An indoor effect picture management method and device, terminal equipment and storage medium

Abstract

The application relates to the technical field of image processing, and discloses an indoor effect picture management method and device, a terminal equipment and a storage medium, which are used for generating an indoor effect picture with an unchanged house type structure and a controllable style migration effect through a style migration model. The method comprises the following steps: acquiring house type data and style information of a target house type; generating a space layout screenshot representing an actual house type structure through a browser automation program based on the house type data; constructing a prompt word containing structure keeping constraints and style migration constraints based on the space layout screenshot and the style information; the structure keeping constraints are used for indicating keeping the space structure represented by the space layout screenshot, and the style migration constraints are used for indicating migrating the visual style represented by the style information; and inputting the space layout screenshot, the style information and the prompt word into the style migration model to obtain an indoor effect picture of the target house type with the migrated visual style.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a method, apparatus, terminal device and storage medium for managing interior renderings. Background Technology

[0002] With the development of science and technology, artificial intelligence has made significant progress in image style transfer scenarios, bringing new technological possibilities to interior design. Users can input floor plan data and style descriptions to automatically generate interior design renderings using artificial intelligence technology, thus lowering the design threshold to some extent.

[0003] Users have two core needs when using AI technology to generate interior design renderings: first, they expect the renderings to maintain the true structure of the original floor plan; second, they expect the model to accurately transfer materials, colors, soft furnishings, and textures based on a reference style. However, the model can easily alter the spatial structure during the generation process, and the style transfer effect is uncontrollable, resulting in a significant deviation between the final interior renderings and the user's expectations.

[0004] Therefore, how to achieve style transfer through models while maintaining the original apartment layout is a problem that needs to be solved. Summary of the Invention

[0005] In view of this, embodiments of this application provide an interior rendering management method, device, terminal equipment, and storage medium, which can effectively solve the problems of easy deformation of the floor plan structure and uncontrollable style transfer effect in interior renderings generated by style transfer using models.

[0006] In a first aspect, embodiments of this application provide a method for managing interior renderings, including: Obtain the floor plan data and style information of the target apartment type; Based on apartment type data, a browser-automated program generates a screenshot of the spatial layout that represents the actual apartment structure. Based on spatial layout screenshots and style information, prompts are constructed that include structural preservation constraints and style transfer constraints; structural preservation constraints are used to indicate the preservation of the spatial structure represented by the spatial layout screenshots, and style transfer constraints are used to indicate the transfer of the visual style represented by the style information. Input the spatial layout screenshot, style information, and prompts into the style transfer model to obtain the interior renderings of the target apartment with the transferred visual style.

[0007] In some embodiments, based on apartment type data, a spatial layout screenshot representing the actual apartment structure is generated by an automated browser program, including: loading a pre-set 3D rendering page through the automated browser program and inputting the apartment type data into the 3D rendering page; constructing a three-dimensional apartment scene based on the apartment type data through the 3D rendering page; and performing a screenshot operation on the 3D rendering page through the automated browser program to capture an image of the three-dimensional apartment scene from the target perspective, thereby obtaining a spatial layout screenshot.

[0008] In some embodiments, structural preservation constraints are used to keep one or more of the following unchanged in the spatial layout screenshot: the spatial position of the wall relationships, the position and size of the door and window openings, the spatial coordinates and orientation of the furniture, the scale ratio and perspective relationship, the ground boundary, the top boundary, and the viewpoint; style transfer constraints are used to transfer one or more of the following extracted from the style information: the material system, color relationship, soft furnishing tone, lighting atmosphere, decorative details, and spatial texture.

[0009] In some embodiments, the method further includes one or more of the following: First, when the target apartment type contains multiple functional spaces, the target apartment type is split according to the spatial dimension to obtain multiple target spaces; based on the apartment type data, a spatial layout screenshot representing the actual apartment type structure is generated by an automated browser program, including: for each target space, based on the apartment type data, a subspace layout screenshot corresponding to the target space is generated by an automated browser program; Second, a corresponding style transfer subtask and subtask identifier are created for each target space; the style transfer subtask includes at least the subspace layout screenshot and style information corresponding to the target space, used to indicate the generation of the interior rendering corresponding to the target space; the subtask identifier is used to query the execution status of the style transfer subtask; Third, in response to a task status query request, if the style transfer subtask is found to have been completed based on the subtask identifier, the interior rendering generated by the style transfer subtask is recorded; if the style transfer subtask is found to have failed based on the subtask identifier, the reason for failure is written and a prompt message is issued, the prompt message being used to indicate that the style transfer subtask has failed.

[0010] In some embodiments, constructing prompt words containing structural preservation constraints and style transfer constraints based on spatial layout screenshots and style information includes: for each target space, constructing prompt words corresponding to the target space based on the layout screenshots and style information of the subspaces corresponding to the target space; inputting the spatial layout screenshots, style information, and prompt words into a style transfer model to obtain the interior rendering of the target apartment with transferred visual style, including: for each target space, selecting a target transfer model from multiple style transfer models; encapsulating the style transfer subtasks corresponding to the target space and the prompt words corresponding to the target space into a target rendering request that matches the target transfer model; submitting the target rendering request to the target transfer model to obtain the interior rendering of the target space; and aggregating the interior renderings corresponding to each target space to obtain the interior rendering of the target apartment.

[0011] In some embodiments, for any target space, the method further includes: obtaining multiple style reference maps; if there is a target reference map in the multiple style reference maps that matches the target space, then the target reference map is used as the style information of the target space; if there is no target reference map in the multiple style reference maps that matches the target space, then one or more of the multiple style reference maps are used as the style information of the target space.

[0012] In some embodiments, the method further includes: for each target space, recording the first time consumed in the process of generating the spatial layout screenshot corresponding to the target space, and the second time consumed in the process of style transfer corresponding to the target space; displaying one or more of the following: the spatial layout screenshot, style information, interior rendering, first time consumed, second time consumed, subtask identifier, prompt words, and model information of the style transfer model corresponding to the target space.

