Urban sewage pipe network simulation construction method based on image recognition

Abstract

The invention discloses an urban sewage pipe network simulation construction method based on image recognition, and relates to the technical field of sewage pipe network simulation, and the method comprises the steps: constructing a miniature shooting robot, and enabling the miniature shooting robot to carry out the picture shooting in an urban sewage pipe network, and obtaining a sewage pipe network picture; the miniature shooting robot obtains a current spatial position and a shooting time point of the miniature shooting robot, generates a marking watermark and obtains a watermark pipe network picture; extracting pipeline picture data and marked watermark data in the watermark pipe network picture, and constructing a simulated urban sewage pipe network; arranging the multiple pieces of pipeline picture data according to a time sequence to obtain a pipeline picture sequence and sewage activity data, and adding simulated sewage activities in the simulated urban sewage pipe network according to the sewage activity data to obtain a dynamic simulated urban sewage pipe network. The method has the effect of improving the accuracy and real-time performance of urban sewage pipe network simulation construction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sewage pipe network simulation, in particular to a city sewage pipe network simulation construction method based on image recognition. BACKGROUND

[0002] With the continuous improvement of urbanization level, problems such as structural damage, blockage and leakage will occur in city sewage pipe network, which seriously threatens the safety of city water environment and city infrastructure.

[0003] In the prior art, in the monitoring process of city sewage pipe network, a sewage pipe network model of city sewage pipe network is established to visualize the distribution status of city underground sewage pipe network. However, in actual use, the sewage pipe network model is generally a static model, and the pipe monitoring data is mostly abstract digital data, which cannot effectively and timely reflect the actual operation of city sewage pipe network. When the pipe network is abnormal, the digital data needs to be analyzed and processed, and the response time is long. On the other hand, with the continuous development of city, city sewage pipe network will change continuously. If the changes of pipe network cannot be recorded in time, there will be a blank in the supervision of pipe network in some areas. SUMMARY

[0004] The purpose of the present application is to provide a city sewage pipe network simulation construction method based on image recognition to solve the problems in the background art.

[0005] The present application provides a city sewage pipe network simulation construction method based on image recognition, which comprises: A miniature shooting robot is constructed, and the miniature shooting robot is released into city sewage pipe network from a sewage pipe opening. The miniature shooting robot shoots pictures in the city sewage pipe network to obtain sewage pipe network pictures. The miniature shooting robot obtains its current spatial position and 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. The pipe picture data and mark watermark data in the watermark pipe network picture are extracted, the pipe spatial position and pipe internal data of the pipe are obtained according to the pipe picture data and the mark watermark data, and a simulated city sewage pipe network is constructed according to the pipe spatial position and the pipe internal data. The plurality of pipe picture data is arranged in time sequence to obtain a pipe picture sequence, a continuous pipe feature change table is obtained according to the pipe picture sequence, sewage activity data is obtained according to the pipe feature change table, simulated sewage activities are added in the simulated city sewage pipe network according to the sewage activity data, and a dynamic simulated city sewage pipe network is obtained.

[0006] Preferably, the step of constructing a miniature shooting robot, releasing the miniature shooting robot into the urban sewer network from the sewer pipe opening, and making a picture of the urban sewer network in the urban sewer network by the miniature shooting robot, specifically comprises: constructing a plurality of miniature shooting robots, and releasing the plurality of miniature shooting robots into the urban sewer network from the sewer pipe opening; after the plurality of miniature shooting robots enter the urban sewer network, making a picture of the urban sewer network in the urban sewer network by the plurality of miniature shooting robots; the plurality of miniature shooting robots perform edge recognition on the picture of the urban sewer network to obtain an extension direction of the sewer network; according to the extension direction, the plurality of miniature shooting robots are automatically grouped and move towards different extension directions, and continuously collect pictures of the urban sewer network at different positions.

