Municipal engineering visual management system and method thereof

Abstract

The invention relates to the technical field of municipal engineering, and particularly discloses a municipal engineering visual management system and method, and the method comprises the steps: constructing a lightweight panoramic three-dimensional model of a municipal engineering building based on laser point cloud data; thirdly, further introducing a data processing technology based on deep learning to perform feature extraction and feature selection on the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model of the municipal engineering so as to mine key three-dimensional space structure features in the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model; and characteristic difference analysis is carried out on the two, so that intelligent inference and visual management of the municipal engineering construction progress are realized. According to the invention, real-time and dynamic monitoring of the municipal engineering progress can be realized, and a visual construction progress view is provided for a manager, so that the manager can adjust and optimize a construction plan according to the construction progress, and the engineering can be completed on time.

Description

Technical Field

[0001] The present application relates to the field of municipal engineering technology, and more specifically, to a municipal engineering visualization management system and method thereof. Background Art

[0002] With the acceleration of urbanization, municipal engineering, as the core part of urban infrastructure construction, has become increasingly complex and large-scale, which has put forward higher requirements for project management. The traditional municipal engineering management model relies on two-dimensional drawings and manual on-site monitoring, which is not only inefficient, but also difficult to fully and accurately reflect the actual situation of the project, especially in large and complex projects. This limitation is particularly obvious. Therefore, exploring and applying advanced visual management methods has become the key to improving the management level of municipal engineering.

[0003] In recent years, 3D scanning technology, especially 3D modeling technology based on laser point cloud, has received extensive attention in the field of municipal engineering due to its high precision and high efficiency. Laser point cloud data can quickly capture detailed information such as topography and building structure at the construction site, and can generate high-precision 3D models after processing with professional software. On this basis, the introduction of BIM (Building Information Modeling) technology has brought revolutionary changes to municipal engineering management. BIM models not only contain geometric information of buildings, but also integrate multi-dimensional information such as materials, costs, and construction progress, realizing the full life cycle management of engineering projects.

[0004] However, although 3D modeling technology and BIM models can provide visual engineering information, in actual application, how to use 3D modeling technology and BIM models for efficient and accurate visualization management of municipal engineering construction progress is still an important issue currently faced. Especially when dealing with large-scale and complex municipal engineering projects, how to reflect the construction progress in real time and dynamically, and timely warn of potential problems so that managers can quickly understand the project status and make decisions is a problem that needs to be solved urgently.

[0005] Therefore, an optimized municipal engineering visualization management system and method thereof are needed to solve the above problems. Summary of the invention

[0006] The present application provides a municipal engineering visual management system and method thereof, which can realize real-time and dynamic monitoring of the progress of municipal engineering projects, and provide managers with an intuitive construction progress view, so that managers can adjust and optimize the construction plan according to the construction progress, and ensure that the project can be completed on time.

[0007] In a first aspect, a method for visual management of municipal engineering is provided, comprising: Obtain laser point cloud data of municipal engineering buildings; Based on the laser point cloud data of the municipal engineering building, a panoramic three-dimensional model of the lightweight building is constructed; Extract the target municipal engineering BIM model; Determine the construction progress based on the panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project; Visually displaying the panoramic three-dimensional model of the lightweight building, the BIM model of the target municipal engineering project and the construction progress on a smart terminal; Among them, based on the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model, the construction progress is determined, including: respectively extracting the architectural model features of the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model to obtain a municipal engineering under-construction building model feature map and a municipal engineering target building model feature map; performing feature selection based on local neighborhood perception on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map; extracting semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map; based on the municipal engineering building model differential semantic coding feature map, the construction progress is determined.

[0008] Preferably, respectively extracting the architectural model features of the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model to obtain a municipal engineering under-construction building model feature map and a municipal engineering target building model feature map, including: The panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project are respectively input into a building model feature extractor based on a three-dimensional convolutional neural network model to obtain a model feature map of the municipal engineering building under construction and a model feature map of the municipal engineering target building.

[0009] Preferably, the feature map of the municipal engineering building model under construction and the feature map of the municipal engineering target building model are respectively subjected to feature selection based on local neighborhood perception to obtain a sparse municipal engineering building model under construction feature map and a sparse municipal engineering target building model feature map, including: Performing feature decoupling and feature flattening on the feature graph of the municipal engineering building under construction model to obtain a set of local feature vectors of the municipal engineering building under construction model; Based on the feature importance of each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction, the set of local feature vectors of the municipal engineering building model under construction is serialized and reorganized to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction; Based on the neighborhood feature activity of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, feature selection and sparse aggregation are performed on the descending sequence of the local feature vectors of the municipal engineering building model under construction to obtain the sparse municipal engineering building model feature map.

[0010] Preferably, based on the feature importance of each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction, the set of local feature vectors of the municipal engineering building model under construction is serialized and reorganized to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction, including: Inputting each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction into an importance measurement module to obtain a set of local feature importance score values ​​of the municipal engineering building model under construction; Based on the set of local feature importance score values ​​of the municipal engineering building models under construction, the set of local feature vectors of the municipal engineering building models under construction is arranged in descending order to obtain a descending sequence of the local feature vectors of the municipal engineering building models under construction.