[0013] Secondly, embodiments of this application provide an interior rendering management device, comprising: an acquisition module for acquiring floor plan data and style information of a target apartment type; and a processing module for: generating a spatial layout screenshot representing the actual apartment structure using an automated browser program based on the floor plan data; constructing prompts containing structural preservation constraints and style transfer constraints based on the spatial layout screenshots and style information; the structural preservation constraints are used to indicate maintaining the spatial structure represented by the spatial layout screenshots, and the style transfer constraints are used to indicate transferring the visual style represented by the style information; and inputting the spatial layout screenshots, style information, and prompts into a style transfer model to obtain an interior rendering of the target apartment type with transferred visual style.

[0014] Thirdly, embodiments of this application provide a terminal device, including a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement any of the indoor rendering management methods in the first aspect.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed on a processor, implements any of the interior rendering management methods in the first aspect.

[0016] The embodiments of this application have the following beneficial effects: Based on apartment layout data, a browser-based automated program generates spatial layout screenshots that represent the actual apartment structure. This transforms abstract apartment layout data into a structural image that intuitively reflects the real wall relationships, door and window locations, and furniture layout, providing a structural benchmark for subsequent style migration.

[0017] By combining spatial layout screenshots and style information, prompts containing structural preservation constraints and style transfer constraints were constructed. In other words, the spatial layout screenshots and style information were functionally separated. The spatial layout screenshots are responsible for providing structural constraints, which can be maintained in the subsequent style transfer model. The style information is responsible for providing the style source, which can be transferred in the subsequent style transfer model.

[0018] Finally, the spatial layout screenshot, style information, and prompts are input into the style transfer model. The prompts impose dual constraints on the style transfer model, enabling precise guidance in the process of generating interior renderings. This ensures that the style transfer model clearly understands "what cannot be changed and what can be changed," effectively reducing the risk of the style transfer model arbitrarily altering the spatial structure. It significantly improves the matching degree between the final generated interior renderings and the actual apartment layout, as well as the controllability of the style transfer effect. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart of an interior rendering management method provided in an embodiment of this application is shown; Figure 2 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 1 ; Figure 3 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 2 ; Figure 4 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 3; Figure 5 This illustration shows an overall flowchart of an interior rendering management method provided in an embodiment of this application. Figure 6 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 4 ; Figure 7 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 5 ; Figure 8 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 6 ; Figure 9 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 7 ; Figure 10 This illustration shows a scenario illustrating an interior rendering management method provided in an embodiment of this application. Figure 8 ; Figure 11 A schematic diagram of an interior rendering management device provided in an embodiment of this application is shown. Detailed Implementation

[0021] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0022] The components of the embodiments of this application described and illustrated in the accompanying drawings can be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0023] In the following text, the terms "comprising," "having," and their cognates, which may be used in various embodiments of this application, are intended only to indicate a particular feature, number, step, operation, element, component, or combination thereof, and should not be construed as primarily excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, or adding the possibility of one or more combinations thereof. Furthermore, the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0024] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of this application pertain. Terms (such as those defined in commonly used dictionaries) shall be interpreted as having the same meaning as in their contextual meaning in the relevant technical field and shall not be construed as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of this application.

[0025] As described in the background section, when using models to generate interior renderings through style transfer, there are problems such as easy deformation of the apartment layout and uncontrollable style transfer effects.

[0026] Based on this, this application provides an interior rendering management method. By acquiring apartment type data and style information, an automated browser program generates spatial layout screenshots based on real apartment type data. The spatial layout screenshots and style information are used to construct prompts containing structural preservation constraints and style transfer constraints. Finally, the style transfer model is input to generate renderings. This method achieves accurate transfer of target visual style while maintaining the original apartment type structure. It solves the technical problems of structural distortion and uncontrollable style in renderings generated by current models, making the final interior renderings have both structural authenticity and aesthetic appeal.

[0027] The method for managing interior renderings will be described in detail below with reference to the accompanying drawings and some specific embodiments. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0028] The indoor rendering management method provided in this application embodiment is applicable to a migration system including a client and a server.

[0029] This involves communication between the client and the server. It should be understood that the connection can be wired or wireless.

[0030] The client application provides an interactive interface for users to interact with the system. For example, the client might display a floor plan selection page or a style selection page, where users can choose their desired floor plan or perform other similar actions.

[0031] The server-side program is equipped with a browser automation program. The server is used to generate spatial layout screenshots through the browser automation program, and also to call the style transfer model to generate interior renderings.

[0032] Style transfer models can be deployed on the server or on other server nodes.

[0033] The client can be a terminal device used by the user, such as a computer, tablet, or mobile phone.

[0034] The server can be a single server or a server cluster consisting of multiple servers.

[0035] Optionally, the client and server can be two independent devices, or they can be integrated together.

[0036] Figure 1 A flowchart of an interior rendering management method provided in an embodiment of this application is shown. This interior rendering management method can be executed by the aforementioned server and includes the following steps: S101. Obtain the floor plan data and style information of the target apartment type.

[0037] In some embodiments, the floor plan data may be in the form of 2D vector floor plan, 3D building model, floor plan text description, etc.

[0038] In some embodiments, style information may take the form of style reference images, textual semantic descriptions, etc. For example, style information may use textual semantic descriptions, including style name, style description, color description, material description, soft furnishing description, etc.

[0039] In some embodiments, the floor plan data and style information of the target apartment type can be entered by the user or selected by the user from a variety of pre-stored floor plan and style sources.

[0040] For example, when a user enters... Figure 2 The floor plan selection page shown allows you to choose a floor plan as your target floor plan, and then proceed to the following steps: Figure 3 The style selection page shown allows users to choose a style. The client then sends the selected apartment type data and style information to the server for subsequent generation of interior renderings.

