Point cloud density adaptive laser scanning method for large-span bridge height measurement

Abstract

The application provides a laser method for large-span bridge height measurement point cloud density self-adaption. Firstly, 360-degree rapid scanning is carried out at a given station to obtain full-field information, and the TLS station pose is estimated by registering the scanning point cloud with the known design model; the measurement object is labeled in the design model in advance, the preset scanning point spacing requirement value is set, the minimum angular resolution of each point of the measurement object is solved based on the laser beam propagation principle, the measurement object point cloud is grouped according to the minimum angular resolution and the angular coordinate, and the angle range and the minimum angular resolution of each group of point sets are extracted; according to the grouping information of the measurement object point cloud, the sectional fine scanning with variable angular resolution is carried out to obtain the measurement object point cloud model, which is used for geometric measurement. The application can significantly improve the scanning efficiency in a large scene and reduce the point cloud volume.

Description

Technical Field

[0001] This invention belongs to the field of bridge inspection, and in particular relates to an adaptive laser scanning method for point cloud density for elevation measurement of long-span bridges. Background Technology

[0002] Monitoring the geometric alignment during the phased construction of long-span bridges is a crucial measure to ensure bridge safety and meet design requirements upon completion. Currently, total stations and levels are primarily used for measuring the alignment of bridge towers and main beams during bridge construction. However, as modern bridges develop towards longer spans and higher piers and towers, increasingly stringent requirements are being placed on bridge construction surveying. The shortcomings of traditional surveying methods, primarily using total stations, are becoming increasingly apparent. These include low surveying efficiency, the large geometric dimensions of the bridge deck and towers and the dense concentration of measuring points on long-span bridges, making point-by-point measurement with total stations time-consuming; and the fact that measurements for critical bridge construction stages are often completed at night or in the early morning, resulting in heavy workloads and high labor intensity.

[0003] Standing laser scanners (TLS) offer advantages in 3D high-resolution and wide-field-of-view measurements; however, current measurement processes often rely on experienced surveying engineers to set scanning parameters and control data acquisition. The fixed angular resolution scanning mode used in TLS can lead to uneven point cloud distribution, with denser points closer to the target area and sparser points further away. This is particularly problematic for long-span bridges, where significant differences in distance between objects within a single site's scanning range can result in excessively dense point clouds and data redundancy, reducing the efficiency of subsequent data transmission and processing. Therefore, it is necessary to establish a point cloud density-adaptive scanning mode for bridge scenarios to improve measurement efficiency and point cloud quality. Summary of the Invention

[0004] The purpose of this invention is to propose an adaptive laser scanning method for point cloud density for elevation measurement of long-span bridges, which can improve measurement efficiency and point cloud quality.

[0005] Technical solution:

[0006] 1. An adaptive laser scanning method for point cloud density for elevation measurement of long-span bridges, comprising the following steps:

[0007] S1: Conduct a rapid coarse scan of the bridge scene to obtain the position and orientation information of the TLS itself;

[0008] S11: Conduct a rapid, rough scan of the bridge scene;

[0009] Set up the TLS and tripod at the given location, set the relevant parameters, and carry out a rapid scan to obtain a coarse point cloud model of the long-span bridge scene at the given site.

[0010] S12: Given a structural design model of a long-span bridge, the main beam region in the model is pre-labeled. The design model is uniformly sampled using the Poisson disk sampling algorithm to obtain the point cloud model of the long-span bridge. The point cloud model of the long-span bridge is registered with the coarse-scanned point cloud model using point cloud geometric homogeneity decomposition and iterative nearest point algorithm. The spatial rigid body transformation matrix M from the TLS coordinate system to the bridge structure coordinate system is solved, and the six-degree-of-freedom position and attitude of the TLS in the bridge structure coordinate system is solved based on M.

[0011] S13: Based on the principle of ray tracing, and using a six-DOF position and attitude model and a structural design model of a long-span bridge based on TLS, laser scanning simulation is conducted to generate a simulated point cloud model.

