An ai-based building decoration pattern design method
By constructing knowledge graphs and generative adversarial networks using AI, intelligent extraction and process adaptation of architectural decorative patterns are achieved, solving the problem of disconnect between pattern design and construction in existing technologies. This enables intelligent and efficient decorative design, connecting historical and modern culture and craftsmanship.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHEJIANG GUANGSHA COLLEGE OF APPLIED CONSTRTECH
- Filing Date
- 2026-03-24
- Publication Date
- 2026-07-21
AI Technical Summary
In existing architectural decoration design, the extraction of patterns lacks an understanding of cultural connotations, and the generation process ignores the feasibility of the technology, resulting in a disconnect between design and construction, and a lack of intelligent integrated design methods.
By using AI technology to construct a knowledge graph and combining it with generative adversarial networks, we can achieve intelligent pattern extraction, cultural semantic analysis, and automatic process adaptation, generating decorative patterns that conform to aesthetics and cultural semantics, and displaying and providing construction guidance through 3D models.
It has enabled intelligent and efficient pattern design, built a digital bridge connecting historical context, contemporary aesthetics and implementation techniques, and provided a systematic solution for cultural inheritance and innovation.
Smart Images
Figure CN122433469A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of architectural decoration design technology, specifically relating to an AI-based method for designing architectural decorative patterns. Background Technology
[0002] Architectural decorative patterns are an important carrier of architectural art, embodying rich cultural semantics, craftsmanship, and aesthetic connotations. Traditional architectural decorative design relies on the experience of designers and manual drawing, resulting in problems such as low efficiency, inaccurate cultural expression, and difficulty in adapting to different techniques. Currently, although some digital design tools exist, there is still a lack of intelligent design systems that can deeply integrate pattern characteristics, craftsmanship, and cultural semantics.
[0003] In existing technologies, pattern extraction is mostly based on image processing, lacking an understanding of cultural connotations; pattern generation focuses on aesthetic form, ignoring technological feasibility; and the design of decorative schemes is disconnected from actual construction, making it difficult to implement the design.
[0004] Therefore, there is an urgent need for an integrated design method that can achieve intelligent pattern extraction, cultural semantic analysis, and automatic process adaptation. Summary of the Invention
[0005] The purpose of this invention is to provide an AI-based method for designing architectural decorative patterns to address the problems mentioned in the background section. The AI-based method for designing architectural decorative patterns provided by this invention features intelligent pattern extraction, cultural semantic analysis, and automatic process adaptation.
[0006] To achieve the above objectives, the present invention provides the following technical solution: an AI-based method for designing architectural decorative patterns, comprising the following steps:
[0007] S1. Collect architectural decorative pattern image data and construct a knowledge graph that includes pattern units, craft attributes, cultural semantics and aesthetic features;
[0008] S2. Using AI image recognition and feature extraction algorithms, extract the texture, structure, color, and topological features of the pattern from the input image;
[0009] S3. Based on knowledge graphs and feature data, pattern reconstruction and innovative design are carried out through generative adversarial networks to generate decorative patterns that conform to architectural aesthetics and cultural semantics.
[0010] S4. Based on the pattern characteristics and architectural space attributes, automatically match the appropriate decorative techniques, material textures, and construction parameters;
[0011] S5. Combine user preferences and spatial scenarios to generate a 3D model of the decoration scheme, and perform virtual simulation and visualization display;
[0012] S6. Outputs digital design documents including pattern drafts, process parameters, material lists, and construction instructions.
[0013] In this invention, further, in S1, the construction of the knowledge graph includes the following steps:
[0014] S11. Collect high-resolution images of architectural decorative patterns, process description text, and material sample data;
[0015] S12. Perform denoising, correction and segmentation processing on the image, and extract pattern units;
[0016] S13. Perform semantic analysis on the text data to extract craft, culture, and aesthetic tags;
[0017] S14. Construct a ternary graph with patterns as entities and craftsmanship, materials, and cultural semantics as relationships.
