A method and system for highway longitudinal and transverse section segmentation integrating multi-scale grid indexing

Abstract

This application discloses a method and system for highway longitudinal and transverse section segmentation based on multi-scale grid indexing, relating to the field of highway surveying and design technology. The method includes: using a multi-scale grid index to divide road object data and road 3D point cloud data into road entity occupancy data; combining the road entity occupancy data, identifying overlapping occupancy in the cross-section cutting zone and longitudinal section extraction zone to obtain overlapping occupancy areas, and analyzing the road entity occupancy status within the overlapping occupancy areas to obtain section ownership conflict data; dividing the overlapping occupancy areas into point cloud ownership boundaries to obtain section ownership boundary data; collecting point clouds belonging to the road to be segmented from the road 3D point cloud data located within the cross-section cutting zone and longitudinal section extraction zone, and performing cross-section segmentation and longitudinal section segmentation to obtain section results. This application improves the accuracy of section segmentation results in complex road scenarios.

Description

Technical Field

[0001] This application relates to the field of highway surveying and design technology, and more specifically, to a method and system for highway longitudinal and transverse section segmentation that integrates multi-scale grid indexing. Background Technology

[0002] Highway longitudinal and transverse section segmentation is a fundamental data processing component in highway surveying and design, reconstruction and expansion re-surveying, construction verification, and maintenance inspection. With the development of vehicle-mounted laser scanning, UAV mapping, airborne LiDAR, and real-scene 3D measurement technologies, roadside terrain, pavement, shoulders, slopes, ditches, under-bridge spaces, and road ancillary structures can be collected in the form of 3D point cloud data. Highway longitudinal and transverse section segmentation typically requires combining road centerline, mileage data, and cross-section setting information to extract the corresponding transverse and longitudinal section data from the 3D point cloud data of the road to be segmented, thus characterizing the geometric shape of the road in terms of lateral structure and elevation along the route.

[0003] Existing methods for segmenting highway longitudinal and transverse sections typically determine the section line, cutting range, or extraction range based on the road centerline, mileage markers, or section setting location. They then combine this with the distance from a point to the section line, the centerline buffer range, the planar grid range, or spatial index results to select point clouds from the road's 3D point cloud data for generating the section results. This type of method can meet general processing needs in scenarios involving a single road entity or relatively simple road spatial relationships. However, in scenarios such as interchanges, parallel sections of main and auxiliary roads, urban elevated roads and underpasses, underpasses, ramps connecting to the main line, and service area connections, the road to be segmented may be close to, partially parallel to, or intersect with other road entities in planar position, or overlap within the planar projection range. In these cases, the point clouds belonging to other road entities may also fall within the section cutting range or extraction range of the road to be segmented, and simultaneously meet the existing selection criteria. This makes it difficult to accurately determine whether a point cloud truly belongs to the road to be segmented when selecting point clouds solely based on spatial location and cutting range.

[0004] When point clouds belonging to other road entities are mistakenly included in the cross-sectional data of the road to be segmented, the cross-sectional results may contain pavement points, shoulder points, slope points, ditch points, or structure points that do not belong to the road to be segmented. This can lead to abnormal cross-sectional widths, duplicate pavement representations, distorted slope contours, ditch location shifts, or incorrect locations of attached structures. For longitudinal section results, if point clouds belonging to other road entities participate in elevation extraction, it may also cause abnormal longitudinal elevation changes, abrupt changes in local elevations, or distorted longitudinal slope representations. Since the mistakenly included point clouds are still geometrically within the cross-sectional processing range, the generated results may still appear as a continuous series of cross-sectional points or a complete cross-sectional contour. During manual verification, it is not easy to directly identify the incorrect road entity attribution, thus affecting the reliability of cut and fill calculations, roadbed width identification, bridge clearance verification, construction re-surveys, and digital road management results.

[0005] The shortcomings of existing technologies lie in the fact that point cloud selection during highway longitudinal and transverse section segmentation primarily revolves around spatial positional relationships. Spatial indexing, grid reading, and section range constraints are mainly used to characterize the geometric proximity between the point cloud and the section processing area, and cannot fully reflect the differences in engineering affiliation between the point cloud and different road entities. In scenarios where multiple road entities are close to each other, have overlapping planar projections, or are intertwined, spatial proximity does not necessarily indicate consistent road entity affiliation. Existing methods still carry the risk of mistakenly including point clouds belonging to other road entities in the longitudinal and transverse sections of the road to be segmented, making it difficult to guarantee the accuracy of highway longitudinal and transverse section segmentation results in complex road scenarios.

[0006] In view of this, this application proposes a method and system for highway longitudinal and transverse section segmentation that integrates multi-scale grid indexes to solve the above problems. Summary of the Invention

[0007] To overcome the aforementioned deficiencies of the prior art and achieve the above objectives, this application provides the following technical solution: a method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes, comprising: Acquire road 3D point cloud data, road alignment data, mileage data, and road entity data; associate the road alignment data, mileage data, and road entity data according to the same road entity to obtain road object data; Using a multi-scale grid index, road entity occupancy is divided into road object data and road 3D point cloud data to obtain road entity occupancy data; Extract road object data to be segmented from road object data, define the cross-sectional cutting range and longitudinal section extraction range of the road object data to be segmented, and obtain the cross-sectional cutting zone and longitudinal section extraction zone. By combining road entity occupancy data, overlapping occupancy is identified in cross-sectional section strips and longitudinal section extraction strips to obtain overlapping occupancy areas. The occupancy status of road entities within the overlapping occupancy areas is then analyzed to obtain cross-section ownership conflict data. By combining the cross-section ownership conflict data, the point cloud ownership boundary of the overlapping occupied area is divided to obtain the cross-section ownership boundary data. By combining the boundary data of the cross-section, the point clouds belonging to the road to be segmented are collected from the three-dimensional point cloud data of the road located within the cross-section cutting zone and the longitudinal section extraction zone, and the belonging point clouds are obtained. The cross-section and longitudinal section segments are then performed on the belonging point clouds to obtain the cross-section results.

[0008] Furthermore, the method for obtaining road entity occupancy data includes: The point cloud points in the road 3D point cloud data are assigned to the corresponding grid regions according to the multi-scale grid index to obtain the point cloud set within each grid region. Extract the road coverage boundary, mileage data and road entity information of each road entity from the road object data. Determine the effective mileage range according to the mileage data and truncate the road coverage boundary. Associate the truncated road coverage boundary with the road entity information to form the road entity occupancy range. The planar projection range of each grid area is compared with the occupancy range of each road entity. If the same grid area intersects with only one road entity occupancy range, the road entity occupancy status is marked as single road entity occupancy status. If the same grid area intersects with the occupied areas of two or more road entities, the road entity occupancy status is marked as multi-road entity occupancy status. By organizing the grid area, the point cloud set within the grid, the road entity occupancy range, and the road entity occupancy status, road entity occupancy data is obtained.

[0009] Furthermore, the method for obtaining the overlapping occupied area includes: Based on road entity occupancy data, the occupancy status of the grid areas covered by the cross section cutting zone and the longitudinal section extraction zone is filtered. The grid areas where the road entity occupancy status is multiple road entity occupancy status and the corresponding road entities include the road to be segmented and other road entities are extracted to obtain candidate overlapping grid areas. Based on the cross-sectional section or longitudinal section extraction zone, the road to be segmented, and other road entities to which the candidate overlapping grid regions belong, the candidate overlapping grid regions are merged. Candidate overlapping grid regions that share boundaries and belong to the same cross-sectional section or longitudinal section extraction zone, the road to be segmented, and other road entities are merged into overlapping occupied regions.