[0007] Preferably, the step of the miniature shooting robot obtaining a current spatial position and a shooting time point of itself, generating a mark watermark on the picture of the urban sewer network according to the current spatial position and the shooting time point, and obtaining a watermark picture of the urban sewer network, specifically comprises: obtaining pipe opening position data of the sewer pipe opening, and generating a spatial origin according to the pipe opening position data; the miniature shooting robot records the spatial origin and constructs a three-dimensional virtual blank space based on the spatial origin; based on the picture of the urban sewer network, the relative position change data of the miniature shooting robot relative to the spatial origin is obtained; according to the relative position change data, the spatial position change of the miniature shooting robot in the three-dimensional virtual blank space is recorded, and the current spatial position of the miniature shooting robot is obtained; the shooting time point of the miniature shooting robot when shooting each picture of the urban sewer network is obtained, a mark watermark is generated by combining the shooting time point and the current spatial position, and the mark watermark is added to each corresponding picture of the urban sewer network to obtain a watermark picture of the urban sewer network.

[0008] Preferably, the step of obtaining the spatial position of the pipe and the internal data of the pipe according to the pipe picture data and the mark watermark data, and constructing a simulated urban sewer network according to the spatial position of the pipe and the internal data of the pipe, specifically comprises: according to the pipe picture data, a plurality of pipe picture data are spliced to obtain an initial pipe cross-section picture with complete pictures; extracting a picture offset of the initial pipe section picture, generating a picture correction value according to the picture offset, correcting the pipe section according to the picture correction value to obtain a target pipe section picture; sequentially arranging the plurality of target pipe section pictures in time sequence to generate a three-dimensional pipe internal image; extracting a connection position between pipes, a pipe inclination angle, and an internal substance picture according to the three-dimensional pipe internal image; obtaining pipe internal data in combination with the connection position, the pipe inclination angle, and the internal substance picture; obtaining a pipe space position of each pipe segment based on the marker watermark and the connection position, and simulating and constructing the pipes in combination with the pipe space position and the pipe internal data to generate a simulated urban sewage pipe network.

[0009] Preferably, after the step of sequentially arranging the plurality of target pipe section pictures in time sequence to generate a three-dimensional pipe internal image, the method further comprises: scanning and identifying a pipe inner wall according to the three-dimensional pipe internal image to obtain an inner wall integrity value and an inner wall diameter data of the pipe inner wall; obtaining a branch pipe connected to the pipe according to the inner wall integrity value, and obtaining pipe diameter change data of the pipe according to the inner wall diameter data; extracting pipe segment diameter data of an adjacent pipe according to the pipe diameter change data, comparing the pipe diameter change data and the pipe diameter data to obtain a diameter difference value, and generating a simulation parameter change value according to the diameter difference value; obtaining a branch inlet according to the branch pipe, extracting an inlet space position of the branch inlet, generating an inlet marker according to the inlet space position, and sending the inlet marker to a subsequent micro-shooting robot.

[0010] Preferably, after the step of sending the inlet marker to the subsequent micro-shooting robot, the method further comprises: the micro-shooting robot enters the branch inlet and shoots a branch pipe picture in the branch pipe after receiving the inlet marker; obtaining a branch pipe diameter of the branch pipe according to the branch pipe picture, and judging 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, marking the branch pipe as an upper-level pipe and increasing a shooting priority of the upper-level pipe; if it is judged that the branch pipe diameter is less than the inner wall diameter data, marking the branch pipe as a lower-level pipe and decreasing a shooting priority of the lower-level pipe. 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.

[0011] Preferably, after the step of simulating and constructing the pipes to generate the simulated urban sewage pipe network, the method further comprises: traversing the simulated urban sewage pipe network to obtain a pipe extension value of each pipe in the simulated urban sewage pipe network and a logical reasonableness value of a pipe connection; judging whether the simulated urban sewage pipe network has abnormal extension according to the pipe 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 pipe connection of the simulated urban sewage pipe network has abnormal connection according to the logical reasonableness value; If it is judged that the pipe 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; According to the abnormal connection type, automatically correcting the abnormal connection.

