Surveying and mapping data integrated building target rapid extraction method and system

Abstract

The invention discloses a surveying and mapping data integrated building target rapid extraction method and system, and relates to the related technical field of building target extraction, and the method comprises the steps: integrating multi-source parallel collection data, and constructing a surveying and mapping integrated database of a target region; acquiring first target surveying and mapping integrated data; extracting a high-dimensional feature vector representing the facade material of the target building; inputting the high-dimensional feature vector into a spectrum-texture collaborative enhancement model to carry out data extraction influence analysis; and if the first influence coefficient is greater than a preset influence coefficient, activating the spectrum-texture collaborative enhancement network to perform data enhancement, obtaining second target surveying and mapping integrated data, and generating an extraction model of the target building. The technical problems of low building recognition rate and poor extraction accuracy caused by difficulty in effectively recognizing the target building formed by different building materials according to material reflection characteristics in the prior art are solved, the accurate extraction model is efficiently generated, and the technical effect of improving the building target recognition rate and accuracy is achieved.

Description

Technical Field

[0001] The present application relates to the technical field related to building target extraction, and specifically to a method and system for rapid extraction of building targets integrated with surveying and mapping data. Background Art

[0002] Because different building materials, such as glass, metal, and masonry, have vastly different light reflection characteristics, the grayscale values and tones presented in optical images are wide and complex. For example, glass may produce specular reflections, causing some areas to be overly bright and losing detail information; metal materials, on the other hand, have unique reflective spectral characteristics that can be easily confused with other materials, making it difficult to effectively distinguish between building targets made of different building materials, resulting in low recognition rates. Furthermore, the textures formed by different materials vary significantly in terms of roughness and pattern complexity, making it difficult to fully explore and utilize the rich information contained in surface textures to assist in recognition. Existing methods struggle to accurately capture these subtle differences, which in turn affects the accuracy of building target recognition.

[0003] Therefore, in the current related technologies, there is a technical problem that it is difficult to effectively identify target buildings made of different building materials based on the material reflection characteristics, resulting in low building recognition rate and poor extraction accuracy. Summary of the Invention

[0004] This application solves the technical problem in the prior art that it is difficult to effectively identify target buildings made of different building materials based on the material reflectance characteristics, resulting in low building recognition rate and poor extraction accuracy, by providing a method and system for rapid extraction of building targets integrated with surveying and mapping data. It efficiently generates an accurate extraction model, achieving the technical effect of improving the recognition rate and accuracy of building targets.

[0005] The present application provides a method for rapid extraction of building targets integrated with surveying and mapping data, the method comprising: integrating multi-source parallel acquisition data to construct a surveying and mapping integrated database of a target area, the target area including a target building to be extracted; obtaining first target surveying and mapping integrated data of the target building in the surveying and mapping integrated database; extracting a high-dimensional feature vector representing the facade material of the target building based on the first target surveying and mapping integrated data, the high-dimensional feature vector consisting of a spectral feature vector and a surface texture feature vector; inputting the high-dimensional feature vector into a spectrum-texture collaborative enhancement model to perform data extraction impact analysis and output a first influence coefficient; if the first influence coefficient is greater than a preset influence coefficient, activating a spectrum-texture collaborative enhancement network to perform data enhancement on the first target surveying and mapping integrated data, obtaining second target surveying and mapping integrated data, and generating an extraction model of the target building based on the second target surveying and mapping integrated data.

[0006] In a possible implementation, the method for rapid extraction of building targets based on surveying and mapping data integration also performs the following processing: the spectrum-texture collaborative enhancement model determines the facade material type of the target building based on the high-dimensional feature vector; obtains the real-time lighting conditions of the multi-source parallel acquisition data, including real-time lighting intensity and real-time lighting angle; establishes a mapping function relationship between the facade material type samples and the lighting condition samples to characterize the degree of influence on data extraction, analyzes the facade material type and the real-time lighting conditions according to the mapping function relationship, and outputs a first influence coefficient.

[0007] In a possible implementation, the method for rapid extraction of building targets based on surveying and mapping data integration further performs the following processing: activating a spectral-texture collaborative enhancement network to identify boundary defects of the first target surveying and mapping integrated data and outputting defect surveying and mapping integrated data; obtaining matching lighting conditions with an influence coefficient less than a preset influence coefficient based on the type of facade material; calling matching surveying and mapping integrated sample data corresponding to the matching lighting conditions to perform defect data enhancement on the defect surveying and mapping integrated data and obtain second target surveying and mapping integrated data.

[0008] In a possible implementation, the method for rapid extraction of building targets based on surveying and mapping data integration further performs the following processing: obtaining high-dimensional feature vector samples, the high-dimensional feature vector samples including spectral feature vector samples and surface texture feature samples of different building materials under different lighting conditions, the spectral feature vector samples including the band reflectance of the spectral image, and the surface texture feature vector samples including contrast, energy, and correlation parameters calculated by GLCM; obtaining label sample data, the label sample data including material categories and corresponding annotated surveying and mapping integrated sample data; performing model training based on the high-dimensional feature vector samples and the label sample data to obtain a spectral-texture collaborative enhancement network trained to convergence.

[0009] In a possible implementation, the method for rapid extraction of building targets based on surveying and mapping data further performs the following processing: initializing a generative adversarial network, including initializing a dual-branch generator and a discriminator; wherein the dual-branch generator includes a spectral branch and a texture branch, the spectral branch simulates the band reflectance changes of different building materials under different lighting conditions through a fully connected layer, and the texture branch simulates the surface texture changes of different building materials under different lighting conditions through a deconvolution layer; after inputting the high-dimensional feature vector sample into the dual-branch generator, a generative high-dimensional feature vector is output, the discriminator receives the generative high-dimensional feature vector output by the dual-branch generator and performs adversarial loss feedback training with the high-dimensional feature vector sample to obtain a trained generative adversarial network; and the trained generative adversarial network is called to supplement the high-dimensional feature vector sample.