[0010] Furthermore, the method for analyzing the occupancy status of road entities within the overlapping occupancy area to obtain cross-section attribution conflict data includes: Extract the road entity hierarchy information of the road to be segmented and other road entities corresponding to the overlapping occupied area, extract the local road centerline segment in the overlapping occupied area, and identify the access status of other road entities relative to the road coverage boundary of the road to be segmented. Based on the road entity hierarchy information, the intersection of local road centerline segments, the minimum included angle of local road centerline segments, and the access of road coverage boundaries, the cross-section ownership conflict type of the overlapping occupied area is identified. The overlapping occupied areas, roads to be divided, other road entities, and cross-section ownership conflict types are organized into cross-section ownership conflict data. The cross-section ownership conflict types include upper and lower layer conflicts, parallel conflicts, or access conflicts.

[0011] Furthermore, the method for obtaining the cross-section attribution boundary data includes: When the cross-section ownership conflict type is upper and lower layer conflict, the point cloud ownership boundary is determined according to the elevation distribution in the overlapping area. When the cross-section assignment conflict type is parallel conflict, the point cloud assignment boundary is determined according to the relative state of the boundary segments between the road to be segmented and other road entities. When the cross-section ownership conflict type is access conflict, the point cloud ownership boundary is determined according to the distance relationship between the centerline of the road to be segmented and other road entities. Based on the point cloud attribution boundary, the scope of the road to be segmented and the scope of other road entities are delineated, and the overlapping occupied areas, cross-section attribution conflict types, point cloud attribution boundaries, scope of the road to be segmented and the scope of other road entities are organized to obtain cross-section attribution boundary data.

[0012] Furthermore, when the cross-section attribution conflict type is an upper-lower layer conflict, the method for determining the point cloud attribution boundary based on the elevation distribution within the overlapping area includes: Within the same cross-sectional section or the same longitudinal section extraction zone, filter out the grid regions that share the boundary with the overlapping occupied region to obtain the elevation reference range of the overlapping occupied region; Based on the elevation reference range of the overlapping occupied area and the road entity occupancy data, the single road entity occupancy status grid area corresponding to the road to be segmented and other road entities is extracted respectively, and the elevation merging statistics of the point cloud set in the grid are performed to obtain the elevation range of the road to be segmented and the elevation range of other road entities. The point cloud set within the grid in the overlapping area is sorted by elevation. The position where the elevation interval between adjacent point cloud points is greater than or equal to the preset inter-layer interval threshold is taken as the elevation break position, and the elevation continuous segment is obtained. For continuous elevation segments that intersect with the elevation range of the road to be segmented and the elevation range of other road entities, an elevation attribution determination quantity is generated based on the degree of overlap of elevation ranges, the proximity of elevation centers, and the consistency of road entity levels. The continuous elevation segments with the largest elevation attribution determination quantity are selected to delineate the road side range to be segmented and the other road entity side ranges. The elevation break position between the two ranges is extracted to obtain the point cloud attribution boundary corresponding to the upper and lower layer conflicts.

[0013] Furthermore, when the cross-section attribution conflict type is a parallel conflict, the method for determining the point cloud attribution boundary based on the relative state of the boundary segments between the road to be segmented and other road entities includes: Along the lateral direction perpendicular to the local road centerline segment corresponding to the road to be divided, extract the side boundary segment close to the road centerline of other road entities from the road coverage boundary of the road to be divided, and extract the side boundary segment close to the road centerline of the road to be divided from the road coverage boundary of other road entities. If there is a gap between the boundary segment of the road to be segmented that is closer to the center line of another road entity and the boundary segment of the road to be segmented that is closer to the center line of the road to be segmented, then the center line of the gap area is taken as the point cloud ownership boundary; if the two are connected, then the connection boundary is taken as the point cloud ownership boundary; if the two partially overlap, then the center line of the partially overlapping part along the lateral direction is taken as the point cloud ownership boundary, thus obtaining the point cloud ownership boundary corresponding to the parallel conflict.

[0014] Furthermore, when the cross-section attribution conflict type is an access conflict, the method for determining the point cloud attribution boundary based on the distance relationship between the centerlines of the road to be segmented and other road entities includes: Extract the entry position of the center line of other road entities into the road coverage boundary of the road to be divided, and the access position where the center line of other road entities connects with the center line of the road to be divided. Within the mileage range defined by the cross section or longitudinal section extraction zone corresponding to the overlapping area, the area between the entry position and the access position is taken as the access transition range. Within the access transition range, calculate the first plane distance from the center position of each grid area to the center line of the road to be segmented and the second plane distance to the center line of other road entities, and delineate the side range of the road to be segmented and the side range of other road entities based on the distance comparison; The boundary positions of adjacent ranges on both sides, as well as the locations of grid areas where the distance between the first plane and the second plane is the same or the absolute value of the difference is less than or equal to the distance calculation accuracy range, are organized into the point cloud belonging boundaries corresponding to the access conflict.

[0015] Furthermore, the method for obtaining the attribution point cloud includes: Candidate point cloud sets for cross sections and longitudinal sections are extracted from the cross section cutting zone and the longitudinal section extraction zone, respectively. For point cloud points in the cross-section candidate point cloud set and the longitudinal section candidate point cloud set that are not located in the overlapping occupied area, if the grid area where the point cloud point is located is occupied by a single road entity and the corresponding road entity is a road to be segmented, then it is assigned to the corresponding cross-section or longitudinal section point cloud. For point cloud points located within overlapping occupied areas, only point cloud points located within the road side range to be divided are assigned to the corresponding cross-section or longitudinal section point cloud; when the same overlapping occupied area corresponds to two or more other road entities, only point cloud points located within the road side range to be divided corresponding to the boundary of each point cloud are assigned to the corresponding cross-section or longitudinal section point cloud. Organize the cross-sectional and longitudinal section attribute point clouds to obtain the attribute point cloud.

[0016] A highway longitudinal and transverse section segmentation system integrating multi-scale grid indexing includes: The road object generation module is used to acquire road 3D point cloud data, road alignment data, mileage data and road entity data, and associate the road alignment data, mileage data and road entity data according to the same road entity to obtain road object data. The entity occupancy partitioning module is used to partition road entity occupancy data and road 3D point cloud data using a multi-scale grid index, thereby obtaining road entity occupancy data. The cross-section range generation module is used to extract road object data to be segmented from road object data, and to define the cross-section cutting range and longitudinal section extraction range of the road object data to be segmented, so as to obtain the cross-section cutting zone and the longitudinal section extraction zone. The attribution conflict generation module is used to combine road entity occupancy data to identify overlapping occupancy of cross-sectional cutting strips and longitudinal section extraction strips, obtain overlapping occupancy areas, and analyze the road entity occupancy status within the overlapping occupancy areas to obtain cross-sectional attribution conflict data. The attribution boundary generation module is used to combine cross-section attribution conflict data to divide the point cloud attribution boundaries of overlapping occupied areas, thereby obtaining cross-section attribution boundary data. The cross-section result generation module is used to combine the cross-section boundary data to collect the point clouds belonging to the road to be segmented from the road 3D point cloud data located within the cross-section cutting zone and the longitudinal section extraction zone, obtain the belonging point cloud, and perform cross-section segmentation and longitudinal section segmentation on the belonging point cloud to obtain the cross-section result.

[0017] Compared with existing technologies, the highway longitudinal and transverse section segmentation method and system of this application, which integrates multi-scale grid indexing, can reduce the situation where point clouds of other road entities are mistakenly included in the longitudinal and transverse sections of the road to be segmented, and improve the accuracy and reliability of highway longitudinal and transverse section segmentation results, for complex road scenarios where multiple road entities are close to each other, have overlapping planar projections, or are intertwined.

[0018] By associating road alignment data, mileage data, and road entity data as the same road entity, and using a multi-scale grid index to partition road object data and 3D point cloud data for road entity occupancy, road entity occupancy data that simultaneously reflects grid spatial distribution and road entity occupancy status can be generated. Compared to processing methods that only select point clouds based on cross-sectional range, planar grid, or spatial proximity, subsequent cross-sectional segmentation can be based on road entity attribution constraints, reducing the risk of misselection of point clouds in complex road scenarios from the source.

