An automatic calibration method for airborne lidar placement angle based on point cloud conjugate matching

Abstract

The present invention discloses an automatic calibration method for the installation angle of an airborne laser radar based on point cloud conjugate matching, which is used to eliminate the installation angle error of the airborne laser radar system, including: extracting conjugate point clouds of buildings with the same name on different routes or at different times; screening the point clouds using cloth filtering and threshold filtering; calculating the corresponding point cloud shapes under different installation angles and displacements; using the point cloud grid area method to calculate the point cloud coverage area; and outputting the heading angle, pitch angle, and roll angle corresponding to the minimum point cloud area, that is, the maximum point cloud overlap. Utilizing the method of the present invention, the efficiency of the installation angle calibration of the airborne laser radar system can be effectively improved, the requirements for the calibration site can be reduced, manual intervention can be reduced to achieve automated system installation angle calibration, and the detection accuracy of the airborne laser radar can be effectively improved.

Description

Technical Field

[0001] The present invention belongs to the field of laser radar measurement technology, and in particular relates to an automatic calibration method for the installation angle of an airborne laser radar based on point cloud conjugate matching. Background Art

[0002] Airborne LiDAR (LiDAR) has become a rapidly developing new measurement technology in recent years, thanks to its wide coverage, high detection efficiency, and high measurement accuracy. The Global Positioning System (GPS) and Inertial Measurement Unit (IMU) modules in an airborne LiDAR system respectively record the current flight position and aircraft attitude information, thereby determining the spatial vector from the laser emission point to the target position and obtaining the current laser footprint coordinates. The coordinate accuracy of the point cloud depends on the measurement and installation errors of each module. To generate high-precision point cloud data, it is necessary to minimize system errors. Due to the constraints of the installation process during the integration of the IMU and laser emission module, the IMU's coordinate system and the laser emission coordinate system do not completely overlap, resulting in slight angular deviations. This deviation is difficult to accurately obtain through direct measurement. However, angular errors have a significant impact on airborne LiDAR data processing and cannot be ignored.

[0003] For example, Chinese patent document CN107621628A discloses a method for calibrating installation angle errors. By selecting flat terrain, tall buildings, and straight roads as the installation angle error calibration field, four mutually perpendicular front, back, left, and right flight routes are designed for scanning and measurement, and manual calibration is performed based on the obtained overlapping point clouds.

[0004] For example, Chinese patent document CN112859052A discloses an airborne lidar system integrated error calibration method based on overlapping flight strip conjugate primitives. By establishing a mathematical model, the conjugate primitives between the reference point cloud and the template point cloud are found, the optimal transformation relationship between the point clouds is solved, and the system integrated error calibration matrix is ​​obtained by inverse solution to complete the system placement angle calibration.

[0005] The current method for calibrating the mounting angle has strict requirements for the calibration site, complex flight path design, and requires manual calibration, which reduces the efficiency of airborne LiDAR. Therefore, there is an urgent need to develop an automatic calibration method for the mounting angle that has low calibration site requirements and fewer flight path requirements, thereby enhancing the efficiency of airborne LiDAR and improving detection accuracy. Summary of the Invention

[0006] In response to the problems of complicated installation angle calibration process, high calibration field requirements and high degree of manual intervention in existing airborne laser radars, the present invention provides an automatic calibration method for the installation angle of airborne laser radars based on point cloud conjugate matching, which can greatly improve the installation angle calibration speed and enhance the detection accuracy of airborne laser radars.

[0007] A method for automatically calibrating the placement angle of an airborne laser radar based on point cloud conjugate matching, comprising:

[0008] (1) Establish the sensor coordinate system according to the elliptical scanning method of the airborne laser radar and calculate the original point cloud;

[0009] (2) Perform heading angle calibration, select a single route covering the characteristic building, and select the conjugate point cloud of the building from the original point cloud;

[0010] (3) performing cloth filtering and threshold filtering on the conjugate point cloud extracted in step (2), and screening the aircraft attitude data corresponding to the point cloud;

[0011] (4) Setting a set of heading angle deviation values, and using the filtered conjugate point cloud and the corresponding aircraft attitude data to calculate the point clouds at two moments under different heading angle deviation values;

[0012] (5) setting the grid resolution according to the point cloud density of the airborne laser radar, and calculating the coverage area of ​​the point cloud after processing in step (4);

