CN109242912A - Join scaling method, electronic equipment, storage medium outside acquisition device - Google Patents

Abstract

The invention provides a method for calibrating external parameters of a collection device, which comprises the steps of: acquiring identification information collected by the collection device; calculating a first coordinate conversion relationship between the identification coordinate system and the collection device coordinate system; obtaining identification information under the identification coordinate system according to the first coordinate conversion relationship The corresponding coordinate points of the marked points in the acquisition device coordinate system, the fitting algorithm is used to fit the coordinate points, the fitting coordinate system is established, and the second coordinate conversion relationship between the fitting coordinate system and the acquisition device coordinate system is calculated; The fitting algorithm corresponding to the preset trajectory fits the coordinate points, generates the fitting trajectory, and calculates the third coordinate transformation relationship between the fitted coordinate system and the movable platform coordinate system; through the second coordinate transformation relationship and the third coordinate transformation relationship , calculate the conversion relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform. The invention can reduce the installation error caused by the uneven installation of the collecting device, and by optimizing the calibration result, the calibration result is more accurate and the calibration error is small.

Description

External parameter calibration method for acquisition device, electronic device, and storage medium

technical field

The invention relates to the technical field of external parameter calibration of a collection device, in particular to a method, electronic equipment and storage medium for the calibration of external parameters of a collection device.

Background technique

Usually, the acquisition device needs to be calibrated in advance in order to obtain accurate acquisition device parameters. The parameters of the acquisition device include internal parameters and external parameters. The internal parameters of the acquisition device are inherent parameters of the acquisition device, such as focal length, optical center position, etc. The external parameters of the acquisition device represent the acquisition device in the world coordinate system (also called the global coordinate system). Position and orientation, from a mathematical point of view, the external parameters of the acquisition device represent the coordinate transformation relationship between the coordinate system of the acquisition device and the specified world coordinate system (also called the global coordinate system). The pose estimation of the acquisition device is to estimate the external parameters of the acquisition device during the relative motion of the acquisition device and the object to be photographed, and the internal parameters of the acquisition device used in the estimation algorithm are usually determined offline by the acquisition device calibration method. The acquisition device-wheel odometer calibration method has high precision requirements for the wheel odometer, and the error is large.

SUMMARY OF THE INVENTION

In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method for calibrating the external parameters of the acquisition device, directly calibrating the motion center, realizing the complete calibration of the objective degrees of freedom, optimizing the calibration result, with high accuracy and small error.

The present invention provides a method for calibrating external parameters of a collection device, comprising the following steps:

Acquiring identification information, acquiring information collected by the collection device at several collection moments, the information being the identification information during the relative movement of the movable platform along the preset track and the identification;

Establish a first coordinate conversion relationship, calculate the conversion relationship between the coordinate system of each collection moment and the coordinate system of the collection device, and record it as the first coordinate conversion relationship;

Establish a second coordinate conversion relationship, obtain the marking points in the marking information on the marking coordinate system, and obtain the corresponding marking points in the coordinate system of the collecting device at each acquisition moment of the marking points according to the first coordinate conversion relationship Several coordinate points of , use a fitting algorithm to fit the coordinate points, establish a fitting coordinate system, calculate the conversion relationship between the fitting coordinate system and the coordinate system of the acquisition device, and record it as the second coordinate conversion relationship;

Establish a third coordinate conversion relationship, fit the coordinate points according to the fitting algorithm corresponding to the preset trajectory, generate a fitted trajectory, calculate the conversion relationship between the fitted coordinate system and the movable platform coordinate system, record is the third coordinate conversion relationship;

In the external parameter calibration of the acquisition device, the conversion relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform is calculated through the second coordinate conversion relationship and the third coordinate conversion relationship.

Further, in the establishment of the first coordinate conversion relationship, the identification information is processed, an image coordinate system is established, the coordinates of the identification on the image coordinate system and the identification coordinate system are obtained, and the collection device is combined. The internal parameters and distortion coefficients are used to calculate the transformation relationship between the identification coordinate system and the acquisition device coordinate system through a pose estimation method.

Further, the preset trajectory includes a rotation trajectory and a straight trajectory generated by the movable platform using the identifier as a reference frame.

Further, the obtaining identification information includes the following steps:

Acquiring a rotating image, acquiring the identification information collected by the collecting device during the in-situ rotation of the movable platform according to the rotating trajectory, and obtaining several rotating images;

Acquiring straight images, acquiring identification information collected by the collecting device during the process of moving the movable platform in a straight line along the coordinate axis of the coordinate system of the movable platform according to the straight trajectory, and obtaining several straight images.

Further, the establishment of the first coordinate conversion relationship includes the following steps:

Generate a rotation pose, process the rotated image, obtain the image coordinates of the marker point, identify the coordinate system of the marker board coordinate system according to the image coordinates of the marker point and the marker point corresponding to the coordinate system, and combine Collecting the internal parameters and distortion coefficients of the device, calculating the pose of the marker plate coordinate system and the marker coordinate system relative to the coordinate system of the collecting device, and obtaining the rotational pose;

Generate a straight pose, process the straight image, obtain the image coordinates of the marker, identify the coordinates of the coordinate system according to the image coordinates of the marker and the marker plate coordinate system corresponding to the marker, and combine The internal parameters and distortion coefficients of the acquisition device are acquired, and the pose of the marker coordinate system of the marker coordinate system relative to the coordinate system of the acquisition device is calculated to obtain the straight-ahead pose.

