A method and system for three-dimensional modeling of building structure based on construction drawings

Abstract

The present invention provides a method and system for three-dimensional modeling of building structures based on construction drawings, which relates to the technical field of building models, including obtaining two-dimensional line vector data of building construction drawings, preprocessing the data, and forming standardized basic data; correcting the standardized basic data according to the wavelet decomposition principle, selecting building component objects, and constructing three-dimensional geometric models of building components based on the corrected standardized basic data; outlining the basic form of the building structure using a deep learning algorithm based on remote sensing images; performing correlation analysis on the three-dimensional geometric model and the basic form of the building structure, merging overlapping or closely spaced vector points and vector lines, and obtaining a complete three-dimensional modeling of the building structure. The present invention can improve the modeling efficiency of the building model, thereby realizing the effective integration of the construction drawings, facilitating the rapid establishment of three-dimensional modeling of the building structure, and managing the equipment network based on the drawings.

Description

Technical Field

[0001] The present invention relates to the technical field of building models, and in particular to a method and system for three-dimensional modeling of building structures based on construction drawings. Background Art

[0002] Architectural design methods include two-dimensional drawing design and BIM (Building Information Modeling) design. The two-dimensional drawing design method is widely adopted by design institutes and construction units. Specifically, it mainly uses the application of two-dimensional plane design drawings, and the local position is expressed through sections to express the design intent.

[0003] At present, when using Revit software for building information modeling, repetitive and mechanized labor has reduced the production efficiency of engineers to a certain extent, especially in some engineering projects with relatively regular structural layouts. The number of drawings for an engineering project ranges from hundreds to thousands, or as few as dozens. The drawings are large in size, and the existing technology fails to fully utilize the information contained in the architectural drawings in the process of identifying architectural drawings. There are too few objects to be identified, and the details of the architectural drawings are not identified, resulting in low recognition accuracy of architectural objects, and the details of the building are difficult to display, which in turn leads to low accuracy of the generated three-dimensional model of the building. When the building is multi-story, the structures of each floor have both differences and similarities, and the existing technology does not pre-optimize the similar parts of each floor when modeling the building, resulting in a large amount of data to be processed for modeling, which in turn leads to low efficiency in building modeling. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for 3D modeling of building structures based on construction drawings to improve the above problems. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] In a first aspect, the present application provides a method for three-dimensional modeling of a building structure based on construction drawings, comprising:

[0006] Obtaining two-dimensional line vector data of building construction drawings and preprocessing them to form standardized basic data, wherein the standardized basic data includes geometric information and non-geometric information of all components;

[0007] The standardized basic data is modified according to the wavelet decomposition principle, the building component objects are selected, and the three-dimensional geometric model of the building component is constructed based on the modified standardized basic data;

[0008] Using deep learning algorithms based on remote sensing images to outline the basic form of building structures, including the location, size and orientation of the building structures;

[0009] The correlation analysis is carried out on the three-dimensional geometric model and the basic form of the building structure, the overlapping or closely spaced vector points and vector lines are merged, and they are integrated according to the actual geographic space location information to obtain a complete three-dimensional model of the building structure.

[0010] Preferably, the two-dimensional line vector data of the building construction drawing is obtained and preprocessed to form standardized basic data, wherein the standardized basic data includes geometric information and non-geometric information of all components, including:

[0011] The first image in the building construction drawing is obtained, and the first image is processed based on the threshold segmentation algorithm, and the calculation formula is as follows:

[0012]

[0013] In the formula, when the gray value of a pixel point in the input image is Gray(i, j) less than the threshold value T, it is the background, and the gray value of the corresponding pixel point Out(i, j) in the output image is set to 0, that is, black; when the gray value of a pixel point is Gray(i, j) greater than or equal to the threshold value T, it is the background, and the gray value of the corresponding pixel point Out(i, j) in the output image is set to 255, that is, white;

[0014] The processed first image is subjected to Gaussian filtering denoising to obtain a denoised second image, and feature extraction is performed on the second image to obtain word frequency feature information of the building construction drawing, and the word frequency feature information is recorded as standardized basic data, which includes adding category filtering conditions according to the geometric information and non-geometric information of the components, and filtering out components belonging to the same category from all the components; by adding different category filtering conditions, all components can be divided into beams, plates, columns and walls.