[0010] In a possible implementation, the method for rapid extraction of building targets based on surveying and mapping data integration further performs the following processing: if there are multiple target buildings to be extracted, multiple target surveying and mapping integration data corresponding to the multiple target buildings are obtained; multiple high-dimensional feature vectors characterizing the facade materials of the multiple target buildings are extracted based on the multiple target surveying and mapping integration data; the multiple high-dimensional feature vectors are input into the spectrum-texture collaborative enhancement model to obtain multiple extraction models corresponding to the multiple target buildings.

[0011] In a possible implementation, the method for rapid extraction of building targets based on surveying and mapping data integration further performs the following processing: the spectrum-texture collaborative enhancement network also includes a dynamic weight adjustment module, which dynamically adjusts the fusion weight of spectral features and texture features according to the facade material type and the real-time lighting conditions; the defect data weight is enhanced on the defective surveying and mapping integrated data according to the fusion weight to obtain second target surveying and mapping integrated data.

[0012] The present application also provides a rapid extraction system for building targets integrated with surveying and mapping data, the system comprising: a surveying and mapping integrated database construction module, for integrating multi-source parallel acquisition data to construct a surveying and mapping integrated database for a target area, the target area including a target building to be extracted; a first target surveying and mapping integrated data acquisition module, for acquiring first target surveying and mapping integrated data of the target building in the surveying and mapping integrated database; a facade feature vector extraction module, for extracting a high-dimensional feature vector characterizing the facade material of the target building based on the first target surveying and mapping integrated data, the high-dimensional feature vector consisting of a spectral feature vector and a surface texture feature vector; a data extraction influence analysis module, for inputting the high-dimensional feature vector into a spectrum-texture collaborative enhancement model for data extraction influence analysis, and outputting a first influence coefficient; a target building extraction model generation module, for activating a spectrum-texture collaborative enhancement network to perform data enhancement on the first target surveying and mapping integrated data if the first influence coefficient is greater than a preset influence coefficient, acquiring second target surveying and mapping integrated data, and generating an extraction model of the target building based on the second target surveying and mapping integrated data.

[0013] The proposed method and system for rapid extraction of building targets using surveying and mapping data integration will integrate multi-source parallel data collection to construct a surveying and mapping integrated database for the target area. The system will also acquire first target surveying and mapping integrated data, extract high-dimensional feature vectors representing the facade materials of the target building, and input the high-dimensional feature vectors into a spectral-texture collaborative enhancement model for data extraction impact analysis. If the first impact coefficient is greater than a preset impact coefficient, the spectral-texture collaborative enhancement network will be activated for data enhancement, and second target surveying and mapping integrated data will be acquired to generate an extraction model for the target building. This method solves the technical problem in the prior art of the difficulty in effectively identifying target buildings made of different building materials based on the material reflectance characteristics, resulting in low building recognition rates and poor extraction accuracy. This method will efficiently generate accurate extraction models, achieving the technical effect of improving the recognition rate and accuracy of building targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the accompanying drawings of the embodiments of the present disclosure are briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in precise order. Instead, various steps may be processed in reverse order or simultaneously as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0015] Figure 1 A flow chart of a method for rapidly extracting building targets from surveying and mapping data integration provided in an embodiment of the present application.

[0016] Figure 2 Schematic diagram of the structure of the building target rapid extraction system based on surveying and mapping data integration provided in an embodiment of the present application.

[0017] Explanation of the reference numerals: surveying and mapping integrated database construction module 10 , first target surveying and mapping integrated data acquisition module 20 , facade feature vector extraction module 30 , data extraction impact analysis module 40 , target building extraction model generation module 50 . DETAILED DESCRIPTION

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0019] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0020] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict, and the terms “first\second” involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or that are inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only.

[0021] The embodiment of the present application provides a method for quickly extracting building targets from surveying and mapping data integration, such as Figure 1 As shown, the method includes: Step S100 , integrating multi-source parallel acquisition data to construct a surveying and mapping integrated database of a target area, wherein the target area includes a target building to be extracted.

[0022] Preferably, data within the same or similar time period of the target area is collected through a variety of different methods to reflect the status of the buildings in the target area under the same time base, so as to avoid data inconsistency caused by changes in building status (such as construction progress, decoration changes, etc.) due to time differences. The target area includes the target building to be extracted. Specifically, optical remote sensing data is obtained by using optical sensors carried by satellites or aircraft to provide large-area image information of the target area, covering the overall layout, general outline and relationship with the surrounding environment of the building, such as clearly showing the arrangement of different buildings; terrestrial laser scanning (TLS) and airborne laser scanning (ALS) obtain laser scanning data of the target building, that is, TLS can perform high-precision scanning of the building at close range, accurately obtain three-dimensional coordinate information of the building surface, and generate detailed Point cloud data is used to obtain the complex facade structure and decorative details of the building, and can accurately measure the size, shape and surface undulations of the building. ALS is used to quickly obtain three-dimensional information of the terrain and buildings in a large area, including the height of the buildings in the target area, the spatial relationship between the terrain and the buildings, etc.; by acquiring images of the target area from multiple angles, oblique photography data is obtained to present multiple sides of the building, providing rich texture information for building a realistic three-dimensional model, making the reconstructed building model more realistic and accurate, which can be used to intuitively display the appearance of the building and assist in identifying the detailed features and structure of the building; surveying and mapping personnel use total stations, levels and other instruments to measure the target building on the spot and obtain field measurement data, including the key dimensional parameters of the target building, such as the length, width, height of the building, the position and size of doors and windows, etc.

[0023] Preferably, the collected multi-source data is format converted and preprocessed so that it can be integrated into a database framework. For example, laser scanning point cloud data and optical image data of different formats are uniformly converted into a format suitable for database storage, and the data is preprocessed by denoising, calibration, and other operations. Then, based on spatial reference information such as geographic coordinates, these data from different sources are accurately aligned to ensure that they completely match in spatial position. For example, the building location in the optical remote sensing image is accurately aligned with the corresponding building location in the laser scanning point cloud data. Select a suitable database management system (DBMS), such as PostgreSQL, to build a surveying and mapping integrated database. Consider the characteristics and storage requirements of different data types and set corresponding data tables and fields. For example, for optical remote sensing data, a data table containing fields such as image file path, shooting time, and band information can be established; for laser scanning point cloud data, a table structure is set to store point cloud coordinates, reflection intensity, and other information to achieve efficient storage and fast query and call of multi-source parallel collected data.