Further, in establishing the second coordinate conversion relationship, a plane fitting algorithm is used to fit the coordinate points, a fitting plane is generated, and a fitting coordinate system is established in the fitting plane.

Further, the generating of the second coordinate conversion relationship includes the following steps:

Calculate the distance between the coordinate point and the fitting plane, when the distance is greater than a threshold, remove the coordinate point, and re-fit the remaining coordinate points;

A fitting coordinate system is established, a first point and a second point are taken in the fitting plane, and the coordinates of the first point and the second point in the coordinate system of the acquisition device are the first coordinate and the second point respectively Coordinate, calculate the first unit vector, the second unit vector, and the third unit vector according to the first coordinate and the second coordinate, and the first unit vector is the unit vector of the difference vector between the first coordinate point and the second coordinate point , the third unit vector is the normal vector of the fitting plane, the second unit vector is the vector product of the first unit vector and the third unit vector, and the first point is the coordinate origin , and establish a fitting coordinate system with the directions of the first unit vector, the second unit vector, and the third unit vector as the directions of the x-axis, the y-axis, and the z-axis;

A coordinate conversion relationship is generated, and a conversion relationship between the fitting coordinate system and the acquisition device coordinate system is generated according to the first unit vector, the second unit vector, the third unit vector, and the first coordinates.

Further, in the establishment of the third coordinate conversion relationship, ellipse fitting is performed on the coordinate points obtained along the rotation trajectory under the fitting coordinate system, and the center of the ellipse in the fitting coordinate system is recorded as the rotation of the movable platform. center to obtain the translation matrix between the fitted plane coordinate system and the movable platform coordinate system; perform straight line fitting on the coordinate points obtained along the straight trajectory under the fitted coordinate system to obtain the fitted plane coordinate system and the movable platform The rotation matrix between the coordinate systems obtains the conversion relationship between the fitting coordinate system and the movable platform coordinate system.

Further, calculating the coordinates of the rotation center of the movable platform under the coordinate system of the collection device through the second coordinate conversion relationship and the third coordinate conversion relationship, and obtaining a translation matrix between the coordinate system of the collection device and the movable platform coordinate system;

Calculate the rotation matrix between the fitting plane coordinate system and the movable platform coordinate system through the third transformation relationship, and calculate the rotation matrix between the acquisition device coordinate system and the movable platform coordinate system through the second transformation relationship;

The transformation relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform is calculated from the translation matrix and the rotation matrix.

Further, it also includes optimizing the calibration result, constructing a cost function through the conversion relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform, and the first coordinate conversion relationship, and inputting the cost function into an optimization library for execution. Calculate to generate optimized calibration results.

Further, the optimization calibration result also includes establishing odometer constraints, inputting the optimization library for calculation, and generating the optimization calibration result.

Further, it also includes establishing one or more optimal calibration results among reprojection constraints, circle constraints, and line constraints.

An electronic device, comprising: a processor;

a memory; and a program, wherein the program is stored in the memory and is configured to be executed by a processor, the program comprising for executing the above-mentioned method for calibrating an external parameter of an acquisition device.

A computer-readable storage medium on which a computer program is stored, the computer program being executed by a processor to perform the above-mentioned method for calibrating an external parameter of an acquisition device.

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

The invention provides a method for calibrating external parameters of a collection device, and the method for calibrating external parameters of a collection device, comprising the following steps: acquiring identification information, and acquiring information collected by a collection device at several collection moments, the information being the relative movement process of the movable platform and the identification along a preset track Establish the first coordinate conversion relationship, calculate the conversion relationship between the logo coordinate system and the acquisition device coordinate system at each collection moment, and record it as the first coordinate conversion relationship; establish the second coordinate conversion relationship, and obtain it from the logo coordinate system For the marked points in the identification information, according to the first coordinate conversion relationship, several coordinate points corresponding to the marked points in the coordinate system of the acquisition device are obtained at each acquisition moment, and a fitting algorithm is used to fit the coordinate points, establish a fitting coordinate system, and calculate the fitting. The conversion relationship between the coordinate system and the coordinate system of the acquisition device is recorded as the second coordinate conversion relationship; the third coordinate conversion relationship is established, the coordinate points are fitted according to the fitting algorithm corresponding to the preset trajectory, the fitting trajectory is generated, and the fitting is calculated. The transformation relationship between the coordinate system and the coordinate system of the movable platform is recorded as the third coordinate transformation relationship; for the external parameter calibration of the acquisition device, the coordinate system of the acquisition device and the movable platform are calculated through the second coordinate transformation relationship and the third coordinate transformation relationship. Transformation relationship between coordinate systems. The invention relates to an electronic device and a readable storage medium, which are used to perform a method for calibrating external parameters of a collection device; the invention performs external parameter calibration of the collection device by adding a fitting plane, so as to reduce mechanical installation errors caused by uneven installation of the collection device , by optimizing the calibration results to make the calibration results more accurate, independent of the high-precision odometer, and the calibration error is small.

