CN117848238B - Rail transit station actual limit measurement method based on laser point cloud - Google Patents
Rail transit station actual limit measurement method based on laser point cloud Download PDFInfo
- Publication number
- CN117848238B CN117848238B CN202410251135.9A CN202410251135A CN117848238B CN 117848238 B CN117848238 B CN 117848238B CN 202410251135 A CN202410251135 A CN 202410251135A CN 117848238 B CN117848238 B CN 117848238B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- station
- points
- track
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000691 measurement method Methods 0.000 title description 3
- 238000000034 method Methods 0.000 claims abstract description 43
- 238000010586 diagram Methods 0.000 claims abstract description 26
- 238000012937 correction Methods 0.000 claims abstract description 23
- 238000004364 calculation method Methods 0.000 claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 7
- 238000000605 extraction Methods 0.000 claims abstract description 5
- 238000005259 measurement Methods 0.000 claims description 28
- 239000011159 matrix material Substances 0.000 claims description 18
- 238000011161 development Methods 0.000 claims description 11
- 238000009499 grossing Methods 0.000 claims description 9
- 229910000831 Steel Inorganic materials 0.000 claims description 4
- 238000003384 imaging method Methods 0.000 claims description 4
- 239000010959 steel Substances 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 19
- 230000018109 developmental process Effects 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 238000003780 insertion Methods 0.000 description 2
- 230000037431 insertion Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 244000007853 Sarothamnus scoparius Species 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Landscapes
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention relates to the field of three-dimensional laser point cloud data processing and calculating methods, in particular to a track traffic platform actual limit measuring method based on laser point cloud, which comprises the following steps: s1, acquiring three-dimensional point cloud data of a station; s2, generating a point cloud gray level expansion diagram through a projection algorithm, and extracting a point cloud section at a measuring position; s4, for the curve segment station, correcting the actual limit value D of the station by using the track super-high value, and obtaining a corrected result as an actual limit value D' of the station of the curve segment station. The invention has the advantages that: the detection efficiency is obviously improved, and the detection labor cost is reduced. And accurately extracting station point clouds at specific positions, and calculating actual limit values. The automation degree of the limit calculation process is high, and the algorithm is accurate and reliable in automatic extraction. For curve segment station, corresponding calculated data correction can be implemented.
Description
Technical Field
The invention relates to the field of three-dimensional laser point cloud data processing and calculating methods, in particular to a track traffic platform actual limit measuring method based on laser point cloud.
Background
1. The subway limit is a graph, a coordinate parameter and a contour dimension line for guaranteeing safe operation of a subway, limiting the cross section dimension of a vehicle, limiting the installation dimension of equipment along the line and determining the effective clearance dimension of a building structure.
The means of limit detection are mainly two types, contact limit detection and non-contact limit detection. The contact type limit detection is a main means of early detection, mainly comprises a cross section method, a comprehensive section method, a track method and the like, and the detection process is mainly carried out by a vehicle-mounted limit frame or tools such as a manual operation level bar, a tape measure, a hanging hammer, a theodolite and the like in a physical contact mode, and has the defects of low precision, low speed, large workload, poor reliability and the like, and can not realize effective continuous detection, and the detection is possibly missed.
The non-contact detection technology has the advantages of high efficiency, high precision and no contact, overcomes the defects of the original detection method, and is rapidly developed in recent years. The main non-contact detection methods include video imaging method, laser scanning method, etc. Video shooting method applies high-speed and high-resolution cameras to carry out limited rapid detection based on mobile binocular vision measurement, but has high requirements on the environment of the light beam, and is limited in application in tunnels. The laser scanning method is a multifunctional rapid detection method aiming at railway tunnel building limit detection, maintenance and management development, and the distance information of surrounding buildings is continuously measured through laser scanning equipment to form a building limit section view, so that the laser scanning method is little interfered by external factors such as light rays and can be used in a complex tunnel environment, as shown in fig. 1.
2. Laser mobile scanning system and achievement point cloud:
The principle of the three-dimensional laser scanning technology is that the three-dimensional coordinates and the reflection intensity of the surface of the measured object are obtained by scanning by emitting laser, so that the three-dimensional laser scanning technology is a non-contact active measurement system. The three-dimensional laser scanning technology has the main advantages that the information such as the space point position coordinates, the texture colors, the echo reflection intensity and the like with high resolution can be rapidly obtained, so that a brand new technical means is provided for rapidly establishing a three-dimensional digital model of an object.
The inventor integrates a three-dimensional laser scanner, an odometer, an industrial personal computer and a mobile trolley to develop a set of 3D laser mobile scanning system, and a system photo is shown in fig. 2. Meanwhile, matched data acquisition and data processing software is developed, and the functions of tunnel convergence calculation, scanning tiling diagram generation, tunnel limit calculation, clearance analysis, ring rotation angle staggering analysis and the like are achieved. The 3D laser mobile scanning system scans the existing tunnel structure rapidly and efficiently, particularly rapidly collects, analyzes and monitors tunnel deformation and diseases, finally forms a whole tunnel real-scene point cloud model, performs further refined processing on data by combining a point cloud fitting method and a split-ring fitting algorithm, and improves data integrity and reliability. All functions are integrated and packaged into a replicable system, so that the subway operation and maintenance detection problem is rapidly, efficiently, conveniently and accurately solved.
3. Recovering three-dimensional space data of a station:
As shown in the 3D laser moving scanning schematic diagram of fig. 3, a scanner laser transmitter is taken as an origin o, a z axis points to a zenith direction, a y axis is taken as a push-broom advancing direction, and a right-hand system is established, and is called o-xyz as a central coordinate system.
The scanner horizontal scale is fixed to ensure that the scanning direction is consistent with the track travelling direction as much as possible. As shown in fig. 4, the small vehicle travels with the scanner perpendicular to the traveling direction, and two-dimensional subway station profile data is collected while the odometer records the data.
