A bridge design scheme generation method and system

Abstract

The embodiment of the present invention provides a bridge design scheme generation method and system, which belongs to the technical field of bridge design. The method includes: obtaining an initial bridge description text corresponding to a target object, and using the initial bridge description text to generate an image to obtain an initial bridge image corresponding to the target object; collecting a target comment text expressed by the target object on the initial bridge image, and adjusting the initial bridge image according to the target comment text to obtain a target bridge image; performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image; performing parameter extraction on each sub-edge information in the target edge information in the target bridge image to obtain a target drawing parameter corresponding to the sub-edge information; and obtaining a bridge design scheme corresponding to the target object in the auxiliary design tool according to the association between the target drawing parameter and the auxiliary design tool.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge design, and in particular to a method and system for generating a bridge design scheme. Background Art

[0002] In bridge design, the use of computer-aided design tools is crucial. Computer-aided design tools can not only help designers clearly present the bridge design effects, but also significantly improve the speed and efficiency of drawing. However, when designers are not familiar with computer-aided design tools, they often face some challenges, such as being unable to quickly transform design ideas into actual design drawings. In this case, the efficiency of bridge design will be seriously affected, and designers need to spend more time and energy to overcome these difficulties. Therefore, there is an urgent need for a method to assist designers in using computer-aided design tools more efficiently to generate bridge designs. Summary of the invention

[0003] The main purpose of the embodiments of the present invention is to provide a bridge design scheme generation method and system, aiming to solve the problem in the related art that when designers are not familiar with computer-aided design tools, they cannot quickly convert design ideas into actual design drawings, thereby reducing the efficiency of bridge design.

[0004] In a first aspect, an embodiment of the present invention provides a method for generating a bridge design scheme, comprising:

[0005] Obtaining an initial bridge description text corresponding to the target object, and using the initial bridge description text to perform image generation to obtain an initial bridge image corresponding to the target object;

[0006] collecting a target comment text expressed by the target object on the initial bridge image, and adjusting the initial bridge image according to the target comment text to obtain a target bridge image;

[0007] Performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image;

[0008] Extracting parameters of each sub-edge information in the target edge information in the target bridge image to obtain target drawing parameters corresponding to the sub-edge information;

[0009] According to the association relationship between the target drawing parameters and the auxiliary design tool, a bridge design scheme corresponding to the target object in the auxiliary design tool is obtained.

[0010] In a second aspect, an embodiment of the present invention provides a bridge design scheme generation system, including:

[0011] An image generation module, used to obtain an initial bridge description text corresponding to a target object, and use the initial bridge description text to perform image generation to obtain an initial bridge image corresponding to the target object;

[0012] An image adjustment module, used for collecting a target comment text expressed by the target object on the initial bridge image, and adjusting the initial bridge image according to the target comment text to obtain a target bridge image;

[0013] An edge detection module, used for performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image;

[0014] A parameter extraction module, used for extracting parameters of each sub-edge information in the target edge information in the target bridge image, and obtaining target drawing parameters corresponding to the sub-edge information;

[0015] The scheme generating module is used to obtain the bridge design scheme corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameters and the auxiliary design tool.

[0016] The embodiment of the present invention provides a bridge design scheme generation method and system, the method comprising: obtaining an initial bridge description text corresponding to a target object, using the initial bridge description text to generate an image to generate a corresponding initial bridge image, thereby ensuring that the initial stage of the design can be based on an accurate description and reflect the design intention of the target object. Next, collecting the feedback of the target object on the initial bridge image to form a target comment text, and then modifying and adjusting the initial bridge image accordingly according to the target comment text, and finally obtaining the target bridge image. The key to this stage is to be able to refine the design through specific feedback so that it is more in line with the actual needs and aesthetic preferences of the target object, thereby laying a solid foundation for subsequent design scheme improvements. Subsequently, edge detection is performed on the target bridge image to extract the target edge information in the image. Then, for each sub-edge information in the target edge information, further analysis and parameter extraction are performed to generate the corresponding target drawing parameters. These parameters are an important basis for the subsequent design scheme drawing, ensuring that the specific details and requirements of the design can be accurately expressed and implemented. Finally, according to the relationship between the target drawing parameters and the auxiliary design tool, the corresponding bridge design scheme is generated. This method solves the problem that designers were unable to quickly transform their design ideas into actual design drawings due to unfamiliarity with computer-aided design tools. This method enables designers to use computer-aided design tools more efficiently, thereby significantly improving the efficiency and quality of bridge design, ensuring that they can quickly and effectively realize their own ideas and concepts in complex design tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 A schematic diagram of a flow chart of a bridge design scheme generation method provided by an embodiment of the present invention;

[0019] Figure 2 A schematic diagram of the module structure of a bridge design scheme generation system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0022] It should be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0023] The embodiment of the present invention provides a bridge design scheme generation method and system. The bridge design scheme generation method can be applied to a terminal device, which can be an electronic device such as a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device. The terminal device can be a server or a server cluster.

[0024] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0025] Please refer to Figure 1 , Figure 1 A schematic flow chart of a bridge design scheme generating method provided in an embodiment of the present invention.

[0026] like Figure 1 As shown, the bridge design scheme generating method includes steps S101 to S105.

[0027] Step S101: obtaining an initial bridge description text corresponding to a target object, and using the initial bridge description text to perform image generation to obtain an initial bridge image corresponding to the target object.

[0028] For example, the target object is a designer or a bridge design user, so the initial bridge description text corresponding to the target object is obtained, and the initial bridge description text is used to fully and accurately express the specific needs of the target object for bridge design. It not only includes the basic requirements for the appearance of the bridge, but also covers multiple aspects such as structural function, design style, and structural requirements.

[0029] For example, key bridge design elements are extracted from the initial bridge description text, such as the type of bridge (such as suspension bridge, arch bridge, etc.), size, material, design style, etc. Based on the extracted information, design specifications for image generation are formulated. These specifications include the appearance requirements, structural characteristics, proportions and dimensions of the bridge, etc., to ensure that the generated image can accurately reflect the target requirements. The design specifications are then input into the image generation model to obtain the initial bridge image corresponding to the target object.