[0024] Step S200: Acquire first target surveying and mapping integrated data of the target building in the surveying and mapping integrated database.

[0025] Preferably, the first target surveying and mapping integrated data of the target building is screened out from the surveying and mapping integrated database. Specifically, the target building is accurately located according to the geographic coordinates and building attributes (building name, number, purpose, etc.) in the surveying and mapping integrated data, and then relevant multi-source surveying and mapping data are extracted from different data tables according to the association relationship. For example, the image of the target building area is extracted from the optical remote sensing data table, from which information such as the building appearance and surrounding environment can be obtained; the point cloud data of the building is retrieved from the laser scanning point cloud data table, which can accurately present the three-dimensional shape and structural details of the building; multi-angle images are obtained from the oblique photography data storage location to provide rich texture for comprehensive analysis of the building appearance; then the multi-source surveying and mapping data (optical images, point cloud data, oblique photography data, etc.) are aligned in time and space to generate the first target surveying and mapping integrated data of the target building, and then the high-dimensional feature vector representing the facade material of the target building can be accurately extracted to accurately describe the material properties and appearance texture of the target building.

[0026] Step S300: extracting a high-dimensional feature vector representing the facade material of the target building based on the first target surveying and mapping integrated data, wherein the high-dimensional feature vector is composed of a spectral feature vector and a surface texture feature vector.

[0027] Preferably, the spectral feature vector and surface texture feature vector of the target building are extracted based on the first target surveying and mapping integrated data, and the spectral feature vector and the surface texture feature vector are combined to form a high-dimensional feature vector characterizing the facade material. This integrates the information of both spectrum and texture, and can more comprehensively and accurately describe the characteristics of the facade material of the target building. Specifically, the spectral feature refers to the reflection, absorption and transmission characteristics of the facade material of the target building for light of different wavelengths. The optical remote sensing data in the first target surveying and mapping integrated data contains rich spectral information, and the reflectance values of different bands are extracted from it, such as the visible light band (such as blue light, green light, red light) and the near-infrared band. The reflectance values of these bands constitute the basic elements of the spectral feature vector. Then, characteristic parameters related to the spectrum are calculated, such as spectral slope, spectral curvature, band ratio, etc., which further enrich the information of the spectral feature vector and help to distinguish different types of building facade materials. For example, the reflectance of metal materials and concrete materials in the near-infrared band is quite different. By analyzing the characteristics of these bands, it is possible to preliminarily determine whether the building facade contains metal components.

[0028] Preferably, the surface texture features reflect the roughness, texture direction, pattern and other characteristics of the target building facade. The laser scanning data, oblique photography data and high-resolution optical images in the first target mapping integrated data are all used to extract surface texture features. Specifically, the texture features of the image data are extracted based on the gray level co-occurrence matrix (GLCM) method, the local binary pattern (LBP) method, etc. For example, the gray level co-occurrence matrix describes the texture features by counting the frequency of occurrence of pixel pairs with different gray values in the image, which can reflect the roughness, contrast, directionality and other information of the texture; from the laser scanning point cloud data, the density distribution, surface curvature and other information of the point cloud are calculated, which can also be used as part of the texture features. For example, the decorative lines or uneven surfaces of the building facade will cause changes in the point cloud density and curvature. By analyzing these changes, the corresponding texture features can be extracted.

[0029] Step S400: input the high-dimensional feature vector into a spectrum-texture collaborative enhancement model to perform data extraction impact analysis and output a first impact coefficient.

[0030] Preferably, the spectrum-texture collaborative enhancement model takes the extracted high-dimensional feature vector containing the spectral feature vector and the surface texture feature vector as input, and is used to simulate and analyze how the spectral information and the texture information interact with each other and synergistically affect the extraction effect of the target building facade material data. Specifically, the spectrum-texture collaborative enhancement model fuses the spectral and texture features in the input high-dimensional feature vector, explores the potential relationship between the two and their comprehensive impact on the accurate extraction of facade material data. For example, the model may analyze whether certain specific spectral features and texture feature combinations can more effectively identify a certain specific facade material, or determine whether certain feature combinations have redundant information; then the input high-dimensional feature vector is compared with the set The standard is compared to evaluate the degree of difference between the current high-dimensional feature vector and the ideal situation, and then analyze its impact on data extraction; finally, the first influence coefficient is output to indicate the degree of influence of the current high-dimensional feature vector on the data extraction of the target building facade material. Generally speaking, if the first influence coefficient is close to 1, it means that the high-dimensional feature vector is more matched with the ideal feature pattern, which has a positive promoting effect on data extraction, can accurately reflect the characteristics of the target building facade material, and help improve the accuracy and reliability of data extraction; if the coefficient is much less than 1, it means that there is a large difference between the high-dimensional feature vector and the ideal pattern, and there may be some interference factors or incomplete features, resulting in blurred boundaries or textures, which has a certain negative impact on data extraction.

[0031] Furthermore, step S400 also includes step S410, wherein the spectrum-texture collaborative enhancement model determines the type of facade material of the target building based on the high-dimensional feature vector; step S420, obtains the real-time lighting conditions of the multi-source parallel collected data, including real-time lighting intensity and real-time lighting angle; step S430, establishes a mapping function relationship between the facade material type sample and the lighting condition sample to characterize the degree of influence of data extraction, analyzes the facade material type and the real-time lighting condition according to the mapping function relationship, and outputs a first influence coefficient.

