Airborne depth sounding laser radar dynamic scanning angle correction method

Abstract

The invention discloses a method for correcting a dynamic scanning angle of an airborne sounding laser radar, which relates to the technical field of airborne sounding and comprises the following steps of: a) laying a site, and erecting a three-dimensional measurement system formed by two calibrated cameras and the sounding laser radar according to a design scheme; b) setting more than six groups of rotational speeds of the airborne depth sounding laser radar motor at a constant speed and carrying out switch scanning, wherein a rotational speed setting sampling interval needs to cover the minimum and maximum rotational speed range of the equipment motor; meanwhile, a scanning track of a corresponding rotating speed laser point is completely shot by using a three-dimensional measurement system; c) calculating the correction amount bn corresponding to the included angle alpha between the rotating shaft and the optical axis of the emitted laser at different rotating speeds according to a static scanning angle correction model, and recording more than six groups of data; the method can achieve the precise acquisition of the dynamic scanning angle of the airborne depth-sounding laser radar, and achieves the high-precision acquisition of the dynamic scanning angle parameters of the depth-sounding laser radar.

Description

Technical Field

[0001] This invention relates to the field of airborne depth sounding technology, specifically a method for dynamic scanning angle correction of airborne depth sounding lidar. Background Technology

[0002] With the rapid increase in my country's maritime activities and the development needs of building a maritime power, integrated surveying of the land-sea transition zone, including coastal zones, islands and reefs, nearshore underwater topography, and the complex intertidal zone, has always been a challenging issue in the field of marine surveying and mapping both domestically and internationally. This is also related to the development of the marine economy and national security. Therefore, conducting marine surveying and mapping in shallow seas and coastal zones to ensure the timely acquisition and updating of marine geospatial information has significant economic and military implications.

[0003] Currently, major international companies have launched their own airborne laser depth sounding systems, such as Optech's CZMIL Nova system, Leica's Eagle Eye and Bat series depth sounding systems, Fugro's LADS HD system, and RI EGL's VQ880-G. However, due to commercial technology secrecy, non-public data formats, high system prices, and the arms embargo against China, the development of my country's marine surveying technology and information security are severely hampered. The most critical component of an airborne depth sounding lidar system is the depth sounding lidar itself. Regarding the scanning angle correction of the depth sounding lidar optical path system, due to limitations in foreign technology, data formats, and the lack of public system parameters, experts and scholars are unable to further study and analyze the system's internal structure. Current research on system scanning angle correction, both domestically and internationally, is limited to the correction of the emitted light beam.

[0004] The domestically produced airborne laser depth sounding system is a 2D lidar system employing a circular axis polarizing mirror oval scanning structure. Despite dynamic balancing tests, an off-axis torque still exists during high-speed mirror rotation, causing dynamic imbalance. This imbalance, in turn, generates centrifugal force on the mirror, leading to dynamic changes in the angle between the rotating axis and the mirror surface during scanning. This dynamic change significantly impacts the accuracy of the scanning points and cannot be quickly and accurately corrected using the methods described above. Therefore, developing a rapid and automated measurement method independent of traditional optical calibration schemes to precisely correct this dynamic scanning angle is a key challenge in the current process of industrializing domestically produced airborne depth sounding systems. Summary of the Invention

[0005] The purpose of this invention is to provide a method for dynamic scanning angle correction of airborne depth sounding lidar, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for dynamic scanning angle correction of an airborne depth sounding lidar is proposed. This method employs a dual-camera stereo measurement principle for model construction. Six sets of airborne depth sounding lidar motor speeds are set at constant velocity, and scanning is performed with the light switched on. The sampling interval for each speed setting should cover the minimum and maximum speed range of the equipment motors. After each speed data acquisition, two photos containing the scanning trajectory of an egg-shaped curve and the corresponding waveform data are obtained using a stereo measurement camera. A stereo image pair constraint is established using the two egg-shaped curves in the two photos. The scanning data trajectory should simultaneously satisfy the collinearity relationship in the two photos and the equation of the egg-shaped curve in the corresponding photos. A conditional adjustment with parameters is used as the adjustment model, and the correction amount bn corresponding to the angle α between the rotation axis and the emitted laser axis is iteratively solved according to the least squares criterion. At least six sets of data are recorded. After data recording, the parameters of the airborne depth sounding lidar dynamic scanning angle correction model are solved using exponential fitting curve equations. During system operation, given the set speed parameters, the correction amount bn corresponding to the angle α given by the dynamic angle correction model is used for correction. The correction parameter α is substituted into the laser point calculation process to obtain more accurate three-dimensional coordinates of the laser point. Specifically, it includes the following steps:

[0008] Step a) Set up the site and, according to the design plan, set up a stereo measurement system consisting of two calibrated cameras and a depth-sounding lidar.

[0009] Step b) Set the rotation speed of the 6 sets of airborne depth-sounding lidar motors at a constant speed and perform scanning with the lights on. The sampling range of the rotation speed setting should cover the minimum and maximum rotation speed range of the equipment motors. At the same time, use a stereo measurement system to completely capture the scanning trajectory of the laser points at the corresponding rotation speeds.

[0010] Step c) Calculate the correction amount bn corresponding to the angle α between the rotating shaft and the emitted laser optical axis at different rotation speeds based on the static scanning angle correction model, and record more than 6 sets of data;

[0011] Step d) After the data recording is completed, the parameters of the dynamic scanning angle correction model of the airborne depth sounding lidar are solved by using the exponential fitting curve equation. During the operation of the system, when the rotation speed parameter is set, the correction amount bn corresponding to the included angle α given by the dynamic angle correction model is used for correction. The correction parameter α is substituted into the laser point calculation process to obtain a more accurate three-dimensional coordinate of the laser point.

[0012] As a further technical solution of the present invention: in the model construction, a flat wall surface is selected as the reference plane for scanning trajectory imaging, and the flatness variation within the scanning range is less than 1mm. An airborne depth sounding radar is placed at a certain vertical distance from the flat wall surface and at a certain height above the ground to ensure that the scanning trajectory can be completely displayed on the wall surface after the equipment is placed.

[0013] As a further technical solution of the present invention: the placement position of the stereo measurement camera needs to ensure that the coordinate system of the stereo measurement camera is approximately parallel to the axis system of the depth-sounding lidar.

[0014] As a further technical solution of the present invention: the model construction requires setting up a total station, establishing an independent local coordinate system, with the coordinate system axis consistent with the coordinate system of the depth-sounding lidar system, and using the total station to measure the initial transformation parameters between the coordinate system of the stereo measuring camera and the coordinate system of the depth-sounding lidar.

[0015] As a further technical solution of the present invention, the data acquisition process is as follows:

[0016] Turn on the airborne depth-sounding lidar and begin data recording. Set the lidar motor speed D (50 ≥ D ≤ 1000 r / min) at intervals (ΔD = 100 r / min) and perform a scanning operation to display the scanning trajectory on the wall. Initially, set D to 50 r / min. Use a stereo camera to capture the complete scanning trajectory and record the radar data at the corresponding times.

[0017] As a further technical solution of the present invention: the dynamic scanning angle parameter is a correction function including the angle α between the rotation axis and the emitted laser optical axis.

[0018] As a further technical solution of the present invention, the dynamic scanning angle correction process is as follows:

[0019] ① Taking the data collected in a certain rotational speed D mode as an example, after the data collection is completed, the waveform file of the data is parsed to obtain the distance analysis file, and then two photos A and B are exported from the stereo measurement camera respectively;

[0020] ② The Hough transform ellipse detection method was used to extract the scanning point trajectory from photos A and B respectively, and the trajectory was fitted to obtain two egg-shaped curve fitting equations F1 and F2.

