A method for simulating and constructing urban sewage pipe networks based on image recognition

Abstract

The application discloses a kind of city sewage pipe network simulation construction methods based on image recognition, it is related to the technical field of sewage pipe network simulation, it includes building miniature shooting robot, miniature shooting robot carries out picture shooting in city sewage pipe network, obtains sewage pipe network picture;Microscopic shooting robot obtains the current space position and shooting time point of itself, generates mark watermark, obtains watermarked pipe network picture;Extract the pipe picture data and mark watermark data in watermarked pipe network picture, construct simulated city sewage pipe network;Multiple pipe picture data is arranged according to time sequence, and pipe picture sequence is obtained, sewage activity data is obtained, simulated sewage activity is added in simulated city sewage pipe network according to sewage activity data, and dynamic simulated city sewage pipe network is obtained.The application has the effect of improving the accuracy and real-time of city sewage pipe network simulation construction.

Description

Technical Field

[0001] This application relates to the technical field of sewage pipe network simulation, and in particular to a method for constructing urban sewage pipe network simulation based on image recognition. Background Technology

[0002] With the continuous improvement of urbanization, problems such as structural damage, blockage, and leakage will occur in urban sewage pipe networks. These problems seriously threaten the safety of urban water environment and urban infrastructure.

[0003] In existing technologies, the monitoring of urban sewage pipe networks often involves establishing sewage pipe network models to visualize the current distribution of underground sewage pipe networks. However, in practical applications, these models are generally static, and the monitoring data is mostly abstract numerical data, which cannot effectively and promptly reflect the actual operation of the urban sewage pipe network. Furthermore, when anomalies occur, the numerical data requires analysis and processing, resulting in a long response time. On the other hand, with continuous urban development, urban sewage pipe networks are constantly changing. If these changes are not recorded in a timely manner, gaps in pipe network monitoring will appear in some areas. Summary of the Invention

[0004] The purpose of this invention is to provide a method for simulating and constructing urban sewage pipe networks based on image recognition, so as to solve the problems mentioned in the background art.

[0005] This application provides a method for simulating and constructing urban sewage pipe networks based on image recognition. The method includes:

[0006] A miniature camera robot is constructed and released from a sewage pipe outlet into the urban sewage pipe network. The miniature camera robot captures images of the urban sewage pipe network to obtain pictures of the sewage pipe network.

[0007] The miniature camera robot acquires its current spatial location and shooting time, and generates a watermark on the sewage pipe network image based on the current spatial location and shooting time to obtain a watermarked pipe network image.

[0008] Extract pipe image data and watermark data from the watermarked pipe network image. Obtain the pipe spatial location and internal pipe data based on the pipe image data and the watermark data. Construct a simulated urban sewage pipe network based on the pipe spatial location and internal pipe data.

[0009] Multiple pipeline image data are arranged in chronological order to obtain a pipeline image sequence. A continuous pipeline feature change table is obtained based on the pipeline image sequence. Sewage activity data is obtained based on the pipeline feature change table. Simulated sewage activity is added to the simulated urban sewage network based on the sewage activity data to obtain a dynamic simulated urban sewage network.

[0010] Preferably, the step of constructing a miniature camera robot, releasing the miniature camera robot from a sewage pipe outlet into the urban sewage pipe network, and having the miniature camera robot capture images of the urban sewage pipe network to obtain images of the sewage pipe network is as follows:

[0011] Construct a miniature camera robot, and release multiple of the miniature camera robots from the sewage pipe outlet into the urban sewage pipe network;

[0012] After the miniature camera robot enters the urban sewage pipe network, it takes pictures of the urban sewage pipe network to obtain images of the sewage pipe network.

[0013] The miniature camera robot performs edge recognition on the sewage pipe network image to obtain the extension direction of the pipe;

[0014] Based on the direction of extension, multiple miniature imaging robots are automatically grouped and move in different directions, continuously collecting images of the sewage pipe network at different locations.

[0015] Preferably, the step of the miniature imaging robot acquiring its current spatial location and shooting time, and generating a watermark on the sewage pipe network image based on the current spatial location and shooting time to obtain a watermarked pipe network image is as follows:

[0016] Obtain the location data of the sewage pipe opening, and generate a spatial origin based on the pipe opening location data;

[0017] The miniature camera robot records the spatial origin and constructs a three-dimensional virtual blank space based on the spatial origin;

[0018] Based on the sewage pipe network images, the relative position change data of the miniature imaging robot relative to the spatial origin is obtained;

[0019] The spatial position change of the miniature camera robot is recorded in the three-dimensional virtual blank space based on the relative position change data, and the current spatial position of the miniature camera robot is obtained.