[0011] Preferably, based on the neighborhood feature activity of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, feature selection and sparse aggregation are performed on the descending sequence of the local feature vectors of the municipal engineering building model under construction to obtain the sparse municipal engineering building model feature map, including: Based on the neighborhood features of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the feature neighborhood activity of each local feature vector of the municipal engineering building model under construction is calculated to obtain a sequence of local feature neighborhood activity of the municipal engineering building model under construction; Based on the sequence of local feature neighborhood activity of the municipal engineering building model under construction, feature selection is performed on the descending sequence of local feature vectors of the municipal engineering building model under construction to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction after selection; The descending sequence of the local feature vectors of the selected municipal engineering building under construction model is reshaped to obtain the sparse municipal engineering building under construction model feature map.

[0012] Preferably, based on the neighborhood features of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the feature neighborhood activity of each local feature vector of the municipal engineering building model under construction is calculated to obtain a sequence of local feature neighborhood activity of the municipal engineering building model under construction, including: Extracting a first local feature vector of a model of a building under construction for a municipal engineering project and a second local feature vector of a model of a building under construction for a municipal engineering project from a descending sequence of local feature vectors of the model of a building under construction for a municipal engineering project; Respectively calculating the internal characteristic attenuation factors of the local characteristic vector of the first municipal engineering building under construction model and the local characteristic vector of the second municipal engineering building under construction model to obtain the local characteristic attenuation factor of the first municipal engineering building under construction model and the local characteristic attenuation factor of the second municipal engineering building under construction model; The ratio of the local feature attenuation factor of the second municipal engineering building model under construction to the local feature attenuation factor of the first municipal engineering building model under construction is calculated to obtain the feature neighborhood activity of the local feature vector of the first municipal engineering building model under construction.

[0013] Preferably, extracting semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map includes: The position difference between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map is calculated to obtain the municipal engineering building model differential semantic coding feature map.

[0014] Preferably, determining the construction progress based on the differential semantic coding feature map of the municipal engineering building model includes: The differential semantic coding feature map of the municipal engineering building model is input into a decoder-based construction progress measurement module to obtain the construction progress.

[0015] In a second aspect, a municipal engineering visual management system is provided, comprising: Point cloud data acquisition module, used to obtain laser point cloud data of municipal engineering buildings; A building panoramic three-dimensional model construction module is used to construct a lightweight building panoramic three-dimensional model based on the laser point cloud data of the municipal engineering building; A target municipal engineering BIM model extraction module is used to extract the target municipal engineering BIM model; A construction progress determination module, used to determine the construction progress based on the panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project; A visualization display module, used for visually displaying the panoramic three-dimensional model of the lightweight building, the BIM model of the target municipal engineering project and the construction progress on an intelligent terminal; Among them, the construction progress determination module includes: a building model feature extraction unit, which is used to extract the building model features of the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model respectively to obtain the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map; a building model feature selection unit, which is used to perform feature selection based on local neighborhood perception on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map respectively to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map; a semantic difference feature extraction unit, which is used to extract the semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map; a construction progress inference unit, which is used to determine the construction progress based on the municipal engineering building model differential semantic coding feature map.

[0016] Preferably, the building model feature extraction unit is used to: The panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project are respectively input into a building model feature extractor based on a three-dimensional convolutional neural network model to obtain a model feature map of the municipal engineering building under construction and a model feature map of the municipal engineering target building.

[0017] This application has at least the following technical effects: The present application provides a municipal engineering visualization management system and method thereof, which first constructs a lightweight panoramic three-dimensional model of a municipal engineering building based on laser point cloud data, and then further introduces deep learning-based data processing technology to perform feature extraction and feature selection on the lightweight panoramic three-dimensional model of the municipal engineering building and the target municipal engineering BIM model, so as to mine the key three-dimensional spatial structural features in the lightweight panoramic three-dimensional model of the building and the target municipal engineering BIM model, and realizes intelligent inference and visualization management of the construction progress of the municipal engineering by performing feature difference analysis on the two, so as to realize real-time and dynamic monitoring of the progress of the municipal engineering, and provide managers with an intuitive construction progress view, so that managers can adjust and optimize the construction plan according to the construction progress, and ensure that the project can be completed on time. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0019] Figure 1 This is a schematic flowchart of the municipal engineering visual management method according to an embodiment of the present application.

[0020] Figure 2 This is a data flow diagram of the municipal engineering visual management method according to an embodiment of the present application.

[0021] Figure 3 This is a schematic flowchart of step S4 in the municipal engineering visual management method of an embodiment of the present application.

[0022] Figure 4 This is a schematic flowchart of step S42 in the municipal engineering visual management method of an embodiment of the present application.

[0023] Figure 5 This is a schematic flowchart of step S423 in the municipal engineering visual management method of an embodiment of the present application.

[0024] Figure 6 This is a schematic block diagram of the municipal engineering visual management system according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0026] It should be noted that the acquisition and processing of all information or data in this application are carried out in compliance with the relevant national data protection laws and policies and with the authorization given by the owner of the corresponding device.

[0027] In the technical solution of this application, if Figure 1 and Figure 2As shown, the municipal engineering visualization management method includes: S1, acquiring laser point cloud data of municipal engineering buildings; S2, constructing a panoramic three-dimensional model of a lightweight building based on the laser point cloud data of the municipal engineering building; S3, extracting a target municipal engineering BIM model; S4, determining a construction progress based on the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model; S5, visually displaying the panoramic three-dimensional model of the lightweight building, the target municipal engineering BIM model and the construction progress on a smart terminal.

[0028] For example, in step S1, laser point cloud data of a municipal engineering building is obtained. It should be understood that the laser point cloud data is obtained by scanning the municipal engineering building with a laser scanning device, contains a large amount of three-dimensional coordinate information on the surface of the building, and can accurately reflect the shape and structure of the building.

