Home decoration plane automatic design method and device

By combining architectural principles and deep learning algorithms, using particle swarm optimization and generative adversarial networks to generate personalized home decoration design solutions, the problem that existing home decoration design tools cannot meet users' personalized needs and budget restrictions is solved, and an intelligent and intuitive design experience is achieved.

CN120296828APending Publication Date: 2025-07-11TIANJIN POLYTECHNIC UNIV
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202510448214.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-04-10
Publication Date
2025-07-11

AI Technical Summary

Technical Problem

Existing home decoration design tools cannot fully consider users' personalized needs and budget limitations. The design process relies on manual adjustments and lacks intelligence. The design plan generation process is not intelligent enough, the user's sense of participation is not strong, and the budget optimization is inaccurate.

Method used

The house space is divided by combining architectural principles and particle swarm optimization algorithms, and a deep learning algorithm is used to generate multiple design solutions. The style transfer is carried out by generating adversarial networks and deep learning algorithms. The budget is optimized by combining reinforcement learning and constraint optimization algorithms, and 3D renderings are generated and virtual roaming experience is provided.

Benefits of technology

Personalized and intelligent home decoration design is realized, ensuring that the design plan meets user needs and budget scope, reducing errors and resource waste in manual design, and improving user participation and intuitiveness of design effects.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120296828A_ABST
    Figure CN120296828A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of home decoration design, and particularly discloses a home decoration plane automatic design method and device, and the method comprises the steps: obtaining home decoration demand data and house basic information inputted by a user, and automatically carrying out the space division of a house room based on an architecture principle and a particle swarm optimization algorithm, and generating a preliminary room layout map; based on the room layout map, analyzing historical design cases and home decoration demand data, and utilizing a deep learning algorithm to generate a plurality of preliminary design schemes; adjusting the generated preliminary design schemes according to style preference input by the user to generate a plurality of optimization design schemes; according to the method, the architecture principle and the particle swarm optimization algorithm are combined, house space division is automatically carried out, the initial room layout diagram is generated, and errors and inconsistency in the manual design process are avoided; historical design cases and home decoration demand data are analyzed by using a deep learning algorithm, a plurality of preliminary design schemes are designed, and the individuation degree and diversity of the design are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of home decoration design, and particularly relates to a method and device for automatically designing home decoration floor plans. Background Art

[0002] With the increasing diversification and personalization of people's requirements for living environments, the home decoration industry has also undergone a transformation from traditional manual design to modern technology-assisted design. Traditional home decoration design often relies on experienced designers for manual design and manual adjustment, with a relatively long design cycle. Moreover, due to the subjectivity of designers, there may be a certain gap between the final design effect and the user's requirements and budget. In addition, it is often difficult to precisely balance aesthetic effects, spatial functions, and construction budgets during the manual design process, resulting in a complex design optimization process and even resource waste.

[0003] In recent years, with the development of computer technology and artificial intelligence, the home decoration industry has gradually transformed towards automation and intelligence. Although some existing home decoration design tools provide functions such as rapid design and effect display, they are mostly limited to specific design stages and fail to fully integrate the user's budget requirements and spatial function requirements. In addition, most existing intelligent home decoration design tools are based on templates, with relatively weak personalization and creativity of design schemes, and often lack effective connection with the later construction links. Currently, many home decoration design platforms on the market provide 3D renderings display, style recommendations, and spatial layout suggestions, but they still have problems such as an insufficiently intelligent design scheme generation process, weak user participation, and inaccurate budget optimization. In order to truly meet the requirements of modern consumers for home design, it is necessary to break the traditional design mode and achieve personalized customization, automated design, and intelligent budget control through intelligent means.

[0004] Therefore, it is necessary to propose a method and device for automatically designing home decoration floor plans to solve the problems in the existing technology, such as the inability to fully consider the user's personalized needs and budget constraints, and the design process relying on manual adjustment and lacking intelligence.

[0005] The above information disclosed in this background art is only used to increase the understanding of the background art of the present invention. Therefore, it may include prior art that is not known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and device for automatically designing home decoration floor plans to solve the problems raised in the above background art.