[0021] ③Use the scanning point positioning formula to obtain the direction vector Vs corresponding to each scanning point;

[0022] ④ Then, the direction vector Vs is adjusted according to the transformation parameters between the coordinate systems of the lidar and the stereo measurement system. ω, κ, Dx, Dy, and Dz are transformed into the stereo camera coordinate system using coordinate transformation to obtain the corresponding direction vector Vc;

[0023] ⑤ Transform the vector Vc to the independent coordinate system of each camera based on the exterior orientation elements of the two cameras in the stereo camera coordinate system to obtain vectors Pa and Pb;

[0024] ⑥Then, the corresponding image points Ca and Cb are obtained by intersecting the light vectors Pa and Pb with the image plane of their respective cameras;

[0025] ⑦ Substitute the two intersection points Ca and Cb into the corresponding egg-shaped curve equations F1 and F2 on each photograph to construct a system containing dynamic scanning angle parameters (correction parameter bn corresponding to the angle α between the rotation axis and the emitted laser optical axis) and coordinate transformation parameters. The constraint equations for ω, κ, Dx, Dy, and Dz;

[0026] ⑧ The conditional adjustment with parameters is used as the adjustment model, and the dynamic scan angle parameter bn is solved iteratively according to the least squares criterion;

[0027] ⑨ Repeat ①-⑧, record the different motor speeds D and the calculated correction amount bn of the α angle, and store them in arrays D[] and Bn[] respectively, recording at least 6 sets of data;

[0028] ⑩ After data collection is complete, the exponential fitting curve equation is shown below:

[0029] bn=p1*exp(-D / p2)+p3+p4*D

[0030] bn is the correction amount of the static scanning angle parameter α, D is the motor speed setting value; p1, p2, p3, and p4 are the parameters of the exponential fitting curve equation.

[0031] Substitute the data stored in arrays D[] and Bn[] in sequence, and use the least squares method to fit the data to obtain the fitting parameters p1, p2, p3, and p4, thus completing the solution of the dynamic correction parameters for the scanning angle.

[0032] During system operation, given the set rotation speed parameters, the correction amount bn corresponding to the included angle α is given by the dynamic angle correction model for correction. The corrected parameter α is then substituted into the laser point calculation process to calculate a more accurate three-dimensional coordinate of the laser point.

[0033] As a further technical solution of the present invention: the method for determining the scanning point positioning model is specifically as follows: using the law of light refraction and the ray tracing method, a scanning point positioning model is established based on the various angles in the scanning part.

[0034] As a further technical solution of the present invention: the automated extraction method of the scanning trajectory is as follows: the scanning trajectory of the blue-green laser depth sounding radar is captured by the camera, and the rotation trajectory of the scanning point is extracted from the image by the ellipse detection method of Hough transform, an automated image recognition algorithm.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] Based on the ray tracing model, this invention utilizes the visible characteristics of blue-green lasers, the egg-shaped curve of the system scanning trajectory, and a stereo measurement camera to construct a correction model for dynamic scanning angle fine calibration, thereby solving the problem of fine correction of optical path system angle decoupling after system integration. Attached Figure Description

[0037] Figure 1 This is a scanning trajectory diagram of a domestically produced airborne laser depth sounding system.

[0038] Figure 2 This is a technology roadmap.

[0039] Figure 3 A diagram is created for the dynamic scanning angle parameters and the radar coordinate system.

[0040] Figure 4 This is a diagram of the actual scanning trajectory of the depth-measuring lidar.

[0041] Figure 5 This is a diagram showing the detection results from a depth-sounding lidar.

[0042] Figure 6 This is a schematic diagram of a dynamic scanning angle correction method for an airborne depth sounding lidar.

[0043] Figure 7 This is a graph showing the relationship between the dynamic correction of the scanning angle and the motor speed.

[0044] Figure 8 This is a diagram showing the relationship between incident and reflected rays. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] The main contents of this invention include:

[0047] ① The scanning point positioning model was determined. Based on the domestic airborne blue-green laser scanning structure, the scanning point positioning model was established using the law of light refraction and the ray tracing method, with each angle in the scanning part as the theoretical basis.

[0048] ②Automatic extraction of scanning trajectory: The scanning trajectory of blue-green laser depth sounding radar is captured by the camera, and the rotation trajectory of the scanning point is extracted from the image using an automated image recognition algorithm, providing technical support for the automated processing of trajectory fitting and scanning angle correction.

