A metro gauge measurement method based on mobile SLAM laser scanning

Abstract

The present application belongs to the field of measurement technology, and relates to a kind of subway clearance measurement method based on mobile SLAM laser scanning.The method includes the following steps: step 1: encrypting control points based on existing underground control points in subway tunnel;step 2: initializing and three-dimensional laser scanning operation;step 3: point cloud SLAM real-time solution and mapping;step 4: trajectory correction and coordinate conversion of point cloud data;step 5: point cloud denoising and resampling;step 6: line parameter design and clearance drawing report generation.The present application quickly obtains three-dimensional space point cloud data of subway tunnel by mobile SLAM three-dimensional laser scanning equipment, measures existing subway tunnel underground control points and encrypted control points, corrects trajectory and converts coordinates of point cloud based on control point coordinates, and then obtains real space position under absolute coordinate system of subway tunnel.The method is efficient, highly applicable, and the result is reliable, providing an efficient solution for subway tunnel clearance measurement.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of measurement, and relates to a subway clearance measurement method based on mobile SLAM laser scanning. BACKGROUND

[0002] As important infrastructure for urban underground space development and construction, subways can efficiently relieve urban traffic congestion, promote regional economic development and improve the quality of citizens' travel. As an essential part of the subway construction process, clearance measurement and analysis is a necessary condition for track laying and equipment installation. Traditional measurement methods cannot accurately and efficiently obtain three-dimensional spatial information of subway tunnels through single-point mode of total station instruments, which greatly restricts the three-dimensional spatial visualization expression of subway tunnels. The point cloud data obtained by the station-mounted three-dimensional laser scanning method is large in quantity and extremely low in operation efficiency, which to some extent limits its application in subway clearance measurement.

[0003] Therefore, exploring an efficient, accurate and visual subway tunnel clearance measurement method has become a problem to be solved in the current subway tunnel construction stage. SUMMARY

[0004] In order to solve the above problems, the application provides a subway clearance measurement method based on mobile SLAM laser scanning. The method constructs a three-dimensional tunnel map in real time through a SLAM algorithm to obtain spatial point cloud, measures existing subway tunnel underground control points and encrypted control points, corrects the SLAM trajectory to obtain point cloud in the absolute coordinate system, and finally obtains optimal tunnel three-dimensional laser point cloud data through point cloud filtering, denoising and resampling processing, so as to obtain subway clearance measurement overbreak and underbreak elements.

[0005] The technical scheme adopted by the application is: a subway clearance measurement method based on mobile SLAM laser scanning, comprising the following steps:

[0006] Step 1: encrypt control points based on existing underground control points of the subway tunnel, the encrypted control points adopt black and white target reflectors, and the three-dimensional coordinates of the reflectors are obtained by the resection method;

[0007] Step 2: collect point cloud data of the subway tunnel, obtain spatial point cloud data by holding a laser radar of a mobile SLAM, record attitude information by an IMU module, and record spatial environment information by a visual sensor;

[0008] Step 3: construct a three-dimensional environment map of the subway tunnel by real-time laser point cloud scanning matching based on the subway tunnel information recorded by the laser radar sensor and the IMU attitude information, and then perform SLAM solving according to the environment information;

[0009] Step 4: Based on the SLAM algorithm, the continuous spatial point cloud is solved in real time, the point cloud data is corrected and the coordinates are converted by using the collected control points and encrypted control points, and then the overall point cloud in the absolute coordinate system is obtained;

[0010] Step 5: The radius filtering algorithm is applied to remove discrete points and noise points in the point cloud, and the point cloud is resampled according to the set density;

[0011] Step 6: Based on the resampled point cloud, the point cloud data is matched with the designed subway tunnel horizontal curve and vertical curve data, the tunnel section three-dimensional space size data is obtained according to the specified mileage or set mileage interval, and then the limit data of the specified position of the section is obtained.