[0007] To achieve the above purpose, the present invention provides the following technical solutions:

[0008] A method for automatically designing home decoration floor plans includes:

[0009] Obtain the home improvement demand data and basic housing information input by the user, and automatically divide the space of the rooms in the house based on architectural principles and the particle swarm optimization algorithm to generate a preliminary room layout plan;

[0010] Based on the room layout plan, generate multiple preliminary design plans by analyzing historical design cases and home improvement demand data using deep learning algorithms;

[0011] Adjust the multiple generated preliminary design plans according to the style preferences input by the user to generate multiple optimized design plans;

[0012] Calculate the various costs required for each optimized design plan and optimize and adjust according to the budget range provided by the user to obtain the corresponding decoration budget list;

[0013] Based on multiple optimized design plans and the corresponding decoration budget list, use 3D modeling technology to generate 3D renderings and real-time rendering images, and display the decoration effects from different angles through virtual roaming.

[0014] Preferably, the obtaining of the home improvement requirements input by the user through an application or platform includes the number of rooms, functional requirements, style preferences, and budget range to obtain home improvement demand data;

[0015] At the same time, obtain the floor plan of the house uploaded by the user through the same application or platform, or the actual size and structural information of the house input by the user to obtain the basic housing information;

[0016] Based on architectural principles, use the particle swarm optimization algorithm to represent different room areas with different particles, find a harmonious and creative spatial layout combination method between different room areas, and generate a preliminary room layout plan. The update formula of the particle swarm optimization algorithm is as follows:

[0017]

[0018] In the formula, is the velocity of particle i, is the position of the particle, w is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random variables, pbest i and gbest are the personal best position and the global best position of the particle respectively.

[0019] Preferably, based on the room layout plan, automatically match each room according to the home improvement demand data to ensure that the functional planning of each room meets the functional requirements;

[0020] According to the room layout plan and home improvement demand data, use a generative adversarial network to generate multiple preliminary design plans, where:

[0021] The generative adversarial network consists of a generator and a discriminator. Using historical design cases as input data, the generator is trained to learn different design styles and layouts;

[0022] The trained generator is used to generate new design drawings based on the home improvement demand data, and the discriminator is used to evaluate whether the generated design drawings meet the functional requirements;

[0023] During the adversarial training process of the generator and the discriminator, the generated design drawings are continuously optimized, and finally multiple preliminary design schemes are generated;

[0024] The basic loss function of the generative adversarial network is as follows:

[0025] L GAN =E x~pdata [logD(x)]+E z~pz [log(1-D(G(z)))]

[0026] In the formula, D(x) is the probability that the discriminator judges the sample x as real, G(z) is the sample generated by the generator, pdata is the real data distribution, pz is the latent space distribution, and z is the input noise.

[0027] Preferably, based on architectural principles, ergonomics, and home aesthetics, the layout and elements of the preliminary design scheme are automatically adjusted to optimize the space usage effect;

[0028] By analyzing style preferences through big data and combining modern design trends to adjust home improvement style suggestions, the balance between aesthetics and function is achieved;

[0029] According to the home improvement style suggestions, style transfer is performed through deep learning algorithms, and while retaining the original structure of multiple preliminary design schemes, the visual features of the new style are incorporated;

[0030] The image conversion formula for different home improvement styles is as follows:

[0031] L total =αL content +βL style

[0032] In the formula, L content is the content loss, L style is the style loss, and α and β are adjustment parameters;

[0033] According to the selected new style, a color scheme is automatically generated, and considering the lighting in the space, the wall color, and the color matching of the furniture, multiple optimized design schemes are obtained;

[0034] Reinforcement learning can also be applied to regard each preliminary design scheme as the exploration process of an agent in the design space to obtain multiple optimized design schemes.

[0035] Preferably, classify the elements in each optimized design solution and set a unit price for each category;

[0036] Use a regression algorithm to perform price prediction based on the set unit price combined with the market price to obtain a predicted unit price;

[0037] Automatically calculate the various costs in each optimized design solution based on the number of elements and the predicted unit price;

[0038] Use a constrained optimization algorithm to dynamically adjust the cost composition in each optimized design solution within the budget;

[0039] Introduce a simulated annealing algorithm to find the optimal balance point within the cost composition and the budget. The formula is as follows:

[0040] E new = E old + ΔE

[0041]

[0042] In the formula, ΔE is the energy difference, T is the current temperature, and P is the probability;

[0043] Based on the adjusted cost composition, use a template or database management tool to automatically generate a cost report and output the corresponding decoration budget list.