[0015] Preferably, the feature extraction of the second image is performed to obtain word frequency feature information of the building construction drawing, which includes:

[0016] Utilizing a scale-invariant feature conversion extraction algorithm, scale-invariant features of all categories of images in a second image are extracted and used as visual vocabulary vectors, wherein the second image includes images of walls, doors and windows, columns, stairs, roofs, and boards;

[0017] The K-Means clustering algorithm is used to merge the words with similar meanings or repeated occurrences in the extracted visual vocabulary vectors to construct a set containing the main standardized basic words;

[0018] quantizing image features of the second image based on a set including main standardized basic words, wherein each image in the second image is replaced by a standardized basic word in the set using a visual word extracted by a scale-invariant feature extraction algorithm;

[0019] The frequencies of occurrence of the main standardized basic words in the second image are counted, and the second image is represented as a K-dimensional numerical vector, namely, the word frequency feature information of the building construction drawing.

[0020] Preferably, the method of outlining the basic form of the building structure based on remote sensing images using a deep learning algorithm, including the location, size and orientation of the building structure, includes:

[0021] Sending a first control command, wherein the first control command is a command for controlling a drone to detect building information, wherein the drone is equipped with a camera system, a distance measurement system, and a GPS positioning system;

[0022] Receive image data sent back by the drone, where the image data includes video information containing ground building structure feature information collected by the camera system and the geographical location of the building structure where the drone is located during the detection process determined by the GPS positioning system;

[0023] Extracting remote sensing images of building structures from image data and establishing a first training set, wherein the first training set includes contour information, style information and size information of the building structures, and annotating them using an annotation tool;

[0024] The first training set is input into the target detection network for multiple iterative calculations to obtain the optimal building structure model, that is, the basic form of the building structure.

[0025] Preferably, the correlation analysis performed on the three-dimensional geometric model and the basic form of the building structure includes:

[0026] The reference sequence for determining the three-dimensional geometric model is recorded as the parent sequence;

[0027] Performing grey correlation analysis on the data information in the three-dimensional geometric model and the basic form of the building structure, including dimensionless processing of the data information in the basic form of the building structure, wherein the dimensionless processing includes averaging and initialization, and obtaining first data information after dimensionless tempering processing;

[0028] Based on the first data information, calculating a grey correlation coefficient value between the parent sequence and the first data information;

[0029] According to the correlation between the three-dimensional geometric model and the data information in the basic form of the building structure calculated by the grey correlation value, the correlation is sorted. The larger the correlation, the better the overlap effect, and then the value is extracted.

[0030] In a second aspect, the present application also provides a building structure three-dimensional modeling system based on construction drawings, comprising:

[0031] Acquisition module: used to acquire the two-dimensional line vector data of the building construction drawings and pre-process them to form standardized basic data, where the standardized basic data includes the geometric information and non-geometric information of all components;

[0032] Construction module: used to modify the standardized basic data according to the wavelet decomposition principle, select the building component object, and construct the three-dimensional geometric model of the building component based on the modified standardized basic data;

[0033] Outlining module: used to outline the basic form of building structures based on remote sensing images using deep learning algorithms, including the location, size and orientation of the building structures;

[0034] Integration module: used to conduct correlation analysis on the three-dimensional geometric model and the basic form of the building structure, merge the overlapping or closely spaced vector points and vector lines, and integrate them according to the actual geographic space location information to obtain a complete three-dimensional model of the building structure.

[0035] In a third aspect, the present application also provides a building structure three-dimensional modeling device based on construction drawings, including:

[0036] Memory for storing computer programs;

[0037] A processor is used to implement the steps of the method for three-dimensional modeling of a building structure based on construction drawings when executing the computer program.

[0038] In a fourth aspect, the present application further provides a readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned method for three-dimensional modeling of a building structure based on construction drawings are implemented.

[0039] The beneficial effects of the present invention are:

[0040] The present invention can improve the modeling efficiency of the building model, thereby realizing the effective integration of the construction drawings, facilitating the rapid establishment of the three-dimensional modeling of the building structure, and managing the equipment network based on the drawings; and the present invention pre-processes the construction drawings, and the generated three-dimensional model of the building has high accuracy and high building modeling efficiency.

[0041] The present invention performs image segmentation on the parts of the construction drawing that are irrelevant to the identification of the first image, so as to facilitate the subsequent processing of the first image and avoid the influence of some irrelevant images on the first image. The optimal segmentation effect is achieved by continuously adjusting the threshold, making the drawing clearer.

[0042] The invention uses Gaussian filtering to reduce noise on the construction drawings to eliminate the influence of image noise. In addition, the edge features of the image can be enhanced after Gaussian filtering, which is beneficial to the subsequent segmentation process.

[0043] The present invention can describe and extract local features in the construction drawings, that is, the second image, through a scale-invariant feature conversion algorithm, and has a certain invariance to image translation, rotation and scaling, and also maintains a certain stability to brightness transformation, affine transformation and image noise; the scale-invariant feature conversion algorithm can be used to perform multi-quantization and rapid extraction of feature vectors of construction drawings, which can meet real-time requirements and is convenient for integration with other feature vectors.

