Multi-camera system on-orbit autonomous pointing methods, systems, media, and computing devices

Abstract

The present application relates to a kind of multi-camera system in-orbit autonomous orientation method, system, medium and computing device, it includes: based on the star point image of multi-camera acquisition, the image plane coordinates of star point and the end mark point of reference scale mark are obtained, and star map identification matching is carried out, find out the corresponding relationship of each star image point and star in star map, obtain the declination angle of each star image point;According to the image plane coordinates of star point and the end mark point of reference scale mark and the declination angle of each star image point, the multi-source data fusion of star image point, reference length image point and reference length is carried out, the high-precision positioning and orientation of multi-camera are realized by beam adjustment, and the relative exterior orientation parameter between each camera is obtained;The spatial coordinates of reference point are reconstructed based on the calibration result of relative exterior orientation parameter, and the coordinate value of reference point in reference coordinate system is used to solve coordinate system transformation parameter, it is compensated to the relative exterior orientation parameter of each camera, and the absolute exterior orientation parameter of each camera in reference coordinate system is obtained.

Description

Methods, systems, media, and computing devices for on-orbit autonomous orientation of multi-camera systems Technical Field

[0001] This invention relates to the field of three-dimensional vision measurement technology, and in particular to an on-orbit autonomous high-precision orientation method, system, medium, and computing device for multi-camera systems suitable for aerospace three-dimensional vision measurement. Background Technology

[0002] In cutting-edge applications such as on-orbit assembly, structural accuracy testing, deformation monitoring, and maintenance of large aerospace equipment, the spatial three-dimensional coordinates of large equipment or components are required. However, the harsh on-orbit environment, difficulty in human intervention, and limited space and resources on spacecraft render some traditional high-precision three-dimensional measurement equipment and technologies unusable. Visual measurement technology, which uses multiple cameras to capture images of the object under test from multiple angles, can achieve three-dimensional spatial measurement. It offers advantages such as non-contact operation, large measurement range, high measurement speed, and simultaneous multi-point measurement, making it highly promising for on-orbit applications. In recent years, the volume of on-orbit aerospace engineering missions has gradually increased, and large space structural components have been widely used in completing aerospace engineering tasks. To ensure the performance and reliable operation of large structural components, high-precision assembly, monitoring, and maintenance are required, making large-size, high-precision measurement technologies in the on-orbit environment particularly important.

[0003] 3D vision measurement uses multiple cameras to simultaneously capture images of the object under test from different positions and angles to analyze spatial 3D information. This technology has advantages such as non-contact operation, wide measurement range, high measurement speed, and simultaneous multi-point measurement, showing great potential in high-precision on-orbit measurement of large structural components. Camera orientation technology, as a key technology in vision measurement, includes determining the spatial position (localization) and attitude angle (orientation) of the camera, and its accuracy directly affects the system's measurement accuracy. Therefore, exploring effective on-orbit calibration methods for the exterior orientation parameters of the camera system is very important and is a prerequisite and guarantee for achieving high-precision 3D coordinate measurement in vision measurement systems. Traditional camera localization and orientation methods mainly include self-calibration bundle adjustment, relative orientation, and absolute orientation. These methods have the following problems: 1) Self-calibration bundle adjustment is used to calibrate the measurement system. In the measurement of high and low temperature thermal deformation of antenna surfaces, it involves taking pictures of a large number of artificial marker points on the antenna surface at different stations, and then using bundle adjustment to calculate the camera's internal and external parameters. However, this method requires a large number of 3D spatial points uniformly distributed in the field of view. 2) Relative orientation utilizes only the image plane coordinate data of corresponding point pairs to solve for the relative exterior orientation parameters between image pairs, relying on the solution and decomposition of the essential matrix. Furthermore, it uses an implicit imaging model of machine vision, suitable only for providing initial values ​​of exterior orientation parameters. Moreover, due to the inability to provide accurate scale information, its accuracy in 3D measurement remains to be evaluated. 3) Absolute orientation is the process of reconstructing the camera's position and attitude at the time of capture, given the 3D coordinates of a single image point and its corresponding control point, given the camera's interior orientation parameters (principal point, focal length, and distortion coefficients). It typically uses resection. Absolute orientation requires selecting a suitable camera position based on the structural characteristics of the object being measured, the measurement space, and the camera's field of view. It requires a large number of spatially evenly distributed reference points with known 3D coordinates, and the orientation accuracy is greatly affected by the accuracy of the reference point coordinates. On-orbit visual measurement differs from measurement in a ground-based simulated environment, lacking structured cooperative targets. Taking antennas as an example, setting markers on the antenna surface affects its performance; the distribution of markers is limited, and the stability and high accuracy of markers cannot be guaranteed in the on-orbit environment. Therefore, the aforementioned traditional methods cannot achieve on-orbit positioning and orientation of the camera system.

