CN114863033B - A cross-section extraction method based on point cloud digital model - Google Patents

Abstract

The invention discloses a section extraction method based on point cloud digital-analog. The existing method for extracting the cross section by utilizing the laser point cloud is low in automation processing degree and low in efficiency. The invention utilizes point cloud data to construct a PLY-format digital elevation model; interpolating the obtained section node coordinates along the section advancing direction through a designated interval to obtain encrypted node coordinates, setting section extraction options and parameters, and interpolating the encrypted node elevation on the triangular surface of the digital elevation model by using a gravity center coordinate smoothing interpolation algorithm to obtain the section; checking section data by using middle pile data, and thinning section nodes; and finally, carrying out format conversion on the section data, and drawing a section chart. The software to be researched of the invention utilizes point cloud digital-analog to extract the section under CAD, and can finish operations such as section data extraction, checking, thinning, format conversion and the like by one key after necessary data preparation and option parameter setting.

Description

Section extraction method based on point cloud digital-analog

Technical Field

The invention belongs to the technical field of mapping, and particularly relates to a section extraction method based on point cloud digital-analog.

Background

The section data acquisition at the present stage mainly adopts three modes of field manual actual measurement, field aerial photography three-dimensional acquisition and laser radar point cloud extraction. The field manual actual measurement needs to run points at characteristic positions of the sections, so that the working efficiency is low, the labor intensity is high, the safety risk in complex and difficult areas is high, and even the characteristic positions cannot be reached. The aerial photographing three-dimensional acquisition in the interior industry relies on visual measurement of characteristic points of operators on a digital measurement workstation, the plane precision of aerial photographing three-dimensional image pairs cannot be corrected through the middle piles in the exterior industry, and the problems of large workload, low automation degree, poor vegetation coverage area precision, low processing efficiency and the like also exist. The laser radar point cloud extraction is an emerging technology, has the advantages of high efficiency, no contact, easiness in realizing automatic processing and the like, and is widely applied in recent years.

The common laser radar point cloud extraction cross section mode in the railway industry is to utilize LiDAR data processing software to finish point cloud classification, screen out ground points, construct a digital elevation model, then import line center lines and cross section lines, manually point and select cross sections, input information such as section names, left-right offset distances, section node step sizes and the like, and output section files pile by pile. And when the vertical section is extracted, corresponding middle pile elevation is interpolated based on the LiDAR point cloud ground points according to the middle pile point coordinates provided by the design specialty, so that a three-dimensional coordinate point set of the vertical section is obtained. The method can extract the transverse and longitudinal sections, but does not consider the problem that the extracted sections are inconsistent with the middle piles, and because the commonly used LiDAR data processing software is general software, an open programming interface is not available, the method cannot be deeply fused with the extraction business of the line engineering section, and has low automation degree and low efficiency.

Disclosure of Invention

In order to make up for the defects of the prior art, the invention provides a section extraction method based on point cloud digital-analog, which enables a railway design specialty to automatically acquire a required transverse section and a required longitudinal section through the point cloud digital-analog provided by the mapping specialty, realizes quick iteration of survey results and accelerates the design process.

In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:

A section extraction method based on point cloud digital-analog specifically comprises the following steps:

s1: constructing a digital elevation model by utilizing the point cloud data;

s2: acquiring a section name and a node plane coordinate;

s3: extracting a section by using a digital elevation model;

S4: checking section data;

s5: thinning section nodes;

S6: converting a section data format;

S7: drawing a sectional view.

Further, the step S1 specifically includes:

Selecting classified airborne LiDAR ground points according to the point cloud data, wherein the data format is LAS; and constructing a digital model by using a two-dimensional Dirony triangulation algorithm provided in a computational geometry algorithm library CGAL, and outputting a triangular network digital elevation model with a PLY format.

Further, the step S3 specifically includes:

And (3) interpolating the section node coordinates obtained in the step (S2) along the section advancing direction by a specified interval to obtain encrypted node coordinates, setting section extraction options and parameters, and interpolating the encrypted node elevation on the triangular surface of the digital elevation model by using a gravity center coordinate smooth interpolation algorithm to obtain the section.

