Urban design scheme image generation method and device

Abstract

The application provides a kind of urban design scheme image generation method and device, with such characteristics, including step S1, according to existing city planning image construction training data set;Step S2, construct planning parameter prediction model and design generation model, and according to training data set respectively training, obtain trained planning parameter prediction model and trained design generation model;Step S3, input specified region image into trained planning parameter prediction model, obtain corresponding predicted planning parameter;Step S4, input predicted planning parameter and specified region image into trained design generation model, obtain design scheme image.In short, the method can generate effective and have specified features and style design scheme image according to specified image.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of urban planning, and particularly relates to a method and device for generating an urban design scheme image. BACKGROUND

[0002] Traditional urban design mainly relies on manual completion of spatial form design. With the development of computer-aided design technology, since 1980, various computer-aided form generation tools have appeared, such as parametric urban design and urban form simulation tools, which have improved the efficiency and scientificity of design generation in different links. However, these classic urban spatial form generation tools mainly rely on artificial setting of form rules to realize automatic modeling of three-dimensional space, and these rules are difficult to completely reflect the logic of urban spatial form, so they are weak in generating urban spatial form, especially in generating block-scale spatial form, which is a more precise and complex task.

[0003] The development of new data environment and artificial intelligence technology brings new perspectives to block-scale form generation. Compared with macro-scale research, block-scale form research has the characteristics of small scale, diversity and discretization, not only focusing on the subtle relationship of urban spatial form, but also reflecting the characteristics and style of design, which brings challenges to the generation model based on preset rules. Artificial intelligence methods based on artificial neural networks, represented by machine learning and deep learning, can effectively identify hidden features in structured data or image data, thus making up for this deficiency. These models usually use pre-training to extract data features from a large number of samples and store them for prediction on new data sets.

[0004] This method can provide fast and personalized scheme suggestions, but there are still problems such as complex training process, weak controllability and difficulty in explanation in practice. Therefore, there is an urgent need for a method that can effectively generate urban design scheme images. SUMMARY

[0005] The present application is made to solve the above problems, and aims to provide a method and device for generating an urban design scheme image.

[0006] The application provides a city design scheme image generation method for generating a corresponding design scheme image according to a specified area image, and has the characteristics that the method comprises the following steps: step S1, constructing a training data set according to an existing city planning image; step S2, constructing a planning parameter prediction model and a design generation model, and training the planning parameter prediction model and the design generation model according to the training data set to obtain a trained planning parameter prediction model and a trained design generation model; step S3, inputting the specified area image into the trained planning parameter prediction model to obtain corresponding predicted planning parameters; and step S4, inputting the predicted planning parameters and the specified area image into the trained design generation model to obtain a design scheme image.

[0007] In the city design scheme image generation method provided by the application, the predicted planning parameters can be parameters related to city design space.

[0008] In the city design scheme image generation method provided by the application, the trained planning parameter prediction model can be provided with the range and distribution characteristics of the planning parameters, and the parameters of all training data in the training data set of the same type as the specified area image are extracted to obtain the range and distribution characteristics of the planning parameters.

[0009] In the city design scheme image generation method provided by the application, the trained design generation model can generate an optimized image according to the specified area image, and adjust the parameters and layout of the buildings in the optimized image according to the predicted planning parameters to obtain the design scheme image.

[0010] In the city design scheme image generation method provided by the application, the existing city planning image can comprise satellite images and aerial images, and step S1 can comprise the following sub-steps: step S1-1, classifying all existing city planning images according to city design space element types to obtain a plurality of sub-classes; step S1-2, slicing each existing city planning image in each sub-class according to a preset cropping frame to obtain a plurality of first samples; step S1-3, respectively shifting the corresponding preset cropping frame of each first sample up, down, left and right by a preset pixel distance to obtain four corresponding second samples; and step S1-4, preprocessing each first sample and second sample to obtain the training data set.

[0011] This invention also provides an urban design scheme image generation device for generating corresponding design scheme images based on images of a specified area. The device comprises: a training data construction module for constructing a training dataset based on existing urban planning images; a model generation module for constructing a planning parameter prediction model and a design generation model, and training them respectively based on the training dataset to obtain trained planning parameter prediction models and trained design generation models; a parameter prediction module for inputting images of the specified area into the trained planning parameter prediction model to obtain corresponding predicted planning parameters; and a scheme generation module for inputting the predicted planning parameters and images of the specified area into the trained design generation model to obtain design scheme images.