The above description is only an overview of the technical solution of the present invention. In order to understand the technical means of the present invention more clearly, and implement it according to the content of the description, the preferred embodiments of the present invention are described in detail below with the accompanying drawings. Specific embodiments of the present invention are given in detail by the following examples and the accompanying drawings.

Description of drawings

The accompanying drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the attached image:

Fig. 1 is the flow chart of the external parameter calibration method of the acquisition device of the present invention;

Fig. 2 is the schematic diagram of aprilTag code in the embodiment of the present invention;

3 is a schematic diagram of a marker board in an embodiment of the present invention;

4 is a schematic diagram of the movement of the acquisition device and the movable platform in the embodiment of the present invention;

5 is a schematic diagram 1 of a fitting plane coordinate system, a collection device coordinate system, and a movable platform coordinate system in an embodiment of the present invention;

6 is a second schematic diagram of a fitting plane coordinate system, a collection device coordinate system, and a movable platform coordinate system in an embodiment of the present invention;

FIG. 7 is a schematic diagram of the movement of the acquisition device and the movable platform in the embodiment of the present invention.

In the figure: 1. Movable platform; 2. Acquisition device; 3. Marker board; 4. Coordinate system of acquisition device; 5. Fitted plane coordinate system; 6. Rotation track; 7. Rotation point; 8. Execution track; 9 , Straight point; 10. Coordinate system of movable platform.

Detailed ways

The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be noted that, on the premise of no conflict, the embodiments or technical features described below can be combined arbitrarily to form new embodiments. .

The external parameter calibration method of the acquisition device, as shown in Figure 1, includes the following steps:

Obtain the identification information, the marking points are on the same plane and regularly arranged on the marking plate, the marking plate is placed on the fixed working platform, the collecting device is installed near the center of the movable platform, facing the fixed working platform, and the collecting device is acquired at several collecting moments The collected information is the identification information during the relative movement of the movable platform along the preset track and the mark, and the identification information stores several marking points. The identification is used for positioning, and can be one or more features with specific shapes, patterns, and two-dimensional codes, which can be identified and coordinate points can be extracted therefrom. The QR code can be AprilTag or other forms. The identification can also be a combination of two-dimensional code and other patterns and colors. Specifically, the identification is obvious and has the characteristics of the identified feature. When it uses a two-dimensional code, it can be arranged in an array of a specific shape. The more QR codes, the more accurate the positioning. The collection device may be a camera, or may be other sensors with identification functions. In this embodiment, the aprilTag code is used as a marker. As shown in Figure 2, the aprilTag code is black, and the background is usually white. Each corner of the aprilTag code is used as a marker point, and each aprilTag code has a fixed ID number and Orientation, as shown in Figure 3, several aprilTag codes are on the same plane and regularly arranged on the marking plate, as shown in Figure 4, in this embodiment, during the movement of the movable platform 1, the collecting device 2 can at least observe the markings The 4 markers on the board 3, the physical coordinate relationship between the markers is known, can be any distinguishing background, can be identified in the image and uniquely determined points, that is, each marker has a certain number in an image. the unique coordinates of .

In an embodiment, preferably, the preset trajectory includes a rotation trajectory and a straight trajectory generated by the movable platform with the identification as the reference frame. Preferably, acquiring the identification information includes the following steps:

Obtain a rotating image, obtain identification information collected by the collection device during the in-situ rotation of the movable platform according to the rotation trajectory, and obtain several rotating images; , the acquisition device can observe at least one aprilTag code, and the acquisition device acquires the image.

A straight image is acquired, the identification information collected by the collecting device is acquired during the process of the movable platform traveling straight along the coordinate axis of the coordinate system of the movable platform according to the straight trajectory, and several straight images are obtained. In this embodiment, the movable platform is made to travel in a straight line along the x-axis of the movable coordinate system for a certain distance. When traveling straight, the acquisition device can observe at least one apriltag code, and the acquisition device acquires an image.

A first coordinate conversion relationship is established, and the conversion relationship between the coordinate system of the marker at each collection moment and the coordinate system of the collection device is calculated, which is recorded as the first coordinate conversion relationship. Preferably, in establishing the first coordinate conversion relationship, the identification information is processed, an image coordinate system is established, and the coordinates of the identification on the image coordinate system and the identification coordinate system are obtained, and combined with the internal parameters and distortion coefficients of the acquisition device, the pose estimation method is used. Calculate the conversion relationship between the identification coordinate system and the acquisition device coordinate system. In this embodiment, the internal parameters and distortion parameters of the acquisition device are calibrated by using the checkerboard calibration board and the acquisition device calibration tool in MATLAB, and the image coordinates of the four marked points on the apriltag code are identified by image processing methods according to the image coordinates and marks of the marked points. The points correspond to the coordinates of the coordinate system of the marker board, and combined with the internal parameters and distortion coefficients of the acquisition device, the transformation relationship between the coordinate system of the marker board and the coordinate system of the acquisition device is calculated by a pose estimation method such as the PNP pose estimation method.

