阅读说明：本技术 一种无人机机载测绘吊舱的摄像头安装误差测定方法 (Method for measuring installation error of camera of airborne surveying and mapping pod of unmanned aerial vehicle ) 是由 裴海龙 林裕峻 于 2020-12-08 设计创作，主要内容包括：本发明公开了一种无人机机载测绘吊舱的摄像头安装误差测定方法,步骤如下：在测绘地面放置标定板,并通过GPS定位模块测量在参考坐标系下的坐标；无人机携带测绘吊舱从多个航向飞过标定板,分别通过IMU得到测绘姿态角和通过GPS定位模块得到测绘位置坐标；根据测绘姿态角和测绘位置坐标计算每张照片对应的参考坐标系到像素坐标系之间的坐标变换关系；通过坐标变换和图像处理得到标定板精确的像素坐标；构建参考坐标系到像素坐标系的包含安装误差参数的坐标重映射函数；利用标定板像素坐标和坐标重映射函数构建重映射误差函数；将重映射误差函数转化为非线性最小二乘优化问题,并利用高斯牛顿法进行求解,得到安装误差参数。(The invention discloses a method for measuring the installation error of a camera of an airborne surveying and mapping pod of an unmanned aerial vehicle, which comprises the following steps: placing a calibration plate on the surveying and mapping ground, and measuring coordinates under a reference coordinate system through a GPS positioning module; the unmanned aerial vehicle carries a surveying and mapping pod to fly over the calibration board from a plurality of courses, and a surveying and mapping attitude angle and a surveying and mapping position coordinate are respectively obtained through an IMU and a GPS positioning module; calculating a coordinate transformation relation between a reference coordinate system corresponding to each photo and a pixel coordinate system according to the surveying attitude angle and the surveying position coordinate; obtaining accurate pixel coordinates of the calibration plate through coordinate transformation and image processing; constructing a coordinate remapping function from a reference coordinate system to a pixel coordinate system, wherein the coordinate remapping function comprises an installation error parameter; constructing a remapping error function by utilizing the pixel coordinates and the coordinate remapping function of the calibration plate; and converting the remapping error function into a nonlinear least square optimization problem, and solving by using a Gauss-Newton method to obtain an installation error parameter.)

1. A method for determining installation errors of a camera of an airborne mapping pod of an unmanned aerial vehicle is characterized by comprising the following steps:

s1, placing a plurality of calibration plates on the surveying and mapping ground, and measuring the coordinate of the center of each calibration plate under a reference coordinate system through a GPS positioning module;

s2, the unmanned aerial vehicle carries a surveying and mapping pod to fly over a calibration plate from a plurality of courses and take a snapshot, and a surveying and mapping attitude angle and a surveying and mapping position coordinate are respectively obtained through an IMU inertial measurement unit and a GPS positioning module;

s3, calculating a coordinate transformation relation between a reference coordinate system and a pixel coordinate system corresponding to each photo according to the surveying attitude angle and the surveying position coordinate;

s4, performing coordinate transformation and image processing through the coordinate transformation relation to obtain the pixel coordinates of the calibration board in a pixel coordinate system;

s5, constructing a coordinate remapping function from a reference coordinate system to a pixel coordinate system, wherein the coordinate remapping function comprises installation error parameters;

s6, constructing a remapping error function by using the pixel coordinate and the coordinate remapping function of the calibration plate in a pixel coordinate system;

and S7, converting the remapping error function into a nonlinear least square optimization problem, and solving by using a Gauss-Newton method to obtain an installation error parameter.

2. The method as claimed in claim 1, wherein 6 2 x 2 black and white calibration plates are placed on the surveying ground in step S1.

3. The method for determining the camera installation error of the unmanned aerial vehicle airborne survey pod as claimed in claim 1, wherein the step S3 is as follows:

mapping attitude angle corresponding to each photoTheta, psi and mapping position coordinates t, set to a point on the ground expressed as P in the reference coordinate system^W＝[x,y,z]^TDenoted P in the coordinate system of the mapping pod^BDenoted P in the camera coordinate system^CExpressed as P in the pixel coordinate system^M＝[u,v]^TThe coordinate transformation between the reference coordinate system and the mapping pod coordinate system is then:

P^W＝R_BWP^B+t_BW

wherein the content of the first and second substances,representing a rotation matrix, R, mapping the nacelle coordinate system to a reference coordinate system_Z(psi) a rotation matrix indicating a rotation angle psi about the Z-axis, R_Y(theta) represents a rotation matrix of rotation angles theta around the Y-axis,indicating angle of rotation about X-axisRotation matrix of t_BWT, representing coordinates of the origin of the coordinate system of the surveying and mapping pod under the reference coordinate system;

coordinate transformation between the mapping pod coordinate system and the camera coordinate system:

P^B＝R_CBP^C+t_CB

wherein R is_CBRotation matrix, t, representing camera coordinate system to mapping pod coordinate system_CBIs the coordinate of the origin of the camera coordinate system under the coordinate system of the surveying and mapping pod, R_CBAnd t_CBIs a constant matrix and a constant vector determined according to the mechanical installation;

coordinate transformation between camera coordinate system and pixel coordinate system:

s＝(KP^C)₍₃₎

wherein K represents the internal reference of the camera and s represents the pixel point P^MCorresponding depth, expression (.)_(1:2)First two-dimensional data representing column vectors, expression (·)₍₃₎Third dimensional data representing a column vector;

coordinate transformation from reference coordinate system to pixel coordinate system:

4. the method as claimed in claim 3, wherein the coordinate transformation and image processing in step S4 are as follows:

for each picture, the coordinates of the plate in the reference coordinate system are calibratedCoordinate transformation is carried out through the coordinate transformation relation from the reference coordinate system to the pixel coordinate system to obtain the calibration plateThe initial pixel coordinates are:

calculating the range of the calibration plate under the pixel coordinate system as

Wherein du is_FAnd dv_FDetermining according to the size of a calibration plate and the surveying and mapping height, graying the image in the range, carrying out window detection on the calibration plate, assuming that pixels with the gray scale larger than a threshold value are characteristic pixels, counting the characteristic pixels in a detection window, and when the number of the characteristic pixels of a left half window is equal to that of the characteristic pixels of a right half window and the number of the characteristic pixels of an upper half window is equal to that of the characteristic pixels of a lower half window, locating the center of the detection window, namely the accurate pixel center of the calibration plate.

5. The method as claimed in claim 4, wherein the step S5 of constructing the coordinate remapping function containing the installation error parameters comprises the following steps:

setting an error rotation matrix caused by the installation error to be delta R and setting an error position offset to be delta t due to the existence of the installation error, and then carrying out coordinate transformation between a mapping pod coordinate system and a camera coordinate system after the installation error is corrected

Wherein Δ R ═ R_Z(γ)R_Y(β)R_X(α), α, β, γ represent mounting error angles due to mounting errors, R_Z(gamma) represents a rotation matrix of rotation angle gamma about the Z axis, R_Y(beta) represents the rotation angle beta around the Y axisOf a rotation matrix R_X(α) represents a rotation matrix of the rotation angle α around the X axis, Δ t ═ Δ X Δ y Δ z]^TΔ x, Δ y, and Δ z represent mounting error offsets caused by mounting errors;

a coordinate remapping equation from a corresponding reference coordinate system to a pixel coordinate system including a mounting error parameter is

6. The method as claimed in claim 5, wherein the step S6 of constructing the remapping error function comprises the following steps:

will P^WTo P^MIs defined as a mapping function f

P^M＝f(P^W,ΔR,Δt)

For n photos taken by the snapshot with multiple headings, the reprojection error of the jth characteristic point in the ith photo is

WhereinDenotes the coordinates of the center of the jth calibration plate in the reference coordinate system, m_ijThe pixel coordinates of the jth block calibration board detected in the ith picture are represented, i is 1,2, …, and n, j is 1,2, …, 6.

7. The method for determining the camera installation error of the unmanned aerial vehicle airborne survey pod as claimed in claim 6, wherein the step S7 is as follows:

let variable xi ═ α β γ Δ x Δ y Δ z]^TConstructing nonlinearity by using the two-norm principle of minimized errorMinimizing the reprojection error problem

And solving by a Gauss-Newton method to obtain the installation error parameters alpha, beta, gamma, delta x, delta y and delta z.

Technical Field

The invention relates to the technical field of error measurement, in particular to a method for measuring installation errors of a camera of an airborne surveying and mapping pod of an unmanned aerial vehicle.