[0044] The present invention adopts the wavelet decomposition principle to represent the building structure vector data and corrects the error, thus laying a solid foundation for the subsequent three-dimensional modeling of the building structure and reducing the error of the building structure data at the same time, thus providing a basis for the subsequent data simplification.

[0045] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 A schematic diagram of a flow chart of a method for three-dimensional modeling of a building structure based on construction drawings described in an embodiment of the present invention;

[0048] Figure 2 It is a schematic diagram of the structure of a building structure three-dimensional modeling system based on construction drawings described in an embodiment of the present invention;

[0049] Figure 3 It is a schematic diagram of the structure of a three-dimensional modeling device for a building structure based on a construction drawing described in an embodiment of the present invention.

[0050] In the figure: 701, acquisition module; 7011, acquisition unit; 7012, processing unit; 70121, extraction unit; 70122, merging unit; 70123, quantification unit; 70124, statistical unit; 702, construction module; 703, outlining module; 7031, sending unit; 7032, receiving unit; 7033, establishment unit; 7034, first calculation unit; 704, integration module; 7041, determination unit; 7042, analysis unit; 7043, second calculation unit; 7044, sorting unit; 800, building structure three-dimensional modeling equipment based on construction drawings; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0052] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0053] Embodiment 1:

[0054] This embodiment provides a method for three-dimensional modeling of a building structure based on construction drawings.

[0055] See also Figure 1 , the figure shows that the method includes step S100, step S200, step S300 and step S400.

[0056] S100, obtaining two-dimensional line vector data of a building construction drawing, and preprocessing the data to form standardized basic data, wherein the standardized basic data includes geometric information and non-geometric information of all components.

[0057] It can be understood that the step S100 includes S101 and S102, wherein:

[0058] S101, obtaining a first image in a building construction drawing, and processing the first image based on a threshold segmentation algorithm, the calculation formula of which is as follows:

[0059]

[0060] In the formula, when the gray value of a pixel point in the input image is Gray(i, j) less than the threshold value T, it is the background, and the gray value of the corresponding pixel point Out(i, j) in the output image is set to 0, that is, black; when the gray value of a pixel point is Gray(i, j) greater than or equal to the threshold value T, it is the background, and the gray value of the corresponding pixel point Out(i, j) in the output image is set to 255, that is, white;

[0061] It should be noted that some of the drawings are backgrounds that are irrelevant to the establishment of the 3D model. Therefore, it is necessary to perform image segmentation on the parts of the construction drawings that are irrelevant to the identification of the first image, so as to process the first image later and avoid the influence of some irrelevant images on the first image. The best segmentation effect is achieved by continuously adjusting the threshold, making the drawings clearer.

[0062] S102, performing Gaussian filtering denoising on the processed first image to obtain a denoised second image, performing feature extraction on the second image to obtain word frequency feature information of the building construction drawing, and recording the word frequency feature information as standardized basic data, which includes adding category screening conditions according to geometric information and non-geometric information of the components, and screening out components belonging to the same category from all the components; by adding different category screening conditions, all components can be divided into beams, plates, columns and walls.

[0063] It should be noted that Gaussian filtering is used for noise reduction to eliminate the influence of image noise. In addition, the edge features of the image can be enhanced after Gaussian filtering, which is conducive to subsequent segmentation processing. According to the geometric information and non-geometric information of the components, category screening is added, including the engineering quantity calculation system for screening enclosure structures, concrete columns, beams, walls, and slabs, so that all parties involved in the project can quickly and accurately obtain engineering quantity and cost data during the project implementation stage, making it possible to finely control engineering costs.

[0064] It should be noted that in step S102, feature extraction is performed on the second image to obtain word frequency feature information of the building construction drawing, which includes S1021, S1022, S1023 and S1024, where:

[0065] S1021, using a scale-invariant feature conversion extraction algorithm to extract scale-invariant features of all categories of images in the second image and use them as visual vocabulary vectors, wherein the second image includes a wall image, a door and window image, a column image, a staircase image, a roof image, and a board image;

[0066] It can be understood that the scale-invariant feature transformation algorithm can describe and extract local features in the construction drawings, that is, the second image, and has a certain invariance to image translation, rotation and scaling, and also maintains a certain stability to brightness transformation, affine transformation and image noise; the scale-invariant feature transformation algorithm can be used to perform multi-quantization and rapid extraction of feature vectors of construction drawings, which can meet real-time requirements and is easy to integrate with other feature vectors.

[0067] S1022, using a K-Means clustering algorithm to merge words with similar meanings or repeated appearances in the extracted visual vocabulary vectors, and construct a set including main standardized basic words;

[0068] It should be noted that since the feature point data extracted from each construction drawing image through the scale-invariant feature transformation algorithm are different, the image is extracted through the bag-of-words model extraction algorithm to ensure that the extracted image features, that is, the dimensions of the repeated words or feature numerical vectors are the same.

[0069] S1023, performing image feature quantization on the second image based on the set including main standardized basic vocabulary, wherein each image in the second image is replaced by a standardized basic vocabulary in the set using a visual vocabulary extracted by a scale-invariant feature extraction algorithm;

[0070] S1024. Count the frequencies of occurrence of main standardized basic words in the second image, and represent the second image as a K-dimensional numerical vector, namely, word frequency feature information of the building construction drawing.

[0071] It should be noted that the K-means clustering algorithm uses K as a parameter to divide the second image set into K categories, and divides them into K categories with the smallest distance from each sample to the category center, that is, the highest similarity. Then the one with the highest similarity is the one that appears most frequently, and they are arranged in order.

[0072] S200, correcting the standardized basic data according to the wavelet decomposition principle, selecting a building component object, and constructing a three-dimensional geometric model of the building component based on the corrected standardized basic data.

[0073] It can be understood that in this step, the wavelet decomposition principle is used to represent the building structure vector data and correct the error, which lays a solid foundation for the subsequent three-dimensional modeling of the building structure and reduces the error of the building structure data, providing a basis for subsequent data simplification.

[0074] S300, based on remote sensing images, uses deep learning algorithms to outline the basic form of the building structure, including the location, size and orientation of the building structure.

[0075] It can be understood that the step S300 includes S301, S302, S303 and S304, wherein:

[0076] S301, sending a first control command, wherein the first control command is a command to control a drone to detect building information, the drone being equipped with a camera system, a distance measurement system, and a GPS positioning system;

[0077] S302, receiving image data transmitted back by the drone, wherein the image data includes video information including ground building structure feature information collected by the camera system and the geographical location of the building structure where the drone is located during the detection process determined by the GPS positioning system;

[0078] S303, intercepting the remote sensing image of the building structure from the image data and establishing a first training set, wherein the first training set includes contour information, style information and size information of the building structure, and annotating using an annotation tool;

[0079] S304: Input the first training set into the target detection network for multiple iterative calculations to obtain an optimal building structure model, that is, a basic form of the building structure.

[0080] It should be noted that the Mask R-CNN network architecture in the target detection network is adopted, and the processed image data is input into the preset neural network for convolution and pooling. Each pixel image in the first training set is processed to obtain the number of set regions of interest, and then the building structure and background are segmented and the border is optimized to obtain the optimized region of interest, which is then classified and regressed through the convolution layer to generate the optimal building structure model.

[0081] S400, performing correlation analysis on the three-dimensional geometric model and the basic form of the building structure, merging overlapping or closely spaced vector points and vector lines, and integrating them according to actual geographic space location information to obtain a complete three-dimensional model of the building structure.

[0082] It can be understood that the step S400 includes S401, S402, S403 and S404, wherein:

[0083] S401, determining a reference sequence of the three-dimensional geometric model as a parent sequence;

[0084] S402, performing grey correlation analysis on the data information in the three-dimensional geometric model and the basic form of the building structure, including performing dimensionless processing on the data information in the basic form of the building structure, wherein the dimensionless processing includes averaging and initialization, to obtain first data information after dimensionless tempering processing;

[0085] S403, calculating a grey correlation coefficient value between the parent sequence and the first data information based on the first data information;

[0086] S404, sorting the correlation between the three-dimensional geometric model and the data information in the basic form of the building structure obtained by calculating the grey correlation value, and extracting the value with a larger correlation value and a better coincidence effect.

[0087] It can be understood that the above steps are to determine the connection between the three-dimensional geometric model and the data information in the basic form of the building structure by calculating the correlation between the two, study the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure (the correlation between the parent sequence and the characteristic sequence), measure the correlation between the data by the correlation (i.e., the correlation), and make auxiliary judgments and decisions. The correlation between the data information in the three-dimensional geometric model and the basic form of the building structure is judged, and the evaluation objects are sorted according to the gray correlation. The correlation sequence of the evaluation objects can be established. The greater the correlation, the better the evaluation result. The better the evaluation result, the better the overlap effect. The overlapping or overlapping or very close vector points and vector lines are merged, and the actual geographic space location information is comprehensively integrated to obtain a complete three-dimensional modeling of the building structure. This method has a small amount of calculation, can ignore the limitations of subjective ideas, is applicable to data in models or forms without rules, can improve the modeling efficiency of the building model, and then realize the effective integration of construction drawings, facilitate the rapid establishment of three-dimensional modeling of building structures, and manage the equipment network based on drawings.

[0088] Embodiment 2:

[0089] like Figure 2 As shown, this embodiment provides a building structure 3D modeling system based on construction drawings, see Figure 2 The system includes an acquisition module 701, a construction module 702, a delineation module 703 and an integration module 704, wherein:

[0090] Acquisition module 701: used to acquire two-dimensional line vector data of building construction drawings, and pre-process the data to form standardized basic data, wherein the standardized basic data includes geometric information and non-geometric information of all components;

[0091] Construction module 702: used to modify the standardized basic data according to the wavelet decomposition principle, select the building component object, and construct the three-dimensional geometric model of the building component based on the modified standardized basic data;

[0092] Outlining module 703: used to outline the basic form of the building structure based on the remote sensing image using a deep learning algorithm, including the location, size and orientation of the building structure;

[0093] Integration module 704: used to perform correlation analysis on the three-dimensional geometric model and the basic form of the building structure, merge the overlapping or closely spaced vector points and vector lines, and integrate them according to the actual geographic space location information to obtain a complete three-dimensional model of the building structure.

[0094] Specifically, the acquisition module 701 includes an acquisition unit 7011 and a processing unit 7012:

[0095] The acquisition unit 7011 is used to acquire the first image in the building construction drawing and process the first image based on the threshold segmentation algorithm, and the calculation formula is as follows:

[0096]

[0097] In the formula, when the gray value of a pixel point in the input image is Gray(i, j) less than the threshold value T, it is the background, and the gray value of the corresponding pixel point Out(i, j) in the output image is set to 0, that is, black; when the gray value of a pixel point is Gray(i, j) greater than or equal to the threshold value T, it is the background, and the gray value of the corresponding pixel point Out(i, j) in the output image is set to 255, that is, white;

[0098] like Figure 2 As shown, the processing unit 7012 of the present invention is used to perform Gaussian filtering and denoising on the processed first image to obtain a denoised second image, perform feature extraction on the second image to obtain word frequency feature information of the building construction drawing, and record the word frequency feature information as standardized basic data, which includes adding category filtering conditions according to the geometric information and non-geometric information of the components, and filtering out components belonging to the same category from all the components; by adding different category filtering conditions, all components can be divided into beams, plates, columns and walls.

[0099] Specifically, the processing unit 7012 includes an extraction unit 70121, a merging unit 70122, a quantization unit 70123 and a statistical unit 70124, wherein:

[0100] Extraction unit 70121: used to extract scale-invariant features of all categories of images in the second image using a scale-invariant feature conversion extraction algorithm and use them as visual vocabulary vectors, wherein the second image includes wall images, door and window images, column images, staircase images, roof images and board images;

[0101] Merging unit 70122: used to merge words with similar meanings or repeated appearances in the extracted visual vocabulary vectors by using K-Means clustering algorithm, and construct a set including main standardized basic words;

[0102] Quantization unit 70123: used for quantizing image features of the second image based on a set including main standardized basic words, wherein each image in the second image is replaced by a standardized basic word in the set using a visual word extracted by a scale-invariant feature extraction algorithm;

[0103] Statistical unit 70124: used to count the frequency of occurrence of main standardized basic vocabulary in the second image, and represent the second image as a K-dimensional numerical vector, that is, the word frequency feature information of the building construction drawing.

[0104] Specifically, the outlining module 703 includes a sending unit 7031, a receiving unit 7032, a establishing unit 7033 and a first calculating unit 7034, wherein:

[0105] The sending unit 7031 is used to send a first control command, wherein the first control command is a command to control the drone to detect building information, and the drone is equipped with a camera system, a distance measurement system and a GPS positioning system;

[0106] Receiving unit 7032: used to receive image data sent back by the drone, wherein the image data includes video information containing ground building structure feature information collected by the camera system and the geographical location of the building structure where the drone is located during the detection process determined by the GPS positioning system;

[0107] Establishing unit 7033: used to capture the remote sensing image of the building structure from the image data and establish a first training set, wherein the first training set includes the outline information, style information and size information of the building structure, and annotates it using an annotation tool;

[0108] The first calculation unit 7034 is used to input the first training set into the target detection network for multiple iterative calculations to obtain the optimal building structure model, that is, the basic form of the building structure.

[0109] Specifically, the integration unit includes a determination unit 7041, an analysis unit 7042, a second calculation unit 7043 and a sorting unit 7044, wherein:

[0110] Determining unit 7041: A reference sequence used to determine the three-dimensional geometric model is recorded as a parent sequence;

[0111] The analysis unit 7042 is used to perform grey correlation analysis on the data information in the three-dimensional geometric model and the basic form of the building structure, including dimensionless processing of the data information in the basic form of the building structure, wherein the dimensionless processing includes averaging and initialization to obtain first data information after dimensionless toughening processing;

[0112] The second calculation unit 7043 is used to calculate the grey correlation coefficient value between the mother sequence and the first data information based on the first data information;

[0113] Sorting unit 7044: used to sort the correlation between the three-dimensional geometric model and the data information in the basic form of the building structure calculated according to the gray correlation value, and the larger the correlation, the better the overlap effect, and then the extraction is performed.

[0114] It should be noted that, regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0115] Embodiment 3:

[0116] Corresponding to the above method embodiment, this embodiment also provides a three-dimensional modeling device for building structures based on construction drawings. The three-dimensional modeling device for building structures based on construction drawings described below and the three-dimensional modeling method for building structures based on construction drawings described above can refer to each other.

[0117] Figure 3 FIG. 8 is a block diagram of a building structure 3D modeling device 800 based on construction drawings according to an exemplary embodiment. Figure 3 As shown, the building structure 3D modeling device 800 based on construction drawings includes: a processor 801 and a memory 802. The building structure 3D modeling device 800 based on construction drawings also includes one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0118] The processor 801 is used to control the overall operation of the building structure three-dimensional modeling device 800 based on construction drawings to complete all or part of the steps in the above-mentioned building structure three-dimensional modeling method based on construction drawings. The memory 802 is used to store various types of data to support the operation of the building structure three-dimensional modeling device 800 based on construction drawings. These data may include, for example, instructions for any application or method operated on the building structure three-dimensional modeling device 800 based on construction drawings, and application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 802 or sent via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules can be a keyboard, a mouse or buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the building structure three-dimensional modeling device 800 based on the construction drawings and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 can include: a Wi-Fi module, a Bluetooth module or an NFC module.

[0119] In an exemplary embodiment, the building structure three-dimensional modeling device 800 based on construction drawings can be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, referred to as ASIC), digital signal processors (Digital Signal Processing Device, referred to as DSP), digital signal processing devices (Digital Signal Processing Device, referred to as DSPD), programmable logic devices (Programmable Logic Device, referred to as PLD), field programmable gate arrays (Field Programmable Gate Array, referred to as FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned building structure three-dimensional modeling method based on construction drawings.

[0120] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, and when the program instructions are executed by a processor, the steps of the above-mentioned method for 3D modeling of a building structure based on construction drawings are implemented. For example, the computer-readable storage medium may be the above-mentioned memory 802 including program instructions, and the above-mentioned program instructions may be executed by the processor 801 of the device 800 for 3D modeling of a building structure based on construction drawings to complete the above-mentioned method for 3D modeling of a building structure based on construction drawings.

[0121] Embodiment 4:

[0122] Corresponding to the above method embodiment, a readable storage medium is also provided in this embodiment. The readable storage medium described below and the three-dimensional modeling method of building structure based on construction drawings described above can refer to each other.

[0123] A computer program is stored on the readable storage medium, and when the computer program is executed by the processor, the steps of the method for three-dimensional modeling of a building structure based on construction drawings of the above method embodiment are implemented.