The scanner acquisition points are two-dimensional points of the station outline, the third dimension is recorded by the odometer, and the pseudo three-dimensional space data of the subway station can be restored by combining the two data. The specific data recovery steps are as follows:
1) Determining the push broom total time: the odometer and the scanner are synchronously recorded, and the odometer acquires data once in 0.1s, so that the head-to-tail odometer time range of the odometer is considered to be a measuring section time interval;
2) Determining the starting time and the ending time of each scanning line: the scanner is considered to uniformly collect each scanning line under fixed parameters, so that the starting point and the end point time of each scanning line can be obtained only through simple uniform distribution;
3) Determining the starting mileage and the scanning line length of each scanning line: interpolating mileage according to the starting time of the scanning line; selecting a minimum mileage range containing the starting point and the end point time of the scanning line, interpolating the scanning line speed, and calculating the scanning line length by combining the scanning line time;
4) Determining mileage value of each scanning line: combining the starting mileage of each scanning line in the step 3), and uniformly distributing the length to each scanning line point;
after the flow, the cloud information of the subway station can be recovered, but the space structure of the subway station is transformed: the shape of the straight line section of the station is unchanged, and the curved section of the station is presented in a straight line. And is shown in particular in figure 7.
The vehicle station cloud coordinates (x, y, z) are calculated as follows:
Wherein: d is the laser measurement distance, θ is the laser scanning angle, and D m is the travel distance calculated by the laser scanning time encoder.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a track traffic platform actual limit measurement method based on laser point cloud, which is used for calculating the actual limit of a station by using 3D mobile laser scanning point cloud.
In order to achieve the above purpose, a method for measuring the actual limit of a rail transit platform based on laser point cloud is designed, which specifically comprises the following steps:
S1, acquiring three-dimensional point cloud data of a station through laser scanning equipment;
S2, generating a point cloud gray level expansion diagram through a projection algorithm, determining a measurement position in the point cloud gray level expansion diagram, and extracting a point cloud section at the measurement position;
S3, in the point cloud section, calculating a track surface perpendicular bisector through identifying and calculating track characteristic points, also calculating a platform screen door outer edge characteristic point at a platform side, and obtaining a straight line section station platform actual limit value D through a point-to-straight line distance formula;
S4, judging whether the station is a curve section station, and for the curve section station, because of the existence of the track superelevation, the perpendicular bisector of the track surface is inconsistent with the actual plumb line, correcting the actual limit value D of the station by using the track superelevation, and obtaining the actual limit value D' of the station of the curve section as a corrected result.
In the step S2, a point cloud gray scale expansion chart is generated through a projection algorithm, which specifically comprises the following steps:
the point cloud is unfolded by taking a track center line o ,x, as a transverse axis and a station section direction o ,y, as a vertical axis, and a subway station plane coordinate system is established; calculating the gray level spread image pixel position of the point cloud according to the point cloud coordinates by using a BMP image with a certain size, filling pixels with the laser reflection value of a laser point, and sequentially obtaining a vehicle-site cloud plane gray level map;
assuming that the width of the station is w, the height of the station is h, correspondingly generating a BMP chart with the width of wp and the height of hp, and recording the maximum and minimum mileage y in the point cloud data as follows: y max、ymin, the maximum and minimum expansion length l coordinates are: l max,lmin, find the position of each point in the pixel as: Wherein wp i、hpi is the corresponding position of the point in the image, respectively, (int) means rounding down;
each pixel corresponds to a certain interval range, the number of points in each pixel is inconsistent, and the gray value of the pixel can be calculated according to the following formula: Wherein n represents the number of points in a pixel, g i is the point gray scale, pg is the pixel gray scale, and g i is the intensity value of a laser point in the point cloud, which is obtained by laser scanning equipment;
The places which are not scanned by the laser and the areas without laser points can cause clear strip black points in the image, in order to filter the black points in the imaging image, smoothing treatment is needed to be carried out on the tiled image, and gray values of pixels without gray points are given in a smoothing mode;
The specific smooth template is as follows: The determination of the smooth template is to determine the corresponding gray value in the middle of the template according to the reciprocal distance between an unknown pixel and an adjacent pixel, the template specification is 5 rows and 5 columns, if no point exists in a certain pixel, the adjacent pixel with the point is used for weighted average, and the condition that the template is unsuitable for use is that: the top two rows of pixels adopt templates as the lower three rows of the weight matrix; the two bottom rows of pixels adopt templates as the upper three rows of the weight matrix; two columns of pixels at the leftmost side adopt a right three-column matrix with a template as a weight matrix; two columns of pixels at the rightmost side adopt a template as a left three-column matrix of the weight matrix;
by means of smoothing the unknown pixel templates, black areas in the images can be eliminated, clearer station point cloud plane gray level images are obtained, and in order to facilitate the viewing of the station point cloud plane gray level images, the gray level lower limit can be set to 180, so that the image brightness is moderate.
In the step S2, a measurement position is determined in the point cloud gray scale development chart, and a point cloud section at the measurement position is extracted, specifically as follows:
marking a designated measurement position in the point cloud gray scale development chart, and obtaining point cloud data of the designated position through the projection corresponding relation between the image and the point cloud;
in data processing software, marking a specific position of a platform shielding door by using a red vertical line in a station point cloud gray scale development diagram, converting the image pixel coordinates of the shielding door into shielding door mileage data, namely extracting a station shielding door point cloud section, fitting a section point to a plane, projecting the point to the plane, and further calculating.