[0020] The system obtains the shooting time of each sewage pipe network image taken by the miniature shooting robot, combines the shooting time with the current spatial location, generates a watermark, and adds the watermark to each corresponding sewage pipe network image to obtain a watermarked pipe network image.

[0021] Preferably, the step of obtaining the spatial location and internal data of the pipeline based on the pipeline image data and the marked watermark data, and constructing a simulated urban sewage pipe network based on the pipeline spatial location and internal data, specifically includes:

[0022] Based on the pipeline image data, multiple pipeline image data are stitched together to obtain a complete initial pipeline cross-section image;

[0023] Extract the image offset of the initial pipe cross-section image, generate an image correction value based on the image offset, and perform image correction on the pipe cross-section based on the image correction value to obtain the target pipe cross-section image;

[0024] Multiple cross-sectional images of the target pipe are arranged sequentially in chronological order to generate a three-dimensional image of the pipe's interior.

[0025] Based on the internal image of the three-dimensional pipe, the connection positions between pipes, the pipe tilt angle, and the internal material image are extracted.

[0026] By combining the connection location, the pipe tilt angle, and the internal material image, the internal data of the pipe is obtained;

[0027] Based on the marked watermark and the connection position, the spatial position of each pipe segment is obtained. Combining the spatial position of the pipe and the internal data of the pipe, the pipe is simulated and constructed to generate a simulated urban sewage pipe network.

[0028] Preferably, after the step of arranging multiple target pipe cross-sectional images sequentially in chronological order to generate a three-dimensional image of the pipe's interior, the method further includes:

[0029] Based on the three-dimensional image of the inside of the pipe, the inner wall of the pipe is scanned and identified to obtain the integrity value of the inner wall and the inner wall diameter data.

[0030] Based on the inner wall integrity value, the branch pipe connected to the main pipe is obtained; based on the inner wall diameter data, the pipe diameter variation data is obtained.

[0031] Based on the pipe diameter change data, extract the pipe segment diameter data of adjacent pipes, compare the pipe diameter change data and the pipe diameter data to obtain the diameter difference value, and generate the simulation parameter change value based on the diameter difference value;

[0032] Based on the branch pipe, the branch entrance is obtained, and the entrance spatial position of the branch entrance is extracted. An entrance marker is generated based on the entrance spatial position, and the entrance marker is sent to the subsequent micro-shooting robot.

[0033] Preferably, after sending the entry marker to the subsequent miniature imaging robot, the method further includes:

[0034] After receiving the entrance marker, the miniature camera robot enters the branch entrance and captures images of the branch pipe within the branch pipe.

[0035] Based on the branch pipe image, the branch pipe diameter is obtained, and it is determined whether the branch pipe diameter is greater than the inner wall diameter data;

[0036] If it is determined that the diameter of the branch pipe is greater than the inner wall diameter data, then the branch pipe is marked as the superior pipe, and the shooting priority of the superior pipe is increased;

[0037] If it is determined that the diameter of the branch pipe is smaller than the inner wall diameter data, then the branch pipe is marked as a lower-level pipe, and the shooting priority of the lower-level pipe is reduced;

[0038] If it is determined that the diameter of the branch pipe is equal to the inner wall diameter data, then the branch pipe is marked as a pipe of the same level, and the shooting priority of the pipe of the same level is not changed.

[0039] Preferably, after the step of simulating and constructing the pipeline to generate a simulated urban sewage pipe network, the method further includes:

[0040] The simulated urban sewage pipe network is traversed to obtain the pipe extension value of each pipe in the simulated urban sewage pipe network, as well as the logical reasonable value of the pipe connection.

[0041] Based on the pipeline extension value, it is determined whether the simulated urban sewage pipe network has experienced abnormal extension;

[0042] If it is determined that the simulated urban sewage pipe network has an abnormal extension, then the abnormal extension segment in the simulated urban sewage pipe network and the abnormal pipe network image corresponding to the abnormal extension segment are extracted.

[0043] Anomaly assessment is performed on the abnormal pipeline network images to determine the cause of the anomaly, and the abnormal extension segment is reconstructed based on the cause of the anomaly.