[0044] Preferably, based on multiple optimized design solutions, adjust the elements therein according to the corresponding preset decoration list to obtain multiple candidate design solutions;

[0045] Use 3D modeling software to create and adjust all the elements in each updated candidate design solution to obtain the 3D structures of all the elements;

[0046] Apply the UV mapping algorithm to map textures or materials onto the 3D structures of all the elements to generate 3D renderings of all the elements;

[0047] Render multiple candidate design solutions through a real-time rendering engine to generate corresponding real-time renderings, where:

[0048] Based on each candidate design solution, automatically label the layout and spatial positions of all the elements to generate the labeling information of all the elements;

[0049] Place, adjust, and configure the corresponding 3D renderings in the virtual environment of the real-time rendering engine according to the labeling information of all the elements;

[0050] Set the light sources and shadows in the virtual environment to simulate different natural light and artificial lighting effects;

[0051] The formula for applying ray tracing rendering is as follows:

[0052] L ray = ∫ Ω f(ω j , ω o )L j (ω j ) max(0, cosθ j ) dω j

[0053] where L ray is the light intensity obtained from the observer's perspective, f(ω j , ω o ) is the bidirectional reflectance distribution function, L j (ω j ) is the incident light intensity from direction ω j , and θ j is the angle between the incident light direction and the surface normal;

[0054] The formula for applying global illumination rendering is as follows:

[0055] E j = ρ j ∫ Ω L j (ω j ) cosθ j dω j

[0056] L out = L direct + ∫ Ω f u (ω j , ω o ) L j (ω j ) cosθ j dω j

[0057] where E j is the radiant energy received by surface j, ρ j is the reflectivity of surface j, Ω represents all possible incident directions, f u (ω j , ω0) describes how surface j reflects the light distribution between the incident light direction ω j and the outgoing direction ω o , L out is the light propagating outward from surface j, and L direct is the direct illumination part, directly irradiated by the light source onto surface j;

[0058] Adjust the rendering details according to the functional requirements in each candidate design solution to generate the corresponding real-time rendering image;

[0059] Integrate the VR experience function, and virtualize and display the decoration effects from different angles through VR devices and AR technology, allowing users to experience the design effects in advance.

[0060] Preferably, the method also collects the feedback information and design suggestions of users, automatically provides personalized adjustment suggestions, and automatically generates complete construction drawings according to the final design scheme selected by the user. The construction drawings will be used as the basis for the subsequent construction plan and for data docking with the construction party.

[0061] An automatic home decoration plane design device, comprising:

[0062] An information input module, used to receive the home decoration demand information and basic housing information of users, and perform basic space planning;

[0063] A design generation module, used to generate multiple preliminary design schemes according to the home decoration demand information and basic space planning by using deep learning algorithms;

[0064] An optimization and adjustment module, used to intelligently optimize multiple preliminary design schemes according to the home decoration demand information to generate multiple optimized design schemes;

[0065] A cost calculation module, used to calculate and adjust the various costs required for each optimized design scheme based on the budget range provided by the user, and generate a corresponding decoration budget list;

[0066] An output display module, used to generate 3D effect drawings and real-time rendering drawings based on each optimized design scheme and the corresponding decoration budget list, and display the decoration effects from different angles through virtual roaming.

[0067] Preferably, the output display module further includes a real-time feedback module, used to receive the feedback information and design suggestions of users, and automatically provide personalized adjustment suggestions.

