Similarity determination method and apparatus, and device, medium and product

Abstract

Disclosed in the present application are a similarity determination method and apparatus, and a device, a medium and a product. The method comprises: firstly, acquiring a rendered image set of a first mesh and a rendered image set of a second mesh; and then, performing similarity analysis processing on the rendered image set of the first mesh and the rendered image set of the second mesh, so as to obtain similarity representation data between the first mesh and the second mesh.

Description

A similarity determination method, device, equipment, medium, product

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese invention patent application with application number 202410182185.6, entitled “A method, device, equipment, medium, product for determining similarity” and filed on February 18, 2024, and the entire application is incorporated herein by reference. Technical Field

[0003] The present application relates to the field of data processing technology, and in particular to a similarity determination method, apparatus, device, medium, and product. Background Art

[0004] A three-dimensional mesh is a complex structure composed of a series of vertices, edges, and faces, and a three-dimensional mesh can be used to represent the geometric shape of a three-dimensional object. Summary of the Invention

[0005] The present application provides a similarity determination method, apparatus, device, medium, and product that can relatively accurately determine the similarity between different three-dimensional grids.

[0006] In order to achieve the above objectives, the technical solutions provided by this application are as follows:

[0007] The present application provides a similarity determination method, the method comprising:

[0008] Obtaining a rendering image set of a first grid and a rendering image set of a second grid; the rendering image set of the first grid is used to describe the state of the first grid under multiple viewing angles; the rendering image set of the second grid is used to describe the state of the second grid under the multiple viewing angles;

[0009] A similarity analysis process is performed on the rendered image set of the first grid and the rendered image set of the second grid to obtain similarity representation data between the first grid and the second grid.

[0010] In a possible implementation manner, the rendered image set includes a grid imaging image at each of the viewing angles and a grid edge image at each of the viewing angles.

[0011] In a possible implementation manner, there is a correspondence between the first image in the rendered image set of the first grid and the second image in the rendered image set of the second grid;

[0012] The process of determining the similarity representation data between the first grid and the second grid includes:

[0013] determining similarity representation data between the first image and the second image based on similarity representation data between image features of the first image and image features of the second image;

[0014] Similarity representation data between the first grid and the second grid is determined based on the similarity representation data between the first image and the second image.

[0015] In one possible implementation, the process of determining the similarity representation data between the first image and the second image includes:

[0016] Acquire a reference image corresponding to the second image, where the reference image is obtained by performing noise processing on the second image;

[0017] Similarity characterization data between the first image and the second image is determined based on the similarity characterization data between the image features of the first image and the image features of the second image, and the similarity characterization data between the image features of the first image and the image features of the reference image.

[0018] In a possible implementation manner, the multiple viewing angles are determined based on at least two spherical positions on a minimum circumscribed sphere of a preset three-dimensional space;

[0019] The process of obtaining the rendered image set includes:

[0020] restricting both the first grid and the second grid to the preset three-dimensional space; wherein the center of the restricted first grid is located at the center of the preset three-dimensional space, and the center of the restricted second grid is located at the center of the preset three-dimensional space;

[0021] performing rendering processing on the restricted first grid at the at least two spherical surface positions to obtain a rendering image set of the first grid;

[0022] The restricted second grid is rendered at the at least two spherical positions to obtain a rendering image set of the second grid.

[0023] In a possible implementation manner, the minimum circumscribed cube of the first grid after the restriction is the preset three-dimensional space, and the minimum circumscribed cube of the second grid after the restriction is the preset three-dimensional space.

[0024] In a possible implementation manner, the preset three-dimensional space is a cube; the at least two spherical positions include spherical positions corresponding to each face in the preset three-dimensional space and spherical positions corresponding to each edge in the preset three-dimensional space.

[0025] The present application provides a similarity determination device, comprising:

[0026] an acquisition unit, configured to acquire a rendering image set of a first grid and a rendering image set of a second grid; the rendering image set of the first grid is used to describe the state of the first grid under multiple viewing angles; the rendering image set of the second grid is used to describe the state of the second grid under the multiple viewing angles;

[0027] An analyzing unit is configured to perform similarity analysis on the rendered image set of the first grid and the rendered image set of the second grid to obtain similarity representation data between the first grid and the second grid.

[0028] The present application provides an electronic device, the device comprising: a processor and a memory;

[0029] The memory is used to store instructions or computer programs;

[0030] The processor is configured to execute the instructions or computer program in the memory, so that the electronic device executes the similarity determination method provided in this application.

[0031] The present application provides a computer-readable medium having instructions or a computer program stored therein. When the instructions or the computer program are executed on a device, the device executes the similarity determination method provided in the present application.

[0032] The present application provides a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program codes for executing the similarity determination method provided by the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0034] FIG1 is a flow chart of a similarity determination method provided in an embodiment of the present application;

[0035] FIG2 is a schematic diagram of a similarity determination process provided in an embodiment of the application;

[0036] FIG3 is a schematic diagram of the structure of a similarity determination device provided in an embodiment of the present application;

[0037] FIG4 is a schematic structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0038] In addition, for some application scenarios, such as 3D mesh comparison scenarios, 3D mesh retrieval scenarios, 3D mesh classification scenarios, etc., these application scenarios may have the following requirements: determining the similarity between different 3D meshes.

[0039] Research has found that for some related schemes for determining similarity, these schemes can determine similarity by calculating the differences between the topological structures of different three-dimensional meshes, such as by calculating the differences between vertex coordinates, face normal vectors, boundary curvature, etc.

[0040] After research, it was found that the above-mentioned related solutions have the defects shown in ①-② below.

[0041] ① The above-mentioned related schemes are relatively inefficient when processing large-scale three-dimensional grids. Specifically, because the related schemes need to perform a large number of calculations and comparison operations to determine the similarity between different three-dimensional grids, the computing time and resources required when the three-dimensional grid has hundreds of thousands or even millions of vertices are relatively large, thus limiting the feasibility of the practical application of the related schemes.

[0042] ② The above-mentioned related solutions show relatively poor robustness on three-dimensional meshes with complex structures. Specifically, in some application scenarios of three-dimensional meshes, three-dimensional meshes with complex structures may be encountered, such as three-dimensional meshes with small geometric details, holes, cracks or discontinuous surfaces. However, the related solutions often show low robustness when facing these three-dimensional meshes with complex structures. For example, when processing three-dimensional meshes with holes or discontinuous surfaces, the related solutions may not be able to correctly identify and pair corresponding vertices, edges or patches. This leads to inaccurate and unstable evaluation results.

[0043] Further research has revealed that in some application scenarios, such as virtual reality (VR), similarity is influenced not only by the geometry of the 3D mesh but also by its visual information. However, the aforementioned approaches focus solely on the geometric differences between different 3D meshes, ignoring differences in visual information. This results in inaccurate similarity.

[0044] Based on the above findings, in order to better improve accuracy, the present application provides a similarity determination method, which includes: first obtaining a rendering image set of a first grid and a rendering image set of a second grid, so that the rendering image set of the first grid is used to describe the state of the first grid under multiple viewing angles, and the rendering image set of the second grid is used to describe the state of the second grid under the multiple viewing angles; then performing similarity analysis on the rendering image set of the first grid and the rendering image set of the second grid to obtain similarity representation data between the first grid and the second grid. In particular, because the above rendering image set is used to describe the state of a three-dimensional grid under multiple viewing angles, the rendering image set can not only represent the geometric characteristics of the three-dimensional grid, but also represent the visual characteristics of the three-dimensional grid, such as edge characteristics, texture characteristics, color distribution characteristics, etc., so that the rendering image set can represent the characteristics of the three-dimensional grid as comprehensively as possible, and further, the similarity representation data determined based on the rendering image set can more accurately represent the degree of similarity between different three-dimensional grids, which is conducive to improving accuracy.