[0018] In this invention, further, in S2, feature extraction includes the following steps:
[0019] S21. Use convolutional neural networks to extract multi-scale texture features;
[0020] S22. Obtain the topological relationships of patterns through skeleton extraction and graph structure analysis;
[0021] S23. Based on color space transformation and histograms, analyze and extract color distribution and cultural semantic mapping.
[0022] In this invention, step S3, generating the decorative pattern includes the following steps:
[0023] S31, texture generation sub-network, responsible for generating pattern details;
[0024] S32, the structure generation sub-network, is responsible for pattern skeleton and layout optimization;
[0025] S33, the color generation sub-network, is responsible for color matching and cultural semantic mapping;
[0026] S34. Through joint training and semantic constraints, ensure that the generated patterns have both aesthetic consistency and cultural distinctiveness.
[0027] In this invention, further, in S4, the decoration process and material adaptation includes the following steps:
[0028] S41. Establish a mapping model between pattern features and process parameters;
[0029] S42. Recommend construction techniques based on pattern complexity, number of color layers, and material properties;
[0030] S43. Combine materials databases to match building decoration materials with suitable texture, durability and cost.
[0031] In this invention, further, in S5, the virtual simulation includes the following steps:
[0032] S51. Map the pattern onto the surface of the 3D architectural model;
[0033] S52, based on physically-based rendering technology, simulates material texture and lighting effects;
[0034] S53 supports interactive adjustment of pattern proportions, positions, colors, and process parameters.
[0035] Furthermore, the present invention also includes:
[0036] Users can input their design ideas through voice, text, images, or sketches;
[0037] The system supports personalized recommendations based on historical plans and user feedback;
[0038] It provides cultural semantic interpretation and aesthetic evaluation to assist in design decisions.
[0039] Furthermore, the present invention also provides an AI-based architectural decorative pattern design system, including a knowledge graph construction module, a pattern feature extraction and recognition module, a pattern generation and reconstruction module, a process and material adaptation module, a three-dimensional visualization and simulation module, and a human-computer interaction and output module, for realizing an AI-based architectural decorative pattern design method.
[0040] Furthermore, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an AI-based architectural decorative pattern design method.
[0041] Furthermore, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the program to implement an AI-based architectural decorative pattern design method.
[0042] Compared with the prior art, the beneficial effects of the present invention are:
[0043] 1. This invention achieves intelligent processing of the entire process from pattern extraction and innovative design to process matching by constructing a pattern knowledge graph, integrating generative adversarial networks and process adaptation models.
[0044] 2. This invention deeply integrates artificial intelligence technology with architecture, arts and crafts, and cultural heritage. It not only realizes the intelligent and efficient design of decorative patterns, but more importantly, it builds a digital bridge connecting "historical context", "contemporary aesthetics" and "implementation technology", providing a practical and systematic solution for the inheritance and innovation of architectural decoration culture. Attached Figure Description
[0045] Figure 1 This is a flowchart of the method of the present invention.
[0046] Figure 2 This is a flowchart illustrating the construction of the knowledge graph for this invention.
[0047] Figure 3 This is a flowchart of the feature extraction process of the present invention.
[0048] Figure 4 The flowchart for generating decorative patterns according to the present invention is shown.
[0049] Figure 5 This is a flowchart illustrating the decorative process and material adaptation for this invention.
[0050] Figure 6 This is a flowchart of the virtual simulation process of the present invention. Detailed Implementation
[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0052] In the description of this invention, unless otherwise explicitly specified and limited, the terms "connected," "linked," and "fixed" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0053] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.
[0054] In the description of this invention, the terms "upper," "lower," "right," and "left," etc., refer to the orientation or positional relationship shown in the accompanying drawings. They are used only for ease of description and simplification of operation, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first" and "second" are used merely for distinction in description and have no special meaning.
[0055] Example 1
[0056] Please see Figures 1-6 This embodiment provides the following technical solution: a complete process of an AI-based architectural decorative pattern design method, which specifically includes the following steps:
[0057] 1. Construction of multidimensional knowledge graph.