Background

Along with the popularization and the popularization of the informatization mapping concept, the importance of geographic spatial information is increasingly enhanced, higher requirements are put forward on the timeliness and the richness of geographic information acquisition, and unmanned aerial vehicle low-altitude mapping is a flexible, intelligent, multifunctional and efficient mapping mode and is widely applied to various scenes, such as building mapping, mine geological mapping, traffic field mapping and the like. The common unmanned aerial vehicle airborne surveying and mapping scheme for low-altitude surveying and mapping of the unmanned aerial vehicle mainly comprises a self-stabilizing cradle head surveying and mapping scheme and an airborne surveying and mapping pod scheme, wherein the self-stabilizing cradle head surveying and mapping scheme mainly depends on a self-stabilizing cradle head to realize stable remote sensing shooting, but the scheme is limited by the load capacity of the cradle head and can only carry small-sized surveying and mapping equipment; the scheme of the surveying and mapping pod overcomes the load limitation, more sensors such as a camera, a laser radar, an IMU (inertial measurement Unit), a GPS (global positioning system) and the like can be carried, and the effect of stable surveying and mapping is achieved through the fusion of various sensors. To unmanned aerial vehicle machine carried survey and drawing performance, survey and drawing precision is an important investigation index, and the error source that influences the survey and drawing precision of unmanned aerial vehicle machine carried survey nacelle mainly comes from sensor installation error, consequently waits to provide an error determination method of camera installation error angle and installation error skew at present urgently, and then through error correction, improves the survey and drawing precision of unmanned aerial vehicle machine carried survey nacelle.

Disclosure of Invention

The invention aims to solve the defects in the prior art, and provides a method for measuring the installation error of a camera of an unmanned aerial vehicle airborne surveying and mapping pod.

The purpose of the invention can be achieved by adopting the following technical scheme:

a method for determining installation errors of a camera of an airborne mapping pod of an unmanned aerial vehicle comprises the following steps:

s1, placing a plurality of calibration plates on the surveying and mapping ground, and measuring the coordinate of the center of each calibration plate under a reference coordinate system through a GPS positioning module;

s2, the unmanned aerial vehicle carries a surveying and mapping pod to fly over a calibration plate from a plurality of courses and take a snapshot, and a surveying and mapping attitude angle and a surveying and mapping position coordinate are respectively obtained through an IMU inertial measurement unit and a GPS positioning module;

s3, calculating a coordinate transformation relation between a reference coordinate system and a pixel coordinate system corresponding to each photo according to the surveying attitude angle and the surveying position coordinate;

s4, performing coordinate transformation and image processing through the coordinate transformation relation to obtain the pixel coordinates of the calibration board in a pixel coordinate system;

s5, constructing a coordinate remapping function from a reference coordinate system to a pixel coordinate system, wherein the coordinate remapping function comprises installation error parameters;

s6, constructing a remapping error function by using the pixel coordinate and the coordinate remapping function of the calibration plate in a pixel coordinate system;

and S7, converting the remapping error function into a nonlinear least square optimization problem, and solving by using a Gauss-Newton method to obtain an installation error parameter.

Further, in the step S1, 6 2 × 2 black and white calibration boards are placed on the surveying and mapping ground.

Further, the step S3 process is as follows:

mapping attitude angle corresponding to each photoTheta, psi and mapping position coordinates t, set to a point on the ground expressed as P in the reference coordinate system^W＝[x,y,z]^TDenoted P in the coordinate system of the mapping pod^BDenoted P in the camera coordinate system^CExpressed as P in the pixel coordinate system^M＝[u,v]^TThe coordinate transformation between the reference coordinate system and the mapping pod coordinate system is then:

P^W＝R_BWP^B+t_BW

wherein the content of the first and second substances,R_Z(psi) a rotation matrix indicating a rotation angle psi about the Z-axis, R_Y(theta) represents a rotation matrix of rotation angles theta around the Y-axis,indicating angle of rotation about X-axisRepresenting a rotation matrix mapping the nacelle coordinate system to a reference coordinate system, t_BWT, representing coordinates of the origin of the coordinate system of the surveying and mapping pod under the reference coordinate system;

coordinate transformation between the mapping pod coordinate system and the camera coordinate system:

P^B＝R_CBP^C+t_CB

wherein R is_CBRotation matrix, t, representing camera coordinate system to mapping pod coordinate system_CBIs the coordinate of the origin of the camera coordinate system under the coordinate system of the surveying and mapping pod, R_CBAnd t_CBIs a constant matrix and a constant vector determined according to the mechanical installation;

coordinate transformation between camera coordinate system and pixel coordinate system:

s＝(KP^C)₍₃₎

wherein K represents the internal reference of the camera and s represents the pixel point P^MCorresponding depth, expression (.)_(1∶2)First two-dimensional data representing column vectors, expression (·)₍₃₎Third dimensional data representing a column vector;

coordinate transformation from reference coordinate system to pixel coordinate system:

further, the coordinate transformation and image processing procedure in step S4 is as follows:

for each picture, the coordinates of the plate in the reference coordinate system are calibratedAnd performing coordinate transformation through a coordinate transformation relation from a reference coordinate system to a pixel coordinate system to obtain an initial pixel coordinate of the calibration plate as follows:

calculating the range of the calibration plate under the pixel coordinate system as

Wherein du is_FAnd dv_FDetermining according to the size of a calibration plate and the surveying and mapping height, graying the image in the range, carrying out window detection on the calibration plate, assuming that pixels with the gray scale larger than a threshold value are characteristic pixels, counting the characteristic pixels in a detection window, and when the number of the characteristic pixels of a left half window is equal to that of the characteristic pixels of a right half window and the number of the characteristic pixels of an upper half window is equal to that of the characteristic pixels of a lower half window, locating the center of the detection window, namely the accurate pixel center of the calibration plate.