[0019] By identifying overlapping areas within the cross-sectional section and longitudinal section extraction zones, and analyzing the occupancy status of road entities within these overlapping areas, cross-section attribution conflict data can be obtained. Thus, areas within the processing range of the road cross-section to be segmented that may contain point clouds of other road entities can be located independently. The unclear attribution issues caused by multiple road entities entering the cross-section processing range can also be transformed into conflict determination criteria for further processing, providing a foundation for subsequent point cloud attribution boundary delineation.

[0020] For conflicts between upper and lower layers, parallel conflicts, and access conflicts, point cloud assignment boundaries are defined based on elevation distribution, relative state of boundary segments, and distance relationship with road centerlines, respectively. This yields cross-sectional assignment boundary data that matches the road weaving pattern. Point clouds within overlapping areas can thus be assigned to the area of ​​the road to be segmented or the area of ​​other road entities, avoiding assignment bias caused by using a single spatial distance rule to handle different types of road conflicts.

[0021] The formation of the attributed point cloud is constrained by both road entity occupancy data and cross-section attribute boundary data. Points in the point cloud not located within overlapping occupancy areas must satisfy the single road entity occupancy status and correspond to the road to be segmented; point points in the point cloud located within overlapping occupancy areas must be within the range of the road to be segmented. Generating cross-section and longitudinal profile data based on this aggregation result can reduce problems such as abnormal cross-section width, duplicate pavement representation, incorrect slope and ditch locations, and local elevation anomalies in the longitudinal profile.

[0022] In summary, this application addresses the problem that point clouds of other road entities in complex road scenarios are easily mixed into the processing range of the road section to be segmented. It constructs a processing chain that connects road entity occupancy division, overlapping occupancy identification, section ownership conflict analysis, point cloud ownership boundary division, and ownership point cloud generation, so that the longitudinal and transverse section segmentation results can more accurately correspond to the road to be segmented itself. Attached Figure Description

[0023] Figure 1 This is a flowchart of a method for segmenting longitudinal and transverse sections of a highway that integrates multi-scale grid indexes, according to an embodiment of this application. Figure 2 This is a schematic diagram of a highway longitudinal and transverse section segmentation system that integrates multi-scale grid indexing, according to an embodiment of this application. Detailed Implementation

[0024] The technical solutions of this application will be described in detail, clearly, and completely below with reference to the accompanying drawings of the embodiments. It should be particularly noted that the specific embodiments described below are only used to better illustrate and explain the technical solutions of this application, and are intended to enable those skilled in the art to better understand and implement this application, and should not be construed as limiting the scope of protection of this application. Without departing from the spirit and substance of this application, those skilled in the art can modify, adjust, or make equivalent substitutions based on the content disclosed in this application, and these modifications, adjustments, or equivalent substitutions should all be considered within the scope of protection of this application.

[0025] Example 1: Please see Figure 1 This embodiment discloses a method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes, including: S1: Acquire road 3D point cloud data, road alignment data, mileage data, and road entity data. Associate the road alignment data, mileage data, and road entity data according to the same road entity to obtain road object data.

[0026] In practice, road 3D point cloud data is obtained through vehicle-mounted laser scanning, UAV laser scanning, ground laser scanning, or existing real-world 3D measurement results. Road 3D point cloud data includes point coordinates. Point coordinates are used to characterize the position of point cloud points along the road in the engineering coordinate system. After the road 3D point cloud data is collected, the point coordinates of the road 3D point cloud data are converted to the engineering coordinate system used by the road alignment data according to the translation, rotation, and scale conversion content recorded in the surveying results, so that the mileage point positions in the road 3D point cloud data, road alignment data, and mileage data all use the same engineering coordinate system.

[0027] Road alignment data is extracted from route design documents, as-built survey documents, road management data, or construction survey results. Road alignment data includes the road centerline and road coverage boundary. The road centerline is used to indicate the extension position of the road entity along the mileage direction. The road coverage boundary is used to indicate the coverage range of the road entity in the plane position. Road alignment data serves as the data source for the division of road entity occupancy in S2, and as the data source for the delineation of cross-section cutting range and longitudinal section extraction range in S3.

[0028] Mileage data is extracted from stationing files, route mileage tables, construction survey results, or as-built survey results. Mileage data includes the starting mileage, ending mileage, mileage direction, and mileage point location corresponding to the road entity. Mileage data is used to represent the correspondence between the road alignment location and the road engineering mileage within the same road entity. When main roads, auxiliary roads, ramp roads, bridge deck roads, under-bridge roads, or underpass roads each have independent mileage systems, the mileage data corresponding to each road entity is retained separately, so that the mileage data of different road entities are retained under the corresponding road entity.

[0029] Road entity data is extracted from road design documents, road management ledgers, as-built data, or road structure classification results. Road entity data includes road entity name, road entity category, and road entity hierarchy information. Road entity name is used to distinguish different road objects. Road entity category is used to distinguish main roads, auxiliary roads, ramp roads, bridge deck roads, under-bridge roads, and underpass roads. Road entity hierarchy information is used to represent the spatial hierarchy of road entities in upper and lower road scenarios. Road entity hierarchy information includes upper-level roads, lower-level roads, or roads on the same level. Road entity data serves as the data source for road entity occupancy division in S2.

[0030] When associating road alignment data, mileage data, and road entity data by the same road entity, the system reads the road entity name and road entity category from the road entity data, and groups the road alignment data with the same road entity name into the same road entity; it reads the mileage direction and mileage point location from the mileage data, and associates the mileage point location with the road alignment data in the same road entity; it then merges and organizes the road entity category and road entity hierarchy information with the road alignment data that has been associated with the mileage to obtain the road object data.

[0031] Road object data includes road entity name, road entity category, road entity hierarchy information, road alignment data, and mileage data; road object data is used to express the ownership information of different road entities at the engineering object level; road object data serves as input data for road entity occupancy division in S2 and as source data for extracting road object data to be segmented in S3.

[0032] S2 utilizes a multi-scale grid index to divide road object data and road 3D point cloud data into road entity occupancy data, thus obtaining road entity occupancy data.

[0033] In specific implementation, a multi-scale grid index is used to spatially organize the road 3D point cloud data; the multi-scale grid index is used to divide the road 3D point cloud data into grid areas according to spatial location; using a multi-scale grid index for point cloud spatial organization is an existing technology in point cloud data processing, and this embodiment does not limit the specific construction algorithm of the multi-scale grid index.

[0034] After completing the spatial organization of the road 3D point cloud data, the coordinates of each point in the road 3D point cloud data are read; according to the grid area in the multi-scale grid index, point cloud points whose coordinates are located in the same grid area are grouped into the same grid point cloud set; the grid point cloud set is used to represent the distribution of road 3D point cloud data in the corresponding grid area.

[0035] Read the road entity name, road entity category, road entity hierarchy information, road alignment data, and mileage data from the road object data; extract the road coverage boundary from the road alignment data, and limit the effective mileage range corresponding to the road coverage boundary according to the mileage data; organize the road coverage boundary and road entity hierarchy information within the effective mileage range to obtain the road entity occupancy range.

[0036] When dividing the road entity occupancy range and grid regions in the multi-scale grid index, grid regions that spatially overlap with the road entity occupancy range are extracted. The spatially overlapping grid regions are then organized with the corresponding road entity name, road entity category, road entity level information, and point cloud set within the grid to obtain the road entity occupancy status corresponding to each grid region.

[0037] The road entity occupancy status is used to indicate whether the same grid area is occupied by one road entity or jointly occupied by two or more road entities. When the same grid area corresponds to one road entity, the same grid area is recorded as single road entity occupancy status. When the same grid area corresponds to two or more road entities, the same grid area is recorded as multi-road entity occupancy status. Multi-road entity occupancy status only indicates that multiple road entities jointly occupy the same grid area, and is not the final result of the road 3D point cloud data.