[0058] The system collects pattern images, historical documents, craft records, and material samples from traditional Chinese architecture (such as palaces, gardens, and residences), local architectural features, and modern decorative examples. The data preprocessing unit performs high-precision noise reduction, geometric correction, and color restoration on the images, and natural language processing on the text to extract craft steps, cultural connotations (e.g., "dragon patterns symbolize imperial power," "the meander pattern symbolizes continuity"), and aesthetic evaluations (e.g., "solemnity," "liveliness"). The knowledge graph generation unit constructs a ternary graph with "pattern entities" as the core, linking "material entities," "craft entities," and "semantic entities." For example, it can form relationships such as ("Forbidden City dragon coffered ceiling pattern," "craftsmanship," "gold leaf gilding"), ("Suzhou garden window ice crack pattern," "implication of semantics," "subtle elegance"), and ("blue brick relief," "suitable materials," "clay-fired blue bricks"), which are stored in the graph database.
[0059] 2. Intelligent pattern extraction and feature analysis.
[0060] Users can upload partial photographs, design sketches, or keyword descriptions of the building to be studied via the system interface. The system accurately separates patterns from complex architectural backgrounds (such as walls, beams, columns, and roof tiles). Subsequently, the feature extraction module is activated.
[0061] Texture and structural feature extraction: A multi-branch convolutional neural network is used. One branch extracts the features of the pattern from microscopic texture (such as wood carving marks and stone texture) to macroscopic layout through convolutional kernels of different sizes. The other branch analyzes the topological connection relationship and structural skeleton of the pattern through skeletonization algorithm and graph convolutional network (GCN).
[0062] Color and semantic feature extraction: The pattern image is converted from RGB to Lab color space, and its main color, secondary color, and color distribution patterns are analyzed. At the same time, a pre-trained BERT model is used in conjunction with a localized cultural corpus to embed the pattern name and descriptive text to generate its cultural semantic feature vector.
[0063] 3. Pattern Reconstruction and Creative Generation. This step is achieved through a hierarchical perceptual generative adversarial network. This network contains three cooperating sub-generators:
[0064] Texture generator: Responsible for generating detailed textures within a given structural contour that conform to the textures of traditional crafts (such as the "smoothness" of lacquerware and the "roughness" of brick carving). Its training employs a multi-scale style loss function based on the Gram matrix to ensure that the statistical characteristics of the generated textures are consistent with the target style.
[0065] Structure Generator: Receives user layout preferences (such as "symmetry," "surround," or "scattered") or learns traditional composition rules (such as "nine-square grid" or "rice-shaped grid") to optimize the layout of the pattern skeleton. Introduces the golden ratio constraint based on the Fibonacci sequence to optimize the spatial distribution and rhythm between elements.
[0066] The color generator retrieves relevant color paradigms from the knowledge graph based on the input cultural semantic instructions (such as "festive" or "solemn"), and generates color schemes that conform to color harmony theory (such as complementary colors or analogous colors). This sub-network is constrained by a color semantic consistency loss function to ensure the matching degree between the color scheme and the cultural intention.
[0067] The outputs of the three sub-networks are fused by a master synthesizer and trained adversarially by a global discriminator to ensure that the final generated patterns are visually harmonious, culturally credible, and stylistically both traditional and innovative.
[0068] 4. Intelligent adaptation of craftsmanship and cultural meaning.
[0069] After the pattern is generated, the adaptation module automatically starts analysis:
[0070] (1) Feature encoding: The visual feature vector of the generated pattern is concatenated with the semantic feature vector to form a comprehensive description vector.
[0071] (2) Process matching: The comprehensive description vector is input into a feature-process mapping model based on a deep neural network. This model has been trained on knowledge graph data and can predict the type of process (such as carving, painting, inlay, casting), process complexity level and key process parameters (such as carving depth and number of painting layers) required to realize the pattern.