[0012] The pre-marked main beam regions in the structural design model of the long-span bridge are mapped onto the simulated point cloud model to obtain the set of visible main beam points {P} in the TLS coordinate system. i (x i ,y i ,z i )}, which is the set of points {P} of the measurement object. i (x i ,y i ,z i )};

[0013] S2: A segmented scanning scheme for adaptive point cloud density calculation;

[0014] S21: TLS for measuring object point set {P} i (x i ,y i ,z i Horizontal scan angular resolution Vertical scanning angular resolution The calculation method;

[0015] Based on the principle of laser beam propagation, TLS in the azimuth angle The lower point spacing The calculation formula is:

[0016]

[0017] TLS generates a point spacing s at a pitch angle Δθ. θ The calculation formula is:

[0018] s θ =Htan(β+Δθ)-H i tan(β) (2)

[0019] Where L and H are the horizontal and vertical distances from the measurement point P(x,y,z) to the TLS scanning station, and β = arctan(L / H) is the azimuth angle. This is the required horizontal scanning angular resolution value, and the pitch angle Δθ is the required vertical scanning angular resolution value.

[0020] in,

[0021] Based on the measurement task requirements, the preset point spacing requirements are determined. and The specific coordinates of the points being measured are also known. Therefore, according to formulas (1) and (2), the coordinates of any point P in the set of points being measured can be calculated. i (x i ,y i ,z i Satisfying the point spacing and Horizontal scanning angular resolution requirement under the given conditions Vertical scanning angular resolution requirement Point P i (x i ,y i ,z i ) Horizontal distance L to TLS scanning site, point P i (x i ,y i ,z i The vertical distance H from the TLS scanning site can be calculated using formula (3);

[0022] S22: Solve for the horizontal scanning angular resolution corresponding to the measurement object point. Minimum value and vertical scan angle resolution Find the minimum value and group them:

[0023] The set of measurement object points {P} in the TLS coordinate system of S13 is... i (x i ,y i ,z i )} by Cartesian coordinate system (x i ,y i ,z i Convert to spherical coordinates That is, distance r i Pitch angle θ i and azimuth

[0024] And calculate the pitch angle θ of all measured object points. i and azimuth

[0025] By comparison, the maximum pitch angle θ was calculated. max Minimum value θ min The maximum value of the azimuth angle Minimum value

[0026] pitch angle θ i The maximum value θ max Minimum value θ min The scanning range is for the vertical angle, and the azimuth angle is... maximum value Minimum value The scanning range is defined by the horizontal angle, with the pitch angle θ as the reference point. i The scanning direction corresponding to the scanning interval is the vertical axis, and the azimuth angle is... The scanning direction corresponding to the scanning interval is the horizontal axis. A preset scanning angle interval value is used to draw a two-dimensional plane containing several grids.

[0027] The measurement object points within each grid are grouped together. According to step S21, the vertical angular resolution Δθ and the horizontal scanning angular resolution corresponding to the measurement object points within each grid are calculated. By comparison, the minimum vertical scanning angular resolution and the minimum horizontal scanning angular resolution within each grid were statistically determined.

[0028] S23: Merge the meshes to form a segmented two-dimensional plane;

[0029] A preset angular resolution difference is set. For each grid, the minimum value of its scanning angular resolution is compared with the minimum value of the scanning angular resolution of its 8 surrounding grids. The difference between the minimum values ​​of the horizontal scanning angular resolution and the minimum value of the vertical scanning angular resolution is calculated. If the sum of the two differences is lower than the preset angular resolution difference, then every two grids that meet the requirements are merged. After merging the grids, the original two-dimensional plane containing several grids is divided into a two-dimensional plane containing several segments.

[0030] By comparison, the minimum vertical scanning angular resolution and the minimum horizontal scanning angular resolution within each segment are determined.

[0031] S24: Save the minimum and maximum values ​​of the azimuth, elevation, and vertical and horizontal scanning angular resolutions within each segment as six scanning parameters. Based on these six parameters, scan the corresponding segment to measure the point set {P}. i (x i ,y i ,z i A segmented scanning scheme is formed.

[0032] S3: Conduct segmented fine scanning of the object being measured.