[0124] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0125] In summary, the present invention can improve the modeling efficiency of the building model, thereby realizing the effective integration of the construction drawings, facilitating the rapid establishment of the three-dimensional modeling of the building structure, and managing the equipment network based on the drawings; and the present invention pre-processes the construction drawings, and the generated three-dimensional model of the building has high accuracy and high building modeling efficiency; the present invention uses Gaussian filtering to reduce noise on the construction drawings to eliminate the influence of image noise. In addition, the edge features of the image can be enhanced after Gaussian filtering, which is beneficial to the subsequent segmentation processing; the present invention can have a certain invariance to the image translation, rotation and scaling in the construction drawings through the scale-invariant feature conversion algorithm, and also maintain a certain stability to the brightness transformation, affine transformation and image noise; it can meet the real-time requirements and is convenient to merge with other feature vectors; it lays a solid foundation for the subsequent three-dimensional modeling of the building structure, and at the same time reduces the error of the building structure data, providing a basis for the subsequent data simplification.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0127] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for three-dimensional modeling of a building structure based on construction drawings, characterized in that: include: Obtaining two-dimensional line vector data of building construction drawings and preprocessing them to form standardized basic data, wherein the standardized basic data includes geometric information and non-geometric information of all components; The standardized basic data is modified according to the wavelet decomposition principle, the building component objects are selected, and the three-dimensional geometric model of the building component is constructed based on the modified standardized basic data; Using deep learning algorithms based on remote sensing images to outline the basic form of building structures, including the location, size and orientation of the building structures; Conduct correlation analysis on the 3D geometric model and the basic form of the building structure, merge the overlapping or closely spaced vector points and vector lines, and integrate them according to the actual geographic space location information to obtain a complete 3D model of the building structure; The method of using deep learning algorithms based on remote sensing images to outline the basic form of the building structure, including the location, size and orientation of the building structure, includes: Sending a first control command, wherein the first control command is a command for controlling a drone to detect building information, wherein the drone is equipped with a camera system, a distance measurement system, and a GPS positioning system; Receive image data sent back by the drone, where the image data includes video information containing ground building structure feature information collected by the camera system and the geographical location of the building structure where the drone is located during the detection process determined by the GPS positioning system; Extracting remote sensing images of building structures from image data and establishing a first training set, wherein the first training set includes contour information, style information and size information of the building structures, and annotating them using an annotation tool; The first training set is input into the target detection network for multiple iterative calculations to obtain the optimal building structure model, that is, the basic form of the building structure; The Mask R-CNN network architecture in the target detection network is used to input the processed image data into the preset neural network for convolution and pooling. Each pixel image in the first training set is processed to obtain the number of regions of interest, segment the building structure and the background, and optimize the border to obtain the optimized region of interest. The convolution layer is used for classification and regression processing to generate the optimal building structure model. The correlation analysis between the three-dimensional geometric model and the basic form of the building structure includes: The reference sequence for determining the three-dimensional geometric model is recorded as the parent sequence; Performing grey correlation analysis on the data information in the three-dimensional geometric model and the basic form of the building structure, including dimensionless processing of the data information in the basic form of the building structure, wherein the dimensionless processing includes averaging and initialization, and obtaining first data information after dimensionless tempering processing; Based on the first data information, calculating a grey correlation coefficient value between the parent sequence and the first data information; According to the correlation between the three-dimensional geometric model and the data information in the basic form of the building structure calculated by the grey correlation value, the correlation is sorted, and the value with a larger correlation has a better coincidence effect, and then it is extracted; Among them, by calculating the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure, the connection between the two is determined, and then the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure is studied, and the degree of correlation between the data is measured by the correlation degree, so as to assist in decision-making; the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure is judged, and the evaluation objects are sorted according to the size of the gray correlation degree, and a correlation sequence of the evaluation objects is established. The greater the correlation degree, the better the evaluation result, the better the evaluation result, and the better the coincidence effect. The overlapping or coincident or very close vector points and vector lines are merged, and comprehensive integration is carried out according to the actual geographic space location information to obtain a complete three-dimensional modeling of the building structure.

2. The method for three-dimensional modeling of a building structure based on construction drawings according to claim 1, characterized in that: The two-dimensional line vector data of the building construction drawing is obtained and preprocessed to form standardized basic data, wherein the standardized basic data includes geometric information and non-geometric information of all components, including: The first image in the building construction drawing is obtained, and the first image is processed based on the threshold segmentation algorithm, and the calculation formula is as follows: In the formula, the gray value of a pixel in the input image is Gray(i, j) When it is less than the threshold T, it is the background, and the pixel corresponding to the output image Out(i,j) The gray value of a pixel is set to 0, that is, black; when the gray value of a pixel is Gray(i, j) When it is greater than or equal to the threshold T, it is the background, and the pixel corresponding to the output image Out(i,j) The grayscale value is set to 255, which is white; The processed first image is subjected to Gaussian filtering denoising to obtain a denoised second image, and feature extraction is performed on the second image to obtain word frequency feature information of the building construction drawing, and the word frequency feature information is recorded as standardized basic data, which includes adding category filtering conditions according to the geometric information and non-geometric information of the components, and filtering out components belonging to the same category from all the components; by adding different category filtering conditions, all components can be divided into beams, plates, columns and walls.

3. The method for three-dimensional modeling of a building structure based on construction drawings according to claim 2, characterized in that: The feature extraction of the second image is performed to obtain word frequency feature information of the building construction drawing, including: Utilizing a scale-invariant feature conversion extraction algorithm, scale-invariant features of all categories of images in a second image are extracted and used as visual vocabulary vectors, wherein the second image includes images of walls, doors and windows, columns, stairs, roofs, and boards; The K-Means clustering algorithm is used to merge the words with similar meanings or repeated occurrences in the extracted visual vocabulary vectors to construct a set containing the main standardized basic words; quantizing image features of the second image based on a set including main standardized basic words, wherein each image in the second image is replaced by a standardized basic word in the set using a visual word extracted by a scale-invariant feature extraction algorithm; The frequencies of occurrence of the main standardized basic words in the second image are counted, and the second image is represented as a K-dimensional numerical vector, namely, the word frequency feature information of the building construction drawing.