[0032] Preferably, since different facade materials have unique performances in spectral reflectance characteristics and surface textures, the spectrum-texture collaborative enhancement model analyzes the high-dimensional feature vector to determine the type of facade material of the target building. For example, glass materials have specific reflection peaks in the spectrum, and their surface texture is usually smoother; while the spectral reflectance distribution and texture characteristics of stone materials are different from those of glass. The spectrum-texture collaborative enhancement model determines the type of facade material of the target building, such as glass, stone, metal, concrete, etc., by comparing the high-dimensional feature vector with the feature templates of various facade materials; during the multi-source parallel data acquisition process, real-time lighting conditions are recorded synchronously, including real-time light intensity and real-time light angle, where light intensity refers to the energy of visible light received per unit area, and light angle refers to the incident angle of light relative to the facade of the target building. Lighting conditions will have a significant impact on the spectral reflectance characteristics of the building facade material. For example, under direct strong light, the reflectivity of the material may increase, resulting in changes in the spectral characteristics. Different lighting angles will also cause different shadows and highlights on the material surface, affecting the performance of texture characteristics.

[0033] Preferably, a large number of facade material type samples (such as glass, stone, metal, concrete, etc.) and corresponding lighting condition samples are collected through field measurements, remote sensing images, laboratory simulations, etc., and each sample data is labeled to clarify its facade material type and corresponding lighting conditions (real-time lighting intensity and real-time lighting angle), and their influence on data extraction is analyzed. A mapping function relationship that characterizes the influence of data extraction is established, and the influence of different facade material types on the accuracy of target building data extraction under different lighting conditions is described. Specifically, features related to facade material type and lighting conditions are extracted. For facade materials, spectral feature vectors and surface Texture feature vector; for lighting conditions, extract real-time light intensity and real-time light angle, and calculate the angle between the lighting direction and the building orientation, the rate of change of light intensity in different time periods, etc.; through correlation analysis, principal component analysis, etc., screen out features that have a significant correlation with the degree of influence on data extraction, remove redundant features, and reduce data dimensions; then select a suitable machine learning model to establish a mapping function relationship. For example, use sample data to train the model (such as linear regression model, decision tree model, support vector machine, neural network, etc.) so that the model can learn the intrinsic mapping relationship between facade material type samples and lighting condition samples to minimize the error between the prediction results and the actual annotation. Finally, after evaluation and verification, the mapping function relationship between facade material type samples and lighting condition samples that characterizes the degree of influence on data extraction is obtained. According to the input facade material type and real-time lighting conditions, the corresponding first influence coefficient can be accurately output to describe the degree of influence on data extraction under these conditions.

[0034] Step S500: If the first influence coefficient is greater than the preset influence coefficient, activate the spectrum-texture collaborative enhancement network to perform data enhancement on the first target surveying and mapping integrated data, obtain second target surveying and mapping integrated data, and generate the extraction model of the target building based on the second target surveying and mapping integrated data.

[0035] Preferably, the preset influence coefficient is a threshold value pre-set based on experience and the requirements for data extraction accuracy. When the first influence coefficient is greater than the preset influence coefficient, the current first target surveying and mapping integrated data has greater interference or uncertainty when used to extract target building information, which may cause inaccurate extraction results. Therefore, the data needs to be enhanced; then the spectrum-texture collaborative enhancement network is activated to perform data enhancement reconstruction of the first target surveying and mapping integrated data. Specifically, for the spectral features in the data, the accuracy of the spectral reflectance is adjusted, the differences between the spectral features of different materials are enhanced, etc., so as to enhance the spectral information. For example, for spectral features that are not obvious enough For building materials, the spectral characteristics of these materials are made more prominent by adjusting the shape of the spectral curve and increasing the weight of specific bands, making them easier to identify. In terms of texture features, the network will optimize the surface texture of the building facade. For example, for blurred texture images, image sharpening and edge enhancement are used to make the texture details clearer. For incomplete texture information, interpolation and completion are used to repair and improve it to improve the recognizability of texture features. The spectrum-texture collaborative enhancement network further enhances the synergy of spectrum and texture information, thereby obtaining the second target mapping integrated data, and can more comprehensively and accurately reflect the facade characteristics of the target building. Then, the spectral and texture information in the second target mapping integrated data is used to build and train a machine learning classification model to determine the extraction model of the target building. Based on the input second target mapping integrated data, the relevant information of the target building, such as the building's outline, facade material type, spatial structure, etc., can be accurately identified and extracted. For example, a deep learning convolutional neural network (CNN) is used, and the second target mapping integrated data is used as input. After network training and learning, the network parameters are adjusted so that the network can output accurate target building extraction results, and finally an extraction model of the target building is generated.

[0036] Furthermore, step S500 also includes step S510, activating the spectrum-texture collaborative enhancement network to identify boundary defects of the first target surveying and mapping integrated data and outputting defect surveying and mapping integrated data; step S520, obtaining matching lighting conditions with an influence coefficient less than a preset influence coefficient based on the type of facade material; step S530, calling the matching surveying and mapping integrated sample data corresponding to the matching lighting conditions to perform defect data enhancement on the defect surveying and mapping integrated data to obtain second target surveying and mapping integrated data.

[0037] Preferably, after the spectrum-texture collaborative enhancement network is activated, the first target mapping integrated data is deeply analyzed to identify and mark boundary defects. Boundary defects in building target mapping may include blurred or incomplete edges of the building facade, or unclear features at the junction of materials. The activated spectrum-texture collaborative enhancement network analyzes the spectral features and finds abnormal changes in the spectral reflectance of certain areas, which may indicate defects in the boundary. Or, starting from the texture features, it detects that the surface texture is broken or discontinuous at the boundary. These defects are then marked to clarify the location and range information of the boundary defects, and finally outputs defect mapping integrated data containing boundary defect identification. Since different facade materials will exhibit different spectral and textural characteristics under different lighting conditions, matching lighting conditions with an influence coefficient less than a preset influence coefficient are obtained based on the type of facade material of the target building. Specifically, when the influence coefficient corresponding to the lighting condition is less than the threshold, it means that the data extraction interference under this lighting condition is smaller, which is more conducive to accurate analysis. For example, for a facade made of glass, under a certain specific light intensity and angle, its spectral reflectance and textural characteristics can be more clearly identified, and the corresponding influence coefficient is less than the preset value, then this lighting condition is identified as a matching lighting condition.