[0033] Furthermore, in step S3, a fine scan is performed on each segment to obtain the point cloud model of the measurement point set of the long-span bridge, and geometric measurements and verification are performed, including the following steps:

[0034] S31: Input the six scanning parameters within each segment into the TLS device and perform a detailed scan one by one;

[0035] S32: Merge the multiple point cloud models generated after scanning, and transform the merged point cloud model from the TLS coordinate system to the bridge structure design coordinate system according to the rigid body transformation matrix M;

[0036] S33: Measuring points are set at fixed intervals along the longitudinal axis of the bridge;

[0037] A point cloud is selected within a 4m×4m rectangular area centered on the measuring point;

[0038] Based on the histogram, the elevation distribution of all measurement points on the bridge surface of the long-span bridge is statistically analyzed. The frequency values ​​of elevation occurrence are smoothed by zero-phase low-pass filtering, and the elevation value corresponding to the peak frequency value after smoothing is recorded as the elevation of the corresponding measurement point.

[0039] Elevation statistics were performed on all measuring points in sequence to obtain the elevation curve of the main beam along the longitudinal axis. The obtained elevation curve was then compared with the design elevation curve.

[0040] Furthermore, in step S11, the horizontal scanning angle resolution and the vertical scanning angle resolution are set to the same value, both below 0.05 degrees, the horizontal scanning angle is 0 to 360 degrees, and the vertical scanning angle is set to -40 to 60 degrees.

[0041] Furthermore, in step S22, the preset scanning angle interval is 2-5 degrees, and the pitch angle θ i and azimuth The scanning angle interval values ​​are the same.

[0042] Furthermore, in step S31, the scanning range is determined by the minimum and maximum values ​​of the azimuth angle within each segment to determine the range of the TLS horizontal scanning angle, and by the minimum and maximum values ​​of the elevation angle within each segment to determine the range of the TLS vertical scanning angle.

[0043] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0044] 1. The position and orientation of the TLS are calibrated by full-field surround scanning, and then the measurement object is located in the scanner's own coordinate system, realizing data acquisition oriented towards the specified measurement object and overcoming the blindness and subjectivity of the scanning process;

[0045] 2. Establish a segmented variable-angle resolution scanning method with adaptive point cloud density to solve the problems of excessively dense near-field point clouds and data redundancy caused by the existing fixed-angle resolution scanning mode. Attached Figure Description

[0046] Figure 1 This is a flowchart of the method of the present invention;

[0047] Figure 2 This is a site layout diagram of the TLS equipment mounted on the crossbeam of the bridge tower;

[0048] Figure 3 It is a rough point cloud map of the site;

[0049] Figure 4 It is a point cloud model of bridge structure design using Poisson disk sampling;

[0050] Figure 5 It involves point cloud registration between the design point cloud model and the coarse-scanned point cloud model;

[0051] Figure 6 This is a schematic diagram of the main beam point set visible under the TLS site;

[0052] Figure 7 It is the point spacing generated by TLS on the plane under horizontal and vertical unit angles;

[0053] Figure 8 It is the two-dimensional grid of azimuth and elevation angles, and the grid to which the set of points of the measurement object belongs;

[0054] Figure 9 It is the grouping result of the measurement object point set in the TLS's own coordinate system;

[0055] Figure 10 It is a merged, finely scanned point cloud model;

[0056] Figure 11 This is a statistical analysis of the point cloud elevation histogram corresponding to the measuring point 103m away from the bridge tower.

[0057] Figure 12 This is an elevation diagram of the main beam along the longitudinal axis. Detailed Implementation

[0058] The following example, using the measurement of the deck elevation of a long-span cable-stayed bridge, illustrates the specific implementation of this invention.

[0059] This invention is based on an adaptive laser scanning method for point cloud density for elevation measurement of long-span bridges, which can be used to measure the elevation of the bridge deck of long-span bridges. A fixed location is pre-selected on the long-span bridge, a tripod and a Riegl Vz-400i TLS device are set up, and the long-span bridge is rapidly scanned at this fixed location to obtain a coarse point cloud model.

[0060] For the given structural design model of the long-span bridge, the main beam region is pre-marked in the structural design model. The design model is transformed into a point cloud model showing the main beam region using a uniform sampling method. The coarse-scanned point cloud model is registered with the design point cloud model showing the main beam region, and the spatial rigid body transformation matrix M from the TLS's own coordinate system to the bridge structure coordinate system, as well as the six-degree-of-freedom position and attitude of the TLS in the bridge structure coordinate system, are solved.

