Method, device, medium and product for generating a random geometry model of a three-dimensional data set

Abstract

The application discloses a random geometry modeling three-dimensional data set generation method, device, medium and product, relates to the technical field of data set generation, and comprises the following steps: for any three-dimensional geometry under the current cycle number, randomly generating a depth, a space parameter and a size in a set numerical range, so as to generate an initial three-dimensional geometry in a three-dimensional scene; randomly generating a noise texture and a target energy value; applying the noise texture to the corresponding initial three-dimensional geometry respectively to obtain an initial three-dimensional scene; in the initial three-dimensional scene, setting a sunlight light source with the target energy value to obtain a target three-dimensional scene; rendering the target three-dimensional scene to obtain an image data group, adding the image data group into an initial data set to obtain a target data set, and performing the next cycle until a maximum cycle number is reached, so that a random geometry modeling three-dimensional data set is obtained. The application realizes the rapid generation of the random geometry modeling three-dimensional data set.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of dataset generation, in particular to a method and device for generating a three-dimensional dataset based on random geometry modeling, a medium and a product. BACKGROUND

[0002] In recent years, research and application in the field of images have attracted widespread attention. For example, through research on image recognition, functions such as facial payment, clock-in, and commodity classification are realized; through semantic segmentation, the efficiency of medical disease diagnosis is improved; pedestrians and vehicles are accurately segmented to promote the development of driverless vehicles; through research on computational holography, naked-eye true 3D stereoscopic display effects are realized.

[0003] With the progress of hardware technology and the significant increase in computing platform computing power, researchers are increasingly inclined to use neural networks to solve the above-mentioned common problems in the field of images. A large amount of three-dimensional dataset is needed as input when training a neural network model. Common image training sets include NYU, MIT-CGH-4K, etc. A suitable dataset is extremely important for network model training. In order to approach the limit performance of the network model and achieve better training effect, the training set used for model training should be as rich and comprehensive as possible. However, the data in the dataset is limited, and it is impossible to provide all possible training data, which leads to the fact that the effect of the trained network model on the test set is often not as good as that on the training set.

[0004] Therefore, there is a need for a method for generating an artificially synthesized three-dimensional dataset that can be quickly generated by a computer, does not rely on external assets, has unlimited possibilities, and is easy to control. SUMMARY

[0005] The purpose of the present application is to provide a method and device for generating a three-dimensional dataset based on random geometry modeling, a medium and a product, which realizes the rapid generation of a three-dimensional dataset based on random geometry modeling.

[0006] To achieve the above-mentioned purpose, the present application provides the following solutions:

[0007] In a first aspect, the present application provides a method for generating a three-dimensional dataset based on random geometry modeling, comprising:

[0008] creating an empty three-dimensional scene under the current number of cycles in the Blender software;

[0009] randomly generating floating-point numbers of a plurality of three-dimensional geometric bodies under the current number of cycles within a set numerical range, and determining the depth of each three-dimensional geometric body based on the floating-point number of the three-dimensional geometric body under the current number of cycles and a uniform depth distribution fitting function;

[0010] randomly generating a spatial parameter, a size and a noise texture of each three-dimensional geometry in the current cycle, wherein the spatial parameter comprises a position, a rotation angle and a scaling ratio;

[0011] generating a plurality of initial three-dimensional geometries in the current cycle based on the depth, the spatial parameter and the size of each three-dimensional geometry in the current cycle in the empty three-dimensional scene in the current cycle;

[0012] applying the noise texture of each three-dimensional geometry in the current cycle to the corresponding initial three-dimensional geometry respectively to obtain a plurality of target three-dimensional geometries in the current cycle, thereby obtaining an initial three-dimensional scene in the current cycle containing the plurality of target three-dimensional geometries;

[0013] randomly generating an energy value of a fluorescent light source in the current cycle to obtain a target energy value in the current cycle, and setting the fluorescent light source with the target energy value in the current cycle in the initial three-dimensional scene in the current cycle to obtain a target three-dimensional scene in the current cycle;

