Water affair pipe network identification method and system based on abnormal detection and rephotographing of unmanned aerial vehicle

Abstract

This invention relates to the field of water pipe network inspection technology, and discloses a method and system for identifying water pipe networks based on UAV-based anomaly detection and supplementary image capture. The method includes: dividing the detection area into district-level sections and matching differentiated inspection routes; the UAV performing inspection tasks according to the differentiated inspection routes to construct an initial full-coverage image dataset; spatially comparing the pipe network features in the initial full-coverage image dataset with GIS data to identify abnormal areas and generate an anomaly supplementary image list; the UAV performing secondary supplementary images based on the supplementary image list to obtain supplementary images; aligning and fusing the initial full-coverage image dataset and the supplementary images; reconstructing the abnormal areas using three-dimensional calculations to invert the anomaly morphology parameters and performing bidirectional calibration to generate a water pipe network anomaly identification list. This invention solves the problems of incomplete data, misjudgment of anomalies, and fuzzy parameters in traditional UAV pipe network detection, achieving the dual goals of accurate detection and efficient operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of water pipeline network detection technology, and more specifically, to a method and system for identifying water pipeline networks based on anomaly detection and re-photographing by unmanned aerial vehicles (UAVs). Background Technology

[0002] As the core carrier of urban water supply and drainage systems, the monitoring of water pipe networks and the investigation of potential hazards directly affect the efficiency of municipal operation and maintenance. Traditional manual inspections and underground pipeline detectors suffer from high labor intensity, low efficiency, and blind spots in deeply buried pipe networks, making it difficult to achieve full coverage. With the popularization of UAV remote sensing mapping technology, pipe network identification methods based on aerial imagery are gradually becoming mainstream, but existing technologies still have the following inherent shortcomings: The inspection routes mostly adopt a unified patrol mode across the entire area, without combining the basic data of water pipe network GIS to carry out differentiated management and control by region. This makes it difficult to balance detection efficiency and image acquisition accuracy in underground pipe network coverage areas, surface facility areas, and key monitoring areas. Initial inspection images are prone to problems such as blurriness, missed shots, and insufficient targeting.

[0003] Anomaly identification often relies on a single comparison between inspection images and GIS data, without simultaneously conducting image quality assessment and multi-dimensional data verification. Influenced by factors such as lighting and occlusion, blurred images are easily misjudged as anomalies or real potential hazards are overlooked, resulting in persistently high false positive and false negative rates. Furthermore, existing methods can only achieve two-dimensional qualitative identification and cannot use three-dimensional calculations to retrieve quantitative parameters such as pipeline burial depth, direction, and offset, making it difficult to support precise operation and maintenance with sufficient positioning accuracy.

[0004] In addition, existing technologies lack a closed-loop anomaly re-shooting mechanism. If the initial images are incomplete or missed, a full-area return inspection is required, which is energy-intensive and time-consuming. Furthermore, the technology has not formed an integrated process of zoned patrol, anomaly identification, targeted re-shooting, and parameter calibration. It has poor standardization and replicability, making it difficult to achieve the coordinated and unified accurate detection and efficient operation and maintenance of large-scale urban pipe networks.

[0005] There are currently no effective solutions to the problems in the relevant technologies. Summary of the Invention

[0006] To address the problems in related technologies, this invention proposes a method and system for identifying water pipe networks based on anomaly detection and re-photographing by unmanned aerial vehicles (UAVs), thereby overcoming the aforementioned technical problems existing in existing related technologies.

[0007] Therefore, the specific technical solution adopted by the present invention is as follows: According to a first aspect of the present invention, a method for identifying water pipe networks based on anomaly detection and re-enhancing images taken by unmanned aerial vehicles (UAVs) is provided, the method comprising: S1. Based on the basic GIS data of the water supply network, the detection area is divided into districts, and differentiated inspection routes are matched according to the district division results. S2. The UAV performs inspection tasks according to the differentiated inspection route and constructs an initial full-coverage image dataset by acquiring visible light images, multispectral images and real-time GPS coordinates of the detection area. S3. Spatially correlate and compare the pipeline network features in the initial full-coverage image dataset with the GIS data, and combine the image clarity and feature integrity to identify abnormal areas in order to generate an abnormal reshoot list. S4. The drone takes a second shot of the abnormal area according to the reshoot list to obtain the reshoot image of the abnormal area; S5. Align and fuse the initial full-coverage image dataset with the re-shot images. Reconstruct the abnormal areas using 3D calculations to invert the burial depth, direction, pipe diameter, and abnormal morphological parameters of the pipeline network. Perform bidirectional calibration with the water pipeline network GIS basic data to generate a list of water pipeline network anomalies.