[0030] For example, the design specifications include the type of bridge (such as arch bridge, suspension bridge), size (length, width, height), material (such as steel, concrete), design style (modern, classical), etc. The design specifications are prepared into an input format acceptable to the image generation model (such as the generative adversarial network GAN). For example, for a text description model, the design specifications may need to be organized into a detailed description; for a parametric model, the data may need to be converted into a specific set of parameters. The organized design specifications are then input into the image generation model to obtain the initial bridge image corresponding to the initial bridge description text.

[0031] In some embodiments, the method of using the initial bridge description text to generate an image to obtain an initial bridge image corresponding to the target object includes: semantically representing the initial bridge description text according to a text encoding layer of an image generation model to obtain a first text vector; keyword representing the initial bridge description text according to a keyword encoding layer of the image generation model to obtain a second text vector; vector fusion of the first text vector with a random noise vector according to a conditional enhancement layer of the image generation model to obtain a target splicing vector; extracting image features using the target splicing vector according to an image feature generation layer of the image generation model to obtain initial image features; generating an image using the initial image features according to a convolutional network layer of the image generation model to obtain a rough bridge image; calculating the correlation value between the second text vector and each sub-region in the rough bridge image according to a correlation calculation layer of the image generation model; and updating the rough bridge image using the correlation value according to an image update layer of the image generation model to obtain the initial bridge image corresponding to the target object.

[0032] Exemplarily, the image generation model includes a text encoding layer, a keyword encoding layer, a conditional enhancement layer, an image feature generation layer, a convolutional network layer, a correlation calculation layer, and an image update layer.

[0033] Exemplarily, the initial bridge description text is input into the text encoding layer in the image generation model to convert the initial bridge description text into a vector representation that can capture its semantic content to obtain a first text vector. The first text vector contains the semantic information in the initial bridge description text, providing a basis for the subsequent image generation process.

[0034] Exemplarily, a keyword encoding layer is used in the image generation model to extract important keywords from the initial bridge description text and encode them. After being processed by the keyword encoding layer, a second text vector is generated. The second text vector reflects the keyword information in the initial bridge description text, which helps to further refine the content of the image generation.

[0035] Exemplarily, the conditional enhancement layer uses a random noise vector to introduce randomness and diversity, and performs vector fusion of the first text vector and a random noise vector to obtain a target splicing vector. The conditional enhancement layer makes the generated image not only conform to the initial bridge description text, but also has a certain diversity. The target splicing vector is used to guide the image generation model to generate richer image features.

[0036] Exemplarily, the image feature generation layer extracts image features from the target splicing vector, and then maps the target splicing vector to the image feature space to generate initial image features, which lay the foundation for the subsequent steps of generating the image. Based on the initial image features extracted from the image feature generation layer, the convolutional network layer gradually converts the initial image features into specific image content through its multi-layer convolution operation to obtain a rough bridge image. The rough bridge image can show the preliminary bridge design, but it may not be perfect in details and accuracy. The correlation calculation layer uses spatial and text image similarity calculation methods to evaluate the correlation value between each sub-region in the rough bridge image and the second text vector. The correlation value is used to describe the degree of match between the features of each sub-region and the keyword vector. The correlation value is used to analyze the degree of conformity between the rough bridge image and the initial bridge description text, providing a basis for subsequent image adjustment.

[0037] Exemplarily, according to the calculated correlation value, the image update layer adjusts and optimizes the rough bridge image. This layer uses the correlation value to correct the parts of the image that do not conform to the text description, ensuring that the image details are consistent with the text requirements. Thus, the image update layer makes careful adjustments so that the final generated image more accurately meets the requirements of the initial bridge description text, and then after the optimization processing of the image update layer, the generated initial bridge image should meet the detail requirements of the initial description text.

[0038] Specifically, the image generation model starts from the initial bridge description text, and after multiple stages such as feature extraction, rough image generation, correlation value calculation and image updating, it finally generates an initial bridge image that meets the target requirements, ensuring that the initial bridge image not only meets the functional and structural requirements of the initial bridge description text, but also presents an ideal visual effect.

[0039] Step S102: collecting target comment text expressed by the target object on the initial bridge image, and adjusting the initial bridge image according to the target comment text to obtain a target bridge image.

[0040] Exemplarily, after obtaining the initial bridge image, the initial bridge image is displayed to the target object through the terminal device, so that the target object can display the initial bridge image through the terminal device, and then after the target object views the initial bridge image, the target comment text expressed by the target object on the initial bridge image is obtained. The target comment text is used to represent the description text of the content that needs to be improved on the initial bridge image by the target object. The target comment text includes but is not limited to pointing out that some parts of the bridge need to be redesigned, or some details need to be enhanced, modifying some image details, adjusting the structural proportions, modifying the color scheme, etc.

[0041] For example, the target review text includes recommendations for structural changes to adjust the design of the bridge. For example, redesigning the bridge's support structure, changing the shape of the bridge, or adding design details. The target review text includes color and texture requests to modify the appearance of the bridge. This may involve changing the color scheme, adding texture effects, or adjusting the appearance of materials. Enhance the details in the initial bridge image to ensure that all details meet the requirements of the target review text. This may include refining the appearance of certain parts, adding decorative elements, or correcting flaws in the image.

[0042] Exemplarily, first, natural language processing technology is used to parse the target comment text to extract key information and adjustment requirements. These requirements may involve aspects such as the shape, color, and material of the bridge. The initial bridge image is feature extracted through a convolutional neural network to obtain features such as the shape, color distribution, and texture of the image. Then, the extracted features are mapped to the requirements in the comment text to determine the specific image features that need to be modified or optimized. A neural network adjustment model is designed that can automatically adjust image features according to the requirements of the target comment text. The neural network adjustment model includes a text understanding module and a feature adjustment module, the former is responsible for parsing the requirements in the comment text, and the latter is responsible for applying the requirements to image feature adjustment. Thus, a generative model such as a generative adversarial network or a variational autoencoder is used to generate a target bridge image based on the adjusted features. Thus, automatic adjustment of the initial bridge image according to the target comment text is achieved based on a neural network.