[0029] In one embodiment, obtaining laser point cloud data of a municipal engineering building includes: first, selecting a suitable laser scanning device, which depends on the specific needs of the project, such as scale, accuracy requirements, and budget. Generally, a terrestrial laser scanner (TLS) is suitable for fixed-position scanning that requires high accuracy, while an airborne laser scanner (ALS) or a mobile laser scanning system (MLS) is more suitable for rapid scanning of large areas or difficult-to-reach areas. Before starting the scan, site preparation must be done and appropriate safety measures must be taken to ensure that the construction site is safe and all relevant parties are notified. Next, during the on-site laser scanning process, the operator will adjust the scanning parameters, such as resolution and coverage, according to the project requirements to ensure that sufficient details can be obtained while maintaining work efficiency. For larger engineering projects, scanning from multiple locations may be required to ensure that all key parts of the structure are captured. Once the scan is completed, the generated raw point cloud data file will contain the spatial coordinate information (x, y, z) of each measured point, as well as additional attributes such as reflection intensity. At the same time, relevant metadata needs to be recorded, including information such as the scanning time, the model of the equipment used, and its parameter settings.

[0030] Exemplarily, in step S2, a lightweight panoramic three-dimensional model of the building is constructed based on the laser point cloud data of the municipal engineering building. It should be understood that considering that the amount of point cloud data is usually very large, direct three-dimensional modeling will consume a lot of computing resources and time. Therefore, in this application, lightweight technology is used to pre-process and simplify the laser point cloud data to construct a lightweight panoramic three-dimensional model of the building that can maintain the main structural features of the building and greatly reduce the amount of data, thereby improving the rendering speed and interactive performance of the model while ensuring the accuracy of the model, and providing more efficient data support for subsequent construction progress inference.

[0031] In one embodiment, after the original point cloud data is obtained, it needs to be preliminarily cleaned to remove noise points (such as outliers caused by scanning equipment or environmental factors) and unnecessary background information (such as surrounding trees, vehicles, etc.). This step can significantly improve the speed and accuracy of subsequent processing. Next, the point cloud is simplified by using uniform sampling, grid sampling or feature retention-based methods to reduce the number of points without losing important geometric characteristics. For flat areas, the point density can be appropriately reduced, while maintaining a high resolution at the edges and details. Then, the discrete point cloud is converted into a continuous surface representation using a surface reconstruction algorithm. Common methods include Delaunay triangulation, Poisson reconstruction, etc. These algorithms can generate polygonal mesh models that approximate the surface of the actual object based on the spatial relationship in the point cloud. In order to further reduce the data burden of the model, model compression techniques such as vertex merging, polygon simplification, and LOD (Level of Detail) hierarchical detail management are applied. At the same time, consider removing invisible faces and retaining only the parts facing the observer, thereby speeding up rendering and improving interactive performance. On the basis of ensuring the lightweight of the model, deep learning or other machine learning algorithms can also be used to identify and annotate the key features of the building (such as door and window locations, cornice lines, etc.), so that the generated 3D model not only looks realistic, but also has rich semantic information, which is convenient for subsequent analysis and application. Finally, the processed lightweight 3D model is integrated into the visualization management system.

[0032] Exemplarily, in step S3, the target municipal engineering BIM model is extracted. It should be understood that in order to obtain the target design information of the municipal engineering building, so as to compare and analyze it with the current lightweight building panoramic three-dimensional model and realize the intelligent evaluation of the municipal engineering construction progress, the present application further extracts the target municipal engineering BIM model. The target municipal engineering BIM model is constructed based on the design drawings and related parameters of the municipal engineering using professional BIM software, and includes the geometric structure, material properties, equipment information and other aspects of the municipal engineering building, providing key data support for subsequent construction progress evaluation and visual management.

[0033] Exemplarily, in step S4, the construction progress is determined based on the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model. It should be understood that the lightweight building panoramic three-dimensional model is a reflection of the actual situation of the current municipal engineering site, while the target municipal engineering BIM model represents the design goal and expected state of the municipal engineering. By comparing and analyzing the two, the deviation and progress of the construction process can be intuitively revealed.

[0034] In one embodiment, Figure 3As shown, based on the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model, the construction progress is determined, including: S41, respectively extracting the architectural model features of the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model to obtain a municipal engineering under-construction building model feature map and a municipal engineering target building model feature map; S42, performing feature selection based on local neighborhood perception on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map; S43, extracting the semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map; S44, determining the construction progress based on the municipal engineering building model differential semantic coding feature map.

[0035] Exemplarily, in step S41, the architectural model features of the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model are extracted respectively to obtain the model feature map of the building under construction of the municipal engineering and the model feature map of the target building of the municipal engineering. It should be understood that, considering that the structure of the municipal engineering building is complex and contains numerous details, such as the pipeline system and structural support inside the building, the traditional overlapping comparison or simple geometric shape matching method is often difficult to accurately capture the differences in the details of the building, and the difference analysis effect is even more limited for the building components and complex structures of non-standard geometric shapes. Therefore, in the present application, a three-dimensional convolutional neural network model is used to construct a building model feature extractor, so as to utilize its powerful feature learning and representation capabilities, and to perform convolution operations on the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model by sliding the convolution kernel in three-dimensional space, so as to capture the spatial structural features of the model, generate the model feature map of the building under construction of the municipal engineering and the model feature map of the target building of the municipal engineering, so as to more accurately reflect the actual structure of the building.