[0045] In addition, the present application does not limit the execution subject of the similarity determination method provided in the embodiment of the present application. For example, the similarity determination method provided in the embodiment of the present application can be applied to a terminal device or a server. For another example, the similarity determination method provided in the embodiment of the present application can also be implemented with the help of the data interaction process between the terminal device and the server. Among them, the terminal device can be a smart phone, a computer, a personal digital assistant (PDA), a tablet computer, etc. The server can be a stand-alone server, a cluster server, or a cloud server.

[0046] In order to help those skilled in the art better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0047] To better understand the technical solution provided by this application, the similarity determination method provided by this application is described below with reference to some drawings. As shown in Figure 1, the similarity determination method provided by the embodiment of this application includes the following S1-S2. Figure 1 is a flow chart of a similarity determination method provided by the embodiment of this application.

[0048] S1: Obtain a rendering image set of a first grid and a rendering image set of a second grid; the rendering image set of the first grid is used to describe the state of the first grid under multiple viewing angles; the rendering image set of the second grid is used to describe the state of the second grid under multiple viewing angles.

[0049] Among them, the first grid refers to a three-dimensional grid that needs to be processed for similarity determination, such as the grid 1 shown in Figure 2; and this application does not limit the implementation method of the first grid. For example, the first grid can refer to a pre-set standard three-dimensional grid.

[0050] For the first mesh described above, the rendered image set of the first mesh is obtained by rendering the first mesh, so that the rendered image set is used to describe the state of the first mesh from multiple perspectives. This allows the rendered image set to describe the overall state of the first mesh and, in turn, to express the characteristics of the first mesh, such as geometric characteristics and visual characteristics, as comprehensively as possible. The multiple perspectives refer to the perspectives described by the two-dimensional images in the rendered image set, so that the multiple perspectives can represent the perspectives required when acquiring the two-dimensional images in the rendered image set.

[0051] The second grid refers to another three-dimensional grid that needs to be processed for similarity determination, such as the grid 2 shown in Figure 2; and this application does not limit the implementation method of the second grid. For example, the second grid can refer to a three-dimensional grid that needs to be quality evaluated.

[0052] For the second mesh described above, the rendered image set of the second mesh is obtained by rendering the second mesh, so that the rendered image set is used to describe the state of the second mesh under multiple perspectives, thereby enabling the rendered image set to describe the overall state of the second mesh, and further enabling the rendered image set to express the characteristics of the second mesh as comprehensively as possible, such as geometric characteristics, visual characteristics, etc.

[0053] It should be noted that this application does not limit the relationship between the multiple perspectives described by the set of rendered images of the second mesh and the multiple perspectives described by the set of rendered images of the first mesh. For example, a pre-defined correspondence between the two is possible. Alternatively, to further improve the accuracy of similarity determination, the two perspectives can be identical, allowing subsequent similarity analysis of the 2D rendered images of different 3D meshes at the same perspective to determine the similarity between the different 3D meshes.

[0054] In addition, the present application does not limit the implementation method of the above S1. For example, it may specifically include the following steps 11 to 14.

[0055] Step 11: Determine a camera placement position corresponding to the first grid. The camera placement position is used to describe a camera placement position referenced when rendering the first grid.

[0056] Among them, the camera placement position corresponding to the first grid refers to the camera placement position required to be referenced when rendering the first grid, so that the camera placement position can indicate where the camera needs to be placed when rendering the first grid, thereby enabling the camera placement position to indicate multiple perspectives referenced when rendering the first grid, so that the imaging of the first grid under the multiple perspectives can be subsequently obtained based on the camera placement position.

[0057] In addition, the present application does not limit the implementation method of the camera placement position corresponding to the first grid above. For example, the camera placement position corresponding to the first grid may include camera placement positions corresponding to multiple perspectives, so that the camera placement position corresponding to each perspective is used to obtain the imaging of the first grid under the corresponding perspective.

[0058] In addition, the present application does not limit the method for determining the camera placement position corresponding to the first grid above. For example, it can be implemented using any existing or future method that can determine the camera placement position of a three-dimensional grid.

[0059] Step 12: Render the first grid at the camera placement position corresponding to the first grid to obtain a rendered image set of the first grid.

[0060] It should be noted that the present application does not limit the implementation method of the above step 12. For example, when the camera placement positions corresponding to the above first grid include camera placement positions corresponding to N perspectives, the step 12 may specifically include the following steps 121-122.

[0061] Step 121: Render the first mesh at the camera placement position corresponding to the nth viewing angle to obtain a rendering result of the first mesh at the nth viewing angle, so that the rendering result can represent the imaging of the first mesh at the nth viewing angle, where n is a positive integer, n≤N, and N is a positive integer.

[0062] The rendering result of the first grid at the n-th viewing angle is used to represent the imaging of the first grid at the n-th viewing angle.

[0063] In addition, the present application does not limit the implementation method of the above "rendering result of the first grid at the nth viewing angle". For example, when the above rendering process is implemented using differentiable rendering technology, the rendering result may include a grid imaging map of the first grid at the nth viewing angle, such as a red, green, and blue (RGB) map. The grid imaging map is used to describe the state of the first grid at the nth viewing angle, such as the color distribution state, geometric structure state, texture state, etc.; and the present application does not limit the implementation method of the grid imaging map. For example, the grid imaging map can be implemented using an RGB map. It should be noted that differentiable rendering technology is a rendering method that can generate gradient information about camera parameters; and the present application does not limit the differentiable rendering technology.

[0064] For example, in some application scenarios, to further improve the accuracy of similarity determination, when the rendering process described above is implemented using differentiable rendering technology, the "rendering result of the first mesh at the nth viewing angle" described above may include a mesh image of the first mesh at the nth viewing angle and a mesh edge map of the first mesh at the nth viewing angle. The mesh edge map is used to describe the edge characteristics of the first mesh at the nth viewing angle.

[0065] Based on the relevant content of step 121 above, it can be known that for any position of the camera placement position corresponding to the first grid, the first grid can be imaged at this position using differentiable rendering technology to obtain the RGB image and grid edge image of the first grid at the viewing angle corresponding to the position, so that these two images can more comprehensively represent the characteristics of the first grid at the viewing angle corresponding to the position.

[0066] Step 122: constructing a rendering image set of the first mesh using the rendering results of the first mesh at various viewing angles, so that the rendering image set includes the rendering results of the first mesh at various viewing angles.

[0067] Based on the relevant content of steps 121 to 122 above, it can be known that for some application scenarios, after obtaining the camera placement position corresponding to the first grid above, such as the camera placement position corresponding to each perspective, the first grid is first imaged at each position using a differentiable rendering technique to obtain an RGB image and a grid edge image of the first grid at the corresponding perspective; then, based on these RGB images and these grid edge images, a rendered image set of the first grid is constructed, so that the rendered image set of the first grid includes the RGB images and the grid edge images at each perspective, so that the rendered image set can represent the characteristics of the first grid as comprehensively as possible.

[0068] Based on the relevant content of step 12 above, it can be known that for the first grid above, after obtaining the camera placement position corresponding to the first grid, the first grid is rendered according to the camera placement position to obtain a rendered image set of the first grid, so that the rendered image set can represent the state of the first grid under multiple perspectives, thereby enabling the rendered image set to represent the characteristics of the first grid as comprehensively as possible.

[0069] Step 13: Determine the camera placement position corresponding to the second grid. The camera placement position is used to describe the camera placement position referenced when rendering the second grid.

