CN116990787A - Scanning platform coordinate system error correction method based on airborne lidar system - Google Patents

Abstract

The invention discloses a scanning platform coordinate system error correction method based on an airborne laser radar system, which belongs to the technical field of laser radar measurement and is used for error correction of the airborne laser radar system, and comprises the steps of carrying the airborne laser radar system on a horizontally advancing crane, enabling the crane to be in a static horizontal suspension state, and vertically calibrating the rotation center of a reflecting mirror downwards; turning on a laser transceiver and a rotating mirror scanning device, reflecting a laser scanning track downwards, and carrying out fixed point calibration on the long and short axis vertexes of the actual oval scanning track according to the actual laser scanning track in a site coordinate system; and processing the rotational distortion, the tensile distortion and the translational distortion to obtain a scanning platform coordinate system error correction formula. The invention corrects the scanning coordinate by the correction equation, thereby realizing the correction and compensation of the influence of the structural error, and the coordinate precision of the laser foot point space coordinate obtained under the corrected scanning platform coordinate system is improved.

Description

Scanning platform coordinate system error correction method based on airborne lidar system

Technical field

The invention discloses a scanning platform coordinate system error correction method based on an airborne lidar system, and belongs to the technical field of lidar measurement.

Background technique

With the expansion of social needs and breakthroughs in lidar-related technologies, airborne lidar detection has developed into an emerging high-tech. As a complex system with highly integrated multi-sensors, the airborne lidar detection system can be mounted on an unmanned aerial vehicle (UAV) platform. Through the fusion processing of data collected by multiple sensors, it can achieve three-dimensional inversion of terrain and landforms. The advantages of airborne lidar's short working cycle and limited detection scenarios give it incomparable advantages in detection work in shallow seas, islands, or areas where ships cannot enter. In order to better apply the circular mirror off-axis marine lidar to actual scenarios, it carries out scanning detection work. Its detection system is more inclined to the development direction of miniaturization and lightweight, which requires each mechanical device to make full use of the internal space for installation. . However, during the process of machining and device installation, some system errors will inevitably be caused, resulting in errors between the actual laser reflected light direction and the designed direction, which will in turn lead to errors in the generated laser foot points and point cloud coordinates. In turn, The accuracy of the performed three-dimensional terrain data is also affected. In the current public airborne lidar error calibration documents, most of them are calibrations of the placement errors of the largest system error sources, and few calibrations of the collimation axis errors caused by the installation of the drive motor shaft and scanning mirror, or It is said that airborne laser scanning platforms are rarely inspected and calibrated. However, the collimation axis error caused by the installation will be fixed and unchanged, and its impact on the detection accuracy of airborne lidar will also increase with the increase of lidar working height. In order to fully improve the detection accuracy of lidar and generate accurate point cloud data, it is necessary and important to calibrate the boresight axis error.

In a circular mirror off-axis marine lidar device, the angle between the normal line of the reflector and the rotation axis of the drive motor is usually 7.5°, and the angle between the rotation axis of the drive motor and the horizontal line is 45°. The laser is incident horizontally and is located at or parallel to the rotation axis of the drive motor. The vertical plane on which it is located. However, when the mechanical device is installed, the installation angle and the parallel accuracy of the incident light cannot be strictly controlled. The error in the actual installation angle and position causes the laser footpoint coordinates obtained in the actual scanning platform coordinate system to be different from the ideal scanning platform coordinate system. There are errors in the laser footpoint coordinates obtained below, which ultimately leads to distortion of the point cloud. Most laser radar error calibration is to calibrate the maximum system error placement error, and rarely calibrate the collimation axis error caused by mechanical installation. However, the data error caused by the collimation axis error cannot pass the placement error calibration equation. Perform data correction. Based on this, the collimation axis error needs to be calibrated before the test. Through the distortion analysis of the actual scanning trajectory and the theoretical trajectory, the collimation axis error of the detection system is reversely deduced, and the scanning platform error equation is constructed to calculate the scanning platform coordinates. Calibrate and generate accurate laser foot point coordinates in the corrected scanning coordinate system.

Contents of the invention

The purpose of the present invention is to provide a scanning platform coordinate system error correction method based on an airborne lidar system to solve the problem of placement errors in the airborne lidar system in the prior art.