[0036] In one embodiment, the architectural model features of the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model are respectively extracted to obtain the model feature map of the municipal engineering building under construction and the model feature map of the municipal engineering target building, including: inputting the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model into the architectural model feature extractor based on the three-dimensional convolutional neural network model to obtain the model feature map of the municipal engineering building under construction and the model feature map of the municipal engineering target building. Specifically, in the training process of the three-dimensional convolutional neural network model, a large amount of municipal engineering architectural model data is first collected, including three-dimensional models of different types of buildings (such as bridges, road ancillary buildings, drainage facilities buildings, etc.), and at the same time, it is ensured that the collected three-dimensional model data covers different construction stages, such as models from the foundation construction stage to the main structure completion stage, and then to the internal facilities installation stage, so that the model can learn the feature differences under different construction progress.

[0037] Next, each building model is annotated in detail. The annotation content includes the type of building components in the model (such as beams, columns, walls, etc.), materials (such as concrete, steel, wood, etc.), sizes, and connection relationships with other components, so as to ensure that the model can accurately understand the structural characteristics of the building. At the same time, the annotated building model data is divided into training set, validation set and test set, for example, it can be divided according to the ratio of 7:2:1 or 8:1:1. Then, the three-dimensional convolutional neural network model is trained using the divided training set data so that it can learn the spatial structural characteristics of the building model under different construction progress. During the training process, the performance of the model is optimized by continuously adjusting the parameters of the three-dimensional convolutional neural network model, such as the weight parameter matrix of the convolution kernel, so that it can more accurately extract the characteristics of the building. At the same time, the performance of the three-dimensional convolutional neural network model is regularly evaluated using the validation set data during the training process to prevent overfitting and ensure that the model has good generalization ability. When the performance of the model on the validation set reaches the preset standard, the training is stopped, and the performance of the model is tested using the test set data to verify the performance and accuracy of the model. In this way, it can be ensured that the trained three-dimensional convolutional neural network model can effectively extract the structural features of municipal engineering buildings and provide a more reliable data basis for subsequent construction progress inference.

[0038] Exemplarily, in step S42, feature selection based on local neighborhood perception is performed on the municipal engineering building model feature map under construction and the municipal engineering target building model feature map, respectively, to obtain a sparse municipal engineering building model feature map under construction and a sparse municipal engineering target building model feature map. It should be understood that, considering that the municipal engineering building model feature map under construction and the municipal engineering target building model feature map usually have a large number of redundant features that do not directly contribute to the construction progress inference, such as subtle size differences of non-critical structural parts, etc., which not only increase the calculation complexity, but also may interfere with the accuracy of the construction progress inference. Therefore, in the present application, feature selection is further performed on the municipal engineering building model feature map under construction and the municipal engineering target building model feature map to remove unnecessary information, reduce the amount of calculation and improve the pertinence of the analysis.

[0039] In one embodiment, Figure 4 As shown, the feature map of the municipal engineering building model under construction and the feature map of the municipal engineering target building model are respectively subjected to feature selection based on local neighborhood perception to obtain a sparse municipal engineering building model under construction feature map and a sparse municipal engineering target building model feature map, including: S421, feature decoupling and feature flattening of the feature map of the municipal engineering building model under construction to obtain a set of local feature vectors of the municipal engineering building model under construction; S422, based on the feature importance of each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction, serializing and reorganizing the set of local feature vectors of the municipal engineering building model under construction to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction; S423, based on the neighborhood feature activity of each local feature vector of the municipal engineering building model under construction in the descending sequence of local feature vectors of the municipal engineering building model under construction, feature selection and sparse aggregation are performed on the descending sequence of local feature vectors of the municipal engineering building model under construction to obtain the sparse feature map of the municipal engineering building model under construction.

[0040] Exemplarily, in step S421, feature decoupling and feature flattening are performed on the feature graph of the municipal engineering building model under construction to obtain a set of local feature vectors of the municipal engineering building model under construction. Specifically, the process can be expressed by the formula: in, represents a characteristic diagram of a building model under construction of the municipal engineering project, Indicates feature decoupling, , , and They represent the first, second, and third dimension of the characteristic graph of the building model under construction in the municipal engineering along the channel dimension. and The local feature matrix of the building model under construction in the municipal engineering project, is the number of channels of the characteristic graph of the building model under construction of the municipal engineering project, Indicates feature flattening. , , and Respectively indicate the , , and stated The corresponding local feature vector of the municipal engineering building under construction model.

[0041] Specifically, taking the feature map of the model of the municipal engineering building under construction as an example, first, in order to perform fine-grained feature selection, the feature map of the model of the municipal engineering building under construction is first subjected to feature decoupling and feature flattening along the channel dimension, so as to decompose the feature map into a set of local feature vectors of the model of the municipal engineering building under construction, ensuring that each local feature vector represents a different independent component, and providing a data basis for subsequent feature selection. Specifically, the process of performing feature selection based on local neighborhood perception on the feature map of the municipal engineering target building model to obtain a sparse municipal engineering target building model feature map can refer to the process of processing the feature map of the model of the municipal engineering building under construction.

[0042] In one embodiment, in step S422, based on the feature importance of each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction, the set of local feature vectors of the municipal engineering building model under construction is serialized and reorganized to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction, including: inputting each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction into an importance measurement module to obtain a set of local feature importance score values ​​of the municipal engineering building model under construction. Specifically, the process can be expressed by the formula: in, and Represent the weight parameter matrix and bias term respectively, represents the importance score reference vector, represents the matrix multiplication operation, Indicates the Importance score of local features of the municipal engineering building under construction model.