[0072] (3) Material Recommendation: The system is linked to the material database and recommends specific materials (such as specific tree species, gypsum types, and metal alloys) based on process requirements, cost budget, durability requirements, and texture performance (such as reflectivity and roughness). The system will output a detailed material list, including specifications, usage estimates, and supplier information for reference.
[0073] (4) Semantic interpretation generation: Based on the pattern characteristics, the system automatically generates a text description to explain the source of its design inspiration, the traditional pattern elements used, the cultural connotations of the colors and composition, and how the appropriate craftsmanship and materials together embody the connotations.
[0074] 5. 3D scene fusion and virtual prototype simulation.
[0075] 3D Model Building: Users can import or create simplified models of target architectural spaces (such as walls, ceilings, and screens). The system provides interactive tools to intelligently map the generated patterns onto the target curved surface and automatically adjust the pattern proportions and shapes according to perspective principles.
[0076] Physical and visual simulation:
[0077] Finite element analysis: For structural decorations (such as reliefs and openwork partitions), the system calls a simplified finite element analysis engine to simulate the stress distribution under gravity and wind loads, and provides early warnings or optimization suggestions for overly fragile structures.
[0078] Physically based rendering: Using the PBR rendering pipeline, appropriate material texture parameters (such as metallicity, roughness, and normal maps) are assigned to the model, and real-time rendering is performed under simulated different lighting environments (natural light, indoor lighting) to intuitively display the final visual effect of the pattern.
[0079] Process animation simulation: It can generate simplified animations of key process steps (such as engraving paths and coating sequences) to help understand the construction process.
[0080] 6. Design output and iterative optimization.
[0081] The system ultimately outputs an integrated project package, including: high-resolution pattern vector graphics, multi-angle renderings, a process instruction manual, a bill of materials, 3D model files (supporting common formats), and a cultural and semantic explanatory document. Furthermore, the system includes a feedback learning module: any adjustments made by the designer to the results (such as changing colors or materials) will be recorded and, after confirmation, used to optimize the knowledge graph and AI model, enabling the system's continuous evolution.
[0082] Example 2
[0083] This embodiment uses a specific example of the modern innovative design of latticework patterns in a modern Chinese tea room to illustrate the specific application of this method:
[0084] 1. Data Input and Intent Understanding: Designers input keywords such as "Jiangnan gardens, latticed windows, modern minimalism, and light and shadow interaction" and upload several reference photos of latticed windows in Suzhou gardens.
[0085] 2. Knowledge Graph Retrieval and Feature Extraction: The system retrieves knowledge graphs and associates typical Jiangnan lattice window patterns such as "ice crack pattern", "begonia pattern" and "bamboo joint pattern", and extracts their spatial philosophical semantics of "transparent but not open" and "changing scenery with each step", as well as their structural features of "delicate", "repetitive" and "geometric".
[0086] 3. Pattern Generation: HP-GAN reconstructs the "ice crack pattern" according to the "modern minimalism" instruction. The structure generator may geometricize the irregular natural ice cracks and regularize them into a more modern combination of broken lines; the color generator abandons the traditional bluish-gray and adopts a monochrome system (such as dark copper); the texture generator simulates the stamping texture of metal plates.
[0087] 4. Process Adaptation: After the system analyzes and generates the pattern, it recommends using the "laser-cut stainless steel plate" process, as it can accurately realize complex geometric shapes, and the material has strong weather resistance and a modern texture. At the same time, the system calculates the appropriate hollowing ratio to ensure rich indoor and outdoor light and shadow effects.
[0088] 5. Virtual simulation: The generated lattice window model is placed into the 3D scene of the tea room to simulate the changes in light and shadow cast on the ground through the lattice window at different times of the day, verifying its design intention of "light and shadow interaction".
[0089] 6. Output: Finally, provide the processing drawings of the stainless steel lattice window, laser cutting path files, detailed installation node drawings, and a design specification, explaining how the design transforms the traditional garden aesthetics into modern architectural language.