[0014] rendering the target three-dimensional scene in the current cycle by setting an orthogonal camera to obtain an image data set in the current cycle, and adding the image data set in the current cycle to an initial data set in the current cycle to obtain a target data set in the current cycle, wherein the image data set comprises a depth image and an intensity image, the initial data set in the current cycle is a target data set in a previous cycle, and an initial data set in an initial cycle is an empty set;

[0015] determining whether the current cycle reaches a maximum cycle;

[0016] if yes, determining the target data set in the current cycle as a random geometry modeling three-dimensional data set;

[0017] if no, updating the current cycle to a next cycle, and returning to "creating an empty three-dimensional scene in the current cycle in the Blender software", until the maximum cycle is reached, thereby obtaining the random geometry modeling three-dimensional data set.

[0018] Optionally, the set value range is [0, 1].

[0019] Optionally, the uniform depth distribution fitting function comprises:

[0020]

[0021] wherein P(r) is the depth of the three-dimensional geometry, and r is a floating point number of the three-dimensional geometry.

[0022] Optionally, the determining process of the uniform depth distribution fitting function comprises:

[0023] constructing a cumulative distribution function; the cumulative distribution function is:

[0024]

[0025] wherein C(n) is a cumulative distribution function value of the nth depth plane in the line of sight of the orthographic camera; N is the total number of depth planes in the line of sight of the orthographic camera; i and k are both layer numbers of the depth planes in the line of sight of the orthographic camera;

[0026] a fitting inverse function of the cumulative distribution function is obtained by value fitting, and the fitting inverse function is determined as the uniform depth distribution fitting function.

[0027] Optionally, the generating process of the noise texture of any three-dimensional geometric body comprises:

[0028] randomly generating an initial height pixel number of a noise image of the three-dimensional geometric body;

[0029] determining an initial width pixel number of the noise image of the three-dimensional geometric body based on the initial height pixel number;

[0030] generating an initial noise image based on the initial height pixel number and the initial width pixel number of the noise image;

[0031] interpolating and enlarging the initial noise image according to a preset height pixel number and a preset width pixel number to obtain a target noise image of the three-dimensional geometric body;

[0032] determining the target noise image of the three-dimensional geometric body as the noise texture of the three-dimensional geometric body.

[0033] Optionally, the randomly generating of the initial height pixel number of the noise image of the three-dimensional geometric body comprises:

[0034] randomly generating a power number of the noise image of the three-dimensional geometric body;

[0035] calculating the initial height pixel number of the noise image of the three-dimensional geometric body according to the power number by using an initial height pixel number calculation formula; the initial height pixel number calculation formula is:

[0036] H noise =2 m ;

[0037] wherein H noise is the initial height pixel number; and m is the power number.

[0038] Optionally, the noise texture is a low-frequency noise texture, a medium-frequency noise texture or a high-frequency noise texture.

[0039] When the noise texture is a low-frequency noise texture, the value range of the power number is [1, 2];

[0040] When the noise texture is a medium-frequency noise texture, the value range of the power number is [5, 6];

[0041] When the noise texture is a high-frequency noise texture, the value range of the power number is [9, 10].

[0042] In a second aspect, the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the generation method of the random geometric modeling three-dimensional data set according to any one of the above.

[0043] In a third aspect, the present application provides a computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the generation method of the random geometric modeling three-dimensional data set according to any one of the above.

[0044] In a fourth aspect, the present application provides a computer program product comprising a computer program, and the computer program is executed by a processor to implement the generation method of the random geometric modeling three-dimensional data set according to any one of the above.

[0045] According to the embodiments provided in the present application, the following technical effects are disclosed:

[0046] The present application discloses a generation method, device, medium and product of a random geometric modeling three-dimensional data set, determines the depth of the corresponding three-dimensional geometric body based on a uniform depth distribution fitting function, can ensure the uniformity of the pixels occupied by different depths, randomly generates the spatial parameters, sizes and noise textures of each three-dimensional geometric body, makes the obtained three-dimensional geometric body and target three-dimensional scene more rich and flexible, does not depend on the pre-existing three-dimensional model or external assets, and realizes the rapid generation of the random geometric modeling three-dimensional data set. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0048] Figure 1 The generation method of the random geometric modeling three-dimensional data set provided by an embodiment of the present application is shown in the flowchart.