The step S3 specifically includes the following steps:
In the section of the extracted point cloud, the characteristic points of the tracks on two sides are rapidly extracted through a minimum optimization sequence algorithm of a binary heap, and the track center point and the track perpendicular bisector are calculated; the track center point is the midpoint of the two feature points, and the track perpendicular bisector is the perpendicular bisector of the connecting line segment of the two feature points;
The specific algorithm for extracting the characteristic points of the steel rail is as follows:
S3.1.1 in the scanner frame, given the approximate positions of the left and right rails, two point clouds A1 and A2 are obtained (here, parameters are customized and optimized in conjunction with the scanner position);
s3.1.2 building KD tree indexes for the point cloud set A1 and the point cloud set A2 respectively;
S3.1.3 process each point cloud in point cloud A1 and point cloud A2, respectively: setting a search radius, judging the number of points in the drawn space sphere, and if the number of points is lower than a threshold value, considering the points as rough difference points and deleting the rough difference points;
s3.1.4 through step S3.1.3, irrelevant noise points may be deleted, and the denoised point cloud set A1 and point cloud set A2 are processed as follows:
S3.1.4.1 constructing a binary heap minimum priority queue in the zenith direction, wherein the number of the set points in the queue is 20;
s3.1.4.2 extracting the maximum 20 points of the Z coordinate from the queue (assuming that the zenith direction is Z coordinate and the section direction is X coordinate), and calculating the mean value to be the Z coordinate of a certain characteristic point of the track;
s3.1.4.3 obtaining a 10 cm thickness point downwards according to the Z coordinate of the characteristic point, and constructing a new binary heap minimum priority queue;
S3.1.4.4 in a new point cloud binary heap minimum priority queue, acquiring 20 points with minimum absolute values of X coordinates, and calculating a mean value to obtain the X coordinates of a certain characteristic point of the track;
s3.1.4.5 two feature points of the left and right rails can be calculated by S3.1.4.1 to S3.1.4.4.
The step S3 specifically includes the following steps:
calculating a rectangular range of the station edge based on the track and station edge position parameters from the track characteristic points calculated in the step S3.1.4.5, searching a plurality of characteristic point sets, wherein the point sets comprise the station edge and other noise points, and carrying out further extraction processing;
For points in the rectangular range, establishing a point cloud index KD tree, and quickly searching for adjacent points of the points; calculating the number of points existing in the radius by setting the position of the search radius, and if the number is less than the set threshold value, considering the points as rough difference points and deleting the rough difference points;
Respectively calculating the distance from the points to the track center line from the rest point set, selecting 20 points with the minimum distance, and taking an average value as a platform edge characteristic point;
The distance from the characteristic point of the platform edge to the center vertical line of the track surface is the station limiting distance D of the section; when the track height is present, the curve height correction is required, and the process proceeds to step S4.
The step S4 specifically includes the following steps:
If the station is positioned on the line straight line section, the perpendicular bisector of the track surface of the scanning point cloud can be regarded as the plumb line of the track center, and the calculated result is the distance from the center line of the real line to the edge of the station;
If the station is positioned on the curve section of the line, the track surface center plumb line is unequal to the track center plumb line due to the existence of the track superelevation, and if the distance between the station platform edge on the curve section and the track center plumb line is required to be obtained, the calculation result is required to be subjected to superelevation correction;
According to the geometric relationship, the ultrahigh correction of the curve segment needs the following parameters: the track gauge L defaults to 1435mm, the track is ultrahigh H, the height difference H between the platform edge characteristic points and the track plane center point, the station limiting distance D, the horizontal distance D between the platform edge characteristic points and the track plane line, and the ultrahigh correction x are adopted, so that the horizontal distance D' between the platform edge and the track center plumb line is calculated;
The curve segment superelevation correction can be divided into the following two cases according to the positions of the rail superelevation and the platform edge (which can be determined by a station line profile):
(1) The track is superhigh and approaching the platform side, and the correction calculation is as follows:
(2) The track is superhigh and far away from the platform side, and the correction calculation is as follows: 。
Compared with the prior art, the invention has the advantages that:
(1) The method is different from the traditional measuring tools and instruments for directly measuring and obtaining the actual limit of the station, the station cloud of the vehicle is obtained through three-dimensional laser mobile scanning, and a set of station actual limit measuring method based on the laser point cloud is researched based on the point cloud characteristics.
(2) Compared with the traditional manual measurement mode, the method has the advantages that the detection efficiency is remarkably improved, and the detection labor cost is reduced.
(3) And marking the section of the platform shielding door by generating a vehicle station cloud plane gray level diagram, accurately extracting station point cloud at a specific position, and calculating an actual limit value.
(4) And a binary heap algorithm and a KD tree algorithm are integrated, feature points are automatically extracted, an actual limit value is calculated, the automation degree of the limit calculation process is high, and the algorithm is accurate and reliable in automatic extraction.
(5) And correcting the calculation result of the platform limit value based on the track superelevation for the curve section station, and correcting the calculation result to be the plumb line distance from the platform edge to the line center.