[0061] Based on the six-degree-of-freedom position and attitude of the TLS in the bridge structural coordinate system and the structural design model of the long-span bridge, a scanning simulation of the structural design model is performed to obtain a simulated point cloud model. Then, the main beam region in the structural design model is mapped onto the simulated point cloud model to obtain the set of main beam points visible under the TLS site {P}. i (x i ,y i ,z i The main beam point set is the measurement object point set.

[0062] The next step is to design a scanning scheme for the object being measured.

[0063] Based on the measurement task requirements, the preset point spacing requirements are determined. and Given the main beam point set {P i (x i ,y i ,z i Given the coordinates of the points {P}, the main beam point set {P} can be calculated using the formula. i (x i ,y i ,z i Horizontal scanning angular resolution at any point in )} that meets the point spacing requirement Vertical scanning angular resolution

[0064] For the main beam point set {P i (x i ,y i ,z i The main beam area is divided into several grids according to the corresponding requirements. All main beam points are assigned to the corresponding grids. Then, the grids are merged according to the requirements. After the grids are merged, the main beam area that was originally several grids becomes a main beam area with several segments.

[0065] Scanning parameters are set for each segment of the main beam region, and each segment is scanned according to different scanning parameters. This forms a scanning scheme for the main beam point set.

[0066] An adaptive laser scanning method for point cloud density for elevation measurement of long-span bridges mainly includes the following steps:

[0067] S1: Conduct a rapid coarse scan of the bridge scene to obtain the position and orientation information of the TLS itself;

[0068] S11: Conduct a rapid, rough scan of the bridge scene;

[0069] On the bridge tower crossbeam, 80m above the bridge deck, tripods and TLS equipment (Riegl Vz-400i) were erected. Figure 2 The horizontal and vertical scanning angular resolution settings are the same, and can be set to below 0.05 degrees. The horizontal scanning angle is set to 0 to 360 degrees, and the vertical scanning angle is set to -40 to 60 degrees. A rapid surround scan of the scene is performed to obtain a coarse point cloud model of the scene. See [link / reference]. Figure 3 ;

[0070] S12: Given a structural design model of a long-span bridge, the main beam region in the model is pre-labeled. A Poisson disk sampling algorithm is used to uniformly sample the design model, resulting in a point cloud model of the design model. (See...) Figure 4 Point cloud geometric homogeneity decomposition and iterative nearest-point algorithm are used to register the point cloud model of the design model with the coarse-scanned point cloud model. (See...) Figure 5 The spatial rigid body transformation matrix M from the TLS's own coordinate system to the bridge structure coordinate system is solved, and the six-degree-of-freedom position and attitude of the TLS in the bridge structure coordinate system are also solved. The position is [2.86, 1.64, 81.36] m, and the Euler angles of the attitude are [-175.42, 0.66, -0.38]. o ;

[0071] S13: Based on the principle of ray tracing, and using the six-DOF position and attitude model of the long-span bridge and the structural design model of the TLS (Through-Loop Tracing), laser scanning simulation is performed on the structural design model of the long-span bridge to generate a simulated point cloud model. The main beam region in the structural design model of the long-span bridge is mapped onto the simulated point cloud model to obtain the set of main beam points visible under the TLS site {P}. i (x i ,y i ,z i )} is considered as the set of points to be measured, see Figure 6 ;

[0072] S2: A segmented scanning scheme for adaptive point cloud density calculation based on the structural design model of a long-span bridge and station pose information;

[0073] S21: TLS for measuring object point set {P} i (x i ,y i ,z iHorizontal scan angular resolution Vertical scanning angular resolution For the calculation method, see Figure 7 ;

[0074] Based on the principle of laser beam propagation, TLS in the azimuth angle The lower point spacing The calculation formula is:

[0075]

[0076] TLS generates a point spacing s at a pitch angle Δθ. θ The calculation formula is:

[0077] s θ =Htan(β+Δθ)-Htan(β) (2)

[0078] Where L and H are the horizontal and vertical distances from the measured object point P(x,y,z) to the TLS scanning station, and β = arctan(L / H), the azimuth angle. This represents the required horizontal scanning angular resolution, while the pitch angle Δθ represents the required vertical scanning angular resolution.