[0070] Among them, the camera placement position corresponding to the second grid refers to the camera placement position required to be referenced when rendering the second grid, so that the camera placement position can indicate where the camera needs to be placed when rendering the second grid, thereby enabling the camera placement position to indicate multiple perspectives referenced when rendering the second grid, so that the imaging of the second grid under the multiple perspectives can be subsequently obtained based on the camera placement position.

[0071] In addition, the present application does not limit the implementation method of the camera placement position corresponding to the second grid above. For example, the camera placement position corresponding to the second grid may include camera placement positions corresponding to multiple perspectives, so that the camera placement position corresponding to each perspective is used to obtain the imaging of the second grid under the corresponding perspective.

[0072] In addition, the present application does not limit the method for determining the camera placement position corresponding to the second grid above. For example, it can be implemented using any existing or future method that can determine the camera placement position of a three-dimensional grid.

[0073] It should be noted that this application does not limit the execution time of the above step 13, as long as the execution time of step 13 is ensured to be earlier than the execution time of the following step 14.

[0074] Step 14: Render the second grid at the camera placement position corresponding to the second grid to obtain a rendered image set of the second grid.

[0075] It should be noted that the present application does not limit the implementation of the above step 14. For example, the implementation of the above step 14 is similar to the implementation of the above step 12. For ease of understanding, the following is an explanation with reference to examples.

[0076] As an example, when the camera placement positions corresponding to the second grid include camera placement positions corresponding to N viewing angles, step 14 may specifically include the following steps 141 and 142.

[0077] Step 141: Render the second grid at the camera placement position corresponding to the nth viewing angle to obtain a rendering result of the second grid at the nth viewing angle, so that the rendering result can represent the imaging of the second grid at the nth viewing angle, where n is a positive integer, n≤N, and N is a positive integer.

[0078] The rendering result of the second grid at the n-th viewing angle is used to represent the imaging of the second grid at the n-th viewing angle.

[0079] In addition, this application does not limit the implementation method of the rendering result of the second grid at the nth viewing angle. For example, when the above rendering process is implemented using differentiable rendering technology, the rendering result may include a grid imaging map of the second grid at the nth viewing angle; the grid imaging map is used to describe the state of the second grid at the nth viewing angle, such as color distribution state, geometric structure state, texture state, etc. For another example, in some application scenarios, in order to better improve the accuracy of similarity determination, when the rendering process is implemented using differentiable rendering technology, the rendering result may include a grid imaging map of the second grid at the nth viewing angle, and a grid edge map of the second grid at the nth viewing angle; the grid edge map is used to describe the edge characteristics of the second grid at the nth viewing angle.

[0080] Based on the relevant content of step 141 above, it can be known that for any position of the camera placement position corresponding to the second grid, the second grid can be imaged at this position using differentiable rendering technology to obtain the RGB image and grid edge image of the second grid at the viewing angle corresponding to the position, so that these two images can more comprehensively represent the characteristics of the second grid at the viewing angle corresponding to the position.

[0081] Step 142: constructing a rendering image set of the second grid using the rendering results of the second grid at each viewing angle, so that the rendering image set includes the rendering results of the second grid at each viewing angle.

[0082] Based on the relevant contents of steps 141 to 142 above, it can be known that for some application scenarios, after obtaining the camera placement position corresponding to the second grid above, such as the camera placement position corresponding to each viewing angle, the second grid is first imaged at each position using a differentiable rendering technique to obtain an RGB image and a grid edge image of the second grid at the corresponding viewing angle; and then, based on these RGB images and these grid edge images, a rendered image set of the second grid is constructed, so that the rendered image set of the second grid includes the RGB images and the grid edge images at each viewing angle, so that the rendered image set can represent the characteristics of the second grid as comprehensively as possible.

[0083] In addition, this application does not limit the execution time of the above step 14, as long as the execution time of the above step 13 is ensured to be later than the execution time of the above step 13.

[0084] It can be seen that for the second grid above, after obtaining the camera placement position corresponding to the second grid, the second grid can be rendered according to the camera placement position to obtain a rendered image set of the second grid, so that the rendered image set can represent the state of the second grid under multiple perspectives, thereby enabling the rendered image set to represent the characteristics of the second grid as comprehensively as possible.

[0085] Based on the relevant content of steps 11 to 14 above, it can be seen that in some application scenarios, due to the different sizes of different 3D grids, the 3D spaces in which the different 3D grids are located are different. Therefore, in order to ensure accuracy, the camera placement position corresponding to each 3D grid can be determined separately; then, for any 3D grid, the 3D grid is rendered according to the camera placement position corresponding to the 3D grid to obtain a rendered image set of the 3D grid.

[0086] In fact, in order to better improve efficiency, the present application also provides a possible implementation of the above S1. Under this implementation, when the above multiple perspectives are determined based on at least two spherical positions on the minimum circumscribed sphere of a preset three-dimensional space, the S1 can specifically include the following steps 21-23.

[0087] Step 21: Confine both the first grid and the second grid to a preset three-dimensional space; the center of the restricted first grid is located at the center of the preset three-dimensional space, and the center of the restricted second grid is located at the center of the preset three-dimensional space.

[0088] Among them, the preset three-dimensional space refers to a pre-set space required to use when restricting different three-dimensional grids to the same three-dimensional space; and this application is not limited to the preset three-dimensional space. For example, the preset three-dimensional space can be implemented using a cube, such as a cube with a side length of 2.

[0089] Furthermore, for the aforementioned preset three-dimensional space, a camera can be placed on the minimum circumscribed sphere of the preset three-dimensional space, such that the camera can sequentially move to at least two spherical positions on the minimum circumscribed sphere, imaging the three-dimensional grid confined within the preset three-dimensional space to obtain images of the three-dimensional grid from multiple perspectives. Thus, in one possible implementation, the multiple perspectives referred to in S1 above may include the camera perspective corresponding to each of the at least two spherical positions.

[0090] In addition, the present application does not limit the implementation method of the above-mentioned "at least two spherical positions on the minimum circumscribed sphere of the preset three-dimensional space". For example, to ensure a more comprehensive description of the three-dimensional grid, the at least two spherical positions can satisfy the following constraint: the union of the camera shooting ranges corresponding to the at least two spherical positions can fully cover the preset three-dimensional space, so that subsequent cameras can traverse the preset three-dimensional space by sequentially moving to the at least two spherical positions. For another example, to better improve efficiency while ensuring comprehensiveness, when the preset three-dimensional space is a cube, such as a cube with a side length of 2, the at least two spherical positions can include the spherical positions corresponding to each face in the preset three-dimensional space and the spherical positions corresponding to each edge in the preset three-dimensional space, so as to obtain comprehensive information of the three-dimensional grid with as few positions as possible, so that subsequent cameras can traverse the preset three-dimensional space with as few movements as possible.

[0091] The restricted first grid refers to a result obtained by restricting the first grid to a preset three-dimensional space, so that subsequent rendering processing of the first grid can be completed based on the restricted first grid and a camera deployed for the preset three-dimensional space.

[0092] In addition, the first grid after the above restriction may satisfy the following constraint: the center of the first grid after the restriction is located at the center position of the preset three-dimensional space.

[0093] In addition, in order to better improve the accuracy, the present application also provides a possible implementation of the first grid after the above-mentioned restriction, under which the minimum circumscribed cube of the first grid after the restriction is a preset three-dimensional space, so that the first grid after the restriction satisfies the following constraints: the center of the first grid after the restriction is located at the center position of the preset three-dimensional space, and the size of the minimum circumscribed cube of the first grid after the restriction is equal to the size of the preset three-dimensional space.

[0094] The restricted second grid refers to the result obtained by restricting the second grid to a preset three-dimensional space, so that subsequent rendering processing of the second grid can be completed based on the restricted second grid and the camera deployed for the preset three-dimensional space.

