Home decoration design system based on artificial intelligence

Abstract

The invention provides a home decoration design system based on artificial intelligence, and the system comprises a house type database module, an intelligent design generation module, a commodity model integration module which is in data interconnection with a commodity management system of a local shopping mall and comprises a three-dimensional home model database of multiple brands; meanwhile, a multi-channel projection unit, a curved surface reflection curtain wall and a gesture capture system are utilized, and the naked eye 3D display module is in data connection with the intelligent design generation module and used for displaying a design scheme in a stereoscopic image mode according to the proportion of 1: 1. The system has the beneficial effects that compared with traditional home decoration design which needs a designer to manually draw a planar graph, model and select materials and consumes several days and even several weeks, the system automatically analyzes house type data through AI and generates a three-dimensional model, and the designer only needs to select a basic framework from multiple schemes recommended by the system and then finely adjust details through an interaction tool.

Description

Technical Field

[0001] The present invention belongs to the field of home decoration design systems, and in particular relates to a home decoration design system based on artificial intelligence. Background Art

[0002] In the field of home decoration design, traditional design processes face significant efficiency bottlenecks and user experience shortcomings. For example, designers need to manually draw floor plans and convert them into 3D models. Modeling complex apartment layouts can take days and is costly to modify. Existing design software lacks intelligent assistance, making it difficult to quickly generate multiple plan comparisons. The furniture, building materials, and other elements used in design plans are mostly virtual models with no direct connection to actual store merchandise, making it difficult to implement the design plans and forcing clients to spend extra time matching products. Traditional renderings rely on 2D flat displays or VR equipment. The former lacks spatial immersion, while the latter requires specialized equipment and struggles to support collaborative review by multiple people. Design data, product information, and construction parameters are scattered across different systems, making modifications impossible to synchronize in real time and prone to version errors.

[0003] In summary, existing technologies have not yet solved the problems of product data integration and immersive interaction, and they also lack a mechanism for connecting with physical shopping malls. Therefore, there is an urgent need for a home decoration design system and method that connects design, products, and display. Summary of the Invention

[0004] In view of this, the present invention aims to propose an artificial intelligence-based home decoration design system to at least solve one problem in the background technology.

[0005] To achieve the above object, the technical solution of the present invention is achieved as follows: An artificial intelligence-based home decoration design system, comprising: A housing database module is configured to store building information model data within the target area, including building floor plans, three-dimensional coordinates of pipeline layouts, and spatial topological relationship matrices, and the module integrates a geographic information system interface; An intelligent design generation module, connected to the apartment database module, including a pre-trained deep neural network, a hard decoration solution library, and a soft decoration matching rule engine; The product model integration module is interconnected with the local shopping mall's product management system data. It includes a database of 3D home furnishing models from multiple brands, with each model bound to a product SKU code, price range, and inventory status tag. It also includes a real-time data synchronization unit that updates product inventory information hourly through a RESTful API interface. The naked-eye 3D display module includes a multi-channel projector group, a curved reflective curtain wall and a gesture capture system. The naked-eye 3D display module is data-connected to the intelligent design generation module and is used to present the design scheme in a 1:1 ratio stereoscopic image.

[0006] Furthermore, the intelligent design generation module includes: A style transfer submodule, which uses a generative adversarial network architecture and includes a feature extractor, a content decoder, and a style discriminator. The feature extractor is configured to separate spatial layout features and material texture features from a reference image uploaded by the user; the content decoder maps the extracted features to the three-dimensional space of the target apartment. The system includes a layout optimization submodule configured to perform the following operations: analyze the weight of the passage paths of each room in the apartment based on a graph neural network; calculate the safety threshold of the furniture spacing based on ergonomic rules; generate three or more alternative layout plans and display a space utilization comparison chart; It includes a material matching submodule, which contains a material conflict detection algorithm. The material conflict detection algorithm is used to identify whether the material color difference of adjacent walls exceeds the visual tolerance threshold of ΔE<3, and automatically adjusts the floor tile UV mapping ratio to avoid texture splicing misalignment.