[0044] Based on the logically reasonable value, determine whether there are any abnormal connections in the pipe connections of the simulated urban sewage pipe network;

[0045] If an abnormal connection is identified in the pipeline connection, images of adjacent abnormal pipelines at both ends of the abnormal connection are extracted, and the type of abnormal connection is determined based on the images of the adjacent abnormal pipelines.

[0046] Based on the type of abnormal connection, the abnormal connection is automatically corrected.

[0047] Preferably, the step of traversing the simulated urban sewage pipe network to obtain the pipe extension value of each pipe in the simulated urban sewage pipe network and the logically reasonable value of the pipe connection is as follows:

[0048] The simulated urban sewage pipe network is traversed to extract the complete network value, flow logic relationship and main branch pipe connection relationship of the simulated urban sewage pipe network;

[0049] Based on the integrity value of the pipeline network, the number of pipeline segment ends in the simulated urban sewage pipeline network is extracted, and the length of each pipeline segment end is extracted. Combining the number of pipeline segment ends and the length of the pipeline segment, the pipeline extension value of each pipeline is obtained.

[0050] Based on the flow direction logic relationship and the main branch pipeline connection relationship, extract the degree of consistency between the flow direction logic relationship and the main branch pipeline connection relationship;

[0051] Based on the degree of agreement, the logical rationality of the main and branch pipeline connection relationship is evaluated to obtain the logical rationality value of the pipeline connection.

[0052] In summary, this application includes at least one of the following beneficial technical effects:

[0053] By constructing miniature imaging robots, these robots are released into the urban sewage network through sewer openings and other sewage pipe openings. Once inside, they photograph the interior of the pipes, producing images of the sewage network. Each robot records its current spatial position in real-time based on its movement path. Combined with the capture time of each sewage network photo, a watermark is generated and added to the sewage network image, creating a watermarked image. This watermarked image is then processed to obtain the spatial position and internal data of each pipe segment, including its inner diameter. Based on this data, a simulated urban sewage network is constructed. Simultaneously, the miniature imaging robots process the sewage network photos in real-time. When a branch pipe is detected, it notifies other miniature imaging robots to enter that branch pipe. The inner diameter of the branch pipe is then compared with the inner diameter of the current pipe to determine its priority. Higher priority branches receive more miniature imaging robots. The miniature imaging robots continuously enter other discovered branch pipes to perform their work. The process continues until enough watermarked pipe network images are collected, and then a simulated urban sewage pipe network is constructed based on these images. The obtained pipe image data is then arranged chronologically to obtain the sewage activity within the pipes, and this sewage activity is reflected in the simulated urban sewage pipe network in real time, resulting in a dynamic simulated urban sewage pipe network. This improves the accuracy and real-time performance of the urban sewage pipe network simulation. Attached Figure Description

[0054] Figure 1 This is a flowchart illustrating the steps of an image recognition-based method for simulating and constructing an urban sewage pipe network, as provided in an embodiment of this application. Detailed Implementation

[0055] The following combination Figure 1 This application will be described in further detail, but the embodiments of the present invention are not limited thereto.

[0056] This application discloses a method for simulating and constructing urban sewage pipe networks based on image recognition.

[0057] In this embodiment, a method for simulating and constructing an urban sewage pipe network based on image recognition is provided, the method comprising:

[0058] S100: Construct a miniature camera robot and release it from the sewage pipe outlet into the urban sewage network. The miniature camera robot will take pictures in the urban sewage network to obtain images of the sewage network.

[0059] S200: The miniature camera robot acquires its current spatial location and shooting time, and generates a watermark on the sewage pipe network image based on the current spatial location and shooting time to obtain a watermarked pipe network image.

[0060] S300: Extract pipe image data and watermark data from watermarked pipe network images, obtain the pipe spatial location and internal pipe data based on the pipe image data and watermark data, and construct a simulated urban sewage pipe network based on the pipe spatial location and internal pipe data.

[0061] S400: Arrange multiple pipeline image data in chronological order to obtain a pipeline image sequence, and obtain a continuous pipeline feature change table based on the pipeline image sequence. Obtain sewage activity data based on the pipeline feature change table, and add simulated sewage activity to the simulated urban sewage network based on the sewage activity data to obtain a dynamic simulated urban sewage network.

[0062] It should be noted that the above process is only the basic steps of this embodiment. In the specific implementation process, some steps may be added, reduced or modified appropriately without affecting the overall implementation effect.

