Underground pipe gallery model segmented loading method, system, equipment and medium

Abstract

The invention relates to the technical field of power system simulation, and discloses an underground pipe gallery model segmented loading method, system and device and a medium, and the method comprises the steps: loading panoramic image data of a target underground pipe gallery to a three-dimensional model of the target underground pipe gallery, and constructing a panoramic three-dimensional model; then, the panoramic three-dimensional model is cut into a plurality of sub-models in the extension direction, and layering processing is carried out according to the layering number and the maximum compression ratio of the sub-models; and for a part needing to be analyzed in the target underground pipe gallery, selecting at least one layer in the multiple layers of sub-models at the corresponding position in the panoramic three-dimensional model for loading and rendering. According to the layering method, the layering number and the maximum compression ratio of the sub-models are comprehensively considered, the performance requirement of equipment and the workload of loading and rendering are effectively reduced, and meanwhile compatible operation of the mobile terminal on the underground pipe gallery model is facilitated. Therefore, the data volume is greatly reduced, the processing efficiency is improved, and the labor cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of power system simulation, and particularly to a method, system, device and medium for segmentally loading an underground pipe gallery model. Background Art

[0002] With the increasing complexity of urban infrastructure, urban power systems are basically constructed in the form of underground pipe gallery pipelines. As the construction of underground pipelines increases, the management and maintenance of underground pipelines become more and more complex. Using digital means, it is becoming increasingly important to globally monitor and manage the operation data of the pipe gallery and refine the management of data such as abnormal alarm risks of the pipe gallery.

[0003] Currently, there are the following problems in the digital management method of the pipe gallery: The underground pipe gallery is managed by combining a BIM model with a panoramic view. The BIM model has problems such as high dependence on computers and high consumption of device performance, and it is difficult to run compatibly through mobile devices. When fusing the panoramic view and the BIM model, there are problems such as large data volume, low processing efficiency, and high labor cost input, which easily makes it difficult to widely use the three-dimensional model of the underground pipe gallery conveniently and at low cost. Summary of the Invention

[0004] In view of this, the present invention provides a method, system, device and medium for segmentally loading an underground pipe gallery model, which solves the technical problems mentioned in the above background art that the underground pipe gallery model is difficult to run compatibly through mobile devices, and when fusing the panoramic view and the BIM model, there are problems such as large data volume, low processing efficiency, and high labor cost input.

[0005] The first aspect of the present invention provides a method for segmentally loading an underground pipe gallery model, including:

[0006] Loading the panoramic image data of the target underground pipe gallery into the three-dimensional model of the target underground pipe gallery to obtain a panoramic three-dimensional model;

[0007] Cutting the panoramic three-dimensional model into multiple sub-models in the extension direction according to preset sub-model segmentation parameters, where the preset sub-model segmentation parameters include the length, number of layers, and maximum compression ratio of each sub-model;

[0008] For each sub-model, layering the sub-model according to the number of layers and the maximum compression ratio of the sub-model to obtain multi-layer sub-models;

[0009] According to the part to be analyzed in the target underground pipe gallery, retrieving the multi-layer sub-models corresponding to the position of the part to be analyzed from the panoramic three-dimensional model, and loading and rendering at least one layer of the retrieved multi-layer sub-models.

[0010] Preferably, loading the panoramic image data of the target underground utility tunnel into the three-dimensional model of the target underground utility tunnel to obtain a panoramic three-dimensional model includes:

[0011] Collecting the point cloud data of the target underground utility tunnel, filtering the point cloud data, and generating the three-dimensional model of the target underground utility tunnel according to the filtered point cloud data;

[0012] Matching the geographical location coordinates of the panoramic image data with the three-dimensional coordinates of the three-dimensional model of the target underground utility tunnel;

[0013] Extracting the panoramic feature parameters in the panoramic image data of the target underground utility tunnel, and loading the panoramic feature parameters into the three-dimensional model of the target underground utility tunnel according to the matching result to obtain a panoramic three-dimensional model.

[0014] Preferably, generating the three-dimensional model of the target underground utility tunnel according to the filtered point cloud data includes:

[0015] Regarding each point in the filtered point cloud data as a vertex to form a vertex array;

[0016] Dividing the vertex array based on the Delaunay triangulation algorithm to obtain a triangular face array;

[0017] Projecting each vertex in the vertex array onto a two-dimensional plane to generate UV coordinates, and creating a UV texture map array based on the UV coordinates;

[0018] Combining the vertex array, the triangular face array, and the UV texture map array to generate the three-dimensional model of the target underground utility tunnel.

[0019] Preferably, this method further includes: optimizing the sub-model segmentation parameters; the optimizing the sub-model segmentation parameters includes:

[0020] Randomly generating multiple groups of sub-model segmentation parameters to form an initial population, and each group of the sub-model segmentation parameters is an individual in the initial population;

[0021] Based on the genetic algorithm, optimizing the initial population according to a preset fitness value to obtain the individual with the optimal preset fitness value, and determining the optimal sub-model segmentation parameters according to the individual with the optimal preset fitness value;

[0022] Among them, the preset fitness value is obtained by weighted calculation according to each preset evaluation index corresponding to the subsection parameters of each group of the sub-models; among them, the preset evaluation indexes include the time consumed by the sub-model layering and subsection operations, the total storage capacity required for each layer of sub-models after layering and subsection, the loading and rendering time of each layer of sub-models after layering and subsection, and the loading and rendering satisfaction of each layer of sub-models after layering and subsection.

[0023] Preferably, the step of cutting the panoramic three-dimensional model into a plurality of sub-models according to the preset sub-model subsection parameters includes:

[0024] Taking the vertex at one end of the panoramic three-dimensional model as the starting point, cutting along the extension direction of the panoramic three-dimensional model according to the length of each sub-model to generate a plurality of cutting surfaces;

[0025] Segmenting the panoramic three-dimensional model based on the plurality of cutting surfaces to generate a plurality of sub-models;

[0026] The step of segmenting the panoramic three-dimensional model based on the plurality of cutting surfaces to generate a plurality of sub-models includes:

[0027] For any subsection of the sub-model, extracting the vertices between the two cutting surfaces adjacent to the left and right of the current subsection of the sub-model from the vertex array as the vertex array of the current subsection of the sub-model, and extracting the UV coordinates corresponding to the vertices in the vertex array of the current subsection of the sub-model from the UV texture map array to obtain the UV texture map array of the current subsection of the sub-model, and extracting at least two triangular faces whose vertices are between the two cutting surfaces adjacent to the left and right of the current subsection of the sub-model from the triangular face array to obtain the triangular face array of the current subsection of the sub-model;

[0028] Combining the vertex array, the triangular face array and the UV texture map array of the current subsection of the sub-model to generate the current subsection of the sub-model.

[0029] Preferably, the step of layering each sub-model according to the layering number and the maximum compression ratio of the sub-model to obtain multiple layers of sub-models includes:

[0030] Determining the contraction ratio of the vertices in each sub-model according to the layering number and the maximum compression ratio;

[0031] Performing vertex contraction on the sub-model according to the contraction ratio to obtain the first layer of sub-model;

[0032] Performing vertex contraction on the first layer of sub-model according to the contraction ratio to obtain the second layer of sub-model;

[0033] Vertex contraction is performed on the second - layer sub - model according to the contraction ratio to obtain a third - layer sub - model, and so on until multi - layer sub - models with the same number of layers as the stratification number are obtained.