[0008] Furthermore, based on the GIS data of the water supply network, the detection area is divided into districts, and differentiated inspection routes are matched according to the district division results, including the following steps: S11. Obtain basic GIS data of water pipe network, extract information on pipe network type, burial depth level and surface marker distribution, as the basis for district-level division; S12. Based on the district-level division criteria, the detection area is divided into the underground pipeline coverage area, the surface facility area, and the key monitoring area; S13. For the detection needs of each district level, set the corresponding cruise parameters and shooting parameters respectively, and generate differentiated inspection routes that match each district level.

[0009] Furthermore, the pipeline network features in the initial full-coverage image dataset are spatially correlated and compared with GIS data. Combined with image clarity and feature integrity, abnormal areas are identified to generate an abnormal re-enhancing list, including the following steps: S31. Extract pipeline network features from the initial full-coverage image dataset, and combine the water pipeline network GIS data to perform spatial location association and morphological matching comparison of the pipeline network features to obtain suspected abnormal points. S32. Analyze the sharpness of each frame in the initial full-coverage image dataset, and filter out invalid image frames that are blurry, occluded, or incomplete to obtain the image defect areas. S33. Integrate suspected abnormal locations and image defect areas, identify abnormal areas, and extract the coordinates, coverage area, and shooting parameters of abnormal areas to generate an abnormal reshoot list.

[0010] Furthermore, pipeline network features are extracted from the initial full-coverage image dataset. These features are then combined with water pipeline network GIS data to perform spatial location correlation and morphological matching comparisons to identify suspected anomaly locations. This process includes the following steps: S311. Extract linear utility tunnel features, point facility features, and surface spectral anomaly features from the initial full-coverage image dataset as the utility network features to be compared. S312. Retrieve the theoretical pipeline vector data, facility coordinates, and attribute information of the corresponding area from the water supply network GIS data as reference data; S313. Unify the pipeline network features to be compared with the reference data into the same spatial coordinate system, perform spatial location overlay analysis and morphological matching comparison, and mark the points with missing features, morphological changes, location offsets and surface anomalies without corresponding GIS markers to form suspected anomaly points.

[0011] Furthermore, the initial full-coverage image dataset is aligned and fused with the re-captured images. Anomalies in the network are reconstructed using 3D calculations, and the abnormal morphological parameters are retrieved. This process is then bidirectionally calibrated with the water supply network GIS basic data to generate a list of anomalies in the water supply network. The steps include: S51. Preprocess the initial full-coverage image dataset and the re-shot images, and construct an enhanced image dataset of abnormal areas through spatiotemporal alignment. S52. Based on the enhanced image dataset, use the spatial solution and dense matching algorithm of multi-view images to perform three-dimensional dense reconstruction of the abnormal area and generate a three-dimensional model of the abnormal area. S53. In the three-dimensional model of the abnormal area, based on the spatial curvature of the pipe gallery edge, the geometric shape of the manhole opening and the abnormal distribution of the ground elevation, the abnormal morphological parameters of the pipe network are obtained through spatial measurement and geometric calculation. S54. Spatially overlay and compare the abnormal morphological parameters with the basic GIS data of the water supply network and perform bidirectional calibration to generate an abnormal identification list of the water supply network.

[0012] Furthermore, based on the enhanced image dataset, spatial computation and dense matching algorithms for multi-view images are used to perform 3D dense reconstruction of the abnormal region, generating a 3D model of the abnormal region, including the following steps: S521. Extract feature points from each image in the enhanced image dataset, perform cross-image feature point matching, and construct corresponding relationships of corresponding points. S522. Based on the correspondence of corresponding points, the image pose parameters of each image are obtained through spatial geometric calculation of multi-view images, and a sparse three-dimensional point cloud is generated. S523. Based on the sparse 3D point cloud and image pose parameters, perform pixel-by-pixel dense matching on the enhanced image to generate a dense 3D point cloud. S524. Perform meshing and texture mapping on the dense 3D point cloud to construct a 3D model of the abnormal region.