[0012] Preferably, the step of traversing the simulated urban sewage pipe network to obtain a pipe extension value of each pipe in the simulated urban sewage pipe network and a logical reasonableness value of a pipe 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 pipe connection relationship of the simulated urban sewage pipe network; According to the pipe network integrity value, the number of pipe network line segment ends existing in the simulated urban sewage pipe network is extracted, and the length of each pipe network line segment end is extracted, and the pipe extension value of each pipe is obtained by combining the number of pipe network line segment ends and the length of the pipe network line segment; According to the flow direction logical relationship and the main branch pipe connection relationship, the degree of coincidence between the flow direction logical relationship and the main branch pipe connection relationship is extracted; According to the degree of coincidence, the logical reasonableness of the main branch pipe connection relationship is evaluated to obtain the logical reasonableness value of the pipe connection.

[0013] In summary, the present application has at least one of the following beneficial technical effects: The micro camera robot is released into the urban sewage pipe network through a sewage pipe opening such as a sewer opening. The micro camera robot takes pictures in the pipe after entering the urban sewage pipe network. The micro camera robot records its current spatial position in real time according to its moving route, combines the shooting time point of each sewage pipe network picture, generates a mark watermark, adds the mark watermark to the sewage pipe network picture, and generates a watermark pipe network picture. Then, the watermark pipe network picture is processed to obtain the spatial position of each pipe and the internal data of the pipe, including the inner diameter of the pipe and other data. The simulated urban sewage pipe network is constructed according to the above data. At the same time, the micro camera robot processes the sewage pipe network pictures in real time, and when a branch pipe is found in the pipe, it notifies subsequent micro camera robots to enter the branch pipe for work. Then, the inner diameter of the branch pipe is compared with the inner diameter of the current pipe, and the priority of the branch pipe is obtained. The higher the priority of the branch pipe, the more micro camera robots enter. At the same time, the micro camera robot will continue to enter other branch pipes found at other positions for work. Until enough watermark pipe network pictures are collected, and then the simulated urban sewage pipe network is constructed according to the watermark pipe network pictures. The pipe picture data is arranged in time sequence, and the sewage activity in the pipe is obtained. The sewage activity is reflected in the simulated urban sewage pipe network in real time, and the dynamic simulated urban sewage pipe network is obtained. The accuracy and real-time performance of the simulated urban sewage pipe network are improved. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 FIG. 1 is a step flow chart of a city sewage pipe network simulation construction method based on image recognition provided by an embodiment of the present application. DETAILED DESCRIPTION

[0015] The present application will be further described in detail below. Figure 1 The embodiments of the present application are not limited to the following, The embodiment of the present application discloses a city sewage pipe network simulation construction method based on image recognition.

[0016] In the embodiment, a city sewage pipe network simulation construction method based on image recognition includes: S100: Construct a micro camera robot, release the micro camera robot into the urban sewage pipe network from a sewage pipe opening, and take pictures in the urban sewage pipe network to obtain sewage pipe network pictures. S200: The micro camera robot obtains its current spatial position and the 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. S300: extract the pipeline picture data and the mark watermark data in the watermarked pipeline network picture, obtain the pipeline spatial position and the pipeline internal data of the pipeline according to the pipeline picture data and the mark watermark data, and construct the simulated urban sewage pipeline network according to the pipeline spatial position and the pipeline internal data; S400: arrange the plurality of pipeline picture data in time sequence to obtain a pipeline picture sequence, obtain a continuous pipeline feature change table according to the pipeline picture sequence, obtain sewage activity data according to the pipeline feature change table, add simulated sewage activity in the simulated urban sewage pipeline network according to the sewage activity data, and obtain a dynamic simulated urban sewage pipeline network.

[0017] It should be pointed out that the above process is only the basic step of the embodiment, and in the specific implementation process, part of the steps can be appropriately added, reduced or modified without affecting the overall implementation effect.