In an embodiment, preferably, establishing the first coordinate conversion relationship includes the following steps:

Generate the rotation pose, process the rotated image, and obtain the image coordinates of the marked point. According to the image coordinates of the marked point and the coordinate system of the marking plate corresponding to the marked point, the coordinates of the coordinate system are identified, and the internal parameters of the acquisition device and the distortion coefficient are used to calculate the marked plate. The coordinate system identifies the coordinate system relative to the pose in the coordinate system of the acquisition device, and obtains the rotation pose; the rotation pose includes the position and the attitude, the position is represented by a translation vector, and the attitude is represented by a rotation vector. During the rotation process, it is assumed that a total of M rotation poses.

Generate the straight pose, process the straight image, and obtain the image coordinates of the marked points. According to the image coordinates of the marked points and the coordinates of the marking plate coordinate system corresponding to the marked points, the coordinates of the coordinate system are identified, and the internal parameters of the acquisition device and the distortion coefficient are used to calculate the marked plate. The coordinate system identifies the coordinate system relative to the pose in the coordinate system of the acquisition device, and obtains the straight-line pose. The straight pose includes position and attitude. The position is represented by a translation vector, and the attitude is represented by a rotation vector. It is assumed that a total of N straight poses are saved.

In one embodiment, the world coordinate system is represented by the symbol O ^w -X ^w Y ^w Z ^w , and the world coordinate system is an absolute coordinate system defined in space, which will not change with the movement of the acquisition device or the object. The coordinate system of the movable platform is represented by the symbol O ^b -X ^b Y ^b Z ^b , and the forward direction of the movable platform is parallel to the positive direction of the x-axis. The coordinate system of the acquisition device is represented by the symbol O ^c -X ^c Y ^c Z ^c . The marker plate coordinate system is denoted by the notation O ^m -X ^m Y ^m Z ^m . The fitted plane coordinate system is denoted by the notation ^Op - X ^p Y ^p Z ^p . Image coordinate system, represented by the symbol O ^I -X ^I Y ^I. Let a rotation matrix be ^a R ^b , which is a 3×3 unit matrix, which represents the rotation relative to the coordinate system b under the coordinate system a, and the rotation vector ^a ψ corresponding to the rotation matrix can be obtained through the Rodrigues change ^b ( ^a α ^b , ^a β ^b , ^a γ ^b ) ^T , where T represents the transpose of the matrix. Similarly, the rotation vector can also be obtained by the Rodrigues formula to obtain the rotation matrix. Let a certain translation vector be ^a t ^b , which is a 3×1 vector, which represents the translation relationship with respect to the coordinate system b under the coordinate system a. Let a certain coordinate be ^a p _i ( ^a x _i , ^a y _i , ^a z _i ), which means the coordinates of the i-th point in the coordinate system a. It is assumed that when the image is acquired for the i-th (i=1,...,M+N) time, the acquisition device observes a total of n markers, wherein the image coordinates of the k-th (k=1,...,n) marker are denoted as ^Ii p _k (u _k ,v _k ), that is, the coordinates of the kth marker point in the image obtained by the ith in the image coordinate system, and the coordinates in the marker plate coordinate system ^mi p _k ( ^mi x _k , ^mi y _k , ^mi z _k ) , that is, the coordinate of the kth point in the image obtained by the ith in the coordinate system of the marker plate, and the conversion between the coordinate system of the acquisition device and the coordinate of the calibration plate is obtained as ^{ct mi} , ^c ψ ^mi , if the pose obtained during the rotation process is also can be expressed as Similarly, the pose obtained in the process of going straight can also be expressed as

Establish a second coordinate conversion relationship, obtain the marked points in the identification information on the identification coordinate system, obtain a number of coordinate points corresponding to the marked point in the coordinate system of the acquisition device at each acquisition moment according to the first coordinate conversion relationship, and use a fitting algorithm to fit the points. Combine the coordinate points, establish a fitting coordinate system, calculate the conversion relationship between the fitting coordinate system and the acquisition device coordinate system, and record it as the second coordinate conversion relationship; preferably, in establishing the second coordinate conversion relationship, a plane fitting algorithm is used Fit the coordinate points, generate a fitting plane, and establish a fitting coordinate system in the fitting plane. Preferably, generating the second coordinate conversion relationship includes the following steps:

Calculate the distance between the coordinate point and the fitting plane. When the distance is greater than the threshold, remove the coordinate point and re-fit the remaining coordinate points; in this embodiment, the threshold is 0.0001m, and when the distance is greater than 0.0001m, the coordinate point is determined to be Discrete points, remove the discrete points, and refit the remaining coordinate points until there are no more discrete points.