[0038] Preferably, after obtaining the matching lighting conditions, the matching surveying and mapping integrated sample data corresponding to the matching lighting conditions is retrieved from the database. This matching surveying and mapping integrated sample data is then applied to the defect surveying and mapping integrated data, and the defect data is enhanced. Specifically, the spectral characteristics of the defective area are corrected and supplemented to make them more consistent with the normal spectral performance of the facade material under the matching lighting conditions; the texture characteristics are repaired and improved to fill in the missing parts of the texture or make the texture clearer and more coherent. Through data enhancement and reconstruction of the boundary defect data, the second target surveying and mapping integrated data is ultimately obtained, improving the data quality and usability, and more accurately extracting relevant information about the target building.

[0039] Furthermore, step S510 also includes step S511, obtaining high-dimensional feature vector samples, the high-dimensional feature vector samples include spectral feature vector samples and surface texture feature samples of different building materials under different lighting conditions, the spectral feature vector samples include the band reflectance of the spectral image, and the surface texture feature vector samples include contrast, energy, and correlation parameters calculated by GLCM; step S512, obtaining label sample data, the label sample data includes material category and corresponding annotated surveying and mapping integrated sample data; step S513, performing model training based on the high-dimensional feature vector samples and the label sample data to obtain a spectral-texture collaborative enhancement network trained to convergence.

[0040] Preferably, a spectrometer or a device with multispectral imaging function is used to collect data on different building materials under different lighting conditions, measure and record the reflectance information of building materials in multiple bands, form a spectral image, extract the reflectance data of each band from the collected spectral image, and the reflectance value of each band is used as a dimension of the feature vector. These reflectance data together constitute the spectral feature vector sample; use a high-resolution camera to capture the surface images of different building materials under different lighting conditions, convert the collected color images into grayscale images, reduce the dimension of the image data, and highlight the texture information of the image For grayscale images, the gray-level co-occurrence matrix is calculated to describe the spatial distribution of grayscale levels in the image. The frequency of pixel pairs with specific grayscale values separated by a certain distance and direction in the image is counted. By changing the distance and direction parameters of the pixel pairs, different GLCMs can be obtained. Parameters such as contrast, energy, and correlation are calculated from the GLCM. Contrast reflects the clarity and roughness of the texture in the image, energy represents the uniformity of the image grayscale distribution, and correlation measures the linear correlation of local grayscales in the image. These parameters serve as elements of the surface texture feature vector and together constitute surface texture feature vector samples. Finally, the spectral feature vector samples and surface texture feature samples are combined to generate high-dimensional feature vector samples for different building materials under different lighting conditions.

[0041] Preferably, the categories of building materials used in different parts of the target building are clearly defined, such as steel, wood, glass, concrete, stone, etc.; then, a comprehensive mapping of the target building is performed, including the building's geometric dimensions, shape, location information, and detailed surface features of the materials. The collected mapping data is then associated and annotated with the corresponding material categories to form corresponding annotated mapping integrated sample data. The material categories and the corresponding annotated mapping integrated sample data are then integrated to form labeled sample data. The prepared high-dimensional feature vector samples and label sample data are input into the spectrum-texture collaborative enhancement network for training. During the training process, the network will try to predict the corresponding material category based on the input feature vector samples, and compare the predicted results with the true values in the label sample data. The performance of the network is evaluated by calculating the error between the predicted value and the true value; based on the error, the network will use optimization algorithms (such as stochastic gradient descent, etc.) to adjust the parameters in the network (such as weights and biases) to gradually reduce the error. When the error is reduced to a certain extent, or when the error no longer decreases significantly in continuous training iterations, the model is considered to have converged, and a spectrum-texture collaborative enhancement network is obtained, which can perform accurate spectrum-texture collaborative analysis on new and unseen data, such as accurately identifying the characteristics of different building materials under various lighting conditions and predicting their material categories.

[0042] Furthermore, step S511 also includes step A, initializing the generative adversarial network, including initializing a dual-branch generator and a discriminator; step B, wherein the dual-branch generator includes a spectral branch and a texture branch, the spectral branch simulates the band reflectance changes of different building materials under different lighting conditions through a fully connected layer, and the texture branch simulates the surface texture changes of different building materials under different lighting conditions through a deconvolution layer; step C, after inputting the high-dimensional feature vector sample into the dual-branch generator, outputting a generative high-dimensional feature vector, the discriminator receives the generative high-dimensional feature vector output by the dual-branch generator and performs adversarial loss feedback training with the high-dimensional feature vector sample to obtain a trained generative adversarial network; step D, calling the trained generative adversarial network to supplement the high-dimensional feature vector sample.

[0043] Preferably, the input end of the spectrum-texture collaborative enhancement network is connected to the generative adversarial network, indicating that the data processed by the generative adversarial network (high-dimensional feature vector samples supplemented by samples) will be used as the input data of the spectrum-texture collaborative enhancement network, forming a data transmission and interaction relationship. The generative adversarial network provides richer input data for the spectrum-texture collaborative enhancement network, which helps to improve the performance of the spectrum-texture collaborative enhancement network. When initializing the generative adversarial network, its core components, the dual-branch generator and the discriminator, are initialized. The dual-branch generator includes a spectrum branch and a texture branch. Specifically, the light The spectral branch uses a fully connected layer to simulate the changes in the band reflectance of different building materials under different lighting conditions. By performing weighted summation on the input features, it learns and captures the laws and characteristics of the spectral reflectance of different building materials under various lighting conditions, thereby generating simulated band reflectance data; the texture branch uses a deconvolution layer to simulate the surface texture changes of different building materials under different lighting conditions. The deconvolution layer has an upsampling function and can recover high-resolution texture information from low-dimensional features. Through training, it learns the characteristics of the surface texture of different building materials under different lighting conditions, and then generates corresponding surface texture data.

[0044] Preferably, high-dimensional feature vector samples are input into a dual-branch generator, where the spectral branch and texture branch process the input separately and then merge to output a generated high-dimensional feature vector, which contains the spectral and texture feature information simulated by the generator. Specifically, the discriminator receives the generated high-dimensional feature vector and the original high-dimensional feature vector samples, and distinguishes whether the input feature vector comes from a real high-dimensional feature vector sample (i.e., real data) or is generated by the generator (i.e., generated data). During the training process, the generator and the discriminator engage in an adversarial game. The generator strives to generate generated high-dimensional feature vectors that can deceive the discriminator, while the discriminator continuously improves its ability to distinguish between real data and generated data.

[0045] Preferably, by calculating an adversarial loss (such as cross-entropy loss), the parameters of the generator and discriminator (such as weights and biases) are adjusted according to the loss value. After multiple rounds of iterative training, when the quality of the data generated by the generator is high enough and the discriminator has difficulty distinguishing between real data and generated data, a trained generative adversarial network is obtained. Finally, the generative adversarial network uses its generative ability to supplement the high-dimensional feature vector samples. That is, the generator generates new samples similar to the original high-dimensional feature vector samples based on the spectral and texture feature patterns of building materials learned during training. For example, when the number of samples of certain building materials in the original dataset is small or samples under certain lighting conditions are missing, the generated new samples can fill these gaps, helping to improve the training effect and performance of the spectrum-texture collaborative enhancement network.

[0046] Furthermore, step S510 also includes step S514, the spectrum-texture collaborative enhancement network also includes a dynamic weight adjustment module, and the dynamic weight adjustment module dynamically adjusts the fusion weight of the spectral features and the texture features according to the facade material type and the real-time lighting conditions; step S515, the defect data weight is enhanced on the defect mapping integrated data according to the fusion weight to obtain the second target mapping integrated data.

[0047] Preferably, the spectral-texture collaborative enhancement network also includes a dynamic weight adjustment module. Since different facade materials have different importance for spectral features and texture features in accurately identifying and extracting building information under different lighting conditions, the dynamic weight adjustment module is used to dynamically adjust the fusion weight of spectral features and texture features according to the facade material type and real-time lighting conditions. For example, the facade of metal material has high reflective properties, and spectral features play a leading role in identifying highly reflective areas. By increasing the weight of spectral features, the spectral reflectance differences of metal materials under different lighting can be more accurately captured, thereby improving the recognition accuracy of highly reflective areas; for facades made of glass, its transparent properties may cause blurred edges in the image, and texture features are more critical for solving the problem of blurred edges of transparent materials. By increasing the weight of texture features, the texture details of the glass surface can be better highlighted, helping to more clearly define the edges of the glass material.

[0048] Preferably, after determining the fusion weights of spectral features and texture features, the defect data weights are enhanced for the defect mapping integrated data. Specifically, according to different fusion weights, the spectral feature part and the texture feature part in the defect mapping integrated data are weighted separately. If the current material is metal and the spectral feature weight is high, the spectral feature part in the data is given a greater weight to enhance its influence in the data; for the texture feature part, it is processed according to the corresponding weight. By adjusting the weights of spectral and texture features, it is possible to highlight important features related to the current facade material and lighting conditions, further optimize the defect mapping integrated data, obtain the second target mapping integrated data, and comprehensively consider the weight changes of spectral and texture features under different facade materials and lighting conditions, and perform more targeted enhancements on the defect part. Compared with the defect mapping integrated data, its quality is higher and it can more accurately reflect the characteristics of the target building, thereby generating a more accurate target building extraction model, which helps to improve the accuracy and reliability of the target building extraction.

[0049] Furthermore, step S500 also includes step S540, if there are multiple target buildings to be extracted, obtaining multiple target surveying and mapping integrated data corresponding to the multiple target buildings; step S550, extracting multiple high-dimensional feature vectors characterizing the facade materials of the multiple target buildings based on the multiple target surveying and mapping integrated data; step S560, inputting the multiple high-dimensional feature vectors into the spectrum-texture collaborative enhancement model to obtain multiple extraction models corresponding to the multiple target buildings.

[0050] Preferably, when there are multiple target buildings to be extracted in the target area, the target surveying and mapping integrated data corresponding to each target building are obtained respectively, wherein the surveying and mapping integrated database contains data collected in parallel from multiple sources (such as optical remote sensing data, laser scanning data, oblique photography data, etc.), and the target surveying and mapping integrated data obtained for each target building can fully reflect the characteristics of the building, such as the location, shape, spectral and texture characteristics of the facade material, etc.; then, the target surveying and mapping integrated data of each target building are extracted to represent the high-dimensional feature vector of the facade material, that is, the spectral feature vector is obtained by analyzing the spectral information, and the surface texture feature vector is extracted from the surface texture information using methods such as the grayscale co-occurrence matrix, and then the two are combined to form a high-dimensional feature vector; for multiple target buildings, multiple corresponding high-dimensional feature vectors are obtained, and each high-dimensional feature vector accurately describes the characteristics of the facade material of the corresponding target building.

[0051] Preferably, the obtained high-dimensional feature vectors of multiple target buildings are input into the spectrum-texture collaborative enhancement model. The model will analyze and process each high-dimensional feature vector, considering the synergistic effect of spectral and texture features and other relevant factors (such as lighting conditions, etc.) to determine the degree of influence of data extraction and other information. It may also be necessary to activate the spectrum-texture collaborative enhancement network to enhance the data. Finally, for each high-dimensional feature vector, a corresponding target building extraction model is generated. The extraction model can accurately identify and extract relevant information of the corresponding target building based on the input data, such as the building's outline, facade material type, etc., thereby ensuring that the relevant information of multiple target buildings can be processed efficiently and accurately.

[0052] In the above, refer to Figure 1 The method for quickly extracting building targets from surveying and mapping data integration according to an embodiment of the present invention is described in detail. Figure 2 A system for rapidly extracting building objects from surveying and mapping data according to an embodiment of the present invention is described.

[0053] The rapid building target extraction system based on surveying and mapping data integration according to the embodiment of the present invention is used to solve the technical problem in the prior art that it is difficult to effectively identify target buildings made of different building materials based on the material reflectance characteristics, resulting in low building recognition rate and poor extraction accuracy. It efficiently generates an accurate extraction model, achieving the technical effect of improving the building target recognition rate and accuracy. Figure 2 As shown, the building target rapid extraction system based on surveying and mapping data integration includes: a surveying and mapping integration database construction module 10, a first target surveying and mapping integration data acquisition module 20, a facade feature vector extraction module 30, a data extraction impact analysis module 40, and a target building extraction model generation module 50.

[0054] A surveying and mapping integrated database construction module 10 is used to integrate multi-source parallel acquisition data to construct a surveying and mapping integrated database for the target area, wherein the target area includes the target building to be extracted; a first target surveying and mapping integrated data acquisition module 20 is used to obtain first target surveying and mapping integrated data of the target building in the surveying and mapping integrated database; a facade feature vector extraction module 30 is used to extract a high-dimensional feature vector characterizing the facade material of the target building based on the first target surveying and mapping integrated data, wherein the high-dimensional feature vector is composed of a spectral feature vector and a surface texture feature vector; a data extraction influence analysis module 40 is used to input the high-dimensional feature vector into a spectrum-texture collaborative enhancement model for data extraction influence analysis and output a first influence coefficient; a target building extraction model generation module 50 is used to activate the spectrum-texture collaborative enhancement network to perform data enhancement on the first target surveying and mapping integrated data if the first influence coefficient is greater than a preset influence coefficient, obtain second target surveying and mapping integrated data, and generate an extraction model of the target building based on the second target surveying and mapping integrated data.

[0055] The specific configuration of the data extraction impact analysis module 40 will be described in detail below. The data extraction impact analysis module 40 further includes: the spectrum-texture collaborative enhancement model determines the facade material type of the target building based on the high-dimensional feature vector; obtains the real-time lighting conditions of the multi-source parallel acquisition data, including real-time lighting intensity and real-time lighting angle; establishes a mapping function relationship between facade material type samples and lighting condition samples to characterize the degree of data extraction influence; analyzes the facade material type and the real-time lighting conditions according to the mapping function relationship, and outputs a first influence coefficient.

[0056] The specific configuration of the target building extraction model generation module 50 will be described in detail below. The target building extraction model generation module 50 further includes: activating a spectral-texture collaborative enhancement network to identify boundary defects in the first target surveying and mapping integrated data and output defect surveying and mapping integrated data; obtaining matching lighting conditions with an influence coefficient less than a preset influence coefficient based on the facade material type; and performing defect data enhancement on the defect surveying and mapping integrated data by calling matching surveying and mapping integrated sample data corresponding to the matching lighting conditions to obtain second target surveying and mapping integrated data.

[0057] The specific configuration of the target building extraction model generation module 50 will be described in detail below. The target building extraction model generation module 50 further includes: obtaining high-dimensional feature vector samples, wherein the high-dimensional feature vector samples include spectral feature vector samples and surface texture feature samples of different building materials under different lighting conditions, wherein the spectral feature vector samples include the band reflectance of the spectral image, and the surface texture feature vector samples include contrast, energy, and correlation parameters calculated by GLCM; obtaining labeled sample data, wherein the labeled sample data includes material category and corresponding annotated surveying and mapping integrated sample data; and performing model training based on the high-dimensional feature vector samples and the labeled sample data to obtain a spectral-texture collaborative enhancement network that has been trained to convergence.

[0058] The specific configuration of the target building extraction model generation module 50 will be described in detail below. The target building extraction model generation module 50 further includes: initializing a generative adversarial network, including initializing a dual-branch generator and a discriminator; wherein, the dual-branch generator includes a spectral branch and a texture branch, the spectral branch simulates the band reflectance changes of different building materials under different lighting conditions through a fully connected layer, and the texture branch simulates the surface texture changes of different building materials under different lighting conditions through a deconvolution layer; after the high-dimensional feature vector sample is input into the dual-branch generator, a generative high-dimensional feature vector is output, the discriminator receives the generative high-dimensional feature vector output by the dual-branch generator and performs adversarial loss feedback training with the high-dimensional feature vector sample to obtain a trained generative adversarial network; calling the trained generative adversarial network to supplement the high-dimensional feature vector sample.

[0059] The specific configuration of the surveying and mapping integrated database construction module 10 will be described in detail below. The surveying and mapping integrated database construction module 10 further includes: if there are multiple target buildings to be extracted, obtaining multiple target surveying and mapping integrated data corresponding to the multiple target buildings; extracting multiple high-dimensional feature vectors representing the facade materials of the multiple target buildings based on the multiple target surveying and mapping integrated data; and inputting the multiple high-dimensional feature vectors into the spectrum-texture collaborative enhancement model to obtain multiple extraction models corresponding to the multiple target buildings.

[0060] The specific configuration of the target building extraction model generation module 50 will be described in detail below. The target building extraction model generation module 50 further includes: the spectrum-texture collaborative enhancement network also includes a dynamic weight adjustment module, which dynamically adjusts the fusion weight of spectral features and texture features based on the facade material type and the real-time lighting conditions; and enhances the defect data weight of the defect mapping integrated data based on the fusion weight to obtain second target mapping integrated data.

[0061] The system for rapid extraction of building targets from surveying and mapping data provided by an embodiment of the present invention can execute the method for rapid extraction of building targets from surveying and mapping data provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0062] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0063] The above specific embodiments do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the scope of protection of this application.