Further, the step S4 specifically includes:

S4.1: when the middle pile level sheet actually measured in the field exists in the target area, comparing the extracted cross section intersection point with middle pile data of the same pile number of the level sheet, designating a flat distance and an altitude difference limit difference, and when the distance between the cross section intersection point and a pile point in the field is recognized as the same point position within the flat distance limit difference, if the distance between the cross section intersection point and the pile point in the field is within the limit difference, using node elevation interpolated by pile height Cheng Tihuan in the field, otherwise, reserving the interpolation elevation and marking as altitude difference overrun;

S4.2: outputting a middle pile replacement condition table, listing the same point location interpolation elevation and field practice height Cheng Chazhi, marking replacement and overrun pile numbers, feeding back the overrun pile numbers to the point cloud number module to serve as a department to search reasons, feeding back field inspection middle pile data if no problem exists, and finally obtaining section data without error inspection.

Further, the step S5 specifically includes:

setting the limit difference of the flat distance and the elevation difference between a certain section node and the front and rear nodes of the section node, and controlling the thinning range; and removing the node from the section data result only when the difference value of the flat distance and the difference value of the height between the point and the front and rear points are within the limit difference and are not transverse and longitudinal intersection points, otherwise, reserving the node.

The invention has the beneficial effects that:

1) The software is used for extracting the section by utilizing the point cloud digital-analog under the AutoCAD platform, and a user can finish operations such as section data extraction, checking, thinning, format conversion and the like by one key after finishing necessary data preparation and option parameter setting, so that the section extraction efficiency is improved;

2) The invention can be fused with a line professional design software platform, and the line flat-vertical design result is directly used for recovering a line design model by a section extraction module, so as to realize parameterization extraction of the transverse section and the vertical section; outputting the section extraction result into a section and land line format required by line design, and assisting the line design;

3) The invention realizes the input of plane coordinates of transverse and longitudinal sections in various modes, meets the use requirements of different professions of railway investigation and design, and can independently extract the required section results after obtaining the point cloud results provided by the surveying and mapping profession, thereby reducing the mutual extraction of data, realizing rapid iteration and accelerating the design process.

Drawings

FIG. 1 is a flow chart of the present invention:

FIG. 2 is a diagram of key data structures used by CGAL generation digital mode;

FIG. 3 is a schematic diagram of a gravity center coordinate smoothing interpolation algorithm adopted by the extraction section of the invention;

FIG. 4 is a format of a land line required for the line art according to the present invention;

FIG. 5 is a cross-sectional format required by the line art of the present invention;

Fig. 6 shows a section format required for tunnel profession according to the present invention.

Detailed Description

The present invention will be described in detail with reference to the following embodiments.

As shown in fig. 1, the present invention specifically includes the following steps:

S1: and selecting the ground point cloud data obtained after data processing links such as checking, resolving, filtering, classifying and the like, wherein the data format is LAS, and the required point cloud data range can be determined through point cloud combined graph retrieval. And constructing a digital model by using a two-dimensional Dirony triangulation algorithm provided in a computational geometry algorithm library CGAL, and outputting a triangular network digital elevation model with a PLY format. The CGAL is a large geometric data structure and algorithm library based on C++, when the Point cloud builds a digital model, the Point cloud is read and stored into the CGAL, wherein the point_set_3< point_3> Point set is traversed by a TIN structure, a two-dimensional Delaunay triangle net is built, and finally the triangle net is written out into PLY format by a binary file stream. The key data structure used by CGAL is shown in fig. 2.

S2: the method comprises the following steps of obtaining a section name and a node plane coordinate:

1) Pattern mode: extracting information by selecting lines and text labels drawn in the plan view;

2) Parameter mode: setting parameters such as a starting mileage, a finishing mileage, a section distance, a section left-right offset distance and the like of a line, and calculating to obtain section node coordinates and pile number name labels through a line design model;

3) Text mode: and calculating to obtain the section node coordinates corresponding to the mileage stake marks through a line design model by using a text file containing the names of the appointed mileage stake marks and the left and right offset distances of the sections.

Lines and texts in the graphic mode are drawn by a designer according to professional requirements and are combined with a plan, and when an auxiliary line design or a section required by the profession cannot be associated with a line model, the mode is adopted, such as a bridge vertical section, a culvert shaft, a tunnel inclined shaft, a debris flow trench vertical section for investigating bad geological conditions and the like are common. Line node coordinates and text labels are obtained by calling a function under the ObjectARX software package autodesk. The Line type can directly acquire the coordinates of the start point and the end point through the readable and writable attribute StartPoint, endPoint; the type of the multi-segment line Polyline can obtain the node coordinates of the designated sequence number through a GetPoint dAT function; a single line of Text is obtained by its readable and writable attribute TextString, and multiple lines of Text is obtained by its read-only attribute Text.