Drawings
FIG. 1 is a schematic view of a bounding transverse dimension;
FIG. 2 is a schematic diagram of a 3D laser mobile scanning system;
FIG. 3 is a schematic diagram of a 3D laser mobile scanning coordinate system definition;
FIG. 4 is a 3D laser mobile scanning top view;
FIG. 5 is a top view of a 3D laser motion scan pseudo three-dimensional restoration;
FIG. 6 is a flow chart of calculation of the actual limit of the mobile laser scanning station;
FIG. 7 is a schematic view of station boundary values in a straight line segment station;
FIG. 8 is an expanded view of vehicle station cloud plan coordinates;
fig. 9 is a station point cloud plane gray scale;
FIG. 10 is a schematic view of three key positions of a marker platform door in a point cloud tiling;
FIG. 11 is a schematic view of track feature point calculation;
FIG. 12 is a schematic view of a station edge feature point calculation;
FIG. 13 is a schematic view of a point cloud profile feature point line and station boundary values;
FIG. 14 is a schematic view of a curve segment station boundary measurement at the superhigh position near the platform side;
FIG. 15 is a schematic view of a curve segment station boundary measurement at an elevated position away from the platform side;
FIG. 16 is a schematic view of three cross-sectional measurement positions of the left, middle and right of the shield door;
FIG. 17 is a graph comparing station boundary values based on laser point clouds with total station measurements;
in the figure: 1.1 station; 1.2 defines a transverse dimension; 1.3 steel rail; 2.1 a main body; 2.2 driving wheel groups; 2.3 driven wheel sets; 2.4 car lights; 2.5 universal wheels; 2.6 A column; 2.7 A scanner; 2.8 displays; 2.9 encoders; 6.1 a first tunnel; 6.2 the direction of travel of the trolley; 6.3 scanner scanning direction; 7.1 a second tunnel; 7.2 pseudo three-dimensional recovery trolley travelling direction; 7.3, restoring the scanning direction of the scanner in a pseudo three-dimensional mode; 10.1 track surface perpendicular bisectors; 10.2 station structure edges; 10.3 a first orbital plane; 14.1 track feature point locations; 15.1 station edge feature points; 17.2 track center plumb line; 17.4 a second orbital plane; 17.5 gauge L;17.6 track super high h;17.7, the height difference H between the platform edge characteristic points and the track plane center point; 17.8 station clearance distance D;17.9, the horizontal distance d from the characteristic point of the platform edge to the perpendicular bisector of the track surface; 17.10 ultra-high correction factor x;18.4 a third track plane; 19.1; 19.2; 19.3.
Detailed Description
The invention is further described below with reference to specific embodiments, and provides a method for measuring actual limit of a rail transit platform based on laser point clouds.
The method comprises the steps of obtaining three-dimensional point cloud data of a station based on a 3D mobile laser scanning system developed by my company, generating a point cloud gray level expansion diagram through a projection algorithm, determining a measurement position in the point cloud gray level expansion diagram, and extracting a point cloud section at the measurement position to be further calculated; in the point cloud section, a track surface perpendicular bisector can be calculated by identifying and calculating track characteristic points, the outer edge characteristic points of the platform screen door are also calculated on the platform side, and an actual limit value D of the station platform of the straight line segment is obtained through a point-to-straight line distance formula, wherein a schematic diagram of the actual limit value D of the station platform of the straight line segment is shown in figure 7. For the curve section station, due to the fact that the track is ultrahigh, the perpendicular bisector of the track surface is inconsistent with the actual plumb line, the track ultrahigh value is used for correcting the actual limit value D of the station in two cases, and the corrected result is the actual limit value D' of the station of the curve section station.
The actual limit flow of the station is calculated by using the 3D mobile laser scanning point cloud as shown in fig. 6:
Specifically, the method comprises the following steps:
S1, acquiring three-dimensional point cloud data of a station through laser scanning equipment;
S2, generating a point cloud gray level expansion diagram through a projection algorithm, determining a measurement position in the point cloud gray level expansion diagram, and extracting a point cloud section at the measurement position;
S3, in the point cloud section, calculating a track surface perpendicular bisector through identifying and calculating track characteristic points, also calculating a platform screen door outer edge characteristic point at a platform side, and obtaining a straight line section station platform actual limit value D through a point-to-straight line distance formula;
S4, judging whether the station is a curve section station, and for the curve section station, because of the existence of the track superelevation, the perpendicular bisector of the track surface is inconsistent with the actual plumb line, correcting the actual limit value D of the station by using the track superelevation, and obtaining the actual limit value D' of the station of the curve section as a corrected result.
In the step S2, a point cloud gray scale expansion chart is generated through a projection algorithm, which specifically comprises the following steps:
The point cloud is unfolded by taking a track center line o ,x, as a transverse axis and a station section direction o ,y, as a vertical axis, and a subway station plane coordinate system is established; calculating the gray level spread image pixel position of the point cloud according to the point cloud coordinates by using a BMP image with a certain size, filling pixels with the laser reflection value of a laser point, and sequentially obtaining a vehicle-station cloud plane gray level map, wherein a schematic diagram is shown in fig. 9;
assuming that the width of the station is w, the height of the station is h, correspondingly generating a BMP chart with the width of wp and the height of hp, and recording the maximum and minimum mileage y in the point cloud data as follows: y max、ymin, the maximum and minimum expansion length l coordinates are: l max,lmin, find the position of each point in the pixel as: Wherein wp i、hpi is the corresponding position of the point in the image, respectively, (int) means rounding down;
each pixel corresponds to a certain interval range, the number of points in each pixel is inconsistent, and the gray value of the pixel can be calculated according to the following formula: Wherein n represents the number of points in a pixel, g i is the point gray scale, pg is the pixel gray scale, and g i is the intensity value of a laser point in the point cloud, which is obtained by laser scanning equipment;
The places which are not scanned by the laser and the areas without laser points can cause clear strip black points in the image, in order to filter the black points in the imaging image, smoothing treatment is needed to be carried out on the tiled image, and gray values of pixels without gray points are given in a smoothing mode;
The specific smooth template is as follows: The determination of the smooth template is to determine the corresponding gray value in the middle of the template according to the reciprocal distance between an unknown pixel and an adjacent pixel, the template specification is 5 rows and 5 columns, if no point exists in a certain pixel, the adjacent pixel with the point is used for weighted average, and the condition that the template is unsuitable for use is that: the top two rows of pixels adopt templates as the lower three rows of the weight matrix; the two bottom rows of pixels adopt templates as the upper three rows of the weight matrix; two columns of pixels at the leftmost side adopt a right three-column matrix with a template as a weight matrix; two columns of pixels at the rightmost side adopt a template as a left three-column matrix of the weight matrix;
by means of smoothing the unknown pixel templates, black areas in the images can be eliminated, clearer station point cloud plane gray level images are obtained, and in order to facilitate the viewing of the station point cloud plane gray level images, the gray level lower limit can be set to 180, so that the image brightness is moderate.