[0013] (6) According to step (5), calculate the point cloud coverage area at two moments before and after under different heading deviation values; calculate the heading deviation angle with the smallest coverage area among a set of heading angle deviation values, that is, the point cloud overlap at different moments in the current scanning trajectory is the highest, and this heading deviation angle is the placement heading angle;

[0014] (7) Perform roll and pitch angle calibration and extract the conjugate point clouds of characteristic buildings on the two routes;

[0015] (8) performing cloth filtering and threshold filtering on the conjugate point cloud extracted in step (7), and screening the point cloud corresponding to the aircraft attitude data;

[0016] (9) setting a set of displacement deviation values, and adding the displacement deviation values ​​to the point cloud processed in step (8);

[0017] (10) setting the grid resolution according to the point cloud density of the airborne laser radar, and calculating the grid area of ​​the point cloud after processing in step (9);

[0018] (11) According to step (10), the point cloud coverage area under different displacement deviation values ​​is calculated, and the displacement deviation with the smallest coverage area among a set of displacement deviation values ​​is further used to calculate the placement pitch angle and roll angle.

[0019] Furthermore, in step (1), an airborne laser radar scanning coordinate system is established, and a sensor coordinate system is established according to a quasi-elliptical scanning method; taking the northeast sky coordinate system as an example, the coordinate origin is the center of the reflector, the X axis is along the direction of the laser emission light, the Y axis is along the flight direction, and the Z axis is vertically upward. The unit vector P of the emission light is expressed as:

[0020]

[0021] in, is the azimuth angle of the outgoing light, ψ is the zenith angle of the outgoing light;

[0022] The aircraft attitude angle matrix Cb is expressed as:

[0023]

[0024] Among them, pitch, roll, and heading are the corresponding pitch angle, roll angle, and heading angle data in the aircraft attitude data;

[0025] The outgoing ray unit vector is expressed as:

[0026] P=Cb′×P′

[0027] Where P′ is the unit vector of the outgoing light after the aircraft attitude angle changes, and P is the outgoing light in the sensor coordinate system.

[0028] In step (2), for the selection of heading angle calibration point clouds, select an area with obvious buildings on a single flight strip. Select the point clouds at two moments before and after the elliptical scanning trajectory as the conjugate point clouds for heading angle calibration.

[0029] In step (3), during the cloth filtering process, the scene model is selected according to the point cloud information, the grid size is set according to the point cloud density, the number of iterations required by the algorithm is set, and the classification threshold between ground points and non-ground points is set; during the threshold filtering process, the appropriate height threshold is set according to the actual height of the selected building point cloud, and all point cloud data greater than the set threshold are extracted; the heading angle, roll angle and pitch angle aircraft attitude angle data corresponding to the point cloud are screened.

[0030] In step (4), the aircraft attitude data extracted in step (3) is substituted into the attitude matrix Cb to obtain the placement angle attitude matrix Cb′, and the point clouds corresponding to different heading angle deviation values ​​are calculated; wherein, the placement angle attitude matrix Cb′ is expressed as:

[0031]

[0032] Wherein, Δp is the pitch angle, Δr is the roll angle, and Δh is the heading angle.

[0033] The outgoing ray unit vector is expressed as:

[0034] P=Cb×Cb′×P″

[0035] Where P″ is the unit vector of the outgoing light after the installation angle is calibrated and the aircraft attitude angle is changed, and P is the outgoing light in the sensor coordinate system. Δp and Δr are set to 0, and Δh is set to the heading angle deviation value, which is then substituted into the installation angle attitude matrix Cb′.

[0036] In step (5), the calculation process of the point cloud coverage area is as follows:

[0037] Set the grid array interval according to the point cloud coordinate range, set the grid resolution according to the point cloud density of the airborne lidar (for example, the grid resolution is set to 1m′×1m), and search each grid to see if there is point cloud data. If so, assign a value of 1, otherwise assign a value of 0.

[0038] Considering that the airborne lidar scans along an elliptical trajectory and the point cloud is uneven, interpolation processing is performed in the direction from the scanning trajectory to the scanning center to complete the missing point cloud data in the middle; the point cloud area grid array is counted to calculate the point cloud coverage area.

[0039] In step (7), two routes that fly in opposite directions and have a certain degree of overlap are selected, and areas with obvious buildings are selected on the routes, and conjugate point clouds of the same building in the two routes are extracted.