[0034] Preferably, the vertex contraction of the sub - model according to the contraction ratio includes:

[0035] Determine the normal angle between two adjacent triangular faces according to the normal directions of each triangular face in the sub - model;

[0036] Obtain two common vertices in two adjacent triangular faces and determine the distance between the two common vertices; wherein, the common vertices are the vertices where adjacent triangular faces overlap, and the two common vertices are respectively the vertex closer to the front and the vertex closer to the back along the extension direction of the panoramic three - dimensional model;

[0037] Based on the contraction ratio, contract the distance between the two common vertices of two adjacent triangular faces that meet the contraction conditions according to the normal angle and the distance between the two common vertices to form new vertices of the triangular faces; wherein, the contraction conditions are that the normal angle between two adjacent triangular faces is less than a preset angle threshold, and the distance between the two common vertices in two adjacent triangular faces is less than the contraction ratio;

[0038] Determine the vertex array of the first - layer sub - model according to the new vertices of each triangular face and the vertices that do not meet the contraction conditions;

[0039] Extract the UV coordinates corresponding to the vertices in the vertex array of the first - layer sub - model from the UV texture map array to obtain the UV texture map array of the first - layer sub - model;

[0040] Based on the Delaunay triangulation algorithm, divide the vertex array of the first - layer sub - model to obtain the triangular face array of the first - layer sub - model;

[0041] Combine the vertex array, the triangular face array, and the UV texture map array of the first - layer sub - model to obtain the first - layer sub - model.

[0042] In a second aspect, the present invention provides an underground utility tunnel model segmented loading system, including:

[0043] A three - mode panoramic modeling module for loading the panoramic image data of the target underground utility tunnel into the three - dimensional model of the target underground utility tunnel to obtain a panoramic three - dimensional model;

[0044] A sub-model cutting module, configured to cut the panoramic three-dimensional model into multiple sub-models in the extension direction according to preset sub-model segmentation parameters, where the preset sub-model segmentation parameters include the length, the number of layers, and the maximum compression ratio of each sub-model;

[0045] A sub-model layering module, configured to layer each sub-model according to the number of layers and the maximum compression ratio of the sub-model to obtain multiple layers of sub-models;

[0046] A rendering and loading module, configured to retrieve multiple layers of sub-models corresponding to the position to be analyzed in the target underground pipe gallery from the panoramic three-dimensional model, and load and render at least one layer of sub-models among the retrieved multiple layers of sub-models.

[0047] In a third aspect, the present invention provides an electronic device, where the electronic device includes a memory and a processor, and a computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of the underground pipe gallery model segmentation and loading method as described in the first aspect.

[0048] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the steps of the underground pipe gallery model segmentation and loading method as described in the first aspect are implemented.

[0049] As can be seen from the above technical solutions, the present invention loads the panoramic image data of the target underground pipe gallery into its three-dimensional model to construct a panoramic three-dimensional model. Subsequently, the panoramic three-dimensional model is cut into multiple sub-models in the extension direction, and layered processing is performed according to the number of layers and the maximum compression ratio of the sub-models. For the part to be analyzed in the target underground pipe gallery, at least one layer of sub-models corresponding to the corresponding position in the panoramic three-dimensional model is selected for loading and rendering. This layering method comprehensively considers the number of layers and the maximum compression ratio of the sub-models, effectively reducing the performance requirements of the device and the workload of loading and rendering, and facilitating the compatible operation of the underground pipe gallery model on the mobile terminal. This significantly reduces the data volume, improves the processing efficiency, and reduces the labor cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0051] Figure 1The application environment of a method for segmental loading of an underground pipe gallery model provided by an embodiment of the present invention;

[0052] Figure 2 The flowchart of a method for segmental loading of an underground pipe gallery model provided by an embodiment of the present invention;

[0053] Figure 3 The schematic diagram of the point cloud data distribution of the underground pipe gallery;

[0054] Figure 4 The schematic diagram of the three-dimensional model generated based on the point cloud data;

[0055] Figure 5 The schematic diagram of cutting the panoramic three-dimensional model based on the cutting plane;

[0056] Figure 6 The schematic diagram of vertex contraction;

[0057] Figure 7 The structural schematic diagram of a system for segmental loading of an underground pipe gallery model provided by an embodiment of the present invention;

[0058] Figure 8 The structural schematic diagram of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0059] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0060] The method for segmental loading of an underground pipe gallery model provided by the embodiments of the present application can be applied as Figure 1In the application environment shown. Among them, the terminal 101 communicates with the server 102 through the network. The data storage system can store the data that the server 102 needs to process. The data storage system can be integrated on the server 102, or placed on the cloud or other network servers. The server 102 obtains a panoramic three-dimensional model by loading the panoramic image data of the target underground utility tunnel into the three-dimensional model of the target underground utility tunnel; cuts the panoramic three-dimensional model into multiple sub-models according to the preset sub-model segmentation parameters in the extension direction, and the preset sub-model segmentation parameters include the length, the number of layers, and the maximum compression ratio of each sub-model; for each sub-model, divides the sub-model into layers according to the number of layers and the maximum compression ratio of the sub-model to obtain multiple layers of sub-models; according to the part to be analyzed in the target underground utility tunnel, retrieves the multiple layers of sub-models corresponding to the position of the part to be analyzed from the panoramic three-dimensional model, and loads and renders at least one layer of the retrieved multiple layers of sub-models.

[0061] Among them, the terminal 101 can be a computer, a mobile terminal, or a server with powerful computing performance.

[0062] Among them, the server 102 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0063] Such as Figure 2 shown, an embodiment of the present application provides a method for segmenting and loading an underground utility tunnel model. Taking this method applied to Figure 1 the terminal 101 or the server 102 in it as an example for illustration, it includes the following steps S1 to S4. Among them:

[0064] Step S1, load the panoramic image data of the target underground utility tunnel into the three-dimensional model of the target underground utility tunnel to obtain a panoramic three-dimensional model.

[0065] Among them, the panoramic image data refers to the panoramic image information of the target underground utility tunnel obtained by a panoramic camera or other panoramic acquisition devices. This panoramic image data can comprehensively display the internal structure and environment of the underground utility tunnel. The panoramic image data has the characteristics of high resolution and panoramic view, and can provide users with an immersive visual experience. When loading into the three-dimensional model, through image processing technology and three-dimensional modeling technology, the panoramic image data is accurately matched with the coordinates of the three-dimensional model to ensure that the panoramic image can accurately fit on the surface of the three-dimensional model, thereby obtaining a panoramic three-dimensional model. The panoramic three-dimensional model not only retains the three-dimensional space information of the underground utility tunnel, but also incorporates the details and textures of the panoramic image, making the model more real and three-dimensional.

[0066] Step S2: Cut the panoramic three-dimensional model into multiple sub-models in the extension direction according to the preset sub-model segmentation parameters. The preset sub-model segmentation parameters include the length, number of layers, and maximum compression ratio of each sub-model.

[0067] Among them, during the construction of the k-th sub-model, the staff has quite a lot of flexibility to define the key parameters of the model. These parameters include the length, number of layers, and maximum compression ratio of the sub-model. First, the length of the sub-model can be set by the staff according to specific needs and input into the system. The value range of the length is strictly limited between 1 meter and 5 meters, ensuring the rationality and feasibility of the model in practical applications. Second, the number of layers is also an important parameter defined by the staff, which determines the complexity of the sub-model. The number of layers must be a positive integer, and its value range is set between 3 and 8. Such a limit not only ensures the fineness of the model but also avoids unnecessary complexity. Finally, the maximum compression ratio is another key parameter, which describes the maximum compression degree that the sub-model can reach during the compression process. The value range of the maximum compression ratio is set between 0% and 100%, which means that the sub-model can be compressed to a state of complete non-compression or complete compression.