[0043] Based on the set of importance score values ​​of the local features of the municipal engineering building model under construction, the set of local feature vectors of the municipal engineering building model under construction is arranged in descending order to obtain a descending sequence of the local feature vectors of the municipal engineering building model under construction. Specifically, the process can be expressed by the formula: in, Represents a sort operation, , , and The first, second, and third in the descending sequence of local feature vectors of the building model under construction in the municipal engineering project and The local feature vector of the model of a building under construction in a municipal engineering project.

[0044] That is, the feature importance of each local feature vector of each municipal engineering building model under construction is evaluated through a neural network to reveal the importance of each model's local features for construction progress inference. Then, according to the obtained local feature importance score, the local feature vectors of each municipal engineering building model under construction are sorted from high to low in importance to form a descending sequence of local feature vectors of municipal engineering building models under construction. In this way, a clear view of the relative importance of features can be provided without changing the feature content, so as to facilitate the subsequent feature selection process.

[0045] In one embodiment, Figure 5 As shown, in step S423, based on the neighborhood feature activity of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the descending sequence of the local feature vectors of the municipal engineering building model under construction is subjected to feature selection sparse aggregation to obtain the sparse municipal engineering building model under construction feature graph, including: S4231, based on the neighborhood features of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the feature neighborhood activity of each local feature vector of the municipal engineering building model under construction is calculated to obtain a sequence of local feature neighborhood activities of the municipal engineering building model under construction. Specifically, the process can be expressed by the formula: in, Indicates the The characteristic scale value of Indicates the Middle The eigenvalues ​​of the positions, and They represent the descending sequence of the local feature vectors of the building model under construction of the municipal engineering project. and The internal characteristic attenuation factor of the local characteristic vector of the municipal engineering under-construction building model, Indicates the The corresponding local feature neighborhood activity of the municipal engineering under-construction building model.

[0046] S4232, based on the sequence of the neighborhood activity of the local feature of the municipal engineering building model under construction, feature selection is performed on the descending sequence of the local feature vectors of the municipal engineering building model under construction to obtain a descending sequence of the local feature vectors of the municipal engineering building model under construction after selection. Specifically, the process can be expressed by the formula: in, represents the feature selection operation, is the preset threshold, , , and They represent the first, second, and third in the descending sequence of the local feature vectors of the municipal engineering under-construction building model after selection. and After the selection, the local feature vector of the municipal engineering under-construction building model is It is the number of feature vectors in the descending sequence of the local feature vectors of the model of the building under construction in the municipal engineering project after the selection.

[0047] S4233, reshaping the descending sequence of the selected local feature vectors of the municipal engineering building under construction model to obtain the sparse municipal engineering building under construction model feature map. Specifically, the process can be expressed by the formula: in, represents the feature shape reshaping, Represents the characteristic map of the sparse municipal engineering under-construction building model.

[0048] That is, in one embodiment, based on the neighborhood characteristics of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the characteristic neighborhood activity of each local feature vector of the municipal engineering building model under construction is calculated to obtain a sequence of local feature neighborhood activities of the municipal engineering building model under construction, including: extracting the first local feature vector of the municipal engineering building model under construction and the second local feature vector of the municipal engineering building model under construction from the descending sequence of the local feature vectors of the municipal engineering building model under construction; calculating the internal feature attenuation factors of the first local feature vector of the municipal engineering building model under construction and the second local feature vector of the municipal engineering building model under construction respectively to obtain the first local feature attenuation factor of the municipal engineering building model under construction and the second local feature attenuation factor of the municipal engineering building model under construction; calculating the ratio between the second local feature attenuation factor of the municipal engineering building model under construction and the first local feature attenuation factor of the municipal engineering building model under construction to obtain the characteristic neighborhood activity of the first local feature vector of the municipal engineering building model under construction.

[0049] Specifically, considering that in a building model, a single feature point carries limited information, and local feature combinations can often more comprehensively and accurately reflect key information such as the building's structural characteristics, spatial relationships, and construction status. That is, although some local feature representations may have a low importance score, their combination with neighboring features may have a significant impact on the inference of the construction progress. Therefore, in the feature selection process, this application further introduces the concept of feature neighborhood activity, and calculates the feature neighborhood activity of each municipal engineering under-construction building model by calculating the degree of interaction between the local feature vector and the neighboring features, revealing the strength of the correlation between the two, thereby providing a more comprehensive perspective for selecting the most effective feature representation by considering the interaction between features.

[0050] Then, based on the obtained feature neighborhood activity, the most valuable local feature vectors of the municipal engineering building model under construction are further screened out, and through feature shape reshaping, they are reorganized back to the original spatial structure to form a sparse municipal engineering building model feature map under construction, maintaining the consistency and compatibility of the model architecture. In this way, feature selection no longer relies solely on the importance score of individual features, but comprehensively considers the correlation between the feature and its neighbors, which helps to more accurately capture the feature representation that has an important impact on the construction progress inference, reduce the interference of redundant information, and improve the accuracy and efficiency of construction progress inference.

[0051] Exemplarily, in step S43, the semantic difference features between the sparse municipal engineering building under construction model feature map and the sparse municipal engineering target building model feature map are extracted to obtain a municipal engineering building model differential semantic coding feature map. It should be understood that in order to further quantify the difference between the building under construction and the target building at the feature level, the present application calculates the positional difference between the sparse municipal engineering building under construction model feature map and the sparse municipal engineering target building model feature map to highlight the structural completion of the building under construction relative to the target building, and generates a municipal engineering building model differential semantic coding feature map, thereby providing a clear data basis for the measurement of construction progress.