[0038] Road entity occupancy data includes grid region, point cloud set within the grid, road entity name, road entity category, road entity hierarchy information, road entity occupancy range, and road entity occupancy status. Road entity occupancy data is used to represent the occupancy correspondence between grid region in multi-scale grid index, road 3D point cloud data, and road entity. Road entity occupancy data serves as input data for overlapping occupancy identification of cross-sectional section strips and longitudinal section extraction strips in S4.

[0039] S3, extract the road object data to be segmented from the road object data, define the cross-sectional cutting range and the longitudinal section extraction range of the road object data to be segmented, and obtain the cross-sectional cutting zone and the longitudinal section extraction zone.

[0040] In practice, the road entity name corresponding to the road to be segmented is read, and records with the same road entity name as the road to be segmented are retrieved from the road object data. The road entity name, road entity category, road entity level information, road alignment data and mileage data are extracted from the retrieved records to obtain the road object data to be segmented.

[0041] The road object data to be segmented is used to define the road entities that need to be segmented by cross section and longitudinal section; the road object data to be segmented serves as the data source for defining the cross section cutting range and the longitudinal section extraction range.

[0042] When defining the cross-sectional cutting range of the road object data to be segmented, the mileage data and road alignment data of the road object data to be segmented are read; the mileage point positions are extracted from the mileage data, and the road centerline is extracted from the road alignment data; the mileage point positions are matched with the road centerline to obtain the center position of the cross section; based on the extension direction of the road centerline at the center position of the cross section, the transverse direction perpendicular to the extension direction of the road centerline is defined as the cross section unfolding direction; according to the cross section cutting width and cross section cutting thickness, the range of the center position of the cross section and the cross section unfolding direction is defined to obtain the cross section cutting zone.

[0043] The cross-sectional cutting width is read from the road coverage boundary in the cross-sectional survey task book, the road design cross-sectional range, or the data of the road object to be segmented; if the cross-sectional cutting width is not recorded in the cross-sectional survey task book or the road design cross-sectional range, the coverage area of ​​the road coverage boundary in the cross-sectional unfolding direction is used as the cross-sectional cutting width; the cross-sectional cutting thickness is read from the cutting thickness requirement in the cross-sectional survey task book; if the cross-sectional cutting thickness is not recorded in the cross-sectional survey task book, the maximum value of the distance between adjacent points in the mileage direction of the road 3D point cloud data and the allowable mileage deviation of the cross-sectional results are read, and the maximum value of the maximum distance between adjacent points and the allowable mileage deviation of the cross-sectional results is used as the cross-sectional cutting thickness.

[0044] The cross-sectional section includes the center position of the cross-section, the direction of cross-section development, the cross-section cutting width, the cross-section cutting thickness, and the corresponding mileage of the cross-section; the cross-sectional section is used to define the range of lateral point cloud extraction of the road to be segmented at a specified mileage position; the cross-sectional section serves as input data for overlapping occupancy identification in S4.

[0045] When defining the longitudinal profile extraction range of the road object data to be segmented, the mileage data and road alignment data of the road object data to be segmented are read; the starting mileage, ending mileage and mileage direction are extracted from the mileage data, and the road centerline is extracted from the road alignment data; the longitudinal range along the extension direction of the road to be segmented is defined according to the starting mileage, ending mileage, mileage direction and road centerline; the longitudinal range is expanded according to the longitudinal profile extraction width to obtain the longitudinal profile extraction zone.

[0046] The longitudinal profile extraction width is obtained from the horizontal range of elevation points recorded in the longitudinal profile surveying task book, or from the road coverage boundary in the road object data to be segmented; if the longitudinal profile surveying task book does not record the longitudinal profile extraction width, the range of points spread out on both sides of the road centerline within the road coverage boundary is taken as the longitudinal profile extraction width; the longitudinal profile extraction zone includes the longitudinal profile mileage range, the longitudinal profile centerline, the longitudinal profile extraction width, and the longitudinal profile mileage direction.

[0047] The longitudinal profile extraction zone is used to define the range of elevation point cloud extraction along the mileage direction of the road to be segmented; the longitudinal profile extraction zone serves as input data for overlapping occupancy identification in S4; the cross section cutting zone and the longitudinal profile extraction zone together define the cross section processing range of the road to be segmented.

[0048] S4. Combining road entity occupancy data, overlapping occupancy is identified in the cross-sectional section and longitudinal section extraction zones to obtain overlapping occupancy areas. The occupancy status of road entities within the overlapping occupancy areas is then analyzed to obtain cross-section ownership conflict data.

[0049] It should be noted that in the existing process of segmenting the longitudinal and transverse sections of highways, the 3D point cloud data of roads located within the section cutting range is usually extracted as candidate point clouds for the current road section. In scenarios such as interchanges, parallel main and auxiliary roads, overlapping planar projections of bridge roads and roads under bridges, and ramps connecting to the main road, the 3D point cloud data of other road entities may also enter the cross section cutting zone or longitudinal section extraction zone of the road to be segmented. If point cloud extraction is performed solely based on whether the 3D point cloud data of the road is within the section range, it is easy to mistakenly include the 3D point cloud data of other road entities into the road section to be segmented. Therefore, this embodiment does not directly use the geometric region obtained solely by the intersection of the road coverage boundary plane as the basis for subsequent point cloud attribution processing. Instead, within the cross section cutting zone and longitudinal section extraction zone, it identifies the area jointly occupied by multiple road entities by combining road entity occupancy data, and then converts the area jointly occupied by multiple road entities into section attribution conflict data for subsequent point cloud attribution boundary division.

[0050] In practice, the cross-sectional section, longitudinal section extraction, and road entity occupancy data are read. Within the grid areas covered by the cross-sectional section and the longitudinal section extraction, grid areas with multiple road entity occupancy states and corresponding road entities including the road to be segmented are extracted to obtain candidate overlapping grid areas. Candidate overlapping grid areas are used to represent grid areas located within the cross-sectional processing range of the road to be segmented and simultaneously occupied by the road to be segmented and other road entities. Other road entities are road entities in the road entity occupancy data other than the road to be segmented and corresponding to the candidate overlapping grid areas.

[0051] Candidate overlapping mesh regions that are within the same cross-sectional section or longitudinal section extraction zone, share boundaries, and correspond to identical road entities are merged to obtain overlapping occupied regions. The overlapping occupied region is not merely a geometric region obtained by the intersection of road coverage boundary planes, but rather a mesh merging region limited to the processing range of the road section to be segmented, corresponding to the road to be segmented and other road entities. The overlapping occupied region is used to represent the spatial location of the 3D point cloud data of other road entities in the current cross-sectional section or longitudinal section extraction zone that enters the candidate range of the road section to be segmented.

[0052] When further analyzing the occupancy status of multiple road entities within the overlapping area, the following steps are taken: read the road entity hierarchy information, road centerline, and road coverage boundary of the road to be segmented and other road entities corresponding to the overlapping area; extract the road centerline segment located within the overlapping area to obtain the local road centerline segment corresponding to the road to be segmented and the local road centerline segment corresponding to other road entities; analyze the differences in road entity hierarchy between the road to be segmented and other road entities, the intersection of local road centerline segments, the minimum included angle of local road centerline segments, and the access status of road coverage boundaries to identify the cross-section ownership conflict type and obtain cross-section ownership conflict data.

[0053] In one possible implementation, the cross-section attribution conflict types include upper-lower layer conflict, parallel conflict, and access conflict. When the road entity hierarchy information of the road to be divided is recorded as an upper-level road and the road entity hierarchy information of other road entities is recorded as a lower-level road, or the road entity hierarchy information of the road to be divided is recorded as a lower-level road and the road entity hierarchy information of other road entities is recorded as an upper-level road, and the road to be divided and other road entities jointly correspond to the same overlapping occupied area, the cross-section attribution conflict type corresponding to the overlapping occupied area is recorded as an upper-lower layer conflict. When the road entity hierarchy information of the road to be divided and other road entities is the same, the centerline segment of the local road corresponding to the road to be divided does not intersect with the centerline segment of the local road corresponding to other road entities, and the minimum included angle between the centerline segments of the local roads is not greater than a preset parallel included angle threshold, the cross-section attribution conflict type corresponding to the overlapping occupied area is recorded as a parallel conflict. When the road centerline of other road entities enters the road coverage boundary of the road to be divided and connects with the road centerline of the road to be divided within the overlapping occupied area, the cross-section attribution conflict type corresponding to the overlapping occupied area is recorded as an access conflict.

