Method for collaborative identification of external riser multi-source data based on spatial topological constraints

Abstract

This invention relates to the field of image processing technology and discloses a collaborative identification method for multi-source data of exterior pipes based on spatial topological constraints. The method includes: establishing a unified coordinate system on the building facade and constructing a spatial topological constraint library for exterior pipes; acquiring building facade image data and building facade point cloud data for spatial registration; extracting linear targets and tubular point cloud clusters to generate theoretical pipe segment trajectories for exterior pipes, and classifying the linear targets, tubular point cloud clusters, and theoretical pipe segments into candidate pipe segments; connecting the candidate pipe segments from bottom to top to generate candidate paths for exterior pipes, and performing topological consistency evaluation on each candidate path; determining the target exterior pipe path set from the remaining candidate paths, determining the exterior pipe type for each exterior pipe, and writing the final exterior pipe identification result into an exterior pipe asset library. This application achieves complete identification and type labeling of exterior pipes, improving the accuracy of exterior pipe identification.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and more specifically, to a method for collaborative identification of multi-source data of external risers based on spatial topological constraints. Background Technology

[0002] Building facades are widely covered with water supply risers, drainage risers, and gas risers. With the increasing stock and height of urban buildings, maintenance departments need to rely on digital means to locate, identify, and manage risers in a unified coordinate system without scaffolding or point-by-point inspection. This is achieved using building facade image data, building facade point cloud data, building information model data, and riser maintenance records. Current engineering practices have developed solutions that utilize UAV oblique imagery, 3D laser scanning point clouds, and panoramic imagery for multi-source data fusion and 3D modeling. Image processing and 3D reconstruction techniques are then used to obtain the geometric shape and surface condition of the building facade.

[0003] However, existing technologies mostly focus on the 3D reconstruction or defect identification of the overall building facade, lacking a dedicated modeling and identification mechanism for linear targets such as external risers. For example, patent document CN119273853B proposes a refined 3D modeling method and system based on multi-source data fusion from air and ground. By fusing UAV oblique photography, scanned point cloud data, and panoramic video data, and performing coarse and fine registration and surface reconstruction on the multi-source data, it achieves refined 3D modeling of complex building environments. However, such solutions typically reconstruct external risers along with other facade components as part of the building surface. They lack a spatial topological constraint library for external risers under a unified coordinate system, which includes vertical continuity, support spacing, floor alignment, and combination arrangement. They also fail to jointly match building facade image data, building facade point cloud data, and external riser maintenance ledger data to generate candidate external riser segments and candidate external riser paths. This makes it difficult to recover the complete path and type information of each external riser in the presence of partial occlusion, data loss, or ledger discrepancies. Furthermore, they cannot output external riser asset library records for external riser inspection and maintenance.

[0004] Therefore, it is necessary to design a collaborative identification method for multi-source data of external risers based on spatial topological constraints to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a multi-source data collaborative identification method for external risers based on spatial topological constraints, which aims to solve the problem of how to accurately reconstruct and identify the path and type of each external riser from multi-source image data and point cloud data when there are obstructions on the building facade, data missing and operation and maintenance records deviations.

[0006] This invention proposes a collaborative identification method for multi-source data of external risers based on spatial topological constraints, comprising: A unified coordinate system is established on the building facade, and a spatial topology constraint library for external risers is constructed based on multi-source data of external risers. The external riser spatial topology constraint library includes external riser vertical continuity constraints, external riser support spacing constraints, external riser floor alignment constraints, and external riser combination arrangement constraints. Acquire building facade image data and building facade point cloud data corresponding to the building, and perform spatial registration in the unified coordinate system; Linear targets are extracted based on the building facade image data, tubular point cloud clusters are extracted based on the building facade point cloud data, and theoretical pipe segments are generated by combining the external pipe maintenance ledger data. Linear targets, tubular point cloud clusters and theoretical pipe segments with horizontal distances less than a preset distance threshold and height overlap greater than a preset overlap threshold are classified as external pipe candidate segments under the unified coordinate system. According to the building facade zoning and vertical direction, the candidate pipe segments of the external riser are connected from bottom to top to generate candidate external riser paths. Based on the external riser spatial topology constraint library, the external riser vertical continuity, external riser support spacing, external riser floor alignment, and external riser combination arrangement are used to evaluate the topology consistency of each candidate external riser path. Candidate external riser paths with topology consistency evaluation results lower than the preset evaluation threshold are eliminated. In the remaining candidate paths of external risers after elimination, the target external riser path set is determined according to the topology consistency evaluation results to form the final external riser identification result, and the external riser type of each external riser in the target external riser path set is determined according to the external riser operation and maintenance ledger data. The final external riser identification result is written into the external riser asset database.

[0007] Furthermore, when establishing a unified coordinate system on the building facade and constructing a spatial topology constraint library for the external risers based on multi-source data, the following steps are included: The multi-source data for external risers includes architectural design drawings, as-built drawings, building information model data, and external riser operation and maintenance ledger data. Using the pre-set reference point on the building facade as the origin, the direction of gravity as the vertical coordinate axis, and the main axis of the building structure as the horizontal coordinate axis, coordinate transformation is performed on the geometric position and floor elevation of the external riser in the multi-source data of the external riser. The geometric relationships and standard parameters of the external risers in the unified coordinate system are converted into external riser vertical continuity constraints, external riser support spacing constraints, external riser floor alignment constraints, and external riser combination layout constraints. Among them, the external riser vertical continuity constraints limit the horizontal offset of the same external riser in adjacent floors to no more than a first distance threshold and the axial direction deviation from the vertical direction to no more than a first angle threshold. The external riser support spacing constraints limit the vertical distance between adjacent supports of the same external riser to be between the minimum support spacing and the maximum support spacing set according to the external riser type. The external riser floor alignment constraints limit the height deviation of the external riser floor crossing position from the corresponding floor elevation to no more than a second distance threshold. The external riser combination layout constraints limit the horizontal distance between different external risers in the same building facade partition to be within a preset spacing range and the arrangement order of water supply external risers, drainage external risers, and gas external risers to conform to a preset combination pattern.

[0008] Furthermore, when acquiring building facade image data and building facade point cloud data corresponding to the building and performing spatial registration in a unified coordinate system, the process includes: Acquire multi-view image data of the building facade covering the building facade, acquire point cloud data of the building facade covering the building facade, set up at least three control points on the building facade and pre-record the spatial coordinates of each control point in the unified coordinate system, detect the image position of each control point in the building facade image data, extract the spatial position of each control point in the building facade point cloud data, determine the registration transformation to align the building facade image data and building facade point cloud data with the unified coordinate system based on the spatial coordinates of the control points in the unified coordinate system and the observation position in the building facade image data and building facade point cloud data, and refine the registration transformation by matching the building facade outline and door and window boundary lines to ensure that the positional deviation of the building facade image data and building facade point cloud data in the unified coordinate system is lower than a preset error threshold.

[0009] Furthermore, when extracting linear targets based on the building facade image data, the process includes: The building facade image data is subjected to grayscale processing and noise reduction processing, edge detection and connected region extraction are performed, and connected regions with a length greater than a first length threshold, a length-to-width ratio greater than a first ratio threshold, and an angle between the main extension direction and the vertical coordinate axis less than a first angle threshold are retained. The center line of each connected region is determined as the linear target.

[0010] Furthermore, when extracting tubular point cloud clusters based on the building facade point cloud data, the process includes: Ground point removal and background point removal are performed on the point cloud data of the building facade. The location point set close to the building facade is extracted. The main extension direction and cross-sectional point cloud distribution range are calculated for each location point set. The point set with the angle between the main extension direction and the vertical coordinate axis less than the second angle threshold and the cross-sectional point cloud distribution radius within the preset radius is retained, and the point set is determined as the tubular point cloud cluster.

[0011] Furthermore, when generating the external riser maintenance log data to create the external riser theoretical pipeline segment trajectory and dividing it into external riser theoretical pipeline segments under the aforementioned unified coordinate system, the process includes: Based on the starting position, ending position and corresponding floor of the external riser recorded in the external riser maintenance ledger data, the starting position and ending position are mapped to the unified coordinate system, a connection path is generated along the vertical coordinate axis and divided into multiple external riser theoretical pipe segments according to the floor height; Under the unified coordinate system, the linear targets, tubular point cloud clusters and external riser theoretical pipe segments are matched according to the horizontal distance threshold and the height interval overlap threshold. Targets that simultaneously meet the distance threshold and height overlap threshold in space with at least two data sources are classified as external riser candidate pipe segments, and targets supported by only a single data source and located between adjacent external riser candidate pipe segments are designated as supplementary external riser candidate pipe segments.

[0012] Furthermore, when generating candidate external riser paths by connecting the candidate external riser segments from bottom to top according to the building facade zoning and vertical direction, the process includes: Within each building facade section, candidate external riser segments are divided into several vertical pipeline groups according to their horizontal position. Within each vertical pipeline group, the candidate external riser segments are sorted from low to high according to their height. Adjacent candidate external riser segments with a height difference less than a third distance threshold and a horizontal offset less than a fourth distance threshold are connected to form the same candidate external riser path. When the height difference between adjacent candidate external riser segments is between the third and fifth distance thresholds, and there is a theoretical external riser segment or a supplementary candidate external riser segment within this height range, the theoretical external riser segment or the supplementary candidate external riser segment is inserted into the candidate external riser path to fill in the gaps and form a continuous candidate external riser path in the vertical direction.