[0049] ③ A method for dynamic scanning angle correction of an airborne depth-sounding lidar: First, different rotational speeds D are set, and the scanning system scans a facade, presenting a complete scanning trajectory curve on the facade. Second, using two digital cameras in a stereoscopic measurement system, camera exposure, shutter speed, and other parameters are set to capture images of the depth-sounding system's scanning trajectory at different rotational speeds D, obtaining scanning patterns. An automated extraction algorithm is used to extract the scanning trajectory from the images. Then, based on the egg-shaped scanning trajectory formed by the intersection of the scanning ray and the plane, and utilizing the relationship that the scanning trajectory, after coordinate transformation, should be collinear with the trajectory captured by the stereoscopic measurement camera, a constraint equation is constructed. The correction amount bn of the scanning angle parameter α at the corresponding rotational speed D is iteratively solved using the least squares criterion. At least six sets of data are recorded to construct a dynamic scanning angle correction model. The least squares method is used for fitting to obtain the fitting parameters p1, p2, p3, and p4, completing the solution for the dynamic scanning angle correction parameters. During system operation, given the set rotation speed parameters, the correction amount bn corresponding to the included angle α is given by the dynamic angle correction model for correction. The corrected parameter α is then substituted into the laser point calculation process to calculate a more accurate three-dimensional coordinate of the laser point.

[0050] This invention studies a dynamic scanning angle correction method for airborne depth sounding lidar, achieving decoupled and precise correction of the dynamic scanning angle after integration with the airborne depth sounding radar. This solves the problem of the inability to decouple and precisely correct the scanning angle during the industrialization of domestically produced airborne blue-green laser depth sounding radars.

[0051] Please see Figures 1-8 This invention discloses a method for dynamic scanning angle correction of airborne depth sounding lidar.

[0052] First, set up the venue:

[0053] a1. Select a flat wall surface as the image reference plane for the scanning trajectory, and ensure that the flatness variation within the scanning range is less than 1mm;

[0054] a2. such as Figure 6 As shown, an airborne depth sounding radar is placed at a certain vertical distance from a flat wall and at a certain height above the ground to ensure that the scanning trajectory can be completely displayed on the wall after the equipment is placed.

[0055] a3. Deploy stereo cameras. The placement of stereo measurement cameras should ensure that the coordinate system of the stereo measurement camera is approximately parallel to the axis system of the depth-sounding lidar. This ensures that the transformation angle parameter between the two coordinate systems is small, which can reduce the requirements for the accuracy of the initial value to a certain extent.

[0056] a4. Set up a total station and establish an independent local coordinate system. The coordinate system axis is basically consistent with the coordinate system of the depth sounding lidar system. Use the total station to measure the initial transformation parameters between the coordinate system of the stereo measuring camera and the coordinate system of the depth sounding lidar.

[0057] Among these, data acquisition is performed.

[0058] b1. Turn on the airborne depth-sounding lidar and begin data recording. Set the airborne depth-sounding lidar motor speed D (50 ≥ D ≤ 1000 r / min) according to the interval rotation speed (△D = 100 r / min) and perform a scanning operation to display a scanning trajectory on the wall, so that the wall (e.g., Figure 4 The scanning trajectory is presented.

[0059] Data was collected more than 6 times. Initially, D was set to 50 r / min, and then ΔD was increased sequentially to the maximum to complete the data collection for each set of rotational speeds D.

[0060] b2. Adjust the exposure parameters and shutter speed of the stereo measurement camera, and use the camera to capture the complete scanning trajectory and record the corresponding lidar data in the D-mode of the airborne depth sounding lidar rotation speed.

[0061] b3. The scan point trajectory is extracted from the image using an ellipse detection method based on Hough transform, such as... Figure 5 .

[0062] b4. Based on the Hough transform method, the trajectory formed by the scanning points captured on the photograph is detected. Using the least squares method and the egg-shaped curve formula, the fitting equations F1 and F2 formed by the scanning trajectory points captured in the two photographs are fitted.

[0063] Perform dynamic scan angle correction bn calculation:

[0064] ① Taking a certain set of data as an example, after the data collection is completed, the waveform file of the data is parsed to obtain the distance analysis file, and then two photos A and B are exported from the stereo measurement camera respectively;

[0065] ② The Hough transform ellipse detection method was used to extract the scanning point trajectory from photos A and B respectively, and the trajectory was fitted to obtain two egg-shaped curve fitting equations F1 and F2.