[0079] in,

[0080] Based on the measurement task requirements, the preset point spacing requirements are determined. and The specific coordinates of the points being measured are also known. Therefore, according to formulas (1) and (2), the coordinates of any point P in the set of points being measured can be calculated. i (x i ,y i ,z i Satisfying the point spacing and Required horizontal scanning angular resolution value Vertical scanning angular resolution requirement Point P i (x i ,y i ,z i ) Horizontal distance L to TLS scanning site, point P i (x i ,y i ,z i The vertical distance H from the TLS scanning site can be calculated using formula (3);

[0081] S22: Solve for the horizontal scanning angular resolution corresponding to the measurement object point. Minimum value and vertical scan angle resolution Find the minimum value and group them:

[0082] The set of measurement object points {P} in the TLS coordinate system of S13 is... i (x i ,y i ,z i )} by Cartesian coordinate system (x i ,y i ,z i Convert to spherical coordinates That is, distance r i Pitch angle θ i and azimuth And calculate the pitch angle θ of all measured object points. i and azimuth The maximum pitch angle θ was calculated. max Minimum value θ min The maximum value of the azimuth angle Minimum value

[0083] pitch angle θ i The maximum value θ max Minimum value θ min The scanning interval for the vertical angle is the azimuth angle. maximum value Minimum value The two scanning intervals are the horizontal angle intervals, which together constitute the angle intervals during TLS scanning.

[0084] With pitch angle θ i The scanning direction corresponding to the scanning interval is the vertical axis (i.e., the length direction of the collapsed bridge), and the azimuth angle is... The scanning direction corresponding to the scanning interval is the horizontal axis (i.e., the width direction of the bridge). A preset scanning angle interval value and a pitch angle θ are used. i and azimuth The scanning angle interval values ​​are the same, and the scanning angle interval value can be selected between 2 and 5 degrees, with a preferred angle interval value of 2 degrees. A two-dimensional plane containing several grids is drawn, see... Figure 8 ;

[0085] The measurement object points within each grid are grouped together. According to step S21, the vertical scanning angular resolution Δθ and the horizontal scanning angular resolution corresponding to the measurement object points within each grid are calculated. By comparison, the minimum vertical scanning angular resolution and the minimum horizontal scanning angular resolution within each grid were statistically determined.

[0086] S23: Merge the meshes to form a segmented two-dimensional plane;

[0087] A preset angular resolution difference is used to compare the minimum scanning angular resolution of each grid with the minimum scanning angular resolution of its eight neighboring grids, and then calculate the minimum horizontal scanning angular resolution. The difference and the minimum vertical scanning angular resolution Δθ min If the sum of the two differences is lower than the preset angular resolution difference, then every two grids that meet the requirement will be merged. After merging the grids, the original two-dimensional plane containing several grids is divided into a two-dimensional plane containing several segments.

[0088] By minimizing the horizontal scan angular resolution within each grid and the minimum vertical scanning angular resolution Δθ min By comparing them, we can obtain the minimum horizontal scanning angular resolution and the minimum vertical scanning angular resolution in each segment of the two-dimensional plane.

[0089] S24: Save the minimum and maximum azimuth angles, minimum and maximum elevation angles, minimum vertical and horizontal scanning angular resolutions, and minimum horizontal scanning angular resolutions within each segment as six scanning parameters. Based on these six scanning parameters, scan the corresponding segments to form a segmented scanning scheme; see [link to relevant documentation]. Figure 9 ;

[0090] S3, Perform segmented fine scanning of the object being measured;

[0091] In step S3, a fine scan is performed on each segment to obtain the point cloud model of the measurement object set of the long-span bridge, and geometric measurements and verification are performed, including the following steps:

[0092] S31: Input the scanning parameters of each segment into the TLS device using a programming language, and perform fine scanning one by one. Determine the range of the TLS horizontal scanning angle by the minimum and maximum values ​​of the azimuth angle in each segment, and determine the range of the TLS vertical scanning angle by the minimum and maximum values ​​of the pitch angle in each segment.