[0004] Currently, spacecraft often use star sensors to observe stars for attitude measurement, achieving accuracy down to the arcsecond level. This allows for on-orbit attitude determination without relying on any measurement accessories, providing some insights for vision systems. However, star points cannot provide camera position information, so high-precision multi-camera positioning remains a challenge in the absence of structured cooperative target points. Summary of the Invention

[0005] To address the aforementioned problems, the purpose of this invention is to provide an on-orbit autonomous orientation method, system, medium, and computing device for multi-camera systems. This method provides accurate and reliable exterior orientation parameters for visual measurement systems to calculate three-dimensional spatial coordinates, thus solving the problem of the lack of artificial reference objects to assist camera orientation in current on-orbit multi-camera visual measurement systems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an on-orbit autonomous orientation method for a multi-camera system, comprising: acquiring the image plane coordinates of the star points and the marker points at both ends of the reference scale based on star point images acquired by multiple cameras, performing star map recognition and matching to find the correspondence between each star image point and the stars in the star map, and obtaining the right ascension and declination angles of each star image point; performing multi-source data fusion of star image points, reference length image points, and reference length based on the image plane coordinates of the star points and the marker points at both ends of the reference scale and the right ascension and declination angles of each star image point, achieving high-precision positioning and orientation of the multiple cameras through beam adjustment, and obtaining the relative exterior orientation parameters between the cameras; reconstructing the spatial coordinates of the reference point based on the calibration results of the relative exterior orientation parameters, and using the coordinate values ​​of the reference point in the reference coordinate system to calculate the coordinate system transformation parameters, compensating them into the relative exterior orientation parameters of each camera, and obtaining the absolute exterior orientation parameters of each camera in the reference coordinate system.

[0007] Furthermore, the image plane coordinates of the star point are the star point imaging equation for the camera's star imaging process:

[0008]

[0009] In the formula, f1 is the implicit star imaging equation; m = 1, 2, 3…M represents the m-th camera; s = 1, 2, 3…S represents the s-th star; the subscript sm represents the s-th star within the field of view of the m-th camera; p is the principal distance of the camera; α and β are the right ascension and declination of the star, respectively; xy sm This represents the coordinate vector of the star point image plane after distortion correction.

[0010] Furthermore, the image plane coordinates of the marker points at both ends of the reference scale are the imaging equations of the reference scale marker points in the relative exterior orientation parameter imaging model:

[0011]

[0012] Where f2 represents the implicit collinearity equation based on relative exterior orientation parameters; m = 1, 2, 3…M represents the m-th camera; n = 1, 2, 3…N represents the n-th length ruler; k = 1, 2 represents the k-th endpoint of the length ruler; xy mnk Represents the vector of coordinate points on the image plane of the length scale; X nk T represents the spatial coordinates of the k-th endpoint of the n-th length ruler; mE represents the camera's translation vector. m This represents the camera's attitude angle vector.

[0013] Furthermore, the relative exterior orientation parameters between each camera are obtained, including:

[0014] Based on the image plane coordinates xy of several star points sm By matching star charts, the right ascension and declination angles of the stars corresponding to these star images are obtained. The linear approximation equations of all star imaging equations on all cameras are combined to obtain the error equation of star imaging.

[0015] For any reference ruler mark point, after a first-order Taylor expansion, a linear approximation equation for the image equation of the reference ruler mark point is obtained; and based on the known length of the reference ruler, a linear approximation equation for the preset spatial length information constraint condition is obtained.

[0016] Based on the linear approximation equation of the imaging equation of the reference scale marker point and the linear approximation equation of the spatial length information constraint, the extended error equation containing the image plane coordinates of the reference length scale endpoint and the spatial distance constraint is obtained.

[0017] Based on the error equation of star point imaging and the extended error equation containing the image plane coordinates and spatial distance constraints of the reference length scale endpoint, a multi-camera extrinsic parameter joint adjustment model containing starlight vectors, length scale endpoints and reference length constraints is established.

[0018] The optimal solution for the relative external orientation parameters and the spatial coordinates of the scale endpoints is obtained through multiple iterations using least squares.