[0054] A preset parallel angle threshold is used to distinguish whether the road to be segmented is in a parallel extension state with other road entities. The preset parallel angle threshold is preferentially read from the parallel layout direction deviation range recorded in the road design document. When the road design document does not record the parallel layout direction deviation range but records the parallel layout relationship, road segments with the same road entity category, the same road entity level information, non-intersecting local road centerline segments, and marked as parallel layout in the road design document are extracted from the road object data and treated as parallel road segments of the same type. When the road design document neither records the parallel layout direction deviation range nor the parallel layout relationship, the threshold is determined from the road management data. Existing road segments with the same road entity category, the same road entity level information, non-intersecting local road centerline segments, and recorded parallel layout relationships are extracted from the data and regarded as parallel road segments of the same type. When parallel road segments of the same type can be extracted, the minimum included angle of the local road centerline segments of the parallel road segments of the same type is counted, and the minimum included angle of the local road centerline segments is sorted from smallest to largest. The maximum value after sorting is used as the preset parallel included angle threshold. When parallel road segments of the same type cannot be extracted, the specified value of the allowable directional deviation of parallel roads in the road design specifications or similar engineering technical standards is read and the specified value is used as the preset parallel included angle threshold.

[0055] The cross-section attribution conflict data includes overlapping occupied areas, roads to be segmented, other road entities, cross-section cutting strips or longitudinal section extraction strips corresponding to overlapping occupied areas, and cross-section attribution conflict types corresponding to overlapping occupied areas; the cross-section attribution conflict data serves as the input data for dividing the point cloud attribution boundaries of overlapping occupied areas in S5.

[0056] In this embodiment, the road entity occupancy status is further constrained to the cross-sectional section and longitudinal section extraction zone by using road entity occupancy data. This enables the identification of overlapping occupancy areas shared by multiple road entities within the cross-sectional processing range of the road to be segmented. By further analyzing the occupancy status of multiple road entities within the overlapping occupancy area, the results of multiple road entities sharing occupancy within the cross-sectional processing range can be converted into cross-sectional attribution conflict data. This allows subsequent processing to no longer rely solely on whether the road 3D point cloud data is within the cross-sectional range for selection, but rather to delineate point cloud attribution boundaries based on the risk of misattribution of road 3D point cloud data corresponding to other road entities.

[0057] S5, combining the cross-section ownership conflict data, performs point cloud ownership boundary division on the overlapping occupied areas to obtain cross-section ownership boundary data.

[0058] It should be noted that the overlapping area is used to represent the area shared by multiple road entities within the cross-sectional section or longitudinal section extraction zone of the road to be segmented; the overlapping area cannot directly indicate the ownership boundary of the road's 3D point cloud data; therefore, this embodiment combines the cross-section ownership conflict data to divide the point cloud ownership boundary between the road to be segmented and other road entities within the overlapping area to obtain the cross-section ownership boundary data.

[0059] In specific implementation, the overlapping occupied area, the road to be segmented, other road entities, the cross-sectional cutting strip or longitudinal section extraction strip corresponding to the overlapping occupied area, and the cross-sectional ownership conflict type corresponding to the overlapping occupied area are read from the cross-sectional ownership conflict data; the point cloud set in the grid corresponding to the overlapping occupied area, the road entity name, road entity category, road entity level information, road entity occupation range, and road entity occupation status are read from the road entity occupation data; the point cloud set in the grid within the overlapping occupied area is used as the point cloud ownership boundary division object; when the cross-sectional ownership conflict data includes two or more other road entities, point cloud ownership boundary division is performed on each other road entity, and corresponding point cloud ownership boundaries are formed respectively.

[0060] When the cross-section belonging conflict type is upper and lower layer conflict, read the road entity level information of the road to be divided and other road entities; determine the grid area that shares the boundary with the overlapping occupied area and is located in the same cross-section cutting zone or the same longitudinal section extraction zone as the elevation reference range of the overlapping occupied area; if no single road entity occupancy status grid area corresponding to the road to be divided is extracted within the elevation reference range of the overlapping occupied area, or if no single road entity occupancy status grid area corresponding to other road entities is extracted, calculate the shortest plane distance between the center position of each single road entity occupancy status grid area and the boundary of the overlapping occupied area within the same cross-section cutting zone or the same longitudinal section extraction zone, and supplement one or more corresponding single road entity occupancy status grid areas with the smallest shortest plane distance to the elevation reference range of the overlapping occupied area.

[0061] From the road entity occupancy data, extract grid regions located within the elevation reference range of the overlapping occupancy area, with a single road entity occupancy status and a road entity name consistent with the road entity name of the road to be segmented. Merge and statistically analyze the point cloud sets within each corresponding grid region to obtain the highest and lowest elevations, and use the interval between the highest and lowest elevations as the elevation range of the road to be segmented. From the road entity occupancy data, extract grid regions located within the elevation reference range of the overlapping occupancy area, with a single road entity occupancy status and a road entity name consistent with the road entity name of other road entities. Merge and statistically analyze the point cloud sets within each corresponding grid region to obtain the highest and lowest elevations, and use the interval between the highest and lowest elevations as the elevation range of other road entities.

[0062] Extract the elevation distribution of the point cloud set within the grid in the overlapping area; sort the point cloud set within the grid according to the elevation value of the point cloud points, and take the position where the elevation interval between adjacent point cloud points is greater than or equal to the preset inter-layer interval threshold as the elevation break position; divide the point cloud set within the grid into at least two continuous elevation segments according to the elevation break position.

[0063] In handling conflicts between upper and lower levels, if the classification of continuous elevation segments is based solely on the degree of overlap in elevation ranges, the classification of continuous elevation segments may still be unstable if both continuous elevation segments partially overlap with the same road entity elevation range. Therefore, an elevation classification determination factor is calculated for continuous elevation segments that intersect with the road entity elevation range. This factor is used to combine the degree of overlap in elevation ranges, the proximity of elevation centers, and the consistency of road entity levels into a single classification criterion.

[0064] Let T be the road to be segmented, O be the other road entities, and the j-th elevation continuous segment in the overlapping area be the j-th elevation continuous segment. For any road entity among the road to be segmented or other road entities, let X be the road entity currently being assigned a location, where X is either T or O.

[0065] When the elevation range of the j-th continuous elevation segment intersects with that of road entity X, calculate the elevation attribution determination of the j-th continuous elevation segment relative to road entity X: ; In the formula, The elevation range of road entity X. Let j be the elevation range of the j-th continuous elevation segment. The elevation center value of road entity X is the elevation range center value. Let j be the center value of the elevation range of the j-th continuous elevation segment. The elevation range and width of road entity X. Let j be the width of the elevation range of the j-th continuous elevation segment. For items with consistent levels; the center value of the elevation range is obtained by averaging the highest and lowest elevations of the corresponding elevation range, and the width of the elevation range is obtained by the difference between the highest and lowest elevations of the corresponding elevation range.

[0066] The hierarchical consistency term is determined based on the road entity hierarchy information and the elevation direction of the elevation continuum; the average of the center value of the elevation range of the road to be divided and the center values ​​of the elevation ranges of other road entities is used as the elevation boundary value; when road entity X is recorded as the upper-level road, and the center value of the elevation range of the j-th elevation continuum is greater than the elevation boundary value, Choose one; when road entity X is recorded as a lower-level road, and the center value of the elevation range of the j-th continuous elevation segment is less than the elevation boundary value, Choose one; in other cases, Take zero.