[0013] Furthermore, when performing a topology consistency evaluation on each candidate path for an external riser based on the external riser spatial topology constraint library, the evaluation includes: For each candidate path of the external riser, the vertical continuity evaluation result of the external riser is calculated based on the horizontal offset and axial direction deviation of each candidate external riser segment under the unified coordinate system; the support spacing evaluation result of the external riser is calculated based on the vertical distance between the support positioning points; the floor alignment evaluation result of the external riser is calculated based on the height difference between the floor crossing position and the corresponding floor elevation; and the combination layout evaluation result of the external riser is calculated based on the horizontal distance and arrangement order between the candidate external riser path and adjacent candidate external riser paths within the same building facade partition. The vertical continuity evaluation result of the external riser, the support spacing evaluation result of the external riser, the floor alignment evaluation result of the external riser, and the combination layout evaluation result of the external riser are weighted and superimposed according to preset weights to obtain the topological consistency evaluation result of the candidate external riser path.

[0014] Furthermore, when determining the target set of external riser paths based on the topology consistency evaluation results from the remaining candidate external riser paths after elimination, the process includes: The topological consistency evaluation result of each candidate external pipe path is used as the path evaluation value. Within each building facade partition, the selection principle is based on maximizing the sum of the path evaluation values. Under the conditions that different candidate external pipe paths do not geometrically overlap in the unified coordinate system and meet the requirements for the number of external pipes and the constraints on the arrangement of external pipe combinations within the same building facade partition, the target set of external pipe paths is selected from the remaining candidate external pipe paths by a combination of stepwise addition and backtracking elimination, thus forming the final external pipe identification result.

[0015] Furthermore, when determining the type of each external riser in the target external riser path set based on the external riser maintenance ledger data and writing the final external riser identification result into the external riser asset database, this includes: The locations of water supply facilities, drainage facilities, and gas facilities, as well as their connection relationships with external risers, recorded in the external riser maintenance ledger data, are mapped to the unified coordinate system. For each external riser in the final external riser identification result, based on its vertical connection relationship with the water supply facilities, drainage facilities, and gas facilities, and the service floor range, the external riser is identified as a water supply external riser, drainage external riser, or gas external riser. In the external riser asset database, the external riser spatial path, external riser type, external riser support positioning point, external riser floor crossing location, and corresponding topology consistency evaluation result are recorded for each external riser.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: By establishing a unified coordinate system on the building facade and constructing a spatial topological constraint library for external pipes, the building facade image data, building facade point cloud data, and external pipe operation and maintenance ledger data are collaboratively processed under the same spatial reference. First, linear targets are extracted from the images, and tubular point cloud clusters are extracted from the point cloud. These are then combined with the ledger data to generate theoretical pipe segment trajectories for external pipes. These trajectories are then fused according to the horizontal distance and height interval overlap relationships to form candidate external pipe segments, which are then connected along the building facade partitions and vertical directions to form candidate external pipe paths. Spatial topological constraints are used to evaluate the topological consistency of each candidate external pipe path, eliminating those with evaluation results lower than [a certain threshold]. The system identifies candidate paths for external risers based on a preset evaluation threshold. Then, it determines the target set of external riser paths based on the topology consistency evaluation results and combines this with external riser maintenance ledger data to determine the external riser type for each path in the target set. Finally, it writes the external riser identification results into the external riser asset database. This allows the system to recover the continuous spatial path and accurate type of each external riser even in complex scenarios with partial occlusion, missing data, and ledger discrepancies, thanks to the synergistic effect of multi-source data and spatial topology constraints. This achieves highly reliable asset-based output for external riser inspection and maintenance, improving the completeness of external riser identification compared to existing technologies that rely solely on 3D reconstruction or local image recognition. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a multi-source data collaborative identification method for external risers based on spatial topological constraints provided in an embodiment of the present invention. Detailed Implementation

[0018] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] In the collaborative processing of building facade image data, point cloud data, and external riser maintenance ledger data, when there is partial occlusion, data loss, or ledger discrepancies, it is difficult to accurately identify the complete spatial path and type information of the external risers, resulting in the inability to generate an external riser asset database that can be used for inspection and maintenance. Specifically, existing technologies lack a dedicated modeling mechanism for linear external risers and have failed to establish a spatial topological constraint database under a unified coordinate system. This makes it impossible to recover the continuous path of the external risers under incomplete data conditions. Furthermore, spatial topological features such as vertical continuity of external risers, support spacing, floor alignment, and combined arrangement are ignored. Consequently, the generation process of candidate pipe segments and candidate paths lacks rule constraints, ultimately leading to an asset database that cannot meet maintenance requirements.

[0020] For example, during facade inspections of high-rise residential buildings, drone-generated tilted images may contain missing data in balcony areas due to obstruction by surrounding trees; 3D laser scanning point clouds may have data gaps at building corners; and the bracket positions recorded in the maintenance logs for external risers may deviate from their actual installation locations. In this scenario, existing methods cannot connect the pipe segments at breakpoints into a complete path, and type identification errors are frequent. Specifically, external risers in partially obstructed areas are segmented into multiple isolated pipe segments, and log discrepancies lead to confusion between gas and water supply external risers, thus affecting the completeness and accuracy of external riser asset information.

[0021] If the above problems are not resolved, the external riser identification results will contain a large number of fragmented pipe segments and incorrect paths, and the asset database records will not reflect the true status. As a result, the operation and maintenance department will not be able to formulate inspection plans based on accurate spatial path information, and facility safety hazards will not be detected in a timely manner, ultimately affecting the reliability of external riser asset management.

[0022] In response, this application proposes a collaborative identification method for multi-source data of external risers based on spatial topological constraints, including: S100: Establish a unified coordinate system on the building facade and construct a spatial topology constraint library for external risers based on multi-source data of external risers. The spatial topology constraint library for external risers includes vertical continuity constraints of external risers, spacing constraints of external riser supports, floor alignment constraints of external risers, and combination layout constraints of external risers. S200: Acquire the building facade image data and building facade point cloud data corresponding to the building, and perform spatial registration in a unified coordinate system; S300: Extract linear targets based on building facade image data, extract tubular point cloud clusters based on building facade point cloud data, and generate theoretical pipe segment trajectories based on external pipe maintenance ledger data. Under a unified coordinate system, linear targets, tubular point cloud clusters, and theoretical pipe segments with horizontal distances less than a preset distance threshold and height overlap greater than a preset overlap threshold are classified as candidate external pipe segments. S400: Connect candidate pipe segments of external risers from bottom to top according to the building facade zoning and vertical direction to generate candidate external riser paths. Based on the external riser spatial topology constraint library, evaluate the topology consistency of each candidate external riser path from the perspectives of external riser vertical continuity, external riser support spacing, external riser floor alignment, and external riser combination arrangement. Remove candidate external riser paths whose topology consistency evaluation results are lower than the preset evaluation threshold. S500: In the remaining candidate paths of external risers after elimination, the target external riser path set is determined according to the topology consistency evaluation results to form the final external riser identification result, and the external riser type of each external riser in the target external riser path set is determined according to the external riser operation and maintenance ledger data, and the final external riser identification result is written into the external riser asset database.

[0023] This application relates to a collaborative identification method for external risers based on spatial topological constraints from multiple sources. It aims to address the problem of accurately identifying the complete spatial path and type information of external risers and generating an external riser asset database when there are partial occlusions, data gaps, or discrepancies in external riser maintenance records in building facade image data and point cloud data. In practical applications, a unified coordinate system can be achieved using a local Cartesian coordinate system with the intersection of the building's ground floor and the ground floor as the origin, or by converting to a geographic coordinate system established through a global navigation satellite system. This primarily serves to provide a geometric reference framework for multi-source data registration. The external riser spatial topological constraint database refers to a parametric database storing external riser layout rules. This can be implemented using numerical settings based on building industry standard manuals, or by deriving typical constraint parameters through statistical analysis of historical external riser engineering data. Its main purpose is to achieve standardized constraints on the geometric relationships of external risers. Specifically, spatial registration can employ automatic correction methods based on feature point matching. For example, accelerating robust feature extraction algorithms can be used to extract corresponding key points between the image and point cloud for transformation matrix calculation, or affine transformation parameters can be calculated by manually selecting building contour feature points. This is primarily to ensure geometric consistency across different data sources within a unified reference framework. Furthermore, extracting linear targets from building facade image data can involve Hough transform detection of straight line segments combined with length filtering, such as retaining line segments with nearly vertical extension directions, or using deep learning segmentation models to identify tubular structure regions. Extracting tubular point cloud clusters from building facade point cloud data can involve using cylindrical fitting algorithms to calculate point cloud distribution characteristics, or identifying subsets of point clouds with tubular geometric characteristics through voxel meshing. Finally, dividing linear targets, tubular point cloud clusters, and theoretical pipe segments into candidate pipe segments for exterior risers can be achieved using spatial clustering analysis methods. For example, target association can be performed based on thresholds for Euclidean distance and height overlap, or a region growing algorithm can be used to merge elements that meet spatial proximity conditions into candidate units. As a preferred implementation method, topology consistency evaluation can employ a multi-index weighted scoring model. For example, constraint indicators such as vertical continuity and support spacing can be quantified into independent scoring items and then synthesized into a total score according to a preset ratio. Alternatively, a rule engine can be used to execute logical judgments to determine path compliance. Therefore, this application overcomes the problems of neglecting external riser features and incorrect path connections in existing technologies by constructing spatial topology constraint rules specific to external risers and coordinating image data, point cloud data, and maintenance ledger data under a unified coordinate system. This enables accurate identification and asset management of external riser spatial paths even under conditions of incomplete data.