Scanning platform coordinate system error correction method based on airborne lidar system, including:

B1. Mount the airborne lidar system on a horizontally advancing crane. The crane is in a static horizontal suspension state, and the mirror rotation center is downward for vertical calibration;

B2. Turn on the laser transceiver device and the rotating mirror scanning device, reflect the laser scanning trajectory downward, and perform fixed-point calibration of the long and short axis vertices of the actual oval scanning trajectory according to the actual laser scanning trajectory in the site coordinate system;

B3. Process rotation distortion;

B4. Deal with tensile distortion;

B5. Process translation distortion;

B6. Obtain the error correction formula of the scanning platform coordinate system.

B1 includes: the calibration point is the coordinate origin O ₂ , the straight line drawn along the horizontal direction of the crane is the Y ₂ axis, the straight line passing through the origin and perpendicular to the Y ₂ axis is the X ₂ axis, and the straight line between the origin and the mirror rotation center is Z ₂ axes, the O ₂ -X ₂ Y ₂ Z ₂ site coordinate system is constructed. The laser scanning trajectory is the long axis b on the Y ₂ axis, the positive half axis of Y is b ₁ , the negative half axis is b ₂ , and on the X ₂ axis The upper part is the minor axis a, the positive half axis of X is a ₁ , and the negative half axis is a ₂ .

B2 includes:

It is measured that the long axis of the actual scanning trajectory is b′, the short axis is a′, and the intersection point of the long and short axes is the actual origin O ₂ ′, and the origin coordinates in the site coordinate system are obtained. , two major axis vertex coordinates/> and/> , two minor axis vertex coordinates/> and/> .

B3 includes:

In the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ , according to the vector and/> Find the angle between two vectors/> :

;

Rotate the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ around the Z ₂ axis by W ₁ to correct the rotation distortion. The rotation distortion correction matrix is :

;

After rotational distortion correction, the image is moved along the X ₂ axis by the X ₀ distance, along the Y ₂ axis by the Y ₀ distance, and along the Z ₂ axis by the Z ₀ distance to obtain the translation distortion correction parameters X ₀ , Y ₀ and Z.

B4 includes:

Based on the known origin and the five coordinate points of the long and short axis vertices in the site coordinate system, draw the long and short axis coordinate diagram. The intersection of the two axes is the origin. Draw the ideal long and short axis diagram and the actual long and short axis diagram of equal specifications, and generate the intersection of the ideal long and short axes. For a rectangular plan view, the actual long and short axes are intersected to generate a parallelogram plan view, and the two images are binarized;

The Sobel operator is used to detect edges in the horizontal and vertical directions of the binary image, and then Radon transform is used based on the edge detection to find the tilt angle in both horizontal and vertical directions.

B4 includes: using Radon transformation to obtain the stretch distortion between the actual long and short axis map and the ideal long and short axis map, obtaining the horizontal tilt angle W ₂ and vertical tilt angle W ₃ of the image, and establishing a rotation matrix to correct the error of the ideal coordinate system;

The obtained horizontal tilt angle W ₂ is the rotation angle of the scanning platform coordinate system around the Y axis, and the correction rotation matrix is , the obtained vertical tilt angle W ₃ is the rotation angle of the scanning platform coordinate system around the X-axis, and the correction rotation matrix is/> :

; /> .

B4 includes:

The algorithm for finding the horizontal tilt angle using Radon transformation is:

T1. Assume that the horizontal tilt range of the image is 0°~180°, and use Radon transformation to project the image after horizontal edge detection from 0°~180°;

T2. Obtain the angle θ1 when the sum of projected non-zero values reaches the maximum;

T3. The horizontal tilt angle of the image should be the supplementary angle of θ1. Subtract θ1 from 90° to get the horizontal tilt angle W ₂ of the image.

B4 includes:

The algorithm for finding the vertical tilt angle using Radon transformation is:

M1. Assume that the vertical tilt range of the image is -45°~45°, and use Radon transformation to project the image after horizontal edge detection at -45°~45°;

M2. Obtain the angle θ2 when the sum of projected non-zero values reaches the maximum;

M3. The vertical tilt range of the image is -45°~45°. Subtract 45° from θ2 to get the vertical tilt angle W ₃ of the image.

B5 includes:

In the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ , the ideal origin is , connect the vertices of the long axis and the vertices of the short axis to get the foot points of two straight lines/> , change the actual origin/> Perform the previous rotation and stretching distortion to obtain the translation distortion parameters:

;

In the formula, is the actual origin coordinate after correction,/> is the rotation matrix, are the uncorrected actual origin coordinates.