[0040] In step (9), the displacement deviation value is divided into two sets of data, one along the direction of aircraft movement, i.e., the Y-axis direction, and the other perpendicular to the direction of aircraft movement, i.e., the X-axis direction. The displacement deviation value is added to the point cloud coordinates processed in step (8) to obtain a new set of point cloud coordinates.

[0041] In step (11), the formula for calculating the placement pitch angle Δp is:

[0042]

[0043] Where Δy is the displacement deviation of the point cloud of the object with the same name in the overlapping flight strip along the flight direction, and H is the flight altitude of the aircraft;

[0044] The formula for calculating the installation roll angle Δr is:

[0045]

[0046] Where Δx is the displacement deviation of the point cloud of the object with the same name in the overlapping flight strip along the direction perpendicular to the flight path, and H is the flight altitude of the aircraft.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. This invention, based on a point cloud conjugate matching method, extracts conjugate point clouds of buildings with the same name from different routes or at different times, filters them with cloth and threshold filters, and uses point cloud area statistics to calculate the degree of overlap of overlapping point clouds. Ultimately, it calculates the heading, roll, and pitch angles of the installation. This allows for automated calibration of the installation angle of the airborne LiDAR system, eliminating the need for excessive manual intervention.

[0049] 2. The present invention uses calibration routes with no specific requirements for building selection and does not require multiple perpendicular and parallel routes, effectively reducing the existing method's demand for calibration sites and alleviating the calibration route design task. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of a method for automatically calibrating the installation angle of an airborne laser radar based on point cloud conjugate matching according to the present invention;

[0051] Figure 2 This is a schematic diagram of the heading angle calibration principle in the present invention;

[0052] Figure 3 is the point cloud result before heading angle calibration in an embodiment of the present invention;

[0053] Figure 4 is the point cloud result after heading angle calibration in an embodiment of the present invention;

[0054] Figure 5 This is a schematic diagram of the roll angle and pitch angle calibration principle in the present invention;

[0055] Figure 6 This is the point cloud result before roll angle and pitch angle calibration in the embodiment of the present invention

[0056] Figure 7 This is the point cloud result after roll angle and pitch angle calibration in the embodiment of the present invention

[0057] Figure 8 This is the point cloud result after calibration in the embodiment of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be described in further detail below with reference to the accompanying drawings and examples. It should be noted that the following examples are intended to facilitate understanding of the present invention and do not have any limiting effect on the present invention.

[0059] like Figure 1As shown in the figure, an automatic calibration method for the installation angle of an airborne lidar based on point cloud conjugate matching is proposed. The method extracts conjugate point clouds of characteristic buildings at different routes or at different times; filters the point clouds using cloth filtering and threshold filtering; calculates the corresponding point cloud shapes under different installation angles and displacements, and uses the point cloud grid area method to calculate the point cloud coverage area; and outputs the heading angle, pitch angle, and roll angle corresponding to the minimum point cloud area and the maximum point cloud overlap, specifically including:

[0060] S01, establish the sensor coordinate system according to the elliptical scanning method of the airborne laser radar and calculate the original point cloud;

[0061] S02, perform heading angle calibration, select a single route covering the characteristic building, and select the conjugate point cloud of the building from the original point cloud;

[0062] S03, performing cloth filtering and threshold filtering on the conjugate point cloud extracted in step S02, and screening the aircraft attitude data corresponding to the point cloud;

[0063] S04, setting a set of heading angle deviation values, and using the filtered conjugate point cloud and the corresponding aircraft attitude data to calculate point clouds at two moments under different heading angle deviation values;

[0064] S05, setting the grid resolution according to the point cloud density of the airborne laser radar, and counting the coverage area of ​​the point cloud processed in step S04;

[0065] S06, according to step S05, calculating the point cloud coverage area at two moments before and after under different heading deviation values; calculating the heading deviation angle with the smallest coverage area among a set of heading angle deviation values, that is, the heading deviation angle with the highest overlap of point clouds at different moments in the current scanning trajectory, and this heading deviation angle is the placement heading angle;

[0066] S07, calibrate the roll and pitch angles and extract the conjugate point clouds of the characteristic buildings on the two routes;

[0067] S08, performing cloth filtering and threshold filtering on the conjugate point cloud extracted in step S07, and screening the point cloud for aircraft attitude data corresponding to the point cloud;

[0068] S09, setting a set of displacement deviation values, and adding the displacement deviation values ​​to the point cloud processed in step S08;

[0069] S10, setting the grid resolution according to the point cloud density of the airborne laser radar, and counting the grid area of ​​the point cloud processed in step S09;

[0070] S11, according to step S10, calculate the point cloud coverage area under different displacement deviation values, and further calculate the placement pitch angle and roll angle for the displacement deviation with the smallest coverage area among a set of displacement deviation values.

[0071] Figure 2 This paper demonstrates the principle of calibration for the installation heading angle employed in this invention. An airborne laser radar maps the terrain using a quasi-elliptical scanning trajectory. As it passes over a target building, the front and rear semicircles of the scanning trajectory pass through the building at different times. If there is heading angle deviation during this time, this will result in deviations in the target building point cloud at different times. Based on this principle, a set of heading angle deviation values ​​is set and substituted into the coordinate calculation formula to obtain the target building point cloud at different heading angle deviations. The area method is then used to calculate the heading angle deviation corresponding to the point cloud with the highest degree of overlap, which is the installation heading angle.

[0072] Figure 3 and Figure 4 The following table shows the building point clouds before and after heading angle calibration. Table 1 shows overlapping point clouds from different routes, and the heading angle deviation of these overlapping point clouds was calculated using the above method. The average heading angle is 1.5°.

[0073] Table 1

[0074] Select point cloud sequence Setting heading angle / ° 1 1.40 2 1.60

[0075] Figure 5 This paper demonstrates the principle behind the calibration of installation pitch and roll angles. By selecting routes with opposite and overlapping flight paths, if the installation roll and pitch angles are present, displacement deviations will be present in the building point clouds in the overlapping area. Based on this principle, a set of displacement deviation arrays is established, and building point clouds are calculated for different displacement deviations. The area method is then used to calculate the displacement deviation corresponding to the point cloud with the highest degree of overlap, and the installation roll and pitch angles are then calculated.

[0076] Figure 6 and Figure 7 The following table shows the building point clouds before and after pitch and roll angle calibration. Table 2 shows overlapping point clouds from different flight paths. The displacement differences between the overlapping point clouds were calculated using the aforementioned method, along with the corresponding roll and pitch angles. The average roll angle was 1.19°, and the average pitch angle was 0.31°.

[0077] Table 2

[0078] Select point cloud sequence Roll angle / ° Installation pitch angle / ° 1 1.19 0.31 2 1.27 0.27 3 1.16 0.31 4 1.22 0.33 5 1.13 0.32

[0079] Figure 8 This demonstration shows point cloud results for multiple overlapping flight paths after calibration of the airborne lidar's positioning angle. The resulting point cloud agrees well with the high-resolution map. Compared to existing calibration techniques, this method reduces the need for calibration sites and flight path design, reduces manual intervention, improves positioning angle calibration efficiency, and enhances detection accuracy.

[0080] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching, characterized in that: include: (1) Establish the sensor coordinate system according to the elliptical scanning method of the airborne laser radar and calculate the original point cloud; (2) Perform heading angle calibration, select a single route covering the characteristic building, and select the conjugate point cloud of the building from the original point cloud; (3) performing cloth filtering and threshold filtering on the conjugate point cloud extracted in step (2), and screening the aircraft attitude data corresponding to the point cloud; (4) Setting a set of heading angle deviation values, and using the filtered conjugate point cloud and the corresponding aircraft attitude data to calculate the point clouds at two moments under different heading angle deviation values; (5) setting the grid resolution according to the point cloud density of the airborne laser radar, and calculating the coverage area of ​​the point cloud after processing in step (4); (6) According to step (5), calculate the point cloud coverage area at two moments before and after under different heading deviation values; calculate the heading deviation angle with the smallest coverage area among a set of heading angle deviation values, that is, the point cloud overlap at different moments in the current scanning trajectory is the highest, and this heading deviation angle is the placement heading angle; (7) Perform roll and pitch angle calibration and extract the conjugate point clouds of characteristic buildings on the two routes; (8) performing cloth filtering and threshold filtering on the conjugate point cloud extracted in step (7), and screening the point cloud corresponding to the aircraft attitude data; (9) setting a set of displacement deviation values, and adding the displacement deviation values ​​to the point cloud processed in step (8); (10) setting the grid resolution according to the point cloud density of the airborne laser radar, and calculating the grid area of ​​the point cloud after processing in step (9); (11) According to step (10), the point cloud coverage area under different displacement deviation values ​​is calculated, and the displacement deviation with the smallest coverage area among a set of displacement deviation values ​​is further used to calculate the placement pitch angle and roll angle.

2. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1 is characterized in that: In step (1), the airborne laser radar scanning coordinate system is established, and the sensor coordinate system is established according to the elliptical scanning method; taking the northeast sky coordinate system as an example, the coordinate origin is the center of the reflector, the X axis is along the direction of the laser light, the Y axis is along the flight direction, and the Z axis is vertically upward. The unit vector P of the light is expressed as: in, is the azimuth angle of the outgoing light, ψ is the zenith angle of the outgoing light; The aircraft attitude angle matrix Cb is expressed as: Among them, pitch, roll, and heading are the corresponding pitch angle, roll angle, and heading angle data in the aircraft attitude data; The outgoing ray unit vector is expressed as: P=Cb×P′ Where P′ is the unit vector of the outgoing light after the aircraft attitude angle changes, and P is the outgoing light in the sensor coordinate system.

3. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1 is characterized in that: In step (2), the point clouds at two moments before and after the elliptical scanning trajectory are selected as conjugate point clouds for heading angle calibration.

4. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1 is characterized in that: In step (3), during the cloth filtering process, the scene model is selected according to the point cloud information, the grid size is set according to the point cloud density, the number of iterations required by the algorithm is set, and the classification threshold between ground points and non-ground points is set; during the threshold filtering process, the appropriate height threshold is set according to the actual height of the selected building point cloud, and all point cloud data greater than the set threshold are extracted; the heading angle, roll angle and pitch angle aircraft attitude angle data corresponding to the point cloud are screened.

5. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1, characterized in that: In step (4), the aircraft attitude data extracted in step (3) is substituted into the attitude matrix Cb to obtain the placement angle attitude matrix Cb′, and the point clouds corresponding to different heading angle deviation values ​​are calculated; wherein, the placement angle attitude matrix Cb′ is expressed as: Wherein, Δp is the pitch angle, Δr is the roll angle, and Δh is the heading angle. The outgoing ray unit vector is expressed as: P=Cb×Cb′×P″ Where P″ is the unit vector of the outgoing light after the installation angle is calibrated and the aircraft attitude angle is changed, and P is the outgoing light in the sensor coordinate system. Δp and Δr are set to 0, and Δh is set to the heading angle deviation value, which is then substituted into the installation angle attitude matrix Cb′.

6. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1, characterized in that: In step (5), the calculation process of the point cloud coverage area is as follows: The grid array interval is set according to the point cloud coordinate range, and the grid resolution is set according to the point cloud density of the airborne lidar. Each grid is searched to see if there is point cloud data. If so, the value is assigned to 1, and if not, the value is assigned to 0. Considering that the airborne lidar scans along an elliptical trajectory and the point cloud is uneven, interpolation processing is performed in the direction from the scanning trajectory to the scanning center to complete the missing point cloud data in the middle; the point cloud area grid array is counted to calculate the point cloud coverage area.

7. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1, characterized in that: In step (7), two routes that fly in opposite directions and have a certain degree of overlap are selected, and areas with obvious buildings are selected on the routes, and conjugate point clouds of the same building in the two routes are extracted.

8. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1, characterized in that: In step (9), the displacement deviation value is divided into two sets of data, one along the direction of aircraft movement, i.e., the Y-axis direction, and the other perpendicular to the direction of aircraft movement, i.e., the X-axis direction. The displacement deviation value is added to the point cloud coordinates processed in step (8) to obtain a new set of point cloud coordinates.

9. The method for automatic calibration of airborne laser radar installation angle based on point cloud conjugate matching according to claim 1, characterized in that: In step (11), the formula for calculating the placement pitch angle Δp is: Where Δy is the displacement deviation of the point cloud of the object with the same name in the overlapping flight strip along the flight direction, and H is the flight altitude of the aircraft; The formula for calculating the installation roll angle Δr is: Where Δx is the displacement deviation of the point cloud of the object with the same name in the overlapping flight strip along the direction perpendicular to the flight path, and H is the flight altitude of the aircraft.

Images