[0007] Furthermore, the product model integration module includes: The model replacement logic unit is configured to: receive a furniture model identifier selected by a user in a naked-eye 3D environment; retrieve a 3D model of a similar product from a product model library, and maintain a coordinate transformation matrix of the original model; A brand information overlay unit configured to dynamically display the following information layers in the stereoscopic projection interface: a floating brand logo and price tag positioned on top of the corresponding object's three-dimensional bounding box; and a store navigation hotspot that, when clicked, retrieves a mall floor plan and generates an optimal route plan. The inventory status synchronization unit is configured to: when the inventory level of the product is lower than a preset threshold, superimpose a semi-transparent warning layer on the surface of the three-dimensional model; and provide a list of recommended alternative products, sorted by brand association and price similarity.

[0008] Furthermore, the naked-eye 3D display module includes: An optical display assembly comprising a laser projector that covers four quadrants of the curved curtain wall in an orthogonal projection manner; and a parallax barrier filter set configured to generate left-eye and right-eye disparity images; An interactive sensing component, comprising a TOF depth camera array and an inertial measurement unit, which is embedded in a handheld controller to detect six-degree-of-freedom pose changes; A multi-user collaborative component is configured to: establish independent rendering viewports, each viewport corresponding to the observation perspective of a specific user; and implement version branch management of object replacement operations at the data layer.

[0009] Furthermore, this solution discloses a home decoration design method, comprising: Unit data loading phase: Receives the community name or map coordinates entered by the user through the touch interface and matches the closest BIM model from the localized unit library. If there is a discrepancy, the unit calibration tool is activated. Intelligent design generation stage: The style selection panel obtains user-specified style tags and reference cases, and calls a pre-trained neural network to generate a design package containing hard decoration solutions, soft decoration matching, and lighting parameters within a set time; 3D interaction stage: Load the design plan in the naked-eye 3D environment, display the product brand information layer simultaneously, capture user gestures, and update the rendering effect after object replacement in real time; Data output stage: Export construction drawing files that meet the standards, and generate material procurement lists and shopping mall navigation information that match the design plan.

[0010] Furthermore, the training process of the neural network specifically includes: Data preparation steps: Collect historical design plans, each set including floor plans, 3D model parameters, and 360° panoramic renderings; vectorize the floor plans to extract wall centerlines, door and window opening coordinates, and room function labels; Model construction steps: The first stage adopts a contrastive learning framework and uses ResNet-101 to extract global features of spatial layout; the second stage adopts the CycleGAN architecture to establish a bidirectional mapping between the material texture domain and the light and shadow effect domain; Optimize the verification process: Introduce a differentiable renderer and minimize the FID between the predicted renderings and real cases through end-to-end training. Add attention masks for irregular-shaped housing units to enhance the model's feature extraction capabilities for unconventional spatial structures.

[0011] Furthermore, the present solution discloses an electronic device, comprising a processor and a memory communicatively connected to the processor and used to store instructions executable by the processor, wherein the processor is used to execute a home decoration design method.

[0012] Furthermore, the present solution discloses a server comprising at least one processor and a memory communicatively connected to the processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the processor so that the at least one processor executes a home decoration design method.

[0013] Furthermore, the present solution discloses a computer-readable storage medium storing a computer program, which implements a home decoration design method when executed by a processor.

[0014] Compared with the existing technology, the artificial intelligence-based home decoration design system described in the present invention has the following beneficial effects: (1) The artificial intelligence-based home decoration design system described in the present invention requires designers to manually draw floor plans, build models, and select materials, which takes days or even weeks. This system uses AI to automatically analyze house type data and generate three-dimensional models. Designers only need to select a basic framework from multiple solutions recommended by the system and then fine-tune the details through interactive tools. (2) The artificial intelligence-based home decoration design system described in the present invention allows customers to view design plans in a 1:1 scale in a naked-eye 3D display space, freely adjust furniture layouts through gestures (such as dragging the position of a sofa or rotating the angle of a lamp), and observe changes in light and shadow in real time (such as the effect of natural light incident at different times). This immersive experience helps customers intuitively understand the design intent and reduce communication misunderstandings. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which constitute part of the present invention, are provided to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are provided to explain the present invention and do not constitute an undue limitation of the present invention. In the accompanying drawings: Figure 1 This is a schematic diagram of an artificial intelligence-based home decoration design system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of an artificial intelligence-based home decoration design method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0016] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0017] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0018] Example 1: Household data processing and intelligent generation When a user enters the apartment type information of a certain community in the system, the system first calls the pre-stored building information model data from the local apartment type library. For example, if the user selects "a 120㎡ apartment type in a real estate project in Pudong New District, Shanghai", the system automatically loads the BIM model data including wall positioning, pipeline direction and spatial topology relationship. If there is a difference between the CAD drawing uploaded by the user and the model in the library (such as a wall offset of 30mm), the system starts the calibration tool, aligns the center line of the load-bearing wall through the edge detection algorithm, and marks the difference area with a highlighted dotted line in the interface for the user to confirm. The system analyzes the room boundaries and door and window positions in the floor plan, identifies functional areas such as the living room (28.5㎡) and master bedroom (18.2㎡), and automatically calculates the access path weight of each area (such as the shortest path from the living room to the kitchen is 2.8m). When the user chooses the "Modern Minimalist" style, the system uses a pre-trained generative adversarial network to transfer the light-colored wood texture features of the reference case to the current apartment type, generating three layout plans. The optimal plan has a space utilization rate of 83%, and automatically detects whether the furniture spacing meets ergonomic standards (for example, an alarm will be triggered if the distance between the sofa and the coffee table is insufficient).

[0019] Example 2: Dynamic replacement of product models and inventory linkage In a naked-eye 3D display, when a user selects a fabric sofa model in the living room, the system searches the local mall's inventory for similar products (length range: 2.0-2.4 meters), filtering out items with fewer than five units in stock (e.g., a certain brand's sofa is marked with a red warning). If the user chooses to replace it with a leather sofa (2.2 meters long) from another brand, the system automatically adjusts the model's coordinates to fit the original position and simultaneously updates the layout of adjacent furniture (e.g., offsetting the coffee table by 100mm). When inventory data is updated via the RESTful interface (e.g., if a product's inventory drops from 10 to 3 units), the system overlays a semi-transparent warning layer on the model and recommends similar products with similar prices (with a price fluctuation of ±10%). When a user clicks a product tag, the system uses the mall floor plan to generate a navigation route (e.g., the shortest route from the elevator on the current floor to the store), displaying the brand logo and real-time price.

[0020] Example 3: Glasses-free 3D collaborative design and real-time rendering The design team uses a multi-channel projector group to view the 1:1 scale design plan on the curved curtain wall. Four laser projectors project images at orthogonal angles, and the parallax grating filter group generates different images for the left and right eyes, allowing users to observe the three-dimensional effect without wearing equipment. When designer A uses gestures to grab the virtual dining table and adjust its position, the TOF depth camera array captures hand movements at a rate of 30 frames per second. The system updates the model coordinates and recalculates the lighting projection in real time. When designer B rotates the view angle using a handheld controller to view the details of the ceiling, the inertial measurement unit detects the six-degree-of-freedom posture changes, and the independent rendering viewports synchronously display different viewing angles. When multiple people modify the wall color and furniture material at the same time, the system records the operation history through version branch management, supporting one-click backtracking to the design plan at any point in time.

[0021] Example 1: Household data processing and intelligent generation When a user enters apartment information for a residential complex into the system, it first retrieves pre-stored Building Information Model (BIM) data from the local apartment library. For example, if a user selects "a 120 square meter apartment in a real estate development in Pudong New District, Shanghai," the system automatically loads BIM model data, including wall positioning, pipeline routing, and spatial topology. If there are discrepancies between the uploaded CAD drawing and the model in the library (e.g., a wall offset by 30 mm), the system activates a calibration tool, aligning the centerlines of the load-bearing walls using an edge detection algorithm and highlighting the discrepancies with a dotted line in the interface for user confirmation. The system analyzes the room boundaries and door and window locations in the floor plan, identifying functional areas such as the living room and master bedroom, and automatically calculates path weights for each area (e.g., the shortest path from the living room to the kitchen is 2.8 meters). If the user selects the "Modern Minimalist" style, the system uses a pre-trained generative adversarial network to transfer the light-colored wood texture features of a reference example to the current apartment, generating three layout options and automatically checking whether furniture spacing meets ergonomic standards (e.g., triggering an alarm if the sofa and coffee table are not spaced sufficiently apart).