[0095] In addition, the second grid after the above restriction may satisfy the following constraint: the center of the second grid after the restriction is located at the center position of the preset three-dimensional space.

[0096] In addition, in order to better improve the accuracy, the present application also provides a possible implementation of the second grid after the above restriction, in which the minimum circumscribed cube of the second grid after the restriction is a preset three-dimensional space, so that the second grid after the restriction satisfies the following constraints: the center of the second grid after the restriction is located at the center position of the preset three-dimensional space, and the size of the minimum circumscribed cube of the second grid after the restriction is equal to the size of the preset three-dimensional space.

[0097] In addition, the present application does not limit the implementation method of the above step 21. For example, it may specifically include the following steps 211 to 214.

[0098] Step 211: Determine a zoom ratio corresponding to the first grid according to the size of the first grid and the size of the preset three-dimensional space.

[0099] The size of the first grid is used to represent the scale of the first grid. The present application does not limit the implementation of the size of the first grid. For example, the size of the first grid may be determined based on the three-dimensional coordinate values ​​of the two farthest nodes in the first grid.

[0100] In addition, for the first grid above, the scaling ratio corresponding to the first grid refers to the size scaling ratio required when restricting the first grid to a preset three-dimensional space; and the present application does not limit the process of determining the scaling ratio. For example, the scaling ratio can be determined based on the ratio between the size of the preset three-dimensional space and the size of the first grid.

[0101] Step 212: According to the scaling ratio corresponding to the first grid, the first grid is restricted to a preset three-dimensional space to obtain a restricted first grid.

[0102] In the present application, for the first grid above, after determining the scaling ratio corresponding to the first grid, the first grid can be restricted to a preset three-dimensional space based on the scaling ratio to obtain the restricted first grid, so that the ratio between the size of the restricted first grid and the size of the first grid before the restriction reaches the scaling ratio, thereby making the size of the minimum circumscribed cube of the restricted first grid equal to the size of the preset three-dimensional space.

[0103] Step 213: Determine a zoom ratio corresponding to the second grid according to the size of the second grid and the size of the preset three-dimensional space.

[0104] The size of the second grid is used to represent the scale of the second grid. The present application does not limit the implementation method of the size of the second grid. For example, the size of the second grid can be determined based on the three-dimensional coordinate values ​​of the two farthest nodes in the second grid.

[0105] In addition, for the second grid above, the scaling ratio corresponding to the second grid refers to the size scaling ratio required when restricting the second grid to the preset three-dimensional space; and the present application does not limit the determination process of the scaling ratio. For example, the scaling ratio can be determined based on the ratio between the size of the preset three-dimensional space and the size of the second grid.

[0106] In addition, the present application does not limit the execution time of the above step 213, as long as the execution time of the step 213 is ensured to be earlier than the execution time of the following step 214.

[0107] Step 214: According to the scaling ratio corresponding to the second grid, the second grid is restricted to a preset three-dimensional space to obtain a restricted second grid.

[0108] In the present application, for the second grid above, after determining the scaling ratio corresponding to the second grid, the second grid can be restricted to a preset three-dimensional space based on the scaling ratio to obtain a restricted second grid, so that the ratio between the size of the restricted second grid and the size of the second grid before the restriction reaches the scaling ratio, thereby making the size of the minimum circumscribed cube of the restricted second grid equal to the size of the preset three-dimensional space.

[0109] In addition, the present application does not limit the execution time of the above step 214, as long as the execution time of the above step 214 is ensured to be later than the execution time of the above step 213.

[0110] Based on the relevant contents of steps 211 to 214 above, it can be seen that in some application scenarios, different three-dimensional grids can be restricted to the same spatial range by means of size scaling to achieve alignment in the spatial range, which is conducive to improving the comparability between different three-dimensional grids.

[0111] Based on the relevant content of step 21 above, it can be known that for some application scenarios, after obtaining the first grid and the second grid, if the size of the first grid is different from the size of the second grid, the first grid and the second grid can be restricted to the preset three-dimensional space to obtain the restricted first grid and the restricted second grid, so that the minimum circumscribed cube of the two grids is the preset three-dimensional space, so that the two grids are aligned in the spatial range, and then the two grids can be compared in the same spatial range, so that the subsequent rendering processing of each grid can be achieved with the help of the camera placement position determined for the preset three-dimensional space, which is conducive to reducing computational complexity and improving computational accuracy.

[0112] Step 22: Rendering the restricted first mesh at the at least two spherical positions to obtain a rendering image set of the first mesh.

[0113] In the present application, for the above-mentioned preset three-dimensional space, a camera is placed on the minimum circumscribed sphere of the preset three-dimensional space, so that the camera can obtain the imaging of the above-mentioned restricted first grid under multiple perspectives by moving to at least two of the above-mentioned spherical positions in sequence; and the process can be specifically as follows: first, for any spherical position, the restricted first grid is rendered to obtain the rendering result corresponding to the spherical position, such as a grid imaging image and a grid edge image; then, using the rendering results corresponding to these spherical positions, a rendering image set of the first grid is constructed, so that the rendering image set includes the rendering results corresponding to these spherical positions, such as 18 grid imaging images and 18 grid edge images.

[0114] Step 23: Rendering the restricted second grid at the at least two spherical positions to obtain a rendering image set of the second grid.

[0115] In the present application, for the above-mentioned preset three-dimensional space, a camera is placed on the minimum circumscribed sphere of the preset three-dimensional space, so that the camera can obtain the imaging of the above-restricted second grid from multiple perspectives by moving to at least two of the above-restricted spherical positions in sequence; and the process can be specifically as follows: first, for any spherical position, the restricted second grid is rendered to obtain the rendering result corresponding to the spherical position, such as a grid imaging image and a grid edge image; then, using the rendering results corresponding to these spherical positions, a rendering image set of the second grid is constructed, so that the rendering image set includes the rendering results corresponding to these spherical positions, such as 18 grid imaging images and 18 grid edge images.

[0116] It should be noted that this application does not limit the correlation between the execution time of step 23 and the execution time of step 22. For example, the former may be earlier than the latter. Another example is that the former is later than the latter. Another example is that the two are the same.

[0117] Based on the relevant content of steps 21 to 23 above, it can be seen that for some application scenarios, when the size of the first grid above is different from the size of the second grid above, after obtaining the first grid and the second grid, both grids can be first restricted to a preset three-dimensional space to ensure that the two grids are aligned in spatial range; then, a camera is arranged for the preset three-dimensional space so that the camera and at least two spherical positions on the minimum circumscribed sphere of the preset three-dimensional space can be used to complete the rendering processing for the two grids, thereby obtaining a rendered image set of the two grids. Among them, because the two three-dimensional grids are aligned in spatial range, the rendered image sets of the two three-dimensional grids are more comparable, thereby making the similarity determined based on the rendered image sets of the two three-dimensional grids more accurate. In addition, since these two three-dimensional grids are restricted to the same spatial range, they can both be rendered using the camera deployed for this spatial range. This helps reduce computational complexity, thereby effectively overcoming the resources consumed by separately determining the camera placement positions of different three-dimensional grids, such as time resources and computing resources, and thus helps reduce resource consumption, such as reducing time overhead and computing resource overhead.

[0118] Based on the relevant content of the rendering image set above, it can be seen that in one possible implementation, the rendering image set may include grid imaging images under each viewing angle and grid edge images under each viewing angle, so that the rendering image set can describe the state of a three-dimensional grid under different viewing angles as comprehensively as possible, thereby helping to improve accuracy.

[0119] Based on the content of the above paragraph, it can be seen that in a possible implementation, when the multiple perspectives involved in S1 above are N perspectives, such as 18 perspectives, the rendering image set of the first grid above may include the grid imaging image under the 1st perspective to the grid imaging image under the Nth perspective, and the grid edge image under the 1st perspective to the grid edge image under the Nth perspective; and the rendering image set of the second grid above may include the grid imaging image under the 1st perspective to the grid imaging image under the Nth perspective, and the grid edge image under the 1st perspective to the grid edge image under the Nth perspective.

[0120] Based on the relevant content of S1 above, it can be known that for the first grid and the second grid that have similarity determination requirements, a rendering image set of the first grid and a rendering image set of the second grid are obtained, so that the rendering image set of the first grid is used to describe the state of the first grid under multiple perspectives, and the rendering image set of the second grid is used to describe the state of the second grid under multiple perspectives, so that the similarity of the two three-dimensional grids can be determined subsequently with the help of similarity analysis processing on the two rendering image sets.

[0121] S2: performing similarity analysis on the rendered image set of the first grid and the rendered image set of the second grid to obtain similarity representation data between the first grid and the second grid.

[0122] The similarity characterization data between the first grid and the second grid is used to characterize the degree of similarity between the first grid and the second grid.

[0123] In addition, the present application does not limit the implementation method of the similarity analysis processing in S2 above. For example, it can be implemented by using any existing or future method that can perform similarity analysis processing on two image sets. For example, in order to better improve the similarity analysis effect, the similarity analysis processing in S2 can be implemented with the help of a pre-built image set similarity analysis model. Among them, the image set similarity analysis model is used to perform similarity analysis processing on the two input image sets of the model; and the present application does not limit the implementation method of the image set similarity analysis model. For example, it can be implemented by using any existing or future machine learning model that has learned how to perform similarity analysis processing on two image sets.

[0124] In addition, in order to better improve the accuracy of similarity determination, the present application also provides a possible implementation of S2 above. Under this implementation, when there is a correspondence between the first image in the rendered image set of the first grid above and the second image in the rendered image set of the second grid above, S2 can specifically include the following steps 31-32.

[0125] Step 31 : Determine similarity representation data between the first image and the second image based on similarity representation data between image features of the first image and image features of the second image.

[0126] The first image is used to represent any two-dimensional image in the set of rendered images of the first grid. For example, when the set of rendered images of the first grid includes 2N images, such as grid imaging images from N viewing angles and grid edge images from N viewing angles, the first image may refer to the i-th image in the set of rendered images of the first grid, where i is a positive integer and i≤2N.

[0127] The second image is used to represent a two-dimensional image corresponding to the first image in the rendered image set of the second grid above; and there is a correspondence between the second image and the first image. The correspondence is used to indicate that the second image and the first image are aligned in at least one dimension, so that the correspondence can indicate that similarity analysis can be performed on the second image and the first image in the future. The at least one dimension refers to the characteristic dimension of the alignment of the two images with the correspondence; and the present application does not limit the at least one dimension. For example, when the rendered image set above only includes grid imaging images under different perspectives, the at least one dimension can include the perspective so that the second image and the first image are aligned in perspective. For another example, when the rendered image set above includes some grid imaging images and some grid edge images, the at least one dimension can include the perspective and image type so that the second image and the first image are aligned in perspective and image type. The image type is used to describe whether an image belongs to a grid imaging image or a grid edge image.

[0128] Based on the content of the above paragraph, it can be seen that in one possible implementation, the following constraints exist between the second image and the first image: the perspective described by the second image is the same as the perspective described by the first image, and the image type to which the second image belongs is the same as the image type to which the second image belongs. The perspective described by the second image refers to the perspective used when acquiring the second image, such as the camera perspective. The perspective described by the first image refers to the perspective used when acquiring the first image, such as the camera perspective. The image type to which the second image belongs is used to indicate that the second image belongs to a grid imaging map or a grid edge map. The image type to which the first image belongs is used to indicate that the first image belongs to a grid imaging map or a grid edge map.

[0129] Based on the relevant content of the above-mentioned second image, it can be known that in some application scenarios, when the rendered image set of the second grid above includes 2N images, and there is a correspondence between the i-th image in the rendered image set of the second grid and the i-th image in the rendered image set of the first grid above, i is a positive integer, i≤2N, if the above-mentioned first image is the i-th image in the rendered image set of the first grid, then the second image can refer to the i-th image in the rendered image set of the second grid, i is a positive integer, i≤2N.

[0130] In addition, for the first image above, the image features of the first image are used to characterize the image information carried by the first image; and the present application does not limit the method for determining the image features. For example, it can be implemented using any existing or future method that can perform image feature extraction processing on an image. For another example, in order to improve the feature extraction effect, the image features of the first image can be determined with the help of a pre-built image feature extraction model. Among them, the image feature extraction model is used to perform image feature extraction processing on the input image of the model; and the present application does not limit the implementation method of the image feature extraction model. For example, it can be implemented with the help of a VGG model. Among them, the VGG model is a classic deep convolutional neural network, and the VGG model has good feature extraction capabilities, so that the VGG model can be used for hierarchical extraction of features of an image.

[0131] In addition, for the second image above, the image features of the second image are used to characterize the image information carried by the second image; and the method for determining the image features of the second image is similar to the method for determining the image features of the first image above. For the sake of brevity, they will not be repeated here.

[0132] Furthermore, for the first image and the second image described above, the similarity characterization data between the image features of the first image and the image features of the second image is used to characterize the degree of similarity between the two features; and this application does not limit the process for determining the similarity characterization data; for example, it can be implemented using any similarity measurement method, such as Euclidean distance, cosine similarity, etc. Thus, in one possible implementation, the similarity characterization data can be determined using the following formulas (1)-(3).

[0133] Where, represents the i-th image in the set of rendered images of the first grid above; represents the i-th image in the rendered image set of the second grid above; f i 1 Indicates that Image features of f i 2 Indicates that Image features; Indicates that f i 1 With the f i 2 The similarity between them represents the data; VGG(·) represents the representation function of the VGG model; MSE(·) represents the Euclidean distance calculation function; i is a positive integer, i≤2N.

[0134] In addition, the present application does not limit the implementation method of the above step 31. For example, it can specifically be: determining the similarity representation data between the image features of the first image and the image features of the second image as the similarity representation data between the first image and the second image.

[0135] In addition, in order to better improve the explainability, the present application also provides a possible implementation of the above step 31. Under this implementation, the step 31 may specifically include the following steps 311-312.

[0136] Step 311: Acquire a reference image corresponding to the second image, where the reference image is obtained by performing noise processing on the second image.

[0137] The reference image corresponding to the second image refers to the image required to be referenced when performing similarity determination processing on the second image and the first image, so that the reference image can represent the characteristics of the three-dimensional grid with the worst quality.

[0138] In addition, for the reference image corresponding to the second image described above, the reference image may be obtained by performing noise processing on the second image; and the present application does not limit the noise processing. For example, it may specifically be: adding K rounds of noise to the second image to obtain a reference image corresponding to the second image, so that the reference image can represent the characteristics of the worst-quality three-dimensional mesh, thereby enabling the subsequent determination of the worst similarity representation data based on the reference image. In this way, the worst similarity representation data corresponding to the second image can be automatically determined, thereby facilitating improved flexibility and accuracy. K is a positive integer.

[0139] Step 312 : Determine similarity representation data between the first image and the second image based on the similarity representation data between the image features of the first image and the image features of the second image, and the similarity representation data between the image features of the first image and the image features of the reference image.

[0140] Among them, the image features of the reference image are used to represent the image information carried by the reference image; and the method for determining the image features of the reference image is similar to the method for determining the image features of the first image above, and for the sake of brevity, it will not be repeated here.