The parameter mode and the text mode are used for directly recovering the line design model through the data of a curve table and a broken link table (if broken links exist) which are imported into the line design result, and after the model is recovered, line mileage coordinate conversion is carried out through the researched software, so that the node coordinates of the target mileage stake mark section are calculated. The line design model results are an Access database file (in mdb format) or a set of text files (in txt format). The parameter mode is suitable for controlling whether pile is added at the broken chain or not through the parameter option when the section with fixed spacing in the specified mileage paragraph is required to be obtained continuously; the text mode is applicable when multiple, non-consecutive mileage paragraph stake marks need to be extracted.

The three input modes are used for uniformly defining the line advancing direction from a small mileage to a large mileage direction, defining the offset of the left paragraph of the cross section in the line advancing direction as a negative value and the offset of the right paragraph as a positive value.

S3: and (3) extracting a section by using a digital elevation model, firstly copying a required point cloud digital-analog file to a project engineering folder, dividing and encrypting the section nodes obtained in the step (2) at fixed intervals, successively taking the encrypted section nodes as points P to be solved, calculating barycentric coordinates (alpha, beta and gamma) corresponding to the points in a two-dimensional plane by using the two-dimensional coordinates of triangle vertexes ABC where the points P are positioned, and calculating the elevation of the points P by using the barycentric coordinates and the triangle vertexes Gao Chengjie, wherein a specific algorithm is shown in figure 3. And outputting the obtained three-dimensional coordinates of the encrypted section node and the offset distance of the node from the transverse and longitudinal intersection points together to obtain the cross section data. If the vertical section is extracted, the three-dimensional coordinates of the node of the encrypted section and the flat distance between the node and the starting point along the line direction are output together, so that the vertical section data are obtained. In this embodiment, the fixed pitch is 0.1m.

S4: section data checking

In the railway line determination stage, the center line measurement is completed in the field by adopting a total station polar coordinate method, a GPS RTK method and other manual actual measurement modes, a horizontal single file containing the plane coordinates and the elevation of the middle pile is obtained, and the horizontal single file is arranged into a text file only containing two data of the pile number and the elevation.

S4.1: setting a flat distance limit difference when extracting the cross section, setting the limit difference to be 0.3m in the embodiment, when the flat distance between the section node and the middle pile point is within the limit difference, determining the same point position, comparing the field actual measurement elevation data of the middle pile point position corresponding to the horizontal single point with the elevation data extracted by the section node by using the ground mould at the moment, when the horizontal single elevation is within the set limit difference, replacing the point cloud with the horizontal single elevation to extract the elevation value, and if the difference is greater than 0.5m, feeding back field verification according to the requirements of the technical rules of railway engineering satellite positioning and remote sensing measurement. Marking the replacement elevation and the pile number and the difference value to be fed back and checked, and outputting the marked replacement elevation and the pile number and the difference value to a text file of the replacement condition of the middle pile for reference;

s4.2: listing the same point location interpolation elevation and field practice height Cheng Chazhi, marking out replacement and overrun pile numbers, feeding back the overrun pile numbers to a point cloud module to make department search reasons, feeding back field inspection middle pile data if no problem exists, and finally obtaining section data with no error in inspection.

S5: in order to keep all the topography characteristic points as much as possible, the fixed distance between the encrypted section nodes in the step S3 is set to be 0.1m, the extracted section data redundant points are more, the processing is inconvenient when the professional use is designed, and the section nodes are required to be thinned;

In addition, certain design professions have limit requirements on the maximum distance of the section nodes, so that the limit difference of the flat distance is 5m and the limit difference of the elevation is 0.2m when the section nodes are thinned in the embodiment. The specific algorithm is to control the thinning range by using the difference value of the flat distance and the elevation between a certain section node and the front and back nodes, and remove the node from the section data result only when the difference value of the flat distance and the elevation between the point and the front and back points are within the limit difference and are not the transverse and longitudinal intersection points. If a transverse intersection point exists when the longitudinal section is extracted, in order to ensure that the intersection point is not thinned, a text file containing the stake number of the intersection point is input, and judgment is carried out by calculating whether the distance between a target point and a starting point along the section direction is equal to the mileage difference from the intersection point to the starting point when thinning, so that the section node of the designated stake number is forcedly reserved. The cross section data obtained after the thinning keeps left and right endpoints and intersection points, and comprises all node offset distances and three-dimensional coordinate values (taking the intersection point as the center, the left offset distance is negative and the right side is positive) from the left endpoint to the right endpoint according to the line advancing direction; the vertical section data is based on the starting point and comprises the offset distance from the section line to the starting point of all the nodes and the three-dimensional coordinate value of the nodes.