[0041] In some embodiments, when the target apartment type contains multiple functional spaces, the target apartment type is split according to spatial dimensions to obtain multiple target spaces. Accordingly, users can select different target styles for different target spaces to perform style transfer and generate interior renderings.

[0042] In some embodiments, for any target space, when multiple style reference maps are obtained, if there is a target reference map in the multiple style reference maps that matches the target space, then the target reference map is used as the style information of the target space; if there is no target reference map in the multiple style reference maps that matches the target space, then one or more of the multiple style reference maps are used as the style information of the target space.

[0043] For example, when a style reference image is uploaded or stored, it carries a space tag. If the target space's identifier matches the space tag, then the style reference image is determined to be a target reference image specifically set up for that target space, and the style reference image is used as the style information for that target space. If the style reference image does not carry a space tag, then the style reference image can be used as a public reference image for style migration across all spaces.

[0044] In another example, the server stores a mapping table between spaces and reference images. Multiple style reference images are stored in the mapping table. The target space is queried in the mapping table based on its identifier. The corresponding reference image found is the target reference image for that target space, and this style reference image is used as the style information for that target space.

[0045] Another example is using one or more style reference images as style information for the target space. This could be the first style reference image, the latest style reference image, the earliest style reference image, or all of the style reference images as style information for the target space.

[0046] In this way, different styles of reference images can be used for different spaces to meet refined design needs. When no dedicated reference image is available, any one or more style reference images can be used as style information to ensure that the subsequent style transfer process is not interrupted due to a lack of style input.

[0047] S102. Based on the apartment type data, generate a screenshot of the spatial layout representing the actual apartment structure through an automated browser program.

[0048] In some embodiments, for each target space, based on the apartment layout data, a browser automation program generates a screenshot of the sub-space layout corresponding to the target space. By aggregating the screenshots of the sub-space layouts corresponding to each target space, a spatial layout screenshot of the target apartment can be obtained.

[0049] Browser automation programs are Node.js-based automation scripts written using the Puppeteer or Playwright browser automation framework and deployed on the server. A browser automation program can include a task acquisition module, a browser control module, a screenshot execution module, and a result feedback module. It can invoke a browser instance started on the server to automatically perform operations such as page navigation, page loading, script invocation, and screenshot capture.

[0050] The browser automation program runs on the server side, performing screenshot tasks by launching a headless browser instance (which has no graphical user interface and only performs page rendering and screenshot operations in the server's memory). Specifically, the browser automation program communicates with the browser instance through the standard DevTools Protocol, controlling the browser to sequentially perform operations such as page navigation, page loading, script invocation, and screenshot capture.

[0051] In some embodiments, step S102 can be specifically implemented as follows: loading a preset 3D rendering page through a browser automation program and inputting the apartment layout data into the 3D rendering page; constructing a three-dimensional apartment layout scene based on the apartment layout data through the 3D rendering page; performing a screenshot operation on the 3D rendering page through a browser automation program to capture the image of the three-dimensional apartment layout scene from the target perspective and obtain a spatial layout screenshot.

[0052] The target viewpoint is obtained by matching the target room from the preset viewpoint library. The camera position and field of view of the preset viewpoint are different for different target rooms.

[0053] The 3D rendering page is pre-deployed on the server, embedding a 3D scene rendering engine. It receives floor plan data from the browser's automated program, loads the corresponding 3D floor plan data, constructs a 3D floor plan scene in real time, and switches to the corresponding preset viewpoint based on the space type. For example, the 3D rendering page can be an SDKFrame page, which is an interactive HTML page pre-deployed on the server with an embedded 3D scene rendering engine (such as Three.js).

[0054] For example, such as Figure 4 As shown, the user has selected the room and style reference image to be generated, and the server has also generated a prompt. The user can click the "Generate Rendering" control on the page to send the data to the server.

[0055] After receiving the data, the server-side layout-auto-image-web (entry service) packages all the raw data submitted by the user into a task_param and stores it in the database. From this, it extracts the following key parameters: floor_plan_id, floor_plan_url, room_type, decorate_style, decorate_style_image_urls, decorate_style_place_image_map, and decorate_style_keywords. The task data is then written to the database table of layoutai-basic (persistence service). Among them, floor_plan_id is used to locate and load specific floor plans, which are then decrypted in the backend and sent to the screenshot page; floor_plan_url is used to save or display the source of the floor plan; room_type indicates the space for which screenshots are to be generated and the number of spaces of each type; decorate_style indicates the style name; decorate_style_image_urls indicates common style reference images; decorate_style_place_image_map indicates the reference image mapping specified by space, such as {"LIVING_ROOM":"https: / / example.com / living-room-style.jpg"}; and decorate_style_keywords indicates style descriptive words.

[0056] The browser automation program in the screenshot service actively receives screenshot tasks (each screenshot task corresponds to the process of generating a spatial layout screenshot for each target space). It launches a headless browser, loads the 3D rendering page (SdkFrame page) and injects authentication status, waiting for the page bridging function to become available. Through the IFrameMsgServerExFuncs bridging object, it calls the inputRoomIdToViewCameraImage function, passing in the floor plan ID, space type, and other floor plan data. After the inputRoomIdToViewCameraImage function is executed, the SdkFrame page loads the complete 3D scene of the floor plan based on floor_plan_id, switches to the preset view of each space according to room_type, and the browser automation program calls the browser's screenshot tool to capture the currently rendered content of the 3D rendering page, obtaining the spatial layout screenshot corresponding to each space. Finally, it returns an imageList containing the screenshot information for each target space.