In the step S2, a measurement position is determined in the point cloud gray scale development chart, and a point cloud section at the measurement position is extracted, specifically as follows:
marking a designated measurement position in the point cloud gray scale development chart, and obtaining point cloud data of the designated position through the projection corresponding relation between the image and the point cloud;
In data processing software, marking a specific position of a platform shielding door by using a red vertical line in a station point cloud gray scale development diagram, converting the image pixel coordinates of the shielding door into shielding door mileage data, namely extracting a station shielding door point cloud section, fitting a section point to a plane, projecting the point to the plane, and further calculating. Schematic diagrams of three key positions of the marking platform door in the point cloud tiling diagram are shown in fig. 10.
The step S3 specifically includes the following steps:
In the section of the extracted point cloud, the characteristic points of the tracks on two sides are rapidly extracted through a minimum optimization sequence algorithm of a binary heap, and the track center point and the track perpendicular bisector are calculated; the track center point is the midpoint of the two feature points, and the track perpendicular bisector is the perpendicular bisector of the connecting line segment of the two feature points;
the specific algorithm for extracting the characteristic points of the steel rail is as follows: the schematic diagram is shown in FIG. 12
S3.1.1 in the scanner frame, given the approximate positions of the left and right rails, two point clouds A1 and A2 are obtained (here, parameters are customized and optimized in conjunction with the scanner position);
s3.1.2 building KD tree indexes for the point cloud set A1 and the point cloud set A2 respectively;
S3.1.3 process each point cloud in point cloud A1 and point cloud A2, respectively: setting a search radius, judging the number of points in the drawn space sphere, and if the number of points is lower than a threshold value, considering the points as rough difference points and deleting the rough difference points;
s3.1.4 through step S3.1.3, irrelevant noise points may be deleted, and the denoised point cloud set A1 and point cloud set A2 are processed as follows:
S3.1.4.1 constructing a binary heap minimum priority queue in the zenith direction, wherein the number of the set points in the queue is 20;
s3.1.4.2 extracting the maximum 20 points of the Z coordinate from the queue (assuming that the zenith direction is Z coordinate and the section direction is X coordinate), and calculating the mean value to be the Z coordinate of a certain characteristic point of the track;
s3.1.4.3 obtaining a 10 cm thickness point downwards according to the Z coordinate of the characteristic point, and constructing a new binary heap minimum priority queue;
S3.1.4.4 in a new point cloud binary heap minimum priority queue, acquiring 20 points with minimum absolute values of X coordinates, and calculating a mean value to obtain the X coordinates of a certain characteristic point of the track;
s3.1.4.5 two feature points of the left and right rails can be calculated by S3.1.4.1 to S3.1.4.4.
The step S3 specifically includes the following steps:
calculating a rectangular range of the station edge based on the track and station edge position parameters from the track characteristic points calculated in the step S3.1.4.5, searching a plurality of characteristic point sets, wherein the point sets comprise the station edge and other noise points, and carrying out further extraction processing;
and for points in the rectangular range, establishing a point cloud index KD tree, and quickly searching for adjacent points of the points. Calculating the number of points existing in the radius by setting the position of the search radius, and if the number is less than the set threshold value, considering the points as rough difference points and deleting the rough difference points;
respectively calculating the distance from the points to the track center line from the rest point set, selecting 20 points with the minimum distance, and taking an average value as a platform edge characteristic point; the characteristic point line of the cross section of the point cloud and the limit value of the station are shown in figure 13
The distance from the characteristic point of the platform edge to the center vertical line of the track surface is the station limiting distance D of the section; the calculation of station limit measurement is shown in figure 14; when the track height is present, the curve height correction is required, and the process proceeds to step S4.
The step S4 specifically includes the following steps:
If the station is positioned on the line straight line section, the perpendicular bisector of the track surface of the scanning point cloud can be regarded as the plumb line of the track center, and the calculated result is the distance from the center line of the real line to the edge of the station;
If the station is positioned on the curve section of the line, the perpendicular track surface perpendicular bisector is different from the track center plumb line due to the existence of the track superelevation, and if the distance between the station platform edge on the curve section and the track center plumb line is required to be obtained, the calculation result needs to be subjected to superelevation correction;
According to the geometric relationship, the ultrahigh correction of the curve segment needs the following parameters: the track gauge L defaults to 1435mm, the track is ultrahigh H, the height difference H between the platform edge characteristic points and the track plane center point is the station limiting distance D, the horizontal distance D between the platform edge characteristic points and the track plane center line is the ultrahigh correction factor x, and therefore the horizontal distance D' between the platform edge and the track center plumb line is calculated;
The curve segment superelevation correction can be divided into the following two cases according to the positions of the rail superelevation and the platform edge (which can be determined by a station line profile):
(1) The track is superhigh and near the platform side, and the correction calculation is as follows, and the schematic diagram is shown in fig. 15:
(2) The track is superhigh and far away from the platform side, and the correction calculation is as follows: 。
in this embodiment, the correlation calculation and adjustment method is as follows:
Minimum priority queuing algorithm for binary heap
The binary heap based minimum priority queue (Binary Heap Minimum Priority Queue) is a common data structure for implementing prioritized element storage and retrieval. In this data structure, each element has an associated priority, and the minimum priority queue ensures that the element with the smallest priority can be accessed and deleted quickly.
Insertion (Insertion): new elements are inserted into the heap and the properties of the heap are maintained. Typically, elements are added to the end of the array and then adjusted to the appropriate position by a floating operation.