[0052] In one embodiment, the semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map are extracted to obtain a differential semantic coding feature map of the municipal engineering building model, including: calculating the positional difference between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain the differential semantic coding feature map of the municipal engineering building model.

[0053] Exemplarily, in step S44, the construction progress is determined based on the differential semantic coding feature map of the municipal engineering building model. In one embodiment, the construction progress is determined based on the differential semantic coding feature map of the municipal engineering building model, including: inputting the differential semantic coding feature map of the municipal engineering building model into a decoder-based construction progress measurement module to obtain the construction progress. That is, in order to convert the difference information between the building under construction and the target building into an intuitive construction progress value. The present application further uses a decoder model to construct a construction progress measurement module, perform feature decoding on the differential semantic coding feature map of the municipal engineering building model, restore the actual meaning represented by the model difference information contained therein, and convert it into a quantitative representation of the construction progress, so as to facilitate project managers to clearly understand the progress status of the project.

[0054] Optionally, in one embodiment of the present application, the differential semantic coding feature map of the municipal engineering building model is input into a construction progress measurement module based on a decoder to obtain the construction progress, including: expanding the differential semantic coding feature map of the municipal engineering building model according to the row vector dimension to obtain a differential semantic coding feature vector of the municipal engineering building model; using a decoder to decode and regress the differential semantic coding feature vector of the municipal engineering building model according to the following formula to obtain the construction progress; Wherein, the formula is: ,in, is the differential semantic encoding feature vector of the municipal engineering building model, is the construction progress, is the weight matrix, Represents matrix multiplication.

[0055] In a preferred example, the differential semantic coding feature map of the municipal engineering building model is passed through a decoder-based construction progress measurement module to generate the construction progress, which includes: Determine the number of non-zero eigenvalues ​​in the differential semantic coding feature map of the municipal engineering building model, and use the scale of the differential semantic coding feature map of the municipal engineering building model, i.e., the width Multiply by height Multiply by the number of channels Subtract the number of non-zero eigenvalues ​​to obtain the zero-dimensional value of the differential semantic encoding of the municipal engineering building model ; Calculate the zero-differentiable value of the differential semantic encoding of the municipal engineering building model after subtracting one from the zero-dimensional value of the differential semantic encoding of the municipal engineering building model Multiply and divide respectively by the zero-dimensional value of the differential semantic encoding of the municipal engineering building model The first municipal engineering building model differential semantic coding field fitting value obtained later and the fitting value of differential semantic coding field of the second municipal engineering building model ; The square root of the sum of squares of all eigenvalues ​​of the differential semantic coding feature map of the municipal engineering building model is calculated to obtain the differential semantic coding mode representation value of the municipal engineering building model: in, Represents the differential semantic encoding feature map of the municipal engineering building model The eigenvalues ​​of the positions, , , Respectively represent the width, height, and number of channels of the differential semantic coding feature map of the municipal engineering building model, Represents the differential semantic coding mode representation value of the municipal engineering building model; Differential semantic encoding of zero differentiable values ​​using the municipal engineering building model Calculate each feature value of the differential semantic encoding feature map of the municipal engineering building model as an index The power function , and multiplied by the first municipal engineering building model differential semantic coding field fitting value To obtain the local representation value of the differential semantic encoding of the municipal engineering building model ; Calculate each eigenvalue of the differential semantic encoding feature map of the municipal engineering building model The fitting value of the differential semantic coding field with the second municipal engineering building model and the differential semantic encoding mode representation value of the municipal engineering building model The product of gets the center pointing value of the differential semantic encoding of the municipal engineering building model ; The municipal engineering building model differential semantic encoding local representation value Differential semantic encoding center orientation value of the municipal engineering building model The weighted difference Composition-optimized differential semantic encoding feature map of municipal engineering building model; The optimized municipal engineering building model differential semantic coding feature map is passed through a decoder-based construction progress measurement module to generate the construction progress.

[0056] Among them, when the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map respectively represent the model semantic features of the municipal engineering under-construction building model and the model semantic features of the municipal engineering target building model, although the feature selection based on feature neighborhood activity can enhance the feature sparsity of both and reduce the amount of calculation, it will also cause the pixel granularity information between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to be misaligned, thereby causing the positional differential expression richness of the municipal engineering building model differential semantic coding feature map, resulting in repetition of unstructured probability mapping, affecting the accuracy of the decoding result.

[0057] Therefore, based on the vector field fitting law of high-dimensional manifolds, the isolated zeros in the differential semantic coding feature map of the municipal engineering building model are used as the differentiable dimension of the high-dimensional manifold, so as to fix the normal direction pointing to the distribution center near the local coordinates represented by the eigenvalues ​​of the differential semantic coding feature map of the municipal engineering building model. In this way, the invalid repetition in the probability mapping process caused by the unstructured characteristics of the rich distribution of the differential semantic coding feature map of the municipal engineering building model is avoided by effectively aligning the feature distribution pattern of the differential semantic coding feature map of the municipal engineering building model with the probability density differentiable field, thereby improving the accuracy of the construction progress obtained by the construction progress measurement module based on the decoder input.

[0058] Exemplarily, in step S5, the panoramic three-dimensional model of the lightweight building, the BIM model of the target municipal engineering project and the construction progress are visualized on the smart terminal. It should be understood that the panoramic three-dimensional model of the lightweight building, the BIM model of the target municipal engineering project and the construction progress are visualized on the smart terminal. Such a visualization tool can greatly improve the efficiency and transparency of project management. By intuitively displaying this information, managers can quickly understand the status of the current project, including completed parts, work in progress, and tasks planned for the future. This not only helps to monitor the progress of the project in real time, but also helps to identify potential problems or bottlenecks, allowing for timely adjustment of construction strategies, optimizing resource allocation, and ensuring that the entire project is completed on time.

