A method for error identification and calibration of an inertial laser scanner system

Abstract

This invention relates to a method for identifying and calibrating system errors in an inertial laser scanner. The specific steps of the calibration method are as follows: Step 1: Determine the system coordinate system; Step 2: Calibrate the tooling head: Obtain the respective point cloud datasets {C}; a. Determine the straight line model in the laser scanner coordinate system based on {C}; b. Repeat step a to form a new subset of interior points {C1}, and obtain the straight line fitting parameters {k}. m1 ,b m1 c. In {C}, remove {C1}, repeat steps a and b, and obtain the line fitting parameters {k}. m2 ,b m2} and point cloud subset {C2}; d. In {C}, remove {C1} and {C2}, repeat steps a and b, and obtain the line fitting parameters {k}. m3 ,b m3} and point cloud subset {C3}; third step, from {C f The first step is to select the part belonging to the head of the calibration fixture from {C1,C2,C3}; the second step is to calculate the initial value of the transformation matrix of the laser scanner measurement result; the third step is to use R0 and t0 as the initial matching values ​​to obtain the final accurate value; the fourth step is to calculate the transformation matrix between the inertial navigation system and the laser scanner. This invention realizes the automatic identification of small system errors.

Description

Technical Field

[0001] This invention belongs to the field of laser image processing technology, specifically relating to a method for identifying and calibrating errors in an inertial laser scanner system. Background Technology

[0002] An inertial / laser scanner is a commonly used combination of measurement equipment. It acquires the pose of a carrier in inertial space through an inertial navigation system, measures the structural shape of the target through laser scanning, identifies the structural features of the target from the scanning results, calculates the relative pose between the target and the inertial navigation system, thereby obtaining the type of structural defects and calculating the location of the defects.

[0003] High-precision laser scanners generally have a small scanning range. Therefore, when integrating systems, it is often necessary to use a configuration of one inertial navigation system with multiple laser scanners. The scanners are installed close to the target being measured and are far away from the inertial navigation system. As a result, there is a geometric transformation relationship between the measurement coordinate systems of the inertial navigation system and the multiple lasers. In order to correctly calculate the spatial positional relationship between the targets being measured, it is necessary to calibrate the geometric transformation relationship between the inertial navigation system and the multiple lasers.

[0004] Approximate values ​​of geometric transformation relationships can be obtained through mechanical measurement, but residual small errors can still have a significant impact on the overall measurement accuracy of the system. Summary of the Invention

[0005] The purpose of this invention is to provide a method for identifying and calibrating system errors in an inertial laser scanner, which enables automatic identification of small system errors.

[0006] The technical solution of this invention is as follows: The specific steps of a method for error identification and calibration of an inertial laser scanner system are as follows:

[0007] Step 1: Determine the system coordinate system: The system coordinate system is located within the cross-section of calibration fixture 3.

[0008] o r —Calibrate the origin of the coordinate system on the left side of the fixture, o rr —Calibrate the origin of the coordinate system on the right side of the fixture, o b —Origin of the inertial conductor coordinate system, o n —The origin of the coordinate system measured by the laser scanner, n = 1, 3; representing the right and left laser scanners respectively. —Origin of the laser scanner reference plane coordinate system —The position of the origin of the reference plane of laser scanner n in the inertial conductor coordinate system;

[0009] Step 2: Locate the head of the calibration fixture in the laser measurement point cloud: Within the cross-section of the calibration fixture 3, two laser scanners scan the calibration fixture from the left and right sides respectively, and obtain their respective point cloud datasets {C}; Steps a to d are performed on the two point cloud datasets {C} respectively.

[0010] a. Randomly select two points from the point cloud dataset {C} acquired by the laser scanner; use these two points to determine the straight line model in the laser scanner coordinate system; set the interior point verification error ε. r , ε r =1~2mm, the inner point is the distance from the line to ε. r Find the points; input all point cloud data in {C} into the linear model and calculate the number of interior points; refit the line using the interior points and record the fitting parameters {k}. i ,b i};k i ,b i Let represent the slope and intercept of the linear model, respectively; i is the fitting order, i = 1, 2, ...

[0011] b. Repeat step a until k. i ,b i Less than the verification error {ε k ,ε b}, and select the set of parameters {k} with the largest number of interior points. m ,b m};Verification error {ε k ,ε b} is one ten-thousandth, where ε k ,ε b The verification errors for the slope and intercept are respectively represented by the parameter {k}, which is less than the error threshold. i ,b i The subset of internal points of {k} m ,b m Merge the sets of all internal points to form a new subset of internal points {C1}, where k i ∈[k m -ε k ,k m +ε k ], b i ∈[b m -ε b ,b m +ε b The new subset of interior points after merging is used to refit the line, obtaining the line fitting parameters {k}. m1 ,b m1}; Complete the fitting of a straight line;

[0012] c. In the original point cloud subset {C}, remove {C1}, and repeat steps a and b to obtain the line fitting parameters {k}. m2 ,b m2} and the point cloud subset {C2}; complete the fitting of the second straight line;

[0013] d. In the original point cloud subset {C}, remove {C1} and {C2}, and repeat steps a and b to obtain the line fitting parameters {k}. m3 ,b m3} and the point cloud subset {C3}; complete the fitting of the third straight line;

[0014] Step 3: From the point cloud subset {C f Select the portion belonging to the head of the calibration fixture from {C1, C2, C3}, excluding {C f The point cloud of line segment S3 that belongs to the S3 line segment is defined as the set of the three fitted lines excluding the lines that belong to the S3 line segment. The set of point clouds excluding the S3 line segments is called {C}. fh};

[0015] Step 4: Calculate the transformation matrix of the laser scanner measurement results in the laser scanner measurement coordinate system. Initial values: This is the transformation matrix between the laser scanner coordinate system and the calibration fixture coordinate system;

[0016]

[0017] Select {C fh The highest point h in} f The highest point refers to o. n The point where the y-axis coordinate of the calibration fixture coordinate system is at its maximum value;

[0018] Select o from the standard point cloud of the calibration tooling r point;

[0019] Calculate the initial value of t, t0 = h f -o r ;

[0020] Calculate the initial value of R.

[0021]

[0022] Step 5: Using R0 and t0 as initial matching values, input them into the ICP algorithm to obtain the final accurate...

[0023] Step 6: Calculate the conversion matrix between the inertial navigation system and the laser scanner.

[0024]

[0025]

[0026] α is the roll angle output by the inertial navigation system. r ob o b In o r Coordinates in a coordinate system.

[0027] The beneficial effects of this invention are that it can effectively identify the characteristics of calibration fixtures under noise interference, and achieve high-precision identification of system installation errors. Attached Figure Description

[0028] Figure 1 This is a schematic diagram illustrating the process of an inertial laser scanner system error identification and calibration method according to the present invention.

[0029] Figure 2 In the calibration tool head, s1 is called the top line, s2 is called the side line, and s3 is called the connecting line.

[0030] Figure 3 This is a schematic diagram illustrating the effect of using the present invention for system error identification; the upper image is the laser measurement image before calibration, and the lower image is the laser measurement image after calibration. Detailed Implementation

[0031] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings.

[0032] like Figure 1 As shown, the present invention provides a method for error identification and calibration of an inertial laser scanner system, the specific steps of which are as follows:

[0033] Step 1: Determine the system coordinate system:

[0034] like Figure 1 As shown, calibration fixture 3 has a concave structure. The concave portion at the center of calibration fixture 3 constitutes the characteristic structural line S3 of the fixture. L is the length of the characteristic structural line S3, which is called the characteristic dimension. The characteristic dimension needs to be measured in advance, and the accuracy of this dimension determines the accuracy of system installation error identification.

[0035] The system coordinate system is defined as being located within the cross-section of calibration fixture 3.

[0036] o r —The origin of the coordinate system on the left side of the calibration fixture is the intersection of the top surface and the left side of calibration fixture 3; the x-axis is the horizontal direction.

[0037] o rr —The origin of the coordinate system on the right side of the calibration fixture is the intersection of the top surface and the right side surface of calibration fixture 3; the x-axis is the horizontal direction.