[0067] When the j-th elevation continuous segment does not intersect with the elevation range of road entity X, the j-th elevation continuous segment is not considered as a candidate elevation continuous segment of road entity X.

[0068] The candidate elevation continuous segment with the largest elevation attribution determination value relative to the road to be segmented is designated as the area of ​​the road to be segmented; the candidate elevation continuous segment with the largest elevation attribution determination value relative to other road entities is designated as the area of ​​other road entities; when two or more candidate elevation continuous segments have the same elevation attribution determination value, if the corresponding road entity level information record is an upper-level road, then the candidate elevation continuous segment with the largest minimum elevation is selected; if the corresponding road entity level information record is a lower-level road, then the candidate elevation continuous segment with the smallest maximum elevation is selected; the elevation boundary range corresponding to the elevation break point between the area of ​​the road to be segmented and the area of ​​other road entities is used as the point cloud attribution boundary.

[0069] The preset inter-layer interval threshold is used to distinguish the separation positions of upper and lower road layers in the elevation direction. The preset inter-layer interval threshold is read in the order of road design documents, bridge clearance data, and underpass as-built data. When the inter-layer interval is not recorded in the road design documents, bridge clearance data, and underpass as-built data, the point cloud elevation interval is statistically analyzed for areas in the same engineering coordinate system where the road entity layer information is recorded as upper road and lower road respectively, and where there is a common coverage area within the plane projection range. During the statistics, the lowest elevation of the upper road point cloud and the highest elevation of the lower road point cloud are extracted, and the difference between the lowest elevation of the upper road point cloud and the highest elevation of the lower road point cloud is used as the point cloud elevation interval. The point cloud elevation intervals are sorted from smallest to largest, and the minimum value after sorting is used as the preset inter-layer interval threshold.

[0070] When the cross-section attribution conflict type is parallel conflict, the following steps are taken: read the local road centerline segment corresponding to the road to be segmented, the local road centerline segments corresponding to other road entities, the road coverage boundary of the road to be segmented, and the road coverage boundary of other road entities; along the lateral direction perpendicular to the local road centerline segment corresponding to the road to be segmented, extract the boundary segment on the side closer to the road centerline of other road entities from the road coverage boundary of the road to be segmented, and extract the boundary segment on the side closer to the road centerline of the road to be segmented from the road coverage boundary of other road entities; when there is a gap between two boundary segments, the centerline position of the gap area is taken as the point cloud attribution boundary; when two boundary segments are connected, the connected boundary position is taken as the point cloud attribution boundary; when there is a partial overlap between two boundary segments, the centerline position of the partially overlapping part along the lateral direction is taken as the point cloud attribution boundary; the side of the point cloud attribution boundary closer to the road centerline of the road to be segmented is defined as the side range of the road to be segmented; the side of the point cloud attribution boundary closer to the road centerline of other road entities is defined as the side range of other road entities.

[0071] When the cross-section conflict type is access conflict, the entry position is obtained by reading the location where the road centerline of other road entities enters the road coverage boundary of the road to be segmented; the access position is obtained by reading the location where the road centerline of other road entities intersects with the road centerline of the road to be segmented; within the mileage range defined by the cross-section cutting zone or longitudinal section extraction zone corresponding to the overlapping area, the area between the entry position and the access position is taken as the access transition range; within the access transition range, the grid center position of each grid area is read, and the first plane distance from the grid center position to the road centerline of the road to be segmented and the second plane distance from the grid center position to the road centerline of other road entities are calculated respectively; the first plane distance is reduced by... The grid area at the second plane distance is defined as the road side range to be segmented; the grid area at the second plane distance less than the first plane distance is defined as the other road entity side range; the boundary positions of the road side range to be segmented and the other road entity side ranges adjacent to each other, as well as the grid areas where the first plane distance and the second plane distance are the same or where the absolute value of the difference between the first plane distance and the second plane distance is within the distance calculation accuracy range, are collectively organized as the point cloud belonging boundary; the distance calculation accuracy range is determined based on the plane coordinate accuracy of the road 3D point cloud data, specifically the plane coordinate accuracy value of the road 3D point cloud data is used as the distance calculation accuracy range; the plane coordinate accuracy value of the road 3D point cloud data is read from the accuracy record of the surveying results.

[0072] The cross-section attribution boundary data includes overlapping occupied areas, cross-section attribution conflict types, point cloud attribution boundaries, the range of the road to be segmented, the range of other road entities, and the cross-section cutting strip or longitudinal section extraction strip corresponding to the overlapping occupied areas; the cross-section attribution boundary data is used to represent the point cloud attribution boundary content between the road to be segmented and other road entities within the overlapping occupied areas; the cross-section attribution boundary data serves as the input data for collecting attribution point clouds in S6.

[0073] In this embodiment, elevation demarcation, lateral adjacency demarcation, and access demarcation are performed on the overlapping occupied area through upper and lower layer conflicts, parallel conflicts, and access conflicts in the cross-section attribution conflict data, respectively. In the upper and lower layer conflict processing, the elevation attribution determination quantity combines the degree of overlap of elevation ranges, the degree of proximity of elevation centers, and the degree of consistency of road entity levels into the same attribution criterion, so that the candidate elevation continuous segments in the overlapping occupied area can be divided according to the attribution characteristics of the corresponding road entities. The road three-dimensional point cloud data in the overlapping occupied area can thus be divided into the road side range to be segmented and other road entity side ranges, so that the formation of the attribution point cloud in S6 is further limited by the cross-section attribution boundary data.

[0074] S6, combining the cross-section boundary data, collect the point clouds belonging to the road to be segmented from the road 3D point cloud data located within the cross-section cutting zone and longitudinal section extraction zone, obtain the belonging point cloud, and perform cross-section segmentation and longitudinal section segmentation on the belonging point cloud to obtain the cross-section result.

[0075] It should be noted that the cross-section ownership boundary data has already formed the scope of the road to be segmented and the scope of other road entities within the overlapping occupied area; therefore, when collecting point clouds belonging to the road to be segmented, cross-section cutting strips, longitudinal section extraction strips, road entity occupancy data and cross-section ownership boundary data are used simultaneously.

[0076] In practice, the following steps are taken: cross-sectional cutting zone, longitudinal section extraction zone, road 3D point cloud data, road entity occupancy data, and cross-section boundary data are read; point cloud points located within the cross-sectional cutting zone are extracted from the road 3D point cloud data to form a cross-sectional candidate point cloud set; point cloud points located within the longitudinal section extraction zone are extracted from the road 3D point cloud data to form a longitudinal section candidate point cloud set; the cross-sectional candidate point cloud set and the longitudinal section candidate point cloud set are used together as the point cloud aggregation objects.

[0077] When aggregating the candidate point cloud set for cross sections, the overlapping occupied area, point cloud ownership boundary, road-side range to be segmented, and other road entity-side ranges corresponding to the cross section boundary data are read from the cross section ownership boundary data. For point cloud points in the candidate point cloud set for cross sections that are not located within the overlapping occupied area, the grid area where the point cloud point is located is located. When the grid area where the point cloud point is located is in a single road entity occupation state in the road entity occupation data, and the road entity name is consistent with the road entity name of the road to be segmented, the point cloud point is included in the cross section ownership point cloud. For point cloud points in the candidate point cloud set for cross sections located within the overlapping occupied area, only point cloud points located within the road-side range to be segmented are included in the cross section ownership point cloud, and point cloud points not located within the road-side range to be segmented are not included in the cross section ownership point cloud.