[0057] The `imageList` is filtered, retaining only images that possess both `placeName` (space name, e.g., living room / bedroom) and `imageUrl` (the URL of the screenshot image). The screenshot results are saved as structured data, including the space name, screenshot URL, screenshot start time, screenshot end time, node IP (the IP address of the server executing the screenshot), and error messages if any are found. After filtering, the browser automation program uses the ` / dy / updateScreenShotTask` interface to write the screenshot results back to the `layout-auto-image-web` service, storing them in the database.

[0058] Example input parameters are: event=life_leads_aigus_generate; content.aigus_id=2ac7039b-72f8-467f-bf3f-c977e760c11f; content.root_life_account_id=7234028844993218563; content.decorate_demand.floor_plan_id=8FFHFKG GKDHIJFLICEIDJJJCDFJFGHF; content.decorate_demand.floor_plan_url=https: / / img3.admin.3vjia.com / UpFile_Building / C00000022 / 3D / BuildingRoomModel / 20241130 / 914253cfe0eb49fabfcd5ecf3910de73_v1.p ng; content.decorate_demand.room_type={"living_room":"1","bedroom":"1","kitchen":"1","washroom":"1"}; content.decorate_demand.decorate_style=Modern minimalist style; content.decorate_demand.decorate_style_image_urls=["https: / / p3-orange.byteorg.com / obj / ad-tetris-site / Image000021.JPG"]; content.decorate_demand.decorate_style_keywords=["Using minimalist lines and white space as the core, the color scheme is mainly white, gray, beige and black, and the materials are mostly stone, wood veneer, glass and metal. The overall look is clean, neat, comfortable and elegant."].

[0059] By opening a pre-set 3D rendering page through a browser automation program and inputting the apartment layout data, a 3D apartment scene is automatically constructed and the target view image is captured. The real apartment structure is solidified from 3D vector data into a 2D bitmap, providing an immutable spatial skeleton benchmark for subsequent model rendering and effectively constraining the model's freedom to arrange walls, doors, windows and furniture.

[0060] S103. Construct prompt words containing structural preservation constraints and style transfer constraints based on spatial layout screenshots and style information.

[0061] Among them, the structure preservation constraint is used to indicate the preservation of the spatial structure represented by the spatial layout screenshot, and the style transfer constraint is used to indicate the transfer of the visual style represented by the style information.

[0062] Specifically, structural preservation constraints include one or more of the following: maintaining the spatial position of the wall relationships in the spatial layout screenshot; maintaining the position and size of the door and window openings in the spatial layout screenshot; maintaining the spatial coordinates and orientation of the furniture in the spatial layout screenshot; maintaining the scale ratio and perspective relationship in the spatial layout screenshot; maintaining the ground boundary in the spatial layout screenshot; maintaining the top boundary in the spatial layout screenshot; and maintaining the viewpoint in the spatial layout screenshot.

[0063] Style transfer constraints include one or more of the following: material system extracted from the transferred style information; color relationships extracted from the transferred style information; soft furnishing tone extracted from the transferred style information; lighting atmosphere extracted from the transferred style information; decorative details extracted from the transferred style information; and spatial texture extracted from the transferred style information.

[0064] By refining and multi-dimensionally deconstructing structural preservation constraints and style transfer constraints semantically, the process of generating interior renderings is upgraded from the vague "refer to this image" to a clear instruction system of "locking in structural elements one by one and transferring style elements one by one." Structural preservation constraints lock in the spatial framework from walls, doors and windows, furniture, scale and perspective to ceiling and floor boundaries. Style transfer constraints guide style reshaping from multiple dimensions, including materials, colors, soft furnishings, lighting, decorations, and texture. These two constraints form clear boundaries between invariance and change, effectively suppressing structural drift and furniture deformation caused by model improvisation, while ensuring the richness and accuracy of style transfer. This allows the model's style transfer results to achieve a controllable balance between the credibility of the real-world floor plan and the expressiveness of the target style. Furthermore, this constraint system is carried out using natural language prompts, allowing it to adapt to different style transfer models without additional training or fine-tuning, thus broadening its applicability.

[0065] In some embodiments, step S103 can be specifically implemented as follows: for each target space, construct the prompt words corresponding to the target space based on the subspace layout screenshot and style information corresponding to the target space.

[0066] In some embodiments, the prompt may also include a user-defined description.

[0067] Optionally, when a user-defined description exists, the prompt words of the user-defined description are input into the subsequent style transfer model for style transfer; when no user-defined description exists, prompt words are generated based on the spatial layout screenshot and style information.

[0068] Optionally, the user-defined description, structure preservation constraints, and style transfer constraints can be concatenated into a complete prompt word input style transfer model.

[0069] The spatial layout screenshot serves as the sole benchmark for spatial structure, basic furniture layout, and camera composition, encompassing the spatial skeleton, wall relationships, door and window positions, main furniture placement, perspective, and camera position constraints. Style information, as the sole source of style reference, serves as the reference for materials, colors, soft furnishings, lighting atmosphere, and local decorative details. The prompt explicitly requires the style transfer model to strictly maintain the wall relationships, ceiling structure, floor boundaries, door and window positions, main furniture placement, scale proportions, perspective, and camera angles from the spatial layout screenshot. Simultaneously, it must transfer only the material system, color relationships, soft furnishing tone, lighting atmosphere, decorative details, and spatial texture from the style information.

[0070] The prompts prioritize style descriptions input by the user or provided by the upstream `decorate_style_keywords`. When the user does not input any style description and the upstream does not provide one, the server-side configuration `ai-render.prompt` is used as the default prompt. When generating interior renderings using different style transfer models, the `ai-render.sysPrompt` system constraint can be added to the above prompts to adapt to the characteristics and requirements of different models.

[0071] When submitting to the style transfer model, a screenshot of the spatial layout is used as the original base image, carried by the `imageUrl` or `initImage` field, while style information is carried by the `extImageUrls` or `extra.extImageUrls` field. This separation of fields ensures that structural and style inputs remain independent at the API level, facilitating adaptation to different style transfer models.