[0038] o b —The origin of the inertial conductor coordinate system is located at the geometric center of inertial navigation system 1; the x-axis is parallel to the inertial conductor reference plane.

[0039] o n —The origin of the laser scanner measurement coordinate system refers to the origin specified by the output data when laser scanner 2 performs measurements; n = 1, 3; representing the right and left laser scanners respectively.

[0040] —The origin of the laser scanner reference plane coordinate system refers to the origin defined by the measuring mechanical reference plane of the laser scanner, which can be obtained by consulting the laser scanner manual;

[0041] —The position of the origin of the reference plane of laser scanner n in the inertial conductor coordinate system;

[0042] Step 2: Locate the head of the calibration fixture in the laser measurement point cloud: Calibrate within the cross-section of calibration fixture 3.

[0043] like Figure 2 As shown, the calibration tool head is defined by a top straight line segment S1 and a side straight line segment S2. However, there are actually three straight lines, namely the feature structure line S3. In some cases, S3 may be misidentified as S1, therefore the conventional RANSAC method is improved.

[0044] The specific calculation process is as follows: Within the cross-section of the calibration fixture 3, two laser scanners scan the calibration fixture from the left and right sides respectively, and obtain their respective point cloud datasets {C}; steps a to d are performed on the two point cloud datasets {C} respectively.

[0045] a. Randomly select two points from the point cloud dataset {C} acquired by the laser scanner; use these two points to determine the straight line model in the laser scanner coordinate system;

[0046] Set the interior point verification error ε r , ε r =1~2mm, the inner point is the distance from the line to ε. r point;

[0047] Substitute all the point cloud data in {C} into the straight line model and calculate the number of interior points;

[0048] Refit the line using interior points and record the fitting parameters {k}. i ,b i};k i ,b i Let represent the slope and intercept of the linear model, respectively; i is the fitting order, i = 1, 2, ...

[0049] b. Repeat step a until k.i ,b i Less than the verification error {ε k ,ε b}, and select the set of parameters {k} with the largest number of interior points. m ,b m};

[0050] Verification error {ε k ,ε b} is one ten-thousandth, where ε k ,ε b The verification errors for the slope and intercept are respectively represented by the parameter {k}, which is less than the error threshold. i ,b i The subset of internal points of {k} m ,b m Merge the sets of all internal points to form a new subset of internal points {C1}, where k i ∈[k m -ε k ,k m +ε k ], b i ∈[b m -ε b ,b m +ε b The new subset of interior points after merging is used to refit the line, obtaining the line fitting parameters {k}. m1 ,b m1}; Complete the fitting of a straight line;

[0051] c. In the original point cloud subset {C}, remove {C1}, and repeat steps a and b to obtain the line fitting parameters {k}. m2 ,b m2} and the point cloud subset {C2}; complete the fitting of the second straight line;

[0052] d. In the original point cloud subset {C}, remove {C1} and {C2}, and repeat steps a and b to obtain the line fitting parameters {k}. m3 ,b m3} and the point cloud subset {C3}; complete the fitting of the third straight line;

[0053] Step 3: From the point cloud subset {C f From {C1, C2, C3}, the portion belonging to the head of the calibration fixture is selected, that is, based on the prior information of the laser scanner pose in the structure, {C1, C2, C3} is excluded. f The point cloud of the S3 line segment in the} is the line that excludes the S3 line segment from the three fitted lines:

[0054] If the laser scanner is located on the right, then the leftmost line segment is S3;

[0055] If the laser scanner is located on the left, then the rightmost line segment is S3.

[0056] The set of point clouds excluding line segment S3 is called {C} fh};

[0057] Step 4: Calculate the transformation matrix of the laser scanner measurement results in the laser scanner measurement coordinate system. Initial values: This is the transformation matrix between the laser scanner coordinate system and the calibration fixture coordinate system;

[0058]

[0059] Select {C fh The highest point h in} f The highest point refers to o. n The point where the y-axis coordinate of the calibration fixture coordinate system is at its maximum value;

[0060] Select o from the standard point cloud of the calibration tooling r point;

[0061] Calculate the initial value of t, t0 = h f -o r ;

[0062] Calculate the initial value of R.