B6 includes:

Error correction formula for the ideal scanning platform:

;

in, are the coordinates in the uncorrected ideal scanning platform coordinate system,/> is the translation correction parameter,/> is the rotation matrix,/> are the corrected ideal coordinate system coordinates;

Use the corrected ideal coordinate system coordinates for accuracy evaluation. If the calibration difference is less than the set accuracy threshold, the scanning platform error calibration equation will be output; if the calibration difference is greater than the set accuracy threshold, a second step will be performed on the basis of the calibration of the scanning platform. For the first correction, the long and short axis coordinates under the corrected scanning platform and the actual long and short axis coordinates are drawn to draw a coordinate diagram and a plane binary image is generated until the correction value of the obtained coordinates and the actual coordinates is obtained under the corrected scanning platform. If it is less than the set accuracy threshold, the accuracy requirements are met:

;

In the formula, for,/> for,/> for;

During secondary calibration, the scanning platform error correction equation is:

;

When calibrating n times, the scanning platform error correction equation is:

;

The second school year is officially underway, It is the coordinate in the coordinate system of the primary correction scanning platform,/> is the secondary translation correction parameter,/> is a quadratic rotation matrix,/> are the ideal coordinate system coordinates after secondary correction;

n times the school is officially in, The coordinates in the coordinate system of the scanning platform are corrected for n-1 times,/> is the n-th translation correction parameter,/> is the n-th rotation matrix,/> are the ideal coordinate system coordinates after n corrections, that is, the coordinates and scanning platform error correction equation that finally meet the accuracy requirements.

Compared with the existing technology, the present invention has the following beneficial effects: by measuring the actual laser scanning trajectory, reversely analyzing and correcting the existing structural errors, and generating an error correction equation. By correcting the scanning coordinates with the correction equation, correction compensation for the impact of structural errors is achieved. The coordinate accuracy of the laser footpoint spatial coordinates obtained under the corrected scanning platform coordinate system will be improved.

Description of the drawings

Figure 1 is a technical flow chart of the present invention;

Figure 2 is the coordinate system diagram of the laser scanning platform;

Figure 3 is a schematic diagram of the test;

Figure 4 is an ideal scanning trajectory diagram;

Figure 5 is the actual scanning trajectory diagram;

Figure 6 is Original image and Radon transformed coordinate system diagram;

Figure 7 is Parallel to/> Linear integral plot of the axis.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention are clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

Scanning platform coordinate system error correction method based on airborne lidar system, including:

B1. Mount the airborne lidar system on a horizontally advancing crane. The crane is in a static horizontal suspension state, and the mirror rotation center is downward for vertical calibration;

B2. Turn on the laser transceiver device and the rotating mirror scanning device, reflect the laser scanning trajectory downward, and perform fixed-point calibration of the long and short axis vertices of the actual oval scanning trajectory according to the actual laser scanning trajectory in the site coordinate system;

B3. Process rotation distortion;

B4. Deal with tensile distortion;

B5. Process translation distortion;

B6. Obtain the error correction formula of the scanning platform coordinate system.

B1 includes: the calibration point is the coordinate origin O ₂ , the straight line drawn along the horizontal direction of the crane is the Y ₂ axis, the straight line passing through the origin and perpendicular to the Y ₂ axis is the X ₂ axis, and the straight line between the origin and the mirror rotation center is Z ₂ axes, the O ₂ -X ₂ Y ₂ Z ₂ site coordinate system is constructed. The laser scanning trajectory is the long axis b on the Y ₂ axis, the positive half axis of Y is b ₁ , the negative half axis is b ₂ , and on the X ₂ axis The upper part is the minor axis a, the positive half axis of X is a ₁ , and the negative half axis is a ₂ .

B2 includes:

It is measured that the long axis of the actual scanning trajectory is b′, the short axis is a′, and the intersection point of the long and short axes is the actual origin O ₂ ′, and the origin coordinates in the site coordinate system are obtained. , two major axis vertex coordinates/> and/> , two minor axis vertex coordinates/> and/> .

B3 includes:

In the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ , according to the vector and/> Find the angle between two vectors/> :

;

Rotate the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ around the Z ₂ axis by W ₁ to correct the rotation distortion. The rotation distortion correction matrix is :

;

After rotational distortion correction, the image is moved along the X ₂ axis by the X ₀ distance, along the Y ₂ axis by the Y ₀ distance, and along the Z ₂ axis by the Z ₀ distance to obtain the translation distortion correction parameters X ₀ , Y ₀ and Z.