[0013] Furthermore, in the three-dimensional model of the abnormal area, based on the spatial curvature of the pipe gallery edge, the geometric shape of the manhole opening, and the abnormal distribution of surface elevation, the abnormal morphological parameters of the pipe network are obtained through spatial measurement and geometric calculation, including the following steps: S531. The three-dimensional model of the abnormal area is segmented to obtain the pipe gallery structure, manhole opening and abnormal surface elevation block; S532. Extract the point cloud of the pipe gallery edge along the pipe gallery structure, and use curvature calculation and trajectory fitting to obtain the pipeline offset and bending parameters; S533. Perform geometric shape fitting on the inspection wellhead, identify the abnormal state of the wellhead, and extract the wellhead coordinates and relative elevation parameters. S534. Measure the elevation difference and delineate the range of the abnormal surface elevation blocks, and invert the surface subsidence parameters and surface leakage parameters by combining the pipeline burial depth data. S535, by integrating pipeline offset, bend parameters, wellhead coordinates, wellhead relative elevation parameters, surface subsidence parameters, and surface leakage parameters, abnormal morphology parameters are obtained.

[0014] According to a second aspect of the present invention, a water pipe network identification system based on unmanned aerial vehicle (UAV) detection and anomaly re-photography is provided, the system comprising: The inspection route planning module is used to divide the detection area into districts based on the basic GIS data of the water pipe network, and match differentiated inspection routes according to the district division results. The initial full-coverage image dataset construction module is used by UAVs to perform inspection tasks according to differentiated inspection routes and to construct an initial full-coverage image dataset by acquiring visible light images, multispectral images and real-time GPS coordinates of the detection area. The abnormal area identification module is used to spatially correlate and compare the pipeline network features in the initial full coverage image dataset with GIS data, and combine image clarity and feature integrity to identify abnormal areas and generate an abnormal reshoot list. The abnormal area reshooting module is used by the drone to reshoot abnormal areas according to the reshooting list, so as to obtain reshoot images of the abnormal areas; The pipeline anomaly identification module is used to align and fuse the initial full-coverage image dataset with the re-captured images. By performing three-dimensional calculation and reconstruction on the abnormal areas, it inverts the burial depth, direction, pipe diameter and abnormal morphological parameters of the pipeline network, and performs bidirectional calibration with the water pipeline network GIS basic data to generate a list of water pipeline network anomalies.

[0015] According to a third aspect of the present invention, a computer device is provided.

[0016] In some embodiments, the computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method described above.

[0017] According to a fourth aspect of the present invention, a computer-readable storage medium is provided.

[0018] In one embodiment, a computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the steps of the above method.

[0019] The beneficial effects of this invention are as follows: 1. This invention establishes a district-level zoning mechanism based on pipeline network GIS data and combines differentiated inspection routes and targeted supplementary shooting rules to achieve refined and efficient pipeline network detection. It can accurately supplement missed or blurred images during the initial inspection, significantly improving the completeness and relevance of pipeline network image acquisition. 2. This invention constructs a triple-linked identification system of spatial feature comparison, image quality judgment, and three-dimensional reconstruction calculation. Through multi-dimensional data verification and analysis, it not only ensures the accuracy of anomaly identification but also complements the shortcomings of images, avoiding the omissions and errors of traditional single inspection and single comparison. 3. This invention solves the problems of incomplete data, misjudgment of anomalies, and fuzzy parameters in traditional UAV pipeline network detection by using zoned differential cruise, anomaly triggering reshoot, and three-dimensional quantitative inversion system. It significantly improves the accuracy of pipeline burial depth inversion, anomaly location accuracy, and inspection efficiency, and achieves the dual goals of accurate detection and efficient operation and maintenance. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of a water pipeline network identification method based on anomaly detection and re-photographing by unmanned aerial vehicles according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a water pipeline network identification system based on unmanned aerial vehicle (UAV) detection and anomaly re-photographing, according to an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of a computer device according to an exemplary embodiment.

[0022] In the picture: 1. Inspection route planning module; 2. Initial full-coverage image dataset construction module; 3. Abnormal area judgment module; 4. Abnormal area re-shooting module; 5. Pipeline network anomaly identification module. Detailed Implementation

[0023] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0024] According to embodiments of the present invention, a method and system for identifying water pipe networks based on anomaly detection and re-photographing by unmanned aerial vehicles (UAVs) are provided.