[0066] ③Use the scanning point positioning formula to obtain the direction vector Vs corresponding to each scanning point;

[0067] ④ Then, the direction vector Vs is adjusted according to the transformation parameters between the coordinate systems of the lidar and the stereo measurement system. ω, κ, Dx, Dy, and Dz are transformed into the stereo camera coordinate system using coordinate transformation to obtain the corresponding direction vector Vc;

[0068] ⑤ Transform the vector Vc to the independent coordinate system of each camera based on the exterior orientation elements of the two cameras in the stereo camera coordinate system to obtain vectors Pa and Pb;

[0069] ⑥Then, the corresponding image points Ca and Cb are obtained by intersecting the light vectors Pa and Pb with the image plane of their respective cameras;

[0070] ⑦ Substitute the two intersection points Ca and Cb into the corresponding egg-shaped curve equations F1 and F2 on each photograph to construct a curve containing the scanning angle parameter (the correction amount bn corresponding to the included angle α) and coordinate transformation parameters. The constraint equations for ω, κ, Dx, Dy, and Dz.

[0071] ⑧ The conditional adjustment with parameters is used as the adjustment model, and the dynamic scan angle correction bn is solved iteratively according to the least squares criterion.

[0072] Solving for the parameters of the dynamic scan angle correction model:

[0073] ① Record at least 6 sets of different motor speeds D and the calculated correction amount bn of angle α and store them in arrays D[] and Bn[] respectively, and record at least 6 sets of data;

[0074] ② After data collection is completed, the exponential fitting curve equation is shown below:

[0075] bn=p1*exp(-D / p2)+p3+p4*D

[0076] bn is the correction amount of the static scanning angle parameter α, D is the motor speed setting value; p1, p2, p3, and p4 are the parameters of the exponential fitting curve equation.

[0077] Substitute the data stored in arrays D[] and Bn[] in sequence, and use the least squares method to fit the data to obtain the fitting parameters p1, p2, p3, and p4, thus completing the solution of the dynamic correction parameters for the scanning angle.

[0078] ③ During system operation, given the set rotation speed parameters, the correction amount bn corresponding to the included angle α is given by the dynamic angle correction model for correction. The corrected parameter α is then substituted into the laser point calculation process to calculate a more accurate three-dimensional coordinate of the laser point.

[0079] The principle is as follows:

[0080] (1) Derivation of the scan point formula

[0081] In the auxiliary coordinate system, we can obtain the following: Figure 3 result;

[0082] Incident ray A i :

[0083]

[0084] α is the angle between the rotation axis and the emitted laser optical axis;

[0085] Normal N:

[0086]

[0087] β is the angle between the rotation axis and the normal of the rotating mirror, and θ is the zero position angle of the rotating mirror;

[0088] First, the incident ray A i Reverse the direction, and then reverse the incident ray -A i Rotate 180 degrees around the normal N to obtain the reflected ray A. r Position and direction, such as Figure 8 As shown:

[0089] Rotation matrix R a for:

[0090]

[0091] in These correspond to the three directional components of the normal.

[0092] The reflected ray is:

[0093] A r =R a *(-A i (4)

[0094] At this point, the direction vector of the reflected ray in the auxiliary coordinate system is obtained. Then, rotate it around the Y-axis by an angle α to the scanning coordinate system, and the direction vector of the reflected ray in the scanning coordinate system is obtained.

[0095]

[0096] Once the distance S from the laser to the target point is obtained, the scanning point P is then... S Positioning formula:

[0097] P S =S*R b *(R a *(-A i (6)

[0098] Incident light A i The rotation matrix R is used to rotate to the reflected ray. a ;

[0099] A i The incident light rays are reversed;

[0100] R b The rotation matrix is ​​used to transform the direction vector of the reflected ray in the auxiliary coordinate system to the corresponding rotation matrix in the scanning coordinate system;

[0101] (2) Formula for egg-shaped curve:

[0102] Depend on Figure 4 As can be seen, the scanning trajectory presents an egg-shaped curve structure on the image, exhibiting an egg-shaped curve that is "larger at one end" and "smaller at the other." An egg-shaped curve function can be constructed as follows:

[0103]

[0104] In the formula a e b e c e d e e e f e These are the parameters for the egg-shaped curve.

[0105] (3) Hough transform detection of the scan trajectory: The scan trajectory shows that the scanning pattern of the domestically produced airborne depth sounding radar is an egg-shaped curve, but it is also close to an ellipse. The actual scan trajectory is as follows: Figure 4 As shown in the figure, by controlling the camera exposure and shutter parameters, the scanning trajectory is clearly visible in the photo and can be distinguished from the background very well. Therefore, the scanning point trajectory can be extracted from the image using the ellipse detection method based on Hough transform.

[0106] The stochastic Hough transform uses a many-to-one mapping, and this random sampling leads to a large amount of unnecessary computation. When the number of points is large, the algorithm's performance drops sharply. The geometric characteristics of an ellipse can be used to reduce the dimensionality of the parameters, improving both efficiency and accuracy. Based on this, improvements to the Hough transform are needed, and the improved dependency theorem is as follows:

[0107] Suppose there is an ellipse on the plane, with point c as the center of the ellipse. If we take any point p on the plane (different from point c), the maximum distance from point p to any point on the ellipse is always greater than the maximum distance from point c to any point on the ellipse.

[0108] Based on the above theorem, firstly, edge detection is performed on the image to obtain a binary edge contour map, and the coordinates of the points on the edge map are stored in array A. Secondly, for each point on the image, the distance to the points in array A obtained in the previous step is calculated to obtain the maximum distance of each point from the points in array A. The point with the smallest maximum distance among all points is the center of the ellipse (P, Q), and this maximum distance is the length a of the major axis of the ellipse. Then, the value of each point in array A and the three ellipse parameters P, Q, and a obtained above are substituted into the ellipse equation (8).

[0109]

[0110] Where: a and b are the major and minor axes of the ellipse, P and Q are the coordinates of the center of the ellipse, and θ is the rotation angle of the ellipse.

[0111] Finally, by statistically analyzing parameters b and θ, a set of parameters whose peak values ​​exceed a certain threshold is identified as an ellipse, thus achieving automated detection of the scanning trajectory. The detection results are as follows: Figure 5 :

[0112] The method for dynamic correction of the scan angle is as follows:

[0113] The lidar positioning equation includes both angle and distance information. To reduce the impact of distance error on angle, a dual-camera stereo measurement principle is used for model construction.

[0114] Turn on the airborne depth sounding lidar and start data recording. Set the airborne depth sounding lidar motor speed D (50≥D≤1000r / min) according to the interval speed (△D=100r / min) and perform a scanning on the wall to present the scanning trajectory. After obtaining two photos containing the scanning trajectory of the egg-shaped curve and the waveform data of the corresponding time period through the stereo camera, establish stereo image pair constraints with the two egg-shaped curves in the two photos. The scanning data trajectory should simultaneously satisfy the collinearity relationship on the two photos and satisfy the egg-shaped curve equations on the corresponding photos.

[0115] The solution for dynamic scan angle correction bn consists of the following steps:

[0116] ① Taking a certain set of data as an example, after the data collection is completed, the waveform file of the data is parsed to obtain the distance analysis file, and then two photos A and B are exported from the stereo measurement camera respectively;

[0117] ② The Hough transform ellipse detection method was used to extract the scanning point trajectory from photos A and B respectively, and the trajectory was fitted to obtain two egg-shaped curve fitting equations F1 and F2.