[0043] In some embodiments, adjusting the initial bridge image according to the target comment text to obtain the target bridge image includes: performing keyword relationship recognition on the target comment text to obtain a target association relationship corresponding to the target comment text; performing data representation on the target association relationship according to the information encoding layer of the image enhancement model to obtain a third text vector; obtaining an image feature vector corresponding to the target association relationship from the initial bridge image using the target association relationship according to the feature extraction layer of the image enhancement model; generating a target contour corresponding to the target comment text using the third text vector and the image feature vector according to the mask generation layer of the image enhancement model; and adjusting the initial bridge image using the target contour according to the image processing layer of the image enhancement model to obtain the target bridge image.

[0044] For example, the target review text is preprocessed by cleaning, word segmentation, part-of-speech tagging, etc. Keywords are extracted from the target review text using a keyword extraction model (such as TF-IDF, Word2Vec, BERT, etc.). Relationship extraction models (such as R-CNN, Bi-LSTM, etc.) are then used to identify the relationship between keywords and obtain the target association relationship.

[0045] Exemplarily, the information encoding layer of the image enhancement model is usually a pre-trained text embedding model, such as BERT or Word2Vec, which then encodes the target comment text into a vector and converts the extracted keywords and target associations into a third text vector. The feature extraction layer of the image enhancement model usually includes one or more convolutional layers to extract features such as color, texture, shape, and edges from the initial bridge image. These features can be used as the basis for subsequent image understanding and adjustment. The feature extraction layer allows the image enhancement model to capture important visual features from the initial bridge image.

[0046] For example, the mask generation layer of the image enhancement model usually contains one or more fully connected layers, which can receive vectors from the information encoding layer and the feature extraction layer and use this information to generate a mask. This mask can guide the pixel operation during the image enhancement process. The main task of the mask generation layer is to generate a guiding layer that defines the area to be adjusted in the initial bridge image. It guides the model to perform operations in a specific area by fusing text information and image features.

[0047] Exemplarily, the image processing layer of the image enhancement model typically includes multiple convolutional layers and possibly one or more deconvolutional layers for adjusting the initial bridge image. In addition to basic convolution operations, normalization, activation functions, and pooling operations may also be included. The initial bridge image is thereby adjusted to ultimately generate a target bridge image that meets the requirements of the target review text. The purpose of the image processing layer is to adjust the initial bridge image using a mask. Through this layer, specific parts of the initial bridge image can be modified while retaining the original image features of other areas, thereby achieving a target bridge image that meets the target review text.

[0048] Specifically, the structure of the image enhancement model is a carefully designed hierarchical system, in which each layer has its specific role and works together to achieve efficient and accurate adjustment from text to image. It can convert the requirements of the target review text into specific image operations and finally generate a target bridge image that meets the specific requirements.

[0049] Step S103: performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image.

[0050] Exemplarily, before edge detection, a series of preprocessing steps are required for the target bridge image to ensure the accuracy and effectiveness of edge detection. First, the image is grayed to convert the color image into a grayscale image, which can reduce the complexity of the data and make subsequent processing more efficient. Graying can focus on the structural features of the image by retaining the brightness information of the image and removing the color information. Next, noise removal is required. The noise in the image may affect the effect of edge detection, so filtering techniques, such as Gaussian filtering or median filtering, should be used to reduce random noise in the image. This step can smooth the image, reduce unnecessary details, and make the actual edge features more prominent.

[0051] Exemplarily, after preprocessing, an edge detection algorithm is used to extract the edge of the target bridge image. Common edge detection algorithms include, but are not limited to, the Sobel operator and the Canny edge detection algorithm. For example, the Canny edge detection algorithm first calculates the gradient of the preprocessed image, then applies non-maximum suppression to accurately determine the edge position, and finally identifies the real edge through dual threshold detection and edge connection to obtain the target edge information corresponding to the target bridge image.

[0052] In some embodiments, the edge detection of the target bridge image to obtain target edge information corresponding to the target bridge image includes: performing grayscale conversion on the target bridge image to obtain a corresponding grayscale bridge image; determining a target window, and sliding the target window on the grayscale bridge image to obtain a window bridge image corresponding to the target window; obtaining a similar image corresponding to the window bridge image, and filtering the window bridge image according to the similar image and a filter to obtain a first bridge image corresponding to the window bridge image; combining the first bridge image according to the position information corresponding to the target window to obtain a second bridge image; and performing edge detection on the second bridge image according to threshold segmentation to obtain the target edge information corresponding to the target bridge image.

[0053] Exemplarily, the target bridge image is converted from a color mode to a grayscale mode, for example, by calculating the brightness value of each pixel in the target bridge image, thereby generating a grayscale image with only brightness information.

[0054] Exemplarily, a target window is selected, and the size and position of the target window are determined according to specific requirements or preset standards. The target window is slid on the grayscale bridge image. Each sliding position generates a corresponding window bridge image. By sliding the target window on the grayscale bridge image, the window bridge image of the local area can be extracted for subsequent processing.

[0055] Exemplarily, features for comparing image change trends are determined. Common features. For the window bridge image and each image in the database, features of the image change trend are extracted, and the features of the image change trend include the gradient, texture features, edge features, etc. of the image. The extracted features are then constructed as feature vectors or feature descriptors. Feature vectors quantify image features into points in a multidimensional space, and then use similarity metrics including Euclidean distance, cosine similarity, Manhattan distance, etc. to quantify the similarity between two feature vectors. Images in the database are then sorted according to the calculated similarity scores. The higher the similarity score, the closer the image is to the window bridge image in terms of change trends. The first image with a similar change trend to the window bridge image is selected as a similar image. Similar images are closer to the window bridge image in the feature space, indicating that they have similar change patterns or structural features.