(1) New elements are added to the end of the array.
(2) If the priority of the new element is less than the priority of its parent node, the new element and the parent node are swapped.
(3) Repeating the step (2) until the new element is no longer smaller than the parent node thereof or reaches the heap top.
Delete min element operation:
(1) The top of heap element is removed and the element at the end of the array is placed at the top of heap position.
(2) And comparing the priority of the heap top element and the priorities of the two child nodes, and finding the smallest child node.
(3) If the heap top element is larger than its smallest child node, they are swapped.
(4) Step 3 is repeated until the heap top element is no longer larger than its child node, or reaches a leaf node.
KD tree algorithm
KD tree (K-dimension tree) is a data structure, which can organize and represent the point set in K-Dimensional space, and is a binary search tree with other condition constraints, and can improve the operation efficiency during interval and neighbor search. The point cloud data is represented in three dimensions, where the root of the KD-tree is separated according to a first dimension (X-coordinate), the next level is separated according to a second dimension (Y-coordinate), each descending level of the KD-tree is separated in the next dimension, and all dimensions are exhausted and then returned to the first dimension. The most efficient way to build a KD-tree is to use a segmentation method like fast classification, put the value of the specified dimension on the root, where the left subtree contains the smaller value and the right subtree contains the larger value, and then repeat this step on the left and right subtrees, respectively, until only one element remains in the last subtree.
The KD tree space structure is adopted to manage the point cloud data, so that the radius neighborhood search and the K neighbor search can be realized relatively quickly, the search time is saved, and the realization of high-speed search in the point cloud data is critical to the subsequent operations of downsampling, point cloud segmentation, point cloud denoising, modeling, deformation analysis and the like.
The specific implementation is as follows:
Based on the method, the inventor spends 3 months to finish 177 scanning operations of the Shanghai rail transit station in a conventional night overhaul time period, the actual limit value of the specified position of 12920 shielding doors is calculated altogether, each shielding door extracts three characteristic positions (the middle and right of the shielding door and the middle of the door frame), the schematic diagram is shown in fig. 16, the distance from the track center line to the edge of the station platform is calculated, the ultra-high data of the curve section track is provided by subway companies, basic data is provided for additionally installing the station platform equipment of the rail transit, and the data is reliable and accurate.
It should be noted that the embodiment of the present application also provides for measurement data accuracy checking:
The method comprises the steps of using a Leica high-precision total station (angle measurement precision 1', prism ranging precision 1 mm+1.5 ppm) to measure two track characteristic points and platform edge characteristic points at corresponding positions of a shielding door on site, measuring three sections on the left, middle and right of the shielding door, calculating the actual limit of a platform, and carrying out data verification on the method. Total station verification measures 11 stations of 6 lines, wherein the number of stations of straight line segments is 5, the number of stations of curve segments is 6, and the limit values of three positions of each shielding door are averaged for comparison during data processing.
Table 1 accuracy analysis table of the method
/>
From fig. 17 and table 1, we can see that the station limit value calculated based on the laser point cloud in the method is compared with the total station, the deviation average value and standard deviation are within 1cm, and the actual limit result precision of the station in the method can reach 1cm.
The above description is only specific to the embodiments of the invention, but the scope of the invention is not limited thereto, and any person skilled in the art who is skilled in the art to which the invention pertains shall apply to the technical solution and the novel concept according to the invention, and shall all be covered by the scope of the invention.
Claims (5)
1. The utility model provides a rail transit platform actual limit measuring method based on laser point cloud, its characterized in that the measuring method includes:
S1, acquiring three-dimensional point cloud data of a station through laser scanning equipment;
S2, generating a point cloud gray level expansion diagram through a projection algorithm, determining a measurement position in the point cloud gray level expansion diagram, and extracting a point cloud section at the measurement position;
S3, in the point cloud section, calculating a track surface perpendicular bisector through identifying and calculating track characteristic points, also calculating a platform screen door outer edge characteristic point at a platform side, and obtaining a straight line section station platform actual limit value D through a point-to-straight line distance formula;
s4, judging whether the station is a curve section station, for the curve section station, because of the existence of the track superelevation, the perpendicular bisector of the track surface is inconsistent with the actual plumb line, correcting the actual limit value D of the station by using the track superelevation value, and obtaining the actual limit value D' of the station of the curve section as a corrected result;
The step S3 specifically includes the following steps:
In the section of the extracted point cloud, the characteristic points of the tracks on two sides are rapidly extracted through a minimum optimization sequence algorithm of a binary heap, and the track center point and the track perpendicular bisector are calculated; the track center point is the midpoint of the two feature points, and the track perpendicular bisector is the perpendicular bisector of the connecting line segment of the two feature points;
The specific algorithm for extracting the characteristic points of the steel rail is as follows:
S3.1.1 in the scanner frame, given the left and right rail positions, two point clouds A1 and A2 are obtained, where parameters are to be customized and optimized in combination with the scanner position;
s3.1.2 building KD tree indexes for the point cloud set A1 and the point cloud set A2 respectively;
S3.1.3 process each point cloud in point cloud A1 and point cloud A2, respectively: setting a search radius, judging the number of points in the drawn space sphere, and if the number of points is lower than a threshold value, considering the points as rough difference points and deleting the rough difference points;
s3.1.4 through step S3.1.3, irrelevant noise points may be deleted, and the denoised point cloud set A1 and point cloud set A2 are processed as follows:
S3.1.4.1 constructing a binary heap minimum priority queue in the zenith direction, wherein the number of the set points in the queue is 20;
S3.1.4.2 from the queue, assuming the zenith direction as a Z coordinate, taking the section direction as an X coordinate, extracting 20 points with the maximum Z coordinate, and calculating the mean value to obtain the Z coordinate of a certain characteristic point of the track;
s3.1.4.3 obtaining a 10 cm thickness point downwards according to the Z coordinate of the characteristic point, and constructing a new binary heap minimum priority queue;
S3.1.4.4 in a new point cloud binary heap minimum priority queue, acquiring 20 points with minimum absolute values of X coordinates, and calculating a mean value to obtain the X coordinates of a certain characteristic point of the track;
s3.1.4.5 two feature points of the left and right rails can be calculated by S3.1.4.1 to S3.1.4.4.