[0118] ③Use the scanning point positioning formula to obtain the direction vector Vs corresponding to each scanning point;

[0119] ④ Then, the direction vector Vs is adjusted according to the transformation parameters between the coordinate systems of the lidar and the stereo measurement system. ω, κ, Dx, Dy, and Dz are transformed into the stereo camera coordinate system using coordinate transformation to obtain the corresponding direction vector Vc;

[0120] ⑤ Transform the vector Vc to the independent coordinate system of each camera based on the exterior orientation elements of the two cameras in the stereo camera coordinate system to obtain vectors Pa and Pb;

[0121] ⑥Then, the corresponding image points Ca and Cb are obtained by intersecting the light vectors Pa and Pb with the image plane of their respective cameras;

[0122] ⑦ Substitute the two intersection points Ca and Cb into the corresponding egg-shaped curve equations F1 and F2 on each photograph to construct a curve containing the scanning angle parameter (the correction amount bn corresponding to the included angle α) and coordinate transformation parameters. The constraint equations for ω, κ, Dx, Dy, and Dz.

[0123] ⑧ The conditional adjustment with parameters is used as the adjustment model, and the dynamic scan angle correction bn is solved iteratively according to the least squares criterion.

[0124] All coordinate transformation matrices use the rotation matrix of the Euler angle model, which is derived from... The three Euler angles ω and κ are represented as follows:

[0125]

[0126] in, Let x be the rotation matrix about the x-axis. R is the rotation angle about the X-axis. y (ω) is the rotation matrix about the y-axis, ω is the rotation angle about the y-axis, and R z (κ) is the rotation matrix about the z-axis, and κ is the rotation angle about the Z-axis. The three independent rotation matrices are shown in the following equation:

[0127]

[0128]

[0129]

[0130] The coordinate transformation model is as follows:

[0131]

[0132] Dx, Dy, and Dz are translation parameters, X i ', Y i ', Z i ' represents the transformed coordinates, X' i Y i Z i Coordinates before transformation;

[0133] Solving for the parameters of the dynamic scan angle correction model:

[0134] ① Record at least 6 sets of different motor speeds D and the calculated correction amount bn of angle α and store them in arrays D[] and Bn[] respectively, and record at least 6 sets of data;

[0135] ② After data collection is completed, the exponential fitting curve equation is shown below:

[0136] bn=p1*exp(-D / p2)+p3+p4*D

[0137] bn is the correction amount of the static scanning angle parameter α, D is the motor speed setting value; p1, p2, p3, and p4 are the parameters of the exponential fitting curve equation.

[0138] Substitute the data stored in arrays D[] and Bn[] into the arrays in turn, and use the least squares method to fit the data to obtain the fitting parameters p1, p2, p3, and p4, thus completing the solution of the dynamic correction parameters for the scanning angle.

[0139] ③ During system operation, given the set rotation speed parameters, the correction amount bn corresponding to the included angle α is given by the dynamic angle correction model for correction. The corrected parameter α is then substituted into the laser point calculation process to calculate a more accurate three-dimensional coordinate of the laser point.

[0140] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0141] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for dynamic scanning angle correction of an airborne depth-sounding lidar, characterized in that, Specifically, it includes the following steps: a) Set up the site and, according to the design plan, install a stereo measurement system consisting of two calibrated cameras and a depth-sounding lidar. b) Set the rotation speed of 6 sets of airborne depth-sounding lidar motors at constant speed and perform scanning with the light on. The sampling range of the rotation speed setting should cover the minimum and maximum rotation speed range of the equipment motors. At the same time, use the stereo measurement system to completely capture the scanning trajectory of the laser point at the corresponding rotation speed. c) Calculate the correction amount bn corresponding to the angle α between the rotating shaft and the emitted laser optical axis at different rotation speeds based on the static scanning angle correction model, and record more than 6 sets of data; d) After the data recording is completed, the parameters of the dynamic scanning angle correction model of the airborne depth sounding lidar are solved by using the exponential fitting curve equation. During the operation of the system, when the rotation speed parameter is set, the correction amount bn corresponding to the included angle α given by the dynamic angle correction model is used for correction. The correction parameter α is substituted into the laser point calculation process to obtain a more accurate three-dimensional coordinate of the laser point.