[0012] The above subway clearance measurement method based on mobile SLAM laser scanning, in step 1, the encrypted control points are based on the existing underground control points, and the spatial distance resection method is used to observe more than three existing underground control points, and the spatial distribution is reasonable and not on the same straight line.

[0013] The above subway clearance measurement method based on mobile SLAM laser scanning, in step 2, the appropriate initialization position refers to a relatively open space environment with low similarity to obtain good initial pose parameters.

[0014] The above subway clearance measurement method based on mobile SLAM laser scanning, in step 3, the real-time laser point cloud scanning matching to construct the three-dimensional environment map of the subway tunnel includes a scanning-based normal distribution transformation algorithm to dynamically update the relative position of the scanner in space.

[0015] The above subway clearance measurement method based on mobile SLAM laser scanning, in step 3, the laser point cloud matching includes a given source point set , voxelization of the target point set , calculation of the mean and variance of the points in each voxel, conversion of the point cloud data in the three-dimensional grid into a continuous differentiable probability density function ;

[0016] The maximum likelihood estimate value of the pose transformation between two frames of point clouds is calculated . .

[0017] The above subway clearance measurement method based on mobile SLAM laser scanning, in step 3, the scanning point cloud matching is performed by iterative calculation of through a nonlinear optimization method, and a three-dimensional environment map of the subway tunnel is constructed in real time.

[0018] The mobile SLAM laser scanning-based metro gauge measurement method, in step 4, the number of collected existing underground control points and encrypted control points is not less than 4, and the point spacing is 40-60 meters.

[0019] The mobile SLAM laser scanning-based metro gauge measurement method, in step 4, the number of collected existing underground control points and encrypted control points is not less than 4, and the point spacing is 40-60 meters. For the laser scanning point cloud, For the control points;

[0020] The least square method is used to calculate The optimal rotation matrix And the translation matrix :

[0021] ; Then the ICP algorithm (iterative closest point algorithm) is used to iteratively calculate the minimum difference between the two sets of point clouds, that is, .

[0022] The mobile SLAM laser scanning-based metro gauge measurement method, in step 5, uses a radius filtering algorithm to remove scattered points and noise points in the point cloud.

[0023] The mobile SLAM laser scanning-based metro gauge measurement method, in step 5, uses a voxel downsampling method to optimize the point cloud density, divides the three-dimensional space where the point cloud is located into a uniform cubic grid (voxel), and each voxel may contain multiple points or be empty. For non-empty voxels, the centroid of the points in the voxel is selected to replace all points in the voxel:

[0024] ;

[0025] In the formula, is the number of points in the voxel, is the coordinate of the th point in the voxel, is the representative point of the voxel.

[0026] The mobile SLAM laser scanning-based metro gauge measurement method, in step 6, the intersection method is used for the design of metro line parameters, the start point, circular curve intersection point and end point coordinates of the horizontal curve are input in turn, and the mileage, flat distance and deviation of the resampled point cloud data can be calculated according to the line design parameters, so that the spatial gauge of the metro tunnel can be analyzed.

[0027] ​The subway clearance measurement method based on mobile SLAM laser scanning has the beneficial effects that: the mobile SLAM three-dimensional laser scanning device is used to quickly obtain three-dimensional space point cloud data of a subway tunnel, the existing underground control points and the encrypted control points of the subway tunnel are measured, the point cloud is corrected and converted based on the control point coordinates, and then the real space position under the absolute coordinate system of the subway tunnel is obtained, and the three-dimensional space point cloud data of the construction tunnel is provided. The mobile SLAM three-dimensional laser scanning is applied to solve the problem of low efficiency of obtaining space data of the subway tunnel; the three-dimensional space point cloud data of the subway tunnel is obtained in real time by mobile three-dimensional laser scanning and based on the SLAM algorithm. The tunnel point cloud data under the absolute coordinate system is matched with the tunnel design parameters, and the subway clearance measurement result can be quickly extracted. The method has high efficiency, strong applicability, reliable result accuracy, and provides an efficient solution for the subway tunnel clearance measurement scene. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 The step flow chart of the subway clearance measurement method based on mobile SLAM laser scanning in the embodiment of the present application is shown in the figure.