[0141] In addition, for the first image and the reference image, the similarity characterization data between the image features of the first image and the image features of the reference image is used to characterize the degree of similarity between the two features; and the process for determining this similarity characterization data is similar to the process for determining the similarity characterization data between the image features of the first image and the image features of the second image described above. Thus, in one possible implementation, this similarity characterization data can be determined using the following formulas (4)-(5).

[0142] Where, represents the reference image corresponding to the i-th image in the rendered image set of the second grid above; f i noise Indicates that Image features of f i 1 represents the image features of the i-th image in the rendered image set of the first grid above; Indicates that f i 1 and The similarity between the characterization data, so that the Can show that in and The worst similarity characterization data required for reference when performing similarity determination processing; i is a positive integer, i≤2N.

[0143] In addition, the present application does not limit the implementation of the above step 312. For example, it may specifically include the following steps 3121 to 3123.

[0144] Step 3121: Calculate similarity representation data between the image features of the first image and the image features of the second image as first similarity representation data, so that the first similarity representation data can represent the degree of similarity between the image features of the first image and the image features of the second image.

[0145] Step 3122: Calculate the similarity representation data between the image features of the first image and the image features of the reference image as the second similarity representation data, so that the second similarity representation data can represent the degree of similarity between the image features of the first image and the image features of the reference image, thereby enabling the second similarity representation data to represent the worst similarity representation data required to be referenced when performing similarity determination processing on the second image and the first image.

[0146] Step 3123: Determine similarity representation data between the first image and the second image based on the ratio between the first similarity representation data and the second similarity representation data.

[0147] It should be noted that the present application does not limit the implementation method of the above step 3123. For example, it can specifically be: directly determining the ratio between the first similarity characterization data and the second similarity characterization data as the similarity characterization data between the first image and the second image.

[0148] For example, in some application scenarios, in order to better improve the interpretability, the present application also provides a possible implementation of the above step 3123. In this implementation, the step 3123 can be specifically as follows: first calculate the ratio between the first similarity representation data and the second similarity representation data; then determine the similarity representation data between the first image and the second image based on the difference between 1 and the ratio, such as the similarity representation data shown in formula (6) below, so that the similarity representation data belongs to the interval [0, 1], and when the similarity representation data is closer to 1, it means that the first image and the second image are more similar, and vice versa. This is conducive to improving interpretability.

[0149] Where, Representing similarity representation data between the i-th image in the set of rendered images of the first grid and the i-th image in the set of rendered images of the second grid,

[0150] Based on the relevant content of steps 3121 to 3123 above, it can be known that for the first image and the second image above, after obtaining the similarity characterization data between the image features of the first image and the image features of the second image, and the similarity characterization data between the image features of the first image and the image features of the reference image corresponding to the second image, the similarity characterization data between the first image and the second image can be determined based on the ratio between the two similarity characterization data, so that the similarity characterization data can better represent the degree of similarity between the first image and the second image, which is conducive to improving interpretability.

[0151] Based on the relevant contents of steps 311 to 312 above, it can be known that in some application scenarios, for the first image and the second image above, after obtaining the similarity representation data between the image features of the first image and the image features of the second image, in order to better improve the interpretability, the similarity representation data can be normalized using the reference image corresponding to the second image, such as the normalization processing shown in formulas (4)-(6) above, to obtain the normalized similarity representation data, so that the similarity representation data can better represent the degree of similarity between the first image and the second image, which is conducive to improving the interpretability.

[0152] Based on step 31 above, after obtaining the rendered image set of the first grid above and the rendered image set of the second grid above, for any pair of corresponding first images and second images existing in the two rendered image sets, first determine the image features of the first image and the image features of the second image; then, based on a preset similarity measurement method, such as Euclidean distance, calculate the similarity representation data between the image features of the first image and the image features of the second image, so that the similarity representation data can represent the degree of similarity between the image features of the two images; then, based on the similarity representation data, determine the similarity representation data between the first image and the second image, so that the similarity representation data can represent the degree of similarity between the image features of the two images.

[0153] Step 32: Determine similarity representation data between the first grid and the second grid based on the similarity representation data between the first image and the second image.

[0154] It should be noted that the present application does not limit the implementation method of the above step 32. For example, when there are 2N pairs of first images and second images with corresponding relationships in the rendering image set of the above first grid and the rendering image set of the above second grid, as shown in the following formula (4), the specific implementation method of step 32 can be: calculating the average value of the similarity representation data between the 2N pairs of first images and second images with corresponding relationships to obtain the similarity representation data between the first grid and the second grid.

[0155] Where S 12 Representing similarity characterization data between the first grid and the second grid; Similarity representation data representing the relationship between the i-th image in the rendered image set of the first grid and the i-th image in the rendered image set of the second grid, and a corresponding relationship exists between the i-th image in the rendered image set of the first grid and the i-th image in the rendered image set of the second grid, such as the same viewing angle and the same image type; i is a positive integer, i≤2N.

[0156] In fact, in order to better improve the accuracy, the present application also provides a possible implementation method of the above step 32. Under this implementation method, when there are 2N pairs of corresponding first images and second images in the rendering image set of the above first grid and the rendering image set of the above second grid, and the 2N pairs of corresponding first images and second images include N grid imaging pairs and N grid edge pairs, the step 32 can specifically include the following steps 321-323.

[0157] Step 321: Calculate the average value of the similarity representation data of the N grid image pairs to obtain a first average value, so that the first average value can represent the degree of similarity between the first grid and the second grid on the grid image.

[0158] The grid imaging image pair refers to an image pair that exists in the rendering image set of the first grid and the rendering image set of the second grid, has a corresponding relationship, and both belong to the grid imaging image.

[0159] In addition, the present application does not limit the implementation methods of the above-mentioned N grid imaging pairs. For example, when the rendering image set of the above-mentioned first grid includes N grid imaging images, and the rendering image set of the above-mentioned second grid includes N grid imaging images, if there is a correspondence between the h-th grid imaging image in the rendering image set of the first grid and the h-th grid imaging image in the rendering image set of the second grid, then the h-th grid imaging pair may include the h-th grid imaging image in the rendering image set of the first grid and the h-th grid imaging image in the rendering image set of the second grid, where h is a positive integer and h≤N.

[0160] In addition, for the h-th grid imaging pair, the similarity representation data of the h-th grid imaging pair is used to represent the similarity representation data between the two images in the h-th grid imaging pair, such as the similarity representation data between the h-th grid imaging image in the rendered image set of the first grid and the h-th grid imaging image in the rendered image set of the second grid, where h is a positive integer, h≤N.

[0161] Step 322: Calculate the average value of the similarity representation data of the N mesh edge graph pairs to obtain a second average value, so that the first average value can represent the degree of similarity between the first mesh and the second mesh on the mesh edge graph.

[0162] The grid edge image pair refers to an image pair that exists in the rendering image set of the first grid and the rendering image set of the second grid, has a corresponding relationship, and both belong to the grid edge image.

[0163] In addition, the present application does not limit the implementation methods of the above-mentioned N grid edge map pairs. For example, when the rendering image set of the above-mentioned first grid includes N grid edge maps, and the rendering image set of the above-mentioned second grid includes N grid edge maps, if there is a correspondence between the j-th grid edge map in the rendering image set of the first grid and the j-th grid edge map in the rendering image set of the second grid, then the j-th grid edge map pair may include the j-th grid edge map in the rendering image set of the first grid and the j-th grid edge map in the rendering image set of the second grid, where j is a positive integer and j≤N.

[0164] In addition, for the j-th grid edge map pair, the similarity representation data of the j-th grid edge map pair is used to represent the similarity representation data between the two images in the j-th grid edge map pair, such as the similarity representation data between the j-th grid edge map in the rendered image set of the first grid and the j-th grid edge map in the rendered image set of the second grid, where j is a positive integer, j≤N.

[0165] Step 323: Perform a weighted summation on the first average value and the second average value to obtain similarity representation data between the first grid and the second grid.

[0166] Based on the relevant contents of steps 321 to 322 above, it can be known that in some application scenarios, when the rendered image set of the first grid above includes 2N images, the N images arranged at the front of the rendered image set of the first grid are all grid imaging images, and the N images arranged at the back of the rendered image set of the first grid are all grid edge images, and the rendered image set of the second grid above includes 2N images, the N images arranged at the front of the rendered image set of the second grid are all grid imaging images, and the N images arranged at the back of the rendered image set of the second grid are all grid edge images, the similarity representation data between the first grid and the second grid can be determined with the help of the following formula (8).

[0167] Where S 12 Represents the similarity characterization data between the first grid and the second grid, S 12 ∈[0,1], and the S 12 The closer it is to 1, the more similar the two 3D grids are, and vice versa. represents the first average value above; α represents the weight corresponding to the first average value; represents the second average value above; 1-α represents the weight corresponding to the second average value. It should be noted that this application does not limit the implementation of α, for example, it can be implemented as 0.7.

[0168] Based on the relevant contents of steps 31 to 32 above, it can be seen that in some application scenarios, for the first grid and the second grid, each corresponding image pair can be found from the rendered image sets of the two grids; then, for any image pair, based on the similarity representation data between the image features of the two images in the image pair, the similarity representation data of the image pair is determined, so that the similarity representation data of the image pair can represent the degree of similarity between the two images in the image pair; then, based on the similarity representation data of all image pairs, the similarity representation data between the first grid and the second grid is determined. In particular, because the rendered image set is used to describe the state of a three-dimensional grid under multiple perspectives, when determining similarity based on the rendered image set, not only the geometric shape of the three-dimensional grid can be referenced, but also the visual information of the three-dimensional grid, such as edge information, color distribution information, texture information, etc., can be referenced, so that the similarity representation data finally determined can more comprehensively and accurately represent the similarity between different three-dimensional grids.

[0169] Based on the relevant contents of S1 to S2 above, it can be seen that for the similarity determination method provided in the embodiment of the present application, a rendering image set of a first grid and a rendering image set of a second grid are first obtained, so that the rendering image set of the first grid is used to describe the state of the first grid under multiple viewing angles, and the rendering image set of the second grid is used to describe the state of the second grid under the multiple viewing angles; then, similarity analysis processing is performed on the rendering image set of the first grid and the rendering image set of the second grid to obtain similarity representation data between the first grid and the second grid. Among them, because the above rendering image set is used to describe the state of a three-dimensional grid under multiple viewing angles, the rendering image set can not only represent the geometric characteristics of the three-dimensional grid, but also represent the visual characteristics of the three-dimensional grid, such as edge characteristics, texture characteristics, color distribution characteristics, etc., so that the rendering image set can represent the characteristics of the three-dimensional grid as comprehensively as possible, and thus the similarity representation data determined based on the rendering image set can more accurately represent the degree of similarity between different three-dimensional grids, which is conducive to improving accuracy.

[0170] Based on the relevant content of the similarity determination method above, it can be seen that the present application determines the similarity between different three-dimensional grids by means of similarity analysis processing between the rendered image sets of different three-dimensional grids, and this implementation scheme has the advantages shown in (1) to (4) below.

[0171] (1) Comprehensiveness and accuracy. Specifically, because the rendered image set is used to describe the state of a three-dimensional mesh under multiple viewing angles, when determining similarity based on the rendered image set, not only the geometric shape of the three-dimensional mesh can be referenced, but also the visual information of the three-dimensional mesh, such as edge information, color distribution information, texture information, etc., can be referenced. As a result, the ultimately determined similarity representation data can more comprehensively and accurately represent the similarity between different three-dimensional meshes, which is conducive to improving the comprehensiveness and accuracy of similarity determination.

[0172] (2) Efficient scalability. Specifically, because the number of images in a rendered image set is far less than the number of nodes, face normal vectors, and boundary curvatures in a three-dimensional mesh, the time overhead of a similarity determination scheme implemented based on the rendered image set is relatively low, thereby improving the efficiency of similarity determination. Furthermore, because the time overhead of the relevant processes of the rendered image set is not significantly correlated with the scale of the three-dimensional mesh, the time overhead of the similarity determination scheme implemented based on the rendered image set hardly changes significantly with the increase in the scale of the three-dimensional mesh. This makes the similarity determination scheme applicable to various scenarios, such as large-scale three-dimensional mesh scenarios, thereby improving the scalability of the similarity determination scheme.

[0173] (3) Robustness and stability, which can be specifically described as follows: because the rendered image set is used to describe the state of a three-dimensional mesh under multiple perspectives, the rendered image set can more accurately represent the overall state of the three-dimensional mesh, making it less susceptible to interference when similarity determination is performed based on the rendered image set, such as interference from noise data, interference from local deformation, etc., thereby making the ultimately determined similarity representation data stable and reliable, which is conducive to improving the robustness and stability of similarity determination.

[0174] (4) Interpretability, which can be specifically described as follows: because the similarity representation data between the first grid and the second grid is limited to the interval [0, 1], the similarity representation data can more accurately represent the degree of similarity between the two three-dimensional grids, which is conducive to improving interpretability.

[0175] In addition, for the similarity determination method provided in this application, this application does not limit the application scenarios of the similarity determination method. For example, the similarity determination method can be applied to three-dimensional grid comparison scenarios, three-dimensional grid retrieval scenarios, three-dimensional grid classification scenarios, and three-dimensional grid analysis scenarios, etc.

[0176] Based on the similarity determination method provided in the embodiment of the present application, the embodiment of the present application also provides a similarity determination device, which is explained and illustrated below in conjunction with Figure 3. Figure 3 is a schematic diagram of the structure of a similarity determination device provided in the embodiment of the present application. It should be noted that for the technical details of the similarity determination device provided in the embodiment of the present application, please refer to the relevant content of the similarity determination method above.

[0177] As shown in FIG3 , the similarity determination device 300 provided in an embodiment of the present application includes:

[0178] An acquisition unit 301 is configured to acquire a set of rendered images of a first grid and a set of rendered images of a second grid; the set of rendered images of the first grid is used to describe the state of the first grid under multiple viewing angles; the set of rendered images of the second grid is used to describe the state of the second grid under the multiple viewing angles;

[0179] The analyzing unit 302 is configured to perform similarity analysis on the rendered image set of the first grid and the rendered image set of the second grid to obtain similarity representation data between the first grid and the second grid.

[0180] In a possible implementation manner, the rendered image set includes a grid imaging image at each of the viewing angles and a grid edge image at each of the viewing angles.

[0181] In a possible implementation manner, there is a correspondence between the first image in the rendered image set of the first grid and the second image in the rendered image set of the second grid;

[0182] The analysis unit 302 is specifically used to: determine the similarity representation data between the first image and the second image based on the similarity representation data between the image features of the first image and the image features of the second image; determine the similarity representation data between the first grid and the second grid based on the similarity representation data between the first image and the second image.

[0183] In one possible implementation, the process of determining the similarity representation data between the first image and the second image includes: obtaining a reference image corresponding to the second image, where the reference image is obtained by performing noise processing on the second image; and determining the similarity representation data between the first image and the second image based on the similarity representation data between the image features of the first image and the image features of the second image, and the similarity representation data between the image features of the first image and the image features of the reference image.