[0063]

[0064] Step 5: Using R0 and t0 as initial matching values, input them into the ICP algorithm to calculate the exact matching result and obtain the final accurate match.

[0065] Step 6: Calculate the conversion matrix between the inertial navigation system and the laser scanner.

[0066]

[0067] in Let be the transformation matrix between the inertial conductor coordinate system and the laser scanner measurement coordinate system, where

[0068]

[0069] α is the roll angle output by the inertial navigation system. r ob o b In o r Coordinates in a coordinate system.

[0070] like Figure 3As shown, the present invention was used for calibration, and the calibration tool feature width was 1399mm. The calibration result was 1399mm, and the calibration result was correct.

Claims

1. A method for error identification and calibration of an inertial laser scanner system, characterized by the following steps: as follows: Step 1: Determine the system coordinate system: The system coordinate system is located within the cross-section of the calibration fixture, which has a concave structure. The head of the calibration fixture consists of the top straight line segment S1 and the side straight line segment S2, and the central concave part consists of the characteristic structural straight line S3 of the fixture. The coordinate system is defined as follows: The origin of the coordinate system on the left side of the calibration fixture is the intersection of the top surface and the left side of the calibration fixture. The origin of the coordinate system on the right side of the calibration fixture is the intersection of the bottom surface and the right side surface of the calibration fixture; Origin of the inertial conductor coordinate system The laser scanner measures the origin of the coordinate system, n=1,3; These represent the right and left laser scanners, respectively. Origin of the laser scanner reference plane coordinate system , —Laser scanner The position of the origin of the reference plane in the inertial conductor coordinate system; The second step involves using two laser scanners to scan the calibration fixture from the left and right sides, respectively, within its cross-section, to obtain their respective point cloud datasets. ; for the two point cloud datasets respectively Proceed through steps a to d; a. Point cloud dataset acquired from a laser scanner Two points are randomly selected; these two points are used to determine the linear model in the laser scanner coordinate system; the interior point verification error is set. , =1~2mm, the inner point is less than the distance from the line. The point; will All point cloud data are input into the linear model to calculate the number of inliers; the inliers are then used to refit the line, and the fitting parameters are recorded. ; These represent the slope and intercept of the linear model, respectively; i is the fitting order, i=1, 2, ...; b. Repeat step a until... Less than the verification error And select the set of parameters with the largest number of interior points. Verification error It is one in ten thousand, of which, The verification errors for slope and intercept are respectively represented by parameters less than the error threshold. Internal subsets and Merge the sets of internal points to form a new subset of internal points. ,in , The new subset of interior points after merging is used to refit the line and obtain the fitting parameters. To fit a straight line; c. In the original point cloud subset Remove Repeat steps a and b to obtain the line fitting parameters. He Dianyun Subset Complete the fitting of the second straight line; d. In the original point cloud subset Remove and Repeat steps a and b to obtain the line fitting parameters. He Dianyun Subset Complete the fitting of the third straight line; Step 3: From point cloud subsets Select the part belonging to the head of the calibration tool and exclude it. Belonging to The point cloud of line segments, i.e., excluding the three fitted lines that belong to... The straight line of the line segment; excluded The set of point clouds of line segments is called ; Step 4: Calculate the transformation matrix of the laser scanner measurement results in the laser scanner measurement coordinate system. Initial values: This is the transformation matrix between the laser scanner coordinate system and the calibration fixture coordinate system; （1） Select The highest point in The highest point refers to The point where the y-axis coordinate of the calibration fixture coordinate system is at its maximum value; Select the standard point cloud of the calibration tooling point; calculate initial value ; calculate initial value, （2） Step 5, with , To match the initial value, it is substituted into the ICP algorithm to obtain the final accurate value. ; Step 6: Calculate the conversion matrix between the inertial navigation system and the laser scanner. : （3） （4） The roll angle output by the inertial navigation system. , for exist Coordinates in a coordinate system.

Images