[0068] Specifically, first determine the extension direction of the panoramic three-dimensional model, which can be set according to the actual trend of the underground utility tunnel or the needs of the staff. Then, starting from one end of the panoramic three-dimensional model, cut along the extension direction according to the length of each sub-model in the preset sub-model segmentation parameters to generate multiple cutting surfaces. These cutting surfaces divide the panoramic three-dimensional model into multiple independent sub-model segments. Next, for each sub-model segment, extract its corresponding vertex array, triangle face array, and UV texture map array, and recombine these information to generate independent sub-models. In this way, multiple sub-models are obtained, and each sub-model contains a part of the structure and information of the underground utility tunnel, facilitating subsequent layering processing and loading and rendering.

[0069] Step S3: For each sub-model, layer the sub-model according to the number of layers and the maximum compression ratio of the sub-model to obtain multiple layers of sub-models.

[0070] Among them, perform layering processing on this sub-model according to the described number of layers of the sub-model and the set maximum compression ratio to obtain a series of hierarchical sub-models.

[0071] Step S4: According to the part to be analyzed in the target underground utility tunnel, retrieve the multiple layers of sub-models corresponding to the position of the part to be analyzed from the panoramic three-dimensional model, and load and render at least one layer of the retrieved multiple layers of sub-models.

[0072] It can be understood that by using a camera, we can accurately identify the specific parts in the underground utility tunnel that need to be loaded and rendered. Subsequently, based on the panoramic image data corresponding to these parts, multiple levels of sub-models associated with this panoramic image data can be further determined. The process of retrieving these relevant models and performing the loading and rendering, given that the loading and rendering technology is already quite mature in this field, will not be elaborated in detail here.

[0073] It should be noted that in the embodiment of the present application, the panoramic image data of the target underground utility tunnel is loaded into its three-dimensional model to construct a panoramic three-dimensional model. Subsequently, the panoramic three-dimensional model is cut into multiple sub-models along the extension direction, and hierarchical processing is performed according to the number of layers and the maximum compression ratio of the sub-models. For the parts in the target underground utility tunnel that need to be analyzed, at least one layer of the multi-layer sub-models at the corresponding positions in the panoramic three-dimensional model is selected for loading and rendering. This hierarchical method comprehensively considers the number of layers and the maximum compression ratio of the sub-models, effectively reducing the performance requirements of the device and the workload of loading and rendering. At the same time, it facilitates the compatible operation of the underground utility tunnel model on the mobile terminal. This significantly reduces the amount of data, improves the processing efficiency, reduces the labor cost, and strongly promotes the wide application of the three-dimensional model of the underground utility tunnel.

[0074] In some embodiments, loading the panoramic image data of the target underground utility tunnel into the three-dimensional model of the target underground utility tunnel in step S1 to obtain a panoramic three-dimensional model includes:

[0075] Step S101, collect the point cloud data of the target underground utility tunnel, perform filtering processing on the point cloud data, and generate a three-dimensional model of the target underground utility tunnel according to the filtered point cloud data.

[0076] Among them, lidar devices such as Trimble X7 three-dimensional laser scanners and Lingguang Lixel L1 handheld SLAM devices can be selected to collect the point cloud data of the underground utility tunnel, as Figure 3 shown. Before data collection, the lidar needs to be accurately calibrated to ensure the accuracy and reliability of the obtained point cloud data.

[0077] As an implementation method, when performing filtering processing on the collected point cloud data, filtering methods such as Gaussian filtering or median filtering can be used to remove the noise points in the point cloud data. Both Gaussian filtering and median filtering are existing filtering technologies and will not be elaborated here.

[0078] Among them, generating a three-dimensional model of the target underground utility tunnel according to the filtered point cloud data includes:

[0079] Step S1011, use each point in the filtered point cloud data as a vertex to form a vertex array.

[0080] Take each point in the filtered point cloud data as a vertex to form a vertex array, and the expression of the vertex array is as follows:

[0081]

[0082] Among them, V represents the vertex array, represents the three-dimensional coordinates of the i-th vertex in the vertex array, (x i , y i , z i ) represents the three-dimensional coordinates of the i-th vertex, that is, the three-dimensional coordinates of the point corresponding to the i-th vertex in the filtered point cloud data. i is the index of the vertex in the vertex array, and i ∈ [1, N], where N represents the number of vertices in the vertex array.

[0083] Step S1012: Segment the vertex array based on the Delaunay triangulation algorithm to obtain a triangular face array.

[0084] Among them, the Delaunay triangulation algorithm refers to performing spatial triangulation on each vertex in the vertex array to form multiple triangular faces, and these triangular faces together constitute the three-dimensional surface of the underground utility tunnel. The Delaunay triangulation algorithm can ensure that the formed triangular faces have the largest minimum angle, that is, the formed triangular faces are as close as possible to equilateral triangles, thereby improving the accuracy and smoothness of the three-dimensional model.

[0085] Specifically, the process of segmenting the vertex array based on the Delaunay triangulation algorithm includes:

[0086] Randomly select three vertices from the vertex array to form an initial point set to generate an initial triangular face;

[0087] Gradually insert the remaining vertices. Each time a vertex is inserted, adjust the surrounding triangular faces to make them meet the Delaunay condition, that is, there is no vertex inside the circumcircle of any triangular face;

[0088] Until all the remaining vertices in the vertex array are inserted, finally obtain a set of multiple triangular faces, which is the triangular face array, and the expression of the triangular face array is as follows;

[0089] T = {(o1, p1, q1), (o2, p2, q2), ···, (o j , p j , q j ), ···, (o M , p M , q M )}

[0090] Among them, T represents the triangular face array, o j , p j,q j They are respectively the three vertices of the j-th triangular face, where j is the index of the triangular face, and i ∈ [1, M], and M represents the number of triangular faces in the triangular face array;

[0091] As another implementation, the triangular face array in the three-dimensional model can also be generated by the Poisson surface reconstruction method, which will not be elaborated here.

[0092] Step S1013: Project each vertex in the vertex array onto a two-dimensional plane to generate UV coordinates, and create a UV texture map array based on the UV coordinates.

[0093] Among them, when projecting each vertex in the vertex array onto a two-dimensional plane to generate UV coordinates, methods such as spherical mapping, cylindrical mapping, or planar mapping can be used. The expression of the texture map array is as follows:

[0094] U = {(u i , v i ) | i = 1, 2, ···, N}

[0095] Among them, U represents the UV texture map array, and (u i , v i ) is the UV coordinate corresponding to the i-th vertex in the vertex array.

[0096] Step S1014: Combine the vertex array, the triangular face array, and the UV texture map array to generate a three-dimensional model of the target underground utility tunnel.

[0097] Among them, the three-dimensional model of the target underground utility tunnel is as Figure 4 shown. By combining the vertex array, the triangular face array, and the UV texture map array to generate a three-dimensional model, the expression of the three-dimensional model is as follows:

[0098] mesh = {V, T, U}

[0099] Among them, mesh represents the three-dimensional model.

[0100] Step S102: Match the geographical location coordinates of the panoramic image data with the three-dimensional coordinates of the three-dimensional model of the target underground utility tunnel.

[0101] Among them, the specific logic for matching the geographical location coordinates of the panoramic image data with the three-dimensional coordinates of the vertices in the three-dimensional model is: find the vertex in the vertex array of the three-dimensional model that is closest to the geographical location coordinates of the panoramic image data. The specific expression is as follows:

[0102]

[0103] Among them, Denote the vertex in the 3D model that is closest to the geographical location coordinates of the panoramic image data, i.e., associate this panoramic image data with the vertex in the 3D image. ||·|| represents the Euclidean distance. Establish an association.