2. The method for measuring the actual limit of the rail transit platform based on the laser point cloud according to claim 1, wherein the step S2 is characterized in that a point cloud gray scale development chart is generated by a projection algorithm, and the method is as follows:
The point cloud is unfolded by taking a track center line o ,x, as a transverse axis and a station section direction o ,y, as a vertical axis, and a subway station plane coordinate system is established; calculating the gray level spread image pixel position of the point cloud according to the point cloud coordinates by using a BMP image with a certain size, filling pixels with the laser reflection value of a laser point, and sequentially obtaining a vehicle-site cloud plane gray level map;
Assuming that the width of the station is w, the height of the station is h, correspondingly generating a BMP chart with the width of wp and the height of hp, and recording the maximum and minimum mileage y in the point cloud data as follows: y max、ymin, the maximum and minimum expansion length l coordinates are: l max,lmin, find the position of each point in the pixel as:
Wherein wp i、hpi is the corresponding position of the point in the image, respectively, (int) means rounding down;
each pixel corresponds to a certain interval range, the number of points in each pixel is inconsistent, and the gray value of the pixel can be calculated according to the following formula:
Wherein n represents the number of points in a pixel, g i is the point gray scale, pg is the pixel gray scale, and g i is the intensity value of a laser point in the point cloud, which is obtained by laser scanning equipment;
The places which are not scanned by the laser and the areas without laser points can cause clear strip black points in the image, in order to filter the black points in the imaging image, smoothing treatment is needed to be carried out on the tiled image, and gray values of pixels without gray points are given in a smoothing mode;
The specific smooth template is as follows:
The determination of the smooth template is to determine the corresponding gray value in the middle of the template according to the reciprocal distance between an unknown pixel and an adjacent pixel, the template specification is 5 rows and 5 columns, if no point exists in a certain pixel, the adjacent pixel with the point is used for weighted average, and the condition that the template is unsuitable for use is that: the top two rows of pixels adopt templates as the lower three rows of the weight matrix; the two bottom rows of pixels adopt templates as the upper three rows of the weight matrix; two columns of pixels at the leftmost side adopt a right three-column matrix with a template as a weight matrix; two columns of pixels at the rightmost side adopt a template as a left three-column matrix of the weight matrix;
The black area in the image is eliminated by a mode of smoothing an unknown pixel template, a clearer station point cloud plane gray level diagram is obtained, and in order to facilitate the viewing of the station point cloud plane gray level diagram, the gray level lower limit is set to 180, so that the image brightness is moderate.
3. The method for measuring the actual limit of the rail transit platform based on the laser point cloud according to claim 2, wherein in the step S2, a measurement position is determined in a point cloud gray scale development chart, and a point cloud section at the measurement position is extracted, specifically as follows:
marking a designated measurement position in the point cloud gray scale development chart, and obtaining point cloud data of the designated position through the projection corresponding relation between the image and the point cloud;
in data processing software, marking a specific position of a platform shielding door by using a red vertical line in a station point cloud gray scale development diagram, converting the image pixel coordinates of the shielding door into shielding door mileage data, namely extracting a station shielding door point cloud section, fitting a section point to a plane, projecting the point to the plane, and further calculating.
4. The method for measuring the actual limit of the rail transit station based on the laser point cloud as set forth in claim 1, wherein the step S3 is specifically as follows:
calculating a rectangular range of the station edge based on the track and station edge position parameters from the track characteristic points calculated in the step S3.1.4.5, searching a plurality of characteristic point sets, wherein the point sets comprise the station edge and other noise points, and carrying out further extraction processing;
for points in the rectangular range, establishing a point cloud index KD tree, and quickly searching for adjacent points of the points; calculating the number of points existing in the radius by setting the position of the search radius, and if the number is less than the set threshold value, considering the points as rough difference points and deleting the rough difference points;
Respectively calculating the distance from the points to the track center line from the rest point set, selecting 20 points with the minimum distance, and taking an average value as a platform edge characteristic point;
The distance from the characteristic point of the platform edge to the center vertical line of the track surface is the station limiting distance D of the section; when the track height is present, the curve height correction is required, and the process proceeds to step S4.