[0025] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, a water supply network identification method based on unmanned aerial vehicle (UAV) detection and anomaly re-enhancing photography includes the following: S1. Based on the basic GIS data of the water supply network, the detection area is divided into districts, and differentiated inspection routes are matched according to the district division results. In this optional embodiment, based on the water pipe network GIS basic data, the detection area is divided into districts, and differentiated inspection routes are matched according to the district division results, including the following steps: S11. Obtain basic GIS data of water pipe network, extract information on pipe network type, burial depth level and surface marker distribution, as the basis for district-level division; S12. Based on the district-level division criteria, the detection area is divided into the underground pipeline coverage area, the surface facility area, and the key monitoring area; Specifically, the underground pipeline coverage area corresponds to areas with no exposed surface facilities and only underground pipelines; the surface facility area corresponds to areas with shallow pipeline burial depth and surface facilities such as inspection wells and valve wells; and the key monitoring area corresponds to high-risk areas such as old pipeline sections, locations prone to leakage and collapse, and pipelines in densely populated areas.

[0026] S13. For the detection needs of each district level, set the corresponding cruise parameters and shooting parameters respectively, and generate differentiated inspection routes that match each district level.

[0027] It should be explained that, for areas covered by underground pipe networks, a full-coverage cruise mode is adopted, setting a first flight altitude and wide-angle lens shooting parameters, and planning a scanning cruise path covering the entire area; for ground facility areas, a facility detailed inspection cruise mode is adopted, setting a second flight altitude and telephoto lens shooting parameters, and planning a straight cruise path along the pipe network; for key monitoring areas, a surround detailed inspection cruise mode is adopted, setting hovering and multi-angle surround flight paths, adjusting the flight altitude and shooting angle according to the target point coordinates, and setting the highest image resolution and highest sampling frame rate for shooting. The first flight altitude must ensure that the entire ground surface is captured without obstruction; the second flight altitude is lower than the first flight altitude and must be closer to the ground facilities to ensure clarity of details.

[0028] S2. The UAV performs inspection tasks according to the differentiated inspection route and constructs an initial full-coverage image dataset by acquiring visible light images, multispectral images and real-time GPS coordinates of the detection area. It should be explained that the initial full-coverage image dataset construction steps are as follows: Using real-time GPS coordinates and attitude parameters, the acquired visible light images and multispectral images are spatially registered frame by frame through collinearity equations or projection transformation algorithms to eliminate parallax caused by differences in sensor installation positions and flight jitter, generating a multi-source image sequence with accurate geocoding; Based on the overlap and geographic coordinates of each image, feature point matching and seamless stitching algorithms are used to geometrically correct, radiometrically even out, and mosaick the discrete single-frame images in a unified spatial coordinate system, fusing them to form the initial full-coverage image dataset.

[0029] S3. Spatially correlate and compare the pipeline network features in the initial full-coverage image dataset with the GIS data, and combine the image clarity and feature integrity to identify abnormal areas in order to generate an abnormal reshoot list. In this optional embodiment, spatial correlation comparison is performed between the pipeline network features in the initial full-coverage image dataset and the GIS data, and abnormal areas are identified by combining image clarity and feature integrity to generate an abnormal re-enhancing list, including the following steps: S31. Extract pipeline network features from the initial full-coverage image dataset, and combine the water pipeline network GIS data to perform spatial location association and morphological matching comparison of the pipeline network features to obtain suspected abnormal points. Specifically, the linear utility tunnel features reflect the planar orientation and continuity of the pipelines, the point-like facility features correspond to the distribution of exposed facilities on the ground, and the surface spectral anomaly features indirectly reflect possible anomalies such as leakage and collapse in the underground pipe network through spectral information. S32. Analyze the sharpness of each frame in the initial full-coverage image dataset, and filter out invalid image frames that are blurry, occluded, or incomplete to obtain the image defect areas. Specifically, sharpness assessment can employ conventional methods in image processing, such as Laplacian variance and grayscale gradient statistics, to calculate a quantifiable value for the sharpness of each frame and compare it with a preset threshold, filtering out image frames that fall below the quality requirements. Simultaneously, by combining the integrity assessment of target features in the image, image defect areas that cannot be accurately assessed are further identified, and these areas are mapped to their corresponding spatial locations.