[0104] Step S103: Extract the panoramic feature parameters in the panoramic image data of the target underground pipe gallery, and load the panoramic feature parameters into the 3D model of the target underground pipe gallery according to the matching result to obtain a panoramic 3D model.

[0105] Among them, when extracting the panoramic feature parameters in the panoramic image data, the panoramic feature parameters include image ID, image URL, geographical location coordinates, compass direction, and acceleration information.

[0106] Among them, calibrate the feature parameter qjList as:

[0107] qjList = {imaID, imagURL, GisPos,, compass, acc}

[0108] In the formula, imaID, imagURL, GisPos,, compass, acc respectively represent image ID, image URL, geographical location coordinates, compass direction, and acceleration information, and GisPos = {la, lon, hight}, where la, lon, hight respectively represent the latitude, longitude, and altitude information in the geographical location coordinates.

[0109] It should be noted that the panoramic image data can be loaded from relevant storage devices or websites. The image ID is the unique identifier of the image, the image URL is the storage path of the image, the geographical location coordinates include latitude, longitude, and altitude, and the acceleration information includes the acceleration values in the x, y, and z directions. Loading the panoramic image data from the storage device or website and extracting the feature parameter group are existing technologies and will not be elaborated here.

[0110] In some embodiments, it is also necessary to define the sub-model segmentation parameters. The sub-model segmentation parameters include defining the length, number of layers, and maximum compression ratio of each sub-model. The expression of the segmentation and layering parameters is as follows;

[0111] CutParam = {(Distance k , layer k , maxCommpress k ) | k = 1, 2, ···, Z}

[0112] Among them, CutParam represents the segmentation and layering parameters, (Distance k , layer k , maxCommpressk ) are respectively the length, the number of layers, and the maximum compression ratio of the k-th sub-model. Z represents the number of segments of the three-dimensional model, which is the number of sub-models. k is the index of the sub-model, and k ∈ [1, Z]. The indices of each sub-model are defined in sequence along the direction from the starting point to the ending point of the three-dimensional model. For example, for the sub-model closest to the starting point, its index is defined as 1, and for the sub-model closest to the ending point, its index is defined as Z;

[0113] It should be noted that the length, the number of layers, and the maximum compression ratio of the k-th sub-model can all be defined and input by the staff. The length is a positive number, with a value between 1 and 5 meters. The number of layers is a positive integer, with a value between 3 and 8. The maximum compression ratio has a value between 0 and 100%.

[0114] In the embodiment of the present application, it is necessary to optimize the sub-model segmentation parameters to improve the accuracy of the sub-model segmentation parameters.

[0115] Specifically, optimizing the sub-model segmentation parameters includes:

[0116] Step S21: Randomly generate multiple groups of sub-model segmentation parameters to form an initial population, and each group of sub-model segmentation parameters is an individual in the initial population.

[0117] Among them, determine the value ranges of the sub-model length, the number of layers, and the maximum compression ratio, and randomly generate multiple groups of sub-model segmentation parameters within the value ranges to form an initial population. Each group of sub-model segmentation parameters is an individual in the initial population.

[0118] Step S22: Based on the genetic algorithm, optimize the initial population according to the preset fitness value, obtain the individual with the optimal preset fitness value, and determine the optimal sub-model segmentation parameters according to the individual with the optimal preset fitness value.

[0119] Among them, calculate the fitness values of each individual in the population, sort the individuals in descending order of fitness, select the individuals ranked in the front as parents for crossover and mutation operations to generate new individuals; until reaching the predetermined number of iterations, select the individual with the largest fitness value as the optimal sub-model segmentation parameters.

[0120] It should be noted that crossover and mutation are conventional operations of the genetic evolution algorithm and will not be elaborated here. In addition, during the mutation operation, it is necessary to ensure the constraint conditions that the length is a positive number, with a value between 1 and 5 meters, the number of layers is a positive integer, with a value between 3 and 8, and the maximum compression ratio has a value between 0 and 100%.

[0121] Among them, the preset fitness value is obtained by weighted calculation according to the respective preset evaluation indicators corresponding to the segmentation parameters of each group of sub-models; among them, the preset evaluation indicators include the time consumed by the sub-model layering and segmentation operations, the total storage capacity required for each layer of sub-models after layering and segmentation, the loading and rendering time of each layer of sub-models after layering and segmentation, and the loading and rendering satisfaction of each layer of sub-models after layering and segmentation.

[0122] Among them, the acquisition process of each preset evaluation indicator includes: based on the segmentation parameters of each group of sub-models, performing segmentation and layering operations on the underground utility tunnel to be analyzed respectively, and recording the time consumed by the segmentation and layering operations and the total storage capacity used to store each layer of sub-models;

[0123] It should be noted that the time consumed by the segmentation and layering operations can be obtained through a timer, and the total storage capacity used to store each layer of sub-models can be obtained by checking the sum of the capacities occupied by each layer of models in the storage device, which will not be elaborated here;

[0124] Apply the model obtained based on the sub-model segmentation parameters to actual loading and rendering multiple times, and record the loading and rendering time and the loading and rendering satisfaction score each time it is applied to actual loading and rendering, and calculate the average loading and rendering time and the average loading and rendering satisfaction score of the model obtained based on each group of sub-model segmentation parameters in multiple actual loading and renderings;

[0125] It should be noted that the loading and rendering time can be obtained through a timer, and the loading and rendering satisfaction score can be obtained by the implementers filling out a loading and rendering satisfaction score form. The loading and rendering satisfaction score form includes a total of five criteria: very dissatisfied, dissatisfied, average, satisfied, and very satisfied. The loading and rendering satisfaction scores corresponding to each criterion are 2, 4, 6, 8, and 10 respectively;

[0126] Perform maximum-minimum normalization processing on all the time consumed by the segmentation and layering operations, all the total storage capacities, all the average loading and rendering times, and all the average loading and rendering satisfaction scores respectively. The maximum-minimum normalization processing is a conventional technical means for those skilled in the art and will not be elaborated here.

[0127] Specifically, the calculation method of the fitness value is as follows:

[0128] 1) Based on the time consumed by the segmentation and layering operations and the total storage capacity, generate a first fitness value for evaluating the difficulty of implementing the segmentation and layering operations. The more difficult the implementation of the segmentation and layering operations, the smaller the first fitness value. The calculation formula is as follows:

[0129]

[0130] Among them, T1 represents the time consumed by the segmentation and layering operation. The longer the time consumed by the segmentation and layering operation, the greater the operation difficulty and computational complexity of segmenting and layering the 3D model, and the more human and material resources required for the segmentation and layering operation, that is, the more difficult the operation of segmenting and layering the 3D model;

[0131] Among them, St represents the total storage capacity. The larger the total storage capacity, the more difficult the storage operation after segmenting and layering the 3D model, and the more storage devices are used to store each layer of sub-models, that is, the more difficult the storage operation after segmenting and layering the 3D model;

[0132] Therefore, the first fitness value is generated based on the time consumed by the segmentation and layering operation and the total storage capacity, so that the first fitness value comprehensively reflects the operation difficulty of segmenting and layering the 3D model and the storage operation difficulty after segmenting and layering the 3D model. Also, since the segmentation and layering of the 3D model belongs to the previous step and has relatively loose requirements for time, while the storage after segmenting and layering the 3D model is a long-term task that requires long-term occupation of storage memory to store each layer of sub-models. Therefore, when defining the generation of the first fitness value, a greater sensitivity is given to the total storage capacity, that is, the time consumed by the segmentation and layering operation is evaluated for the first fitness value in the form of and the total storage capacity is evaluated for the first fitness value in the form of A smaller weight coefficient ω1 is given to

[0133] and a larger weight coefficient ω2 is given to

[0134] Based on ensuring the consistency of the relative ratio and sum of the weights, on the basis of 0 < ω1 < ω2 < 1, it is limited that ω1 + ω2 = 1, and the specific values of ω1 and ω2 are set by the staff according to the actual situation and are not limited here;

[0135] 2) Based on the average value of the loading and rendering time and the average value of the loading and rendering satisfaction score, generate a second fitness value for evaluating the application convenience degree of the model after segmentation and layering. The less convenient the application of the model after segmentation and layering, the smaller the second fitness value. The calculation formula is as follows:

[0136] SYD2 = ω3 * √MYD - ω4 * ln(1 + T2)

[0137] Among them, T2 represents the average loading and rendering time. The larger the average loading and rendering time is, the longer the waiting time required during the real-time loading and rendering process is, which is more unfavorable for the real-time detection and management of the underground pipe gallery, that is, it indicates that the application of the model after segmentation and stratification is less convenient;

[0138] Among them, MYD represents the average loading and rendering satisfaction score. The larger the average loading and rendering satisfaction score is, the better the loading and rendering result is during the real-time loading and rendering process, the more satisfied the staff is with the loading and rendering result, and the more conducive it is to the detection and management of the underground pipe gallery, that is, it indicates that the application of the model after segmentation and stratification is more convenient;

[0139] Therefore, a second fitness value is generated based on the average loading and rendering time and the average loading and rendering satisfaction score, so that the second fitness value comprehensively reflects the advantages and disadvantages of the loading and rendering waiting time and the loading and rendering result. Also, because the loading and rendering result directly affects the accuracy of subsequent detection and management of the underground pipe gallery and is more important than the loading and rendering waiting time, when defining and generating the second fitness value, a greater sensitivity is given to the average loading and rendering satisfaction score, that is, the average loading and rendering satisfaction score is used to evaluate the second fitness value in the form of, and the average loading and rendering time is used to evaluate the second fitness value in the form of -ln(1 + T2), and a larger weight coefficient ω3 is given to, and a smaller weight coefficient ω4 is given to -ln(1 + T2). And to ensure the consistency of the relative proportion and the sum of the weights, on the basis of 0 < ω4 < ω3 < 1, it is limited that ω3 + ω4 = 1, and the specific values of ω3 and ω4 are set by the staff according to the actual situation and are not restricted here;

[0140] As an implementation, the value range of ω3 is 0.6 - 0.9, and the value range of ω4 is 0.1 - 0.4;

[0141] Among them, SYD2 represents the second fitness value. The longer the average loading and rendering time is and the smaller the average loading and rendering satisfaction score is, it indicates that the waiting time required for loading and rendering is longer and the loading and rendering result is worse, which is more unfavorable for the real-time detection and management of the underground pipe gallery, that is, it indicates that the application of the model after segmentation and stratification is less convenient, and the second fitness value is smaller;

[0142] 3) Generate a third fitness value for evaluating the waiting duration based on the time consumed by the segmentation and stratification operation and the average loading and rendering time. The longer the waiting duration is, the smaller the third fitness value is. The calculation formula is as follows:

[0143]

[0144] Among them, SYD3 represents the third fitness value, which is jointly composed of the time consumed by the segmentation and layering operation and the average value of the loading and rendering time. Thus, the superiority and inferiority of the sub-model segmentation parameters are evaluated in terms of the waiting time. Moreover, the longer the time consumed by the segmentation and layering operation and the average value of the loading and rendering time, the longer the required waiting time, the worse the performance of the sub-model segmentation parameters in terms of the waiting time, and the smaller the third fitness value. Also, since segmenting and layering the 3D model is a previous step and the requirement for time is relatively loose, the average value of the loading and rendering time directly affects the real-time detection and management of the underground pipe gallery. Therefore, a greater sensitivity is given to the average value of the loading and rendering time, that is, the time consumed by the segmentation and layering operation is used to evaluate the third fitness value in the form of and the average value of the loading and rendering time is used to evaluate the third fitness value in the form of , and a smaller weight coefficient ω5 is assigned to , and a larger weight coefficient ω6 is assigned to . And to ensure the consistency of the relative ratio and the sum of the weights, it is limited that ω5 + ω6 = 1 on the basis of 0 < ω5 < ω6 < 1. The specific values of ω5 and ω6 are set by the staff according to the actual situation and are not restricted here;

[0145] As an implementation manner, the value range of ω5 is 0.25 - 0.45, and the value range of ω6 is 0.55 - 0.75;

[0146] 4) Based on the first fitness value, the second fitness value, and the third fitness value, a fitness value for evaluating the superiority and inferiority of the sub-model segmentation parameters is generated. The better the sub-model segmentation parameters, the larger the fitness value. The calculation formula is as follows:

[0147] SYD = ω7 * SYD1 + ω8 * SYD2 + ω9 * SYD3

[0148] Among them, SYD represents the fitness value, which is jointly composed of the first fitness value, the second fitness value, and the third fitness value. Moreover, the larger the first fitness value, the second fitness value, and the third fitness value are, the simpler the segmented layering operation based on the segmentation parameters of the sub-model is, the more convenient the application of the model after segmentation and layering is, and the shorter the waiting time is. That is to say, the better the segmentation parameters of the sub-model are, the larger the fitness value is. Also, since the first fitness value reflects the previous work of the inspection and management of the underground utility tunnel, the second fitness value reflects the real-time work of the inspection and management of the underground utility tunnel, and the third fitness value takes into account both the previous and real-time work of the inspection and management of the underground utility tunnel. Therefore, the second fitness value is the most important, the third fitness value is the second, and the first fitness value is even less important. Therefore, a minimum weight coefficient ω7 is assigned to the first fitness value, a maximum weight coefficient ω8 is assigned to the second fitness value, and an intermediate weight coefficient ω9 is assigned to the third fitness value. And to ensure the consistency of the relative ratio and the sum of the weights, on the basis of 0 < ω7 < ω9 < ω8 < 1, it is limited that ω7 + ω8 + ω9 = 1. The specific values of ω7, ω8, and ω9 are set by the staff according to the actual situation and are not limited here;

[0149] As an implementation manner, the value range of ω7 is 0.1 - 0.2, the value range of ω8 is 0.5 - 0.7, and the value range of ω9 is 0.2 - 0.3.

[0150] In some embodiments, the panoramic three-dimensional model is cut into multiple sub-models according to the preset sub-model segmentation parameters, including:

[0151] Step S201: Starting from the vertex at one end of the panoramic three-dimensional model, cut along the extension direction of the panoramic three-dimensional model according to the length of each sub-model to generate multiple cutting planes.

[0152] Among them, select the vertex at one end of the three-dimensional model as the starting point, and use the vertex at the other end of the three-dimensional model as the ending point. According to the defined length of each sub-model, starting from the starting point, along the extension direction of the three-dimensional model, generate multiple cutting planes. The expression is as follows:

[0153]

[0154] Among them, plane[k] represents the cutting plane of the k-th sub-model, represents the starting point, represents the extension direction vector of the three-dimensional model. The extension direction vector of the three-dimensional model can be specifically obtained through the coordinate relationship of each vertex in the vertex array and will not be elaborated here.

[0155] Step S202: Segment the panoramic three-dimensional model based on the multiple cutting planes to generate multiple sub-models.

[0156] Specifically, as Figure 5 shown, the panoramic three-dimensional model is segmented based on multiple cutting planes to generate multiple sub-models, including:

[0157] Step S2021: For any sub-model of a segment, extract the vertices between the two adjacent cutting planes of the current segment of the sub-model from the vertex array as the vertex array of the current segment of the sub-model, and extract the UV coordinates corresponding to the vertices in the vertex array of the current segment of the sub-model from the UV texture map array to obtain the UV texture map array of the current segment of the sub-model, and extract the triangular faces with at least two vertices between its two adjacent cutting planes from the triangular face array to obtain the triangular face array of the current segment of the sub-model;

[0158] Step S2022: Combine the vertex array, triangular face array, and UV texture map array of the current segment of the sub-model to generate the current segment of the sub-model.

[0159] Exemplarily, the process of segmenting the panoramic three-dimensional model based on multiple cutting planes includes:

[0160] 1) For the first sub-model, extract the vertices between the starting point and the cutting plane from the vertex array to form the vertex array of the first sub-model, and extract the UV coordinates corresponding to the vertices in this vertex array from the UV texture map array to form the UV texture map array of the first sub-model, and extract the triangular faces with at least two vertices between the starting point and the cutting plane from the triangular face array to form the triangular face array of the first sub-model;

[0161] 2) For the sub-models in the middle, extract the vertices between the cutting plane of this segment of the sub-model and the cutting plane of the previous segment of the sub-model to form the vertex array of this segment of the sub-model, and extract the UV coordinates corresponding to the vertices in this vertex array from the UV texture map array to form the UV texture map array of this segment of the sub-model, and extract the triangular faces with at least two vertices between the cutting plane of this segment of the sub-model and the cutting plane of the previous segment of the sub-model from the triangular face array to form the triangular face array of this segment of the sub-model;

[0162] 3) For the last sub-model, extract the vertices between the end point and the cutting plane to form the vertex array of the Kth segment of the sub-model, and extract the UV coordinates corresponding to the vertices in this vertex array from the UV texture map array to form the UV texture map array of the Kth segment of the sub-model, and extract the triangular faces with at least two vertices between the end point and the cutting plane from the triangular face array to form the triangular face array of the Kth segment of the sub-model;

[0163] Among them, the last (the kth segment) sub-model is labeled as Model k , and Model k = {V k , T k , Uk}, V k , T k , U k are respectively the vertex array, triangle face array, and UV texture mapping array of the k-th sub-model.

[0164] Among them,

[0165] herein, the i'-th vertex, j'-th triangle face, and the UV coordinates corresponding to the i'-th vertex in the k-th sub-model are respectively defined as:

[0166]

[0167] In the formula, i' is the index of the vertex in the sub-model, and i' ∈ [1, N k , N k is the number of vertices in the k-th sub-model, j' is the index of the triangle face in the sub-model, and j' ∈ [1, M k , M k is the number of triangle faces in the k-th sub-model.

[0168] In some embodiments, for each sub-model, the sub-model is hierarchically divided according to the number of hierarchical levels and the maximum compression ratio of the sub-model to obtain a multi-level sub-model, including:

[0169] Step S301: Determine the contraction ratio of the vertices within each sub-model according to the number of hierarchical levels and the maximum compression ratio.

[0170] Among them, based on the number of hierarchical levels and the maximum compression ratio, calculate the contraction ratio of the vertices in the sub-model, and the calculation formula is as follows:

[0171]

[0172] Among them, K k represents the contraction ratio of the vertices in the k-th sub-model.

[0173] Step S302: Perform vertex contraction on the sub-model according to the contraction ratio to obtain the first-level sub-model;

[0174] Among them, perform vertex contraction on the sub-model according to the contraction ratio to obtain the structure after vertex contraction as shown in Figure 6 Specifically, the process of performing vertex contraction on the sub-model according to the contraction ratio includes:

[0175] Step S3021: Determine the normal angle between two adjacent triangle faces according to the normal direction of each triangle face in the sub-model.

[0176] Among them, determine the normal direction of each triangle face in the sub-model and calculate the normal angle between two adjacent triangle faces, and the calculation formula is as follows:

[0177]

[0178] where n k,j' is the normal direction of the j'-th triangular face in the k-th sub-model, × represents the vector cross product, and θ k,j',j'+1 is the angle between the j'-th triangular face and the (j'+1)-th triangular face in the k-th sub-model, and the (j'+1)-th triangular face is the triangular face adjacent to the j'-th triangular face.

[0179] Step S3022: Obtain two common vertices in two pairwise adjacent triangular faces, and determine the distance between the two common vertices; where the common vertices are the vertices where adjacent triangular faces overlap, and the two common vertices are respectively the vertex closer to the front and the vertex closer to the back along the extension direction of the panoramic three-dimensional model.

[0180] Obtain two common vertices in two pairwise adjacent triangular faces, and calculate the distance between them. The calculation formula is as follows:

[0181]

[0182] where L k,j',j'+1 represents the distance between two common vertices in the j'-th triangular face and the (j'+1)-th triangular face of the k-th sub-model, and v' k,j',j'+1 , v” k,j',j'+1 respectively represent two common vertices in the j'-th triangular face and the (j'+1)-th triangular face of the k-th sub-model, and v' k,j',j'+1 is the vertex closer to the front along the extension direction of the three-dimensional model, and v” k,j',j'+1 is the vertex closer to the back along the extension direction of the three-dimensional model.

[0183] Step S3023: Based on the contraction ratio, contract the distance between the two common vertices of two pairwise adjacent triangular faces that meet the contraction condition according to the normal angle and the distance between the two common vertices, to form new vertices of the triangular faces; where the contraction condition is that the normal angle between two pairwise adjacent triangular faces is less than a preset angle threshold, and the distance between the two common vertices in two pairwise adjacent triangular faces is less than the contraction ratio.

[0184] In an example, when the normal angle between two pairwise adjacent triangular faces is less than a preset angle threshold, and the distance between the two common vertices in two pairwise adjacent triangular faces is less than the contraction ratio, contract the two common vertices in the adjacent two triangular faces to generate new vertices. The calculation formula for the coordinates of the new vertices is as follows:

[0185]

[0186] where v k,j',j'+1Represents the new vertex coordinates generated after contracting two common vertices in two adjacent triangular faces, W1' k,j',j'+1 , W2' k,j',j'+1 Respectively represent the areas of the j'-th triangular face and the (j'+1)-th triangular face in the k-th sub-model segment, and W1' k,j',j'+1 Is the area of the triangular face closer to the front along the extension direction of the 3D model, W2' k,j',j'+1 Is the area of the triangular face closer to the back along the extension direction of the 3D model.

[0187] When the contraction condition is not met, the two common vertices in the corresponding pair of adjacent triangular faces are not contracted.

[0188] Step S3024: Determine the vertex array of the first-layer sub-model based on the new vertices of each triangular face and the vertices that do not meet the contraction condition;

[0189] Step S3025: Extract the UV coordinates corresponding to the vertices in the vertex array of the first-layer sub-model from the UV texture map array to obtain the UV texture map array of the first-layer sub-model;

[0190] Step S3026: Divide the vertex array of the first-layer sub-model based on the Delaunay triangulation algorithm to obtain the triangular face array of the first-layer sub-model;

[0191] Step S3027: Combine the vertex array, triangular face array, and UV texture map array of the first-layer sub-model to obtain the first-layer sub-model.

[0192] Step S303: Perform vertex contraction on the first-layer sub-model according to the contraction ratio to obtain the second-layer sub-model;

[0193] Step S304: Perform vertex contraction on the second-layer sub-model according to the contraction ratio to obtain the third-layer sub-model, and so on until a multi-layer sub-model with the same number of layers as the number of layers is obtained.

[0194] Among them, after obtaining a multi-layer sub-model with the same number of layers as the number of layers, the multi-layer sub-model is stored.

[0195] Based on the same inventive concept, the embodiments of the present application also provide an underground pipe gallery model segmented loading system for implementing the underground pipe gallery model segmented loading method involved above.

[0196] The implementation solution provided by this system to solve the problem is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more embodiments of the underground pipe gallery model segmented loading system provided below can refer to the limitations on the underground pipe gallery model segmented loading method in the above text, and will not be repeated here.

[0197] Such as Figure 7As shown in the figure, an underground pipe gallery model segmented loading system provided by an embodiment of the present application includes:

[0198] A three-mode panoramic modeling module 100, configured to load panoramic image data of a target underground pipe gallery into a three-dimensional model of the target underground pipe gallery to obtain a panoramic three-dimensional model;

[0199] A sub-model cutting module 200, configured to cut the panoramic three-dimensional model into multiple sub-models according to a preset sub-model segmentation parameter in the extension direction, and the preset sub-model segmentation parameter includes the length, the number of layers, and the maximum compression ratio of each sub-model;

[0200] A sub-model layering module 300, configured to layer each sub-model according to the number of layers and the maximum compression ratio of the sub-model to obtain a multi-layer sub-model;

[0201] A rendering loading module 400, configured to retrieve multi-layer sub-models corresponding to the position to be analyzed in the panoramic three-dimensional model according to the part to be analyzed in the target underground pipe gallery, and perform loading rendering on at least one layer of the retrieved multi-layer sub-models.

[0202] In some embodiments, the three-mode panoramic modeling module 100 is configured to:

[0203] Collect point cloud data of the target underground pipe gallery, perform filtering processing on the point cloud data, and generate a three-dimensional model of the target underground pipe gallery according to the filtered point cloud data;

[0204] Match the geographical location coordinates of the panoramic image data with the three-dimensional coordinates of the three-dimensional model of the target underground pipe gallery;

[0205] Extract panoramic feature parameters in the panoramic image data of the target underground pipe gallery, and load the panoramic feature parameters into the three-dimensional model of the target underground pipe gallery according to the matching result to obtain a panoramic three-dimensional model.

[0206] In some embodiments, generating a three-dimensional model of the target underground pipe gallery according to the filtered point cloud data includes:

[0207] Regarding each point in the filtered point cloud data as a vertex to form a vertex array;

[0208] Segment the vertex array based on the Delaunay triangulation algorithm to obtain a triangular face array;

[0209] Project each vertex in the vertex array onto a two-dimensional plane to generate UV coordinates, and create a UV texture map array based on the UV coordinates;

[0210] Combine the vertex array, the triangular face array, and the UV texture map array to generate a three-dimensional model of the target underground pipe gallery.

[0211] In some embodiments, the system further includes: a parameter optimization module for optimizing the sub-model segmentation parameters; optimizing the sub-model segmentation parameters includes:

[0212] Randomly generating multiple groups of sub-model segmentation parameters to form an initial population, and each group of sub-model segmentation parameters is an individual in the initial population;

[0213] Based on the genetic algorithm, optimizing the initial population according to a preset fitness value to obtain an individual with the optimal preset fitness value, and determining the optimal sub-model segmentation parameters according to the individual with the optimal preset fitness value;

[0214] Wherein, the preset fitness value is obtained by weighted calculation according to each preset evaluation index corresponding to each group of sub-model segmentation parameters; wherein, the preset evaluation indexes include the time consumed by the sub-model layering and segmentation operations, the total storage capacity required for each layer of sub-models after layering and segmentation, the loading and rendering time of each layer of sub-models after layering and segmentation, and the loading and rendering satisfaction of each layer of sub-models after layering and segmentation.

[0215] In some embodiments, the sub-model cutting module 200 is used for:

[0216] Starting from the vertex at one end of the panoramic three-dimensional model, cutting along the extension direction of the panoramic three-dimensional model according to the length of each sub-model to generate multiple cutting planes;

[0217] Segmenting the panoramic three-dimensional model based on the multiple cutting planes to generate multiple sub-models;

[0218] Segmenting the panoramic three-dimensional model based on the multiple cutting planes to generate multiple sub-models, including:

[0219] For any segment of the sub-model, extracting the vertices between the two cutting planes adjacent to the left and right of the current segment of the sub-model from the vertex array as the vertex array of the current segment of the sub-model, and extracting the UV coordinates corresponding to the vertices in the vertex array of the current segment of the sub-model from the UV texture map array to obtain the UV texture map array of the current segment of the sub-model, and extracting the triangular faces with at least two vertices between the two cutting planes adjacent to the left and right of it from the triangular face array to obtain the triangular face array of the current segment of the sub-model;

[0220] Combining the vertex array, triangular face array and UV texture map array of the current segment of the sub-model to generate the current segment of the sub-model.

[0221] In some embodiments, the sub-model layering module 300 is used for:

[0222] Determining the contraction ratio of the vertices in each sub-model according to the number of layers and the maximum compression ratio;

[0223] Vertex contraction is performed on the sub-model according to the contraction ratio to obtain the first-layer sub-model;

[0224] Vertex contraction is performed on the first-layer sub-model according to the contraction ratio to obtain the second-layer sub-model;

[0225] Vertex contraction is performed on the second-layer sub-model according to the contraction ratio to obtain the third-layer sub-model, and so on until a multi-layer sub-model with the same number of layers as the number of layers is obtained.

[0226] In some embodiments, performing vertex contraction on the sub-model according to the contraction ratio includes:

[0227] According to the normal direction of each triangular face in the sub-model, determine the normal angle between two adjacent triangular faces;

[0228] Obtain two common vertices in two adjacent triangular faces and determine the distance between the two common vertices; wherein, the common vertex is the vertex where the adjacent triangular faces overlap, and the two common vertices are respectively the vertex closer to the front and the vertex closer to the back along the extension direction of the panoramic three-dimensional model;

[0229] Based on the contraction ratio, contract the distance between the two common vertices of two adjacent triangular faces that meet the contraction condition according to the normal angle and the distance between the two common vertices to form a new vertex of the triangular face; wherein, the contraction condition is that the normal angle between two adjacent triangular faces is less than a preset angle threshold, and the distance between the two common vertices in two adjacent triangular faces is less than the contraction ratio;

[0230] Determine the vertex array of the first-layer sub-model according to the new vertices of each triangular face and the vertices that do not meet the contraction condition;

[0231] Extract the UV coordinates corresponding to the vertices in the vertex array of the first-layer sub-model from the UV texture map array to obtain the UV texture map array of the first-layer sub-model;

[0232] Based on the Delaunay triangulation algorithm, divide the vertex array of the first-layer sub-model to obtain the triangular face array of the first-layer sub-model;

[0233] Combine according to the vertex array, triangular face array and UV texture map array of the first-layer sub-model to obtain the first-layer sub-model.

[0234] As Figure 8 shown, an embodiment of the present application provides an electronic device. The electronic device 10 includes a memory 20 and a processor 30. A computer program is stored in the memory 20. When the computer program is executed by the processor 30, the processor 30 is caused to execute the steps of the underground utility tunnel model segmented loading method in the above embodiment.

[0235] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the steps of the method for segmentally loading an underground pipe gallery model in the above embodiment are implemented.

[0236] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described system, electronic device, and computer storage medium can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0237] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order different from those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0238] In several embodiments provided by the present invention, it can be understood that each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code includes one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in an order different from that marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved.

[0239] In several embodiments provided by the present invention, it should be understood that the disclosed system, electronic device, computer storage medium, and method can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other may be through some interfaces, and the indirect couplings or communication connections of the devices or units may be in electrical, mechanical, or other forms.

[0240] The unit described as a separation component may or may not be physically separated. The component displayed as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0241] In addition, each functional unit in various embodiments of the present invention may be integrated into a processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0242] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (English full name: Read-Only Memory, English abbreviation: ROM), random access memories (English full name: Random Access Memory, English abbreviation: RAM), magnetic disks, or optical discs that can store program codes.