[0056] Exemplarily, filters such as Gaussian filters (for smoothing and denoising) or mean filters (for smoothing) should be selected according to processing requirements. The selection of filters should take into account the characteristics of the image and the processing objectives. Ensure that the similar image is spatially aligned with the window bridge image, and fuse the filtered window bridge image with the similar image. The purpose of fusion is to use the characteristic information of the similar image to improve the window bridge image, thereby obtaining a first bridge image. The data-processed first bridge image is recombined to the corresponding position of the original image according to the position of the target window. In this way, each filtered window image is superimposed on its original position to form a second bridge image.

[0057] Exemplarily, threshold segmentation is performed on the second bridge image, and one or more thresholds are set to divide the pixels in the second bridge image into different regions, which is usually used to distinguish the foreground from the background. For example, an edge detection algorithm (such as Sobel operator, Canny edge detection, etc.) is used to identify and extract edge information based on the gradient change in the second bridge image to obtain the target edge information of the target bridge image.

[0058] In some embodiments, filtering the window bridge image according to the similar image and the filter to obtain the first bridge image corresponding to the window bridge image includes: determining an initial filtering coefficient corresponding to the window bridge image according to the filter and the similar image; performing gradient processing on the window bridge image to obtain a gradient bridge image corresponding to the window bridge image; calculating weight information corresponding to the window bridge image according to the gradient bridge image; adjusting a smoothness parameter in the initial filtering coefficient according to the weight information to obtain a target filtering parameter; and filtering the window bridge image according to the target filtering parameter to obtain the first bridge image corresponding to the window bridge image.

[0059] Exemplarily, based on the similar images and the characteristics of the filter used, the filter coefficients for the window bridge image are initially determined. The selection of the initial filter coefficients may depend on the statistical characteristics of the similar images or empirical rules. For example, if the similar images show specific image features, the corresponding filter parameters may be selected to preserve these features.

[0060] Exemplarily, the window bridge image is subjected to gradient processing. Gradient processing involves calculating the brightness changes of the image in all directions, typically including horizontal gradients and vertical gradients. This can be done by calculating the gradient of the image through its pixel values, such as using the Sobel operator. After gradient processing, a gradient bridge image is generated, which shows the intensity distribution of edges and details in the window bridge image. The gradient bridge image can reveal key structural information in the image and facilitate weight calculation in subsequent steps. Using the gradient bridge image, the weight information of the window bridge image is calculated. The weight information typically indicates the relative importance or feature strength of different regions in the image. The weight can be determined by analyzing the size of the gradient value in the gradient image, and high gradient value areas may require more filtering and smoothing to remove noise and retain important features.

[0061] Exemplarily, the smoothness parameters in the initial filter coefficients are adjusted based on the calculated weight information. The weight information helps determine which areas require more smoothing and which areas can retain more details. For example, in areas with larger gradients, higher smoothness may be required to reduce the impact of noise; in areas with smaller gradients, more details can be retained. The adjusted parameters will form the target filter parameters, which can process the image more accurately and achieve better filtering effects.

[0062] Exemplarily, the window bridge image is filtered using the determined target filtering parameters. The filtering operation applies the target filtering parameters to smooth the image, remove noise, and retain important structures and features. After filtering, a first bridge image of the window bridge image is obtained.

[0063] In some embodiments, determining the initial filter coefficient corresponding to the window bridge image according to the filter and the similar image includes: determining the transformation relationship between the similar image and the window bridge image in the filter; determining the filter target corresponding to the filter according to the transformation relationship, and determining the target cost function corresponding to the window bridge image according to the filter target; using the similar image and the window bridge image to perform parameter solution on the target cost function to obtain the initial filter coefficient, wherein the initial filter coefficient includes a first parameter and a second parameter, and the first parameter and the second parameter are calculated according to the following formula:

[0064]

[0065]

[0066] in, represents the first parameter corresponding to the kth window bridge image, represents the second parameter corresponding to the k-th window bridge image, represents the number of pixels corresponding to the kth window bridge image, M represents the number of horizontal pixels corresponding to the window bridge image, and N represents the number of vertical pixels corresponding to the window bridge image. represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the similar image, represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the window bridge image, represents the pixel mean corresponding to the k-th similar image, represents the pixel mean corresponding to the k-th window bridge image, represents the pixel variance corresponding to the k-th similar image, represents the regularization parameter.

[0067] Exemplarily, the similar image and the window bridge image have similar edge information, and the similar image has clear edge information. The similar image is used as a guide image of the filter, the window bridge image is determined as the input of the filter, and the local linear relationship between the input of the filter and the output of the filter. That is, the transformation relationship is a local linear relationship.

[0068] Exemplarily, a target cost function is determined, which is used to evaluate the performance of the filter in achieving the filtering goal. The target cost function may include the difference between the noise level of the image and the quality of the image after filtering. The target cost function is established by combining the features of the similar image and the window bridge image. The parameters of the filter are used as the input of the target cost function to quantify the filtering effect. And the mathematical expression of the target cost function is determined, which should reflect the influence of the filter parameters on the image processing effect. Through optimization algorithms (such as gradient descent, Newton's method, etc.), the filter parameters are adjusted to minimize the target cost function. During the optimization process, the parameters are adjusted by applying the contrast filter to the effects of the similar image and the window bridge image. The filter parameters obtained after optimization are thus the initial filter coefficients. Among them, the transformation relationship is a local linear relationship, then the initial filter coefficients include the first parameter and the second parameter, and the first parameter and the second parameter can be calculated according to the following formula:

[0069]

[0070]

[0071] in, represents the first parameter corresponding to the k-th window bridge image, represents the second parameter corresponding to the k-th window bridge image, represents the number of pixels corresponding to the kth window bridge image, M represents the number of horizontal pixels corresponding to the window bridge image, and N represents the number of vertical pixels corresponding to the window bridge image. Represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the similar image. Represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the window bridge image, represents the pixel mean corresponding to the k-th similar image, represents the pixel mean corresponding to the bridge image in the kth window, represents the pixel variance corresponding to the k-th similar image, represents the regularization parameter.