[0049] Figure 2A diagram for setting an orthogonal camera for rendering a target three-dimensional scene;

[0050] Figure 3 A diagram for modeling a three-dimensional data set by random geometry;

[0051] Figure 4 A diagram for modeling an intensity image with a power of 1 in a three-dimensional data set by random geometry;

[0052] Figure 5 A diagram for modeling a depth image with a power of 2 in a three-dimensional data set by random geometry;

[0053] Figure 6 A diagram for modeling an intensity image with a power of 5 in a three-dimensional data set by random geometry;

[0054] Figure 7 A diagram for modeling a depth image with a power of 6 in a three-dimensional data set by random geometry;

[0055] Figure 8 A diagram for modeling an intensity image with a power of 9 in a three-dimensional data set by random geometry;

[0056] Figure 9 A diagram for modeling a depth image with a power of 10 in a three-dimensional data set by random geometry;

[0057] Figure 10 A structural diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0059] The purpose of the present application is to provide a generation method, device, medium and product of a three-dimensional data set modeled by random geometry, aiming to realize fast generation of the three-dimensional data set modeled by random geometry.

[0060] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0061] In an exemplary embodiment, as shown in Figure 1 the generation method of the three-dimensional data set modeled by random geometry in the embodiment includes:

[0062] Step 01: Create an empty three-dimensional scene in the current cycle number in the Blender software.

[0063] Step 02: Randomly generate a series of floating-point numbers of a plurality of three-dimensional geometric bodies in the current cycle number within a set numerical range, and determine the depth of each three-dimensional geometric body based on the floating-point number of the three-dimensional geometric body in the current cycle number and the uniform depth distribution fitting function.

[0064] As an optional implementation, the set numerical range is [0, 1].

[0065] As an optional implementation, the uniform depth distribution fitting function includes:

[0066]

[0067] Wherein, P(r) is the depth of the three-dimensional geometric body; r is the floating-point number of the three-dimensional geometric body.

[0068] As an optional implementation, the determination process of the uniform depth distribution fitting function in step 02 includes:

[0069] Step 021: Construct a cumulative distribution function; the cumulative distribution function is:

[0070]

[0071] Wherein, C(n) is the cumulative distribution function value of the nth layer of depth plane in the line of sight of the orthographic camera; N is the total number of depth planes in the line of sight of the orthographic camera; i and k are both the layer number of the depth plane in the line of sight of the orthographic camera.

[0072] Specifically, the construction process of the cumulative distribution function includes:

[0073] Let a series of variables z0, z1, …, z N represent the 0th layer to the Nth layer of depth plane in the line of sight of the orthographic camera respectively. When a continuous geometric body scene is converted into a multi-depth scene, the three-dimensional geometric body pixels of the (n-1)th layer of depth plane z n-1 and the nth layer of depth plane z n in the line of sight of the orthographic camera are projected and compressed to z n-1 In order to realize the uniformity of the depth distribution, theoretically, the projection of the three-dimensional geometric body pixels on each depth layer needs to be statistically equal, that is:

[0074]

[0075] Wherein, p zn is the probability density function value of the nth layer of depth plane in the line of sight of the orthographic camera.

[0076] However, due to the occlusion of the three-dimensional geometry in different positions, the corresponding occlusion probability needs to be calculated, and the pixel proportion probability of the projection of the three-dimensional geometry in z0, z1, …, z N The pixel proportion probability of the projection of the three-dimensional geometry in z0, z1, …, z 1, and further calculation can obtain a probability density function considering occlusion, which is used to determine the depth of the object in the scene. The probability density function considering occlusion is defined as:

[0077]

[0078] Where p(n) is the probability density function value considering occlusion of the nth layer of depth plane in the line of sight of the orthogonal camera.

[0079] Further, the cumulative distribution function is obtained by cumulative integration of the probability density function considering occlusion.

[0080] Step 022: The fitting inverse function of the cumulative distribution function is obtained by value fitting, and the fitting inverse function is determined as the uniform depth distribution fitting function.