[0068] Compared with the prior art, the beneficial effects of the present invention are:

[0069] The present invention combines architectural principles with the particle swarm optimization algorithm to automatically divide the housing space and generate a preliminary room layout diagram, avoiding errors and inconsistencies in the manual design process; uses deep learning algorithms to analyze historical design cases and home decoration demand data to design multiple preliminary design schemes, improving the personalization and diversity of the design; adjusts multiple schemes according to the user's style preferences to ensure that the design is more in line with the actual needs of users, and at the same time, through optimizing the design scheme, ensures completion within the budget, avoiding the over-budget problem that may occur in traditional designs. In addition, by generating 3D effect drawings and real-time rendering drawings, combined with virtual roaming to display the decoration effects from different angles, users can more intuitively experience the design results, reducing the deviation between the design and the actual decoration. Brief Description of the Drawings

[0070] Figure 1 It is a flowchart of the automatic home decoration plane design method of the present invention;

[0071] Figure 2 It is a framework diagram of the automatic home decoration plane design device of the present invention. Detailed Embodiments

[0072] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0073] Embodiment 1:

[0074] Please refer to Figure 1 As shown, an automatic home decoration plane design method includes:

[0075] Obtain the home decoration demand data and basic housing information input by the user, and automatically divide the space of the housing rooms based on architectural principles and the particle swarm optimization algorithm to generate a preliminary room layout diagram;

[0076] Based on the room layout diagram, generate multiple preliminary design plans by analyzing historical design cases and home decoration demand data using deep learning algorithms;

[0077] Adjust the generated multiple preliminary design plans according to the style preferences input by the user to generate multiple optimized design plans;

[0078] Calculate the various costs required for each optimized design plan, and perform optimized adjustment according to the budget range provided by the user to obtain the corresponding decoration budget list;

[0079] Based on multiple optimized design plans and the corresponding decoration budget list, use 3D modeling technology to generate 3D renderings and real-time rendering images, and display the decoration effects from different angles through virtual roaming.

[0080] Obtain the home decoration demands input by the user through an application or platform, including the number of rooms, functional requirements, style preferences, and budget range, to obtain home decoration demand data;

[0081] At the same time, obtain the floor plan of the house uploaded by the user through the same application or platform, or the actual size and structural information of the house input by the user, to obtain the basic housing information;

[0082] Based on architectural principles, the particle swarm optimization algorithm is used, with different particles representing different room areas, to find harmonious and creative spatial layout combinations between different room areas and generate a preliminary room layout diagram.

[0083] Based on architectural principles and the particle swarm optimization algorithm, the present invention can automatically optimize the division of house space according to the user's home decoration needs and the basic information of the house, thus greatly reducing manual intervention and avoiding the limitations of traditional manual design. The algorithm intelligently searches for the optimal spatial layout method, making the design scheme more in line with the actual living needs and aesthetic standards.

[0084] Based on the room layout diagram, automatically match each room according to the home decoration demand data to ensure that the functional planning of each room meets the functional requirements;

[0085] According to the room layout diagram and the home decoration demand data, use a generative adversarial network to generate multiple preliminary design schemes, where:

[0086] The generative adversarial network consists of a generator and a discriminator. Using historical design cases as input data, train the generator to learn different design styles and layouts;

[0087] Use the trained generator to generate new design drawings according to the home decoration demand data, and use the discriminator to evaluate whether the generated design drawings meet the functional requirements;

[0088] During the adversarial training process of the generator and the discriminator, continuously optimize the generated design drawings, and finally generate multiple preliminary design schemes.

[0089] Compared with traditional design methods, the present invention uses an adversarial network algorithm. By analyzing historical design cases and home decoration demand data, it can automatically generate multiple preliminary design schemes. The application of the deep learning model makes the matching of the preliminary design schemes more accurate in terms of style and function, thus providing users with more personalized choices.

[0090] Based on architectural principles, ergonomics, and home aesthetics, automatically adjust the layout and elements of the preliminary design scheme to optimize the space utilization effect;

[0091] Through big data analysis of style preferences, combine modern design trends to adjust the home decoration style suggestions to achieve the balance between aesthetics and function;

[0092] According to the home decoration style suggestions, perform style transfer through a deep learning algorithm, retaining the original structure of multiple preliminary design schemes while integrating the visual features of the new style;

[0093] Automatically generate a color scheme according to the selected new style, and consider the lighting in the space, the color of the walls, and the color matching of the furniture to obtain multiple optimized design schemes;

[0094] Reinforcement learning can also be applied to treat each preliminary design scheme as an exploration process of an intelligent agent in the design space, and multiple optimized design schemes can be obtained.

[0095] After multiple preliminary design schemes are generated, the present invention uses style transfer combined with modern design trends to adjust and optimize the styles of the preliminary design schemes, balance aesthetics and functions, and improve the overall design quality.