[0024] Specifically, after establishing a unified coordinate system on the building facade, all data are incorporated into the same geometric reference frame. The origin of the coordinate system can be specifically set at a fixed reference point at the main entrance on the ground floor of the building to ensure spatial consistency in the processing. Furthermore, a spatial topology constraint library for external risers, constructed based on multi-source data, is generated. This constraint library includes constraints on the vertical continuity of external risers, the spacing of external riser supports, the alignment of external risers between floors, and the arrangement of external risers in combination. These constraints transform the external riser engineering specifications into a quantifiable set of rules. For example, the vertical continuity constraint limits the horizontal offset between adjacent floors to no more than a specific distance threshold. After acquiring the building facade image data and point cloud data, they are spatially registered in a unified coordinate system through control point matching, ensuring that positional deviations are controlled within millimeter-level accuracy. When extracting linear targets from image data, regions with a length greater than a preset threshold and a main extension direction close to vertical are retained after grayscale processing and edge detection. When extracting tubular point cloud clusters from point cloud data, background points are removed, and point sets that meet the vertical angle and radius range are selected by analyzing the main extension direction. The theoretical pipe segment trajectory of the external riser generated by combining the maintenance and operation data of the external riser is matched with linear targets and pipe point cloud clusters in a unified coordinate system. Targets with a horizontal distance of less than 0.25 meters and an overlap of more than 80% in height intervals are classified as candidate pipe segments of the external riser. This achieves complementary collaboration of multi-source data and addresses the blind spots in identification caused by occlusion or missing data from a single data source.

[0025] Based on the building facade zoning and vertical orientation, candidate external riser segments are connected from bottom to top to form candidate external riser paths. For example, in the facade zoning of a residential building, candidate segments are grouped according to horizontal position and then connected by height. According to the external riser spatial topology constraint library, the vertical continuity, support spacing, floor alignment, and combination layout characteristics of each candidate path are quantitatively evaluated. The evaluation result of external riser support spacing is obtained by calculating the conformity of the vertical distance between adjacent supports with the specification range; paths with evaluation results below a preset threshold are eliminated. Among the remaining paths, the target external riser path set is determined based on the topology consistency evaluation results, forming the final identification result. The external riser type is determined based on the facility connection relationships mapped from the external riser operation and maintenance ledger data. A complete record including spatial path, type, and evaluation results is written into the external riser asset library.

[0026] This method, through a multi-source data collaboration mechanism under a unified coordinate system, complements the sensitivity of building facade image data to surface linear structures with the precise descriptive ability of point cloud data to three-dimensional tubular structures. It can still generate reliable candidate pipe segments even when there is partial occlusion or missing data. The introduction of a spatial topology constraint library transforms the engineering specifications for external risers into executable evaluation rules, strictly filtering paths that do not meet the requirements of vertical continuity or floor alignment, thus avoiding the problem of external riser features being ignored in general reconstruction methods. The final output asset library records integrate spatial path and type information, providing a structured digital management foundation for the operation and maintenance department, and solving the problem of identifying complete external riser paths and managing assets under data deviation conditions.

[0027] In practical applications, some of the embodiments described above in this application propose to construct a spatial topology constraint library for external risers to achieve accurate identification of external risers. However, in the process of implementation, the geometric position and floor elevation of external risers cannot be unified due to coordinate differences among multi-source data, which leads to deviations in the definition of spatial topology constraints, thereby reducing the accuracy and reliability of external riser identification.

[0028] In response, this application further proposes establishing a unified coordinate system on the building facade and constructing a spatial topology constraint library for the external risers based on multi-source data, including: The multi-source data for external risers includes architectural design drawings, as-built drawings, building information model data, and external riser operation and maintenance ledger data. Using the pre-set reference point on the building facade as the origin, the direction of gravity as the vertical coordinate axis, and the main axis of the building structure as the horizontal coordinate axis, coordinate transformation is performed on the geometric position and floor elevation of the external riser in the multi-source data of the external riser. The geometric relationships and standard parameters of the external risers in the unified coordinate system are converted into external riser vertical continuity constraints, external riser support spacing constraints, external riser floor alignment constraints, and external riser combination layout constraints. Among them, the external riser vertical continuity constraints limit the horizontal offset of the same external riser in adjacent floors to no more than a first distance threshold and the axial direction deviation from the vertical direction to no more than a first angle threshold. The external riser support spacing constraints limit the vertical distance between adjacent supports of the same external riser to be between the minimum support spacing and the maximum support spacing set according to the external riser type. The external riser floor alignment constraints limit the height deviation of the external riser floor crossing position from the corresponding floor elevation to no more than a second distance threshold. The external riser combination layout constraints limit the horizontal distance between different external risers in the same building facade partition to be within a preset spacing range and the arrangement order of water supply external risers, drainage external risers, and gas external risers to conform to a preset combination pattern.

[0029] Among them, multi-source data for external risers refers to an authoritative data set covering the entire life cycle of a building. This data can include architectural design drawings, as-built drawings, building information model data, and external riser operation and maintenance records. Its purpose is to provide comprehensive and complementary information sources, avoiding the limitations of a single data source and laying a complete foundation for coordinate transformation and constraint definition. Using a pre-set benchmark point on the building facade as the coordinate origin refers to selecting a physical reference point with engineering stability, which can be achieved using key nodes of the building structure or permanent marker points. Its purpose is to establish a reproducible coordinate benchmark. Using the direction of gravity as the vertical coordinate axis refers to determining the vertical benchmark based on the direction of the Earth's gravity field. This can be determined using plumb line measurements or an inertial navigation system, aiming to ensure vertical consistency conforms to physical laws. Using the main axis of the building structure as the horizontal coordinate axis refers to the geometric symmetry axis based on the main building structure, which can be marked on structural drawings. The main axis of the external riser is determined by either the main axis of the main axis or the main axis of the on-site measurement. Its purpose is to provide a stable horizontal reference frame. The vertical continuity constraint of the external riser refers to the engineering rules that regulate the vertical continuity of the external riser. It can set the threshold for horizontal offset and angular deviation based on the geometric relationship between adjacent floors. Its purpose is to ensure the smoothness of the external riser path and its anti-obstruction ability. The external riser support spacing constraint refers to the standard requirements that limit the vertical spacing of the supports. It can dynamically adjust the spacing range according to the type of external riser. Its purpose is to adapt to the mechanical properties of different pipe materials. The external riser floor alignment constraint refers to the control of the matching accuracy between the external riser and the floor elevation. It can set the allowable range of height deviation. Its purpose is to improve the spatial consistency between the external riser and the building structure. The external riser combination arrangement constraint refers to the rules that regulate the relative positions of multiple external risers. It can limit the horizontal spacing and arrangement order. Its purpose is to meet safety regulations and improve maintenance efficiency.

[0030] Specifically, the proposed solution establishes a unified coordinate system using a pre-defined reference point on the building facade as the origin, the direction of gravity as the vertical coordinate axis, and the main axis of the building structure as the horizontal coordinate axis. This maps the geometric positions and floor elevations of the external risers from multiple sources to the same physical reference frame, eliminating coordinate differences between different data sources. Based on this, the geometric relationships and engineering specification parameters of the external risers under the unified coordinate system are quantified into vertical continuity constraints, support spacing constraints, floor alignment constraints, and combined arrangement constraints, ensuring that the definition of spatial topological constraints is strictly based on measurable physical quantities. Vertical continuity constraints ensure the vertical continuity of the external riser path by limiting the horizontal offset and axial angle deviation between adjacent floors. Support spacing constraints dynamically set the spacing range according to the type of external riser to adapt to the mechanical requirements of different pipe materials. Floor alignment constraints ensure the matching accuracy between the external riser and the building elevation by controlling the height deviation of the floor crossing positions. Combined arrangement constraints regulate the spatial layout of multiple external risers by limiting the horizontal spacing and arrangement order. These constraints together constitute the topological constraint library for external risers, providing accurate quantitative basis for the topological consistency evaluation of candidate paths for external risers, thereby enabling collaborative identification of multi-source data under a unified coordinate system.

[0031] As a specific implementation method, this application selects the main entrance corner as a preset reference point when establishing a unified coordinate system on the building facade; the vertical coordinate axis is determined by on-site plumb line measurement to determine the gravity direction; the horizontal coordinate axis is based on the main axis direction in the building structural drawings. In the process of constructing the external riser spatial topology constraint library, the external riser positions in the building design drawings, the floor elevations in the as-built drawings, the geometric information of the building information model data, and the pipe segment records in the external riser maintenance log are mapped to this unified coordinate system through coordinate transformation. For example, the external riser vertical continuity constraint is defined as the horizontal offset between adjacent floors not exceeding a typical distance threshold and the axis deviation from the vertical not exceeding a typical angle threshold, ensuring that the external riser path presents a continuous vertical state; in the external riser combination layout constraint, the arrangement order of water supply external risers, drainage external risers, and gas external risers is set according to safety specifications to be arranged sequentially from left to right, with the horizontal spacing maintained within a typical range. This implementation method utilizes a unified coordinate system to integrate multi-source data, giving the construction process of the external riser spatial topology constraint library clear physical meaning and engineering operability.

[0032] Through the above solution, this application solves the problem of inconsistent geometric position and floor elevation caused by coordinate differences in multi-source data of external risers, making the definition of spatial topology constraints more accurate and reliable, thereby improving the accuracy of external riser identification and engineering applicability.

[0033] In practical applications, some of the embodiments described above in this application propose spatial registration to align building facade image data and point cloud data with a unified coordinate system. However, in its implementation, due to the lack of a precise registration mechanism, the image data and point cloud data have a large positional deviation in the unified coordinate system. This makes it difficult to accurately align multi-source data in cases of partial occlusion, missing data, or complex building structures, thereby affecting the reliable division of candidate pipe segments for external risers and the accuracy of topology consistency evaluation.

[0034] In response, this application further proposes the following steps: acquiring multi-view building facade image data covering the building facade, acquiring building facade point cloud data covering the building facade, setting up at least three control points on the building facade and pre-recording the spatial coordinates of each control point in a unified coordinate system, detecting the image position of each control point in the building facade image data, extracting the spatial position of each control point in the building facade point cloud data, determining the registration transformation to align the building facade image data and building facade point cloud data to the unified coordinate system based on the spatial coordinates of the control points in the unified coordinate system and their observation positions in the building facade image data and building facade point cloud data, and refining the registration transformation by matching the building facade outline and door and window boundary lines, so that the positional deviation of the building facade image data and building facade point cloud data in the unified coordinate system is lower than a preset error threshold.