[0012] The role and effect of invention

[0013] According to the urban design scheme image generation method and apparatus of the present invention, on the one hand, a planning parameter prediction model is trained using a training dataset, enabling the model to learn the range and distribution of predicted planning parameters for various types of buildings in existing urban planning images. This allows the model to generate predicted planning parameters for a specified area image that conform to the characteristics of the training data. On the other hand, by designing a generation model to learn from existing urban planning images in the training dataset and performing pixel-by-pixel learning, layout features and other characteristics are obtained. This optimizes the layout planning in the specified area image, and then, based on the predicted planning parameters, the parameters for each building in the optimized image are generated, thus obtaining the design scheme image. Therefore, the urban design scheme image generation method and apparatus of the present invention can generate effective design scheme images with specified characteristics and styles based on a specified image. Attached Figure Description

[0014] Figure 1 This is a block diagram of the urban design scheme image generation device in an embodiment of the present invention;

[0015] Figure 2 This is a schematic diagram of the process of constructing the training dataset in an embodiment of the present invention;

[0016] Figure 3 This is a flowchart illustrating the method for generating images of urban design schemes in an embodiment of the present invention. Detailed Implementation

[0017] To make the technical means, creative features, objectives and effects of the present invention easy to understand, the following embodiments, in conjunction with the accompanying drawings, will specifically illustrate the urban design scheme image generation method and apparatus of the present invention.

[0018] This embodiment provides a device for generating urban design scheme images, used to generate corresponding design scheme images based on images of a specified region.

[0019] Figure 1 This is a block diagram of an urban design scheme image generation device in an embodiment of the present invention.

[0020] like Figure 1 As shown, the urban design scheme image generation device 100 includes a training data construction module 10, a model generation module 20, a parameter prediction module 30, a scheme generation module 40, and a control module 50 that controls the operation of the above modules.

[0021] Training data construction module 10 is used to construct a training dataset based on existing urban planning images.

[0022] Figure 2 This is a schematic diagram of the process of constructing a training dataset in an embodiment of the present invention.

[0023] like Figure 2 As shown, the training data construction module 10 constructs a training dataset based on existing urban planning images, including the following steps:

[0024] Step S1-1: Classify all existing urban planning images according to the type of urban design spatial elements to obtain multiple subcategories.

[0025] Steps S1-2 involve slicing each existing urban planning image within a subclass according to a preset cropping frame to obtain multiple first samples. In this embodiment, the preset cropping frame size is 800m*800m.

[0026] Steps S1-3: For each first sample, shift the corresponding preset cropping box up, down, left, and right by a preset pixel distance to capture the corresponding four second samples.

[0027] Steps S1-4 involve preprocessing each first and second sample to obtain the training dataset. In this embodiment, preprocessing sets the non-building-covered areas in the samples to white and sets the corresponding building-covered areas to grayscale based on the building height, thus obtaining a grayscale image as training data.

[0028] The model generation module 20 is used to construct a planning parameter prediction model and a design generation model, and trains them separately based on the training dataset to obtain the trained planning parameter prediction model and the trained design generation model. In this embodiment, both the planning parameter prediction model and the design generation model are relatively lightweight models, which can be obtained by adjusting existing lightweight models such as linear models, decision trees, support vector machines, and lightweight neural networks. During the training process in this embodiment, the planning parameter prediction model and the design generation model are optimized through methods such as parameter pruning, quantization, and model distillation, thereby reducing the model size and computational requirements while maintaining the model performance as much as possible.

[0029] The parameter prediction module 30 is used to input the image of the specified area into the trained planning parameter prediction model to obtain the corresponding predicted planning parameters.

[0030] The trained planning parameter prediction model includes settings for the range and distribution characteristics of planning parameters. These ranges and characteristics are obtained by extracting parameters from all training data of the same image type as the specified region within the training dataset. The predicted planning parameters are spatially relevant to urban design; in this embodiment, they include waterfront landscape openness, open space ratio, development intensity, and average building height.

[0031] The scheme generation module 40 is used to input the predicted planning parameters and the image of the specified area into the trained design generation model to obtain the design scheme image.

[0032] The trained design generation model generates optimized images based on images of a specified region, and adjusts the parameters and layout of buildings in the optimized images according to predicted planning parameters to obtain the design scheme image.

[0033] The control module 50 stores the control program that controls the operation of each module.

[0034] The process of generating urban design scheme images using the urban design scheme image generation device 100 will be described below with reference to the accompanying drawings.

[0035] Figure 3 This is a flowchart illustrating the method for generating images of urban design schemes in an embodiment of the present invention.

[0036] like Figure 3 As shown, the method for generating images of urban design schemes includes the following steps:

[0037] Step S1: Use training data construction module 10 to construct a training dataset based on existing urban planning images.

[0038] Step S2: The planning parameter prediction model and the design generation model are constructed using the model generation module 20, and trained respectively based on the training dataset to obtain the trained planning parameter prediction model and the trained design generation model.

[0039] Step S3: The parameter prediction module 30 inputs the image of the specified area into the trained planning parameter prediction model to obtain the corresponding predicted planning parameters.