[0029] Figure 2 The on-site operation schematic diagram of the present application is shown in the figure.

[0030] Figure 3 The space distance resection schematic diagram is shown in the figure.

[0031] Figure 4 The subway tunnel line design parameter schematic diagram is shown in the figure. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the technical scheme of the present application, the specific implementation of the present application will be described below in combination with the drawings and embodiments.

[0033] In recent years, with the development of three-dimensional laser scanning and SLAM technology, the mobile SLAM three-dimensional laser scanner has become an important tool for obtaining underground space data. In order to solve the problem that the traditional means of subway tunnel clearance measurement is low in efficiency, it is difficult to obtain real three-dimensional space data of the subway tunnel accurately and comprehensively, and sometimes repeated measurement is necessary, the present application proposes a mobile SLAM three-dimensional laser method for quickly obtaining three-dimensional space point cloud data of the subway tunnel, the point cloud data is corrected and converted in combination with the existing underground control points and the encrypted control points of the subway tunnel, and then the point cloud data under the absolute coordinate system is quickly obtained, and the clearance measurement result of any mileage position can be obtained based on the resampled point cloud and the tunnel line design parameter. The method at least includes the following steps:

[0034] 1. Underground encryption control point measurement. Based on the existing underground control points of the subway tunnel, the spatial distance resection method is used to encrypt the underground control points, and the control point interval should not exceed 60 meters, and the control points should not be on a straight line.

[0035] 2. Initialization and three-dimensional laser scanning operation. Select a place with low spatial similarity to start the mobile SLAM scanning device for initialization, the initialization time is ≥20s, and scanning along the planned route is performed. When encountering underground control points and encryption control points, control point collection is performed until the scanning is completed.

[0036] 3. In the process of mobile scanning, the laser radar collects three-dimensional spatial point cloud data of the subway tunnel, the IMU records the attitude information, and the visual sensor records the environmental image information. By extracting environmental feature points through the visual sensor, combining three-dimensional laser point cloud data, establishing spatial feature information association, and then based on the normal distribution transformation algorithm of scanning, the relative position of the scanner in space is dynamically updated to construct a three-dimensional spatial map of the subway tunnel.

[0037] 4. Trajectory correction and coordinate conversion are performed using least squares and ICP algorithm to obtain three-dimensional point cloud in absolute coordinate system. Least squares method is used for point cloud data conversion, and the collected underground control point coordinates are used as absolute reference. ICP algorithm is used for point cloud data trajectory correction to obtain the best point cloud data.

[0038] 5. Point cloud data denoising and resampling. Radius filtering algorithm is used to remove discrete points and noise points in the point cloud to obtain clean and usable point cloud data. Voxelization downsampling algorithm is used for point cloud data resampling to obtain lightweight point cloud data, which is convenient for subsequent limit data analysis.

[0039] 6. Limit analysis diagram and report production. Based on the resampled point cloud, subway tunnel line parameter design is performed in the tunnel point cloud special processing software TMO. The point cloud coordinates can be matched to the line to obtain the mileage, flat distance and deviation of any point of the point cloud data, and then the subway limit measurement overbreak and underbreak section drawing and subway limit analysis measurement report are generated.

[0040] Example 2, as shown in Figures 1-3 , a subway limit measurement method based on mobile SLAM laser scanning, comprising the following steps:

[0041] Step 1: Based on the existing underground control points of the subway tunnel, the spatial distance resection measurement method is used to encrypt the underground control points. The underground encryption control points are made of 2cm×2cm black and white reflective targets, and the encryption control point coordinates are obtained by resection. The coordinates of three known control points 、 、 , three existing control points are not on the same straight line, as shown in Figure 3 , the obtained encrypted control point coordinates are:

[0042] (1);

[0043] In formula (1), ;

[0044] , , are the internal angles of the triangle formed by the known control points, respectively , , are three horizontal angles observed at point.