[0030] S33. Integrate suspected abnormal locations and image defect areas, identify abnormal areas, and extract the coordinates, coverage area, and shooting parameters of abnormal areas to generate an abnormal reshoot list.

[0031] It should be explained that if a suspected anomaly is located within an image defect area, it means that the image quality of that location is insufficient to support an accurate judgment, and a new high-quality image needs to be captured for verification. If a suspected anomaly is located within an image quality acceptable area, it will be directly listed as an anomaly to be verified. If there are no suspected anomalies within an image defect area, the area still needs to be re-captured to fill the data blind spot and avoid missing real potential problems due to image quality issues.

[0032] In this optional embodiment, extracting pipeline network features from the initial full-coverage image dataset and combining them with water pipeline network GIS data to perform spatial location correlation and morphological matching comparison of the pipeline network features to obtain suspected anomaly locations includes the following steps: S311. Extract linear utility tunnel features, point facility features, and surface spectral anomaly features from the initial full-coverage image dataset as the utility network features to be compared. S312. Retrieve the theoretical pipeline vector data, facility coordinates, and attribute information of the corresponding area from the water supply network GIS data as reference data; S313. Unify the pipeline network features to be compared with the reference data into the same spatial coordinate system, perform spatial location overlay analysis and morphological matching comparison, and mark the points with missing features, morphological changes, location offsets and surface anomalies without corresponding GIS markers to form suspected anomaly points.

[0033] It should be explained that points meeting any of the following conditions will be marked as suspected anomalies: 1. There are obvious surface anomalies in the image but no corresponding risk markers in the GIS; 2. The facilities marked in the GIS are completely missing or severely obscured in the image; 3. The direction of pipelines in the image deviates from the GIS path by more than a preset threshold; 4. The form of facilities in the image is significantly different from the standard form recorded in the GIS.

[0034] S4. The drone takes a second shot of the abnormal area according to the reshoot list to obtain the reshoot image of the abnormal area; It's important to clarify that secondary reshoots are not simply repetitive shots, but rather targeted data collection based on the anomaly types and image defect characteristics listed in the anomaly reshoot list. For example, for defective areas caused by image blur, reshoots improve sharpness by reducing flight speed, increasing exposure time, or adjusting focus parameters; for feature loss due to a single viewpoint, reshoots supplement the viewpoint by using multi-angle surround shooting; and for areas requiring 3D reconstruction, reshoots ensure that the overlap between adjacent images meets the requirements for 3D calculation.

[0035] S5. Align and fuse the initial full-coverage image dataset with the re-shot images. Reconstruct the abnormal areas using 3D calculations to invert the burial depth, direction, pipe diameter, and abnormal morphological parameters of the pipeline network. Perform bidirectional calibration with the water pipeline network GIS basic data to generate a list of water pipeline network anomalies.

[0036] In this optional embodiment, the initial full-coverage image dataset is aligned and fused with the re-captured images. Anomalies in the pipeline network are then reconstructed using 3D calculations of the abnormal areas to retrieve the abnormal morphological parameters. Finally, a bidirectional calibration is performed with the water pipeline network GIS basic data to generate a list of anomalies in the water pipeline network. This process includes the following steps: S51. Preprocess the initial full-coverage image dataset and the re-shot images, and construct an enhanced image dataset of abnormal areas through spatiotemporal alignment. Specifically, the initial full-coverage image dataset and the re-captured images undergo preprocessing, mainly including basic processing such as image distortion correction and radiometric normalization, to eliminate the influence of lens distortion and differences in lighting conditions between different batches of images. Spatiotemporal alignment is based on the precise GPS coordinates and timestamp information recorded in each frame of the image, to perform spatial location matching and temporal sequence sorting of all images of the same anomalous area.

[0037] S52. Based on the enhanced image dataset, use the spatial solution and dense matching algorithm of multi-view images to perform three-dimensional dense reconstruction of the abnormal area and generate a three-dimensional model of the abnormal area. S53. In the three-dimensional model of the abnormal area, based on the spatial curvature of the pipe gallery edge, the geometric shape of the manhole opening and the abnormal distribution of the ground elevation, the abnormal morphological parameters of the pipe network are obtained through spatial measurement and geometric calculation. S54. Spatially overlay and compare the abnormal morphological parameters with the basic GIS data of the water supply network and perform bidirectional calibration to generate an abnormal identification list of the water supply network.