[0081] Step 03: Randomly generate the space parameters, size and noise texture of each three-dimensional geometry in the current loop number; the space parameters include position, rotation angle and scaling ratio.

[0082] Specifically, when the position (x, y) is randomly generated, the value range of the x-axis component of the position is [-1.92, 1.92], and the value range of the y-axis component of the position is [-1.08, 1.08]. The scaling ratio can also be set according to actual needs. In order to make the volume of the three three-dimensional geometries, i.e. cube, tetrahedron and sphere, basically the same, and better ensure the uniformity of the depth distribution, the scaling ratio of the three is set to 1.61199:3:1. The size of the three-dimensional geometry should not be too large in actual assignment, and should be less than half of the target three-dimensional scene to be rendered, and should not be too small to reduce the rendering efficiency.

[0083] As an optional implementation, in step 03, the generation process of the noise texture of any three-dimensional geometry includes:

[0084] Step 031: Randomly generate the initial height pixel number of the noise image of the three-dimensional geometry.

[0085] As an optional implementation, step 031 includes:

[0086] Step 0311: Randomly generate the power number of the noise image of the three-dimensional geometry.

[0087] Step 0312: calculating the initial height pixel number of the noise image of the three-dimensional geometric body according to the power number by using an initial height pixel number calculation formula; the initial height pixel number calculation formula is:

[0088] H noise =2 m .

[0089] wherein H noise is the initial height pixel number; and m is the power number.

[0090] Step 032: determining the initial width pixel number of the noise image of the three-dimensional geometric body based on the initial height pixel number.

[0091] Step 033: generating an initial noise image based on the initial height pixel number and the initial width pixel number of the noise image.

[0092] Step 034: performing interpolation enlargement on the initial noise image according to a preset height pixel number and a preset width pixel number to obtain a target noise image of the three-dimensional geometric body.

[0093] Step 035: determining the target noise image of the three-dimensional geometric body as the noise texture of the three-dimensional geometric body.

[0094] As an optional implementation, the noise texture is a low-frequency noise texture, a medium-frequency noise texture or a high-frequency noise texture.

[0095] When the noise texture is a low-frequency noise texture, the value range of the power number is [1, 2].

[0096] When the noise texture is a medium-frequency noise texture, the value range of the power number is [5, 6].

[0097] When the noise texture is a high-frequency noise texture, the value range of the power number is [9, 10].

[0098] Step 04: generating a plurality of initial three-dimensional geometric bodies in the current cycle number of the empty three-dimensional scene based on the depth, the spatial parameter and the size of each three-dimensional geometric body in the current cycle number.

[0099] Step 05: respectively applying the noise texture of each three-dimensional geometric body in the current cycle number to the corresponding initial three-dimensional geometric body to obtain a plurality of target three-dimensional geometric bodies in the current cycle number, thereby obtaining an initial three-dimensional scene in the current cycle number containing the plurality of target three-dimensional geometric bodies.

[0100] Step 06: randomly generate an energy value of the fluorescent light source in the current cycle, obtain a target energy value in the current cycle, and set the fluorescent light source with the target energy value in the current cycle in the initial three-dimensional scene in the current cycle to obtain a target three-dimensional scene in the current cycle.

[0101] Specifically, when the target energy value is randomly generated, the value range is [5, 20], and too small or too large target energy value will result in more low-frequency components, reducing the generalization ability of the random geometric modeling three-dimensional data set.

[0102] Step 07: render the target three-dimensional scene in the current cycle with an orthogonal camera to obtain an image data set in the current cycle, and add the image data set in the current cycle to the initial data set in the current cycle to obtain a target data set in the current cycle.

[0103] The image data set includes a depth image and an intensity image, the initial data set in the current cycle is the target data set in the last cycle, and the initial data set in the initial cycle is an empty set.

[0104] Specifically, the schematic diagram of rendering the target three-dimensional scene with an orthogonal camera is shown in Figure 2 The orthogonal multiple of the orthogonal camera is 3.84, and the orthogonal multiple is the width of the depth image and the intensity image in the rendered image data set. According to the 4K ratio, the height of the depth image and the intensity image can be obtained as 2.16. The position of the orthogonal camera is set as (0, 0, 1) to ensure that it will not be blocked by the three-dimensional geometric body.

