formula, is the point cloud vertex coordinate, A is the first plane equation parameter, Bis the parameter of the second plane equation, C is the parameter of the third plane equation, D is the parameter of the fourth plane equation, x i is the horizontal axis coordinate of the point cloud vertex, y i is the vertical axis coordinate of the point cloud vertex, z i is the vertical axis coordinate of the point cloud vertex, i is the ordinal number of the point cloud vertex, N 2 is the total number of point cloud vertices;

[0118] In this embodiment, the main task of geometric adjustment is to correct the geometric shape deviation in the model to ensure that it conforms to the shape of the actual object. The commonly used geometric adjustment method is the least squares fitting, which adjusts the shape of the object by minimizing the fitting error. For example, when fitting a plane, the least squares method calculates the distance from each point to the plane and minimizes the sum of the squares of these distances to optimize the fitting result.

[0119] In S32, the formula of the Bezier curve interpolation method Specifically:

[0120]

[0121] In the formula, P 0 is the first control point, P 1 is the second control point, P 2 is the third control point, t is the interpolation parameter, .

[0122] In this embodiment, boundary repair mainly solves the jagged edges and discontinuous object contours in the model. The present invention uses the Bezier curve interpolation method to smooth these boundaries, making the edges of the model more smooth, and the control points of the Bezier curve can be adjusted according to actual needs.

[0123] S4 includes the following sub-steps:

[0124] S41. Calculate the geometric center of the monomer model;

[0125] S42. Calculate the geometric properties of the monomer model, including volume and surface area;

[0126] S43. Add spatial attributes and geometric features to the monomer model. The geometric features include the geometric center and geometric properties, and the spatial attributes include name, use, hierarchical relationship, and geographical coordinates.

[0127] In S41, calculating the geometric center of the monomer model E The specific expression of

[0128]

[0129] In the formula, is the j coordinate of the n-th vertex, and

[0130] is the total number of vertices;

[0131] In this embodiment, the geometric center of each object is obtained by calculating the average value of all vertex coordinates. The geometric center is the "center of gravity" of the object and plays a role in positioning in space. V In S42, the expression of the volume

[0132]

[0133] is specifically: is the area of the m-th triangular patch, is the vertical height of the M m-th triangular patch to the reference plane, and

[0134] is the total number of triangular patches; S The expression of the surface area

[0135] .

[0136] In this embodiment, the geometric properties include volume and surface area, which are the basic physical characteristics of the model and can reflect the size and surface characteristics of the object. The volume is calculated by the area of each triangular patch and its height to the reference plane, and the surface area is the sum of the areas of all patches.

[0137] S5 includes the following sub-steps:

[0138] S51. Organize the monomer model using a tree structure. The tree structure includes several root nodes and sub-nodes subordinate to the root nodes. Each sub-node stores the spatial attributes and geometric features of the monomer model;

[0139] In this embodiment, the tree structure T is specifically:

[0140]

[0141] In the formula, is the root node, is the first set of sub-nodes, is the k m-th set of sub-nodes. Establishing the tree structure enables hierarchical management of the model, facilitating storage and access.

[0142] S52. The tree structure adopts a way that supports dynamic expansion and adjustment, allowing the addition, deletion, or modification of root nodes and child nodes to meet the requirements of different business scenarios;

[0143] S53. The tree structure is output in a standardized format to adapt to the business requirements of multiple scenarios and complete the method for image monomerization based on real-scene three-dimensional.

[0144] The beneficial effects of the present invention are as follows: The present invention provides a method for image monomerization based on real-scene three-dimensional. By preprocessing data, the image quality is improved. The AI image recognition technology is used in combination with user annotation to optimize the segmentation accuracy, and geometric adjustment and topological repair are performed on the model to ensure the integrity and accuracy of the model. At the same time, spatial attributes are bound to the model, and the compilation and dynamic reorganization of the tree structure are supported, significantly enhancing the organization ability and application flexibility of the model. It is applicable to the requirements of multiple fields such as architectural design, urban planning, and urban management, and has high practical value and economic benefits. It solves the problems of insufficient data quality, insufficient segmentation accuracy, insufficient model optimization, and weak attribute organization ability in the prior art, and improves the applicability and practical value of the model.

[0145] The monomerization method of the present invention based on real-scene three-dimensional images can efficiently and accurately generate and manage monomer models, providing data support for further spatial analysis and applications.

[0146] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by terms such as "center", "thickness", "upper", "lower", "horizontal", "top", "bottom", "inner", "outer", "radial", etc. is based on the orientation or positional relationship shown in the drawings, and is 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 thus cannot be understood as a limitation to the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of technical features. Therefore, the features defined by "first", "second", and "third" may explicitly or implicitly include one or more of such features.