[0072] Step S104: extracting parameters of each sub-edge information in the target edge information in the target bridge image to obtain target drawing parameters corresponding to the sub-edge information.

[0073] Exemplarily, the target drawing parameters include direction information, position information and size information corresponding to the sub-edge information. The target bridge image is placed in the target coordinate system, the coordinate vector of the sub-edge information in the target coordinate system is obtained, and then the direction information, position information and size information corresponding to the sub-edge information are determined according to the coordinate vector.

[0074] Exemplarily, a target coordinate system suitable for the target bridge image is determined. This target coordinate system should match the geometry and analysis requirements of the target bridge image. Typically, the target coordinate system is a two-dimensional Cartesian coordinate system, where the x-axis represents the horizontal position and the y-axis represents the vertical position. The target bridge image is placed in the target coordinate system. For example, the origin (upper left corner or other reference point) of the target bridge image is aligned to the origin of the coordinate system or other known position.

[0075] Exemplarily, the sub-edge information is converted from the image coordinate system (usually in pixels) to the target coordinate system. To this end, the pixel coordinates in the target bridge image are mapped to the actual positions in the target coordinate system, and a coordinate vector of each sub-edge information is generated in the target coordinate system. The coordinate vector usually includes the positions of the starting point and the ending point of the sub-edge information, or a bounding box describing the sub-edge area.

[0076] Exemplarily, the direction of the sub-edge information is calculated based on the coordinate vector of the sub-edge information in the target coordinate system. The direction information is usually the inclination angle of the sub-edge segment, or the angle between the edge and the horizontal line. Based on the coordinate vector of the sub-edge information, its position in the target coordinate system is determined. This usually includes the coordinates of the center point of the sub-edge or the coordinates of its starting and ending points. Based on the coordinate vector of the sub-edge information, its size information in the target coordinate system is calculated. For a line segment edge, the size information may include the length of the line segment; for an area edge, the size information may include the width and height of the area.

[0077] In some embodiments, the extracting parameters of each sub-edge information in the target edge information in the target bridge image to obtain the target drawing parameters corresponding to the sub-edge information includes: identifying the line type of the sub-edge information to obtain the target type corresponding to the sub-edge information; obtaining the edge length corresponding to each sub-edge information in the target edge information in the target bridge image, and determining the size information corresponding to the sub-edge information based on the edge length; identifying intersections using the position information corresponding to each sub-edge information in the target edge information in the target bridge image to obtain the corresponding connection relationship between the sub-edge information; and determining the target drawing parameters corresponding to the sub-edge information based on the target type, the size information and the connection relationship.

[0078] For example, the line type, such as a straight line, a curve, a broken line, etc., is determined according to the characteristics of the target bridge image. For each sub-edge information, its geometric features, such as shape, curvature, angle, etc., are analyzed. These features are used to determine the type of the line. For example, a straight line may appear as a simple line segment with two end points connected by straight lines, while a curve may appear as a smooth curve.

[0079] Exemplarily, each sub-edge information is classified into a corresponding line type by matching a template of a known line type or using a classification algorithm according to the extracted features.

[0080] Exemplarily, the coordinate vector of the sub-edge information is determined, and the starting point coordinates and the ending point coordinates of the sub-edge information in the target bridge image are obtained. The edge length of the sub-edge information is calculated using the Euclidean distance formula according to the starting point coordinates and the ending point coordinates of the sub-edge information.

[0081] Exemplarily, the size information of the sub-edge information is determined based on the calculated edge length. For example, if the edge length represents the actual length of the line, this edge length is the size information of the sub-edge information. If necessary, the edge length can be converted from the image coordinate unit to the actual physical size unit, such as meter or centimeter. This usually requires the scale or calibration information of the image.

[0082] Exemplarily, the position information of the sub-edge information in the target bridge image is used to identify the intersections between the sub-edge information. The intersection is the location where multiple edges intersect, which can be achieved by calculating the intersection of the edges. Based on the intersection information, the connection relationship between the sub-edge information is determined. The connection relationship indicates which sub-edges are connected to each other and their relative positions.

[0083] Exemplarily, all relevant data are integrated according to the line type, edge length, size information and connection relationship corresponding to each sub-edge information to determine the target drawing parameter corresponding to the sub-edge information. The line type is used to determine the line style used when drawing; the size information is used to determine the line width, length, etc. when drawing; the connection relationship is used to determine the overall structure and connection method of the graphic.

[0084] Specifically, by identifying the line type of the sub-edge information, the nature of each edge can be accurately determined, such as a straight line, curve, or polyline. This identification helps provide a basis for the selection of auxiliary design tools. In addition, identifying the intersections between sub-edge information can reveal the connection relationship between different components. This is crucial for the construction and optimization of bridge structures, because accurate intersection identification ensures the accuracy of the sub-edge information connection, thereby providing reliable support for the rapid generation of bridge design solutions.

[0085] Step S105: obtaining a bridge design scheme corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameters and the auxiliary design tool.

[0086] Exemplarily, the interface function description and interface parameter description of the auxiliary design tool are obtained. The interface function description and interface parameter description describe the functions and required parameters of each interface, and how to use these interfaces to complete specific tasks. Matching the interface function description with the line type in the target drawing parameter ensures that the selected interface can meet the requirements of the target drawing parameter. For example, if the target drawing parameter contains a specific line type (such as a straight line, a curve, or a polyline), it is necessary to ensure that the matched interface can handle these line types. Once the interface function description successfully matches the line type in the target drawing parameter, the target drawing parameter can be matched with the interface parameter description. Through this correspondence, it can be determined how each target drawing parameter is mapped to the interface parameter to ensure that each parameter is correctly passed to the auxiliary design tool. Therefore, according to these mapping relationships, the corresponding interface in the auxiliary design tool is called, and then the bridge design scheme of the target object in the auxiliary design tool is obtained by using these interfaces. This process ensures a seamless conversion from the target drawing parameter to the actual design scheme, making the design work more efficient and accurate, and also realizing the accurate conversion from the target drawing parameter to the actual bridge design scheme.