2. The method for dynamic scanning angle correction of an airborne depth-sounding lidar according to claim 1, characterized in that, Step a includes the following sub-steps: a1. Select a flat wall surface as the image reference plane for the scanning trajectory, and ensure that the flatness variation within the scanning range is less than 1mm; a2. Place the airborne depth sounding radar at a certain vertical distance from the flat wall and at a certain height above the ground to ensure that the scanning trajectory can be completely displayed on the wall after the equipment is placed. a3. Deploy stereo cameras. The placement of stereo measurement cameras should ensure that the coordinate system of the stereo measurement camera is approximately parallel to the axis system of the depth-sounding lidar. This ensures that the transformation angle parameter between the two coordinate systems is small, which can reduce the requirements for the accuracy of the initial value to a certain extent. a4. Set up a total station and establish an independent local coordinate system. The coordinate system axis is basically consistent with the coordinate system of the depth sounding lidar system. Use the total station to measure the initial transformation parameters between the coordinate system of the stereo measuring camera and the coordinate system of the depth sounding lidar.

3. The method for dynamic scanning angle correction of an airborne depth-sounding lidar according to claim 1, characterized in that, Step b includes the following sub-steps: b1. Turn on the airborne depth-sounding lidar and start data recording. Set the airborne depth-sounding lidar motor speed D according to the interval speed △D, and perform a light-on scan to show the scanning trajectory on the wall. 50≥D≤1000r / min, △D=100r / min, and set D to 50r / min for the first time. b2. Adjust the exposure parameters and shutter speed of the stereo measurement camera, and use the camera to capture the complete scanning trajectory and record the corresponding lidar data in the D-mode of the airborne depth sounding lidar rotation speed; b3. Extract the scan point trajectory from the image using an ellipse detection method based on Hough transform; b4. Based on the Hough transform method, the trajectory formed by the scanning points captured in the photograph is detected. Using the least squares method and the egg-shaped curve formula, the fitting equations F1 and F2 formed by the scanning trajectory points captured in the two photographs are fitted.

4. The method for dynamic scanning angle correction of an airborne depth-sounding lidar according to claim 1, characterized in that, Step c includes the following sub-steps: c1. Using the scanning point positioning formula, obtain the direction vector Vs corresponding to each scanning point; c2. Then, the direction vector Vs is adjusted according to the transformation parameters between the lidar and the stereo measurement system coordinate system. ω, κ, Dx, Dy, and Dz are transformed into the stereo camera coordinate system using coordinate transformation to obtain the corresponding direction vector Vc; c3. Transform the vector Vc to the independent coordinate system of each camera based on the exterior orientation elements of the two cameras in the stereo camera coordinate system to obtain vectors Pa and Pb; c4. Then, intersect the light vectors Pa and Pb with the image plane of their respective cameras to solve for the corresponding image points Ca and Cb; c5. Substitute the two intersection points Ca and Cb into the corresponding egg-shaped curve equations F1 and F2 on each photograph, respectively, to construct a curve containing the scan angle parameter and coordinate transformation parameter. The constraint equations for ω, κ, Dx, Dy, and Dz, and the scanning angle parameters include the angle α between the rotation axis and the emitted laser optical axis, the angle β between the rotation axis and the normal of the rotating mirror, and the zero position angle θ of the rotating mirror; c6. Use conditional adjustment with parameters as the adjustment model, and iteratively solve the correction amount bn of the static scan angle parameter α according to the least squares criterion; C7. Repeat c1-c6, and record the different motor speeds D and the calculated correction amount bn of the α angle in arrays D[] and Bn[] respectively, recording at least 6 sets of data.

5. The method for dynamic scanning angle correction of an airborne depth-sounding lidar according to claim 1, characterized in that, Step d includes the following sub-steps: d1. After data collection is complete, the exponential fitting curve equation is shown below: bn=p1*exp(-D / p2)+p3+p4*D bn is the correction amount of the static scanning angle parameter α, D is the motor speed setting value, and p1, p2, p3, and p4 are the parameters of the exponential fitting curve equation. Substitute the data stored in arrays D[] and Bn[] in sequence, and use the least squares method to fit the data to obtain the fitting parameters p1, p2, p3, and p4, thus completing the solution of the dynamic correction parameters for the scanning angle. d2. During system operation, given the set rotation speed parameters, the correction amount bn corresponding to the included angle α is given by the dynamic angle correction model for correction. The corrected parameter α is substituted into the laser point calculation process to calculate a more accurate three-dimensional coordinate of the laser point.

Images