[0078] When collecting the candidate point cloud set of longitudinal profiles, the overlapping occupied area, point cloud ownership boundary, road-side range to be segmented, and other road entity-side range corresponding to the longitudinal profile extraction zone are read from the cross-section ownership boundary data. For point cloud points in the candidate point cloud set of longitudinal profiles that are not located in the overlapping occupied area, the grid area where the point cloud point is located is located. When the grid area where the point cloud point is located is in a single road entity occupation state in the road entity occupation data, and the road entity name is consistent with the road entity name of the road to be segmented, the point cloud point is included in the longitudinal profile ownership point cloud. For point cloud points in the candidate point cloud set of longitudinal profiles that are located in the overlapping occupied area, only the point cloud points located within the road-side range to be segmented are included in the longitudinal profile ownership point cloud, and the point cloud points not located within the road-side range to be segmented are not included in the longitudinal profile ownership point cloud.

[0079] When the same overlapping area corresponds to two or more other road entities, point cloud attribution is determined for the cross-section attribution boundary data corresponding to each other road entity; point cloud points that are simultaneously located within the range of the road to be segmented corresponding to each point cloud attribution boundary are assigned to the cross-section attribution point cloud or the longitudinal section attribution point cloud; point cloud points located within the range of any other road entity are excluded from the cross-section attribution point cloud or the longitudinal section attribution point cloud.

[0080] The cross-sectional and longitudinal section point clouds are respectively used as components of the assigned point cloud to obtain the assigned point cloud. The assigned point cloud is used to represent the 3D point cloud data of the road that belongs to the road to be segmented after being jointly defined by the road entity occupancy data and the cross-sectional boundary data. The assigned point cloud does not include the 3D point cloud data of the road that is included in the scope of other road entities in the cross-sectional boundary data. For point cloud points that are not located in the overlapping occupancy area, if the grid area where the point cloud point is located is not in a single road entity occupancy state, or if the road entity name is inconsistent with the road entity name of the road to be segmented, the point cloud point is not included in the assigned point cloud.

[0081] When performing cross-sectional segmentation on the attributed point cloud, the attributed point cloud of the cross section is classified by mileage according to the corresponding mileage of the cross section in the cross section cutting zone; the attributed point clouds of the cross section under the corresponding mileage of the same cross section are sorted according to the cross section unfolding direction to form the cross section data of the road to be segmented; the cross section data is used to represent the transverse cross section shape of the road to be segmented at the corresponding mileage position.

[0082] When performing longitudinal profile segmentation on the assigned point cloud, the longitudinal profile assigned point cloud is sorted by mileage according to the longitudinal profile mileage direction and longitudinal profile mileage range in the extracted longitudinal profile zone; elevation points in the longitudinal profile assigned point cloud are extracted according to the longitudinal profile mileage direction to form longitudinal profile data of the road to be segmented; the longitudinal profile data is used to represent the elevation changes of the road to be segmented along the mileage direction.

[0083] The cross-sectional results include cross-sectional data and longitudinal cross-sectional data; the cross-sectional results are used to represent the longitudinal and cross-sectional segmentation results of the road to be segmented based on the point cloud.

[0084] In this embodiment, the assigned point cloud is jointly defined by the cross-sectional section strip, the longitudinal section extraction strip, the road entity occupancy data, and the cross-sectional assignment boundary data. Point cloud points not located within the overlapping occupancy area need to be confirmed by the road entity occupancy data to be in a single road entity occupancy state in the grid area where the point cloud point is located, and the road entity name must be consistent with the road entity name of the road to be segmented. Point cloud points located within the overlapping occupancy area need to be located within the range of the road to be segmented in the cross-sectional assignment boundary data. This allows the 3D point cloud data of roads that are within the cross-sectional processing range but belong to other road entities to be excluded from the assigned point cloud, so that the cross-sectional and longitudinal section data of the road to be segmented are jointly defined by the cross-sectional processing range and the cross-sectional assignment boundary data.

[0085] Example 2: Please see Figure 2 This embodiment provides a highway longitudinal and transverse section segmentation system that integrates multi-scale grid indexes, used to implement the aforementioned highway longitudinal and transverse section segmentation method that integrates multi-scale grid indexes. The system includes: The road object generation module is used to acquire road 3D point cloud data, road alignment data, mileage data and road entity data, and associate the road alignment data, mileage data and road entity data according to the same road entity to obtain road object data. The entity occupancy partitioning module is used to partition road entity occupancy data and road 3D point cloud data using a multi-scale grid index, thereby obtaining road entity occupancy data. The cross-section range generation module is used to extract road object data to be segmented from road object data, and to define the cross-section cutting range and longitudinal section extraction range of the road object data to be segmented, so as to obtain the cross-section cutting zone and the longitudinal section extraction zone. The attribution conflict generation module is used to combine road entity occupancy data to identify overlapping occupancy of cross-sectional cutting strips and longitudinal section extraction strips, obtain overlapping occupancy areas, and analyze the road entity occupancy status within the overlapping occupancy areas to obtain cross-sectional attribution conflict data. The attribution boundary generation module is used to combine cross-section attribution conflict data to divide the point cloud attribution boundaries of overlapping occupied areas, thereby obtaining cross-section attribution boundary data. The cross-section result generation module is used to combine the cross-section boundary data to collect the point clouds belonging to the road to be segmented from the road 3D point cloud data located within the cross-section cutting zone and the longitudinal section extraction zone, obtain the belonging point cloud, and perform cross-section segmentation and longitudinal section segmentation on the belonging point cloud to obtain the cross-section result.

[0086] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0087] Finally: The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for segmenting longitudinal and transverse sections of highways by integrating multi-scale grid indexing, characterized in that, include: Acquire road 3D point cloud data, road alignment data, mileage data, and road entity data; associate the road alignment data, mileage data, and road entity data according to the same road entity to obtain road object data; Using a multi-scale grid index, road entity occupancy is divided into road object data and road 3D point cloud data to obtain road entity occupancy data; Extract road object data to be segmented from road object data, define the cross-sectional cutting range and longitudinal section extraction range of the road object data to be segmented, and obtain the cross-sectional cutting zone and longitudinal section extraction zone. By combining road entity occupancy data, overlapping occupancy is identified in cross-sectional section strips and longitudinal section extraction strips to obtain overlapping occupancy areas. The occupancy status of road entities within the overlapping occupancy areas is then analyzed to obtain cross-section ownership conflict data. By combining the cross-section ownership conflict data, the point cloud ownership boundary of the overlapping occupied area is divided to obtain the cross-section ownership boundary data. By combining the boundary data of the cross-section, the point clouds belonging to the road to be segmented are collected from the three-dimensional point cloud data of the road located within the cross-section cutting zone and the longitudinal section extraction zone, and the belonging point clouds are obtained. The cross-section and longitudinal section segments are then performed on the belonging point clouds to obtain the cross-section results.

2. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 1, characterized in that, The method for obtaining road entity occupancy data includes: The point cloud points in the road 3D point cloud data are assigned to the corresponding grid regions according to the multi-scale grid index to obtain the point cloud set within each grid region. Extract the road coverage boundary, mileage data and road entity information of each road entity from the road object data. Determine the effective mileage range according to the mileage data and truncate the road coverage boundary. Associate the truncated road coverage boundary with the road entity information to form the road entity occupancy range. The planar projection range of each grid area is compared with the occupancy range of each road entity. If the same grid area intersects with only one road entity occupancy range, the road entity occupancy status is marked as single road entity occupancy status. If the same grid area intersects with the occupied areas of two or more road entities, the road entity occupancy status is marked as multi-road entity occupancy status. By organizing the grid area, the point cloud set within the grid, the road entity occupancy range, and the road entity occupancy status, road entity occupancy data is obtained.

3. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 2, characterized in that, The method for obtaining the overlapping occupied area includes: Based on road entity occupancy data, the occupancy status of the grid areas covered by the cross section cutting zone and the longitudinal section extraction zone is filtered. The grid areas where the road entity occupancy status is multiple road entity occupancy status and the corresponding road entities include the road to be segmented and other road entities are extracted to obtain candidate overlapping grid areas. Based on the cross-sectional section or longitudinal section extraction zone, the road to be segmented, and other road entities to which the candidate overlapping grid regions belong, the candidate overlapping grid regions are merged. Candidate overlapping grid regions that share boundaries and belong to the same cross-sectional section or longitudinal section extraction zone, the road to be segmented, and other road entities are merged into overlapping occupied regions.

4. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 3, characterized in that, The method for analyzing the occupancy status of road entities within overlapping occupancy areas to obtain cross-section attribution conflict data includes: Extract the road entity hierarchy information of the road to be segmented and other road entities corresponding to the overlapping occupied area, extract the local road centerline segment in the overlapping occupied area, and identify the access status of other road entities relative to the road coverage boundary of the road to be segmented. Based on the road entity hierarchy information, the intersection of local road centerline segments, the minimum included angle of local road centerline segments, and the access of road coverage boundaries, the cross-section ownership conflict type of the overlapping occupied area is identified. The overlapping occupied areas, roads to be divided, other road entities, and cross-section ownership conflict types are organized into cross-section ownership conflict data. The cross-section ownership conflict types include upper and lower layer conflicts, parallel conflicts, or access conflicts.

5. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 4, characterized in that, The method for obtaining the cross-section boundary data includes: When the cross-section ownership conflict type is upper and lower layer conflict, the point cloud ownership boundary is determined according to the elevation distribution in the overlapping area. When the cross-section assignment conflict type is parallel conflict, the point cloud assignment boundary is determined according to the relative state of the boundary segments between the road to be segmented and other road entities. When the cross-section ownership conflict type is access conflict, the point cloud ownership boundary is determined according to the distance relationship between the centerline of the road to be segmented and other road entities. Based on the point cloud attribution boundary, the scope of the road to be segmented and the scope of other road entities are delineated, and the overlapping occupied areas, cross-section attribution conflict types, point cloud attribution boundaries, scope of the road to be segmented and the scope of other road entities are organized to obtain cross-section attribution boundary data.

6. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 5, characterized in that, When the cross-section attribution conflict type is an upper-lower layer conflict, the method for determining the point cloud attribution boundary based on the elevation distribution within the overlapping area includes: Within the same cross-sectional section or the same longitudinal section extraction zone, filter out the grid regions that share the boundary with the overlapping occupied region to obtain the elevation reference range of the overlapping occupied region; Based on the elevation reference range of the overlapping occupied area and the road entity occupancy data, the single road entity occupancy status grid area corresponding to the road to be segmented and other road entities is extracted respectively, and the elevation merging statistics of the point cloud set in the grid are performed to obtain the elevation range of the road to be segmented and the elevation range of other road entities. The point cloud set within the grid in the overlapping area is sorted by elevation. The position where the elevation interval between adjacent point cloud points is greater than or equal to the preset inter-layer interval threshold is taken as the elevation break position, and the elevation continuous segment is obtained. For continuous elevation segments that intersect with the elevation range of the road to be segmented and the elevation range of other road entities, an elevation attribution determination quantity is generated based on the degree of overlap of elevation ranges, the proximity of elevation centers, and the consistency of road entity levels. The continuous elevation segments with the largest elevation attribution determination quantity are selected to delineate the road side range to be segmented and the other road entity side ranges. The elevation break position between the two ranges is extracted to obtain the point cloud attribution boundary corresponding to the upper and lower layer conflicts.

7. The method for segmenting highway longitudinal and transverse sections by incorporating multi-scale grid indexes according to claim 5, characterized in that, When the cross-section attribution conflict type is a parallel conflict, the method for determining the point cloud attribution boundary based on the relative state of the boundary segments between the road to be segmented and other road entities includes: Along the lateral direction perpendicular to the local road centerline segment corresponding to the road to be divided, extract the side boundary segment close to the road centerline of other road entities from the road coverage boundary of the road to be divided, and extract the side boundary segment close to the road centerline of the road to be divided from the road coverage boundary of other road entities. If there is a gap between the boundary segment of the road to be segmented that is closer to the center line of another road entity and the boundary segment of the road to be segmented that is closer to the center line of the road to be segmented, then the center line of the gap area is taken as the point cloud ownership boundary; if the two are connected, then the connection boundary is taken as the point cloud ownership boundary; if the two partially overlap, then the center line of the partially overlapping part along the lateral direction is taken as the point cloud ownership boundary, thus obtaining the point cloud ownership boundary corresponding to the parallel conflict.

8. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 5, characterized in that, When the cross-section attribution conflict type is an access conflict, the method for determining the point cloud attribution boundary based on the distance relationship between the centerline of the road to be segmented and other road entities includes: Extract the entry position of the center line of other road entities into the road coverage boundary of the road to be divided, and the access position where the center line of other road entities connects with the center line of the road to be divided. Within the mileage range defined by the cross section or longitudinal section extraction zone corresponding to the overlapping area, the area between the entry position and the access position is taken as the access transition range. Within the access transition range, calculate the first plane distance from the center position of each grid area to the center line of the road to be segmented and the second plane distance to the center line of other road entities, and delineate the side range of the road to be segmented and the side range of other road entities based on the distance comparison; The boundary positions of adjacent ranges on both sides, as well as the locations of grid areas where the distance between the first plane and the second plane is the same or the absolute value of the difference is less than or equal to the distance calculation accuracy range, are organized into the point cloud belonging boundaries corresponding to the access conflict.

9. The method for segmenting highway longitudinal and transverse sections by integrating multi-scale grid indexes according to claim 1, characterized in that, The method for obtaining the attributed point cloud includes: Candidate point cloud sets for cross sections and longitudinal sections are extracted from the cross section cutting zone and the longitudinal section extraction zone, respectively. For point cloud points in the cross-section candidate point cloud set and the longitudinal section candidate point cloud set that are not located in the overlapping occupied area, if the grid area where the point cloud point is located is occupied by a single road entity and the corresponding road entity is a road to be segmented, then it is assigned to the corresponding cross-section or longitudinal section point cloud. For point cloud points located within overlapping occupied areas, only point cloud points located within the road side range to be divided are assigned to the corresponding cross-section or longitudinal section point cloud; when the same overlapping occupied area corresponds to two or more other road entities, only point cloud points located within the road side range to be divided corresponding to the boundary of each point cloud are assigned to the corresponding cross-section or longitudinal section point cloud. Organize the cross-sectional and longitudinal section attribute point clouds to obtain the attribute point cloud.

10. A highway longitudinal and transverse section segmentation system integrating multi-scale grid indexing, used to implement the highway longitudinal and transverse section segmentation method integrating multi-scale grid indexing as described in any one of claims 1-9, characterized in that, include: The road object generation module is used to acquire road 3D point cloud data, road alignment data, mileage data and road entity data, and associate the road alignment data, mileage data and road entity data according to the same road entity to obtain road object data. The entity occupancy partitioning module is used to partition road entity occupancy data and road 3D point cloud data using a multi-scale grid index, thereby obtaining road entity occupancy data. The cross-section range generation module is used to extract road object data to be segmented from road object data, and to define the cross-section cutting range and longitudinal section extraction range of the road object data to be segmented, so as to obtain the cross-section cutting zone and the longitudinal section extraction zone. The attribution conflict generation module is used to combine road entity occupancy data to identify overlapping occupancy of cross-sectional cutting strips and longitudinal section extraction strips, obtain overlapping occupancy areas, and analyze the road entity occupancy status within the overlapping occupancy areas to obtain cross-sectional attribution conflict data. The attribution boundary generation module is used to combine cross-section attribution conflict data to divide the point cloud attribution boundaries of overlapping occupied areas, thereby obtaining cross-section attribution boundary data. The cross-section result generation module is used to combine the cross-section boundary data to collect the point clouds belonging to the road to be segmented from the road 3D point cloud data located within the cross-section cutting zone and the longitudinal section extraction zone, obtain the belonging point cloud, and perform cross-section segmentation and longitudinal section segmentation on the belonging point cloud to obtain the cross-section result.

Images