[0105] The rendering resolution is 3840x2160, and the rendered image data is obtained. The depth image and the intensity image in the rendered image data set are both presented in RGB mode.

[0106] Step 08: determine whether the current cycle reaches the maximum cycle.

[0107] Step 09: if yes, the target data set in the current cycle is determined as the random geometric modeling three-dimensional data set.

[0108] Step 10: if no, the current cycle is updated to the next cycle, and returns to step 01 until the maximum cycle is reached, and the random geometric modeling three-dimensional data set is obtained.

[0109] The random geometric modeling three-dimensional data set is shown in Figure 3The intensity image with the power of 1, the depth image with the power of 2, the intensity image with the power of 5, the depth image with the power of 6, the intensity image with the power of 9, and the depth image with the power of 10 in the random geometry modeling three-dimensional data set are as shown in FIGS. 1-6. Figures 4-9

[0110] The present application introduces a uniform depth distribution fitting function to control the depth of the three-dimensional geometry, ensuring the uniformity of the pixels at different depths, so that the depth distribution law of the three-dimensional geometry in the target three-dimensional scene can be better used for neural network training, preventing the neural network from being biased towards any frequently occurring depth and producing undesirable results on those sparse depths when uneven pixel depth distribution occurs. The random geometry modeling three-dimensional data set has more details such as angles, edges, occlusions, and background information of lighting, compared to traditional data sets, and contains more image information. The distribution of high-frequency and low-frequency components in the frequency spectrum can also be flexibly adjusted. It allows flexible design of three-dimensional scenes by setting parameters, without relying on pre-existing three-dimensional models or external assets. Through Python scripting in Blender, automation of the entire data set generation is achieved, including random generation of three-dimensional geometry, rendering and saving of images, and the entire process is automatically executed by the program. This automated control greatly reduces the time and effort required for human intervention, and through parameterized settings, users can easily adjust the characteristics of the generated three-dimensional geometry and three-dimensional scene. These parameter settings make the generated geometry modeling scene not only have rich diversity, but also can flexibly adjust the scene complexity and geometry properties according to specific needs. By randomly generating the position, size and rotation angle of the three-dimensional geometry, the diversity and unpredictability of the random geometry modeling three-dimensional data set are ensured, which is particularly important for the training of deep learning models. The application of random noise texture and the setting of random position and rotation angle make each rendered target three-dimensional scene unique, thereby avoiding the homogenization problem caused by repetitive data in traditional methods.

[0111] In an exemplary embodiment, a computer device is provided, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor executing the computer program to implement the method for generating a random geometry modeling three-dimensional data set.

[0112] In an exemplary embodiment, a computer readable storage medium is provided, having stored thereon a computer program, which, when executed by a processor, implements the method for generating a random geometry modeling three-dimensional data set.

[0113] In an exemplary embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the method for generating a random geometry modeling three-dimensional data set.​

[0114] In an exemplary embodiment, a computer device, which can be a server or a terminal, has an internal structure as shown in Figure 10 The computer device includes a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with external terminals through a network connection. The computer program is executed by the processor to implement a method for generating a three-dimensional data set based on random geometry modeling.

[0115] Those skilled in the art can understand that Figure 10 The structure shown in the above description is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0116] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of the relevant data need to comply with relevant regulations.