[0038] Specifically, for areas where the measured parameters are more accurate than the GIS data and there are significant differences, such as discrepancies between the actual pipeline routes and GIS records, or the absence of recorded new facilities, the measured data is used to update the pipeline coordinates and attribute information in the GIS—this is forward calibration. For areas where the measured data has low confidence due to occlusion or image quality issues, such as partial vegetation cover causing missing point clouds, the complete pipeline topology in the GIS is used to logically repair the missing parts in the 3D model—this is reverse calibration. Through bidirectional calibration, the final generated list of water network anomalies includes both the anomaly information discovered in this survey and the latest calibrated network status data, achieving synchronous updates of the survey results and basic data.

[0039] In this optional embodiment, based on the enhanced image dataset, the three-dimensional dense reconstruction of the abnormal region is performed using spatial calculation and dense matching algorithms of multi-view images to generate a three-dimensional model of the abnormal region, including the following steps: S521. Extract feature points from each image in the enhanced image dataset, perform cross-image feature point matching, and construct corresponding relationships of corresponding points. S522. Based on the correspondence of corresponding points, the image pose parameters of each image are obtained through spatial geometric calculation of multi-view images, and a sparse three-dimensional point cloud is generated. S523. Based on the sparse 3D point cloud and image pose parameters, perform pixel-by-pixel dense matching on the enhanced image to generate a dense 3D point cloud. S524. Perform meshing and texture mapping on the dense 3D point cloud to construct a 3D model of the abnormal region.

[0040] In this optional embodiment, in the three-dimensional model of the abnormal area, based on the spatial curvature of the pipe gallery edge, the geometric shape of the manhole opening, and the abnormal distribution of surface elevation, the abnormal morphological parameters of the pipe network are obtained through spatial measurement and geometric calculation, including the following steps: S531. The three-dimensional model of the abnormal area is segmented to obtain the pipe gallery structure, manhole opening and abnormal surface elevation block; Specifically, the structure of the utility tunnel reflects the pipeline itself; the manholes reflect the surface facilities; and the abnormal ground elevation areas reflect surface deformation.

[0041] S532. Extract the point cloud of the pipe gallery edge along the pipe gallery structure, and use curvature calculation and trajectory fitting to obtain the pipeline offset and bending parameters; S533. Perform geometric shape fitting on the inspection wellhead, identify the abnormal state of the wellhead, and extract the wellhead coordinates and relative elevation parameters. S534. Measure the elevation difference and delineate the range of the abnormal surface elevation blocks, and invert the surface subsidence parameters and surface leakage parameters by combining the pipeline burial depth data. Specifically, elevation differences are measured in areas of surface elevation anomalies to extract the elevation values ​​of the highest and lowest points within the area, calculate the surface subsidence depth or uplift height, and delineate the boundaries of the anomaly range. Based on this, and combined with the pipe gallery burial depth data obtained from S532 (i.e., the difference between the pipe gallery center point elevation and the surface elevation), the correlation between surface subsidence and underground pipe networks is inverted: if the subsidence area is directly above the pipe gallery and its depth exceeds a threshold, it is determined to be surface subsidence caused by pipe network leakage; if the subsidence area is accompanied by spectral anomalies such as vegetation death, leakage parameters are further correlated.

[0042] S535, by integrating pipeline offset, bend parameters, wellhead coordinates, wellhead relative elevation parameters, surface subsidence parameters, and surface leakage parameters, abnormal morphology parameters are obtained.

[0043] According to one embodiment of the present invention, such as Figure 2 As shown, a water pipe network identification system based on drone-based anomaly detection and re-photography is also provided. This system includes: The inspection route planning module 1 is used to divide the detection area into districts based on the basic data of water pipe network GIS, and match differentiated inspection routes according to the district division results. The initial full-coverage image dataset construction module 2 is used by UAVs to perform inspection tasks according to differentiated inspection routes and to construct an initial full-coverage image dataset by acquiring visible light images, multispectral images and real-time GPS coordinates of the detection area. The abnormal area judgment module 3 is used to spatially correlate and compare the pipeline network features in the initial full coverage image dataset with the GIS data, and combine the image clarity and feature integrity to identify abnormal areas and generate an abnormal reshoot list. The abnormal area reshooting module 4 is used by the drone to reshoot abnormal areas according to the reshooting list, so as to obtain reshoot images of the abnormal areas. The pipeline anomaly identification module 5 is used to align and fuse the initial full-coverage image dataset with the re-shot images. By performing three-dimensional calculation and reconstruction on the abnormal areas, it inverts the burial depth, direction, pipe diameter and abnormal morphological parameters of the pipeline network, and performs bidirectional calibration with the water pipeline network GIS basic data to generate a water pipeline network anomaly identification list.