[0022] Example 2: Dynamic replacement of product models and inventory linkage In the naked-eye 3D display, when a user selects a fabric sofa model in the living room, the system searches the local mall's inventory for similar products, filtering out items with fewer than five units in stock (e.g., a certain brand of sofa is marked with a red warning). If the user chooses to replace the sofa with a leather sofa from another brand, the system automatically adjusts the model's coordinates to fit the original position and simultaneously updates the layout of adjacent furniture (e.g., offsetting the coffee table by 100mm). When inventory data is updated via the RESTful interface (e.g., if a product's inventory drops from 10 units to 3 units), the system overlays a semi-transparent warning layer on the model's surface and recommends similar products at similar prices. When a user clicks a product tag, the system uses the mall floor plan to generate a navigation path (e.g., the shortest route from the elevator on the current floor to the store), displaying the brand logo and real-time price.

[0023] Example 3: Glasses-free 3D collaborative design and real-time rendering The design team uses a multi-channel projector group to view the 1:1 scale design plan on the curved curtain wall. Four laser projectors project images at orthogonal angles, and the parallax grating filter group generates different images for the left and right eyes, allowing users to observe the three-dimensional effect without wearing equipment. When designer A uses gestures to grab the virtual dining table and adjust its position, the TOF depth camera array captures hand movements at a rate of 30 frames per second. The system updates the model coordinates and recalculates the lighting projection in real time. When designer B rotates the view angle using a handheld controller to view the details of the ceiling, the inertial measurement unit detects the six-degree-of-freedom posture changes, and the independent rendering viewports synchronously display different viewing angles. When multiple people modify the wall color and furniture material at the same time, the system records the operation history through version branch management, supporting one-click backtracking to the design plan at any point in time.

[0024] Example 4: Construction drawing generation and data integration Once the design plan is confirmed, the system generates construction drawings that comply with national drafting standards and annotate key parameters. A comprehensive diagram of water and electricity pipelines is generated using the BIM model, noting the requirements for cross-shielding of power and weak current. The material purchase list automatically links to the store's product library, listing product information. When a particular item is out of stock, an alternative is marked in the list. Construction workers use AR glasses to view the 3D model and the on-site results. The system compares the construction progress in real time and prompts corrections for any deviations.

[0025] Example 5: AI model training and optimization verification The system trains a deep neural network using historical design examples. It first vectorizes floor plans and extracts spatial layout features using ResNet-101. In the second phase, the CycleGAN architecture is used to transfer the "New Chinese" style to unusually shaped apartments (such as curved living rooms) and optimize material texture mapping.

[0026] Those skilled in the art will appreciate that the units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0027] In the several embodiments provided in this application, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the division of the units described above is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. The above-mentioned units may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiment of the present invention.

[0028] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and description of the present invention.

[0029] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A home decoration design system based on artificial intelligence, characterized by: include: A housing database module is configured to store building information model data within the target area, including building floor plans, three-dimensional coordinates of pipeline layouts, and spatial topological relationship matrices, and the module integrates a geographic information system interface; An intelligent design generation module, connected to the apartment database module, including a pre-trained deep neural network, a hard decoration solution library, and a soft decoration matching rule engine; The product model integration module is interconnected with the local shopping mall's product management system data. It includes a database of 3D home furnishing models from multiple brands, with each model bound to a product SKU code, price range, and inventory status tag. It also includes a real-time data synchronization unit that updates product inventory information hourly through a RESTful API interface. The naked-eye 3D display module includes a multi-channel projector group, a curved reflective curtain wall and a gesture capture system. The naked-eye 3D display module is data-connected to the intelligent design generation module and is used to present the design scheme in a 1:1 ratio stereoscopic image.