[0096] Classify the elements in each optimized design scheme and set a unit price for each category;

[0097] Use a regression algorithm to perform price prediction according to the set unit price combined with the market price to obtain a predicted unit price;

[0098] Automatically calculate the various expenses in each optimized design scheme according to the number of elements and the predicted unit price;

[0099] Use a constraint optimization algorithm to dynamically adjust the cost composition in each optimized design scheme within the budget;

[0100] Introduce a simulated annealing algorithm to find the optimal balance point within the cost composition and the budget;

[0101] Based on the adjusted cost composition, use a template or a database management tool to automatically generate a cost report and output the corresponding decoration budget list.

[0102] Combined with market price prediction and budget range, automatically control and adjust the cost composition of each optimized design scheme through a constraint optimization algorithm, introduce a simulated annealing algorithm to find the optimal balance point, ensure that the budget is controlled within a reasonable range, and provide a more accurate decoration budget list.

[0103] Based on the room function and space size, automatically recommend suitable furniture sizes and styles. Users can also choose their favorite furniture brands and types, which will be automatically matched. And it is recommended to use environmentally friendly materials and energy-saving furniture to ensure that the design conforms to the principle of sustainability. At the same time, provide an online procurement service for furniture, home appliances, and building materials, and users can directly purchase the required items on one platform.

[0104] Based on multiple optimized design schemes, adjust the elements therein according to the corresponding decoration preset list to obtain multiple candidate design schemes;

[0105] Use 3D modeling software to create and adjust all the elements in each updated candidate design scheme to obtain the 3D structures of all the elements;

[0106] Apply the UV mapping algorithm to map textures or materials onto the 3D structures of all the elements to generate 3D renderings of all the elements;

[0107] Render multiple candidate design solutions through a real-time rendering engine to generate corresponding real-time renderings, where:

[0108] Based on each candidate design solution, automatically annotate the layout and spatial positions of all elements to generate annotation information for all elements;

[0109] Place, adjust, and configure the corresponding 3D renderings in the virtual environment of the real-time rendering engine according to the annotation information of all elements;

[0110] Set the light sources and shadows in the virtual environment to simulate different natural light and artificial lighting effects;

[0111] Adjust the rendering details according to the functional requirements in each candidate design solution to generate the corresponding real-time renderings;

[0112] Integrate the VR experience function, and virtualize and display the decoration effects from different angles through VR devices and AR technology to experience the design effects in advance.

[0113] Based on the optimized design solution, the present invention uses 3D modeling technology and a real-time rendering engine to generate detailed 3D renderings and real-time renderings, and displays the decoration effects from different angles through virtual roaming. In this way, users can intuitively view the decoration effects from multiple angles, extract and experience the final effects of the design solutions, and make real-time adjustments and optimizations.

[0114] Collect the feedback information and design suggestions of users, automatically provide personalized adjustment suggestions, and automatically generate complete construction drawings according to the final design solution selected by the user. The construction drawings will be used as the basis for the subsequent construction plan for data docking with the construction party.

[0115] Embodiment 2:

[0116] Application example: A certain decoration design company conducts the process of home decoration plane design for users.

[0117] (1) The user fills in or uploads the home decoration requirements, including:

[0118] Number of rooms: For example, 3 bedrooms, 1 living room, 1 kitchen, 2 bathrooms, etc.

[0119] Functional requirements: Such as the bedroom needs an independent work area, the living room needs an open design, etc.

[0120] Style preference: The user selects their preferred design style (modern, European, simple, etc.).

[0121] Budget range: The decoration budget range input by the user (for example, 100,000 to 150,000).

[0122] The user uploads the floor plan of the house or manually enters the size and structural information of the house.

[0123] Based on architectural principles and known house data, spatial planning is initially carried out.

[0124] (2) Analyze the user's needs and historical design case data to generate multiple preliminary design schemes.

[0125] The design schemes include:

[0126] The functional division and layout of each room (bedroom, living room, kitchen, etc.).

[0127] The specific size and space utilization rate of each room.

[0128] Suggestions on spatial layout and elements in different styles.