[0035] Among them, control points refer to reference marks with known spatial coordinates pre-laid on the building facade. They can be implemented using reflective stickers, QR codes, or physical convex point structures. The purpose is to provide a stable coordinate benchmark to establish an accurate correspondence between image data and point cloud data. Registration transformation refers to a mathematical transformation model that maps different data sources to a unified coordinate system. It can be implemented using rigid body transformation or affine transformation. The purpose is to eliminate coordinate differences and ensure spatial consistency. Outline and door / window boundary lines refer to the geometric feature lines of the building facade. They can be implemented using edge detection algorithms or feature point matching technology. The purpose is to use the inherent features of the building as additional constraint points to optimize registration accuracy and compensate for the limitations of control point distribution.

[0036] Specifically, the solution of this application first acquires multi-view building facade image data to cover complete visual information and avoid the influence of local occlusion under a single viewpoint. At the same time, it acquires building facade point cloud data to provide accurate three-dimensional geometric structure, forming data complementarity. Then, control points are set up on the building facade and their spatial coordinates in a unified coordinate system are recorded to establish a stable mapping relationship between coordinate systems. Next, the image position of the control points is detected in the image data to associate visual observation with spatial position. The spatial position of the control points is extracted in the point cloud data to achieve direct connection between the point cloud and the unified coordinate system. Then, the registration transformation is calculated based on the known coordinates of the control points and the dual observation positions to eliminate coordinate transformation errors. Finally, by matching the building facade outline and door and window boundary lines, the registration transformation is refined by utilizing the geometric consistency of the building's inherent features to ensure that the positional deviation is below a preset error threshold, thereby providing a high-precision spatial data foundation for the division of candidate pipe segments for external risers.

[0037] As a specific implementation method, control points can be placed at window sills and corners on the building facade using high-contrast reflective markers; building facade image data is obtained by taking multi-angle photos of the building via a drone equipped with a high-resolution camera flying around the building; building facade point cloud data is collected by a ground-based laser scanner; in the image data, the SIFT feature detection algorithm is used to identify the image positions of the control points; in the point cloud data, the spatial positions of the control points are extracted through point cloud segmentation; the registration transformation calculation uses the least squares method to fit the coordinate correspondence of the control points; the matching of the contour lines and door / window boundary lines is optimized by detecting the corresponding points of the image edges and the point cloud contours using Hough transform.

[0038] The above scheme reduces the positional deviation between building facade image data and point cloud data in a unified coordinate system, improves the spatial consistency of multi-source data, thereby ensuring the reliability of candidate pipe segment division for external risers and providing an accurate data basis for topology consistency evaluation.

[0039] Specifically, in some of the embodiments described above in this application, a method is proposed to extract linear targets from building facade image data for generating candidate pipe segments for external risers. However, in the implementation process, non-tubular interference elements such as window frames and decorative lines, as well as image noise, are commonly found in building facade images. This results in the directly extracted linear targets containing a large number of misidentification results that do not conform to the characteristics of external risers, making it impossible to distinguish between real external risers and background interference, thereby affecting the accuracy of external riser candidate paths.

[0040] In this regard, this application further proposes a step for extracting linear targets based on the building facade image data, including: The building facade image data is subjected to grayscale processing and noise reduction processing, edge detection and connected region extraction are performed, and connected regions with a length greater than a first length threshold, a length-to-width ratio greater than a first ratio threshold, and an angle between the main extension direction and the vertical coordinate axis less than a first angle threshold are retained. The center line of each connected region is determined as the linear target.

[0041] Among these, grayscale processing refers to the process of converting a color image into a single-channel grayscale image, which can be achieved using RGB weighted averaging or maximum value methods, aiming to simplify image data dimensions and highlight edge information; denoising can be understood as applying filtering algorithms to suppress random noise interference, specifically using median filtering or Gaussian filtering, aiming to reduce false edges caused by noise; edge detection refers to the technique of identifying regions with abrupt changes in brightness in an image, specifically using Canny edge detectors or Sobel operators, aiming to locate potential linear structures; connected component extraction refers to the method of clustering adjacent edge pixels into complete regions, specifically achieved through region growing or connected component analysis, aiming to shape... The candidate set of linear targets; length threshold filtering refers to retaining connected regions whose length exceeds a preset threshold, with the aim of eliminating discontinuous interference such as short cracks or local decorative lines; ratio threshold filtering refers to retaining connected regions whose length-to-width ratio is greater than a preset threshold, with the aim of filtering wide wall areas or large component outlines to ensure that the target has linear geometric attributes; angle threshold filtering refers to retaining connected regions whose main extension direction is close to vertical, with the aim of eliminating interference elements with inconsistent directions such as horizontal window sill lines or inclined decorative lines; centerline extraction refers to the operation of simplifying the region data into a single-pixel wide geometric representation, which can be implemented through skeletonization algorithms, with the aim of providing a standardized linear model for spatial matching.

[0042] Specifically, the proposed solution first reduces image data complexity and suppresses noise interference through grayscale processing and denoising, providing a high-quality input foundation for edge detection. Based on this, edge detection and connected region extraction collaboratively locate potential line structures and form a region set, establishing a data framework that distinguishes the riser from other linear elements. Subsequently, a multi-level filtering mechanism using length thresholds, ratio thresholds, and angle thresholds is applied sequentially to the connected regions, precisely filtering based on the riser's continuity, slender characteristics, and vertical installation specifications, eliminating interfering elements that do not conform to geometric features. Finally, centerline extraction transforms the filtered regions into a precise linear geometric representation, preserving the core path information of the riser while avoiding the influence of boundary fluctuations. Each step of this process is executed sequentially and interconnected, jointly constructing a refined filtering mechanism oriented towards the riser's features, ensuring that the extraction results strictly conform to the physical characteristics of the riser.

[0043] As a specific implementation method, the solution of this application is implemented as follows: In the processing of building facade image data, firstly, grayscale conversion is applied to convert RGB images into grayscale images, and median filtering is used for noise suppression; then, Canny edge detector is used to extract edge features, and edge pixels are clustered into region sets through connected component analysis; next, connected regions with a length greater than the background interference elements, a slenderness that conforms to the tubular characteristics, and a direction that is close to vertical are selected; finally, skeletonization algorithm is applied to the selected regions to extract the center line, which is used as a linear target for generating candidate pipe segments for external risers.

[0044] Through the above-mentioned scheme, this application suppresses the influence of non-tubular interference elements such as window frames and decorative lines, as well as image noise in building facade images, and improves the accuracy of linear target extraction, thereby providing reliable input for the generation of candidate paths for external risers and enhancing the overall reliability of external riser recognition results.

[0045] Specifically, in some of the embodiments described above in this application, a method is proposed to extract tubular point cloud clusters based on building facade point cloud data to generate candidate segments for external risers. However, in the implementation process, the original point cloud data contains non-building structural interference such as ground points and background points, and the external riser, as a vertical tubular target, has specific geometric features. If the point cloud data is processed directly, walls, decorative components, or noise are easily misidentified as tubular structures, causing the extracted tubular point cloud clusters to deviate from the actual position of the external riser or contain invalid data, thereby affecting the accuracy and reliability of the external riser candidate segment division.

[0046] In response, this application further proposes methods for extracting tubular point cloud clusters from building facade point cloud data, including: Ground point removal and background point removal are performed on the point cloud data of the building facade. The location point set close to the building facade is extracted. For each location point set, the main extension direction and cross-sectional point cloud distribution range are calculated. The point set with the angle between the main extension direction and the vertical coordinate axis less than the second angle threshold and the cross-sectional point cloud distribution radius within the preset radius is retained and the point set is determined as a tubular point cloud cluster.

[0047] Ground point removal and background point removal refer to removing non-building structure points related to the ground, vegetation, or pedestrians from the point cloud data. This can be achieved using segmentation algorithms based on elevation thresholds or deep learning classification models, with the aim of eliminating external interference and ensuring that the data focuses on the main building structure. Extracting the point set near the building facade can be understood as filtering the point cloud region attached to the building surface. Specifically, this can be done based on a distance threshold between the point cloud and the building facade outline, with the aim of reducing background object interference and providing a clean data foundation for tubular structure extraction. Calculating the main extension direction refers to determining the main extension direction of the point set, which can be achieved using principal component analysis or direction vector estimation, with the aim of quantifying the geometric characteristics of the point set to match the vertical characteristics of the external pipe. The cross-sectional point cloud distribution range specifically measures the distribution size of the point set on a cross section perpendicular to the main extension direction. For example, calculations can be performed based on point cloud density distribution or minimum circumcircle fitting to verify whether the point set conforms to the cross-sectional characteristics of a tubular structure. Retaining point sets where the angle between the main extension direction and the vertical coordinate axis is less than the second angle threshold refers to filtering structures close to the vertical direction. A reasonable angle tolerance range can be set to highlight the vertical continuity of the external riser and filter out tilted non-tubular structures. Filtering where the cross-sectional point cloud distribution radius is within a preset radius range can be understood as limiting the point set cross-sectional size to within the standard external riser diameter range. Specifically, the radius range can be set based on industry standards to exclude excessively thick or thin non-target structures and ensure that the extraction results conform to the actual external riser dimensions. Determining the filtered point set as a tubular point cloud cluster means outputting high-confidence tubular structure data, aiming to provide reliable input for matching with linear targets in the image and theoretical pipe segments.

[0048] Specifically, the proposed solution establishes a point cloud processing workflow tailored to the characteristics of external risers by sequentially performing ground and background point removal, extracting location point sets, calculating geometric features, and applying angle and radius filtering. First, by removing ground and background points, non-building-related interference is isolated, ensuring the analysis focuses on the building structure. Second, location point sets near the building facade are extracted, and based on the physical characteristics of the risers attached to the building surface, a subset of point clouds closely related to the target area is selected. Next, the main extension direction and cross-sectional distribution are calculated for each location point set, quantifying its geometric attributes to reflect the inherent characteristics of the riser's vertical extension and circular cross-section. Finally, a dual filtering mechanism using angle thresholds and radius ranges accurately identifies structures conforming to the geometric characteristics of the risers. This phased filtering mechanism ensures seamless integration of each step, progressively improving data purity and feature matching, thereby reliably extracting tubular point cloud clusters in complex point cloud environments.