[0243] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of various embodiments of the present invention.

Claims

1. A segmented loading method for an underground pipe gallery model, characterized in that: include: Loading the panoramic image data of the target underground utility corridor into the three-dimensional model of the target underground utility corridor to obtain a panoramic three-dimensional model; The panoramic three-dimensional model is cut into a plurality of sub-models in an extension direction according to preset sub-model segmentation parameters, wherein the preset sub-model segmentation parameters include the length, the number of layers and the maximum compression rate of each sub-model; For each of the sub-models, the sub-models are layered according to the number of layers of the sub-models and the maximum compression rate to obtain a multi-layer sub-model; According to the part to be analyzed in the target underground utility tunnel, a multi-layer sub-model corresponding to the part to be analyzed is retrieved from the panoramic three-dimensional model, and at least one layer of the retrieved multi-layer sub-model is loaded and rendered.

2. The underground pipe gallery model segmented loading method according to claim 1 is characterized in that: The method of loading the panoramic image data of the target underground utility corridor into the three-dimensional model of the target underground utility corridor to obtain the panoramic three-dimensional model includes: Collecting point cloud data of the target underground utility corridor, filtering the point cloud data, and generating a three-dimensional model of the target underground utility corridor according to the filtered point cloud data; Matching the geographical location coordinates of the panoramic image data with the three-dimensional coordinates of the three-dimensional model of the target underground pipe gallery; Panoramic feature parameters in the panoramic image data of the target underground utility corridor are extracted, and the panoramic feature parameters are loaded into the three-dimensional model of the target underground utility corridor according to the matching result to obtain a panoramic three-dimensional model.

3. The underground pipe gallery model segmented loading method according to claim 2 is characterized in that: Generating the three-dimensional model of the target underground pipe gallery according to the filtered point cloud data includes: Each point in the filtered point cloud data is taken as a vertex to form a vertex array; The vertex array is segmented based on a Delaunay triangulation algorithm to obtain a triangular face array; Projecting each vertex in the vertex array onto a two-dimensional plane, generating UV coordinates, and creating a UV mapping array based on the UV coordinates; The three-dimensional model of the target underground pipe gallery is generated by combining the vertex array, the triangle array and the UV map array.

4. The underground pipe gallery model segmented loading method according to claim 1 is characterized in that: Also includes: Optimize the sub-model segmentation parameters; The optimizing of the sub-model segmentation parameters comprises: Randomly generate multiple groups of sub-model segmentation parameters to form an initial population, each group of the sub-model segmentation parameters is an individual in the initial population; Based on a genetic algorithm, the initial population is optimized according to a preset fitness value to obtain an individual with the best preset fitness value, and an optimal sub-model segmentation parameter is determined according to the individual with the best preset fitness value; Among them, the preset fitness value is obtained by weighted calculation based on the preset evaluation indicators corresponding to each group of sub-model segmentation parameters; wherein, the preset evaluation indicators include the time consumed by sub-model stratification and segmentation operations, the total storage capacity required for each layer of sub-models after stratification and segmentation, the loading and rendering time of each layer of sub-models after stratification and segmentation, and the loading and rendering satisfaction of each layer of sub-models after stratification and segmentation.

5. The underground pipe gallery model segmented loading method according to claim 3 is characterized in that: The step of cutting the panoramic three-dimensional model into a plurality of sub-models in the extension direction according to the preset sub-model segmentation parameters includes: Starting from a vertex at one end of the panoramic three-dimensional model as a starting point, cutting is performed along an extension direction of the panoramic three-dimensional model according to the length of each sub-model to generate a plurality of cutting surfaces; Segmenting the panoramic three-dimensional model based on the plurality of cutting planes to generate a plurality of sub-models; The step of segmenting the panoramic three-dimensional model based on the plurality of cutting planes to generate a plurality of sub-models comprises: For any segment of the sub-model, extract the vertices between the two adjacent cutting surfaces of the sub-model of the current segment from the vertex array as the vertex array of the sub-model of the current segment, extract the UV coordinates corresponding to the vertices in the vertex array of the sub-model of the current segment from the UV map array to obtain the UV map array of the sub-model of the current segment, and extract at least two triangular faces whose vertices are located between the two adjacent cutting surfaces from the triangle face array to obtain the triangle face array of the sub-model of the current segment; The sub-model of the current segment is generated by combining the vertex array, the triangle array and the UV map array of the sub-model of the current segment.

6. The underground pipe gallery model segmented loading method according to claim 5 is characterized in that: For each of the sub-models, the sub-models are layered according to the number of layers and the maximum compression rate of the sub-model to obtain a multi-layer sub-model, including: Determining the contraction ratio of the vertices in each of the sub-models according to the number of layers and the maximum compression ratio; Performing vertex shrinkage on the sub-model according to the shrinkage ratio to obtain a first layer of sub-model; Performing vertex contraction on the first layer sub-model according to the contraction ratio to obtain a second layer sub-model; The second layer sub-model is subjected to vertex contraction according to the contraction ratio to obtain a third layer sub-model, and so on, until a multi-layer sub-model having the same number of layers as the number of layers is obtained.

7. The underground pipe gallery model segmented loading method according to claim 6 is characterized in that: The step of performing vertex shrinkage on the sub-model according to the shrinkage ratio includes: Determine the normal angles of two adjacent triangular faces according to the normal directions of the triangular faces in the sub-model; Obtain two common vertices in two adjacent triangular faces, and determine the distance between the two common vertices; wherein the common vertices are vertices of the adjacent triangular faces overlapping each other, and the two common vertices are respectively a front vertex and a rear vertex along the extension direction of the panoramic three-dimensional model; Based on the shrinkage ratio, the distance between two common vertices of the two adjacent triangular faces that meet the shrinkage condition is shrunk according to the normal angle and the distance between the two common vertices to form new vertices of the triangular faces; wherein the shrinkage condition is that the normal angle of the two adjacent triangular faces is less than a preset angle threshold, and the distance between the two common vertices of the two adjacent triangular faces is less than the shrinkage ratio; Determine the vertex array of the first layer sub-model according to the new vertices of each triangular face and the vertices that do not meet the contraction condition; Extracting UV coordinates corresponding to vertices in the vertex array of the first layer sub-model from the UV mapping array to obtain the UV mapping array of the first layer sub-model; Segmenting the vertex array of the first layer sub-model based on the Delaunay triangulation algorithm to obtain a triangular face array of the first layer sub-model; The first layer sub-model is obtained by combining the vertex array, the triangle array and the UV map array of the first layer sub-model.

8. A segmented loading system for an underground pipe gallery model, characterized in that: include: A three-mode panoramic modeling module is used to load the panoramic image data of the target underground pipeline corridor into the three-dimensional model of the target underground pipeline corridor to obtain a panoramic three-dimensional model; A sub-model cutting module, used for cutting the panoramic three-dimensional model into a plurality of sub-models in the extension direction according to preset sub-model segmentation parameters, wherein the preset sub-model segmentation parameters include the length, the number of layers and the maximum compression rate of each sub-model; A sub-model layering module, used for layering each sub-model according to the layer number and the maximum compression rate of the sub-model to obtain a multi-layer sub-model; A rendering loading module is used to retrieve a multi-layer sub-model corresponding to the position to be analyzed in the target underground pipeline corridor from the panoramic three-dimensional model, and load and render at least one layer of the retrieved multi-layer sub-model.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, wherein a computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of the underground corridor model segmented loading method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the steps of the underground pipeline gallery model segmented loading method as described in any one of claims 1 to 7 are implemented.

Images