B4 includes:

Based on the known origin and the five coordinate points of the long and short axis vertices in the site coordinate system, draw the long and short axis coordinate diagram. The intersection of the two axes is the origin. Draw the ideal long and short axis diagram and the actual long and short axis diagram of equal specifications, and generate the intersection of the ideal long and short axes. For a rectangular plan view, the actual long and short axes are intersected to generate a parallelogram plan view, and the two images are binarized;

The Sobel operator is used to detect edges in the horizontal and vertical directions of the binary image, and then Radon transform is used based on the edge detection to find the tilt angle in both horizontal and vertical directions.

B4 includes: using Radon transformation to obtain the stretch distortion between the actual long and short axis map and the ideal long and short axis map, obtaining the horizontal tilt angle W ₂ and vertical tilt angle W ₃ of the image, and establishing a rotation matrix to correct the error of the ideal coordinate system;

The obtained horizontal tilt angle W ₂ is the rotation angle of the scanning platform coordinate system around the Y axis, and the correction rotation matrix is , the obtained vertical tilt angle W ₃ is the rotation angle of the scanning platform coordinate system around the X-axis, and the correction rotation matrix is/> :

; /> .

B4 includes:

The algorithm for finding the horizontal tilt angle using Radon transformation is:

T1. Assume that the horizontal tilt range of the image is 0°~180°, and use Radon transformation to project the image after horizontal edge detection from 0°~180°;

T2. Obtain the angle θ1 when the sum of projected non-zero values reaches the maximum;

T3. The horizontal tilt angle of the image should be the supplementary angle of θ1. Subtract θ1 from 90° to get the horizontal tilt angle W ₂ of the image.

B4 includes:

The algorithm for finding the vertical tilt angle using Radon transformation is:

M1. Assume that the vertical tilt range of the image is -45°~45°, and use Radon transformation to project the image after horizontal edge detection at -45°~45°;

M2. Obtain the angle θ2 when the sum of projected non-zero values reaches the maximum;

M3. The vertical tilt range of the image is -45°~45°. Subtract 45° from θ2 to get the vertical tilt angle W ₃ of the image.

B5 includes:

In the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ , the ideal origin is , connect the vertices of the long axis and the vertices of the short axis to get the foot points of two straight lines/> , change the actual origin/> Perform the previous rotation and stretching distortion to obtain the translation distortion parameters:

;

In the formula, is the actual origin coordinate after correction,/> is the rotation matrix, are the uncorrected actual origin coordinates.

B6 includes:

Error correction formula for the ideal scanning platform:

;

in, are the coordinates in the uncorrected ideal scanning platform coordinate system,/> is the translation correction parameter,/> is the rotation matrix,/> are the corrected ideal coordinate system coordinates;

Use the corrected ideal coordinate system coordinates for accuracy evaluation. If the calibration difference is less than the set accuracy threshold, the scanning platform error calibration equation will be output; if the calibration difference is greater than the set accuracy threshold, a second step will be performed on the basis of the calibration of the scanning platform. For the first correction, the long and short axis coordinates under the corrected scanning platform and the actual long and short axis coordinates are drawn to draw a coordinate diagram and a plane binary image is generated until the correction value of the obtained coordinates and the actual coordinates is obtained under the corrected scanning platform. If it is less than the set accuracy threshold, the accuracy requirements are met:

;

In the formula, for,/> for,/> for;

During secondary calibration, the scanning platform error correction equation is:

;

When calibrating n times, the scanning platform error correction equation is:

;

The second school year is officially underway, It is the coordinate in the coordinate system of the primary correction scanning platform,/> is the secondary translation correction parameter,/> is a quadratic rotation matrix,/> are the ideal coordinate system coordinates after secondary correction;

n times the school is officially in, The coordinates in the coordinate system of the scanning platform are corrected for n-1 times,/> is the n-th translation correction parameter,/> is the n-th rotation matrix,/> are the ideal coordinate system coordinates after n corrections, that is, the coordinates and scanning platform error correction equation that finally meet the accuracy requirements.

The technical process of the present invention is shown in Figure 1, which includes obtaining the oval origin and vertex coordinates of the actual oval scanning trajectory, scanning trajectory distortion, setting the accuracy threshold, and determining if it is greater than the accuracy threshold, then generating a binary image and performing sobel operator edge Detection, perform radon transformation to find the tilt angle, correct the scanning platform coordinates, calculate the vertex coordinates of the laser corner point, obtain the oval origin and vertex coordinates of the design and oval scanning trajectory, re-execute the scanning trajectory distortion amount, judge that it is less than the accuracy threshold, and output the final Error correction equation.

The schematic diagram of the test process of the present invention is shown in Figure 3. The formula is calculated according to the scanning angle in the calculation process. , when the mirror normal is located on the optical path plane, the mirror rotation angle/> or/> , the scanning angle reaches the maximum/> , the formula for finding the long axis is/> , when the reflector rotates/> or/> , the scan angle reaches the minimum/> , the formula for calculating the short axis is/> .

In B2, there are distortions such as rotation, translation, and stretching between the actual scanning trajectory and the ideal scanning trajectory. The structural errors of the ideal laser scanning platform are back calculated through the distortion of the actual scanning trajectory. In B3, when the motor shaft and reflection mirror are installed, the motor shaft is not coplanar with the ideal scanning coordinate system OYZ, and there is an error angle between the two, resulting in rotational distortion of the laser scanning trajectory in the O ₂ X ₂ Y ₂ plane. In B4, when the motor shaft and reflection mirror are installed, the actual angle between the motor shaft and the reflection mirror is not equal to the design value, resulting in tensile distortion of the laser scanning trajectory in the O ₂ X ₂ Y ₂ plane. Stretch distortion tilt modes include horizontal tilt, vertical tilt and mixed tilt. Because the installation is not deterministic, there may be mixed tilt distortion, so calibration needs to consider corrections in both horizontal and vertical directions.

The Sobel operator performs edge detection on the binarized image in both horizontal and vertical directions. The Sobel operator refers to taking a certain pixel as the center and performing a weighted operation on the pixel value within its neighborhood to determine whether the point is in an extreme state. If the point is in an extreme state, the point is the edge of the image. . Its advantages are simple calculation and fast speed.

The specific implementation is as follows: Indicates point/> pixel value. /> for point/> The gradient in the horizontal direction is used to detect horizontal edges. /> is calculated as follows:

;

for point/> Gradient in the vertical direction, used to detect vertical edges. /> is calculated as follows:

;

Set the threshold T=1, and compare the threshold with Judge by size/> Whether it is an edge point. When detecting horizontal edges, set the threshold for each pixel in the image/> The average of the squares. When detecting vertical edges, set the threshold for each pixel in the image/> The average of the squares. If the threshold ratio/> small, then/> is the edge point.

Use Radon transformation to find the tilt angle of the image. Radon transformation means that the image is projected in a certain angle direction. When the projection distance is at the extreme value, the corresponding angle is the tilt angle of the image. Smeared or unbalanced images will not affect the tilt angle obtained by Radon transform. Due to the influence of the actual installation position of the motor shaft and the reflection mirror, the long and short axes of the oval scanning trajectory can form a parallelogram plane on the O ₂ X ₂ Y ₂ plane, as shown in Figures 4 and 5. The actual deformation is detected and corrected step by step. First, the horizontal rotation correction is performed, the horizontal rotation angle is detected, and the correction is performed by rotating the image. For miscutting deformation in the vertical direction, the wrong shearing angle is detected, and affine transformation is used to correct the miscutting.

The projection calculation formula generated by the two-dimensional Radon transform is:

;

in:

;

In the formula, are the ideal long and short axis chart coordinates,/> are the actual long and short axis chart coordinates,/> is the tilt angle.

The Radon transform is shown in Figure 6, Figure 6 is/> The binary image and the coordinate system generated by Radon transformation. Figure 7 is/> Parallel to/> Linear integral of the axis. When/> When it is an image matrix, Radon transformation refers to the projection of the image matrix at a certain angle.

In B5, according to the ideal origin, and the actual origin after correction/> The positional relationship of the two points intuitively reflects the translation distortion of the image. In B6, the rotation angle W ₁ is obtained through rotation distortion correction, the translation amounts X ₀ and Y ₀ are obtained through translation distortion correction, and the rotation angles W ₂ and W ₃ are obtained through tensile distortion correction. Through the obtained correction parameters, the ideal scanning platform coordinate system is corrected, and the actual scanning platform coordinate system that meets the actual conditions is obtained through parameter correction.

In B6, the obtained error correction equation is calibrated to the ideal scanning platform coordinate system, the error correction is performed to the laser foot point coordinates obtained under the scanning platform, and the corrected long and short axis coordinates are calibrated with the actual long and short axis coordinates. If the calibration difference is less than the set accuracy threshold, the scanning platform error calibration equation will be output; if the calibration difference is greater than the set accuracy threshold, the calibration will be corrected again based on the calibration of the scanning platform, and the long and short axis coordinates under the corrected scanning platform will be With the actual long and short axis coordinates, draw a coordinate diagram and generate a plane binary image. Repeat the previous steps until the calibration value between the obtained coordinates and the actual coordinates is less than the set accuracy threshold under the corrected scanning platform. , to meet the accuracy requirements.

The error correction method of the scanning platform coordinate system based on the airborne lidar system is used to calculate the spatial coordinates of the laser foot points, including:

S1. Establish the laser scanning platform coordinate system and auxiliary coordinate system. The laser scanning platform coordinate system is shown in Figure 2;

S2. Obtain distance data H and mirror angle through the airborne lidar detection system , through distance H and corner/> Obtain the spatial coordinates of the laser footpoint;

S3. Analyze the impact of collimation axis error, and use the relationship between the degree of error impact and the working height to be proportional. After the laser radar reaches a fixed height, the impact of structural error on the laser scanning trajectory is amplified. By measuring the actual laser scanning trajectory, Reversely analyze the structural errors existing in the calibration and generate error correction equations; correct the scanning coordinates through the correction equations to achieve correction compensation for the impact of structural errors; obtain the spatial coordinates of the laser foot points in the corrected scanning platform coordinate system, Improve coordinate accuracy.

S1 includes: establishing a laser scanning platform coordinate system with the mirror rotation center as the coordinate origin O. The Y axis is the forward direction of the lidar vehicle, the Z axis is vertical upward, and the X axis is perpendicular to the Y and Z axes, forming a right-hand coordinate system O-XYZ. ;

Using _the X _- axis as the rotation _{axis ,} reverse rotation _forms the , the Y ₁ axis is the forward direction of the lidar vehicle, the Z ₁ axis is the rotation axis of the drive motor, and the X ₁ axis is perpendicular to the Y ₁ and Z ₁ axes, forming a right-hand coordinate system OX ₁ Y ₁ Z ₁ .

S2 includes: In the laser radar device, the angle between the normal line of the designed reflector and the rotation axis of the drive motor is , in the embodiment, is 7.5 degrees, the angle between the drive motor shaft and the horizontal line is/> , in the embodiment,/> is 45 degrees, the laser is incident horizontally and is located at or parallel to the vertical plane where the drive motor shaft is located;

S2.1. In the auxiliary coordinate system OX ₁ Y ₁ Z ₁ , find the incident light direction vector F1:

;

Mirror normal vector F2:

;

According to the vector angle formula, the angle between the incident light and the normal of the mirror for:

;

.

S2.2. When the incident light of the laser, the normal of the mirror and the reflected light of the laser are coplanar, the plane where the laser is located is called the optical path plane. Under the auxiliary coordinate system, the normal vector of the optical path plane is obtained. for:

;

;

The incident light F1, the mirror normal F2 and the light path plane normal are known. and the positional relationship between the reflected light F3 and establish a relationship to solve the reflected light F3:

.

S2.3. Rotate the reflected light F3 from the auxiliary coordinate system O ₁ -X ₁ Y ₁ Z ₁ to the laser scanning platform coordinate system O-XYZ. The reflected light unit vector is in the O-XYZ coordinate system. for:

;

in, is the direction cosine of the reflected light vector,/> Indicates that the obtained reflected light F3 is rotated in the opposite direction around the X-axis/> .

S2.4. The angle between the reflected light and the vertical downward direction is the scanning angle, which is also the zenith angle. :

;

Express/> is about/> The function;

The central turning angle of the laser on the plane, that is, the azimuth angle is :

;

Express/> is about/> The function;

Spatial coordinates of laser footpoints for:

;

In the formula, ,/> ,/> The three-dimensional coordinate values representing the spatial coordinates of the laser foot points;

The spatial coordinates of the laser footpoint are determined by distance H and zenith angle q ( ) and azimuth angle y(/> ) is expressed as:

.

Under the coordinate system of the ideal scanning platform, the spatial coordinates of the laser foot points obtained are the coordinates under the ideal design values. However, when the motor shaft and reflective mirror are installed, due to the uncertainty of the installation process, they cannot be installed perfectly according to the design values. To a certain extent, the angle A between the normal line of the mirror and the drive motor shaft is inconsistent with the drive motor shaft and the horizontal line. Angle B is not equal to the set angle 7.5° and 45°, and there is an error between the design value and the actual value. This error exists fixedly and does not change with the motion state, which will lead to errors in obtaining the spatial coordinates of the laser foot points, and the degree of the error increases with the increase in the working height of the laser radar. Therefore, it is necessary to complete the calibration of this error and generate a correction equation before the lidar test.

However, due to the uncertainty of the installation process, it is not feasible to directly detect the result error caused by the installation process, and it is very difficult to measure. This invention utilizes the relationship between the degree of error influence and the working height being proportional. After the laser radar reaches a fixed height, the influence of structural error on the laser scanning trajectory is amplified.

;

;/> is the total rotation matrix;

;

;

but ;

In the formula, is the intermediate parameter.

The above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments. The recorded technical solutions may be modified, or some or all of the technical features may be equivalently replaced, but these modifications or substitutions shall not cause the essence of the corresponding technical solutions to depart from the scope of the technical solutions of each embodiment of the present invention.

Claims (10)

1. The scanning platform coordinate system error correction method based on the airborne laser radar system is characterized by comprising the following steps of:

B1. the method comprises the steps of carrying an airborne laser radar system on a horizontally advancing crane, wherein the crane is in a static horizontal suspension state, and vertically calibrating the rotation center of a reflector downwards;

B2. turning on a laser transceiver and a rotating mirror scanning device, reflecting a laser scanning track downwards, and carrying out fixed point calibration on the long and short axis vertexes of the actual oval scanning track according to the actual laser scanning track in a site coordinate system;

B3. processing rotational distortion;

B4. processing stretching distortion;

B5. processing translational distortion;

B6. and obtaining a scanning platform coordinate system error correction formula.

2. The method for correcting an error in a scanning platform coordinate system based on an airborne laser radar system according to claim 1, wherein B1 comprises: calibration point as origin of coordinates O ₂ The straight line along the horizontal advancing direction of the crane is Y ₂ An axis passing through the origin and perpendicular to Y ₂ The straight line of the axis is X ₂ The straight line of the axis, the origin and the rotation center of the reflector is Z ₂ Shaft, build O ₂ -X ₂ Y ₂ Z ₂ A field coordinate system, a laser scanning track is in Y ₂ The axis is a long axis b, and the Y positive half axis is b ₁ The negative half shaft is b ₂ At X ₂ The axis is a short axis a, and the X positive half axis is a ₁ The negative half shaft is a ₂ 。

3. The method for correcting an error in a scanning platform coordinate system based on an airborne laser radar system according to claim 2, wherein B2 comprises:

the long axis of the actual scanning track is measured as b ', the short axis is measured as a', and the intersection point of the long axis and the short axis is measured as the actual origin O ₂ ' obtaining the origin coordinate in the site coordinate systemTwo long axis vertex coordinates-> and />Two short axis vertex coordinates-> and />。

4. The method for correcting an error of a scanning platform coordinate system based on an airborne laser radar system according to claim 3, wherein B3 comprises:

site coordinate system O ₂ -X ₂ Y ₂ Z ₂ Based on the vector and />Obtaining the included angle +.>：

；

Will be the site coordinate system O ₂ -X ₂ Y ₂ Z ₂ Around Z ₂ Shaft rotation W ₁ Realizes the correction of the rotation distortion quantity, and the rotation distortion correction matrix is that：

；

After rotational distortion correction, the image is corrected along X ₂ Axis movement X ₀ Distance along Y ₂ Axis movement Y ₀ Distance along Z ₂ Axial movement Z ₀ Distance-derived translational distortion correction parameter X ₀ 、Y ₀ and Z.

5. The method for error correction of a scanning platform coordinate system based on an airborne lidar system of claim 4, wherein B4 comprises:

drawing a long-short axis coordinate graph according to 5 coordinate points of a known origin and a long-short axis vertex in a site coordinate system, wherein two axes are intersected to form the origin, drawing an ideal long-short axis graph and an actual long-short axis graph with equal specifications, intersecting the ideal long-short axis to form a rectangular plan graph, intersecting the actual long-short axis to form a parallelogram plan graph, and carrying out binarization treatment on the two graphs;

and (3) performing edge detection on the binarized image in two directions of horizontal and vertical by utilizing a Sobel operator, and then solving the inclination angles in the two directions of horizontal and vertical by using Radon transformation on the basis of the edge detection.

6. The method for error correction of a scanning platform coordinate system based on an airborne lidar system of claim 5, wherein B4 comprises: obtaining the stretching distortion between the actual long-short axis image and the ideal long-short axis image by using Radon transformation, and obtaining the horizontal inclination angle W of the image ₂ And vertical tilt angle W ₃ Establishing a rotation matrix to perform error correction on an ideal coordinate system;

the resulting horizontal tilt angle W ₂ Correcting the rotation matrix to be the rotation angle around the Y axis of the scanning platform coordinate systemThe resulting vertical tilt angle W ₃ For the rotation angle of the scanning platform coordinate system around the X-axis, the rotation matrix is corrected to +.>：

； />。

7. The method for error correction of a scanning platform coordinate system based on an airborne lidar system of claim 6, wherein B4 comprises:

the horizontal tilt angle algorithm is calculated by Radon transformation:

t1, assuming that the horizontal inclination range of an image is 0-180 degrees, projecting the image subjected to horizontal edge detection at 0-180 degrees by using Radon transformation;

t2, obtaining an angle θ1 when the addition of the non-zero values after projection reaches the maximum;

t3. the horizontal tilt angle of the image should be the complement of θ1, and subtracting θ1 from 90 ° will obtain the horizontal tilt angle W of the image ₂ 。

8. The method for error correction of a scanning platform coordinate system based on an airborne lidar system of claim 7, wherein B4 comprises:

the vertical tilt algorithm using Radon transform is:

the method comprises the steps that M1, assuming that the vertical inclination range of an image is-45 degrees, carrying out projection on the image subjected to horizontal edge detection at-45 degrees by using Radon transformation;

m2, obtaining an angle theta 2 when the addition of the non-zero values after projection reaches the maximum;

m3, the vertical inclination range of the image is-45 degrees to 45 degrees, and the vertical inclination angle W of the image is obtained by subtracting 45 degrees from theta 2 ₃ 。

9. The method for correcting errors in a scanning platform coordinate system based on an airborne laser radar system according to claim 8, wherein B5 comprises:

site coordinate system O ₂ -X ₂ Y ₂ Z ₂ The ideal origin isConnecting the long axis vertex with the short axis vertex to obtain two straight line foot points +.>The actual origin +.>Performing the rotation and the rotation of the previousStretching distortion to obtain translational distortion parameters:

；

in the formula ,for corrected actual origin coordinates, +.>For rotating matrix +.>The actual origin coordinates are uncorrected.

10. The method for error correction of a scanning platform coordinate system based on an airborne lidar system of claim 9, wherein B6 comprises:

an error correction formula is carried out on an ideal scanning platform:

；

wherein ,for uncorrected coordinates in the ideal scanning platform coordinate system, +.>In order to translate the correction parameters,for rotating matrix +.>The corrected ideal coordinate system coordinates;

performing precision evaluation by using the corrected ideal coordinate system coordinates, and outputting a scanning platform error calibration equation if the calibration difference is smaller than a set precision threshold; if the correction difference is larger than the set precision threshold, carrying out secondary correction on the basis of the corrected scanning platform, drawing a coordinate graph of the long and short axis coordinates and the actual long and short axis coordinates under the corrected scanning platform, and generating a plane binarization image until the correction value of the obtained coordinates and the actual coordinates is smaller than the set precision threshold under the corrected scanning platform, and meeting the precision requirement:

；

in the formula ,for (I)>For (I)>Is that;

during secondary correction, scanning a platform error correction equation:

；

and (3) scanning a platform error correction equation during n times of correction:

；

in the secondary correction type, the correction amount is set,for correcting the coordinates in the scanning platform coordinate system once, +.>Correction parameters for secondary translation, < >>Is a quadratic rotation matrix +.>The ideal coordinate system coordinate after the secondary correction;

in the n-time correction type, the correction is performed,correcting the coordinates of the scanning platform in the coordinate system for n-1 times,/for>Correction parameters for n translations,>for n rotation matrices +.>And the coordinates of the ideal coordinate system after n times of correction are the coordinates which finally meet the precision requirement and the error correction equation of the scanning platform.