Further, the process of constructing the coordinate remapping function containing the installation error parameters in step S5 is as follows:

setting an error rotation matrix caused by the installation error to be delta R and setting an error position offset to be delta t due to the existence of the installation error, and then carrying out coordinate transformation between a mapping pod coordinate system and a camera coordinate system after the installation error is corrected

Wherein Δ R ═ R_Z(γ)R_Y(β)R_X(α), α, β, γ represent mounting error angles due to mounting errors, R_Z(gamma) represents a rotation matrix of rotation angle gamma about the Z axis, R_Y(beta) represents a rotation matrix of rotation angle beta about the Y axis, R_X(α) represents a rotation matrix of the rotation angle α around the X axis, Δ t ═ Δ X Δ y Δ z]^TΔ x, Δ y, and Δ z represent mounting error offsets caused by mounting errors;

a coordinate remapping equation from a corresponding reference coordinate system to a pixel coordinate system including a mounting error parameter is

Further, the process of constructing the remapping error function in step S6 is as follows:

will P^WTo P^MIs defined as a mapping function f

P^M＝f(P^W，ΔR，Δt)

For n photos taken by the snapshot with multiple headings, the reprojection error of the jth characteristic point in the ith photo is

WhereinDenotes the coordinates of the center of the jth calibration plate in the reference coordinate system, m_ijThe pixel coordinates of the jth calibration board detected in the ith picture are represented, i is 1, 2.

Further, the step S7 process is as follows:

let variable xi ═ α β γ Δ x Δ y Δ z]^TThe nonlinear minimization reprojection error problem is constructed by utilizing the two-norm principle of the minimization error

And solving by a Gauss-Newton method to obtain the installation error parameters alpha, beta, gamma, delta x, delta y and delta z.

Compared with the prior art, the invention has the following advantages and effects:

according to the invention, the calibration plate is shot at high altitude, nonlinear optimization is carried out based on the minimum reprojection error target, the installation error angle and the installation error offset of the camera are obtained through calculation, and the surveying and mapping precision of the unmanned aerial vehicle airborne surveying and mapping pod can be improved through error correction.

Drawings

FIG. 1 is a flowchart of a method for determining a camera mounting error of an airborne surveying and mapping pod of an unmanned aerial vehicle according to an embodiment of the present invention;

fig. 2 is a schematic view of a calibration plate used in an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Examples

As shown in fig. 1, the method for determining the installation error of the camera of the unmanned aerial vehicle airborne surveying and mapping pod disclosed in the embodiment mainly comprises the following steps: placing a calibration plate and measuring coordinates, taking an aerial photograph of the calibration plate and recording measurement data of an IMU inertial measurement unit and a GPS positioning module, calculating a coordinate transformation relation of each photo, detecting a pixel coordinate of the calibration plate of each photo, constructing a coordinate remapping function containing installation error parameters, constructing a coordinate remapping error function, converting the coordinate remapping error function into a nonlinear least square optimization problem and solving the problem. The following detailed description is made with reference to fig. 1:

and S1, placing calibration plates and measuring coordinates, placing 6 black and white calibration plates shown in the specification if 2 on the surveying and mapping ground, and measuring the coordinates of the center of each calibration plate in a reference coordinate system through a GPS positioning module.

And S2, aerial photographing the calibration board, recording the measurement data of the IMU inertial measurement unit and the GPS positioning module, controlling the unmanned aerial vehicle to carry the surveying and mapping pod to fly above the calibration board from multiple courses and take photos, and respectively obtaining a surveying and mapping attitude angle through the current IMU inertial measurement unit and a surveying and mapping position coordinate through the GPS positioning module.