[0093] S32: Merge the multiple point cloud models generated after scanning, and perform geometric transformations according to the rigid body transformation matrix M to transform the merged point cloud model from the TLS coordinate system to the bridge structure design coordinate system; see Figure 10 ;

[0094] S33: Elevation measuring points are set at 5m intervals along the longitudinal axis of the bridge; a point cloud is selected within a 4m×4m rectangular area centered on the measuring point, and the distribution of elevations of all points is statistically analyzed based on the histogram. Zero-phase low-pass filtering is used to smooth the frequency values ​​of elevation occurrences, and the elevation value corresponding to the peak frequency after smoothing is recorded as the elevation of that measuring point. See [link to relevant documentation]. Figure 11Elevation statistics were performed on all measuring points sequentially to obtain the elevation curve of the main beam along the longitudinal axis, as shown in the figure. Figure 12 The obtained elevation curves are then compared with the design elevation curves.

Claims

1. A point cloud density adaptive laser scanning method for elevation measurement of long-span bridges, characterized in that, Includes the following steps: S1: Conduct a rapid coarse scan of the bridge scene to obtain the position and orientation information of TLS itself; S11: Conduct a rapid, rough scan of the bridge scene; Set up the TLS and tripod at the given location, set the relevant parameters, and carry out a rapid scan to obtain a coarse point cloud model of the long-span bridge scene at that given location. S12: Given a structural design model of a long-span bridge, the main beam region in the model is pre-labeled. The design model is uniformly sampled using the Poisson disk sampling algorithm to obtain the point cloud model of the design model. The point cloud model of the design model is registered with the coarse-scanned point cloud model using point cloud geometric homogeneity decomposition and iterative nearest point algorithm. The spatial rigid body transformation matrix M from the TLS's own coordinate system to the bridge structure coordinate system is solved, and the six-degree-of-freedom position and attitude of the TLS in the bridge structure coordinate system is solved based on M. S13: Based on the principle of ray tracing, and using a six-DOF position and attitude model and a structural design model of a long-span bridge based on TLS, laser scanning simulation is conducted to generate a simulated point cloud model. Mapping the pre-marked main girder regions in the structural design model of the long-span bridge onto the simulated point cloud model yields the set of visible main girder points {P} in the TLS coordinate system. i (x i , y i , z i )}, which is the set of points {P} of the measurement object. i (x i , y i , z i )}; S2: A segmented scanning scheme for adaptive point cloud density calculation; S21: TLS for measuring object point set { P i (x i , y i , z i Horizontal scan angular resolution Vertical scanning angular resolution The calculation method; Based on the principle of laser beam propagation, TLS generates a point spacing s at an azimuth angle Δφ. φ The calculation formula is: s φ =2Lsin( / 2) (1) TLS generates a point spacing s at a pitch angle Δθ. θ The calculation formula is: s θ =Htan(β+Δθ)-Htan(β) (2) Where L and H are the horizontal and vertical distances from the measurement object point P(x, y, z) to the TLS scanning station, β = arctan(L / H), the azimuth angle Δφ is the required horizontal scanning angular resolution, and the elevation angle Δθ is the required vertical scanning angular resolution. Where L= H = |z| (3); Based on the measurement task requirements, the preset point spacing requirements are determined. and Then the specific coordinates of the measurement object point are also known. Therefore, according to formulas (1) and (2), the coordinates of any point P in the set of measurement object points can be solved. i (x i , y i , z i Satisfying the point spacing and Horizontal scanning angular resolution requirement under the given conditions Vertical scanning angular resolution requirement value Point P i (x i , y i , z i ) Horizontal distance L to TLS scanning site, point P i (x i , y i , z i The vertical distance H from the TLS scanning site can be calculated using formula (3); S22: Solve for the horizontal scanning angular resolution corresponding to the measurement object point. Minimum value and vertical scan angle resolution Find the minimum value and group them: The set of measurement object points {P} in the TLS coordinate system of S13 is... i (x i , y i , z i )} by Cartesian coordinate system (x i , y i ,z i Convert to spherical coordinates (r) i , θ i , φ i ), that is, distance r i Pitch angle θ i and azimuth φ i , And calculate the pitch angle θ of all measured object points. i and azimuth φ i , By comparison, the maximum pitch angle θ was calculated. max Minimum value θ min The maximum value of the azimuth angle φ max Minimum value φ min ; pitch angle θ i The maximum value θ max Minimum value θ min The scanning range is the vertical angle, and the azimuth angle is φ. i maximum value φ max Minimum value φ min The scanning range is defined by the horizontal angle, with the pitch angle θ as the reference point. i The scanning direction corresponding to the scanning interval is the vertical axis, and the azimuth angle φ i The scanning direction corresponding to the scanning interval is the horizontal axis. A preset scanning angle interval value is used to draw a two-dimensional plane containing several grids. The measurement object points in each grid are grouped together. According to step S21, the vertical scanning angle resolution Δθ and the horizontal scanning angle resolution Δφ corresponding to the measurement object points in each grid are calculated. By comparison, the minimum value of the vertical scanning angle resolution and the minimum value of the horizontal scanning angle resolution in each grid are statistically determined. S23: Merge the meshes to form a segmented two-dimensional plane; A preset angular resolution difference is set. For each grid, the minimum value of its scanning angular resolution is compared with the minimum value of the scanning angular resolution of its 8 surrounding grids. The difference between the minimum values ​​of the horizontal scanning angular resolution and the minimum value of the vertical scanning angular resolution is calculated. If the sum of the two differences is lower than the preset angular resolution difference, then every two grids that meet the requirements are merged. After merging the grids, the original two-dimensional plane containing several grids is divided into a two-dimensional plane containing several segments. By comparison, the minimum vertical scanning angular resolution and the minimum horizontal scanning angular resolution within each segment are determined. S24: Save the minimum and maximum values ​​of the azimuth, elevation, and vertical and horizontal scanning angular resolutions within each segment as six scanning parameters. Based on these six scanning parameters, scan the corresponding segment to measure the point set {P}. i (x i , y i ,z i A segmented scanning scheme is formed. S3: Conduct segmented fine scanning of the object being measured.