[0040] Step S4: The scheme generation module 40 inputs the predicted planning parameters and the image of the specified area into the trained design generation model to obtain the design scheme image.

[0041] In this embodiment, an image of a waterfront area of ​​a city is used as the designated area image, and an image of a city with excellent landscape design is selected as the existing urban planning image. A corresponding design scheme image is generated using the urban design scheme image generation device 100. Compared to the original image, this design scheme image adds multiple waterfront open spaces and corridors extending from the shoreline into the hinterland. Therefore, the design scheme image can optimize the design of existing areas and enhance their landscape value. In other embodiments, by selecting suitable existing urban planning images to construct a training dataset, design scheme images with corresponding characteristics can be generated.

[0042] The role and effect of the embodiments

[0043] According to the urban design scheme image generation method and apparatus involved in this embodiment, on the one hand, a planning parameter prediction model is trained using a training dataset, enabling the model to learn the range and distribution of predicted planning parameters for various types of buildings in existing urban planning images. This allows the trained model to generate predicted planning parameters for a specified area image that conform to the characteristics of the training data. On the other hand, by designing a generation model to learn from existing urban planning images in the training dataset and performing pixel-by-pixel learning, layout features and other characteristics are obtained. This optimizes the layout planning in the specified area image, and then, based on the predicted planning parameters, the parameters for each building in the optimized image are generated, thus obtaining the design scheme image. In summary, this method can generate effective design scheme images with specified characteristics and styles based on a given image.

[0044] Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to this invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A city design scheme image generation method for generating a corresponding design scheme image according to a specified area image, characterized in that, The method comprises the following steps: Step S1, constructing a training data set according to existing urban planning images; Wherein, the existing urban planning images include satellite images and aerial images, The step S1 includes the following sub-steps: Step S1-1, classifying all the existing urban planning images according to the city design space element type to obtain multiple sub-classes; Step S1-2, for each of the sub-classes, slicing each of the existing urban planning images in the sub-class according to a preset intercepting frame to obtain multiple first samples; Step S1-3, for each of the first samples, respectively shifting the corresponding preset intercepting frame up, down, left and right by a preset pixel distance to obtain four corresponding second samples; Step S1-4, preprocessing each of the first samples and the second samples to obtain the training data set; Step S2, constructing a planning parameter prediction model and a design generation model, and training them respectively according to the training data set to obtain a trained planning parameter prediction model and a trained design generation model; Step S3, inputting the specified area image into the trained planning parameter prediction model to obtain the corresponding predicted planning parameter; Step S4, inputting the predicted planning parameter and the specified area image into the trained design generation model to obtain the design scheme image; Wherein, the trained design generation model generates an optimized image according to the specified area image, and adjusts the parameters and layout of the buildings in the optimized image according to the predicted planning parameter to obtain the design scheme image.

2. The urban design scheme image generation method according to claim 1, wherein: wherein The predicted planning parameter is a parameter related to urban design space.

3. The urban design scheme image generation method according to claim 1, wherein: wherein The trained planning parameter prediction model is provided with the range and distribution characteristics of the planning parameter, The parameters of all training data in the training data set that are of the same type as the specified area image are extracted to obtain the range and distribution characteristics of the planning parameter.

4. The urban design scheme image generation method according to claim 1, wherein: wherein The trained design generation model generates an optimized image according to the specified area image, and adjusts the parameters and layout of the buildings in the optimized image according to the predicted planning parameter to obtain the design scheme image.

5. An urban design plan image generation device for generating a corresponding design plan image from a designated area image, characterized by comprising: It comprises: A training data construction module for constructing a training data set according to existing urban planning images; wherein the existing urban planning images include satellite images and aerial images, The training data construction module includes the following units: A classification unit for classifying all the existing urban planning images according to the city design space element type to obtain multiple sub-classes; A slicing unit for slicing each of the existing urban planning images in each of the sub-classes according to a preset intercepting frame to obtain multiple first samples; An intercepting unit for shifting the corresponding preset intercepting frame of each of the first samples up, down, left and right by a preset pixel distance to obtain four corresponding second samples; The pre-processing unit pre-processes each of the first sample and the second sample to obtain the training data set; The model generation module is configured to construct a planning parameter prediction model and a design generation model, and train the planning parameter prediction model and the design generation model according to the training data set to obtain a trained planning parameter prediction model and a trained design generation model; The parameter prediction module is configured to input the specified region image into the trained planning parameter prediction model to obtain corresponding predicted planning parameters; The scheme generation module is configured to input the predicted planning parameters and the specified region image into the trained design generation model to obtain the design scheme image. The trained design generation model generates an optimized image according to the specified region image, and adjusts parameters and layout of buildings in the optimized image according to the predicted planning parameters to obtain the design scheme image.

Images