[0049] As a preferred embodiment, the solution of this application is implemented as follows: In the point cloud processing stage, a robust segmentation algorithm based on elevation threshold is used to remove ground points, and a spatial index structure is used to quickly extract point sets within a certain range from the building facade outline; for each location point set, the principal extension direction vector is calculated through principal component analysis, and the point cloud distribution is fitted on a plane perpendicular to this direction to determine the cross-sectional size; reasonable angle tolerance and radius range thresholds are set to select point sets that conform to the geometric characteristics of the external riser as tubular point cloud clusters.

[0050] By employing the above-mentioned method, this application eliminates non-building structure interference in point cloud data, accurately identifies tubular structures with vertical continuity and standard pipe diameter characteristics, improves the accuracy and reliability of tubular point cloud cluster extraction, and thus provides a high-quality data foundation for the division of candidate pipe segments for external risers.

[0051] In practical applications, some of the embodiments described above in this application propose to generate the trajectory of the external riser pipe section by combining the maintenance ledger data of the external riser to assist in the identification of the external riser. However, in the process of its implementation, due to the possible positional deviation of the ledger data and the lack of clear quantitative rules when matching multiple data sources in space, it is impossible to accurately divide the candidate pipe sections of the external riser in the case of partial occlusion or missing data. Especially when a single data source has noise or incomplete coverage, it is easy to cause the candidate pipe sections to be missed or misjudged, thereby affecting the continuity of the external riser path and the accuracy of type identification.

[0052] In response, this application further proposes that when generating the theoretical pipeline trajectory of the external pipeline by combining the external pipeline maintenance ledger data and dividing it into external pipeline theoretical pipeline segments under the unified coordinate system, the following steps are included: Based on the starting position, ending position and corresponding floor of the external riser recorded in the external riser maintenance ledger data, the starting position and ending position are mapped to the unified coordinate system, a connection path is generated along the vertical coordinate axis and divided into multiple external riser theoretical pipe segments according to the floor height; Under the unified coordinate system, the linear targets, tubular point cloud clusters and external riser theoretical pipe segments are matched according to the horizontal distance threshold and the height interval overlap threshold. Targets that simultaneously meet the distance threshold and height overlap threshold in space with at least two data sources are classified as external riser candidate pipe segments, and targets supported by only a single data source and located between adjacent external riser candidate pipe segments are designated as supplementary external riser candidate pipe segments.

[0053] Specifically, the external riser maintenance ledger data refers to an engineering database that records the physical location and attribute information of external risers. This database can utilize a Building Information Modeling (BIM) database or spreadsheet files to store the starting and ending points of the external risers and their corresponding floor information. Its purpose is to provide fundamental data support for the theoretical trajectory of the external risers. Mapping the starting and ending points to a unified coordinate system means transforming the ledger coordinates to the building facade spatial reference. This can be achieved using coordinate transformation matrices or spatial registration algorithms, aiming to eliminate coordinate differences between multi-source data. In practical applications, generating connection paths along the vertical coordinate axis and dividing the external riser into multiple theoretical pipe segments according to floor height refers to segmenting the trajectory based on building structural characteristics. This can be achieved by combining floor elevation data with vertical projection, aiming to ensure a natural fit between the theoretical pipe segments and the building floor elevations. Specifically, the horizontal distance threshold and height interval overlap threshold refer to the criteria for quantifying spatial consistency. The standard can be established by using a dynamic threshold range based on the type of external riser, with the aim of establishing objective matching rules. Matching linear targets, tubular point cloud clusters, and theoretical segments of external risers refers to establishing a correspondence between multi-source observation data and theoretical trajectories. This can be achieved using spatial indexing structures or geometric similarity calculations, with the aim of identifying multi-source data segments representing the same external riser. In practical applications, classifying targets that simultaneously meet thresholds with at least two data sources as candidate external riser segments involves screening reliable segments through cross-validation mechanisms. This can be achieved using confidence-weighted or voting decision-making methods, with the aim of improving the confidence of the identification results. Specifically, using targets supported by only a single data source and located between adjacent candidate external riser segments as supplementary candidate external riser segments refers to a completion strategy for data gaps. This can be achieved using spatial continuity constraints or contextual reasoning, with the aim of maintaining the integrity of the external riser trajectory.

[0054] Specifically, the proposed solution generates segmented theoretical pipe segments by mapping ledger data to a unified coordinate system, providing a structured benchmark framework for multi-source data matching. Based on this, it uses horizontal distance thresholds and height interval overlap thresholds to quantitatively match linear targets, tubular point cloud clusters, and theoretical pipe segments. A multi-source data cross-validation mechanism is used to filter out high-confidence candidate pipe segments for external risers. Simultaneously, for areas with missing data, it intelligently completes broken paths based on the spatial relationships of adjacent candidate pipe segments, using targets supported by a single data source and located in gaps as supplementary candidate pipe segments. This synergistic effect of structured mapping and hierarchical matching strategies maintains the accuracy and completeness of external riser candidate segment division even under conditions of ledger location deviation, partial occlusion, or data loss, thus providing a reliable data foundation for external riser path generation.

[0055] As a preferred embodiment, the specific implementation of the solution in this application is as follows: In a facade recognition project of a commercial complex, firstly, the maintenance log data of the external riser is extracted from the building information model database to obtain the starting position of the drainage external riser located at the ground elevation of the first floor and the ending position located at the roof elevation of the top floor, and the elevation of each floor it passes through is recorded; these location data are mapped to a unified coordinate system through a coordinate transformation matrix, and a straight connection path from the starting point to the ending point is generated along the vertical coordinate axis, and the path is divided into theoretical pipe segments corresponding to each floor according to the building floor elevation data; subsequently, under the unified coordinate system, the vertical linear targets extracted by image edge detection and the tubular point cloud clusters extracted by point cloud main direction analysis are spatially matched with the above-mentioned theoretical pipe segments, and reasonable horizontal distance thresholds and height interval overlap thresholds are adopted to determine the pipe segments that simultaneously meet the spatial matching conditions with the image linear targets and the point cloud tubular clusters as candidate external riser pipe segments; for tubular point cloud clusters that are only supported by point cloud data on the middle floors and are located between adjacent candidate pipe segments, they are included in the path generation process as supplementary candidate external riser pipe segments.

[0056] Through the above-described scheme, this application can accurately classify candidate pipe segments for external risers even when there are location discrepancies, partial obstructions, or missing data in the ledger data. This avoids the omission or misjudgment of candidate pipe segments due to noise or incomplete coverage from a single data source, thereby improving the accuracy of external riser path continuity and type identification. Specifically, in some of the embodiments described above in this application, a method is proposed to connect candidate pipe segments of external risers from bottom to top according to the building facade zoning and vertical direction to generate candidate external riser paths. However, in the implementation process, due to partial occlusion of the building facade, missing point cloud data, or image recognition errors, there may be height gaps between adjacent candidate pipe segments of external risers, making it impossible to directly connect them to form a continuous path, thereby affecting the identification and asset management of complete external riser paths.

[0057] In this regard, this application further proposes that when generating candidate external riser paths by connecting the candidate external riser segments from bottom to top according to the building facade zoning and vertical direction, the following steps are included: Within each building facade section, candidate external riser segments are divided into several vertical pipeline groups according to their horizontal position. Within each vertical pipeline group, the candidate external riser segments are sorted from low to high according to their height. Adjacent candidate external riser segments with a height difference less than a third distance threshold and a horizontal offset less than a fourth distance threshold are connected to form the same candidate external riser path. When the height difference between adjacent candidate external riser segments is between the third and fifth distance thresholds, and there is a theoretical external riser segment or a supplementary candidate external riser segment within this height range, the theoretical external riser segment or the supplementary candidate external riser segment is inserted into the candidate external riser path to fill in the gaps and form a continuous candidate external riser path in the vertical direction.

[0058] Among them, the building facade zoning refers to dividing the building facade into independent management units according to structural characteristics or functional areas. This can be achieved by dividing according to the main axis of the building structure or by the distribution density of external risers. The purpose is to decompose the complex facade into independently manageable areas and avoid incorrect cross-regional connections. Vertical pipeline groups can be understood as sets of candidate external riser segments grouped within the building facade zoning based on horizontal proximity. This can be achieved using Euclidean distance clustering or grouping based on external riser type. The purpose is to group segments that may physically belong to the same external riser together. The third distance threshold is a threshold parameter used to determine whether the height difference between adjacent candidate external riser segments meets the direct connection condition. Different numerical ranges can be set according to the external riser type and floor height to ensure the physical continuity of connected segments in the vertical direction. The fourth distance threshold is used to determine whether the horizontal offset between adjacent candidate external riser segments meets the connection condition. The threshold parameter can be set according to the spacing of the external riser supports and the accuracy of the building structure. Its purpose is to ensure that the connected pipe segments will not be misconnected due to excessive horizontal offset. The fifth distance threshold is the upper limit threshold used to define the range within which height gaps can be filled. It can be set based on the maximum spacing of the external riser supports and the building floor height. Its purpose is to determine within what range of height gaps theoretical data can be used for intelligent filling. The external riser theoretical pipe segment specifically refers to the theoretical external riser path segment generated based on the external riser maintenance ledger data. It can be achieved by mapping the start and end positions of the ledger records to a unified coordinate system and generating a connection path along the vertical coordinate axis. Its purpose is to provide prior knowledge support for the external riser path. The supplementary external riser candidate pipe segment can be understood as the target pipe segment supported by only a single data source and located between adjacent external riser candidate pipe segments. It can be obtained by filtering the matching results of linear targets, tubular point cloud clusters and theoretical pipe segments. Its purpose is to fill the path gaps using limited data source information.