2. The point cloud density adaptive laser scanning method for elevation measurement of long-span bridges according to claim 1, characterized in that, In step S3, a fine scan is performed on each segment to obtain the point cloud model of the measurement object set of the long-span bridge, and geometric measurements and verification are performed, including the following steps: S31: Input the six scanning parameters within each segment into the TLS device and perform a detailed scan one by one; S32: Merge the multiple point cloud models generated after scanning, and transform the merged point cloud model from the TLS coordinate system to the bridge structure design coordinate system according to the rigid body transformation matrix M; S33: Measuring points are set at fixed intervals along the longitudinal axis of the bridge; Select a point cloud within a 4 m × 4 m rectangular area centered on the measuring point; Based on the histogram, the elevation distribution of all measurement points on the bridge surface of the long-span bridge is statistically analyzed. The frequency values ​​of elevation occurrence are smoothed by zero-phase low-pass filtering, and the elevation value corresponding to the peak frequency value after smoothing is recorded as the elevation of the corresponding measurement point. Elevation statistics were performed on all measuring points in sequence to obtain the elevation curve of the main beam along the longitudinal axis. The obtained elevation curve was then compared with the design elevation curve.

3. The point cloud density adaptive laser scanning method for elevation measurement of long-span bridges according to claim 1, characterized in that, In step S11, the horizontal scanning angular resolution and the vertical scanning angular resolution are set to the same value, both below 0.05 degrees. The horizontal scanning angle is from 0 to 360 degrees, and the vertical scanning angle is from -40 to 60 degrees.

4. The point cloud density adaptive laser scanning method for elevation measurement of long-span bridges according to claim 1, characterized in that, In step S22, the preset scanning angle interval is 2-5 degrees, and the pitch angle θ i and azimuth φ i The scanning angle interval values ​​are the same.

5. The point cloud density adaptive laser scanning method for elevation measurement of long-span bridges according to claim 2, characterized in that, In step S31, the scanning range is determined by the minimum and maximum values ​​of the azimuth angle within each segment to determine the range of the TLS horizontal scanning angle, and by the minimum and maximum values ​​of the elevation angle within each segment to determine the range of the TLS vertical scanning angle.

Images