[0059] In one embodiment, the intelligent terminal integrates a data interface to obtain the latest lightweight building panoramic three-dimensional model and BIM model constructed by laser point cloud data from the back-end server. The data of these two types of models are usually stored in a cloud database, and the front-end application obtains the updated model file through API calls and renders it. In order to ensure user experience, the technologies involved here may include but are not limited to WebGL, Three.js or other 3D rendering technologies suitable for mobile devices. For the display of construction progress, the specific value of the current construction progress can be clearly displayed by adding a dynamic progress bar at the key position of the three-dimensional model or directly marking the percentage value. The progress bar can be arranged along the main axis of the building, or placed on a specific work surface, making it easier for users to associate progress data with the actual physical location.

[0060] In summary, the method for visual management of municipal engineering according to the embodiment of the present application is explained, which first constructs a lightweight panoramic three-dimensional model of a municipal engineering building based on laser point cloud data, and then further introduces data processing technology based on deep learning to perform feature extraction and feature selection on the lightweight panoramic three-dimensional model of the municipal engineering building and the target municipal engineering BIM model, so as to mine the key three-dimensional spatial structural features in the lightweight panoramic three-dimensional model of the building and the target municipal engineering BIM model, and by analyzing the feature differences between the two, the intelligent inference and visual management of the construction progress of the municipal engineering can be realized. In this way, real-time and dynamic monitoring of the progress of the municipal engineering project can be achieved, providing managers with an intuitive view of the construction progress, so that managers can adjust and optimize the construction plan according to the construction progress to ensure that the project can be completed on time.

[0061] Figure 6 Schematic diagram of the municipal engineering visual management system of the embodiment of the present application. Figure 6As shown, the municipal engineering visualization management system 100 includes: a point cloud data acquisition module 110, which is used to acquire laser point cloud data of municipal engineering buildings; a building panoramic three-dimensional model construction module 120, which is used to construct a lightweight building panoramic three-dimensional model based on the laser point cloud data of the municipal engineering building; a target municipal engineering BIM model extraction module 130, which is used to extract the target municipal engineering BIM model; a construction progress determination module 140, which is used to determine the construction progress based on the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model; a visualization display module 150, which is used to visualize the lightweight building panoramic three-dimensional model, the target municipal engineering BIM model and the construction progress on a smart terminal.

[0062] In one embodiment, the construction progress determination module includes: a building model feature extraction unit, which is used to extract the building model features of the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model respectively to obtain a municipal engineering under-construction building model feature map and a municipal engineering target building model feature map; a building model feature selection unit, which is used to perform feature selection based on local neighborhood perception on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map respectively to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map; a semantic difference feature extraction unit, which is used to extract the semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map; a construction progress inference unit, which is used to determine the construction progress based on the municipal engineering building model differential semantic coding feature map.

[0063] In one embodiment, the building model feature extraction unit is used to: input the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model into a building model feature extractor based on a three-dimensional convolutional neural network model respectively to obtain a model feature map of the municipal engineering under-construction building and a model feature map of the municipal engineering target building.

[0064] The specific operations of each module and unit in the above municipal engineering visual management system have been referenced above. Figures 1 to 5 The description of the visual management method of municipal engineering projects has been introduced in detail, and therefore, its repeated description will be omitted.

[0065] An embodiment of the present application further provides a computer program product, which includes a computer program code. When the computer program code runs on a computer, the computer implements the methods in the above embodiments of the present application.

[0066] An embodiment of the present application further provides a computer-readable storage medium, which stores computer instructions. When the computer instructions are executed on a computer, the computer implements the methods in the above embodiments of the present application.

[0067] An embodiment of the present application also provides a chip, including a circuit, for executing the methods in the above embodiments of the present application.

[0068] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0069] The prefixes such as "first" and "second" used in the embodiments of the present application are only used to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. The use of prefixes such as ordinal numbers used to distinguish description objects in the embodiments of the present application does not constitute a limitation on the described objects. For the statement of the described objects, please refer to the description in the context of the claims or embodiments, and the use of such prefixes should not constitute an unnecessary limitation.

[0070] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0071] In the various embodiments of the present application, unless otherwise specified or logically conflicting, the terms and / or descriptions between the various embodiments are consistent and can be referenced to each other, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0072] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0073] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0074] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A visual management method for municipal engineering projects, characterized in that: include: Obtain laser point cloud data of municipal engineering buildings; Based on the laser point cloud data of the municipal engineering building, a panoramic three-dimensional model of the lightweight building is constructed; Extract the target municipal engineering BIM model; Determine the construction progress based on the panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project; Visually displaying the panoramic three-dimensional model of the lightweight building, the BIM model of the target municipal engineering project and the construction progress on a smart terminal; Wherein, based on the panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project, the construction progress is determined, including: Extracting architectural model features of the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model respectively to obtain a model feature map of the municipal engineering building under construction and a model feature map of the municipal engineering target building; Performing feature selection based on local neighborhood perception on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map respectively to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map; Extracting semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map; Determining the construction progress based on the differential semantic coding feature map of the municipal engineering building model; The feature selection based on local neighborhood perception is performed on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map, including: Performing feature decoupling and feature flattening on the feature graph of the municipal engineering building under construction model to obtain a set of local feature vectors of the municipal engineering building under construction model; Based on the feature importance of each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction, the set of local feature vectors of the municipal engineering building model under construction is serialized and reorganized to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction; Based on the neighborhood feature activity of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, feature selection and sparse aggregation are performed on the descending sequence of the local feature vectors of the municipal engineering building model under construction to obtain the sparse municipal engineering building model feature map.