[0087] In some embodiments, obtaining a bridge design scheme corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameters and the auxiliary design tool includes: determining a mapping relationship between a line type and a drawing tool in the auxiliary design tool; obtaining a target drawing tool corresponding to the sub-edge information in the auxiliary design tool according to the mapping relationship and the target type; and obtaining the bridge design scheme using the target drawing tool according to the size information and the connection relationship.

[0088] For example, determine the object type that needs to be identified (e.g., straight line, curve, polyline, etc.). Match each object type with the corresponding drawing tool in the auxiliary design tool. Create a mapping table or document to clarify which object types correspond to which drawing tools. For example, a straight line may correspond to a "straight line drawing tool" and a curve may correspond to a "curve drawing tool."

[0089] Exemplarily, the mapping relationship between the line type and the drawing tool in the auxiliary design tool is obtained by querying in the mapping table, and the target drawing tool that should be used for each sub-edge information in the auxiliary design tool is first determined. This process involves matching each sub-edge information (such as a straight line, curve, or polyline) in the target design with the drawing tool available in the auxiliary design tool. Through this matching, it can be ensured that the selected drawing tool can accurately handle different line types and geometric features in the target design.

[0090] For example, the target drawing tool is used for drawing according to the size information and the connection relationship. The size information includes parameters such as the actual length and curvature radius of each sub-edge information, while the connection relationship involves how each sub-edge information is connected and combined with each other in space. The determined target drawing tool is used to obtain the bridge design scheme by applying the size information and the connection relationship in the auxiliary design tool.

[0091] In some embodiments, obtaining the bridge design scheme using the target drawing tool according to the size information and the connection relationship includes: determining a drawing order corresponding to the sub-edge information using the connection relationship; and determining the bridge design scheme according to the drawing order using the target drawing tool in combination with the size information.

[0092] Exemplarily, the connection relationship between the sub-edge information is analyzed in detail, and a diagram or flowchart is created based on the connection relationship to show the connection relationship of all sub-edge information. This will help to intuitively understand the drawing order of each edge. Each node in the diagram or flowchart is a sub-edge information, and the connection between the nodes is associated through the connection relationship between the sub-edge information. Based on the connection diagram, starting from the starting point or starting edge in the connection diagram, the drawing order is gradually determined according to the connection relationship. This usually starts from the basic part of the design and gradually expands to more complex parts. Ensure that the previous part of each edge has been drawn when drawing, so as to facilitate the connection of subsequent parts. Therefore, the bridge design scheme is determined according to the drawing order using the target drawing tool combined with the size information.

[0093] Specifically, using the connection relationship to determine the drawing order of sub-edge information helps to improve the accuracy of design and the drawing efficiency.

[0094] See also Figure 2 , Figure 2 A bridge design scheme generation system 200 is provided in an embodiment of the present application. The bridge design scheme generation system 200 includes an image generation module 201, an image adjustment module 202, an edge detection module 203, a parameter extraction module 204, and a scheme generation module 205, wherein the image generation module 201 is used to obtain an initial bridge description text corresponding to a target object, and use the initial bridge description text to perform image generation to obtain an initial bridge image corresponding to the target object; the image adjustment module 202 is used to collect a target comment text expressed by the target object on the initial bridge image, and adjust the initial bridge image according to the target comment text to obtain a target bridge image; the edge detection module 203 is used to perform edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image; the parameter extraction module 204 is used to perform parameter extraction on each sub-edge information in the target edge information in the target bridge image to obtain a target drawing parameter corresponding to the sub-edge information; the scheme generation module 205 is used to obtain a bridge design scheme corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameter and the auxiliary design tool.

[0095] In some implementations, the image generation module 201, during the process of using the initial bridge description text to generate an image to obtain the initial bridge image corresponding to the target object, performs:

[0096] Performing semantic representation on the initial bridge description text according to the text encoding layer of the image generation model to obtain a first text vector;

[0097] Perform keyword characterization on the initial bridge description text according to the keyword encoding layer of the image generation model to obtain a second text vector;

[0098] According to the conditional enhancement layer of the image generation model, the first text vector and the random noise vector are vector-fused to obtain a target splicing vector;

[0099] The image feature generation layer of the image generation model uses the target splicing vector to extract image features to obtain initial image features;

[0100] Performing image generation using the initial image features according to the convolutional network layer of the image generation model to obtain a rough bridge image;

[0101] Calculate the correlation value between the second text vector and each sub-region in the rough bridge image according to the correlation calculation layer of the image generation model;

[0102] The image updating layer of the image generation model uses the correlation value to update the rough bridge image to obtain the initial bridge image corresponding to the target object.

[0103] In some implementations, during the process of adjusting the initial bridge image according to the target comment text to obtain the target bridge image, the image adjustment module 202 performs:

[0104] Perform keyword relationship recognition on the target review text to obtain a target association relationship corresponding to the target review text;

[0105] Performing data representation on the target association relationship according to the information encoding layer of the image enhancement model to obtain a third text vector;

[0106] Obtaining an image feature vector corresponding to the target association relationship from the initial bridge image using the target association relationship according to the feature extraction layer of the image enhancement model;

[0107] Generate a target contour corresponding to the target comment text using the third text vector and the image feature vector according to the mask generation layer of the image enhancement model;

[0108] The target bridge image is obtained by adjusting the initial bridge image using the target contour according to the image processing layer of the image enhancement model.

[0109] In some implementations, the edge detection module 203 performs, during the process of performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image:

[0110] Performing grayscale conversion on the target bridge image to obtain a corresponding grayscale bridge image;

[0111] Determine a target window, and slide the target window on the grayscale bridge image to obtain a window bridge image corresponding to the target window;

[0112] Obtaining a similar image corresponding to the window bridge image, and performing filtering processing on the window bridge image according to the similar image and a filter to obtain a first bridge image corresponding to the window bridge image;

[0113] Combining the first bridge images according to the position information corresponding to the target window to obtain a second bridge image;

[0114] The target edge information corresponding to the target bridge image is obtained by performing edge detection on the second bridge image according to threshold segmentation.