[0018] Constructing a miniature shooting robot, releasing the miniature shooting robot from the sewage pipeline opening into the urban sewage pipeline network, the miniature shooting robot performing picture shooting in the urban sewage pipeline network to obtain the sewage pipeline network picture, specifically comprising: Constructing a miniature shooting robot, releasing a plurality of miniature shooting robots from the sewage pipeline opening into the urban sewage pipeline network; After the miniature shooting robot enters the urban sewage pipeline network, the miniature shooting robot performs picture shooting in the urban sewage pipeline network to obtain the sewage pipeline network picture; The miniature shooting robot performs edge recognition on the sewage pipeline network picture to obtain the extension direction of the pipeline; According to the extension direction, the plurality of miniature shooting robots are automatically grouped, respectively move forward towards different extension directions, and continuously collect sewage pipeline network pictures at different positions.

[0019] In an example, a plurality of miniature camera robots are designed and manufactured, each with a diameter of 10 cm and a height of 15 cm, equipped with a high-definition camera and a waterproof shell. These robots are released into the sewer network from the sewer pipe opening in the central area of a city. Upon entering the network, the robots immediately begin filming the inside of the pipe, obtaining a first set of sewer network pictures. Subsequently, the robots perform edge recognition processing on the first picture taken, identifying the contour line of the pipe and calculating the extension direction of the pipe. For example, in an actual application, the robot identified the extension direction of the current pipe as northeast, with an angle offset of 15 degrees. Based on this result, the robots are automatically divided into three groups: the first group moves in the northeast direction, the second group moves in the southeast direction, and the third group moves in the north direction. Each group of robots moves at a speed of 5 meters per minute in the pipe, while taking a new sewer network picture every 1 meter. During this process, the robots continuously collect pictures at different positions, for example, the first group of robots takes 50 pictures after moving 50 meters, the second group takes 40 pictures after moving 40 meters, and the third group takes 60 pictures after moving 60 meters. In this way, the robots efficiently cover multiple branch areas in the network, ensuring the comprehensiveness of the data.

[0020] The miniature camera robot obtains its current spatial position and the time point of taking pictures, generates a mark watermark on the sewer network picture according to the current spatial position and the time point of taking pictures, and obtains a watermarked network picture, specifically as follows: Obtain the pipe opening position data of the sewer pipe opening, and generate a spatial origin according to the pipe opening position data; The miniature camera robot records the spatial origin and constructs a three-dimensional virtual blank space based on the spatial origin; Based on the sewer network pictures, obtain the relative position change data of the miniature camera robot relative to the spatial origin; According to the relative position change data, record the spatial position change of the miniature camera robot in the three-dimensional virtual blank space, and obtain the current spatial position of the miniature camera robot; Obtain the time point of taking pictures of the miniature camera robot for each sewer network picture, combine the time point of taking pictures and the current spatial position, generate a mark watermark, and add the mark watermark to each corresponding sewer network picture to obtain a watermarked network picture.

[0021] For example, a robot obtains the precise location data of a sewer pipe opening in a certain city. This data is obtained through GPS positioning, with coordinates of (North Latitude 30.5 degrees, East Longitude 120.3 degrees). The robot records this location as the spatial origin and constructs a three-dimensional virtual blank space model in the internal system based on this origin. Subsequently, while moving within the pipe, the robot records its displacement data relative to the spatial origin in real time through the built-in inertial navigation system. For example, a certain robot records that after moving 30 meters, it has an offset of 25 meters in the X-axis, 15 meters in the Y-axis, and a 2-meter drop in the Z-axis. At the same time, the robot records the shooting time point of each picture, such as the shooting time of the first picture being June 1, 2025, 10:00:00, and the second picture being 10:00:30. Combining the current spatial position (such as X=25, Y=15, Z=-2) and the shooting time point, the robot generates a watermark with the format "Position: (25, 15, -2); Time: 2025-06-01 10:00:00". This watermark is automatically added to the lower right corner of the corresponding sewer network picture, generating a watermarked sewer network picture. For example, the first picture is saved as "Pic_001.jpg" after adding the watermark, and the second picture is "Pic_002.jpg". In this way, each picture carries precise spatial and temporal information, providing key data for subsequent pipe network modeling.