[0045] Step 2: scanning preparation and initialization. Select a suitable starting position (a relatively open space environment with low similarity), perform pre-scanning preparation, connect the laser scanning host with the controller, image acquisition device and controller, start the device for initialization scanning, obtain the feature points and image data of the environment around the starting position, and the static time is not less than 20s. After initialization, the handheld SLAM laser scanner completes scanning along the predetermined route, and when encountering underground control points and encrypted control points, the SLAM scanning device is placed on the control points for recording. The SLAM laser scanner moves to collect subway tunnel space point cloud data, the camera collects image data, and the visual sensor records environmental image information.

[0046] Step 3: real-time point cloud SLAM solving and mapping. Feature points in the subway tunnel environment are extracted through the visual sensor, combined with laser radar point cloud features and IMU attitude data, a relative coordinate system associated with spatial features is established, and SLAM solving is performed according to real-time point cloud feature matching. The establishment of the relative coordinate system associated with spatial features includes a scanning-based normal distribution transformation algorithm, dynamic updating of the relative position of the scanner in space, and construction of a three-dimensional space map of the subway tunnel.

[0047] During data acquisition, the laser radar collects spatial environment point cloud data, the IMU collects attitude data during walking, and the visual sensor records environmental feature graphic information. Through the visual sensor, feature points and point cloud features are collected, combined with IMU attitude data, and a spatial environment correlation is established. In this process, the positioning core is to solve the real-time matching problem of laser point cloud features. Laser point cloud matching includes given source point set and target point set , voxelization of the target point set , calculation of the mean and variance of the points inside each voxel (The transpose symbol) transforms point cloud data in a 3D mesh into a continuously differentiable probability density function. (2);

[0048] Determine the pose transformation between two frames of point clouds. Maximum likelihood estimate (3).

[0049] The nonlinear optimization method was used to... Iterative calculations are performed to match the scanned point cloud and construct a 3D environmental map of the subway tunnel in real time.

[0050] Step 4: Trajectory Correction and Coordinate Transformation Based on Control Points. Existing underground control points and densified control points acquired during the mobile scanning process can be used as 3D spatial references to correct the trajectory of the real-time constructed 3D spatial point cloud data, obtaining the optimal point cloud in the absolute coordinate system. The number of existing underground control points and densified control points acquired should be no less than four, with a spacing of 40-60 meters.

[0051] make For laser scanning point clouds, For control points; calculate using the least squares method. Switch to Optimal rotation matrix Translation matrix :

[0052] (4).

[0053] Then, the ICP algorithm (Iterative Closest Point Algorithm) is used to iteratively calculate the minimum difference between corresponding points in the two point clouds, i.e.: (5).

[0054] Step 5: Point Cloud Denoising and Resampling. A radius filtering algorithm is applied to remove discrete and noise points from the obtained point cloud. For each point in the point cloud... Calculation points Compared with other points in the point cloud Euler distance:

[0055] (6).

[0056] Specifically, the neighborhood point threshold is set to... Each time, take the current point as the center and determine a radius of... Given a sphere. Calculate the number of neighboring points within the current sphere; if the number is greater than... If the condition is met, the point is retained; otherwise, it is discarded.

[0057] A voxelization downsampling method is used to optimize the point cloud density. The 3D space of the point cloud is divided into a uniform cubic mesh (voxels). Each voxel may contain multiple points or be empty. For non-empty voxels, the centroid of the points within the voxel is used to replace all points within the voxel.

[0058] (7);

[0059] In equation (4), It is the number of intravoxel points. It is the first voxel. The coordinates of the points These are the representative points of voxels. All voxel representative points are output as downsampled point cloud data.

[0060] Step 6: Line Parameter Design and Clearance Map Report Generation. The subway tunnel line parameters are designed in the dedicated tunnel point cloud processing software TMO. Horizontal curves are input sequentially as follows: Figure 4 The system inputs the coordinates of straight lines or ZH points, circular curve intersections (JD), and endpoints. For vertical curves, it inputs the elevation and radius of the slope change points, along with parameters such as centerline offset and chainage breakage. The resampled point cloud is then imported into the software for line parameter matching. Based on the designed clearance section type, the point cloud data is segmented to extract clearance data at arbitrary mileage and location, generating clearance analysis maps and reports.