2. The artificial intelligence-based home decoration design system according to claim 1, characterized in that: The intelligent design generation module includes: A style transfer submodule, which uses a generative adversarial network architecture and includes a feature extractor, a content decoder, and a style discriminator. The feature extractor is configured to separate spatial layout features and material texture features from a reference image uploaded by the user; the content decoder maps the extracted features to the three-dimensional space of the target apartment. The system includes a layout optimization submodule configured to perform the following operations: analyze the weight of the passage paths of each room in the apartment based on a graph neural network; calculate the safety threshold of the furniture spacing based on ergonomic rules; generate three or more alternative layout plans and display a space utilization comparison chart; It includes a material matching submodule, which contains a material conflict detection algorithm. The material conflict detection algorithm is used to identify whether the material color difference of adjacent walls exceeds the visual tolerance threshold of ΔE<3, and automatically adjusts the floor tile UV mapping ratio to avoid texture splicing misalignment.

3. The artificial intelligence-based home decoration design system according to claim 1, characterized in that: The commodity model integration module includes: The model replacement logic unit is configured to: receive a furniture model identifier selected by a user in a naked-eye 3D environment; retrieve a 3D model of a similar product from a product model library, and maintain a coordinate transformation matrix of the original model; A brand information overlay unit configured to dynamically display the following information layers in the stereoscopic projection interface: a floating brand logo and price tag positioned on top of the corresponding object's three-dimensional bounding box; and a store navigation hotspot that, when clicked, retrieves a mall floor plan and generates an optimal route plan. The inventory status synchronization unit is configured to: when the inventory level of the product is lower than a preset threshold, superimpose a semi-transparent warning layer on the surface of the three-dimensional model; and provide a list of recommended alternative products, sorted by brand association and price similarity.

4. The artificial intelligence-based home decoration design system according to claim 1, characterized in that: The naked eye 3D display module includes: An optical display assembly comprising a laser projector that covers four quadrants of the curved curtain wall in an orthogonal projection manner; and a parallax barrier filter set configured to generate left-eye and right-eye disparity images; An interactive sensing component, comprising a TOF depth camera array and an inertial measurement unit, which is embedded in a handheld controller to detect six-degree-of-freedom pose changes; A multi-user collaborative component is configured to: establish independent rendering viewports, each viewport corresponding to the observation perspective of a specific user; and implement version branch management of object replacement operations at the data layer.

5. A home decoration design method, characterized in that: include: Unit data loading phase: Receives the community name or map coordinates entered by the user through the touch interface and matches the closest BIM model from the localized unit library. If there is a discrepancy, the unit calibration tool is activated. Intelligent design generation stage: The style selection panel obtains user-specified style tags and reference cases, and calls a pre-trained neural network to generate a design package containing hard decoration solutions, soft decoration matching, and lighting parameters within a set time; 3D interaction stage: Load the design plan in the naked-eye 3D environment, display the product brand information layer simultaneously, capture user gestures, and update the rendering effect after object replacement in real time; Data output stage: Export construction drawing files that meet the standards, and generate material procurement lists and shopping mall navigation information that match the design plan.

6. A home decoration design method according to claim 5, characterized in that: The training process of the neural network specifically includes: Data preparation steps: Collect historical design plans, each set including floor plans, 3D model parameters, and 360° panoramic renderings; vectorize the floor plans to extract wall centerlines, door and window opening coordinates, and room function labels; Model construction steps: The first stage adopts a contrastive learning framework and uses ResNet-101 to extract global features of spatial layout; the second stage adopts the CycleGAN architecture to establish a bidirectional mapping between the material texture domain and the light and shadow effect domain; Optimize the verification process: Introduce a differentiable renderer and minimize the FID between the predicted renderings and real cases through end-to-end training. Add attention masks for irregular-shaped housing units to enhance the model's feature extraction capabilities for unconventional spatial structures.

7. An electronic device comprising a processor and a memory in communication with the processor and configured to store instructions executable by the processor, wherein: The processor is used to execute a home decoration design method as described in any one of claims 5-6 above.

8. A server, characterized in that: It includes at least one processor and a memory communicatively connected to the processor, wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the processor so that the at least one processor executes a home decoration design method as described in any one of claims 5-6.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the home decoration design method according to any one of claims 5-6 is implemented.

Images