Claims

1. A method for image singulation based on real scene three-dimensional, characterized in that: The following steps are involved: S1, acquiring oblique photography data, and preprocessing the oblique photography data to obtain preprocessed oblique photography data; S2. Perform segmentation training on the pre-processed oblique photography data through AI image recognition technology to obtain a segmentation mask map; S3, performing geometric adjustment and topological repair on the segmented target object in the segmentation mask image to establish a monomer model; S4, spatial attributes and geometric features of the bound monomer model; S5. Build a tree structure of a monomer model to adapt to multi-scenario business needs and complete the image monomerization method based on real scene 3D; The S3 comprises the following sub-steps: S31, obtaining a model of the segmented target object in the segmentation mask image, fitting the geometric shape of the model by the least square method, correcting the segmentation error, and calculating the deviation between the vertex and the fitting surface, and correcting the vertex whose deviation is greater than the deviation threshold; S32, using the Bezier curve interpolation method to repair the model boundary to ensure the continuity of the model boundary; S33. Use the topology optimization algorithm to repair the non-manifold edges in the model, remove isolated vertices, and merge redundant faces to ensure the topological structure of the model is complete and establish a single model. In S31, the expression of the geometric shape of the least squares fitting model is specifically: In the formula, is the vertex coordinate of the point cloud, A is the first plane equation parameter, B is the second plane equation parameter, C is the third plane equation parameter, D is the fourth plane equation parameter, x i is the horizontal axis coordinate of the point cloud vertex, y i is the vertical coordinate of the point cloud vertex, z i is the vertical axis coordinate of the point cloud vertex, i is the ordinal number of the point cloud vertex, N 2 is the total number of point cloud vertices; In S32, the formula of the Bezier curve interpolation method is Specifically: In the formula, P 0 is the first control point, P 1 is the second control point, P 2 is the third control point, t is the interpolation parameter, .

2. The image singulation method based on real scene three-dimensional according to claim 1, characterized in that: In S1, the method for pre-processing comprises the following steps: S11, using a Gaussian filter algorithm to perform noise reduction processing on the oblique photography data to smooth out noise interference; S12, based on the image after noise reduction, extracting the image edge through the Laplacian operator to enhance the image details; S13, eliminating color distortion through a linear mapping function to obtain pre-processed oblique photography data.

3. The image singulation method based on real scene three-dimensional according to claim 2, characterized in that: In S11, the expression of the Gaussian filtering algorithm is specifically: in, is the pixel value of the original image after noise reduction processing, is the pixel value of the original image, is the filter window, is the Gaussian weight; In S12, the expression of the image after detail enhancement is specifically: In the formula, To enhance the image pixel value after detail enhancement, λ is the enhancement coefficient, which is used to control the degree of edge prominence. is the edge gradient of the image, and its specific expression is: In the formula, is the second-order derivative of the image in the horizontal direction, reflecting the rate of change of the brightness of the image in this direction. It is the second-order derivative of the image in the vertical direction, reflecting the brightness change rate of the image in this direction; In S13, the pre-processed oblique photography data The specific expression is: In the formula, is the gain coefficient, This is the color cast correction value.

4. The image singulation method based on real scene three-dimensional according to claim 1, characterized in that: The S2 comprises the following sub-steps: S21, inputting the preprocessed oblique photography data and the corresponding label data into the deep learning network, and optimizing the deep learning network through multiple rounds of training; S22, correcting the output of the deep learning network by manual annotation to obtain a trained deep learning network; S23. Output a segmentation mask image through the trained deep learning network, and each pixel in the segmentation mask image is assigned a corresponding category label.

5. The image singulation method based on real scene three-dimensional according to claim 4, characterized in that: In S21, the parameters of the deep learning network are optimized by the cross entropy loss function. The cross entropy loss function The specific expression is: In the formula, is the true label for each pixel, is the model prediction value, N 1 is the total number of samples.

6. The image singulation method based on real scene three-dimensional according to claim 1, characterized in that: The S4 comprises the following sub-steps: S41, calculating the geometric center of the monomer model; S42, calculating geometric properties of the monomer model, including volume and surface area; S43. Add spatial attributes and geometric features to the monomer model. The geometric features include geometric center and geometric attributes. The spatial attributes include name, purpose, hierarchical relationship and geographic coordinates.

7. The image singulation method based on real scene three-dimensional according to claim 6, characterized in that: In S41, the geometric center of the monomer model is calculated E The specific expression is: In the formula, For the j The vertex coordinates, is the total number of vertices; In S42, the volume V The specific expression is: In the formula, For the The area of ​​the triangle patch, For the The vertical height of the triangle patch to the reference plane, M is the total number of triangle patches; Surface Area S The specific expression is: 。 8. The image singulation method based on real scene three-dimensional according to claim 1, characterized in that: The S5 comprises the following sub-steps: S51, using a tree structure to organize the monomer model, the tree structure includes a number of root nodes, a number of child nodes under the root nodes, each child node stores the spatial attributes and geometric features of the monomer model; S52. The tree structure supports dynamic expansion and adjustment, allowing the addition, deletion or modification of root nodes and child nodes to meet the needs of different business scenarios; S53. The tree structure is output in a standardized format to meet the business needs of multiple scenarios and complete the image monomerization method based on real-scene three-dimensional.

Images