[0090] As can be seen from the above embodiments, the method of the present invention deeply integrates artificial intelligence technology with architecture, arts and crafts, and cultural heritage. It not only realizes the intelligent and efficient design of decorative patterns, but more importantly, it builds a digital bridge connecting "historical context", "contemporary aesthetics" and "implementation technology", providing a practical and systematic solution for the inheritance and innovation of architectural decoration culture.
[0091] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. An AI-based method for designing architectural decorative patterns, characterized in that, Includes the following steps: S1. Collect architectural decorative pattern image data and construct a knowledge graph that includes pattern units, craft attributes, cultural semantics and aesthetic features; S2. Using AI image recognition and feature extraction algorithms, extract the texture, structure, color, and topological features of the pattern from the input image; S3. Based on knowledge graphs and feature data, pattern reconstruction and innovative design are carried out through generative adversarial networks to generate decorative patterns that conform to architectural aesthetics and cultural semantics. S4. Based on the pattern characteristics and architectural space attributes, automatically match the appropriate decorative techniques, material textures, and construction parameters; S5. Combine user preferences and spatial scenarios to generate a 3D model of the decoration scheme, and perform virtual simulation and visualization display; S6. Outputs digital design documents including pattern drafts, process parameters, material lists, and construction instructions.
2. The AI-based architectural decorative pattern design method according to claim 1, characterized in that: In step S1, the construction of the knowledge graph includes the following steps: S11. Collect high-resolution images of architectural decorative patterns, process description text, and material sample data; S12. Perform denoising, correction and segmentation processing on the image, and extract pattern units; S13. Perform semantic analysis on the text data to extract craft, culture, and aesthetic tags; S14. Construct a ternary graph with patterns as entities and craftsmanship, materials, and cultural semantics as relationships.
3. The AI-based architectural decorative pattern design method according to claim 1, characterized in that: In step S2, feature extraction includes the following steps: S21. Use convolutional neural networks to extract multi-scale texture features; S22. Obtain the topological relationships of patterns through skeleton extraction and graph structure analysis; S23. Based on color space transformation and histograms, analyze and extract color distribution and cultural semantic mapping.
4. The AI-based architectural decorative pattern design method according to claim 1, characterized in that: In step S3, generating the decorative pattern includes the following steps: S31, texture generation sub-network, responsible for generating pattern details; S32, the structure generation sub-network, is responsible for pattern skeleton and layout optimization; S33, the color generation sub-network, is responsible for color matching and cultural semantic mapping; S34. Through joint training and semantic constraints, ensure that the generated patterns have both aesthetic consistency and cultural distinctiveness.
5. The AI-based architectural decorative pattern design method according to claim 1, characterized in that: In step S4, the adaptation of decorative techniques and materials includes the following steps: S41. Establish a mapping model between pattern features and process parameters; S42. Recommend construction techniques based on pattern complexity, number of color layers, and material properties; S43. Combine materials databases to match building decoration materials with suitable texture, durability and cost.
6. The AI-based architectural decorative pattern design method according to claim 1, characterized in that: In S5, the virtual simulation includes the following steps: S51. Map the pattern onto the surface of the 3D architectural model; S52, based on physically-based rendering technology, simulates material texture and lighting effects; S53 supports interactive adjustment of pattern proportions, positions, colors, and process parameters.
7. The AI-based architectural decorative pattern design method according to claim 1, characterized in that, Also includes: Users can input their design ideas through voice, text, images, or sketches; The system supports personalized recommendations based on historical plans and user feedback; It provides cultural semantic interpretation and aesthetic evaluation to assist in design decisions.
8. An AI-based architectural decorative pattern design system, used to implement the method described in any one of claims 1-7, characterized in that: It includes a knowledge graph construction module, a pattern feature extraction and recognition module, a pattern generation and reconstruction module, a process and material adaptation module, a 3D visualization and simulation module, and a human-computer interaction and output module.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.
10. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-7.