[0059] Specifically, the proposed solution first divides candidate external riser pipe segments into several vertical pipeline groups based on their horizontal position within the building facade partitions. This ensures that pipe segments with similar physical locations are grouped together, preventing incorrect association of pipe segments from different external risers. Within each vertical pipeline group, pipe segments are sorted from low to high height, utilizing the continuous upward trajectory of the external risers to provide a logical sequence for connections. By setting a dual-dimensional condition that the height difference is less than a third distance threshold and the horizontal offset is less than a fourth distance threshold, adjacent pipe segments meeting the conditions are connected as the same external riser candidate path, ensuring the physical rationality of the connection and avoiding path breaks or misconnections due to local obstruction or measurement errors. When the height difference between adjacent pipe segments is between the third and fifth distance thresholds and there are theoretical external riser pipe segments or supplementary external riser candidate segments, these segments are inserted into the external riser candidate path to fill in the gaps. This fully utilizes the prior knowledge of the external riser maintenance log to intelligently restore path continuity in areas with missing data. This mechanism, which combines hierarchical processing, dual-dimensional verification, and intelligent completion, ultimately forms a continuous candidate path for external risers in the vertical direction, providing reliable input for topology evaluation and asset library construction.

[0060] As a specific implementation method, when processing the facade of a high-rise building, this application first divides the building facade into three zones: east, south, and west. Within the south facade zone, candidate external riser pipe segments are divided into two groups of vertical pipelines based on their horizontal position: one group is located near the east side of the building, and the other near the west side. In the east vertical pipeline group, four candidate external riser pipe segments are obtained by sorting them from low to high height. The height difference between the first and second segments is less than the third distance threshold, and their horizontal offset is less than the fourth distance threshold, therefore they are directly connected. The height difference between the second and third segments is between the third and fifth distance thresholds, and a theoretical external riser pipe segment exists within this height range; therefore, this theoretical segment is inserted as a supplement. The third and fourth segments meet the direct connection condition, ultimately forming a continuous candidate external riser pipe path. This implementation method, through reasonable area division and an intelligent completion mechanism, successfully recovers some missing pipe segments caused by balcony obstruction, achieving the identification of a complete external riser pipe path.

[0061] Through the above solution, this application solves the problem of height gap between adjacent candidate pipe segments of external risers caused by partial occlusion of building facade, missing point cloud data, or image recognition errors. It ensures the vertical continuity of candidate paths of external risers, improves the accuracy and reliability of complete path identification of external risers, and provides a solid technical foundation for the asset management of external risers.

[0062] Specifically, in some embodiments of this application, a method for generating candidate paths for external risers is proposed to connect candidate segments of external risers. However, in its implementation, due to partial occlusion, registration errors, or data loss in building facade image data or point cloud data, the generated candidate paths may deviate from the spatial topological constraints of the actual external risers in terms of vertical continuity, support spacing, floor alignment, or combination arrangement. For example, there may be unnatural bends, support spacing exceeding the standard range, floor crossing positions offset from the elevation, or disordered arrangement of different types of external risers. As a result, invalid paths are retained in the identification process, affecting the accuracy of the final external riser asset database.

[0063] In this regard, this application further proposes a step for evaluating the topological consistency of each candidate path for an external riser based on the external riser spatial topological constraint library, including: For each candidate path of the external riser, the evaluation result of the vertical continuity of the external riser is calculated based on the horizontal offset and axial direction deviation of each candidate external riser segment in a unified coordinate system. The evaluation results of the spacing between external riser supports are calculated based on the vertical distance between the support positioning points; The evaluation results of external riser floor alignment are calculated based on the height difference between the floor crossing location and the corresponding floor elevation. The evaluation results of the external riser combination layout are calculated based on the horizontal distance and arrangement order between the candidate external riser paths and adjacent candidate external riser paths within the same building facade partition. The topological consistency evaluation results of candidate external riser paths are obtained by weighting and superimposing the evaluation results of vertical continuity of external riser, external riser support spacing, external riser floor alignment, and external riser combination layout according to preset weights.

[0064] The evaluation results for the vertical continuity of external risers assess the degree of continuity of the external risers in the vertical direction. This is achieved by calculating the horizontal offset of the centerline of adjacent candidate external riser segments and the angular deviation between the axial direction and the vertical coordinate axis, aiming to ensure that the external riser path has no significant vertical deviation. The evaluation results for the external riser support spacing assess whether the support installation spacing conforms to engineering specifications. This is achieved by measuring the vertical distance between support positioning points on the candidate external riser path and comparing it with preset minimum and maximum support spacing ranges, aiming to prevent structural safety hazards caused by excessively dense or sparse supports. The evaluation results for the external riser floor alignment assess the accuracy of the external riser's floor crossing locations. This is achieved by calculating the vertical distance between the centerline of adjacent candidate external riser segments and the angular deviation between the axial direction and the vertical coordinate axis, aiming to ensure that the external riser path has no significant vertical deviation. The evaluation of external risers is achieved by measuring the difference between the actual height of the floor crossing point and the corresponding floor elevation in the building information model. This ensures precise alignment between the external risers and the building structure. The evaluation result of the external riser combination layout specifically assesses the spatial relationship between different types of external risers. This can be achieved by analyzing whether the horizontal spacing and arrangement order of adjacent external riser candidate paths within the same building facade partition conform to the preset combination pattern of water supply, drainage, and gas external risers. This aims to avoid maintenance difficulties caused by chaotic external riser layout. The topology consistency evaluation result is a comprehensive index obtained by weighting and summing the above evaluation results according to preset weights. This can be achieved using linear weighting or nonlinear combination methods. The purpose is to provide an objective and quantitative standard for screening external riser candidate paths.

[0065] Specifically, this application's solution quantifies the compliance of each candidate external riser path in four dimensions—vertical continuity, support spacing, floor alignment, and combined arrangement—within a unified coordinate system framework. Based on the precise geometric benchmark provided by the unified coordinate system, the vertical continuity evaluation result is first calculated using the horizontal offset and axial direction deviation of the candidate external riser segments to ensure that the vertical continuity of the path is not affected by missing local data. Secondly, combined with the support specification parameters in the external riser maintenance log, the support spacing evaluation result is calculated by the vertical distance between support positioning points to prevent invalid paths due to data noise. Simultaneously, relying on the prior knowledge of floor elevations in the building information model, the floor alignment evaluation result is calculated by the height difference between the floor crossing location and the floor elevation to reduce deviations caused by texture interference or point cloud noise. Furthermore, based on the spatial relationship constraints of different types of external risers in the engineering specifications, the combined arrangement evaluation result is calculated by the horizontal distance and arrangement order of adjacent candidate external riser paths to avoid cross-misalignment that does not conform to the installation mode in dense areas. Finally, the evaluation results are weighted and superimposed according to the preset weights to form a comprehensive topology consistency evaluation index. This index can objectively reflect the overall degree of conformity between the candidate path and the actual spatial topology constraints, thereby accurately eliminating invalid paths under the condition of data uncertainty.

[0066] As a preferred embodiment, the specific implementation of this application is as follows: Consider a candidate path for an external riser within a building facade section, which is formed by connecting multiple candidate external riser segments from bottom to top. First, calculate the evaluation result of the vertical continuity of the external riser. By analyzing the horizontal offset of the centerline and the axial direction deviation of each segment, it is confirmed that there is no non-physical bending in the vertical direction of the path. Second, calculate the evaluation result of the external riser support spacing. Measure the vertical distance between the support positioning points on the path to verify that it is within the spacing range allowed by the specifications. Next, calculate the evaluation result of the external riser floor alignment. Compare the actual height of the path at the floor crossing point with the floor elevation to ensure that the height difference is within a reasonable tolerance. Then, calculate the evaluation result of the external riser combination layout. Check the horizontal spacing and arrangement order of the path with the adjacent drainage external riser candidate paths to confirm that it conforms to the preset combination pattern of water supply, drainage, and gas external risers. Finally, the above four evaluation results are weighted and superimposed according to preset weights. If the comprehensive evaluation result is higher than the preset threshold, the path is retained as an external riser path; otherwise, it is eliminated.

[0067] Through the above scheme, this application can eliminate invalid candidate paths for external risers caused by partial occlusion, registration errors or data loss in building facade image data or point cloud data, and ensure that the final external riser identification results strictly follow the spatial topological constraints of vertical continuity, bracket spacing, floor alignment and combination arrangement, thereby improving the accuracy and reliability of path information and type records in the external riser asset library in complex engineering scenarios.

[0068] Specifically, in some of the embodiments described above in this application, a set of target external riser paths is proposed to be determined based on the topology consistency evaluation results. However, in the process of its implementation, the remaining candidate external riser paths may have spatial geometric overlap, fail to meet the requirements for the number of internal and external risers in the building facade zoning, or violate the constraints on the combination and arrangement of external risers, resulting in the inability to generate accurate final external riser identification results that conform to actual operation and maintenance specifications.

[0069] In response, this application further proposes that when determining the target set of external risers from the remaining candidate paths after elimination, based on the topology consistency evaluation results, the following steps are taken: the topology consistency evaluation result of each candidate external riser path is used as the path evaluation value; within each building facade partition, the selection principle is based on maximizing the sum of path evaluation values; and under the conditions that different candidate external risers do not geometrically overlap in a unified coordinate system and meet the requirements for the number of external risers and the constraints on the arrangement of external risers within the same building facade partition, the target set of external risers is selected from the remaining candidate external risers through a combination of stepwise addition and backtracking elimination, thus forming the final external riser identification result.

[0070] Among them, the path evaluation value refers to the quantification index of the topology consistency evaluation result. It can be achieved by directly using the original evaluation value or through linear normalization. The purpose is to objectively inherit the previous evaluation results of the spatial topological rationality of candidate paths and avoid the bias introduced by subjective judgment. The principle of maximizing the sum of path evaluation values ​​can be understood as the optimization objective of maximizing the overall path quality under the constraints. It can be implemented using integer programming or heuristic algorithm frameworks. The purpose is to prioritize the retention of high-quality paths to improve the reliability of the final identification result. Constraining different external pipe candidate paths to avoid geometric overlap in a unified coordinate system can be specifically implemented as a spatial collision detection mechanism in practical applications, such as by calculating the path The minimum distance threshold of the centerline is used for mutual exclusion verification. The purpose is to enforce compliance with the non-intersecting characteristic of the physical installation of external risers and prevent logical contradictions in the identification results. To meet the requirements for the number of external risers and the constraints on the arrangement of external risers, specific measures can be taken, such as checking the preset quantity range and matching the arrangement order rules. For example, verifying whether the horizontal spacing and relative position of water supply external risers, drainage external risers and gas external risers meet the specifications. The purpose is to ensure that the identification results are consistent with the building operation and maintenance specifications. The method of progressive addition and backtracking elimination can be combined. In practical applications, an algorithm combining depth-first search and pruning strategy can be used. The purpose is to dynamically balance path quality and constraint satisfaction, avoid the high complexity of exhaustive calculation, and ensure the optimality of the solution.

[0071] Specifically, the proposed solution transforms topology consistency evaluation results into quantified path evaluation values. Within the building facade partitions, the objective function is to maximize the sum of path evaluation values. Simultaneously, it enforces spatial geometric exclusivity, quantity specifications, and combination layout constraints. An optimization mechanism involving progressive addition and backtracking elimination dynamically constructs a set of target external riser paths. This mechanism first prioritizes high-quality paths based on their evaluation values, adding them to the candidate set. Then, during the addition process, it verifies in real-time whether new paths cause geometric overlap, exceed quantity limits, or violate combination order rules. If a constraint conflict is detected, the current path is immediately backtracked and eliminated, and a second-best option is attempted. This ensures that all operational constraints are met while maintaining overall path quality, ultimately generating external riser identification results that are free of spatial conflicts and conform to specifications.

[0072] As a specific implementation method, in a certain building facade partition, after calculating the path evaluation value of each candidate path for external risers, an empty target path set is initialized. The algorithm first selects the candidate path with the highest evaluation value and adds it to the set. Then, it checks the spatial position relationship between this path and the existing paths in the set. If there is geometric overlap, the path is backtracked and eliminated. If there is no overlap, it further verifies whether the total number of internal and external risers in the partition exceeds the preset upper limit and whether the arrangement order of water supply external risers, drainage external risers, and gas external risers conforms to the preset pattern. If the constraints are violated, they are backtracked and adjusted. If all constraints are satisfied, the path is retained and the next candidate path with the second highest evaluation value is processed. This process is iterated until all candidate paths are evaluated, and finally, a target external riser path set that meets the requirements of spatial mutual exclusion, quantity specifications, and combination layout, and has the largest sum of path evaluation values ​​is formed.

[0073] Through the above technical solution, this application can eliminate spatial geometric conflicts between candidate external riser paths, ensure that the final identification results strictly comply with the specifications for the number and combination of internal and external risers in the building facade zoning, and thus generate an accurate set of external riser paths that can be directly used for external riser asset management.

[0074] In some of the embodiments described above in this application, a method for determining the target set of external riser paths is proposed. However, in this process, the specific functional types of external risers are not clearly distinguished, and there is a lack of a mechanism to structurally store the identification results in the asset database. This makes it impossible for the operation and maintenance department to conduct accurate management of external risers based on digital means, especially when there is partial obstruction or discrepancies in the ledger data, it is difficult to accurately associate facility functions with spatial paths.

[0075] In response, this application further proposes that when determining the type of each external riser in the target external riser path set based on the external riser maintenance ledger data and writing the final external riser identification result into the external riser asset database, the following steps are taken: mapping the locations of water supply facilities, drainage facilities, and gas facilities recorded in the external riser maintenance ledger data, as well as their connection relationships with external risers, to a unified coordinate system; for each external riser in the final external riser identification result, based on the connection relationship between the external riser and the water supply facilities, drainage facilities, and gas facilities in the vertical direction, as well as the service floor range, the external riser is identified as a water supply external riser, a drainage external riser, or a gas external riser; and the external riser spatial path, external riser type, external riser support positioning point, external riser floor crossing location, and corresponding topology consistency evaluation result are recorded for each external riser in the external riser asset database.

[0076] Among them, the external riser maintenance ledger data refers to the facility management database accumulated by the maintenance department over a long period of time. It can be implemented using a relational database or unstructured document storage format. Its purpose is to integrate historical maintenance information and provide a data foundation for type determination. The external riser asset database can be understood as a dedicated database system for asset management. It can be implemented using a distributed database architecture or cloud storage service. Its purpose is to build a traceable and verifiable asset information management framework. The vertical connection relationship specifically refers to the physical connection topology between the external riser and the facility in the direction of gravity. It can be implemented using spatial proximity analysis or geometric projection matching methods. Its purpose is to establish the mapping logic between functional type and spatial location.

[0077] Specifically, the proposed solution maps the locations of water supply, drainage, and gas facilities, as well as their connections to the external risers, recorded in the external riser maintenance ledger data, to a unified coordinate system. This ensures accurate alignment of data from different sources within a unified spatial framework, avoiding facility association errors caused by coordinate differences. Based on this, for each external riser in the final identification result, its type is determined according to its vertical connection to the water supply, drainage, and gas facilities, as well as the service floor range, forming a dual judgment mechanism based on vertical connection relationships and service floors. Finally, by structurally recording the external riser spatial path, type, support location, floor crossing location, and corresponding topological consistency evaluation results in the external riser asset database, spatial geometric information, functional attributes, and quality assessment indicators are organically integrated, thereby constructing a complete external riser asset information chain.

[0078] As a specific implementation method, the solution of this application is implemented as follows: In the process of identifying the facade of a high-rise building, the location of the pump room, the coordinates of the drainage inspection well, and the installation point of the gas pressure regulating box recorded in the operation and maintenance log are first mapped to a unified coordinate system; for the identified external riser path, the spatial connection relationship between its bottom and the water supply facility and the service floor range at the top are analyzed. When the bottom of an external riser is directly connected to the water outlet of the pump room and the service floor covers the domestic water area, it is determined to be a water supply external riser; at the same time, a record entry is established in the external riser asset database to store the three-dimensional spatial trajectory data, type identifier, bracket installation point coordinates, floor crossing elevation value, and topology consistency score of the external riser. This asset database can be specifically implemented using a PostgreSQL database that supports spatial data expansion and provides a visual query function through a geographic information system interface.

[0079] Through the above technical solutions, this application has achieved accurate identification of the functional types of external risers and systematic management of asset information, solved the problem of accurate association between facility functions and spatial paths under conditions of partial obstruction or deviation in ledger data, provided the operation and maintenance department with traceable and verifiable external riser asset data support, and improved the practicality and reliability of digital management of external risers.

[0080] It is understood that, in the various embodiments of this application, the threshold parameters such as the first distance threshold, the second distance threshold, the horizontal distance threshold, the third distance threshold, the fourth distance threshold, the fifth distance threshold, the height interval overlap threshold, the preset overlap threshold, the preset error threshold, and the minimum and maximum support spacing in the external riser support spacing constraints can all be set based on engineering specifications and actual application scenarios of the external riser through engineering experience, and can be adjusted within a reasonable range to balance recognition accuracy and calculation stability.

[0081] Specifically, the minimum and maximum support spacing in the external riser support spacing constraints can be set with reference to relevant building water supply, drainage and gas pipeline specifications and existing engineering experience. For external risers of ordinary civil buildings, it is preferable to make the vertical distance between adjacent supports fall within the range of one to two meters to ensure the structural stability of the external riser and the recognizability of support features in images and point clouds. In the external riser combination arrangement constraints, the horizontal distance between different external risers can be set with reference to the safety distance requirements and in combination with the building facade scale. Generally, the horizontal distance between different external risers can be in the range of tens of centimeters to about one meter to avoid misidentifying adjacent components as the same external riser.

[0082] The first, second, third, fourth, and fifth distance thresholds used for path connection and floor alignment can be determined proportionally based on the building's floor clear height and the diameter of the external riser. A preferred approach is to control the allowable height difference between adjacent external riser candidate segments within a certain percentage range of the corresponding floor clear height, for example, not exceeding 5% to 20% of the floor clear height, and to control the allowable horizontal offset within a range of several times the external riser diameter, so as to allow slight offsets in actual installation while avoiding misconnecting different risers to the same path. The height interval overlap threshold and the preset overlap threshold can be set according to the vertical coverage length of the external riser candidate segments. Preferably, the height interval overlap length corresponding to different data sources is not less than a certain proportion of the shorter segment length, for example, not less than 30% to 50% of that length, to ensure that multi-source data matching has a sufficient real overlap area.

[0083] The first and second angle thresholds used for image extraction and point cloud extraction can be set according to the allowable deviation of the building's external riser from the vertical direction during actual installation. Generally, the deviation angle of the riser axis relative to the vertical direction is small. It is preferable to set the first and second angle thresholds in the range of 10° to 20° to exclude obviously tilted linear targets and point cloud targets from the candidate set. The preset error threshold is used to evaluate the registration accuracy of the building facade image data and the building facade point cloud data in a unified coordinate system. It can be determined according to the building height and point cloud density. One option is to control the preset error threshold between several thousandths and several hundredths of the building facade height so that the registration error is within an acceptable range relative to the overall building scale, thereby ensuring the spatial accuracy of subsequent external riser candidate segment matching and external riser candidate path generation.