[0129] Automatically generate design drawings that meet the functional requirements through a generative adversarial network, continuously optimize and generate multiple design schemes for the user to choose.

[0130] (3) The user selects their favorite preliminary design scheme, and it is automatically adjusted according to the user's selection.

[0131] Through deep learning and style transfer techniques, automatically provide style suggestions for each preliminary scheme to optimize the aesthetic effect.

[0132] Automatically adjust the color tone combination, furniture layout, lighting design, etc., so that the design better meets the user's needs.

[0133] Further optimize the space utilization and functional layout to ensure that the functional requirements of each room are met.

[0134] (4) Classify the elements in each optimized design scheme and set unit prices (e.g., floor, furniture, wall materials, etc.).

[0135] Use regression algorithms combined with market prices to predict the unit prices in each design scheme.

[0136] Apply constraint optimization algorithms according to the user's budget to adjust the cost composition of the design scheme to ensure that the design scheme does not exceed the budget.

[0137] Automatically generate a decoration budget list, list the costs of each item, and provide detailed cost estimates.

[0138] (5) According to each optimized design scheme, use 3D modeling technology to create and adjust the three-dimensional structures of all elements.

[0139] Map textures or materials onto the three-dimensional structures through the UV mapping algorithm to generate 3D renderings.

[0140] Use a real-time rendering engine for lighting simulation to generate real-time renderings from different angles.

[0141] Roam in the virtual environment to display the design effect.

[0142] Users can conduct virtual tours by combining AR technology with VR devices to experience the decoration effect in advance.

[0143] (6) Users view multiple design plans and 3D renderings and provide feedback, such as suggesting modifications to the color, function, etc. of a certain design.

[0144] Automatically adjust the design plan according to user feedback and display the new modified plan.

[0145] Users can provide feedback again according to the new design adjustments, and automatic optimization adjustments will be made until the final design plan is determined.

[0146] (7) After the user confirms the final design plan, the platform automatically generates detailed construction drawings.

[0147] The construction drawings include the layout plan of each room, electrical circuit diagram, water pipe layout diagram, material specifications, etc.

[0148] Provide the construction drawings to the construction party for data docking and support digital management and collaboration.

[0149] Example 3:

[0150] Please refer to Figure 2 As shown, a home improvement plane automatic design device includes:

[0151] An information input module for receiving the home improvement demand information and basic housing information of users and conducting basic space planning;

[0152] A design generation module for generating multiple preliminary design plans using deep learning algorithms based on the home improvement demand information and basic space planning;

[0153] An optimization adjustment module for intelligently optimizing multiple preliminary design plans according to the home improvement demand information to generate multiple optimized design plans;

[0154] A cost calculation module for calculating and adjusting the various costs required for each optimized design plan based on the budget range provided by the user to generate a corresponding decoration budget list;

[0155] An output display module for generating 3D renderings and real-time renderings using 3D modeling technology based on each optimized design plan and the corresponding decoration budget list, and displaying the decoration effects from different angles through virtual roaming;

[0156] A real-time feedback module for receiving feedback information and design suggestions from users and automatically providing personalized adjustment suggestions.

[0157] An embodiment of the present invention also provides a computer-readable storage medium. A program for implementing the automatic home improvement plane design method described in any one of the above is stored on the computer-readable storage medium. When the program is executed by a processor, it realizes each process of the above automatic design method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described in detail here. Among them, the computer-readable storage medium, such as a read-only memory (Read-Only Memory, abbreviated as ROM), a random access memory (Random Access Memory, abbreviated as RAM), a magnetic disk or an optical disc, etc.

[0158] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0159] In the drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments of the present invention are involved. Other structures can refer to the general design. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other.

[0160] The flowcharts shown in the drawings are only illustrative examples, and do not necessarily include all the contents and operations / steps, nor do they necessarily need to be executed in the described order. For example, some operations / steps can also be decomposed, combined or partially merged, so the actual execution order may change according to the actual situation.