[0044] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores static and dynamic information data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements the steps in the above method embodiments.

[0045] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device to which the present invention is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0046] In addition, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0047] In addition, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0048] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0049] This invention is not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this invention is limited only by the appended claims.

Claims

1. A water service pipe network identification method based on abnormality detection and re-shooting by a UAV, characterized in that, The method includes: S1. Based on the basic GIS data of the water supply network, the detection area is divided into districts, and differentiated inspection routes are matched according to the district division results. S2. The UAV performs inspection tasks according to the differentiated inspection route and constructs an initial full-coverage image dataset by acquiring visible light images, multispectral images and real-time GPS coordinates of the detection area. S3. Spatially correlate and compare the pipeline network features in the initial full-coverage image dataset with the GIS data, and combine the image clarity and feature integrity to identify abnormal areas in order to generate an abnormal reshoot list. S4. The drone takes a second shot of the abnormal area according to the reshoot list to obtain the reshoot image of the abnormal area; S5. Align and fuse the initial full-coverage image dataset with the re-shot images. Reconstruct the abnormal areas using 3D calculations to invert the burial depth, direction, pipe diameter, and abnormal morphological parameters of the pipeline network. Perform bidirectional calibration with the water pipeline network GIS basic data to generate a list of water pipeline network anomalies.

2. The water service pipe network identification method based on abnormality detection and re-shooting of a UAV according to claim 1, characterized in that, The process of dividing the detection area into districts based on the GIS data of the water supply network and matching differentiated inspection routes according to the district division results includes the following steps: S11. Obtain basic GIS data of water pipe network, extract information on pipe network type, burial depth level and surface marker distribution, as the basis for district-level division; S12. Based on the district-level division criteria, the detection area is divided into the underground pipeline coverage area, the surface facility area, and the key monitoring area; S13. For the detection needs of each district level, set the corresponding cruise parameters and shooting parameters respectively, and generate differentiated inspection routes that match each district level. 3.The water service pipe network identification method based on abnormality detection and rephotographing of a UAV according to claim 1, characterized in that, The process of spatially correlating and comparing the pipeline network features in the initial full-coverage image dataset with GIS data, and combining image clarity and feature integrity to identify abnormal areas and generate an abnormal re-enhancing list includes the following steps: S31. Extract pipeline network features from the initial full-coverage image dataset, and combine the water pipeline network GIS data to perform spatial location association and morphological matching comparison of the pipeline network features to obtain suspected abnormal points. S32. Analyze the sharpness of each frame in the initial full-coverage image dataset, and filter out invalid image frames that are blurry, occluded, or incomplete to obtain the image defect areas. S33. Integrate suspected abnormal locations and image defect areas, identify abnormal areas, and extract the coordinates, coverage area, and shooting parameters of abnormal areas to generate an abnormal reshoot list.

4. The water service pipe network identification method based on abnormality detection and re-shooting of a UAV according to claim 3, characterized in that, The process of extracting pipeline network features from the initial full-coverage image dataset and combining them with water supply network GIS data to perform spatial location correlation and morphological matching comparison of the pipeline network features to obtain suspected anomaly locations includes the following steps: S311. Extract linear utility tunnel features, point facility features, and surface spectral anomaly features from the initial full-coverage image dataset as the utility network features to be compared. S312. Retrieve the theoretical pipeline vector data, facility coordinates, and attribute information of the corresponding area from the water supply network GIS data as reference data; S313. Unify the pipeline network features to be compared with the reference data into the same spatial coordinate system, perform spatial location overlay analysis and morphological matching comparison, and mark the points with missing features, morphological changes, location offsets and surface anomalies without corresponding GIS markers to form suspected anomaly points.