[0084] In summary, this application establishes a unified coordinate system on the building facade and constructs a spatial topological constraint library for external pipes. Under the same spatial reference, it collaboratively processes building facade image data, building facade point cloud data, and external pipe operation and maintenance ledger data. First, linear targets are extracted from the images, and tubular point cloud clusters are extracted from the point cloud. Combined with the ledger, the theoretical pipe segment trajectories of external pipes are generated. Then, based on the horizontal distance and height interval overlap relationship, these are fused into candidate external pipe segments and connected along the building facade partitions and vertical directions to form candidate external pipe paths. Spatial topological constraints are used to evaluate the topological consistency of each candidate external pipe path, and those with evaluation results below a preset evaluation threshold are eliminated. The candidate paths for external risers are determined, and based on the topology consistency evaluation results, the set of target external riser paths is determined. Combined with the external riser operation and maintenance ledger data, the external riser type of each external riser in the target external riser path set is determined. The final external riser identification results are written into the external riser asset database. Thus, even in complex scenarios with partial occlusion, data loss, and ledger discrepancies, the continuous spatial path and accurate type of each external riser can still be restored by relying on the synergistic effect of multi-source data and spatial topology constraints. This achieves highly reliable asset output for external riser inspection and maintenance, improving the completeness of external riser identification compared to existing technologies that rely solely on 3D reconstruction or local image recognition.

[0085] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for collaborative identification of multi-source data of external risers based on spatial topological constraints, characterized in that, include: A unified coordinate system is established on the building facade, and a spatial topology constraint library for external risers is constructed based on multi-source data of external risers. The external riser spatial topology constraint library includes external riser vertical continuity constraints, external riser support spacing constraints, external riser floor alignment constraints, and external riser combination arrangement constraints. Acquire building facade image data and building facade point cloud data corresponding to the building, and perform spatial registration in the unified coordinate system; Linear targets are extracted based on the building facade image data, tubular point cloud clusters are extracted based on the building facade point cloud data, and theoretical pipe segments are generated by combining the external pipe maintenance ledger data. Linear targets, tubular point cloud clusters and theoretical pipe segments with horizontal distances less than a preset distance threshold and height overlap greater than a preset overlap threshold are classified as external pipe candidate segments under the unified coordinate system. According to the building facade zoning and vertical direction, the candidate pipe segments of the external riser are connected from bottom to top to generate candidate external riser paths. Based on the external riser spatial topology constraint library, the external riser vertical continuity, external riser support spacing, external riser floor alignment, and external riser combination arrangement are used to evaluate the topology consistency of each candidate external riser path. Candidate external riser paths with topology consistency evaluation results lower than the preset evaluation threshold are eliminated. In the remaining candidate paths of external risers after elimination, the target external riser path set is determined according to the topology consistency evaluation results to form the final external riser identification result, and the external riser type of each external riser in the target external riser path set is determined according to the external riser operation and maintenance ledger data. The final external riser identification result is written into the external riser asset database.

2. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 1, characterized in that, When establishing a unified coordinate system on the building facade and constructing a spatial topology constraint library for the external risers based on multi-source data, the following steps are included: The multi-source data for external risers includes architectural design drawings, as-built drawings, building information model data, and external riser operation and maintenance ledger data. Using the pre-set reference point on the building facade as the origin, the direction of gravity as the vertical coordinate axis, and the main axis of the building structure as the horizontal coordinate axis, coordinate transformation is performed on the geometric position and floor elevation of the external riser in the multi-source data of the external riser. The geometric relationships and standard parameters of the external risers in the unified coordinate system are converted into external riser vertical continuity constraints, external riser support spacing constraints, external riser floor alignment constraints, and external riser combination layout constraints. Among them, the external riser vertical continuity constraints limit the horizontal offset of the same external riser in adjacent floors to no more than a first distance threshold and the axial direction deviation from the vertical direction to no more than a first angle threshold. The external riser support spacing constraints limit the vertical distance between adjacent supports of the same external riser to be between the minimum support spacing and the maximum support spacing set according to the external riser type. The external riser floor alignment constraints limit the height deviation of the external riser floor crossing position from the corresponding floor elevation to no more than a second distance threshold. The external riser combination layout constraints limit the horizontal distance between different external risers in the same building facade partition to be within a preset spacing range and the arrangement order of water supply external risers, drainage external risers, and gas external risers to conform to a preset combination pattern.

3. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 1, characterized in that, When acquiring building facade image data and building facade point cloud data corresponding to the building and performing spatial registration in a unified coordinate system, the following steps are included: Acquire multi-view image data of the building facade covering the building facade, acquire point cloud data of the building facade covering the building facade, set up at least three control points on the building facade and pre-record the spatial coordinates of each control point in the unified coordinate system, detect the image position of each control point in the building facade image data, extract the spatial position of each control point in the building facade point cloud data, determine the registration transformation to align the building facade image data and building facade point cloud data with the unified coordinate system based on the spatial coordinates of the control points in the unified coordinate system and the observation position in the building facade image data and building facade point cloud data, and refine the registration transformation by matching the building facade outline and door and window boundary lines to ensure that the positional deviation of the building facade image data and building facade point cloud data in the unified coordinate system is lower than a preset error threshold.

4. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 1, characterized in that, Extracting linear targets based on the building facade image data includes: The building facade image data is subjected to grayscale processing and noise reduction processing, edge detection and connected region extraction are performed, and connected regions with a length greater than a first length threshold, a length-to-width ratio greater than a first ratio threshold, and an angle between the main extension direction and the vertical coordinate axis less than a first angle threshold are retained. The center line of each connected region is determined as the linear target.

5. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 4, characterized in that, Extracting tubular point cloud clusters based on the building facade point cloud data includes: Ground point removal and background point removal are performed on the point cloud data of the building facade. The location point set close to the building facade is extracted. The main extension direction and cross-sectional point cloud distribution range are calculated for each location point set. The point set with the angle between the main extension direction and the vertical coordinate axis less than the second angle threshold and the cross-sectional point cloud distribution radius within the preset radius is retained, and the point set is determined as the tubular point cloud cluster.

6. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 5, characterized in that, When generating the theoretical pipeline trajectory of the external pipeline by combining the external pipeline maintenance ledger data and dividing it into external pipeline theoretical pipeline segments under the unified coordinate system, the process includes: Based on the starting position, ending position and corresponding floor of the external riser recorded in the external riser maintenance ledger data, the starting position and ending position are mapped to the unified coordinate system, a connection path is generated along the vertical coordinate axis and divided into multiple external riser theoretical pipe segments according to the floor height; Under the unified coordinate system, the linear targets, tubular point cloud clusters and external riser theoretical pipe segments are matched according to the horizontal distance threshold and the height interval overlap threshold. Targets that simultaneously meet the distance threshold and height overlap threshold in space with at least two data sources are classified as external riser candidate pipe segments, and targets supported by only a single data source and located between adjacent external riser candidate pipe segments are designated as supplementary external riser candidate pipe segments.

7. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 1, characterized in that, When connecting the candidate external riser pipe segments from bottom to top to generate candidate external riser paths according to the building facade zoning and vertical direction, the following is included: Within each building facade section, candidate external riser segments are divided into several vertical pipeline groups according to their horizontal position. Within each vertical pipeline group, the candidate external riser segments are sorted from low to high according to their height. Adjacent candidate external riser segments with a height difference less than a third distance threshold and a horizontal offset less than a fourth distance threshold are connected to form the same candidate external riser path. When the height difference between adjacent candidate external riser segments is between the third and fifth distance thresholds, and there is a theoretical external riser segment or a supplementary candidate external riser segment within this height range, the theoretical external riser segment or the supplementary candidate external riser segment is inserted into the candidate external riser path to fill in the gaps and form a continuous candidate external riser path in the vertical direction.

8. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 7, characterized in that, When performing topology consistency evaluation on each candidate path of an external riser based on the external riser spatial topology constraint library, the following is included: For each candidate path of the external riser, the vertical continuity evaluation result of the external riser is calculated based on the horizontal offset and axial direction deviation of each candidate external riser segment under the unified coordinate system; the support spacing evaluation result of the external riser is calculated based on the vertical distance between the support positioning points; the floor alignment evaluation result of the external riser is calculated based on the height difference between the floor crossing position and the corresponding floor elevation; and the combination layout evaluation result of the external riser is calculated based on the horizontal distance and arrangement order between the candidate external riser path and adjacent candidate external riser paths within the same building facade partition. The vertical continuity evaluation result of the external riser, the support spacing evaluation result of the external riser, the floor alignment evaluation result of the external riser, and the combination layout evaluation result of the external riser are weighted and superimposed according to preset weights to obtain the topological consistency evaluation result of the candidate external riser path.

9. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 1, characterized in that, When determining the target set of external riser paths from the remaining candidate paths after elimination, based on the topology consistency evaluation results, the following steps are included: The topological consistency evaluation result of each candidate external pipe path is used as the path evaluation value. Within each building facade partition, the selection principle is based on maximizing the sum of the path evaluation values. Under the conditions that different candidate external pipe paths do not geometrically overlap in the unified coordinate system and meet the requirements for the number of external pipes and the constraints on the arrangement of external pipe combinations within the same building facade partition, the target set of external pipe paths is selected from the remaining candidate external pipe paths by a combination of stepwise addition and backtracking elimination, thus forming the final external pipe identification result.

10. The method for collaborative identification of multi-source data of external risers based on spatial topological constraints according to claim 9, characterized in that, When determining the external riser type for each external riser in the target external riser path set based on the external riser maintenance ledger data and writing the final external riser identification result into the external riser asset database, the following steps are included: The locations of water supply facilities, drainage facilities, and gas facilities, as well as their connection relationships with external risers, recorded in the external riser maintenance ledger data, are mapped to the unified coordinate system. For each external riser in the final external riser identification result, based on its vertical connection relationship with the water supply facilities, drainage facilities, and gas facilities, and the service floor range, the external riser is identified as a water supply external riser, drainage external riser, or gas external riser. In the external riser asset database, the external riser spatial path, external riser type, external riser support positioning point, external riser floor crossing location, and corresponding topology consistency evaluation result are recorded for each external riser.

Images