[0117] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0118] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0119] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0120] The principles and implementation modes of the present application are described by applying specific examples herein, and the above-mentioned embodiments are only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range will be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A method for generating a random geometry model of a three-dimensional data set, characterized in that, The generation method of the random geometry modeling three-dimensional data set comprises the following steps: creating an empty three-dimensional scene under the current cycle number in the Blender software; randomly generating floating numbers of a plurality of three-dimensional geometries under the current cycle number within a set numerical range, and determining depths of the corresponding three-dimensional geometries based on the floating numbers of the three-dimensional geometries under the current cycle number and a uniform depth distribution fitting function; randomly generating spatial parameters, sizes and noise textures of the three-dimensional geometries under the current cycle number; the spatial parameters comprise a position, a rotation angle and a scaling ratio; generating a plurality of initial three-dimensional geometries under the current cycle number in the empty three-dimensional scene based on the depths, spatial parameters and sizes of the three-dimensional geometries under the current cycle number; applying the noise textures of the three-dimensional geometries under the current cycle number to the corresponding initial three-dimensional geometries respectively to obtain a plurality of target three-dimensional geometries under the current cycle number, thereby obtaining an initial three-dimensional scene under the current cycle number containing the target three-dimensional geometries; randomly generating an energy value of a daylight lamp light source under the current cycle number to obtain a target energy value under the current cycle number, and setting the daylight lamp light source with the target energy value under the current cycle number in the initial three-dimensional scene under the current cycle number to obtain a target three-dimensional scene under the current cycle number; setting an orthogonal camera for rendering on the target three-dimensional scene under the current cycle number to obtain an image data set under the current cycle number, and adding the image data set under the current cycle number to an initial data set under the current cycle number to obtain a target data set under the current cycle number; the image data set comprises a depth image and an intensity image, the initial data set under the current cycle number is a target data set under a previous cycle number, and an initial data set under an initial cycle number is an empty set; determining whether the current cycle number reaches a maximum cycle number; if yes, determining the target data set under the current cycle number as the random geometry modeling three-dimensional data set; if no, updating the current cycle number to a next cycle number, and returning to the step of creating the empty three-dimensional scene under the current cycle number in the Blender software until the maximum cycle number is reached, and obtaining the random geometry modeling three-dimensional data set.

2. The method of claim 1, wherein, The set numerical range is [0, 1].

3. The method of claim 1, wherein, The uniform depth distribution fitting function comprises: wherein P(r) is the depth of the three-dimensional geometry, and r is the floating number of the three-dimensional geometry.

4. The method of claim 3, wherein, The determination process of the uniform depth distribution fitting function comprises the following steps: constructing a cumulative distribution function; the cumulative distribution function is: wherein C(n) is a cumulative distribution function value of an nth depth plane in a line of sight of the orthogonal camera; N is a total number of depth planes in the line of sight of the orthogonal camera; i and k are both layer numbers of the depth planes in the line of sight of the orthogonal camera; the fitting inverse function of the cumulative distribution function is obtained through value fitting, and the fitting inverse function is determined as the uniform depth distribution fitting function.

5. The method of claim 1, wherein, The generation process of the noise texture of any three-dimensional geometry comprises the following steps: randomly generating an initial height pixel number of a noise image of the three-dimensional geometry; determine an initial width pixel number of a noise image of the three-dimensional geometric body based on the initial height pixel number; generate an initial noise image based on the initial height pixel number and the initial width pixel number of the noise image; perform interpolation enlargement on the initial noise image according to a preset height pixel number and a preset width pixel number to obtain a target noise image of the three-dimensional geometric body; determine the target noise image of the three-dimensional geometric body as a noise texture of the three-dimensional geometric body.

6. The method of claim 5, wherein, randomly generating an initial height pixel number of a noise image of a three-dimensional geometric body comprises: randomly generating an exponent number of a noise image of a three-dimensional geometric body; calculating the initial height pixel number of the noise image of the three-dimensional geometric body according to the exponent number by using an initial height pixel number calculation formula; the initial height pixel number calculation formula is: H noise ＝2 m ； wherein H noise is the initial height in pixels; m is the power.

7. The method of claim 6, wherein, the noise texture is a low-frequency noise texture, a medium-frequency noise texture or a high-frequency noise texture; when the noise texture is a low-frequency noise texture, the value range of the exponent number is [1, 2]; when the noise texture is a medium-frequency noise texture, the value range of the exponent number is [5, 6]; when the noise texture is a high-frequency noise texture, the value range of the exponent number is [9, 10].

8. A computer apparatus comprising: A memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the generation method of the random geometric modeling three-dimensional data set in any one of claims 1-7.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the generation method of the random geometric modeling three-dimensional data set in any one of claims 1-7.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the generation method of the random geometric modeling three-dimensional data set in any one of claims 1-7.

Images