Establish a fitting plane coordinate system, take the first point and the second point in the fitting plane, the first point is represented by p1, the second point is represented by p2, the coordinates of the first point and the second point in the coordinate system of the acquisition device are respectively is the first coordinate and the second coordinate, the first coordinate is The second coordinate is Calculate the first unit vector, the second unit vector, and the third unit vector according to the first coordinate and the second coordinate. The first unit vector is the unit vector of the difference vector between the first coordinate point and the second coordinate point, and the third unit vector is the fitting The normal vector of the plane, the second unit vector is the vector product of the first unit vector and the third unit vector, with the first point as the coordinate origin, and the directions of the first unit vector, the second unit vector, and the third unit vector are The directions of the x-axis, y-axis, and z-axis establish a fitting coordinate system; specifically, the first unit vector third unit vector second unit vector vector The direction is parallel to the p ₂ p ₁ direction, the vector direction is the plane normal vector direction, p ₂ p ₁ lies in the fitted plane, the vector and vector perpendicular to each other, according to the principle of orthogonality vector, i.e. vector The directions are perpendicular to the vector and vector As shown in FIG. 5 , if the z-axis of the movable platform coordinate system 10 and the z-axis of the acquisition device coordinate system 4 are in the same direction, the point p ₁ is taken as the coordinate origin, and the positive direction of the x-axis is set along the unit vector direction, the positive y-axis is along the unit vector direction, the positive z-axis is along the unit vector The direction establishes the fitting plane coordinate system 5, and obtains the conversion relationship between the fitting plane coordinate system 5 and the acquisition device coordinate system 4:

As shown in FIG. 6 , if the z-axis of the movable platform coordinate system 10 is opposite to the z-axis of the acquisition device coordinate system ₄ , the point p1 is taken as the coordinate origin, and the positive direction of the x-axis is set along the unit vector direction, the negative y-axis is along the unit vector direction, the negative z-axis is along the unit vector The direction establishes the fitting plane coordinate system 5, and obtains the conversion relationship between the fitting plane coordinate system 5 and the acquisition device coordinate system 4:

Wherein, ^c R ^p and ^c t ^p respectively represent the rotation change matrix and translation matrix of the coordinate system of the acquisition device relative to the fitted plane coordinate system. Then the rotation change matrix and translation matrix of the fitted plane coordinate system relative to the acquisition device coordinate system are:

^p R ^c =( ^c R ^p ) ^T , ^p t ^c =-( ^p R ^c · ^c t ^p )

A coordinate conversion relationship is generated, and a conversion relationship between the fitted plane coordinate system and the acquisition device coordinate system is generated according to the first unit vector, the second unit vector, the third unit vector and the first coordinates.

Establish a third coordinate conversion relationship, fit the coordinate points according to the fitting algorithm corresponding to the preset trajectory, generate a fitting trajectory, calculate the conversion relationship between the fitted coordinate system and the movable platform coordinate system, and record it as the third coordinate conversion relationship ; For example, the plane is fitted according to the plane fitting algorithm, the ellipse is fitted according to the ellipse fitting algorithm, and the straight line is fitted according to the straight line fitting algorithm. For example, the coordinates of the marker plate coordinate system corresponding to the marker points in the rotating image and the straight image are converted into the coordinates under the coordinate system of the acquisition device through the rotation pose and the straight pose, respectively, and the coordinate points under the coordinate system of the acquisition device are obtained. The ellipse obtained by the platform rotation is a 2D ellipse, and the collected mark points are 3D points. The plane fitting algorithm is used to fit the coordinate points. The coordinates of the coordinate points are equivalent to the translation vector. The fitting plane equation is:

A _p x+B _p y+C _p z+D _p =0

Among them, (A _p , B _p , C _p ) is the plane normal vector, then the plane intersects the x-axis at intersects the y-axis at intersects the z-axis at A _p , B _p , C _p , D _p can be obtained from the fitted plane. It is also possible to transform the coordinate system of the acquisition device into other coordinate systems determined as required according to the acquisition device model and the mapping relationship.

In one embodiment, preferably, in establishing the third coordinate conversion relationship, ellipse fitting is performed on the coordinate points obtained along the rotation trajectory under the fitting coordinate system, and the center of the ellipse in the fitting coordinate system is the rotation of the movable platform. center to obtain the translation matrix between the fitted plane coordinate system and the movable platform coordinate system; perform straight line fitting on the coordinate points obtained along the straight trajectory under the fitted coordinate system to obtain the fitted plane coordinate system and the movable platform coordinate system The rotation matrix between the fitting coordinate system and the movable platform coordinate system is obtained.

The external parameters of the acquisition device are calibrated, and the conversion relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform is calculated through the second coordinate conversion relationship and the third coordinate conversion relationship. Preferably, the coordinates of the rotation center of the movable platform in the coordinate system of the acquisition device are calculated through the second coordinate conversion relationship and the third coordinate conversion relationship, so as to obtain a translation matrix between the coordinate system of the acquisition device and the coordinate system of the movable platform; The transformation relationship calculates the rotation matrix between the fitting plane coordinate system and the movable platform coordinate system, and calculates the rotation matrix between the acquisition device coordinate system and the movable platform coordinate system through the second transformation relationship; calculates the acquisition device coordinates from the translation matrix and the rotation matrix. The transformation relationship between the system and the coordinate system of the movable platform. As shown in FIG. 7 , according to the conversion relationship between the acquisition device coordinate system 4 and the fitting plane coordinate system 5 , the coordinates of the i-th point in the acquisition device coordinate system 4 are set as Then its coordinates in the fitting plane coordinate system 5 are:

^p p _i = ^p R ^c · ^c p _i + ^p t ^c

Since the discrete points have been removed before fitting, the z-coordinate of the rotation point 7 or the straight point 9 in the fitting plane coordinate system 5 is close to 0, and it can be considered that the coordinate points in the fitting plane coordinate system 5 are all located on the xy plane , take the x and y-axis coordinates of the rotation point 7 under the fitting plane coordinate system 5 to perform ellipse fitting, and generate a rotation trajectory 6. The ellipse fitting method adopts the ellipse fitting function to fit, and the output result is that the center of the ellipse is at the corresponding coordinate The coordinates in the coordinate system, assuming that it is (x _c , y _c ), then the coordinates of the point in the fitting plane coordinate system are (x _c , y _c , 0). According to the principle of relative motion, the center of the ellipse is identified as the movable platform The rotation center of , and the coordinates of the rotation center in the coordinate system of the acquisition device are obtained through the coordinate change:

^c t ^b = ^c R ^p ·[x _c ,y _c ,0] ^T + ^c t ^p

Take the x and y-axis coordinates of the straight point 9 under the fitting plane coordinate 5 to perform straight line fitting, and generate a straight line trajectory 8. The straight line fitting method uses a straight line fitting function for fitting, and the output result is the straight line equation coefficient, which is represented by the straight line equation. for:

B _l ·y=A _l ·x+C _l

According to the coefficients A _l and B _l of the straight line equation, the angle between the straight line and the x-axis is obtained, that is, the angle between the forward direction and the x-axis:

The straight line direction is the forward direction of the movable platform, and the rotation matrix between the fitted plane coordinate system and the movable platform coordinate system is obtained:

Then the rotation relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform:

^c R ^b = ^c R ^p · ^p R ^b

That is, the conversion relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform: the rotation matrix is ^c R ^b , and the translation matrix is ^c t ^b . In order to facilitate the next process optimization, the rotation matrix is converted into a rotation vector ^c ψ ^b ( ^c α ^b , ^c β ^b , ^c γ ^b ).

In one embodiment, it is preferable to further include steps to optimize the calibration result, according to the principle of relative motion, construct a cost function through the conversion relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform, and the first coordinate conversion relationship, and convert the cost function into the cost function. Input the optimization library for calculation, and generate the optimization calibration result. Preferably, optimizing the calibration result further includes establishing an odometer constraint, inputting it into an optimization library for calculation, and generating an optimized calibration result. Preferably, it also includes establishing one or more optimal calibration results among reprojection constraints, circle constraints, and line constraints. The conversion between the coordinate system of the collecting device and the coordinate system of the movable platform is: ^c t ^b ( ^c x ^b , ^c y ^b ), ^c ψ ^b ( ^c α ^b , ^c β ^b , ^c γ ^b ), that is, the light of the collecting device The positional relationship between the center and the coordinates of the movable platform; in the process of rotating straight, the conversion between the coordinate systems of the marker plate under the coordinate system of the acquisition device is: In order to distinguish between the rotation and the straight travel process, the conversion between the coordinate systems of the marker plate under the coordinate system of the acquisition device during the rotation process is: The conversion between the coordinate systems of the marker plate under the coordinate system of the acquisition device during the straight travel is: The coefficients of the straight line are: A _l , B _l , the radius of the circle is r, and the total degree of freedom of the optimization variable is 6M+6N+9. In this embodiment, the cost function is specifically:

1. Straight line constraints, N equations can be obtained:

2. Circle constraints, M equations can be obtained:

3. Reprojection constraints, 2(M+N) equations can be obtained:

Assuming that when the image is acquired for the ith (i=1,...,M+N) time, the acquisition device observes n markers in total, wherein the image coordinates of the jth (j=1,...,n) marker are denoted as ^Ii p _j (u _j , v _j ), the coordinates ^mi p _k ( ^mi x _k , ^mi y _k , ^mi z _k ) under the coordinate system of the marker board, the conversion between the coordinate system of the acquisition device and the coordinate system of the calibration board is obtained as ^c t ^mi , ^c ψ ^mi . In the reprojection process, first calculate the coordinates of the jth marker point in the coordinate system of the acquisition device:

Secondly, the distorted image coordinates are obtained through the distortion model. In theory, the observed image coordinates should be equal to the image coordinates obtained by reprojection:

dist=1+r·(k ₁ +k ₂ ·r)

4. Odometer constraint (trajectory constraint), 6(M+N) equations can be obtained:

For the odometer, the position and attitude of each moment in the world coordinate system can be directly obtained,

That is ( ^{wt b0} , ^w R ^b0 ), ..., ( ^{wt bi} , ^w R ^bi ), ( ^{wt bj} , ^w R ^bj ), the rotational translation of the j-th trajectory point relative to the i-th trajectory point is obtained:

Translation: ^{bit bj} = ( ^w R ^bi ) ^T · ( ^w t ^bj - ^w t ^bi )

Rotation: ^bi R ^bj = ( ^w R ^bi ) ^T · ^w R ^bj

For the acquisition device, the poses ( ^c t ^m0 , cR ^m0 ), ( ^c t ^mi , ^c R ^mi ), ( ^c t ^mj , ^c R ^mj ) between the coordinate system of the acquisition device and the marker plate coordinate system are obtained. , if the calibration plate coordinate system is fixed, the above pose is equivalent to ( ^c0 t ^m , ^c0 R ^m ), ..., ( ^ci t ^m , ^ci R ^m ), ( ^cj t ^m , ^cj R ^m ), through the acquisition device The coordinate system and the movable platform coordinate system get the above trajectory:

Translation: ^bit t ^bj = ^b R ^c · ( ^ci R ^m · ( ^b R ^c · ^cj R ^m ) ^T · (- ^b t ^c )+ ^ci R ^m · ( ^cj R ^m ) ^T · (- ^cj t ^m ) + ^ci t ^m )+ ^b t ^c

Rotation: ^bi R ^bj = ^b R ^c · ^ci R ^m · ( ^b R ^c · ^cj R ^m ) ^T

A total of 9(M+N) equations are obtained, and the degrees of freedom of the variables are only 6N+6, so the optimization conditions are satisfied. The four constraints are used to obtain the final result with the optimization library, and the optimized calibration result is obtained.

An electronic device, comprising: a processor;

a memory; and a program, wherein the program is stored in the memory and is configured to be executed by the processor, the program includes a method for performing the above-mentioned external parameter calibration method of the acquisition device.

A computer-readable storage medium on which a computer program is stored, and the computer program is executed by a processor to perform the above-mentioned method for calibrating external parameters of an acquisition device.

The invention provides a method for calibrating external parameters of a collection device, and the method for calibrating external parameters of a collection device, comprising the following steps: acquiring identification information, and acquiring information collected by a collection device at several collection moments, the information being the relative movement process of the movable platform and the identification along a preset track Establish the first coordinate conversion relationship, calculate the conversion relationship between the logo coordinate system and the acquisition device coordinate system at each collection moment, and record it as the first coordinate conversion relationship; establish the second coordinate conversion relationship, and obtain it from the logo coordinate system For the marked points in the identification information, according to the first coordinate conversion relationship, several coordinate points corresponding to the marked points in the coordinate system of the acquisition device are obtained at each acquisition moment, and a fitting algorithm is used to fit the coordinate points, establish a fitting coordinate system, and calculate the fitting. The conversion relationship between the coordinate system and the coordinate system of the acquisition device is recorded as the second coordinate conversion relationship; the third coordinate conversion relationship is established, the coordinate points are fitted according to the fitting algorithm corresponding to the preset trajectory, the fitting trajectory is generated, and the fitting is calculated. The transformation relationship between the coordinate system and the coordinate system of the movable platform is recorded as the third coordinate transformation relationship; for the external parameter calibration of the acquisition device, the coordinate system of the acquisition device and the movable platform are calculated through the second coordinate transformation relationship and the third coordinate transformation relationship. Transformation relationship between coordinate systems. The invention relates to an electronic device and a readable storage medium, which are used to perform a method for calibrating external parameters of a collection device; the invention performs external parameter calibration of the collection device by adding a fitting plane, so as to reduce mechanical installation errors caused by uneven installation of the collection device , by optimizing the calibration results to make the calibration results more accurate, independent of the high-precision odometer, and the calibration error is small.

The above are only preferred embodiments of the present invention, and do not limit the present invention in any form; any person of ordinary skill in the industry can smoothly implement the present invention as shown in the accompanying drawings and above; however, any Those skilled in the art, without departing from the scope of the technical solution of the present invention, make use of the above-disclosed technical content to make some changes, modifications and equivalent changes of evolution are equivalent embodiments of the present invention; at the same time, Any alteration, modification and evolution of any equivalent changes made to the above embodiments according to the essential technology of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims (14)