[0063] The steps involved in constructing a miniature imaging robot, releasing it from a sewage pipe outlet into the urban sewage network, and having the robot capture images of the sewage network are as follows:

[0064] Construct miniature filming robots and release multiple miniature filming robots from sewage pipe outlets into the urban sewage pipe network;

[0065] After entering the city's sewage pipe network, the miniature camera robot takes pictures of the sewage pipe network.

[0066] A miniature camera robot performs edge recognition on images of sewage pipe networks to determine the direction of pipe extension;

[0067] Based on the direction of extension, multiple miniature shooting robots are automatically grouped and move in different directions, continuously collecting images of the sewage pipe network at different locations.

[0068] In this application, taking a city's sewage pipe network as an example, multiple miniature imaging robots were designed and manufactured. Each robot measures 10 cm in diameter and 15 cm in height, equipped with a high-definition camera and a waterproof shell. These robots were released into the pipe network from a sewage pipe outlet in the city center. Upon entering the network, the robots immediately began capturing images of the pipe's interior, obtaining the first set of sewage pipe network images. Subsequently, the robots performed edge recognition processing on the first captured image, identifying the pipe's outline using image processing algorithms and calculating its direction of extension. For example, in one practical application, the robot identified the current pipe's direction as northeast, with an angle deviation of 15 degrees. Based on this result, the robots automatically divided into three groups: the first group moved northeast, the second group moved southeast, and the third group moved due north. Each group of robots moved within the pipe at a speed of 5 meters per minute, capturing a new sewage pipe network image every meter moved. During this process, the robots continuously collected images from different locations. For example, the first group of robots took 50 images after advancing 50 meters, the second group took 40 images after advancing 40 meters, and the third group took 60 images after advancing 60 meters. In this way, the robots efficiently covered multiple branch areas in the pipeline network, ensuring the comprehensiveness of the data.

[0069] The miniature imaging robot acquires its current spatial location and the time of the shot, and generates a watermark on the sewage pipe network image based on the current spatial location and the time of the shot, resulting in a watermarked pipe network image. The specific steps are as follows:

[0070] Obtain the location data of the sewage pipe outlet, and generate the spatial origin based on the pipe outlet location data;

[0071] A miniature camera robot records the spatial origin and constructs a three-dimensional virtual blank space based on the spatial origin;

[0072] Based on images of the sewage pipe network, data on the relative position change of the miniature imaging robot relative to the spatial origin were obtained.

[0073] The spatial position change of the miniature camera robot is recorded in a three-dimensional virtual blank space based on the relative position change data, and the current spatial position of the miniature camera robot is obtained.

[0074] The system obtains the shooting time of each sewage pipe network image taken by the miniature shooting robot, combines the shooting time and current spatial location to generate a watermark, and adds the watermark to each corresponding sewage pipe network image to obtain a watermarked pipe network image.

[0075] In practical application, taking a city's sewage pipe network as an example, the robot acquires precise location data of the sewage pipe openings. This data is obtained through GPS positioning, with coordinates (30.5 degrees North latitude, 120.3 degrees East longitude). The robot records this location as the spatial origin and constructs a three-dimensional virtual blank space model within its internal system based on this origin. Subsequently, as the robot moves within the pipe, it records its displacement data relative to the spatial origin in real time through its built-in inertial navigation system. For example, after moving 30 meters, the robot records a 25-meter offset on the X-axis, a 15-meter offset on the Y-axis, and a 2-meter descent on the Z-axis. Simultaneously, the robot records the time of each image capture; for example, the first image was captured at 10:00:00 on June 1, 2025, and the second image at 10:00:30. Combining the current spatial location (e.g., X=25, Y=15, Z=-2) and the time of capture, the robot generates a watermark in the format "Location: (25, 15, -2); Time: 2025-06-01 10:00:00". This watermark is automatically added to the lower right corner of the corresponding sewage network image, generating a watermarked network image. For example, the first image with the watermark is saved as "Pic_001.jpg", and the second as "Pic_002.jpg". In this way, each image carries precise spatial and temporal information, providing crucial data for subsequent network modeling.

[0076] Based on the pipeline image data and watermark data, the spatial location and internal data of the pipeline are obtained. The specific steps for constructing a simulated urban sewage network based on the pipeline's spatial location and internal data are as follows:

[0077] Based on the pipeline image data, multiple pipeline image data are stitched together to obtain a complete initial pipeline cross-section image;

[0078] Extract the image offset of the initial pipe cross-section image, generate an image correction value based on the image offset, and perform image correction on the pipe cross-section based on the image correction value to obtain the target pipe cross-section image;

[0079] Multiple target pipe cross-section images are arranged sequentially in time to generate a three-dimensional image of the pipe's interior.

[0080] Based on the internal images of the 3D pipe, extract the connection points between pipes, the pipe tilt angle, and the internal material images;

[0081] By combining the connection location, pipe tilt angle, and internal material images, we can obtain internal pipe data.

[0082] Based on the watermark and connection location, the spatial location of each pipe segment is obtained. Combining the spatial location of the pipe and the internal data of the pipe, the pipe is simulated and constructed to generate a simulated urban sewage pipe network.

[0083] In application, taking a city's sewage pipe network as an example, the system collects pipe image data from all watermarked pipe network images, such as 2000 images of pipe cross-sections at different locations. These images are then stitched together; for example, images numbered 001 to 010 are stitched together to form a complete initial pipe cross-section image, covering a 50-meter-long pipe area. During the stitching process, some images are found to have image offsets; for example, image 005 has an offset of 3 pixels horizontally and 2 pixels vertically. The system calculates an image correction value of (-3, -2) and applies this correction value to adjust the images, obtaining the corrected target pipe cross-section image. Subsequently, all the corrected images are arranged in chronological order, for example, images 001 (time 10:00) to 200 (time 12:00), and a 3D reconstruction algorithm is used to generate a 3D image of the pipe's interior. The image clearly shows the connection points between pipes (e.g., pipe A connects to pipe B at coordinates (30, 20, -5)), the pipe inclination angles (e.g., pipe B has an inclination angle of 5 degrees), and the internal material composition (e.g., the sludge on the inner wall of pipe C is 2 cm thick). Combining this data, the system generates a data table of the pipe's internal structure, including connection type, inclination angle, and material distribution. Finally, based on the location data in the watermark (e.g., the starting coordinates of pipe A are (0, 0, 0), and the ending coordinates are (50, 0, -3)), the system constructs a simulation model of the pipes in three-dimensional space, generating a complete simulated urban sewage pipe network.

[0084] After arranging multiple target pipe cross-sectional images sequentially in chronological order to generate a 3D image of the pipe's interior, the following steps are also included:

[0085] Based on the three-dimensional image of the inside of the pipe, the inner wall of the pipe is scanned and identified to obtain the integrity value of the inner wall and the inner wall diameter data.

[0086] Based on the inner wall integrity value, the branch pipes connected to the main pipe are obtained; based on the inner wall diameter data, the pipe diameter variation data are obtained.

[0087] Based on the pipe diameter change data, extract the pipe segment diameter data of adjacent pipes, compare the pipe diameter change data and the pipe diameter data to obtain the diameter difference value, and generate the simulation parameter change value based on the diameter difference value;

[0088] Based on the branch pipe, the branch entrance is obtained, and the entrance spatial position of the branch entrance is extracted. An entrance marker is generated based on the entrance spatial position, and the entrance marker is sent to the subsequent miniature shooting robot.

[0089] In application, taking a city's sewage pipe network as an example, the system performs inner wall recognition on 3D images. For instance, it identifies a section of pipe with an inner wall integrity value of 95% (indicating 5% damage) and an inner wall diameter of 1.2 meters. Based on this integrity value, the system detects a branch pipe connected to this pipe at coordinates (40, 10, -4). Simultaneously, based on the inner wall diameter data, it finds that the branch pipe's diameter is 0.8 meters, creating a diameter difference from the main pipe's 1.2 meters. The system extracts the diameter data of adjacent pipe sections (e.g., upstream pipe diameter 1.2 meters, downstream pipe diameter 1.0 meter), calculating a diameter difference of 0.4 meters (1.2 - 0.8). Based on this difference, the system generates simulation parameter change values ​​for subsequent pipe network flow simulation. Furthermore, the system records the branch pipe's inlet spatial location as (40, 10, -4), generates an inlet marker "Branch Inlet: Coordinates (40, 10, -4)," and sends this marker to a miniature imaging robot that subsequently enters the pipe network. For example, after receiving a marker, robot R-001 adjusts its route to the branch entrance to take a picture.

[0090] After sending the entry marker to the subsequent miniature camera robot, the process also includes:

[0091] After receiving the entrance marker, the miniature camera robot enters the branch entrance and takes pictures of the branch pipes within the branch pipes;

[0092] Based on the branch pipe image, obtain the branch pipe diameter and determine whether the branch pipe diameter is greater than the inner wall diameter data;

[0093] If it is determined that the diameter of the branch pipe is greater than the inner wall diameter, then the branch pipe is marked as the superior pipe, and the shooting priority of the superior pipe is increased.

[0094] If it is determined that the diameter of the branch pipe is smaller than the inner wall diameter, the branch pipe is marked as a lower-level pipe, and the shooting priority of the lower-level pipe is reduced.

[0095] If it is determined that the diameter of the branch pipe is equal to the inner wall diameter, the branch pipe will be marked as a pipe of the same level, and the shooting priority of pipes of the same level will not be changed.

[0096] In application, taking a city's sewage pipe network as an example, after receiving a branch entrance marker, the miniature imaging robot performs the following operations: Robot R-001 enters the branch entrance at coordinates (40, 10, -4) and captures images of the branch pipe. Image analysis reveals that the branch pipe diameter is 0.8 meters, while the main pipe's inner wall diameter is 1.2 meters. The system determines that the branch pipe diameter (0.8 meters) is smaller than the main pipe's inner wall diameter (1.2 meters), therefore marking the branch pipe as a "lower-level pipe" and adjusting its imaging priority from "high" to "medium." Simultaneously, the robot continues to capture images within the branch pipe, for example, advancing 20 meters and capturing 20 new images. If the branch pipe diameter is 1.5 meters (greater than the main pipe's 1.2 meters), it is marked as a "higher-level pipe," its priority is raised to "highest," and more robots are dispatched to the branch. If the diameters are equal (e.g., both 1.2 meters), it is marked as a "same-level pipe," and its priority remains unchanged. Through priority adjustment, the system ensures efficient coverage of critical pipes.

[0097] After the steps of simulating and constructing the pipeline to generate a simulated urban sewage pipe network, the following steps are also included:

[0098] The simulated urban sewage pipe network is traversed to obtain the pipe extension value of each pipe in the simulated urban sewage pipe network, as well as the logical reasonable value of the pipe connection.

[0099] Based on the pipeline extension value, determine whether there is abnormal extension in the simulated urban sewage pipe network;

[0100] If it is determined that there is an abnormal extension in the simulated urban sewage pipe network, then extract the abnormal extension section in the simulated urban sewage pipe network and the abnormal pipe network image corresponding to the abnormal extension section.

[0101] Anomaly assessment is performed on abnormal pipeline images to determine the cause of the anomaly, and the abnormal extension section is reconstructed based on the cause of the anomaly.

[0102] Based on logically reasonable values, determine whether there are any abnormal connections in the simulated urban sewage pipe network.

[0103] If an abnormal connection is identified in the pipeline connection, extract the images of the adjacent abnormal pipelines at both ends of the abnormal connection, and determine the type of abnormal connection based on the images of the adjacent abnormal pipelines.

[0104] Automatic correction is performed on abnormal connections based on their type.

[0105] In practice, taking a city's sewage pipe network as an example, after simulating the construction of the urban sewage pipe network, the system performs a traversal check: First, it calculates the extension value of each pipe. For example, the extension value of pipe A is 100 meters (from the starting point (0,0,0) to the ending point (100,0,-10)), and pipe B is 80 meters. If an abnormal extension value of a pipe is found (such as the extension value of pipe C being 200 meters, far exceeding the average), the system extracts the corresponding abnormal pipe network images (such as Pic_150 to Pic_160). Anomaly assessment is performed on these images, revealing that the anomaly is caused by an image stitching error leading to an inflated pipe length. The system automatically reconstructs the pipe segment, correcting the extension value to 50 meters. Second, it checks the logically reasonable values ​​at pipe connections. For example, pipes D and E connect at coordinates (80,30,-15), but the flow logic indicates that pipe E should be upstream. The system extracts images (Pic_170 and Pic_171) from both ends of the connection, identifies the abnormal connection type as "flow direction reversed", and automatically corrects the connection relationship to ensure that pipe E is located upstream.

[0106] The steps for traversing the simulated urban sewage pipe network to obtain the pipe extension value of each pipe in the simulated urban sewage pipe network, as well as the logically reasonable values ​​of the pipe connections, are as follows:

[0107] The simulated urban sewage pipe network is traversed to extract the complete network value, flow logic relationship and main branch pipe connection relationship of the simulated urban sewage pipe network;

[0108] Based on the integrity value of the pipeline network, the number of pipeline segment ends in the simulated urban sewage pipeline network is extracted, and the length of each pipeline segment end is extracted. Combining the number of pipeline segment ends and the length of the pipeline segment, the pipeline extension value of each pipeline is obtained.

[0109] Based on the logical relationship of flow direction and the connection relationship of main branch pipeline, extract the degree of consistency between the logical relationship of flow direction and the connection relationship of main branch pipeline;

[0110] Based on the degree of agreement, the logical rationality of the connection relationship between the main and branch pipelines is evaluated to obtain the logical rationality value of the pipeline connection.

[0111] In application, taking a city's sewage pipe network as an example, the system extracts a network integrity value of 98% (missing 2% coverage). The flow direction logic is "main road → branch," and the main and branch pipe connections include 10 main nodes and 20 branch nodes. The system calculates the number of pipe segment ends (e.g., 5 end points) and the pipe length at each end (e.g., end A is 15 meters long, end B is 20 meters long). Combining the number and length, the extension value of each pipe is obtained (e.g., the main pipe extension value is 120 meters). Subsequently, the system analyzes the consistency between the flow direction logic and the main-branch connections (e.g., consistency 90%). If the consistency is below 85% (e.g., a node has a consistency of 80%), the logical reasonable value is judged as "low," and the connection relationship is automatically corrected (e.g., the branch node is moved to a position consistent with the flow direction). In this way, the system ensures the logical reasonableness and structural integrity of the pipe network model.

[0112] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. An image recognition-based simulation construction method for a municipal sewer network, characterized by, The method comprises the following steps: constructing a miniature shooting robot, releasing the miniature shooting robot into a municipal sewer network from a sewer pipe opening, and performing picture shooting in the municipal sewer network to obtain a sewer network picture; the miniature shooting robot acquires its current spatial position and shooting time point, generates a marked watermark on the sewer network picture according to the current spatial position and the shooting time point, and obtains a watermark network picture; extracting pipe picture data and marked watermark data in the watermark network picture, acquiring pipe spatial position and pipe internal data of the pipe according to the pipe picture data and the marked watermark data, and constructing a simulated municipal sewer network according to the pipe spatial position and the pipe internal data; arranging a plurality of pipe picture data in time sequence to obtain a pipe picture sequence, and acquiring a continuous pipe feature change table according to the pipe picture sequence, obtaining sewer activity data according to the pipe feature change table, adding simulated sewer activity in the simulated municipal sewer network according to the sewer activity data, and obtaining a dynamic simulated municipal sewer network; the step of acquiring pipe spatial position and pipe internal data of the pipe according to the pipe picture data and the marked watermark data, and constructing a simulated municipal sewer network according to the pipe spatial position and the pipe internal data, specifically comprises: according to the pipe picture data, performing data splicing on a plurality of pipe picture data to obtain an initial pipe cross-section picture with complete pictures; extracting a picture offset of the initial pipe cross-section picture, generating a picture correction value according to the picture offset, performing picture correction on the pipe cross-section according to the picture correction value, and obtaining a target pipe cross-section picture; continuously arranging a plurality of target pipe cross-section pictures in time sequence to generate a three-dimensional pipe internal image; according to the three-dimensional pipe internal image, extracting a connection position between pipes, a pipe inclination angle, and an internal substance picture; combining the connection position, the pipe inclination angle, and the internal substance picture to obtain pipe internal data; based on the marked watermark and the connection position, acquiring pipe spatial position of each pipe, and combining the pipe spatial position and the pipe internal data to simulate and construct the pipe to generate a simulated municipal sewer network.

2. The method according to claim 1, wherein, The step of constructing a miniature shooting robot, releasing the miniature shooting robot into a municipal sewer network from a sewer pipe opening, and performing picture shooting in the municipal sewer network to obtain a sewer network picture, specifically comprises: constructing a plurality of miniature shooting robots, and releasing the miniature shooting robots into a municipal sewer network from a sewer pipe opening; after the miniature shooting robots enter the municipal sewer network, performing picture shooting in the municipal sewer network to obtain a sewer network picture; the miniature shooting robot performs edge recognition on the sewer network picture to obtain an extension direction of the pipe; According to the extension direction, a plurality of micro shooting robots are automatically grouped, respectively move towards different extension directions, and continuously collect pictures of the sewage pipe network at different positions.