S6: and carrying out format conversion on the section intermediate result processed by the steps according to a professional design software interface. Traffic route selection CAD system software for route design, the ground line format of which is shown in FIG. 4: the first line of marks 2 indicates that the section data is in two columns, the second and third lines indicate stake marks and intersection elevations of two pieces of cross section data, and the last line is the text data end mark. The cross-sectional format is shown in fig. 5: the first line of mark 1 indicates that the section data is absolute distance and absolute elevation, then the intersection pile number and elevation of each section, the intersection point and elevation of the left side of the section, the left side section node ending mark '0 0', the intersection point and elevation of the right side of the section, the right side section node ending mark '0 0' and finally the identification file of '0 00 0' are recorded in sequence. The cross-sectional format required by the tunnel profession is shown in fig. 6, all section data are organized and stored as an Excel worksheet named as a ground line, each section comprises two rows of records, the first row of records stores the section name and the offset of each node, and the second row of records stores the elevation value of each node. The data interfaces required by the road profession are cross sections and land line formats defined by the weft road traffic aided design system software (HintCAD). The conversion of various professional interfaces between text formats is mainly finished through format arrangement after arithmetic operation, and the conversion of the text formats into an Excel list is mainly finished through calling an API in Microsoft. Office. Intep. Excel. Dll to process the Excel worksheet, and the cell Cells of the text formats are filled with values according to format requirements according to lines and rows.

S7: storing the thinned transverse and vertical section result data under a specified folder in a text file format, and setting parameters such as a drawing data catalog, a drawing scale, a table width height, a transverse and vertical section layout of a section diagram, a layer, colors, character patterns and the like to draw the section diagram by using drawing tool software developed under an AutoCAD platform.

The content of the invention is not limited to the examples listed, and any equivalent transformation to the technical solution of the invention that a person skilled in the art can take on by reading the description of the invention is covered by the claims of the invention.

Claims (2)

1. A section extraction method based on point cloud digital-analog is characterized by comprising the following steps: the method specifically comprises the following steps:

s1: constructing a digital elevation model by utilizing the point cloud data;

s2: acquiring a section name and a node plane coordinate;

s3: extracting a section by using a digital elevation model;

S4: checking section data;

s5: thinning section nodes;

S6: converting a section data format;

s7: drawing a sectional view;

the step S3 specifically comprises the following steps:

Interpolating the section node coordinates obtained in the step S2 along the section advancing direction through a designated interval to obtain encrypted node coordinates, setting section extraction options and parameters, and interpolating the encrypted node elevation on the triangular surface of the digital elevation model by using a gravity coordinate smoothing interpolation algorithm to obtain the section;

the step S4 specifically includes:

S4.1: when the middle pile level sheet actually measured in the field exists in the target area, comparing the extracted cross section intersection point with middle pile data of the same pile number of the level sheet, designating a flat distance and an altitude difference limit difference, and when the distance between the cross section intersection point and a pile point in the field is recognized as the same point position within the flat distance limit difference, if the distance between the cross section intersection point and the pile point in the field is within the limit difference, using node elevation interpolated by pile height Cheng Tihuan in the field, otherwise, reserving the interpolation elevation and marking as altitude difference overrun;

S4.2: outputting a middle pile replacement condition table, listing the same point location interpolation elevation and field practice height Cheng Chazhi, marking replacement and overrun pile numbers, feeding back the overrun pile numbers to a point cloud module to serve as a department to find reasons, feeding back field inspection middle pile data if no problem exists, and finally obtaining section data without error in inspection;

the step S5 specifically comprises the following steps:

setting the limit difference of the flat distance and the elevation difference between a certain section node and the front and rear nodes of the section node, and controlling the thinning range; and removing the node from the section data result only when the difference value of the flat distance and the difference value of the height between the point and the front and rear points are within the limit difference and are not transverse and longitudinal intersection points, otherwise, reserving the node.

2. The method for extracting the cross section based on the point cloud digital-analog according to claim 1, wherein the method comprises the following steps: the step S1 specifically comprises the following steps:

Selecting classified airborne LiDAR ground points according to the point cloud data, wherein the data format is LAS; and constructing a digital model by using a two-dimensional Dirony triangulation algorithm provided in a computational geometry algorithm library CGAL, and outputting a triangular network digital elevation model with a PLY format.