[0161] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An automatic home decoration plane design method, characterized in that, Including: Obtain the home improvement demand data and basic housing information input by the user, and automatically perform spatial division of the rooms in the house based on architectural principles and the particle swarm optimization algorithm to generate a preliminary room layout diagram; Based on the room layout diagram, generate multiple preliminary design schemes by analyzing historical design cases and the home improvement demand data using a deep learning algorithm; Adjust the multiple generated preliminary design schemes according to the style preference input by the user to generate multiple optimized design schemes; Calculate the various costs required for each of the optimized design schemes and perform optimized adjustment according to the budget range provided by the user to obtain the corresponding decoration budget list; Based on the multiple optimized design schemes and the corresponding decoration budget list, use 3D modeling technology to generate 3D renderings and real-time renderings, and display the decoration effects from different angles through virtual roaming.

2. The automatic home decoration plane design method according to claim 1, characterized in that: The obtaining of the home improvement demand data and basic housing information input by the user, and automatically performing spatial division of the rooms in the house based on architectural principles and the particle swarm optimization algorithm to generate a preliminary room layout diagram includes: Obtain the home improvement demands input by the user through an application or platform, including the number of rooms, functional requirements, the style preference, and the budget range, to obtain the home improvement demand data; At the same time, obtain the floor plan of the house uploaded by the user through the same application or platform, or the actual size and structural information of the house input by the user, to obtain the basic housing information; Based on architectural principles, use the particle swarm optimization algorithm with different particles representing different room areas to find a harmonious and creative spatial layout combination method between different room areas, and generate the preliminary room layout diagram. The update formula of the particle swarm optimization algorithm is as follows: In the formula, is the velocity of particle i, is the position of the particle, w is the inertia weight, c1 and c2 are acceleration constants, r1 and r2 are random variables, pbest i and gbest are the personal best position and the global best position of the particle, respectively.

3. The automatic home decoration plane design method according to claim 2, wherein: The generating of multiple preliminary design schemes by analyzing historical design cases and the home improvement demand data using a deep learning algorithm based on the room layout diagram includes: Based on the room layout diagram, automatically match each room according to the home improvement demand data to ensure that the functional planning of each room meets the functional requirements; According to the room layout diagram and the home improvement demand data, use a generative adversarial network to generate multiple preliminary design schemes, where: The generative adversarial network consists of a generator and a discriminator. Use the historical design cases as input data to train the generator to learn different design styles and layouts; Use the trained generator to generate new design drawings according to the home improvement demand data, and use the discriminator to evaluate whether the generated design drawings meet the functional requirements; During the adversarial training process of the generator and the discriminator, continuously optimize the generated design drawings, and finally generate multiple preliminary design schemes; The basic loss function of the generative adversarial network is as follows: L GAN = E x~pdata [logD(x)] + E z~pz [log(1 - D(G(z)))] In the formula, D(x) is the probability that the discriminator judges the sample x as real, G(z) is the sample generated by the generator, pdata is the real data distribution, pz is the latent space distribution, and z is the input noise.

4. The automatic home decoration plane design method according to claim 3, characterized in that: The adjustment of the multiple generated preliminary design schemes according to the style preference input by the user to generate multiple optimized design schemes includes: Based on architectural principles, ergonomics, and home aesthetics, automatically adjust the layout and elements of the preliminary design plan to optimize the space utilization effect; Analyze the style preferences through big data, and adjust the home decoration style suggestions in combination with modern design trends to achieve the balance between aesthetics and functionality; Perform style transfer through deep learning algorithms according to the home decoration style suggestions, while retaining the original structures of multiple preliminary design plans and integrating the visual features of the new style; The image conversion formulas for different home decoration styles are as follows: L total = αL content + βL style where L content is the content loss, and L style is the style loss, and α and β are adjustment parameters; Automatically generate color matching schemes according to the selected new style, and consider the lighting in the space, the color of the walls, and the color matching of the furniture to obtain multiple optimized design plans; Reinforcement learning can also be applied to regard each preliminary design plan as an exploration process of an agent in the design space to obtain multiple optimized design plans.