[0072] For example, the structural preservation prompt could be: Please strictly maintain the apartment structure, wall relationships, door and window positions, ceiling boundaries, floor boundaries, main furniture positions, scale proportions, perspective relationships, and target angles shown in the space layout screenshot. Do not add partition walls, do not change room openings, do not move main furniture such as sofas, beds, and cabinets, and do not change the spatial depth and perspective.

[0073] Another example of a style migration suggestion could be: Please refer only to the style information to migrate the material system and soft furnishing atmosphere of the modern minimalist style. The colors are mainly white, light gray, beige, and natural wood color, with black metal lines used in some areas; the materials include wood veneer, light gray fabric, matte stone, transparent glass, and a small amount of metal; the lighting uses no main light, soft warm white light, and local linear light strips, with an overall clean, transparent, and restrained feel.

[0074] Another example, the complete prompt could be: Aiming for a modern minimalist style, while maintaining the spatial structure and furniture layout of the screenshot, the style should be adapted by referencing the white and gray wood tones, wood veneer, light gray fabrics, no-main-light atmosphere, and simple soft furnishings in the style information; no changes should be made to the walls, doors, windows, ceiling, floor boundaries, main furniture positions, scale proportions, perspective, or camera positions; no new partitions or changes to the space's purpose should be added; output realistic interior renderings with natural images, soft lighting, and delicate materials.

[0075] The complete prompt can also be: Please use the space layout screenshot as the sole basis for the spatial structure, keeping the wall, door, window, ceiling, floor boundaries, main furniture positions, perspective relationships, and target angles unchanged; please use the style information as the sole style reference, transferring the white, gray, and natural wood tones, wood veneer, light gray fabrics, warm white light without main lights, simple curtains, low-saturation decorative paintings, and green plant atmosphere of the modern minimalist style to the space corresponding to the space layout screenshot; do not change the number of rooms, opening positions, furniture dimensions, or space usage, and output realistic, natural, and detailed interior renderings.

[0076] The constraints and objectives in the prompts can be expressed in various natural language ways. For example, color can be expressed through natural language color names (such as white, light gray, beige, natural wood color), color scheme descriptions (such as warm gray tones, earth tones), or standard color codes (such as #F5F1EA); material can be expressed through material names (such as wood veneer, stone, glass, metal, cotton and linen fabrics), surface finishes (such as matte, glossy, brushed), and texture characteristics (such as wood grain, stone grain, woven texture); and the style of soft furnishings can be expressed through furniture (such as the style and material of sofas, beds, and dining tables) and curtains (such as material, color, and hanging style). The atmosphere can be expressed through various elements, including hanging methods, carpets (material, pattern, and coverage), lighting fixtures (type, shape, and color temperature), decorative paintings (content, color, and mounting method), and green plants (type, size, and placement). Atmosphere can be conveyed through lighting methods (natural light and artificial light), color temperature (warm white light and neutral light), light and shadow levels (bright and transparent, light and shadow contrast), and lighting styles (main light, no main light, linear light strips, and spotlights). Texture can be comprehensively constrained by overall descriptive terms (such as delicate, warm, transparent, light luxury, and retro) combined with specific content from style reference images.

[0077] The structural preservation constraint provides a negative constraint for the style transfer model, while the style transfer constraint provides a positive objective for the style transfer model. The prompt words not only set an insurmountable bottom line for the subsequent style transfer model, but also indicate the target path of style change for the style transfer model, so that the final generated interior rendering can achieve a controllable balance between structural fidelity and style transfer.

[0078] S104. Input the spatial layout screenshot, style information, and prompt words into the style transfer model to obtain the interior rendering of the target apartment with the transferred visual style.

[0079] In some embodiments, the server can invoke multiple style transfer models, and the specific style transfer model invoked can be switched according to the configuration. For example, the style transfer model can be a family of models such as gemini, gpt-image, seedream, and doubao.

[0080] Developers can pre-train different style transfer models for different space types. For example, model 1 can be trained for toilet space and model 2 can be trained for living room space. Then, for the toilet space and living room space in the target house type, model 1 can generate the interior rendering of the toilet space and model 2 can generate the interior rendering of the living room space.

[0081] When the system configuration changes, subsequent style transfer tasks will automatically use the changed model.

[0082] Optionally, the same spatial layout screenshot can be reused between different style transfer models. You can switch between the self-developed createImage link and the external link by configuration, and store fields such as model, renderWay, ai_render_task_ids, and ai_render_task_result in the database. This will facilitate the subsequent review of the performance of different models on the same structure map and the same style map, so as to compare the effects, control costs, and upgrade the models.

[0083] In some embodiments, when the target apartment type includes multiple target spaces, a corresponding style migration subtask and a subtask identifier can be created for each target space. The style migration subtask includes at least a screenshot of the subspace layout and style information corresponding to the target space, which is used to indicate the generation of the interior rendering of the target space. The subtask identifier is used to query the execution status of the style migration subtask.

[0084] When the target apartment type includes multiple target spaces, step S104 can be specifically implemented as follows: for each target space, select a target migration model from multiple style migration models; encapsulate the style migration subtask corresponding to the target space and the prompt words corresponding to the target space into a target rendering request that matches the target migration model; submit the target rendering request to the target migration model to obtain the interior rendering of the target space; aggregate the interior renderings corresponding to each target space to obtain the interior rendering of the target apartment type.

[0085] When any style transfer subtask is completed, the corresponding interior rendering is recorded. When all style transfer subtasks are completed, the interior renderings generated by each style transfer subtask are aggregated.

[0086] Optionally, when any style transfer subtask is completed, the generated interior renderings can be returned and displayed to the user first.

[0087] By constructing style transfer subtasks for each target space, multiple spaces in a house type can be matched with the most suitable style to generate interior renderings. Furthermore, the failure of the style transfer subtask for one space does not affect the normal rendering of other spaces. Multiple style transfer models can process style transfer subtasks in parallel, thereby improving the overall generation effect.

[0088] In some embodiments, in response to a task status query request, if the style transfer subtask is found to have been completed based on the subtask identifier, the interior rendering generated by the style transfer subtask is recorded; if the style transfer subtask is found to have failed based on the subtask identifier, the reason for failure is written and a prompt message is issued, which is used to indicate that the style transfer subtask has failed.

[0089] Task status query requests can be periodically issued by the server after submitting the target output request, so as to track the execution progress of style transfer subtasks in real time.

[0090] The subtask identifier allows for individual querying of the real-time processing status of each space, rather than just the overall status. When a space fails to generate a map or produces unsatisfactory results, the subtask identifier can be used to directly locate that space. Based on the written failure reason and prompt information, the process can proceed to manual troubleshooting. Manual checks can be performed on screenshots, prompt words, or model parameters without needing to review all data. Furthermore, the map generation request for that space can be resubmitted for that style transfer subtask without affecting other spaces that have already been successfully executed.

[0091] For example, after the browser automation program completes the screenshots of each space layout, it sends the screenshot results back to the server through a callback interface. The server then iterates through the screenshot callback imageList and filters out invalid images that have no placeName or no imageUrl.

[0092] The server identifies each functional space (living room, bedroom, kitchen, bathroom, etc.) by placeName and constructs an independent model rendering request for each space (not a simple string concatenation). Based on the space name, a reference image is parsed. If `decorate_style_place_image_map` matches the current space, that space-specific reference image is used; otherwise, it falls back to the `decorate_style_image_urls` public style map. An independent style transfer subtask ID is generated for each space and written to `ai_render_task_ids`, establishing the association between the style transfer subtask and the main task.

[0093] The server reads the current model parameters from the configuration: ai-render.task-type (task type), ai-render.model (model family), ai-render.source (source value), and ai-render.usei2i (link type). If the page entry provides inputs such as model, source, and size, the backend determines the final model parameters to be used by prioritizing input parameters and using configuration fallback.

[0094] The server encapsulates the imageUrl (spatial layout screenshot), extImageUrls (style reference image), prompt (prompt word), model (model parameters), and size parameters of each spatial subtask into a unified model output request.

[0095] The server calls the encapsulated large model rendering interface, submits complete parameters such as imageUrl, extImageUrls, prompt, and model, saves the style transfer subtask ID returned by the rendering service, and updates the task status of the style transfer subtask to "submitted model rendering".

[0096] Each space can generate interior renderings independently, or they can be submitted to the polling manager and queried in batches.

[0097] The server periodically queries the execution status of each style transfer subtask by polling the manager and assigning it to `renderTaskId`. When a style transfer subtask for a specific space is completed, the corresponding interior rendering image for that space is written to the server, and the status of the entire task (referring to the complete process from business submission, screenshotting, model rendering to result writeback) is set to "Model Rendering". When the number of returned interior rendering images reaches the number of `renderTasks` in this batch, all subtasks are considered to have successfully returned interior rendering images, and the overall style transfer task is considered complete. The model rendering results for all spaces are then aggregated and written to `ai_render_task_result`, and the `ai_render_task_status` and `dy_callback_task_status` of the original task are updated.

[0098] Once all style transfer subtasks are completed, the model rendering results for all spaces are aggregated, the model rendering time and total task time are calculated, and the model rendering is considered complete. If a failure or timeout occurs, the reason for the failure is written, and an alarm or manual troubleshooting process is initiated.

[0099] The final interior rendering verification includes the task status, task ID, list of returned images, and timeout / failure handling. If the generated interior rendering shows significant changes in wall, door, window, furniture dimensions, or camera position, automatic structural repair will not be performed on the submitted style migration subtask. Users can resubmit the rendering request on the client according to the changed spatial status, and the server will process it as a new rendering task.

[0100] In some embodiments, for each target space, the server records the first time taken to generate the spatial layout screenshot of the target space and the second time taken to generate the style transfer process of the target space; and displays one or more of the following: spatial layout screenshot, style information, interior rendering, first time taken, second time taken, subtask identifier, prompt words, and model information of the style transfer model of the target space.

[0101] By jointly displaying this information, the generated interior renderings can be previewed, reviewed, and statistically analyzed, facilitating comparisons of the effects of different models and prompts, and also accumulating vertical datasets for model fine-tuning.

[0102] The following describes the interior rendering management solution provided in this application from an overall process perspective, such as... Figure 5 As shown: 1. Data Input: Data is submitted through the business entry point (user front-end page), for example... Figure 6 The floor plan shown and Figure 7 The style reference image shown; 2. Spatial layout screenshot generation: layout-auto-image-web (server-side interface layer) parses content.decorate_demand → layoutai-basic (server-side core layer) writes to layout_screen_shot_task → screenshotworker (browser automation program) calls / dy / claimNextLayoutScreenShotTask to claim the task → loads SDKFrame (3D rendering page) and completes authentication → calls inputRoomIdToViewCameraImage to load the floor plan and spatial perspective → returns imageList screenshot results → / dy / updateScreenShotTask writes back the screenshot results. The generated spatial layout screenshot is as follows: Figure 8 As shown; 3. Parameter assembly and model selection: Spatial layout screenshots constrain spatial structure and composition; style reference images transfer materials, colors, and atmosphere; select the target style transfer model from multiple models; layout-auto-image-web assembles imageUrl, extImageUrls, prompt, and model to generate style transfer subtasks; 4. Interior Rendering Generation: Execute the style transfer subtask to generate interior renderings, such as... Figure 9 As shown; the same spatial layout screenshot can be reused in different style transfer models; 5. Result Aggregation: Poll the renderTaskId and aggregate the results; 6. Results Display and Review: The `layout_screen_shot_task_result` function saves the `ai_render_task_result`, displaying style reference images, spatial layout screenshots, generated model renderings, and time consumption information for each stage. A detailed display page can be found as follows: Figure 10 The diagram shown (style reference image and time taken at each stage are not shown) is used for result display and review.

[0103] In summary, this application forms a complete closed loop through "reusing intermediate screenshots + dual reference inputs of structure and style + independent style transfer subtasks + result data accumulation".

[0104] The browser-automated program generates spatial layout screenshots based on real apartment data, real spatial perspectives, and business rules, serving as structural anchors for subsequent model style transfer. These screenshots pre-define the apartment structure, basic furniture layout, and camera positions as image input, preventing spatial drift caused by the model freely imagining the apartment structure from text.

[0105] Using the spatial layout screenshot as the structural baseline, the relationships between walls, the positions of doors and windows, the layout of furniture, the scale and proportions, and the target perspective are locked. Style information provides visual style sources such as materials, colors, soft furnishings, and lighting. Through the structural maintenance constraints and style transfer constraints in the prompts, the style transfer model can complete the style transfer while maintaining the original floor plan structure.

[0106] The overall task can also be broken down into multiple independent style transfer subtasks based on functional space dimensions (living room, bedroom, kitchen, bathroom, etc.). Each style transfer subtask independently matches reference images (space-specific reference images take priority, and public reference images are rolled back), independently submits models, and independently queries execution status. The images of each space are generated in parallel without affecting each other.

[0107] Record the original input data, spatial layout screenshots, model parameters, task ID, interior renderings, time consumption, and reasons for failure. This allows for iterative optimization of prompts, reference image strategies, and model configurations based on real task data, forming a data-driven closed loop for continuous optimization.

[0108] Spatial layout screenshots solidify the actual apartment structure into intermediate products, which serve as structural anchors for inputting the style transfer model. Independent style transfer subtasks enable parallel processing and fine-grained control across multiple spaces, and the accumulated results support subsequent strategy iteration and optimization. Intermediate products and results generated at any stage of the overall process can be reused in subsequent tasks, forming a complete closed loop of "screenshot anchoring of structure → model style transfer → data feedback optimization".

[0109] Figure 11 A schematic diagram of an interior rendering management device according to an embodiment of this application is shown. Exemplarily, the interior rendering management device 200 includes an acquisition module 201 and a processing module 202.

[0110] The acquisition module 201 is used to acquire the floor plan data and style information of the target apartment type; the processing module 202 is used to: generate a spatial layout screenshot representing the actual floor plan structure through an automated browser program based on the floor plan data; construct prompt words containing structure preservation constraints and style transfer constraints based on the spatial layout screenshot and style information; the structure preservation constraint is used to indicate that the spatial structure represented by the spatial layout screenshot is preserved, and the style transfer constraint is used to indicate that the visual style represented by the style information is transferred; input the spatial layout screenshot, style information and prompt words into the style transfer model to obtain the interior rendering of the target apartment type with transferred visual style.

[0111] In some embodiments, the processing module 202 is specifically used to: load a preset 3D rendering page through a browser automation program and input the floor plan data into the 3D rendering page; construct a three-dimensional floor plan scene based on the floor plan data through the 3D rendering page; and perform a screenshot operation on the 3D rendering page through a browser automation program to capture the image of the three-dimensional floor plan scene from the target perspective, thereby obtaining a spatial layout screenshot.

[0112] In some embodiments, the processing module 202 is further configured to: First, when the target apartment type contains multiple functional spaces, split the target apartment type according to the spatial dimension to obtain multiple target spaces; specifically, for each target space, based on the apartment type data, generate a screenshot of the sub-space layout corresponding to the target space through a browser automation program; Second, create a corresponding style migration sub-task and a sub-task identifier for each target space; the style migration sub-task includes at least the screenshot of the sub-space layout and style information corresponding to the target space, used to indicate the generation of the interior rendering corresponding to the target space; the sub-task identifier is used to query the execution status of the style migration sub-task; Third, in response to the task status query request, if the style migration sub-task is found to have been completed based on the sub-task identifier, record the interior rendering generated by the style migration sub-task; if the style migration sub-task is found to have failed based on the sub-task identifier, write the failure reason and issue a prompt message, the prompt message being used to indicate that the style migration sub-task has failed.

[0113] In some embodiments, the processing module 202 is specifically used to: for each target space, construct prompt words corresponding to the target space based on the layout screenshots and style information of the subspaces corresponding to the target space; for each target space, select a target migration model from multiple style migration models; encapsulate the style migration subtasks corresponding to the target space and the prompt words corresponding to the target space into a target rendering request that matches the target migration model; submit the target rendering request to the target migration model to obtain the interior rendering of the target space; and aggregate the interior renderings corresponding to each target space to obtain the interior rendering of the target apartment type.

[0114] In some embodiments, the acquisition module 201 is further configured to acquire multiple style reference maps; the processing module 202 is further configured to: if there is a target reference map in the multiple style reference maps that matches the target space, then use the target reference map as the style information of the target space; if there is no target reference map in the multiple style reference maps that matches the target space, then use one or more of the multiple style reference maps as the style information of the target space.

[0115] In some embodiments, the processing module 202 is further configured to: for each target space, record the first time consumed in the process of generating the spatial layout screenshot corresponding to the target space, and the second time consumed in the process of style transfer corresponding to the target space; and display one or more of the following: the spatial layout screenshot, style information, interior rendering, first time consumed, second time consumed, subtask identifier, prompt words, and model information of the style transfer model corresponding to the target space.

[0116] It is understood that the device in this embodiment corresponds to the indoor rendering management method in the above embodiment, and the options in the above embodiment are also applicable to this embodiment, so they will not be described again here.

[0117] This application also provides a terminal device, exemplary of which includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to enable the terminal device to perform the functions of the various modules in the above-described interior rendering management method or the above-described interior rendering management device.

[0118] The processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including at least one of a Central Processing Unit (CPU), Graphics Processing Unit (GPU), Network Processor (NP), Digital Signal Processor (DSP), Application-Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0119] The memory can be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory is used to store computer programs, and the processor can execute the computer programs accordingly after receiving execution instructions.

[0120] This application also provides a computer-readable storage medium for storing the computer program used in the aforementioned terminal device. For example, the computer-readable storage medium may include, but is not limited to, various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0121] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that, in alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0122] In addition, the functional modules or units in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0123] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0124] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for managing interior renderings, characterized in that, include: Obtain the floor plan data and style information of the target apartment type; Based on the apartment type data, a screenshot of the spatial layout representing the actual apartment structure is generated through an automated browser program. Based on the spatial layout screenshot and the style information, a prompt word containing structure preservation constraints and style transfer constraints is constructed; the structure preservation constraint is used to indicate that the spatial structure represented by the spatial layout screenshot is preserved, and the style transfer constraint is used to indicate that the visual style represented by the style information is transferred. The spatial layout screenshot, the style information, and the prompt words are input into the style transfer model to obtain the interior rendering of the target apartment type with the transferred visual style.

2. The method according to claim 1, characterized in that, The step of generating a spatial layout screenshot representing the actual apartment structure using a browser automation program based on the apartment type data includes: The browser's automated program loads a pre-set 3D rendering page and inputs the apartment layout data into the 3D rendering page. The 3D rendering page constructs a three-dimensional apartment scene based on the apartment data. The browser's automated program performs a screenshot operation on the 3D rendered page, capturing an image of the 3D apartment scene from the target perspective to obtain the spatial layout screenshot.

3. The method according to claim 1, characterized in that, The structural constraints are used to maintain one or more of the following in the spatial layout screenshot: the spatial position of the wall relationships, the position and size of the door and window openings, the spatial coordinates and orientation of the furniture, the scale and perspective relationships, the ground boundary, the top boundary, and the viewpoint. The style transfer constraint is used to transfer one or more of the material system, color relationship, soft furnishing tone, lighting atmosphere, decorative details, and spatial texture extracted from the style information.

4. The method according to claim 1, characterized in that, The method also includes one or more of the following: The first step is to split the target apartment type into multiple target spaces when the target apartment type contains multiple functional spaces according to the spatial dimension. The step of generating a spatial layout screenshot representing the actual apartment structure using a browser automation program based on the apartment type data includes: For each target space, based on the apartment layout data, the browser automation program generates a screenshot of the sub-space layout corresponding to the target space; The second step involves creating a corresponding style transfer subtask and a subtask identifier for each target space. The style transfer subtask includes at least a screenshot of the subspace layout and style information corresponding to the target space, used to indicate the generation of an interior rendering of the target space. The subtask identifier is used to query the execution status of the style transfer subtask. Thirdly, in response to a task status query request, if the style transfer subtask is found to have been completed based on the subtask identifier, the interior rendering generated by the style transfer subtask is recorded. If the style transfer subtask fails to execute based on the subtask identifier, write the reason for the failure and issue a prompt message. The prompt message is used to indicate that the style transfer subtask has failed to execute.

5. The method according to claim 4, characterized in that, The prompt words constructed based on the spatial layout screenshot and the style information, which include structural preservation constraints and style transfer constraints, include: For each target space, based on the layout screenshot and style information of the subspace corresponding to the target space, construct the prompt words corresponding to the target space; The process of inputting the spatial layout screenshot, the style information, and the prompt words into a style transfer model to obtain an interior rendering of the target apartment type with the transferred visual style includes: For each target space, a target transfer model is selected from multiple style transfer models; The style transfer subtask corresponding to the target space and the prompt words corresponding to the target space are encapsulated into a target output request that matches the target transfer model; The target rendering request is submitted to the target migration model to obtain the interior rendering of the target space; By aggregating the interior renderings corresponding to each of the target spaces, the interior rendering of the target apartment type is obtained.

6. The method according to claim 4, characterized in that, For any one of the target spaces, the method further includes: Multiple style reference images were obtained; If there is a target reference map in the plurality of style reference maps that matches the target space, then the target reference map is used as the style information of the target space; If there is no target reference map matching the target space in the plurality of style reference maps, then one or more of the plurality of style reference maps are used as the style information of the target space.

7. The method according to claim 4, characterized in that, The method further includes: For each target space, record the first time consumed in the process of generating the spatial layout screenshot corresponding to the target space, and the second time consumed in the process of style transfer corresponding to the target space; Display one or more of the following: a screenshot of the spatial layout corresponding to the target space, the style information, the interior rendering, the first time taken, the second time taken, the subtask identifier, the prompt words, and the model information of the style transfer model.

8. An interior rendering management device, characterized in that, include: The acquisition module is used to acquire the floor plan data and style information of the target apartment type; Processing module, used for: Based on the apartment type data, a screenshot of the spatial layout representing the actual apartment structure is generated through an automated browser program. Based on the spatial layout screenshot and the style information, a prompt word containing structure preservation constraints and style transfer constraints is constructed; the structure preservation constraint is used to indicate that the spatial structure represented by the spatial layout screenshot is preserved, and the style transfer constraint is used to indicate that the visual style represented by the style information is transferred. The spatial layout screenshot, the style information, and the prompt words are input into the style transfer model to obtain the interior rendering of the target apartment type with the transferred visual style.

9. A terminal device, characterized in that, The terminal device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the interior rendering management method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed on a processor, implements the interior rendering management method according to any one of claims 1-7.

Images