[0061] The above embodiments are merely illustrative of the ideas and features of the present invention, intended to enable those skilled in the art to fully understand the content of the present invention and implement it accordingly, and should not be construed as limiting the scope of protection of the present invention. Any equivalent substitutions or modifications made according to the concept and content of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A metro gauge measurement method based on mobile SLAM laser scanning, characterized in that, The method comprises the following steps: Step 1: encrypting control points based on existing underground control points in a subway tunnel, the encrypted control points adopting black-and-white target reflectors, and the three-dimensional coordinates of the reflectors being obtained by resection; Step 2: selecting an open space environment with low similarity for initialization, collecting point cloud data of the subway tunnel, obtaining spatial point cloud data by a handheld mobile SLAM laser radar, recording attitude information by an IMU module, and recording spatial environment information by a visual sensor; Step 3: constructing a three-dimensional environment map of the subway tunnel by real-time laser point cloud scanning matching based on the subway tunnel information recorded by the laser radar sensor and the IMU attitude information, and then performing SLAM calculation according to the environment information; Step 4: continuously solving spatial point clouds by the SLAM algorithm, correcting the trajectory and converting the coordinates of the point cloud data by using the collected existing control point and encrypted control point coordinates, and then obtaining overall point clouds in an absolute coordinate system; Step 5: removing discrete points and noise points in the point clouds by applying a radius filtering algorithm, and resampling the point clouds according to a set density; Step 6: matching the point cloud data with designed subway tunnel horizontal curve and vertical curve data based on the resampled point clouds, obtaining three-dimensional spatial size data of a tunnel section according to a specified mileage or a set mileage interval, and then obtaining limit data at a specified position of the section.

2. The metro gauge measurement method based on mobile SLAM laser scanning according to claim 1, characterized in that: In step 1, the encrypted control points are based on the existing underground control points, and three or more existing underground control points are observed by a spatial distance resection method, and the control points are reasonably distributed in space and not on the same straight line.

3. The metro gauge measurement method based on mobile SLAM laser scanning according to claim 1, characterized in that: In step 3, the laser point cloud matching includes voxelizing the target point set given a source point set , computing the mean and variance of the points inside each voxel , converting the point cloud data in a three-dimensional grid into a continuous differentiable probability density function ; Finding maximum likelihood estimate of pose transformation between two frames of point clouds .​ 4. The metro gauge measurement method based on mobile SLAM laser scanning according to claim 3, characterized in that: In step 3, a nonlinear optimization method is used to optimize the... Iterative calculations are performed to match the scanned point cloud and construct a 3D environmental map of the subway tunnel in real time.

5. The metro gauge measurement method based on mobile SLAM laser scanning according to claim 1, characterized in that: In step 4, the number of the collected existing underground control points and encrypted control points is not less than four, and the point spacing is 40-60 meters.

6. The metro gauge measurement method based on mobile SLAM laser scanning according to claim 1, characterized in that: In step 4, trajectory correction and coordinate conversion are performed based on the control points, and the control points are obtained by are laser scanning point clouds, are control points; least squares method transformed to optimal rotation matrix and translation matrix : ； An iterative closest point algorithm is used to iteratively calculate the minimum difference between the corresponding points of the two sets of point clouds, i.e.: .

7. The metro gauge measurement method based on mobile SLAM laser scanning according to claim 1, characterized in that: In step 5, the point cloud density is optimized by a voxel downsampling method, the three-dimensional space where the point cloud is located is divided into uniform cubic grids, each voxel may contain multiple points or be empty, and for a non-empty voxel, the centroid of the points in the voxel is selected to replace all the points in the voxel: ； wherein is the number of points within the voxel, is the coordinate of the point within the voxel, is the representative point of the voxel.

Images