5. The automatic home decoration plane design method according to claim 4, characterized in that: Calculate the various costs required for each optimized design plan, and perform optimization adjustments according to the budget range provided by the user to obtain the corresponding decoration budget list, including: Classify the elements in each optimized design plan and set a unit price for each category; Use regression algorithms to predict prices based on the set unit prices and market prices to obtain predicted unit prices; Automatically calculate the various costs in each optimized design plan according to the number of elements and the predicted unit prices; Use constraint optimization algorithms to dynamically adjust the cost composition in each optimized design plan within the budget range; Introduce a simulated annealing algorithm to find the optimal balance point within the cost composition and the budget range, and its formula is as follows: E new = E old + ΔE In the formula, ΔE is the energy difference, T is the current temperature, and P is the probability; Based on the adjusted cost composition, use templates or database management tools to automatically generate cost reports and output the corresponding decoration budget list.

6. The automatic home decoration plane design method according to claim 5, wherein: Based on multiple optimized design plans and the corresponding decoration budget list, use 3D modeling technology to generate 3D renderings and real-time renderings, and display the decoration effects from different angles through virtual tours, including: Based on multiple optimized design plans, adjust the elements therein according to the corresponding decoration preset list to obtain multiple candidate design plans; Use 3D modeling software to create and adjust all the elements in each updated candidate design plan to obtain the 3D structures of all the elements; Apply the UV mapping algorithm to map textures or materials onto the 3D structures of all the elements to generate the 3D renderings of all the elements; Render multiple candidate design plans through a real-time rendering engine to generate the corresponding real-time renderings, where: Based on each candidate design plan, automatically label the layout and spatial positions of all the elements to generate the annotation information of all the elements; Place, adjust, and configure the corresponding 3D renderings in the virtual environment of the real-time rendering engine according to the annotation information of all the elements; Set the light sources and shadows in the virtual environment to simulate different natural light and artificial lighting effects; The formula for applying ray tracing rendering is as follows: L ray = ∫ Ω f(ω j , ω o )L j (ω j ) max(0, cosθ j ) dω j where L ray is the light intensity obtained from the observer's perspective, f(ω j , ω o ) is the bidirectional reflectance distribution function, L j (ω j ) is the incident light intensity from direction ω j , and θ j is the angle between the incident direction of the light ray and the surface normal; The formula for applying global illumination rendering is as follows: E j = ρ j ∫ Ω L j (ω j ) cosθ j dω j L out = L direct + ∫ Ω f u (ω j , ω o ) L j (ω j ) cos θ j dω j where E j is the radiative energy received by surface j, ρ j is the reflectivity of surface j, Ω represents all possible incident directions, f u (ω j , ω0) describes how surface j reflects the distribution of light between the incident direction ω j and the outgoing direction ω o , L out is the light propagating outward from surface j, L direct is the direct illumination part, directly irradiated by the light source onto surface j; Adjust the rendering details according to the functional requirements in each of the candidate design solutions to generate the corresponding real-time rendering; Integrate the VR experience function, and virtualize and display the decoration effects from different angles through VR devices and AR technology to experience the design effects in advance.

7. A method for automatically designing a home improvement floor plan according to claim 6, characterized in that, The method also collects the feedback information and design suggestions of the user, automatically provides personalized adjustment suggestions, and automatically generates complete construction drawings according to the final design solution selected by the user. The construction drawings will be used as the basis for the subsequent construction plan for data docking with the construction party.

8. An automatic home decoration plane design device, characterized in that, It includes: An information input module, which is used to receive the user's home decoration requirement information and the basic information of the house, and perform basic space planning; A design generation module, which is used to generate multiple preliminary design solutions using deep learning algorithms according to the home decoration requirement information and the basic space planning; An optimization and adjustment module, which is used to intelligently optimize multiple preliminary design solutions according to the home decoration requirement information to generate multiple optimized design solutions; A cost calculation module, which is used to calculate and adjust the various costs required for each optimized design solution based on the budget range provided by the user to generate the corresponding decoration budget list; An output display module, which is used to generate 3D renderings and real-time renderings using 3D modeling technology based on each optimized design solution and the corresponding decoration budget list, and display the decoration effects from different angles through virtual roaming.

9. The automatic home decoration plane design device according to claim 8, wherein: The output display module also includes a real-time feedback module, which is used to receive the feedback information and design suggestions of the user and automatically provide personalized adjustment suggestions.

Citation Information

Cited By

  • Furniture layout space optimization design method and system based on artificial intelligence

    CN120951573A

  • Indoor layout semantic constraint generation method based on multi-modal alignment

    CN121234302A