5. The water service pipe network identification method based on abnormality detection and re-shooting of a UAV according to claim 1, characterized in that, The process of aligning and fusing the initial full-coverage image dataset with the re-captured images, reconstructing abnormal areas in three dimensions, retrieving abnormal morphological parameters of the pipeline network, and performing bidirectional calibration with the water pipeline network GIS basic data to generate a list of water pipeline network anomalies includes the following steps: S51. Preprocess the initial full-coverage image dataset and the re-shot images, and construct an enhanced image dataset of abnormal areas through spatiotemporal alignment. S52. Based on the enhanced image dataset, use the spatial solution and dense matching algorithm of multi-view images to perform three-dimensional dense reconstruction of the abnormal area and generate a three-dimensional model of the abnormal area. S53. In the three-dimensional model of the abnormal area, based on the spatial curvature of the pipe gallery edge, the geometric shape of the manhole opening and the abnormal distribution of the ground elevation, the abnormal morphological parameters of the pipe network are obtained through spatial measurement and geometric calculation. S54. Spatially overlay and compare the abnormal morphological parameters with the basic GIS data of the water supply network and perform bidirectional calibration to generate an abnormal identification list of the water supply network.

6. The water pipe network identification method based on UAV detection and anomaly re-photographing as described in claim 5, characterized in that, The process of using enhanced image datasets and spatial computation and dense matching algorithms for multi-view images to perform 3D dense reconstruction of abnormal regions and generate 3D models of abnormal regions includes the following steps: S521. Extract feature points from each image in the enhanced image dataset, perform cross-image feature point matching, and construct corresponding relationships of corresponding points. S522. Based on the correspondence of corresponding points, the image pose parameters of each image are obtained through spatial geometric calculation of multi-view images, and a sparse three-dimensional point cloud is generated. S523. Based on the sparse 3D point cloud and image pose parameters, perform pixel-by-pixel dense matching on the enhanced image to generate a dense 3D point cloud. S524. Perform meshing and texture mapping on the dense 3D point cloud to construct a 3D model of the abnormal region.

7. The water pipe network identification method based on UAV detection and anomaly re-photographing as described in claim 6, characterized in that, In the three-dimensional model of the abnormal area, the abnormal morphology parameters of the pipeline network are obtained through spatial measurement and geometric calculation based on the spatial curvature of the pipeline corridor edge, the geometric shape of the manhole opening, and the abnormal distribution of surface elevation. The steps include: S531. The three-dimensional model of the abnormal area is segmented to obtain the pipe gallery structure, manhole opening and abnormal surface elevation block. S532. Extract the point cloud of the pipe gallery edge along the pipe gallery structure, and use curvature calculation and trajectory fitting to obtain the pipeline offset and bending parameters; S533. Perform geometric shape fitting on the inspection wellhead, identify the abnormal state of the wellhead, and extract the wellhead coordinates and relative elevation parameters. S534. Measure the elevation difference and delineate the range of the abnormal surface elevation blocks, and invert the surface subsidence parameters and surface leakage parameters by combining the pipeline burial depth data. S535, by integrating pipeline offset, bend parameters, wellhead coordinates, wellhead relative elevation parameters, surface subsidence parameters, and surface leakage parameters, abnormal morphology parameters are obtained.

8. A water pipe network identification system based on UAV detection and re-photographing, used to implement the water pipe network identification method based on UAV detection and re-photographing as described in any one of claims 1-7, characterized in that, The system includes: The inspection route planning module is used to divide the detection area into districts based on the basic GIS data of the water pipe network, and match differentiated inspection routes according to the district division results. The initial full-coverage image dataset construction module is used by UAVs to perform inspection tasks according to differentiated inspection routes and to construct an initial full-coverage image dataset by acquiring visible light images, multispectral images and real-time GPS coordinates of the detection area. The abnormal area identification module is used to spatially correlate and compare the pipeline network features in the initial full coverage image dataset with GIS data, and combine image clarity and feature integrity to identify abnormal areas and generate an abnormal reshoot list. The abnormal area reshooting module is used by the drone to reshoot abnormal areas according to the reshooting list, so as to obtain reshoot images of the abnormal areas; The pipeline anomaly identification module is used to align and fuse the initial full-coverage image dataset with the re-captured images. By performing three-dimensional calculation and reconstruction on the abnormal areas, it inverts the burial depth, direction, pipe diameter and abnormal morphological parameters of the pipeline network, and performs bidirectional calibration with the water pipeline network GIS basic data to generate a list of water pipeline network anomalies.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images