3. The method according to claim 2, wherein, The micro shooting robot obtains its current spatial position and a shooting time point, generates a mark watermark on the sewage pipe network picture according to the current spatial position and the shooting time point, and obtains a watermark pipe network picture. Obtain the pipe opening position data of the sewage pipe opening, and generate a spatial origin according to the pipe opening position data; The micro shooting robot records the spatial origin and constructs a three-dimensional virtual blank space based on the spatial origin; Based on the sewage pipe network picture, the relative position change data of the micro shooting robot relative to the spatial origin is obtained; According to the relative position change data, the spatial position change of the micro shooting robot in the three-dimensional virtual blank space is recorded, and the current spatial position of the micro shooting robot is obtained; Obtain the shooting time point of the micro shooting robot when shooting each sewage pipe network picture, combine the shooting time point and the current spatial position, generate a mark watermark, and add the mark watermark to each corresponding sewage pipe network picture to obtain a watermark pipe network picture.

4. The method according to claim 3, wherein, After the step of arranging a plurality of target pipe section pictures in time sequence to generate a three-dimensional pipe internal image, the method further includes: According to the three-dimensional pipe internal image, the inner wall of the pipe is scanned and identified to obtain an inner wall integrity value and an inner wall diameter data of the pipe; According to the inner wall integrity value, a branch pipe connected with the pipe is obtained, and according to the inner wall diameter data, a variable diameter data of the pipe is obtained; According to the pipe variable diameter data, the pipe segment diameter data of the adjacent pipe is extracted, the pipe variable diameter data and the pipe diameter data are compared to obtain a diameter difference value, and a simulation parameter change value is generated according to the diameter difference value; According to the branch pipe, a branch entrance is obtained, and an entrance spatial position of the branch entrance is extracted, an entrance mark is generated according to the entrance spatial position, and the entrance mark is sent to the subsequent micro shooting robot.

5. The method according to claim 4, wherein, After the step of sending the entrance mark to the subsequent micro shooting robot, the method further includes: After receiving the entrance mark, the micro shooting robot enters the branch entrance and shoots a branch pipe picture in the branch pipe; According to the branch pipe picture, a branch pipe diameter of the branch pipe is obtained, and it is judged whether the branch pipe diameter is greater than the inner wall diameter data; If it is judged that the branch pipe diameter is greater than the inner wall diameter data, the branch pipe is marked as an upper-level pipe, and the shooting priority of the upper-level pipe is increased; If it is judged that the branch pipe diameter is less than the inner wall diameter data, the branch pipe is marked as a lower-level pipe, and the shooting priority of the lower-level pipe is reduced; If it is judged that the branch pipe diameter is equal to the inner wall diameter data, the branch pipe is marked as a same-level pipe, and the shooting priority of the same-level pipe is not changed.

6. The method according to claim 5, wherein, The step of simulating and constructing the pipeline to generate the simulated urban sewage pipe network further comprises: traversing the simulated urban sewage pipe network to obtain a pipeline extension value of each pipeline in the simulated urban sewage pipe network and a logically reasonable value of a pipeline connection; judging whether the simulated urban sewage pipe network has abnormal extension according to the pipeline extension value; if it is judged that the simulated urban sewage pipe network has abnormal extension, extracting an abnormal extension section in the simulated urban sewage pipe network and an abnormal pipe network picture corresponding to the abnormal extension section; performing abnormal evaluation on the abnormal pipe network picture to obtain an abnormal reason, and reconstructing the abnormal extension section according to the abnormal reason; judging whether the pipeline connection of the simulated urban sewage pipe network has abnormal connection according to the logically reasonable value; if it is judged that the pipeline connection has abnormal connection, extracting adjacent abnormal pipe network pictures at both ends of the abnormal connection, and obtaining an abnormal connection type according to the adjacent abnormal pipe network pictures; automatically correcting the abnormal connection according to the abnormal connection type.

7. The method according to claim 6, wherein, The step of traversing the simulated urban sewage pipe network to obtain a pipeline extension value of each pipeline in the simulated urban sewage pipe network and a logically reasonable value of a pipeline connection comprises: traversing the simulated urban sewage pipe network to extract a pipe network integrity value, a flow direction logical relationship and a main branch pipeline connection relationship of the simulated urban sewage pipe network; extracting a number of pipe network line segment ends existing in the simulated urban sewage pipe network according to the pipe network integrity value, and extracting a pipe network line segment length of each pipe network line segment end, and obtaining a pipeline extension value of each pipeline in combination with the number of pipe network line segment ends and the pipe network line segment length; extracting a degree of coincidence between the flow direction logical relationship and the main branch pipeline connection relationship according to the flow direction logical relationship and the main branch pipeline connection relationship; performing logical reasonableness evaluation on the main branch pipeline connection relationship according to the degree of coincidence to obtain a logically reasonable value of a pipeline connection.

Images