5. The method for measuring the actual limit of the rail transit station based on the laser point cloud as set forth in claim 1, wherein the step S4 is specifically as follows:
If the station is positioned on the line straight line section, the perpendicular bisector of the track surface of the scanning point cloud can be regarded as the plumb line of the track center, and the calculated result is the distance from the center line of the real line to the edge of the station;
If the station is positioned on the curve section of the line, the track surface center plumb line is different from the track center plumb line due to the existence of the track superelevation, and if the distance between the station platform edge on the curve section and the track center plumb line is required to be obtained, the calculation result needs to be subjected to superelevation correction;
According to the geometric relationship, the ultrahigh correction of the curve segment needs the following parameters: the track gauge L defaults to 1435mm, the track is ultrahigh H, the height difference H between the platform edge characteristic points and the track plane center point is the station limiting distance D, the horizontal distance D between the platform edge characteristic points and the track plane center line is the ultrahigh correction factor x, and therefore the horizontal distance D' between the platform edge and the track center plumb line is calculated;
According to the positions of the rail superelevation and the platform edge, the curve section superelevation correction can be determined by a station line flattening graph and can be divided into the following two conditions:
(1) The track is superhigh and approaching the platform side, and the correction calculation is as follows:
(2) The track is superhigh and far away from the platform side, and the correction calculation is as follows:
。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410251135.9A CN117848238B (en) | 2024-03-06 | 2024-03-06 | Rail transit station actual limit measurement method based on laser point cloud |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410251135.9A CN117848238B (en) | 2024-03-06 | 2024-03-06 | Rail transit station actual limit measurement method based on laser point cloud |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117848238A CN117848238A (en) | 2024-04-09 |
| CN117848238B true CN117848238B (en) | 2024-05-14 |
Family
ID=90536336
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410251135.9A Active CN117848238B (en) | 2024-03-06 | 2024-03-06 | Rail transit station actual limit measurement method based on laser point cloud |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117848238B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110398199A (en) * | 2019-07-05 | 2019-11-01 | 内蒙古能建数字信息科技有限公司 | A kind of track clearance detection method |
| CN110986877A (en) * | 2019-12-03 | 2020-04-10 | 中铁第一勘察设计院集团有限公司 | Railway engineering clearance detection method based on high-precision vehicle-mounted laser mobile measurement system |
| CN116128834A (en) * | 2023-02-01 | 2023-05-16 | 河南理工大学 | Rail transit platform limit rapid detection method based on three-dimensional laser scanning |
| WO2023168888A1 (en) * | 2022-03-11 | 2023-09-14 | 中国铁路设计集团有限公司 | Rail superelevation and longitudinal level measurement method based on vehicle-mounted mobile laser point clouds |
| CN117492026A (en) * | 2023-12-29 | 2024-02-02 | 天津华铁科为科技有限公司 | Railway wagon loading state detection method and system combined with laser radar scanning |
-
2024
- 2024-03-06 CN CN202410251135.9A patent/CN117848238B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110398199A (en) * | 2019-07-05 | 2019-11-01 | 内蒙古能建数字信息科技有限公司 | A kind of track clearance detection method |
| CN110986877A (en) * | 2019-12-03 | 2020-04-10 | 中铁第一勘察设计院集团有限公司 | Railway engineering clearance detection method based on high-precision vehicle-mounted laser mobile measurement system |
| WO2023168888A1 (en) * | 2022-03-11 | 2023-09-14 | 中国铁路设计集团有限公司 | Rail superelevation and longitudinal level measurement method based on vehicle-mounted mobile laser point clouds |
| CN116128834A (en) * | 2023-02-01 | 2023-05-16 | 河南理工大学 | Rail transit platform limit rapid detection method based on three-dimensional laser scanning |
| CN117492026A (en) * | 2023-12-29 | 2024-02-02 | 天津华铁科为科技有限公司 | Railway wagon loading state detection method and system combined with laser radar scanning |
Non-Patent Citations (3)
| Title |
|---|
| 地铁隧道横向变形的激光扫描检测方法及应用;吴昌睿;黄宏伟;邵华;;地下空间与工程学报;20200615(03);全文 * |
| 基于移动激光扫描三维点云的隧道限界检测方法;周世明;;四川建材;20180110(01);全文 * |
| 某地铁车站S3-S4轴东西两侧站台侵限原因分析及处理;陶琛;;建材与装饰;20160909(37);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117848238A (en) | 2024-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110647798B (en) | Automatic track center line detection method based on vehicle-mounted mobile laser point cloud | |
| CN110986878B (en) | Method for automatically extracting rail section based on mobile measurement system | |
| CN113870123B (en) | Automatic detection method for contact net leading height and pulling value based on vehicle-mounted mobile laser point cloud | |
| CN111855664B (en) | Adjustable three-dimensional tunnel defect detection system | |
| Yu et al. | 3D reconstruction of road surfaces using an integrated multi-sensory approach | |
| CN107792115A (en) | One kind automatically extracts both wired rail crest level methods using three-dimensional laser point cloud | |
| CN114312877B (en) | Railway comprehensive inspection system | |
| Nex et al. | Automatic roof outlines reconstruction from photogrammetric DSM | |
| CN110660094A (en) | Subway tunnel mobile scanning point cloud fine division method based on image recognition | |
| CN114379598B (en) | Railway comprehensive inspection system | |
| CN114379607B (en) | Comprehensive railway inspection method | |
| CN116465327A (en) | Bridge line shape measurement method based on vehicle-mounted three-dimensional laser scanning | |
| CN115690138A (en) | Road boundary extraction and vectorization method fusing vehicle-mounted image and point cloud | |
| CN110986877B (en) | Railway engineering clearance detection method based on high-precision vehicle-mounted laser mobile measurement system | |
| Du et al. | Dislocation detection of shield tunnel based on dense cross-sectional point clouds | |
| CN112184725A (en) | Structured light strip center extraction method for asphalt pavement image | |
| CN112964195B (en) | Power supply rail geometric parameter comprehensive detection method and system based on laser triangulation method | |
| Neubert et al. | Extraction of railroad objects from very high resolution helicopter-borne LiDAR and ortho-image data | |
| CN114897777A (en) | Full-automatic extraction method of laser point cloud of overhead line system supporting facility considering spatial relationship | |
| CN117848238B (en) | Rail transit station actual limit measurement method based on laser point cloud | |
| CN113722796B (en) | Vision-laser radar coupling-based lean texture tunnel modeling method | |
| CN114119355B (en) | Method and system for early warning of blocking dropping risk of shield tunnel | |
| Jang et al. | Deep learning-based 3D digital damage map of vertical-type tunnels using unmanned fusion data scanning | |
| CN108805790A (en) | A kind of Information acquisition system of the drive way and the lane information processing method using the system | |
| CN111524133B (en) | Window statistical vertex extraction method and system for honeycomb regularity detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |