Three-dimensional scene generation method and device, equipment and storage medium

Abstract

The invention provides a three-dimensional scene generation method and device, equipment and a storage medium. The method comprises the steps of obtaining an original three-dimensional scene in response to a three-dimensional scene obtaining command; wherein the original three-dimensional scene comprises a plurality of fission elements; in response to the style label acquisition command, acquiring a style label; determining a plurality of fission options according to the fission elements and the style label; wherein the fission elements comprise a plurality of original three-dimensional models and original display effects; generating massive three-dimensional scene generation schemes according to the fission options and the generation rules; obtaining a three-dimensional scene pool according to the massive three-dimensional scene generation scheme and the original three-dimensional scene; wherein the three-dimensional scene pool comprises massive candidate three-dimensional scenes; and performing clustering according to the three-dimensional scene pool to obtain a target three-dimensional scene with style difference, and displaying the target three-dimensional scene. According to the scheme, the efficiency of generating the three-dimensional scene can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to technical fields such as three-dimensional scene modeling and three-dimensional scene fission. Background Art

[0002] The software production tools for 3D scene production are relatively mature. When using software production tools to produce 3D scenes, manual model design, lighting design and other work are usually required, which consumes a lot of manpower costs and has a long production cycle. Therefore, the efficiency of 3D scene generation is low. Summary of the invention

[0003] The present disclosure provides a three-dimensional scene fission method, device, equipment and storage medium to solve or alleviate one or more technical problems in the prior art.

[0004] In a first aspect, the present disclosure provides a three-dimensional scene generation method, comprising:

[0005] In response to a three-dimensional scene acquisition command, an original three-dimensional scene is acquired; wherein the original three-dimensional scene includes a plurality of fission elements;

[0006] In response to a style tag acquisition command, acquiring a style tag;

[0007] Determining a plurality of fission options according to the plurality of fission elements and the style tag; wherein the plurality of fission elements include a plurality of original three-dimensional models and original display effects;

[0008] Generate a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules;

[0009] A 3D scene pool is obtained according to the massive 3D scene generation scheme and the original 3D scene; wherein the 3D scene pool includes a massive 3D scene candidate;

[0010] Clustering is performed according to the three-dimensional scene pool to obtain a number of target three-dimensional scenes and display them; wherein the similarity between the several target three-dimensional scenes is lower than a threshold.

[0011] In a second aspect, the present disclosure provides a three-dimensional scene generation device, comprising:

[0012] A scene acquisition module, used to acquire an original three-dimensional scene in response to a three-dimensional scene acquisition command; wherein the original three-dimensional scene includes a plurality of fission elements;

[0013] A style acquisition module, used for acquiring a style tag in response to a style tag acquisition command;

[0014] A determination module, used to determine a plurality of fission options according to the plurality of fission elements and the style tag; wherein the plurality of fission elements include a plurality of original three-dimensional models and original display effects;

[0015] A solution generation module, used to generate a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules;

[0016] A scene generation module, used to obtain a 3D scene pool according to the massive 3D scene generation scheme and the original 3D scene; wherein the 3D scene pool includes a massive number of candidate 3D scenes;

[0017] The clustering module is used to cluster the three-dimensional scene pool to obtain and display a plurality of target three-dimensional scenes, wherein the similarity between the plurality of target three-dimensional scenes is lower than a threshold.

[0018] In a third aspect, an electronic device is provided, including:

[0019] at least one processor; and

[0020] a memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute any method in the embodiments of the present disclosure.

[0022] In a fourth aspect, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute any method according to the embodiments of the present disclosure.

[0023] In a fifth aspect, a computer program product is provided, comprising a computer program, which implements any method according to the embodiments of the present disclosure when executed by a processor.

[0024] The beneficial effects of the technical solution provided by the present disclosure include at least: based on the original three-dimensional scene, a target three-dimensional scene that meets the style label and matching constraints can be automatically generated through fission and displayed to the user, thereby improving the efficiency of generating three-dimensional scenes.

[0025] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the multiple drawings represent the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments provided according to the present disclosure and should not be regarded as limiting the scope of the present disclosure.

[0027] Figure 1 is a flow chart of a three-dimensional scene generation method according to an embodiment of the present disclosure;

[0028] Figure 2 is a flow chart of a three-dimensional scene generation method according to another embodiment of the present disclosure;

[0029] Figure 3 is a flow chart of a three-dimensional scene generation method according to another embodiment of the present disclosure;

[0030] Figure 4 is a flow chart of a three-dimensional scene generation method according to another embodiment of the present disclosure;

[0031] Figure 5 is a flowchart of a method for generating a three-dimensional scene according to an embodiment of the present disclosure;

[0032] Figure 6 is a structural schematic diagram of a three-dimensional scene generation device according to an embodiment of the present disclosure;

[0033] Figure 7 is a structural schematic diagram of a three-dimensional scene generating device according to another embodiment of the present disclosure;

[0034] Figure 8 is a block diagram of an electronic device for implementing an embodiment of the present disclosure. DETAILED DESCRIPTION

[0035] The present disclosure will be further described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0036] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present disclosure.

[0037] Figure 1 1 is a flow chart of a method for generating a three-dimensional scene according to an embodiment of the present disclosure. In one implementation, the method includes:

[0038] S101, in response to a 3D scene acquisition command, acquiring an original 3D scene; wherein the original 3D scene includes a plurality of fission elements;

[0039] S102, in response to a style tag acquisition command, acquiring a style tag;

[0040] S103, determining a plurality of fission options according to the plurality of fission elements and the style tag; wherein the plurality of fission elements include a plurality of original three-dimensional models and original display effects;

[0041] S104, generating a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules;

[0042] S105, obtaining a 3D scene pool according to the massive 3D scene generation scheme and the original 3D scene; wherein the 3D scene pool includes a massive 3D scene candidate;

[0043] S106, clustering is performed according to the three-dimensional scene pool to obtain a plurality of target three-dimensional scenes and displaying them; wherein the similarity between the plurality of target three-dimensional scenes is lower than a threshold value.

[0044] In the disclosed embodiments, a three-dimensional scene may include a virtual environment with three dimensions (length, width, and height) simulated and presented by computer technology. For example, an original three-dimensional scene may include a virtual environment of a three-dimensional apartment with one or more of furniture models, decoration models, lighting effects, and display filters. For example, an original three-dimensional scene may include a virtual environment of a restaurant with a chandelier model, four chair models, a dining table model, a set of wall models, a set of floor models, and warm-toned lighting effects.

[0045] In the embodiment of the present disclosure, when a user opens the fission interface of the design software, the design software can obtain the original three-dimensional scene input by the user or the default of the fission interface in response to the three-dimensional scene acquisition command. For example, when a user opens the fission interface, the design software can obtain the original three-dimensional scene in response to the default three-dimensional scene acquisition command. For another example, after the user selects a preset three-dimensional scene or inputs a custom three-dimensional scene in the fission interface, the design software can select or input the original three-dimensional scene in the fission interface in response to the three-dimensional scene acquisition command corresponding to the user's selection operation or input operation.

[0046] In the disclosed embodiment, the style tag may include keywords of the three-dimensional scene required by the user, such as simple, warm, serious, and energetic. The style tag can be extracted based on the demand description text input by the user. The user can enter the description text into the demand description text box of the fission interface of the design software, and click options such as "Confirm". In this case, the design software can respond to the style tag acquisition command to obtain the style tag required by the user. For example, after the user opens the fission interface in the design software, he can enter his text description of the decoration style he wants to generate by fission in the demand description text box. The text description can include a complete sentence or paragraph, or a single or multiple words. After the user confirms the input demand description text, the design software generates a style tag acquisition command. In response to the style tag acquisition command corresponding to the confirmation operation, the style tag in the demand description text is extracted. One or more style tags can be extracted from a section of demand description text.

[0047] In the disclosed embodiment, a 3D scene generation scheme for the original 3D scene can be generated according to the original 3D scene, the style label, and the set matching constraints. The matching constraints may include one or more items such as Cartesian products and matching rules. For example, the matching rules may include one or more items such as color matching rules, scene matching rules, and model matching rules. The 3D scene generation scheme may include one or more items such as 3D models replaced in the original 3D scene and various display effects. For example, the original 3D scene may include the original 3D model and the original display effect, and an optional 3D model of the same type as the original 3D model that meets the requirements of the style label can be obtained; and an optional display effect that meets the requirements of the style label can also be obtained. By combining the optional 3D model and the optional display effect according to the matching constraints, a large number of 3D scene generation schemes can be obtained.

[0048] In the disclosed embodiment, the 3D scene generation scheme can be used to guide the use of optional 3D models and optional display effects to update or replace the corresponding original 3D models and original display effects in the original 3D scene. After the original 3D scene is updated or replaced using the 3D scene generation scheme, a candidate 3D scene can be obtained.

[0049] In the disclosed embodiment, after obtaining the candidate three-dimensional scene corresponding to each three-dimensional scene generation scheme, all candidate three-dimensional scenes can be clustered to obtain clustering results. The candidate three-dimensional scene corresponding to all the cluster center points in the clustering results can be selected as the target three-dimensional scene. The target three-dimensional scene and / or the two-dimensional image of the target three-dimensional scene can be displayed to the user in the result display area of ​​the fission interface. During the clustering process, the candidate three-dimensional scene can be converted into corresponding points, and all points representing the candidate three-dimensional scene can be clustered to obtain a cluster consisting of points corresponding to several candidate three-dimensional scenes. The clustering results may include one or more clusters. The display effects and / or three-dimensional models of the corresponding candidate three-dimensional scenes of the points in the same cluster are similar. The cluster center point refers to the central position point of each cluster during the clustering process, and the center point can correspond to a candidate three-dimensional scene.

[0050] According to the disclosed embodiments, a target 3D scene that satisfies style tags and matching constraints can be automatically generated and displayed to the user through fission based on the original 3D scene, thereby improving the efficiency of generating 3D scenes. For example, multiple optional decoration plans can be provided to the user and previews can be provided, thereby improving the convenience of the user in decoration design and increasing the user's satisfaction and experience.

[0051] In one implementation, S101 acquires an original 3D scene in response to a 3D scene acquisition command, including at least one of the following:

[0052] In response to a 3D scene selection command, selecting the original 3D scene from a 3D scene library;

[0053] In response to a three-dimensional scene creation command, create the original three-dimensional scene;

[0054] The default 3D scene is used as the original 3D scene.

[0055] In the embodiment of the present disclosure, the 3D scene acquisition command may include a 3D scene selection command and a 3D scene creation command, etc. The fission interface of the design software may include a candidate scene area, a user-defined scene area, and a scene import area, etc.

[0056] In the embodiment of the present disclosure, a plurality of preset candidate 3D scenes or 2D images of candidate 3D scenes may be displayed in the candidate scene area. The user may select one of the candidate 3D scenes in the candidate scene area as the original 3D scene. The design software may obtain the original 3D scene selected by the user in response to the 3D scene selection command corresponding to the operation of the user selecting the scene. The original 3D scene or its 2D image may be displayed in the fission interface of the design software.

[0057] In the embodiment of the present disclosure, in the user-defined scene area, the user can create and / or modify the three-dimensional scene. For example, the design software can open the three-dimensional scene editing interface in response to the three-dimensional scene creation command corresponding to the user's three-dimensional scene creation operation. In the three-dimensional scene editing interface, the original three-dimensional model can be created and edited according to the commands corresponding to various operations (such as creating, adding, modifying, and deleting) in the process of the user creating the original three-dimensional scene.

[0058] In the disclosed embodiment, in the scene import area, the user can import an existing three-dimensional scene. When importing a three-dimensional scene, the user can directly import the three-dimensional scene or import a three-dimensional scene configuration file. In the case where the user directly imports the three-dimensional scene, the design software can respond to the three-dimensional scene creation command corresponding to the import three-dimensional scene operation and use the scene imported by the user as the original three-dimensional scene. In the case where the user imports a three-dimensional scene configuration file, the design software responds to the three-dimensional scene creation command corresponding to the import three-dimensional scene configuration file operation and generates an original three-dimensional scene according to the three-dimensional scene configuration file.

[0059] In the embodiment of the present disclosure, when the user does not select or input an original 3D scene, the design software can also use the software's default 3D scene as the original 3D scene. For example, the first candidate 3D scene in the candidate scene area can be used as the original 3D scene, or a candidate 3D scene can be randomly selected from the candidate scene area as the original 3D scene.

[0060] In one implementation, S102 obtains the style tag in response to the style tag obtaining command, including at least one of the following:

[0061] Responding to the input fission requirement description, the style tag is obtained;

[0062] Use the default tag as the style tag.

[0063] In the disclosed embodiment, the fission interface of the design software may include a user requirement input area. The user may input his or her requirements in the form of text in the user requirement input area to obtain a requirement description text. The design software may respond to the style tag acquisition command corresponding to the operation in which the user inputs the requirement description text and confirms it, and perform keyword matching in the style tag library according to the requirement description text to obtain a style tag. When performing keyword matching in the style tag library, fuzzy matching may be supported. In the case where the user does not enter a description text or the description text entered by the user is invalid, a default tag may be used as the style tag. For example, the most used style tag or multiple style tags with a large number of uses in the history records may be used as the default tag, or one or more style tags may be randomly selected from the style tag library as the default tag.

[0064] In one implementation, S103 determines a plurality of fission options according to the plurality of fission elements and the style tags, including at least one of the following:

[0065] Retrieving a model of the same category as the original three-dimensional model in the model library, and filtering the models of the same category according to the style tag to obtain the optional three-dimensional model;

[0066] The optional display effect is filtered in the display effect template library according to the style tag.

[0067] In the disclosed embodiment, the design software can obtain fission elements based on the original three-dimensional scene. The fission elements may include three-dimensional models and display effects that can be replaced in the original scene. The design software may use all three-dimensional models in the original three-dimensional scene as fission elements, or may select some three-dimensional models from the original three-dimensional scene as fission elements. For example, an original three-dimensional scene includes chair A, chair B, table C, chandelier D, and display effect I. In one extraction method, the extracted fission elements may include chair A, chair B, table C, chandelier D, and display effect I. In another extraction method, the extracted fission elements only include chair A, table C, and display effect I.

[0068] In the embodiment of the present disclosure, when the fission element is obtained, the design software can search the 3D model library according to the type of 3D model in the original scene to obtain the 3D model search result. The 3D model search result can include one or more categories, and each category can include one or more 3D models. When the 3D model search result is obtained, the design software can filter the 3D model search result according to the style label, and retain the 3D model that meets the style label as the fission option.

[0069] In the disclosed embodiment, when the fission elements and style labels are obtained, the design software can perform a joint search in the 3D model library according to the type and style label of the 3D model in the original scene, and obtain a 3D model of the same type as the fission element and that meets the style label as a fission option.

[0070] In the disclosed embodiment, the design software can select display effects that meet the style tags in the display effect template library as fission options based on the style tags.

[0071] In the embodiment of the present disclosure, when the fission option is obtained, the design software can obtain a combination template according to the combination mode of the fission elements in the original three-dimensional scene. The design software can combine the three-dimensional model and the display effect in the fission option according to the matching constraint according to the combination template to obtain a three-dimensional scene generation scheme. For example, the design software can calculate the Cartesian product of the set of various three-dimensional models and the set of display effects in the fission option to obtain a three-dimensional scene generation scheme. For another example, the design software can also combine the set of various three-dimensional models and the set of display effects in the fission option according to at least one of the color matching rule, the scene matching rule, the model matching rule, etc. to obtain a three-dimensional scene generation scheme. For another example, the design software can also filter the calculation results according to at least one of the color matching rule, the scene matching rule, the model matching rule, etc. after calculating the Cartesian product calculation result of the set of various three-dimensional models and the set of display effects in the fission option to obtain a three-dimensional scene generation scheme.

[0072] In the disclosed embodiment, the design software can first search in the model library according to the respective types of the three-dimensional models in the fission elements in the original three-dimensional scene, and obtain other models under the type. In the case of obtaining other models, the design software can filter the search results according to the style tag and retain the three-dimensional models that meet the requirements of the style tag. After the design software searches and filters all types of three-dimensional models in the fission element according to the above operations, all optional three-dimensional models can be obtained. For example, the design software can first determine that the fission element includes 4 original three-dimensional models according to the original three-dimensional scene, namely model A, model B, model C and model D, wherein the types of model A and model B are both type 1, the type of model C is type 2, and the type of model D is type 3. According to type 1, type 2 and type 3, the model library is searched, and models A1, A2, A3, B1, B2 and B3 under type 1, models C1, C2 and C3 under type 2, and models D1, D2 and D3 under type 3 can be obtained. Filtering is performed according to the style label n to obtain optional three-dimensional models A1, A2, B2, B3, C1, C3, D1, D2 and D3 that meet the style label n.

[0073] In the disclosed embodiment, the design software can search the display effect template library according to the style tag to obtain the optional display effect that meets the requirements of the style tag. For example, the design software can search the display effect template library according to the style tag n to obtain optional display effect I, optional display effect II and optional display effect III.

[0074] In one implementation, S104 generates a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules, including:

[0075] According to the optional three-dimensional model and the optional display effect in the fission options, the Cartesian product combination result is calculated, and the Cartesian product combination result is screened according to the model matching rule to obtain the three-dimensional scene generation solution.

[0076] In the disclosed embodiment, the optional 3D models corresponding to the same original 3D model in the original 3D scene may be taken as a set. For example, the optional 3D models A1 and A2 corresponding to the original model A may be taken as a set, and the optional 3D models B2 and B3 corresponding to the original model B may be taken as a set. The optional display effects may also be taken as a set.

[0077] In the disclosed embodiment, the design software can calculate the Cartesian product for each optional model set and the optional display effect set, and obtain each possible combination as the Cartesian product combination result. For example, the Cartesian product is calculated for the optional three-dimensional model set {A1, A2}, {B2, B3} and the optional display effect {effect I, effect II}, and the combination of {A1, B2, effect I}, {A1, B2, effect II}, {A1, B3, effect I}, {A1, B3, effect II}, {A2, B2, effect I}, {A2, B2, effect II}, {A2, B3, effect I} and {A2, B3, effect II} can be obtained. The design software can use one or more of the above combinations as the Cartesian product combination result.

[0078] In the disclosed embodiment, the design software can also filter the Cartesian product combination results obtained by the above-mentioned Cartesian product calculation according to preset matching rules such as color matching rules, scene matching rules and model matching rules to obtain a three-dimensional scene generation scheme. For example, if the constraint includes "A1 and B3 cannot be combined", then the screening results are {A1, B2, effect I}, {A1, B2, effect II}, {A2, B2, effect I}, {A2, B2, effect II}, {A2, B3, effect I} and {A2, B3, effect II}. The design software can use the above combination as a three-dimensional scene generation scheme.

[0079] Figure 2 is a flow chart of a 3D scene generation method according to another embodiment of the present disclosure. The method may include one or more features of the above 3D scene generation method. In one embodiment, in one embodiment, S105 obtains a 3D scene pool according to the massive 3D scene generation scheme and the original 3D scene, including at least one of the following:

[0080] S201, using the optional 3D model in the 3D scene generation solution to replace the original 3D model in the original 3D scene;

[0081] S202: Use the optional display effects in the three-dimensional scene generation solution to replace the original display effects in the original three-dimensional scene.

[0082] In the disclosed embodiment, after obtaining the 3D scene generation scheme, the design software can use the optional 3D model in the 3D scene generation scheme to replace the original 3D model in the original 3D scene. The attributes of the 3D model may include size and position. When replacing, the design software can set the size of the optional 3D model to the size of the original 3D model, set the optional 3D model at the position of the original 3D model, and delete the original 3D model. For example, when the design software uses the 3D scene generation scheme {A1, B2, effect I} to replace the model of the original scene, the size of the original 3D model A is obtained as {length: 0.45, width: 0.45, height: 1.2}, and the center point position of model A is (0.7, 0.8, 0.6). The optional 3D model corresponding to the original 3D model A is A1. Then, the size of the optional 3D model A1 can be set to {length: 0.45, width: 0.45, height: 1.2}, and the center point of the optional 3D model A1 can be set at the (0.7, 0.8, 0.6) position of the original 3D scene. Then delete the original 3D model A from the original 3D scene to complete the replacement.

[0083] In the disclosed embodiment, after obtaining the three-dimensional scene generation scheme, the design software may also use the optional display effects in the three-dimensional scene generation scheme to replace the original display effects in the original three-dimensional scene. Display effects may include lighting effects and filters, etc. For example, the lighting effects may include one or more of the number, position, color, etc. of light sources. When replacing, the design software may set the number and position of light sources in the lighting effects in the optional display effects to the number and position of light sources in the original lighting effects; the original filters may be replaced with optional filters. For example, when the design software uses the three-dimensional scene generation scheme {A1, B2, effect I} to replace the display effects of the original scene, the number of light sources of the original lighting effect obtained is 3, and the positions of the light sources are (1.3, 1.3, 2.7), (2.5, 2.5, 2.7) and (1.3, 2.5, 2.7) respectively. The color of the lighting effect in the optional display effect I is light yellow. The number of light sources in the optional display effect can be set to 3, and the positions of the 3 light sources can be set to (1.3, 1.3, 2.7), (2.5, 2.5, 2.7) and (1.3, 2.5, 2.7) respectively. Use the light yellow light effect with the set number and position of light sources to replace the light effect in the original 3D scene. Then use the optional filter "Noon Sunshine" in the optional display effect to replace the original filter in the original display effect.

[0084] In the disclosed embodiment, an optional three-dimensional model may be used to replace part or all of the original three-dimensional models in the original three-dimensional scene, or an optional display effect may be used to replace part or all of the original display effects in the original three-dimensional scene to obtain a candidate three-dimensional scene. For example, when the design software uses the three-dimensional scene generation scheme {A1, B2, effect I} for replacement, only the optional three-dimensional models A1 and B2 may be used to replace the original three-dimensional models A and B of the original three-dimensional scene to obtain a candidate three-dimensional scene. It is also possible to only use the optional display effect I to replace the original display effect of the original three-dimensional scene to obtain a candidate three-dimensional scene. It is also possible to use the optional three-dimensional model A1 to replace the original three-dimensional model A of the original three-dimensional scene, and use the optional display effect I to replace the original display effect of the original three-dimensional scene to obtain a candidate three-dimensional scene.

[0085] In one embodiment, if Figure 2 As shown, S105 obtains a 3D scene pool according to the massive 3D scene generation scheme and the original 3D scene, and also includes:

[0086] S203: Correct the position and / or adsorption relationship of the optional three-dimensional model according to the bounding box of the optional three-dimensional model.

[0087] In the disclosed embodiment, after the design software replaces the original three-dimensional model with an optional three-dimensional model, the optional three-dimensional model replaced into the three-dimensional scene may have unreasonable situations such as being through the mold or hanging in the air. This is because different models have different center point positions when the model size is consistent. In the event of a through-the-mold or hanging in the air, the design software can correct the position of the optional three-dimensional model based on the bounding box of the original three-dimensional model, the position of the center point of the original three-dimensional model, the bounding box of the optional three-dimensional model, and the position of the center point of the optional three-dimensional model. The correction method may include displacing the optional three-dimensional model. For example, taking the example of a model penetrating the mold on the ground, set the ground plane to correspond to the x0y plane in the three-dimensional coordinate system. In the case of a model penetrating the mold on the ground, the design software can perform a displacement correction on the center of the model in the z-axis direction. The displacement amount and direction can be calculated according to the following formula:

[0088] Δz=-(a`.center().z()-a.center().z())+v(a`.height() / 2-a.height() / 2)

[0089] Among them, the value of Δz can represent the amount of displacement, the sign of Δz can represent the direction of displacement, "+" can represent displacement along the positive direction of the z-axis, and "-" can represent displacement along the negative direction of the z-axis. a represents the original 3D model, and a` can represent the optional 3D model. a.center().z() can represent the value of the center point of the original 3D model on the z-axis, and a`.center().z() can represent the value of the center point of the optional 3D model on the z-axis. v can represent the relative position of the optional 3D model to the original 3D model on the z-axis, which can be determined by the relative distance between the center point of the original 3D model and the center point of the optional 3D model. a.height() can represent the height of the original 3D model, and a`.height() can represent the height of the optional 3D model.

[0090] In the disclosed embodiment, after the design software replaces the original 3D model with the optional 3D model, the optional 3D model replaced into the 3D scene may conflict with other optional 3D models. In the case of conflicts such as penetration between 3D models, the design software can correct the position of the optional 3D model based on the bounding box of the original 3D model, the position of the center point of the original 3D model, the bounding box of the optional 3D model, and the position of the center point of the optional 3D model. The correction method can refer to the above example.

[0091] According to the embodiments of the present disclosure, after the original three-dimensional model is replaced by the optional three-dimensional model, the position of the optional three-dimensional model can be further adjusted so that the optional three-dimensional model is in a reasonable position in the candidate three-dimensional scene, conforming to the real physical laws and design aesthetic requirements.

[0092] In one embodiment, a 3D scene pool is obtained based on the massive 3D scene generation scheme and the original 3D scene, and further includes: obtaining the 3D scene pool based on the candidate 3D scene with the correct position and / or adsorption relationship of the optional 3D model.

[0093] Figure 3 1 is a flow chart of a method for generating a three-dimensional scene according to another embodiment of the present disclosure. The method may include one or more features of the above three-dimensional scene generation method. In one embodiment, S106 performs clustering based on the three-dimensional scene pool to obtain a plurality of target three-dimensional scenes and displays them, including:

[0094] S301, respectively acquiring a plurality of two-dimensional full-angle images of the candidate three-dimensional scenes to obtain a plurality of fissioned two-dimensional full-angle images;

[0095] S302, extracting respective image features from the multiple candidate two-dimensional full-angle images to obtain multiple candidate image features;

[0096] S303, clustering the multiple candidate image features to obtain a center point of a first target class;

[0097] S304: Display the target three-dimensional scene obtained according to the center point of the first target class family.

[0098] In the disclosed embodiment, the fission two-dimensional full-angle image is obtained by rendering the design software according to the candidate three-dimensional scene at an angle that can display the most three-dimensional models in the candidate three-dimensional scene. After determining the rendering angle of the fission two-dimensional full-angle image of a candidate three-dimensional scene, the remaining candidate three-dimensional scenes can be rendered according to the above rendering angle to obtain their respective fission two-dimensional full-angle images. For example, for all candidate three-dimensional scenes of an original three-dimensional scene, the angle of the three-dimensional model that can be displayed by one of the candidate three-dimensional scenes includes the front view angle. The design software can use the front view angle as the rendering angle, and render all candidate three-dimensional scenes to obtain the corresponding fission two-dimensional full-angle images.

[0099] In the disclosed embodiment, after determining the fissioned two-dimensional full-angle images of all candidate three-dimensional scenes, the design software can perform dimensionality reduction processing on the fissioned two-dimensional full-angle images through an embedding algorithm to obtain fissioned image features.

[0100] In the embodiment of the present disclosure, after obtaining the fission image features of the fission two-dimensional full-angle image, the design software can input all the fission image features into the clustering algorithm for clustering to obtain the clustering results. The clustering algorithm may include a K-means clustering algorithm, a mean shift clustering algorithm, a density-based spatial clustering of applications with noise (DBSCAN) algorithm, etc., which is not limited in the present disclosure. After clustering the fission image features, the design software can obtain a clustering result containing multiple clusters. In the clustering results, the cluster center point can represent the candidate three-dimensional scene that best represents the cluster, and the three-dimensional model and display effect of the candidate three-dimensional scene corresponding to each cluster center point are quite different. The design software can use the candidate three-dimensional scene corresponding to all cluster center points as the target three-dimensional scene. For example, after the design software clusters the fission two-dimensional full-angle image corresponding to the candidate three-dimensional scene of an original three-dimensional scene, three clusters are obtained. The center points of the three clusters are cluster center point M, cluster center point N and cluster center point O respectively. The design software can obtain the candidate three-dimensional scene 12, the candidate three-dimensional scene 17 and the candidate three-dimensional scene 20 corresponding to the cluster center point M, the cluster center point N and the cluster center point O respectively as the target three-dimensional scene.

[0101] In the disclosed embodiment, the design software can display the target three-dimensional scene in the result display area of ​​the fission interface. The design software can directly display the candidate three-dimensional scene, or it can display the fission two-dimensional full-angle image corresponding to the target three-dimensional scene.

[0102] Figure 4 1 is a flow chart of a method for generating a three-dimensional scene according to another embodiment of the present disclosure. The method may include one or more features of the above three-dimensional scene generation method. In one embodiment, S106 performs clustering based on the three-dimensional scene pool to obtain a plurality of target three-dimensional scenes and displays them, and further includes:

[0103] S401, respectively acquiring the original 3D scene and the 2D full-angle images of the plurality of candidate 3D scenes to obtain the original 2D full-angle image and the plurality of candidate 2D full-angle images;

[0104] S402, extracting image features from the original two-dimensional full-angle image and the multiple candidate two-dimensional full-angle images respectively, to obtain original image features and multiple candidate image features;

[0105] S403, clustering the original image feature and the plurality of candidate image features to obtain cluster center points;

[0106] S404, according to the similarity distance between the cluster center point and the original image point, the second target cluster center point is obtained by screening according to the set interval;

[0107] S405: Display the target three-dimensional scene obtained according to the center point of the second target class family.

[0108] In the disclosed embodiment, the design software can also render the original three-dimensional scene into an original two-dimensional full-angle image according to the rendering angle of the candidate three-dimensional scene. The design software can also use the embedding algorithm to perform dimensionality reduction processing on the original two-dimensional full-angle image to obtain the original image features. The design software can input the original image features and the fission image features into the clustering algorithm for clustering to obtain the clustering results. The clustering results may include the points of the original image features. The design software can screen the class family center points whose distances from the original image features are greater than a set threshold according to the distances between the center points of each class family and the points of the original image features to obtain the target class family center points. The design software can send the candidate three-dimensional scene corresponding to each target class family center point to the fission result interface for display. For example, after the design software clusters the fission two-dimensional full-angle image and the original two-dimensional full-angle image corresponding to the original three-dimensional scene, three class families are obtained. The center points of the three class families are class family center point M, class family center point N and class family center point O, respectively. The design software obtains the distances of the cluster center point M, the cluster center point N and the cluster center point O from the point Q corresponding to the original three-dimensional scene, which are MQ=1.2, NQ=1.7 and OQ=2.3 respectively. If the threshold is 1.5, the target cluster center points can be screened according to the threshold as the cluster center point N and the cluster center point O, and the corresponding target candidate three-dimensional scenes are candidate three-dimensional scene 17 and candidate three-dimensional scene 20. This embodiment can make the candidate three-dimensional scene displayed to the user have a large style difference from the original three-dimensional scene.

[0109] Figure 5 is a flow chart of a method for generating a three-dimensional scene according to an embodiment of the present disclosure, such as Figure 5 As shown, the method may include the following steps:

[0110] S501. The user selects the original 3D scene at the foreground: In the tool, the three-dimensional (3D) scene provided by the platform can be used, or a 3D scene can be built by the user as the original scene for fission. The newly generated scenes for subsequent fission will be based on this scene as a template.

[0111] S502, the user describes the overall style of the desired scene on the front end: the overall style of the desired scene can be described in text form, and the platform automatically extracts style tags. If no valid extraction is done, the platform default configuration is used.

[0112] S503, generate a 3D scene generation plan: determine the scene fission elements and fission options according to the style tags obtained in S502. Specifically include:

[0113] Determine the scene fission elements: The platform supports models and overall effects to participate in fission. Fission elements can include fission models and fission effect templates. Fission elements can be determined based on the original 3D scene. Models in 3D scenes have their own classification labels (category, style, color, etc.). Determine the model elements suitable for fission according to the rules configured by the platform. The overall effect template refers to the lighting and effect parameters in the 3D scene, which determines the overall visual effect and style.

[0114] Determine fission options: Fission options are used to replace the fission elements determined by S502. Fission options need to match the style tags obtained by S502. The platform provides a large model library, and selects suitable optional models according to the category tags of the fission model. The platform provides a variety of overall effect templates, such as "Caramel", "Black and White", "Morning Sunshine", "Noon Sunshine", etc.

[0115] Generate 3D scene generation plan: Cartesian product combination of the options of fission elements, each combination is a 3D scene generation plan. The 3D scene generation plan will be filtered through the combination rules configured by the platform, such as avoiding conflicts in model color and style.

[0116] S504, generating fission scenes: obtaining a three-dimensional scene generation scheme through S503, and generating 3D scenes on a large scale. Specifically, it may include:

[0117] Perform model replacement, keep the real size of the replaced model, and use the original model's position information.

[0118] Perform effect template replacement to replace all lighting and effect configurations.

[0119] S505, correction of spatial relationship of fission scene: The 3D scene obtained by S504 may have unreasonable spatial relationship of models due to model replacement, and there may be phenomena such as model penetration and hanging, which need to be automatically corrected. Specifically, it may include:

[0120] The collision and attachment relationship between objects is obtained through the object BBox, and the collision is avoided and the attachment relationship is maintained through displacement. Let a be the original model and a` be the replacement model. The appropriate displacement of a` needs to be calculated. The displacement includes the x, y, and z axes. Taking the x-axis direction as an example, the following operations can be performed:

[0121] Let v be the position of a` at a in the x-axis direction, -1 in the positive direction and 1 in the negative direction. Then the required x-axis correction displacement can be calculated according to the following formula:

[0122] Δx=-(a`.center().x()-a.center().x())+v(a`.width() / 2-a.width() / 2)

[0123] The value of Δx can represent the amount of displacement, the sign of Δx can represent the direction of displacement, "+" can represent displacement along the positive direction of the x-axis, and "-" can represent displacement along the negative direction of the x-axis. a.center().x() can represent the value of the center point of the original 3D model on the x-axis, and a`.center().x() can represent the value of the center point of the optional 3D model on the x-axis. a.width t() can represent the width of the original 3D model, and a`.width() can represent the width of the optional 3D model.

[0124] S506, fission scene difference screening: The spatial relationship of the 3D scene obtained through S505 is relatively reasonable, but because the fission elements are combined through Cartesian products to obtain a three-dimensional scene generation scheme, there are many similar redundant scenes in the three-dimensional scene generation scheme, which requires further difference filtering. Specific filtering operations may include:

[0125] Make a full-view rendering of the fissioned 3D scene to obtain a rendering image.

[0126] The rendered image is embedded and reduced in dimension to obtain the image feature vector.

[0127] All image feature vectors are clustered, and the cluster center of the cluster is the effective 3D scene rendering.

[0128] Filtering renderings means filtering out valid 3D scenes.

[0129] S507, the user selects the scene that meets the expectation at the front desk: through S506, the platform automatically selects the 3D scene, and the user can make further manual selection on the platform.

[0130] Figure 6 1 is a schematic diagram of a structure of a three-dimensional scene generation device according to an embodiment of the present disclosure. In one implementation, the method includes:

[0131] The scene acquisition module 601 is used to acquire an original three-dimensional scene in response to a three-dimensional scene acquisition command; wherein the original three-dimensional scene includes a plurality of fission elements;

[0132] The style acquisition module 602 is used to acquire the style tag in response to the style tag acquisition command;

[0133] A determination module 603 is used to determine a plurality of fission options according to the plurality of fission elements and the style tags; wherein the plurality of fission elements include a plurality of original three-dimensional models and original display effects;

[0134] A solution generation module 604, used to generate a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules;

[0135] A scene generation module 605 is used to obtain a 3D scene pool according to the massive 3D scene generation schemes and the original 3D scenes; wherein the 3D scene pool includes a massive number of candidate 3D scenes;

[0136] The clustering module 606 is used to perform clustering according to the three-dimensional scene pool to obtain and display a plurality of target three-dimensional scenes; wherein the similarity between the plurality of target three-dimensional scenes is lower than a threshold.

[0137] In one implementation, the scene acquisition module 601 is further used for at least one of the following:

[0138] In response to a 3D scene selection command, selecting the original 3D scene from a 3D scene library;

[0139] In response to a three-dimensional scene creation command, create the original three-dimensional scene;

[0140] The default 3D scene is used as the original 3D scene.

[0141] In one implementation, the style acquisition module 602 is further used for at least one of the following:

[0142] Responding to the input fission requirement description, the style tag is obtained;

[0143] Use the default tag as the style tag.

[0144] In one embodiment, the fission module 603 is further used for at least one of the following:

[0145] Retrieving a model of the same category as the original three-dimensional model in the model library, and filtering the models of the same category according to the style tag to obtain the optional three-dimensional model;

[0146] The optional display effect is filtered in the display effect template library according to the style tag.

[0147] In one implementation, the solution generation module 604 is further used to perform matching according to the matching constraints based on the optional three-dimensional model and the optional display effect in the fission options to obtain the three-dimensional scene generation solution.

[0148] In one implementation, the scene generation module 605 is further used for at least one of the following:

[0149] Using the optional 3D model in the 3D scene generation solution to replace the original 3D model in the original 3D scene;

[0150] The optional display effect in the three-dimensional scene generation solution is used to replace the original display effect in the original three-dimensional scene.

[0151] Figure 7 6 is a schematic diagram of a structure of a three-dimensional scene generation device according to another embodiment of the present disclosure. The device may include one or more features of the above three-dimensional scene generation device. In one implementation, the scene generation module 605 further includes:

[0152] The correction submodule 701 is used to correct the position and / or adsorption relationship of the optional three-dimensional model according to the bounding box of the optional three-dimensional model.

[0153] In one implementation, the scene generation module 605 is further configured to obtain the three-dimensional scene pool according to the candidate three-dimensional scene with the correct position and / or adsorption relationship of the optional three-dimensional model.

[0154] In one embodiment, Figure 7 As shown, the clustering module 606 includes:

[0155] The two-dimensional image acquisition submodule 702 is used to respectively acquire the two-dimensional full-angle images of the plurality of candidate three-dimensional scenes to obtain a plurality of candidate two-dimensional full-angle images;

[0156] The feature extraction submodule 703 is used to extract respective image features from the multiple candidate two-dimensional full-angle images to obtain multiple candidate image features;

[0157] A clustering submodule 704 is used to cluster the multiple candidate image features to obtain a first target cluster center point;

[0158] The display submodule 705 is used to display the target three-dimensional scene obtained according to the center point of the first target class family.

[0159] In one embodiment, the 2D image acquisition submodule 702 is further used to respectively acquire the 2D full-angle images of the original 3D scene and the plurality of candidate 3D scenes, to obtain the original 2D full-angle image and the plurality of candidate 2D full-angle images;

[0160] The feature extraction submodule 703 is further used to extract respective image features from the original two-dimensional full-angle image and the multiple candidate two-dimensional full-angle images to obtain original image features and multiple candidate image features;

[0161] The clustering submodule 704 is further used to cluster the original image features and the plurality of candidate image features to obtain cluster center points;

[0162] like Figure 7 As shown, the clustering module 606 further includes:

[0163] A screening submodule 706 is used to screen and obtain a second target cluster center point according to the set interval based on the similarity distance between the cluster center point and the original image point;

[0164] The display submodule 705 is further used to display the target three-dimensional scene obtained according to the center point of the second target class family.

[0165] For the description of specific functions and examples of each module and submodule of the device in the embodiment of the present disclosure, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0166] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0167] Figure 8 FIG. 1 is a structural block diagram of an electronic device according to an embodiment of the present disclosure. Figure 8 As shown, the electronic device includes: a memory 810 and a processor 820, and the memory 810 stores a computer program that can be run on the processor 820. The number of the memory 810 and the processor 820 can be one or more. The memory 810 can store one or more computer programs. When the one or more computer programs are executed by the electronic device, the electronic device executes the method provided by the above method embodiment. The electronic device may also include: a communication interface 830, which is used to communicate with external devices and perform data exchange transmission.

[0168] If the memory 810, the processor 820 and the communication interface 830 are implemented independently, the memory 810, the processor 820 and the communication interface 830 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 8 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0169] Optionally, in a specific implementation, if the memory 810, the processor 820 and the communication interface 830 are integrated on a chip, the memory 810, the processor 820 and the communication interface 830 can communicate with each other through an internal interface.

[0170] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the Advanced RISC Machines (ARM) architecture.

[0171] Further, optionally, the above-mentioned memory may include a read-only memory and a random access memory, and may also include a non-volatile random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (DR RAM).

[0172] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the process or function described in the embodiment of the present disclosure is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (for example: coaxial cable, optical fiber, data subscriber line (Digital Subscriber Line, DSL)) or wireless (for example: infrared, Bluetooth, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access, or a data storage device such as a server or data center that includes one or more available media integrations. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a digital versatile disc (DVD)), or a semiconductor medium (e.g., a solid state disk (SSD)), etc. It is worth noting that the computer-readable storage medium mentioned in the present disclosure may be a non-volatile storage medium, in other words, a non-transient storage medium.

[0173] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0174] In the description of the embodiments of the present disclosure, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0175] In the description of the embodiments of the present disclosure, unless otherwise specified, " / " means or, for example, A / B can mean A or B. "And / or" in this article is only a way to describe the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0176] In the description of the embodiments of the present disclosure, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, unless otherwise specified, "plurality" means two or more.

[0177] The above description is only an exemplary embodiment of the present disclosure and is not intended to limit the present disclosure. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A three-dimensional scene generation method, comprising: In response to a three-dimensional scene acquisition command, an original three-dimensional scene is acquired; wherein the original three-dimensional scene includes a plurality of fission elements; In response to a style tag acquisition command, acquiring a style tag; Determining a plurality of fission options according to the plurality of fission elements and the style tags; wherein the plurality of fission elements include a plurality of original three-dimensional models and original display effects; Generate a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules; According to the massive 3D scene generation scheme and the original 3D scene, a 3D scene pool is obtained; wherein the 3D scene pool includes a massive 3D scene candidate; Clustering is performed according to the three-dimensional scene pool to obtain a number of target three-dimensional scenes and display them; wherein the similarity between the several target three-dimensional scenes is lower than a threshold.

2. The method according to claim 1, wherein: In response to the 3D scene acquisition command, an original 3D scene is acquired, including at least one of the following: In response to a 3D scene selection command, selecting the original 3D scene from a 3D scene library; In response to a three-dimensional scene creation command, create the original three-dimensional scene; The default 3D scene is used as the original 3D scene.

3. The method according to claim 1, wherein: In response to the style tag acquisition command, the style tag is acquired, including at least one of the following: Acquire the style tag in response to the input fission requirement description; Sets the default tag as the style tag.

4. The method according to any one of claims 1 to 3, wherein: Determining a plurality of fission options according to the plurality of fission elements and the style tags includes at least one of the following: Retrieving models of the same category as the original three-dimensional model in a model library, and selecting optional three-dimensional models from the models of the same category according to the style label; Selectable display effects are screened in a display effect template library according to the style tag.

5. The method according to any one of claims 1 to 4, wherein: According to the multiple fission options and generation rules, a large number of three-dimensional scene generation solutions are generated, including: According to the optional three-dimensional model and the optional display effect in the fission options, the Cartesian product combination result is calculated, and the Cartesian product combination result is screened according to the model matching rule to obtain the three-dimensional scene generation solution.

6. The method according to any one of claims 1 to 5, wherein: According to the massive 3D scene generation scheme and the original 3D scene, a 3D scene pool is obtained, including at least one of the following: Using the optional 3D model in the 3D scene generation solution to replace the original 3D model in the original 3D scene; The optional display effects in the three-dimensional scene generation solution are used to replace the original display effects in the original three-dimensional scene.

7. The method according to any one of claims 1 to 6, wherein: According to the massive 3D scene generation scheme and the original 3D scene, a 3D scene pool is obtained, which also includes: According to the bounding box of the optional three-dimensional model, the position and / or adsorption relationship of the optional three-dimensional model is corrected.

8. The method according to any one of claims 1 to 7, wherein: According to the massive 3D scene generation scheme and the original 3D scene, a 3D scene pool is obtained, which also includes: The three-dimensional scene pool is obtained according to the candidate three-dimensional scenes with correct positions and / or adsorption relationships of the optional three-dimensional models.

9. The method according to any one of claims 1 to 8, wherein: Clustering is performed according to the three-dimensional scene pool to obtain and display a number of target three-dimensional scenes, including: Respectively acquiring two-dimensional full-angle images of each of the plurality of candidate three-dimensional scenes to obtain a plurality of candidate two-dimensional full-angle images; Extracting respective image features from the plurality of candidate two-dimensional full-angle images to obtain a plurality of candidate image features; Clustering the multiple candidate image features to obtain a first target class center point; The target three-dimensional scene obtained according to the center point of the first target class family is displayed.

10. The method according to any one of claims 1 to 8, wherein: Clustering is performed according to the three-dimensional scene pool to obtain a number of target three-dimensional scenes and display them, and further comprising: Respectively acquiring the original three-dimensional scene and the two-dimensional full-angle images of the plurality of candidate three-dimensional scenes to obtain the original two-dimensional full-angle image and a plurality of candidate two-dimensional full-angle images; Extracting respective image features from the original two-dimensional full-angle image and the plurality of candidate two-dimensional full-angle images to obtain original image features and a plurality of candidate image features; Clustering the original image features and the plurality of candidate image features to obtain cluster center points; According to the similarity distance between the cluster center point and the original image point, a second target cluster center point is obtained by screening according to a set interval; The target three-dimensional scene obtained according to the center point of the second target class family is displayed.

11. A three-dimensional scene generation device, comprising: A scene acquisition module, configured to acquire an original three-dimensional scene in response to a three-dimensional scene acquisition command; wherein the original three-dimensional scene includes a plurality of fission elements; A style acquisition module, used for acquiring a style tag in response to a style tag acquisition command; A determination module, used to determine a plurality of fission options according to the plurality of fission elements and the style tags; wherein the plurality of fission elements include a plurality of original three-dimensional models and original display effects; A solution generation module, used to generate a large number of three-dimensional scene generation solutions according to the multiple fission options and generation rules; A scene generation module, used to obtain a 3D scene pool according to the massive 3D scene generation schemes and the original 3D scenes; wherein the 3D scene pool includes a massive number of candidate 3D scenes; A clustering module is used to cluster the three-dimensional scene pool to obtain and display a plurality of target three-dimensional scenes; wherein the similarity between the plurality of target three-dimensional scenes is lower than a threshold.

12. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

13. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-10.

14. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 10.

Images