[0115] In some implementations, the edge detection module 203 performs, during the process of filtering the window bridge image according to the similar image and the filter to obtain the first bridge image corresponding to the window bridge image:

[0116] Determine an initial filter coefficient corresponding to the window bridge image according to the filter and the similar image;

[0117] Performing gradient processing on the window bridge image to obtain a gradient bridge image corresponding to the window bridge image;

[0118] Calculate weight information corresponding to the window bridge image according to the gradient bridge image;

[0119] Adjusting the smoothness parameter in the initial filter coefficient according to the weight information to obtain a target filter parameter;

[0120] The window bridge image is filtered according to the target filtering parameters to obtain the first bridge image corresponding to the window bridge image.

[0121] In some implementations, the edge detection module 203, during the process of determining the initial filter coefficient corresponding to the window bridge image according to the filter and the similar image, performs:

[0122] Determine the corresponding transformation relationship between the similar image and the window bridge image in the filter;

[0123] Determine a filtering target corresponding to the filter according to the transformation relationship, and determine a target cost function corresponding to the window bridge image according to the filtering target;

[0124] The target cost function is parameterized using the similar image and the window bridge image to obtain the initial filter coefficient, wherein the initial filter coefficient includes a first parameter and a second parameter, and the first parameter and the second parameter are calculated according to the following formula:

[0125]

[0126]

[0127] in, represents the first parameter corresponding to the kth window bridge image, represents the second parameter corresponding to the k-th window bridge image, represents the number of pixels corresponding to the kth window bridge image, M represents the number of horizontal pixels corresponding to the window bridge image, and N represents the number of vertical pixels corresponding to the window bridge image. represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the similar image, represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the window bridge image, represents the pixel mean corresponding to the k-th similar image, represents the pixel mean corresponding to the k-th window bridge image, represents the pixel variance corresponding to the k-th similar image, represents the regularization parameter.

[0128] In some implementations, the parameter extraction module 204 performs, in the process of extracting parameters from each sub-edge information in the target edge information in the target bridge image and obtaining the target drawing parameters corresponding to the sub-edge information:

[0129] Performing line type recognition on the sub-edge information to obtain a target type corresponding to the sub-edge information;

[0130] Obtaining the edge length corresponding to each sub-edge information in the target edge information in the target bridge image, and determining the size information corresponding to the sub-edge information according to the edge length;

[0131] Using each of the sub-edge information in the target edge information to identify intersections at the position information corresponding to the target bridge image, to obtain corresponding connection relationships between the sub-edge information;

[0132] The target drawing parameter corresponding to the sub-edge information is determined according to the target type, the size information and the connection relationship.

[0133] In some implementations, the solution generation module 205, in the process of obtaining the bridge design solution corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameter and the auxiliary design tool, performs:

[0134] Determining a mapping relationship between line types and drawing tools in the auxiliary design tool;

[0135] Obtaining a target drawing tool corresponding to the sub-edge information in the auxiliary design tool according to the mapping relationship and the target type;

[0136] The bridge design scheme is obtained by using the target drawing tool according to the size information and the connection relationship.

[0137] In some implementations, the solution generation module 205, in the process of obtaining the bridge design solution by using the target drawing tool according to the size information and the connection relationship, executes:

[0138] Determining a drawing order corresponding to the sub-edge information by using the connection relationship;

[0139] A bridge design scheme is determined according to the drawing sequence using the target drawing tool in combination with the size information.

[0140] In some implementations, the bridge design solution generating system 200 may be applied to a terminal device.

[0141] It should be noted that technicians in the relevant field can clearly understand that for the convenience and simplicity of description, the specific working process of the bridge design scheme generation system 200 described above can refer to the corresponding process in the aforementioned bridge design scheme generation method embodiment, and will not be repeated here.

[0142] An embodiment of the present invention also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any bridge design scheme generation method provided in the description of the embodiment of the present invention.

[0143] The storage medium may be an internal storage unit of the terminal device described in the foregoing embodiment, such as a hard disk or memory of the terminal device. The storage medium may also be an external storage device of the terminal device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc., equipped on the terminal device.

[0144] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware embodiment, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transient medium). As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0145] It should be understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0146] The serial numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A bridge design scheme generation method, characterized in that: The method comprises: Obtaining an initial bridge description text corresponding to a target object, and using the initial bridge description text to perform image generation to obtain an initial bridge image corresponding to the target object, wherein the target object is a designer or a bridge design user; Collecting the target comment text expressed by the target object on the initial bridge image, adjusting the initial bridge image according to the target comment text to obtain the target bridge image, wherein the target comment text is used to represent the description text of the target object's improvement content on the initial bridge image, and the target comment text includes suggestions for structural changes, adjusting the design, color and texture requirements of the bridge, modifying the appearance of the bridge, refining the appearance of components, adding decorative elements or correcting defects in the image; Performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image; Extracting parameters of each sub-edge information in the target edge information in the target bridge image to obtain target drawing parameters corresponding to the sub-edge information; According to the association relationship between the target drawing parameters and the auxiliary design tool, a bridge design scheme corresponding to the target object in the auxiliary design tool is obtained.

2. The method according to claim 1, characterized in that The step of using the initial bridge description text to generate an image to obtain an initial bridge image corresponding to the target object includes: Performing semantic representation on the initial bridge description text according to the text encoding layer of the image generation model to obtain a first text vector; Perform keyword characterization on the initial bridge description text according to the keyword encoding layer of the image generation model to obtain a second text vector; According to the conditional enhancement layer of the image generation model, the first text vector and the random noise vector are vector-fused to obtain a target splicing vector; The image feature generation layer of the image generation model uses the target splicing vector to extract image features to obtain initial image features; Performing image generation using the initial image features according to the convolutional network layer of the image generation model to obtain a rough bridge image; Calculate the correlation value between the second text vector and each sub-region in the rough bridge image according to the correlation calculation layer of the image generation model; The image updating layer of the image generation model uses the correlation value to update the rough bridge image to obtain the initial bridge image corresponding to the target object.