1. a method for calibrating external parameters of an acquisition device, characterized in that it comprises the following steps: The collection device obtains information collected at several collection moments, and the information is the identification information during the relative movement of the movable platform and the identification along the preset track; Calculate the conversion relationship between the coordinate system of each collection moment and the coordinate system of the collection device, and record it as the first coordinate conversion relationship; Acquire the marker points in the marker information on the marker coordinate system, obtain a number of coordinate points corresponding to the marker points in the coordinate system of the collecting device at each collection moment according to the first coordinate conversion relationship, and use an approximate Fitting the coordinate points with a combined algorithm, establishing a fitting coordinate system, and calculating the conversion relationship between the fitting coordinate system and the acquisition device coordinate system, which is recorded as the second coordinate conversion relationship; Fitting the coordinate points according to the fitting algorithm corresponding to the preset trajectory, generating a fitting trajectory, and calculating the conversion relationship between the fitting coordinate system and the movable platform coordinate system, which is recorded as a third coordinate conversion relationship; Through the second coordinate transformation relationship and the third coordinate transformation relationship, the transformation relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform is calculated. 2. The method for calibrating external parameters of a collection device according to claim 1, wherein: in the establishment of the first coordinate conversion relationship, the identification information is processed, an image coordinate system is established, and the identification is obtained in the The image coordinate system and the coordinates on the identification coordinate system, combined with the internal parameters and distortion coefficients of the acquisition device, calculate the transformation relationship between the identification coordinate system and the acquisition device coordinate system through a pose estimation method. 3 . The method for calibrating external parameters of an acquisition device according to claim 2 , wherein the preset trajectory comprises a rotation trajectory and a straight trajectory generated by the movable platform with the identification as a reference frame. 4 . 4. The method for calibrating external parameters of an acquisition device as claimed in claim 3, wherein the acquiring identification information comprises the following steps: Acquire the identification information collected by the acquisition device during the in-situ rotation of the movable platform according to the rotation trajectory, and acquire several rotating images; The identification information collected by the collecting device is acquired when the movable platform travels in a straight line along the coordinate axis of the coordinate system of the movable platform according to the straight trajectory, and several straight images are obtained. 5. The method for calibrating external parameters of an acquisition device as claimed in claim 4, wherein the establishing the first coordinate conversion relationship comprises the following steps: The rotated image is processed to obtain the image coordinates of the marker points, and the coordinates of the coordinate system are identified according to the image coordinates of the marker points and the marker plate coordinate system corresponding to the marker points, and combined with the internal parameters of the acquisition device and distortion coefficient, calculate the pose of the marker plate coordinate system and the marker coordinate system relative to the coordinate system of the acquisition device, and obtain the rotational pose; Process the straight image to obtain the image coordinates of the marked points, and identify the coordinates of the coordinate system according to the image coordinates of the marked points and the marking plate coordinate system corresponding to the marked points, and combine the internal parameters of the acquisition device and the distortion coefficient, calculate the pose of the marker plate coordinate system marked coordinate system relative to the coordinate system of the acquisition device, and obtain the straight-ahead pose. 6 . The method for calibrating external parameters of a collection device according to claim 1 , wherein: in the establishment of the second coordinate conversion relationship, a plane fitting algorithm is used to fit the coordinate points, and a fitting plane is generated. 6 . A fitting coordinate system is established in the fitting plane. 7. The method for calibrating external parameters of an acquisition device as claimed in claim 6, wherein the generating the second coordinate conversion relationship comprises the following steps: Calculate the distance between the coordinate point and the fitting plane, when the distance is greater than a threshold, remove the coordinate point, and re-fit the remaining coordinate points; A first point and a second point are taken in the fitting plane, and the coordinates of the first point and the second point in the coordinate system of the acquisition device are the first coordinate and the second coordinate, respectively. A coordinate and the second coordinate calculate a first unit vector, a second unit vector, and a third unit vector, the first unit vector is the unit vector of the difference vector between the first coordinate point and the second coordinate point, and the third unit vector The vector is the normal vector of the fitting plane, the second unit vector is the vector product of the first unit vector and the third unit vector, and the first point is the origin of the coordinates, respectively The directions of the first unit vector, the second unit vector, and the third unit vector are the directions of the x-axis, the y-axis, and the z-axis to establish a fitting coordinate system; A conversion relationship between the fitting coordinate system and the collecting device coordinate system is generated according to the first unit vector, the second unit vector, the third unit vector and the first coordinate. 8 . The method for calibrating external parameters of an acquisition device according to claim 3 , wherein in the establishment of the third coordinate conversion relationship, ellipse fitting is performed on the coordinate points obtained along the rotation trajectory under the fitting coordinate system. 9 . Combine, record the center of the ellipse in the fitting coordinate system as the rotation center of the movable platform, and obtain the translation matrix between the fitting plane coordinate system and the movable platform coordinate system; The coordinate points are fitted with a straight line to obtain a rotation matrix between the fitted plane coordinate system and the movable platform coordinate system, and the conversion relationship between the fitted coordinate system and the movable platform coordinate system is obtained. 9. The method for calibrating external parameters of an acquisition device as claimed in claim 8, wherein the coordinates of the rotation center of the movable platform under the coordinate system of the acquisition device are calculated through the second coordinate conversion relationship and the third coordinate conversion relationship, Obtain the translation matrix between the coordinate system of the acquisition device and the coordinate system of the movable platform; Calculate the rotation matrix between the fitting plane coordinate system and the movable platform coordinate system through the third transformation relationship, and calculate the rotation matrix between the acquisition device coordinate system and the movable platform coordinate system through the second transformation relationship; The transformation relationship between the coordinate system of the acquisition device and the coordinate system of the movable platform is calculated from the translation matrix and the rotation matrix. 10 . The method for calibrating external parameters of a collection device according to claim 1 , further comprising optimizing the calibration result, through the conversion relationship between the coordinate system of the collection device and the coordinate system of the movable platform, the first A cost function is constructed with a coordinate transformation relationship, and the cost function is input into an optimization library for calculation to generate an optimized calibration result. 11 . The method for calibrating external parameters of an acquisition device according to claim 10 , wherein the optimal calibration result further comprises establishing an odometer constraint, inputting it into an optimization library for calculation, and generating an optimal calibration result. 12 . 12 . The method for calibrating external parameters of an acquisition device according to claim 11 , further comprising establishing one or more optimal calibration results among reprojection constraints, circle constraints, and line constraints. 13 . 13. An electronic device, comprising: a processor; a memory; and a program, wherein the program is stored in the memory and configured to be executed by a processor, the program comprising for performing the method of any one of claims 1-12. 14. A computer-readable storage medium on which a computer program is stored, wherein the computer program is executed by a processor to execute the method according to any one of claims 1-12.