[0022] According to the pipe picture data and the marked watermark data, the pipe space position and the pipe internal data are obtained, and according to the pipe space position and the pipe internal data, the steps of simulating the urban sewer network are constructed, specifically: According to the pipe picture data, multiple pipe picture data are spliced to obtain an initial pipe cross-section picture with complete pictures; Extract the picture offset of the initial pipe cross-section picture, generate a picture correction value according to the picture offset, correct the pipe cross-section according to the picture correction value, and obtain a target pipe cross-section picture; Arrange multiple target pipe cross-section pictures in chronological order to generate a three-dimensional pipe internal image; According to the three-dimensional pipe internal image, extract the connection position between pipes, the pipe inclination angle, and the internal material picture; Combine the connection position, the pipe inclination angle, and the internal material picture to obtain the pipe internal data; Based on the marked watermark and the connection position, the pipe space position of each pipe is obtained, and the pipe is simulated and constructed based on the pipe space position and the pipe internal data to generate a simulated urban sewer network.

[0023] In use, the system collects all pipe screen data in the watermarked pipe network pictures, such as 2000 pipe cross-section pictures in different locations in a certain city. The system processes the data splicing of these pictures, such as splicing pictures numbered 001 to 010 into a complete initial pipe cross-section picture covering a 50-meter pipe area. During the splicing process, it is found that some pictures have screen offset, such as picture 005 with a horizontal offset of 3 pixels and a vertical offset of 2 pixels. The system calculates the screen correction value as (-3, -2) and applies the correction value to adjust the picture to obtain the corrected target pipe cross-section picture. Subsequently, the corrected pictures are arranged in chronological order, such as picture 001 (time 10:00) to picture 200 (time 12:00), and a three-dimensional reconstruction algorithm is used to generate a three-dimensional pipe interior image. The image clearly shows the connection position between pipes (such as the connection of pipe A and pipe B at coordinate (30, 20, -5)), the pipe inclination angle (such as the inclination angle of pipe B being 5 degrees), and the internal material picture (such as the thickness of the sludge attached to the inner wall of pipe C being 2 cm). Combined with these data, the system generates a pipe interior data table including connection type, inclination angle, and material distribution. Finally, based on the position data in the marker watermark (such as the starting coordinate of pipe A being (0, 0, 0) and the ending coordinate being (50, 0, -3)), the system constructs a simulation model of the pipe in three-dimensional space to generate a complete simulation of the urban sewage pipe network.

[0024] After the step of arranging the plurality of target pipe cross-section pictures in chronological order to generate a three-dimensional pipe interior image, the method further comprises: According to the three-dimensional pipe interior image, scanning and identifying the inner wall of the pipe to obtain the inner wall integrity value and the inner wall diameter data of the inner wall of the pipe; According to the inner wall integrity value, obtaining the branch pipe connected to the pipe, and according to the inner wall diameter data, obtaining the variable diameter data of the pipe; According to the pipe variable diameter data, extracting the pipe segment diameter data of the adjacent pipe, comparing the pipe variable diameter data and the pipe diameter data to obtain the diameter difference value, and generating the simulation parameter change value according to the diameter difference value; According to the branch pipe, obtaining the branch inlet and extracting the inlet space position of the branch inlet, generating an inlet marker according to the inlet space position, and sending the inlet marker to the subsequent miniature shooting robot.

[0025] For example, the system identifies that the inner wall of a certain section of the pipeline is 95% complete (indicating that 5% of the area is damaged) and the inner wall diameter is 1.2 meters. According to the inner wall completeness value, the system detects that the pipeline is connected to a branch pipeline at the coordinate (40, 10, -4). At the same time, according to the inner wall diameter data, it is found that the diameter of the branch pipeline is 0.8 meters, which forms a diameter difference with the main pipeline of 1.2 meters. The system extracts the diameter data of adjacent pipe sections (such as the upstream pipe section diameter of 1.2 meters and the downstream pipe section diameter of 1.0 meters), calculates the diameter difference value of 0.4 meters (1.2-0.8). According to this difference value, the system generates a simulated parameter change value for subsequent pipe network flow simulation. In addition, the system records the entry space position of the branch pipeline as (40, 10, -4), generates an entry marker "branch entry: coordinate (40, 10, -4)", and sends the marker to the subsequent miniature shooting robot in the pipe network. For example, after receiving the marker, the robot R-001 adjusts the route to the branch entry for shooting.

[0026] After the step of sending the entry marker to the subsequent miniature shooting robot, it further includes: After receiving the entry marker, the miniature shooting robot enters the branch entry and shoots a branch pipeline picture in the branch pipeline; According to the branch pipeline picture, the branch pipeline diameter of the branch pipeline is obtained, and it is judged whether the branch pipeline diameter is greater than the inner wall diameter data; If it is judged that the branch pipeline diameter is greater than the inner wall diameter data, the branch pipeline is marked as an upper-level pipeline, and the shooting priority of the upper-level pipeline is increased; If it is judged that the branch pipeline diameter is less than the inner wall diameter data, the branch pipeline is marked as a lower-level pipeline, and the shooting priority of the lower-level pipeline is reduced; If it is judged that the branch pipeline diameter is equal to the inner wall diameter data, the branch pipeline is marked as a same-level pipeline, and the shooting priority of the same-level pipeline is not changed.

[0027] In use, taking a municipal sewer network of a certain city as an example, after receiving the branch entrance mark, the miniature shooting robot performs the following operations: the robot R-001 enters the branch entrance with coordinates (40, 10, -4), and shoots the picture of the branch pipe. Through image analysis, it is obtained that the diameter of the branch pipe is 0.8 meters, and the current inner wall diameter of the main pipe is 1.2 meters. The system judges that the diameter of the branch pipe (0.8 meters) is smaller than the inner wall diameter of the main pipe (1.2 meters), so the branch pipe is marked as “subordinate pipe”, and the shooting priority is adjusted from “high” to “medium”. At the same time, the robot continues to shoot the picture in the branch pipe, for example, advances 20 meters in the branch pipe and shoots 20 new pictures. If the diameter of the branch pipe is 1.5 meters (larger than the main pipe 1.2 meters), it is marked as “superior pipe”, the priority is raised to “highest”, and more robots are dispatched to enter the branch. If the diameters are equal (such as both 1.2 meters), it is marked as “same level pipe”, and the priority remains unchanged. Through priority adjustment, the system ensures the coverage efficiency of key pipes.

[0028] After the step of simulating and constructing the pipes to generate the simulated municipal sewer network, further comprising: traversing the simulated municipal sewer network to obtain a pipe extension value of each pipe in the simulated municipal sewer network, and a logically reasonable value of a pipe connection; judging whether the simulated municipal sewer network has abnormal extension according to the pipe extension value; if it is judged that the simulated municipal sewer network has abnormal extension, extracting an abnormal extension section in the simulated municipal sewer 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 pipe connection of the simulated municipal sewer network has abnormal connection according to the logically reasonable value; if it is judged that the pipe 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.

[0029] In use, taking the urban sewer network of a certain city as an example, after the urban sewer network is simulated to be constructed, the system performs traversal inspection: first, the extension value of each pipe is calculated, 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 the extension value of pipe B is 80 meters. If it is found that the extension value of a certain pipe is abnormal (for example, the extension value of pipe C is 200 meters, far exceeding the average value), the system extracts the corresponding abnormal pipe network picture (for example, Pic_150 to Pic_160). Abnormal evaluation is performed on these pictures, and it is found that the reason for the abnormality is that the picture splicing error leads to the virtual increase of the pipe length. The system automatically reconstructs the pipe section, and the extension value is corrected to 50 meters. Secondly, the logical reasonable value of the pipe connection is checked, for example, pipe D and pipe E are connected at the coordinate (80, 30, -15), but the flow direction logic shows that pipe E should be upstream. The system extracts the pictures at both ends of the connection (Pic_170 and Pic_171), identifies the abnormal connection type as “flow direction reversal”, and automatically corrects the connection relationship to ensure that pipe E is located upstream.

[0030] The steps of traversing the simulated urban sewer network to obtain the pipe extension value of each pipe in the simulated urban sewer network and the logical reasonable value of the pipe connection are as follows: Traverse the simulated urban sewer network, extract the pipe network complete value of the simulated urban sewer network, the flow direction logical relationship, and the main branch pipe connection relationship; According to the pipe network complete value, the number of pipe network line segment ends existing in the simulated urban sewer network is extracted, and the pipe network line segment length of each pipe network line segment end is extracted, and the pipe network line segment length is combined with the pipe network line segment end number to obtain the pipe extension value of each pipe; According to the flow direction logical relationship and the main branch pipe connection relationship, the degree of coincidence between the flow direction logical relationship and the main branch pipe connection relationship is extracted; According to the degree of coincidence, the logical reasonableness of the main branch pipe connection relationship is evaluated to obtain the logical reasonable value of the pipe connection.

[0031] In use, taking the urban sewer network of a certain city as an example, the system extracts the network integrity value as 98% (2% of the coverage is missing), the flow direction logical relationship is "main road → branch", the main branch pipe connection relationship includes 10 main nodes and 20 branch nodes. The number of pipe network line ends (such as 5 end points) is calculated, the length of each end pipe (such as end A is 15 meters long and end B is 20 meters long), and the extension value of each pipe is obtained by combining the number and length (such as the main pipe extension value is 120 meters). Then, the consistency of the flow direction logic and the main branch connection is analyzed (such as the consistency is 90%). If the consistency is lower than 85% (such as the consistency of a certain node is 80%), it is determined that the logic reasonable value is "low", and the connection relationship is automatically corrected (such as the branch node is shifted to the position consistent with the flow direction). In this way, the system ensures the logical rationality and structural integrity of the network model.

[0032] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application, therefore: any equivalent changes made according to the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method for simulating and constructing urban sewage pipe networks based on image recognition, characterized in that, include: 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. 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. 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. 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.

2. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 1, characterized in that, The steps of constructing a miniature imaging robot, releasing the miniature imaging robot from a sewage pipe outlet into the urban sewage pipe network, and having the miniature imaging robot capture images of the urban sewage pipe network to obtain images of the sewage pipe network are as follows: Construct a miniature camera robot, and release multiple of the miniature camera robots from the sewage pipe outlet into the urban sewage pipe network; 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. The miniature camera robot performs edge recognition on the sewage pipe network image to obtain the extension direction of the pipe; 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.

3. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 2, characterized in that, 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. The specific steps are as follows: Obtain the location data of the sewage pipe opening, and generate a spatial origin based on the pipe opening location data; The miniature camera robot records the spatial origin and constructs a three-dimensional virtual blank space based on the spatial origin; Based on the sewage pipe network images, the relative position change data of the miniature imaging robot relative to the spatial origin is obtained; 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. 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.

4. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 3, characterized in that, The steps for 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, are as follows: Based on the pipeline image data, multiple pipeline image data are stitched together to obtain a complete initial pipeline cross-section image; 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; 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. 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. By combining the connection location, the pipe tilt angle, and the internal material image, the internal data of the pipe is obtained; 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.

5. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 4, characterized in that, 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: 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. 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. 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; 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.

6. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 5, characterized in that, Following the step of sending the entry marker to the subsequent miniature camera robot, the method further includes: After receiving the entrance marker, the miniature camera robot enters the branch entrance and captures images of the branch pipe within the branch pipe. 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; 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; 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; 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.

7. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 6, characterized in that, After the steps of simulating and constructing the pipeline to generate a simulated urban sewage pipe network, the following steps are also included: 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. Based on the pipeline extension value, it is determined whether the simulated urban sewage pipe network has experienced abnormal extension; 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. Anomaly assessment is performed on the abnormal pipeline images to determine the cause of the anomaly, and the abnormal extension section is reconstructed based on the cause of the anomaly. Based on the logically reasonable value, determine whether there are any abnormal connections in the pipe connections of the simulated urban sewage pipe network; 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. Based on the type of abnormal connection, the abnormal connection is automatically corrected.

8. The method for simulating and constructing an urban sewage pipe network based on image recognition according to claim 7, characterized in that, The steps 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 are as follows: 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; 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. 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; 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.

Images