Claims

1. A rapid extraction method for building targets based on surveying and mapping data integration, characterized by: The method comprises: Integrate multi-source parallel acquisition data to build a surveying and mapping integrated database for the target area, wherein the target area includes the target building to be extracted; Acquire first target surveying and mapping integrated data of the target building in the surveying and mapping integrated database; Extracting a high-dimensional feature vector representing the facade material of the target building according to the first target surveying and mapping integrated data, wherein the high-dimensional feature vector is composed of a spectral feature vector and a surface texture feature vector; Inputting the high-dimensional feature vector into a spectrum-texture collaborative enhancement model to perform data extraction impact analysis and output a first impact coefficient; If the first influence coefficient is greater than the preset influence coefficient, the spectrum-texture collaborative enhancement network is activated to perform data enhancement on the first target surveying and mapping integrated data, obtain second target surveying and mapping integrated data, and generate an extraction model of the target building based on the second target surveying and mapping integrated data.

2. The method for rapid extraction of building targets from surveying and mapping data integration according to claim 1, wherein: Inputting the high-dimensional feature vector into the spectrum-texture collaborative enhancement model to perform data extraction impact analysis and outputting a first impact coefficient, the method includes: The spectrum-texture collaborative enhancement model determines the facade material type of the target building according to the high-dimensional feature vector; Acquire real-time illumination conditions of the multi-source parallel collected data, including real-time illumination intensity and real-time illumination angle; A mapping function relationship is established between facade material type samples and lighting condition samples to characterize the degree of influence on data extraction, and the facade material type and the real-time lighting condition are analyzed according to the mapping function relationship to output a first influence coefficient.

3. The method for rapid extraction of building targets from surveying and mapping data integration as claimed in claim 2, characterized in that: Activating the spectrum-texture collaborative enhancement network to perform data enhancement on the first target surveying and mapping integrated data, the method comprising: activating a spectrum-texture collaborative enhancement network to identify boundary defects of the first target surveying and mapping integrated data and output defect surveying and mapping integrated data; According to the type of the facade material, obtaining a matching lighting condition with an influence coefficient less than a preset influence coefficient; The matching surveying and mapping integrated sample data corresponding to the matching lighting condition is called to perform defect data enhancement on the defect surveying and mapping integrated data to obtain second target surveying and mapping integrated data.

4. The method for rapid extraction of building targets from surveying and mapping data integration as claimed in claim 2, characterized in that: Methods for training spectral-texture collaborative enhancement networks include: Obtaining high-dimensional feature vector samples, the high-dimensional feature vector samples including spectral feature vector samples and surface texture feature samples of different building materials under different lighting conditions, the spectral feature vector samples including the band reflectance of the spectral image, and the surface texture feature vector samples including contrast, energy, and correlation parameters calculated by GLCM; Acquire label sample data, wherein the label sample data includes material categories and corresponding labeled surveying and mapping integrated sample data; Model training is performed based on the high-dimensional feature vector samples and the label sample data to obtain a spectrum-texture collaborative enhancement network that has been trained to convergence.

5. The method for rapid extraction of building targets from surveying and mapping data integration as claimed in claim 4, characterized in that: The input end of the spectrum-texture collaborative enhancement network is connected to the generative adversarial network, and the method includes: Initialize the generative adversarial network, including initializing the two-branch generator and discriminator; The dual-branch generator includes a spectral branch and a texture branch. The spectral branch simulates the changes in the band reflectance of different building materials under different lighting conditions through a fully connected layer, and the texture branch simulates the changes in the surface texture of different building materials under different lighting conditions through a deconvolution layer. After the high-dimensional feature vector sample is input into the dual-branch generator, a generative high-dimensional feature vector is output; the discriminator receives the generative high-dimensional feature vector output by the dual-branch generator and performs adversarial loss feedback training on the high-dimensional feature vector sample to obtain a trained generative adversarial network; The trained generative adversarial network is called to supplement the high-dimensional feature vector samples.

6. The method for rapid extraction of building targets from surveying and mapping data integration according to claim 1, wherein: Constructing a surveying and mapping integrated database for the target area, the method further includes: If there are multiple target buildings to be extracted, multiple target surveying and mapping integrated data corresponding to the multiple target buildings are obtained; Extracting a plurality of high-dimensional feature vectors representing the facade materials of the plurality of target buildings according to the plurality of target surveying and mapping integrated data; The multiple high-dimensional feature vectors are input into the spectrum-texture collaborative enhancement model to obtain multiple extraction models corresponding to the multiple target buildings.

7. The method for rapid extraction of building targets from surveying and mapping data integration as claimed in claim 3, characterized in that: The spectrum-texture collaborative enhancement network further includes a dynamic weight adjustment module, which dynamically adjusts the fusion weight of the spectrum feature and the texture feature according to the facade material type and the real-time lighting conditions; Defect data weight enhancement is performed on the defect surveying and mapping integrated data according to the fusion weight to obtain second target surveying and mapping integrated data.

8. The rapid extraction system of building targets based on surveying and mapping data integration is characterized by: The system is used to implement the method for rapidly extracting building targets from surveying and mapping data integration according to any one of claims 1 to 7, and the system comprises: A surveying and mapping integrated database construction module is used to integrate multi-source parallel acquisition data to construct a surveying and mapping integrated database for a target area, wherein the target area includes a target building to be extracted; A first target surveying and mapping integrated data acquisition module is used to acquire first target surveying and mapping integrated data of the target building in the surveying and mapping integrated database; a facade feature vector extraction module, configured to extract a high-dimensional feature vector representing the facade material of the target building based on the first target surveying and mapping integrated data, wherein the high-dimensional feature vector is composed of a spectral feature vector and a surface texture feature vector; A data extraction impact analysis module, configured to input the high-dimensional feature vector into a spectrum-texture collaborative enhancement model to perform data extraction impact analysis and output a first impact coefficient; The target building extraction model generation module is used to activate the spectrum-texture collaborative enhancement network to perform data enhancement on the first target mapping integrated data if the first influence coefficient is greater than the preset influence coefficient, obtain second target mapping integrated data, and generate the extraction model of the target building based on the second target mapping integrated data.

Images