4. A three-dimensional modeling system for building structures based on construction drawings, characterized in that: include: Acquisition module: used to acquire the two-dimensional line vector data of the building construction drawings and pre-process them to form standardized basic data, where the standardized basic data includes the geometric information and non-geometric information of all components; Construction module: used to modify the standardized basic data according to the wavelet decomposition principle, select the building component object, and construct the three-dimensional geometric model of the building component based on the modified standardized basic data; Outlining module: used to outline the basic form of building structures based on remote sensing images using deep learning algorithms, including the location, size and orientation of the building structures; Integration module: used to conduct correlation analysis on the 3D geometric model and the basic form of the building structure, merge the overlapping or closely spaced vector points and vector lines, and integrate them according to the actual geographic space location information to obtain a complete 3D modeling of the building structure; The outline module includes: A sending unit: used for sending a first control command, wherein the first control command is a command for controlling a drone to detect building information, wherein the drone is provided with a camera system, a distance measurement system and a GPS positioning system; Receiving unit: used to receive the image data sent back by the drone, where the image data includes the video information containing the characteristic information of the ground building structure collected by the camera system and the geographical location of the building structure where the drone is located during the detection process determined by the GPS positioning system; Establishment unit: used for intercepting the remote sensing image of the building structure from the image data and establishing a first training set, wherein the first training set includes the outline information, style information and size information of the building structure, and annotates it using an annotation tool; The first calculation unit is used to input the first training set into the target detection network for multiple iterative calculations to obtain the optimal building structure model, that is, the basic form of the building structure; The integration module includes: Determine unit: The reference sequence used to determine the three-dimensional geometric model is recorded as the parent sequence; Analysis unit: used for performing grey correlation analysis on the data information in the three-dimensional geometric model and the basic form of the building structure, including dimensionless processing of the data information in the basic form of the building structure, wherein the dimensionless processing includes averaging and initialization to obtain the first data information after dimensionless toughening processing; A second calculation unit is used to calculate a grey correlation coefficient value between the mother sequence and the first data information based on the first data information; Sorting unit: used to sort the correlation between the three-dimensional geometric model and the data information in the basic form of the building structure calculated according to the gray correlation value, and the larger the correlation, the better the overlap effect, and then the value is extracted; Among them, by calculating the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure, the connection between the two is determined, and then the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure is studied, and the degree of correlation between the data is measured by the correlation degree, so as to assist in decision-making; the correlation between the data information in the three-dimensional geometric model and the basic form of the building structure is judged, and the evaluation objects are sorted according to the size of the gray correlation degree, and a correlation sequence of the evaluation objects is established. The greater the correlation degree, the better the evaluation result, the better the evaluation result, and the better the coincidence effect. The overlapping or coincident or very close vector points and vector lines are merged, and comprehensive integration is carried out according to the actual geographic space location information to obtain a complete three-dimensional modeling of the building structure.

5. The building structure three-dimensional modeling system based on construction drawings according to claim 4 is characterized in that: The acquisition module includes: The acquisition unit is used to acquire the first image in the building construction drawing and process the first image based on the threshold segmentation algorithm. The calculation formula is as follows: In the formula, the gray value of a pixel in the input image is Gray(i, j) When it is less than the threshold T, it is the background, and the pixel corresponding to the output image Out(i,j) The gray value of a pixel is set to 0, that is, black; when the gray value of a pixel is Gray(i, j) When it is greater than or equal to the threshold T, it is the background, and the pixel corresponding to the output image Out(i,j) The grayscale value is set to 255, which is white; Processing unit: used for performing Gaussian filtering denoising on the processed first image to obtain a denoised second image, performing feature extraction on the second image to obtain word frequency feature information of the building construction drawing, and recording the word frequency feature information as standardized basic data, including adding category screening conditions according to the geometric information and non-geometric information of the components, and screening out components belonging to the same category from all the components; by adding different category screening conditions, all components can be divided into beams, plates, columns and walls.

6. The building structure three-dimensional modeling system based on construction drawings according to claim 5 is characterized in that: The processing unit includes: Extraction unit: used for extracting scale-invariant features of all categories of images in the second image using a scale-invariant feature conversion extraction algorithm and using them as visual vocabulary vectors, wherein the second image includes images of walls, doors and windows, columns, stairs, roofs and boards; Merging unit: used to merge the words with similar meanings or repeated appearances in the extracted visual vocabulary vectors using the K-Means clustering algorithm to construct a set containing the main standardized basic words; A quantization unit: configured to quantize image features of the second image based on a set comprising main standardized basic words, wherein each image in the second image is replaced by a standardized basic word in the set using a visual word extracted by a scale-invariant feature extraction algorithm; Statistical unit: used to count the frequency of occurrence of main standardized basic words in the second image, and represent the second image as a K-dimensional numerical vector, that is, the word frequency feature information of the building construction drawing.

Images