2. The method for visual management of municipal engineering according to claim 1, characterized in that: Extracting architectural model features of the panoramic three-dimensional model of the lightweight building and the target municipal engineering BIM model respectively to obtain a model feature map of the municipal engineering building under construction and a model feature map of the municipal engineering target building, including: The panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project are respectively input into a building model feature extractor based on a three-dimensional convolutional neural network model to obtain a model feature map of the municipal engineering building under construction and a model feature map of the municipal engineering target building.

3. The method for visual management of municipal engineering according to claim 2, characterized in that: Based on the feature importance of each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction, the set of local feature vectors of the municipal engineering building model under construction is serialized and reorganized to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction, including: Inputting each local feature vector of the municipal engineering building model under construction in the set of local feature vectors of the municipal engineering building model under construction into an importance measurement module to obtain a set of local feature importance score values ​​of the municipal engineering building model under construction; Based on the set of local feature importance score values ​​of the municipal engineering building models under construction, the set of local feature vectors of the municipal engineering building models under construction is arranged in descending order to obtain a descending sequence of the local feature vectors of the municipal engineering building models under construction.

4. The method for visual management of municipal engineering according to claim 3, characterized in that: Based on the neighborhood feature activity of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, feature selection and sparse aggregation are performed on the descending sequence of the local feature vectors of the municipal engineering building model under construction to obtain the sparse municipal engineering building model feature map, including: Based on the neighborhood features of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the feature neighborhood activity of each local feature vector of the municipal engineering building model under construction is calculated to obtain a sequence of local feature neighborhood activity of the municipal engineering building model under construction; Based on the sequence of local feature neighborhood activity of the municipal engineering building model under construction, feature selection is performed on the descending sequence of local feature vectors of the municipal engineering building model under construction to obtain a descending sequence of local feature vectors of the municipal engineering building model under construction after selection; The descending sequence of the local feature vectors of the selected municipal engineering building under construction model is reshaped to obtain the sparse municipal engineering building under construction model feature map.

5. The method for visual management of municipal engineering according to claim 4, characterized in that: Based on the neighborhood features of each local feature vector of the municipal engineering building model under construction in the descending sequence of the local feature vectors of the municipal engineering building model under construction, the feature neighborhood activity of each local feature vector of the municipal engineering building model under construction is calculated to obtain a sequence of local feature neighborhood activity of the municipal engineering building model under construction, including: Extracting a first local feature vector of a model of a building under construction for a municipal engineering project and a second local feature vector of a model of a building under construction for a municipal engineering project from a descending sequence of local feature vectors of the model of a building under construction for a municipal engineering project; Respectively calculating the internal characteristic attenuation factors of the local characteristic vector of the first municipal engineering building under construction model and the local characteristic vector of the second municipal engineering building under construction model to obtain the local characteristic attenuation factor of the first municipal engineering building under construction model and the local characteristic attenuation factor of the second municipal engineering building under construction model; The ratio of the local feature attenuation factor of the second municipal engineering building model under construction to the local feature attenuation factor of the first municipal engineering building model under construction is calculated to obtain the feature neighborhood activity of the local feature vector of the first municipal engineering building model under construction.

6. The method for visual management of municipal engineering according to claim 5, characterized in that: Extracting semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map, including: The position difference between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map is calculated to obtain the municipal engineering building model differential semantic coding feature map.

7. The method for visual management of municipal engineering according to claim 6, characterized in that: Determining the construction progress based on the differential semantic coding feature map of the municipal engineering building model includes: The differential semantic coding feature map of the municipal engineering building model is input into a decoder-based construction progress measurement module to obtain the construction progress.

8. Municipal engineering visual management system, including: Point cloud data acquisition module, used to obtain laser point cloud data of municipal engineering buildings; A building panoramic three-dimensional model construction module is used to construct a lightweight building panoramic three-dimensional model based on the laser point cloud data of the municipal engineering building; A target municipal engineering BIM model extraction module is used to extract the target municipal engineering BIM model; A construction progress determination module, used to determine the construction progress based on the panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project; A visualization display module, used for visually displaying the panoramic three-dimensional model of the lightweight building, the BIM model of the target municipal engineering project and the construction progress on an intelligent terminal; Among them, the construction progress determination module includes: a building model feature extraction unit, which is used to extract the building model features of the lightweight building panoramic three-dimensional model and the target municipal engineering BIM model respectively to obtain the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map; a building model feature selection unit, which is used to perform feature selection based on local neighborhood perception on the municipal engineering under-construction building model feature map and the municipal engineering target building model feature map respectively to obtain a sparse municipal engineering under-construction building model feature map and a sparse municipal engineering target building model feature map; a semantic difference feature extraction unit, which is used to extract the semantic difference features between the sparse municipal engineering under-construction building model feature map and the sparse municipal engineering target building model feature map to obtain a municipal engineering building model differential semantic coding feature map; a construction progress inference unit, which is used to determine the construction progress based on the municipal engineering building model differential semantic coding feature map.

9. The municipal engineering visual management system according to claim 8, characterized in that: The building model feature extraction unit is used to: The panoramic three-dimensional model of the lightweight building and the BIM model of the target municipal engineering project are respectively input into a building model feature extractor based on a three-dimensional convolutional neural network model to obtain a model feature map of the municipal engineering building under construction and a model feature map of the municipal engineering target building.

Images