Step S3, calculating the coordinate transformation relation of each photo, obtaining the coordinate transformation relation between the corresponding reference coordinate system and the pixel coordinate system by calculating the corresponding mapping attitude angle and mapping position coordinate for each photo, and obtaining the mapping attitude angle corresponding to each photoTheta, psi and mapping position coordinates t, set to a point on the ground expressed as P in the reference coordinate system^W＝[x，y，z]^TDenoted P in the coordinate system of the mapping pod^BDenoted P in the camera coordinate system^CExpressed as P in the pixel coordinate system^M＝[u，v]^TCoordinate transformation between the reference coordinate system and the coordinate system of the mapping pod

P^W＝R_BWP^B+t_BW (1)

Wherein the content of the first and second substances,R_Z(psi) a rotation matrix indicating a rotation angle psi about the Z-axis, R_Y(theta) represents a rotation matrix of rotation angles theta around the Y-axis,indicating angle of rotation about X-axisRotational moment ofAn array representing a rotation matrix mapping the nacelle coordinate system to a reference coordinate system, t_BWT, showing the coordinate of the origin of the coordinate system of the surveying and mapping pod under the reference coordinate system; coordinate transformation between mapping pod coordinate system and camera coordinate system

P^B＝R_CBP^C+t_CB (2)

Wherein R is_CBRotation matrix, t, representing camera coordinate system to mapping pod coordinate system_CBIs the coordinate of the origin of the camera coordinate system under the coordinate system of the surveying and mapping pod, R_CBAnd t_CBIs a constant matrix and a constant vector determined according to the mechanical installation; coordinate transformation between camera coordinate system and pixel coordinate system

s＝(KP^C)₍₃₎ (6)

Wherein K represents the internal reference of the camera, and s represents the pixel point P^MCorresponding depth, expression (.)_(1∶2)First two-dimensional data representing column vectors, expression (·)₍₃₎Third dimensional data representing a column vector; the coordinate transformation from the reference coordinate system to the pixel coordinate system can be obtained by combining equations (1) to (6)

s＝(KR_CB ^-1(R_BW ^-1(P^W-t_BW)-t_CB))₍₃₎ (8)

Step S4, detecting the pixel coordinates of the calibration board of each photo, and for each photo, assuming the coordinates of the calibration board in the reference coordinate system asSubstituting equations (7) - (8) can obtain the initial pixel coordinate of the calibration plate as

Calculating the range of the calibration plate under the pixel coordinate system as

Wherein du is_FAnd dv_FDetermining according to the size of the calibration plate and the surveying and mapping height, graying the image in the range, detecting the calibration plate, and detecting the adopted window; and assuming that pixels with the gray scale larger than the threshold value are characteristic pixels, counting the characteristic pixels in the detection window, and when the number of the characteristic pixels of the left half window is equal to that of the characteristic pixels of the right half window and the number of the characteristic pixels of the upper half window is equal to that of the characteristic pixels of the lower half window, determining the center of the detection window, namely the accurate pixel center of the position calibration plate.

Step S5, constructing a coordinate remapping function containing installation error parameters, setting an error rotation matrix caused by installation errors as delta R and setting error position offset as delta t due to the existence of the installation errors, and transforming coordinates between a coordinate system of the surveying and mapping pod and a coordinate system of the camera after the installation errors are corrected

Wherein Δ R ═ R_Z(γ)R_Y(β)R_X(α), α, β, γ denote mounting error angles due to mounting errors, and Δ t ═ Δ x Δ y Δ z]^TΔ x, Δ y andΔ z represents a mounting error deviation due to a mounting error; combining the equations (1), (3) - (6), (9) can obtain the coordinate remapping equation containing the installation error parameter from the reference coordinate system to the pixel coordinate system as

s＝(KR_CB ^-1ΔR^-1(R_BW ^-1(P^W-t_BW)-t_CB-Δt))₍₃₎ (11)

Step S6, a coordinate remapping error function is constructed, and the remapping function from the reference coordinate system to the pixel coordinate system is redefined as a mapping function f, namely

P^M＝f(P^W，ΔR，Δt)

Then, for n photographs taken in a plurality of courses, the reprojection error of the j (j) 1,2, 6) feature point in the i (i) 1,2

Wherein the content of the first and second substances,denotes the coordinates of the center of the jth calibration plate in the reference coordinate system, m_ijAnd the pixel coordinates of the jth block calibration board detected in the ith picture are shown.

Step S7, converting the angle alpha, beta, gamma and deviation delta x, delta y, delta z into nonlinear least square optimization problem and solving, constructing variable xi with the angle alpha, beta, gamma and deviation delta x, delta y, delta z,

i.e. ξ ═ α β γ Δ x Δ y Δ z]^TThe nonlinear minimization reprojection error problem is constructed by utilizing the two-norm principle of the minimization error

And solving by a Gauss-Newton method to obtain the installation errors alpha, beta, gamma, delta x, delta y and delta z of the target measurement, and realizing the method of the invention.

The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.