3. The method according to claim 1, characterized in that The step of adjusting the initial bridge image according to the target comment text to obtain a target bridge image includes: Perform keyword relationship recognition on the target review text to obtain a target association relationship corresponding to the target review text; Performing data representation on the target association relationship according to the information encoding layer of the image enhancement model to obtain a third text vector; Obtaining an image feature vector corresponding to the target association relationship from the initial bridge image using the target association relationship according to the feature extraction layer of the image enhancement model; Generate a target contour corresponding to the target comment text using the third text vector and the image feature vector according to the mask generation layer of the image enhancement model; The target bridge image is obtained by adjusting the initial bridge image using the target contour according to the image processing layer of the image enhancement model.

4. The method according to claim 1, characterized in that The performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image includes: Performing grayscale conversion on the target bridge image to obtain a corresponding grayscale bridge image; Determine a target window, and slide the target window on the grayscale bridge image to obtain a window bridge image corresponding to the target window; Obtaining a similar image corresponding to the window bridge image, and performing filtering processing on the window bridge image according to the similar image and a filter to obtain a first bridge image corresponding to the window bridge image; Combining the first bridge images according to the position information corresponding to the target window to obtain a second bridge image; The target edge information corresponding to the target bridge image is obtained by performing edge detection on the second bridge image according to threshold segmentation.

5. The method according to claim 4, characterized in that The filtering process of the window bridge image according to the similarity image and the filter to obtain a first bridge image corresponding to the window bridge image includes: Determine an initial filter coefficient corresponding to the window bridge image according to the filter and the similar image; Performing gradient processing on the window bridge image to obtain a gradient bridge image corresponding to the window bridge image; Calculate weight information corresponding to the window bridge image according to the gradient bridge image; Adjusting the smoothness parameter in the initial filter coefficient according to the weight information to obtain a target filter parameter; The window bridge image is filtered according to the target filtering parameters to obtain the first bridge image corresponding to the window bridge image.

6. The method according to claim 5, characterized in that The step of determining the initial filter coefficient corresponding to the window bridge image according to the filter and the similar image comprises: Determine the corresponding transformation relationship between the similar image and the window bridge image in the filter; Determine a filtering target corresponding to the filter according to the transformation relationship, and determine a target cost function corresponding to the window bridge image according to the filtering target; The target cost function is parameterized using the similar image and the window bridge image to obtain the initial filter coefficient, wherein the initial filter coefficient includes a first parameter and a second parameter, and the first parameter and the second parameter are calculated according to the following formula: ; ; in, represents the first parameter corresponding to the kth window bridge image, represents the second parameter corresponding to the k-th window bridge image, represents the number of pixels corresponding to the kth window bridge image, M represents the number of horizontal pixels corresponding to the window bridge image, and N represents the number of vertical pixels corresponding to the window bridge image. represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the similar image, represents the pixel value corresponding to the i-th horizontal pixel and the j-th vertical pixel of the window bridge image, represents the pixel mean corresponding to the k-th similar image, represents the pixel mean corresponding to the k-th window bridge image, represents the pixel variance corresponding to the k-th similar image, represents the regularization parameter.

7. The method according to claim 1, characterized in that The extracting parameters of each sub-edge information in the target edge information in the target bridge image to obtain the target drawing parameters corresponding to the sub-edge information includes: Performing line type recognition on the sub-edge information to obtain a target type corresponding to the sub-edge information; Obtaining the edge length corresponding to each sub-edge information in the target edge information in the target bridge image, and determining the size information corresponding to the sub-edge information according to the edge length; Using each of the sub-edge information in the target edge information to identify intersections at the position information corresponding to the target bridge image, to obtain corresponding connection relationships between the sub-edge information; The target drawing parameter corresponding to the sub-edge information is determined according to the target type, the size information and the connection relationship.

8. The method according to claim 7, characterized in that The step of obtaining a bridge design scheme corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameter and the auxiliary design tool comprises: Determining a mapping relationship between line types and drawing tools in the auxiliary design tool; Obtaining a target drawing tool corresponding to the sub-edge information in the auxiliary design tool according to the mapping relationship and the target type; The bridge design scheme is obtained using the target drawing tool according to the size information and the connection relationship.

9. The method according to claim 8, characterized in that The obtaining the bridge design scheme by using the target drawing tool according to the size information and the connection relationship comprises: Determining a drawing order corresponding to the sub-edge information by using the connection relationship; A bridge design scheme is determined according to the drawing sequence using the target drawing tool in combination with the size information.

10. A bridge design scheme generation system, characterized in that: include: An image generation module, used to obtain an initial bridge description text corresponding to a target object, and use the initial bridge description text to perform image generation to obtain an initial bridge image corresponding to the target object, wherein the target object is a designer or a bridge design user; An image adjustment module is used to collect the target comment text expressed by the target object on the initial bridge image, and adjust the initial bridge image according to the target comment text to obtain a target bridge image, wherein the target comment text is used to represent the description text of the content that needs to be improved on the initial bridge image by the target object, and the target comment text includes suggestions for structural changes, adjusting the design, color and texture requirements of the bridge, modifying the appearance of the bridge, refining the appearance of components, adding decorative elements or correcting defects in the image; An edge detection module, used for performing edge detection on the target bridge image to obtain target edge information corresponding to the target bridge image; A parameter extraction module, used for extracting parameters of each sub-edge information in the target edge information in the target bridge image, and obtaining target drawing parameters corresponding to the sub-edge information; The scheme generating module is used to obtain the bridge design scheme corresponding to the target object in the auxiliary design tool according to the association relationship between the target drawing parameters and the auxiliary design tool.

Images