[0184] In a possible implementation manner, the multiple viewing angles are determined based on at least two spherical positions on a minimum circumscribed sphere of a preset three-dimensional space;

[0185] The process of acquiring the rendered image set includes: restricting both the first grid and the second grid to the preset three-dimensional space; the center of the restricted first grid is located at the center position of the preset three-dimensional space, and the center of the restricted second grid is located at the center position of the preset three-dimensional space; rendering the restricted first grid at the at least two spherical positions to obtain a rendered image set of the first grid; rendering the restricted second grid at the at least two spherical positions to obtain a rendered image set of the second grid.

[0186] In a possible implementation manner, the minimum circumscribed cube of the first grid after the restriction is the preset three-dimensional space, and the minimum circumscribed cube of the second grid after the restriction is the preset three-dimensional space.

[0187] In a possible implementation manner, the preset three-dimensional space is a cube; the at least two spherical positions include spherical positions corresponding to each face in the preset three-dimensional space and spherical positions corresponding to each edge in the preset three-dimensional space.

[0188] Based on the relevant contents of the similarity determination device 300 described above, it can be seen that for the similarity determination device 300 provided in the embodiment of the present application, a rendering image set of a first grid and a rendering image set of a second grid are first obtained, so that the rendering image set of the first grid is used to describe the state of the first grid under multiple viewing angles, and the rendering image set of the second grid is used to describe the state of the second grid under the multiple viewing angles; then similarity analysis processing is performed on the rendering image set of the first grid and the rendering image set of the second grid to obtain similarity representation data between the first grid and the second grid. Among them, because the above rendering image set is used to describe the state of a three-dimensional grid under multiple viewing angles, the rendering image set can not only represent the geometric characteristics of the three-dimensional grid, but also represent the visual characteristics of the three-dimensional grid, such as edge characteristics, texture characteristics, color distribution characteristics, etc., so that the rendering image set can represent the characteristics of the three-dimensional grid as comprehensively as possible, and further, the similarity representation data determined based on the rendering image set can more accurately represent the degree of similarity between different three-dimensional grids, which is conducive to improving accuracy.

[0189] In addition, an embodiment of the present application also provides an electronic device, which includes a processor and a memory: the memory is used to store instructions or computer programs; the processor is used to execute the instructions or computer programs in the memory, so that the electronic device executes any implementation of the similarity determination method provided in the embodiment of the present application.

[0190] Referring to FIG4 , a schematic diagram of the structure of an electronic device 400 suitable for implementing embodiments of the present disclosure is shown. Terminal devices in embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. The electronic device shown in FIG4 is merely an example and should not limit the functionality or scope of use of embodiments of the present disclosure.

[0191] As shown in Figure 4, electronic device 400 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage device 408 into a random access memory (RAM) 403. Various programs and data required for the operation of electronic device 400 are also stored in RAM 403. Processing device 401, ROM 402, and RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to bus 404.

[0192] Typically, the following devices may be connected to the I / O interface 405: an input device 406 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 407 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 408 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 409. The communication device 409 may allow the electronic device 400 to communicate with other devices wirelessly or by wire to exchange data. Although FIG4 shows the electronic device 400 with various devices, it should be understood that not all of the devices shown are required to be implemented or present. More or fewer devices may alternatively be implemented or present.

[0193] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 409, or installed from the storage device 408, or installed from the ROM 402. When the computer program is executed by the processing device 401, the above-mentioned functions defined in the method of the embodiment of the present disclosure are performed.

[0194] The electronic device provided by the embodiment of the present disclosure and the method provided by the above embodiment belong to the same inventive concept. For technical details not fully described in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0195] An embodiment of the present application further provides a computer-readable medium, in which instructions or computer programs are stored. When the instructions or computer programs are executed on a device, the device executes any implementation of the similarity determination method provided in the embodiment of the present application.

[0196] It should be noted that the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0197] In some embodiments, the client and server can communicate using any currently known or later developed network protocol, such as HTTP (Hypertext Transfer Protocol), and can be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or later developed network.

[0198] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0199] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device can perform the method.

[0200] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including, but not limited to, object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0201] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0202] The units involved in the embodiments described in this disclosure may be implemented in software or hardware, wherein the name of a unit / module does not, in some cases, limit the unit itself.

[0203] The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0204] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0205] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems or devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0206] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0207] It should also be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0208] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0209] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A similarity determination method, wherein the method comprises: Obtain a rendering image set of a first grid and a rendering image set of a second grid; The rendered image set of the first grid is used to describe the state of the first grid at multiple viewing angles; The rendered image set of the second grid is used to describe the state of the second grid under the multiple viewing angles; A similarity analysis process is performed on the rendered image set of the first grid and the rendered image set of the second grid to obtain similarity representation data between the first grid and the second grid. 2 . The method according to claim 1 , wherein the rendered image set comprises a grid imaging image at each of the viewing angles and a grid edge image at each of the viewing angles.

3. The method according to claim 1, wherein there is a correspondence between the first image in the rendered image set of the first grid and the second image in the rendered image set of the second grid; The process of determining the similarity representation data between the first grid and the second grid includes: determining similarity representation data between the first image and the second image based on similarity representation data between image features of the first image and image features of the second image; Similarity representation data between the first grid and the second grid is determined based on the similarity representation data between the first image and the second image.

4. The method according to claim 3, wherein the process of determining the similarity characterization data between the first image and the second image comprises: Acquire a reference image corresponding to the second image, where the reference image is obtained by performing noise processing on the second image; Similarity characterization data between the first image and the second image is determined based on the similarity characterization data between the image features of the first image and the image features of the second image, and the similarity characterization data between the image features of the first image and the image features of the reference image.

5. The method according to claim 1, wherein the multiple viewing angles are determined based on at least two spherical positions on a minimum circumscribed sphere of a predetermined three-dimensional space; The process of obtaining the rendered image set includes: restricting both the first grid and the second grid to the preset three-dimensional space; wherein the center of the restricted first grid is located at the center of the preset three-dimensional space, and the center of the restricted second grid is located at the center of the preset three-dimensional space; performing rendering processing on the restricted first grid at the at least two spherical surface positions to obtain a rendering image set of the first grid; The restricted second grid is rendered at the at least two spherical positions to obtain a rendering image set of the second grid. 6 . The method according to claim 5 , wherein the minimum circumscribed cube of the first restricted grid is the preset three-dimensional space, and the minimum circumscribed cube of the second restricted grid is the preset three-dimensional space.

7. The method according to claim 5, wherein the preset three-dimensional space is a cube; The at least two spherical positions include spherical positions corresponding to each surface in the preset three-dimensional space and spherical positions corresponding to each edge in the preset three-dimensional space.

8. A similarity determination device, wherein the device comprises: An acquiring unit, configured to acquire a rendering image set of a first grid and a rendering image set of a second grid; The rendered image set of the first grid is used to describe the state of the first grid at multiple viewing angles; The rendered image set of the second grid is used to describe the state of the second grid under the multiple viewing angles; An analysis unit is configured to perform similarity analysis on the rendered image set of the first grid and the rendered image set of the second grid to obtain similarity representation data between the first grid and the second grid.

9. An electronic device, wherein the device comprises: processor and memory; The memory is used to store instructions or computer programs; The processor is configured to execute the instructions or computer program in the memory, so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable medium, wherein the computer-readable medium stores instructions or a computer program, and when the instructions or the computer program are executed on a device, the device is caused to execute the method according to any one of claims 1 to 7.

11. A computer program product, wherein the program product comprises a computer program carried on a non-transitory computer-readable medium, the computer program comprising program code for executing the method according to any one of claims 1 to 7.

Images