[0019] Furthermore, the preset spatial length information constraint condition is as follows:

[0020]

[0021] Where n1 and n2 represent the two target points of the nth length ruler, Ln represents the distance between the two ends of the length ruler, and (X, Y, Z) are the spatial coordinates.

[0022] Furthermore, before fusing multi-source data of stellar image points, reference length image points, and reference length, the weights of stellar image points, reference length image points, and reference length are determined.

[0023] Furthermore, the absolute exterior orientation parameters of each camera in the reference coordinate system include:

[0024] The three-dimensional coordinates of the reference point in the measurement coordinate system are obtained by using the relative exterior orientation parameters of each camera and the image plane coordinates of the reference point on each camera through the forward intersection of the light beams.

[0025] Using the coordinates of the reference point in the measurement coordinate system and the reference coordinate system, the six-degree-of-freedom transformation parameters from the reference coordinate system to the measurement coordinate system are solved by nonlinear least squares method;

[0026] The obtained six-degree-of-freedom transformation parameters are compensated into the relative exterior orientation parameters to obtain the absolute exterior orientation parameters of each camera relative to the reference coordinate system.

[0027] A multi-camera system for on-orbit autonomous orientation includes: a first processing module, which acquires the image plane coordinates of the star points and the marker points at both ends of a reference scale based on star point images acquired by multiple cameras, performs star map recognition and matching, finds the correspondence between each star image point and the stars in the star map, and obtains the right ascension and declination angles of each star image point; a second processing module, which performs multi-source data fusion of star image points, reference length image points, and reference length based on the image plane coordinates of the star points and the marker points at both ends of the reference scale and the right ascension and declination angles of each star image point, and achieves high-precision positioning and orientation of the multiple cameras through bundle adjustment, and obtains the relative exterior orientation parameters between the cameras; and a third processing module, which reconstructs the spatial coordinates of the reference point based on the calibration results of the relative exterior orientation parameters, calculates the coordinate system transformation parameters using the coordinate values ​​of the reference point in the reference coordinate system, compensates for them in the relative exterior orientation parameters of each camera, and obtains the absolute exterior orientation parameters of each camera in the reference coordinate system.

[0028] A computer-readable storage medium storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform any of the methods described above.

[0029] A computing device includes: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the methods described above.

[0030] The present invention has the following advantages due to the adoption of the above technical solutions:

[0031] This invention eliminates the coupling between the extrinsic parameters of each camera based on relative extrinsic parameter imaging. Simultaneously, it achieves absolute camera positioning and orientation based on reference target points through joint relative orientation adjustment processing of a multi-camera system using multi-source (stellar image points, reference length image points, and reference length) data fusion. This solves the problem of current on-orbit multi-camera visual measurement systems lacking artificial reference objects to assist camera orientation. This invention is applicable to on-orbit visual measurement in space. Attached Figure Description

[0032] Figure 1a is a schematic diagram of the measuring device in an embodiment of the present invention;

[0033] Figure 1b is a schematic diagram of the orientation device in an embodiment of the present invention;

[0034] Figure 2 is a flowchart of the on-orbit autonomous orientation method of the multi-camera system in an embodiment of the present invention;

[0035] Figure 3 is the celestial coordinate system in an embodiment of the present invention;

[0036] Figure 4 is a schematic diagram of relative external orientation imaging in an embodiment of the present invention;

[0037] Figure 5a is a diagram showing the X-coordinate error distribution in an embodiment of the present invention;

[0038] Figure 5b is a diagram showing the Y-coordinate error distribution in an embodiment of the present invention;

[0039] Figure 5c is a Z-coordinate error distribution diagram in an embodiment of the present invention;

[0040] 1-Star, 2-Camera, 3-Camera mount, 4-Rotation mechanism, 5-Length reference ruler, 6-Reflection point, 7-Object under test, 8-Object under test connector, 9-Reference marker, 10-Computer, 11-Synchronization trigger, 12-Data transmission line. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0042] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0043] In visual measurement, reference length rulers are often used to provide accurate scaling of measurement results or for accuracy evaluation. Their main component is a carbon fiber or Invar alloy rod with an extremely low coefficient of thermal expansion. Reflection target points are fixed at both ends of the rod, and the distance between these target points is calibrated using high-precision instruments such as coordinate measuring machines or laser interferometers. Given that reference length rulers are easy to measure with high precision and maintain length stability during launch and on-orbit operation, they are an easily deployable type of stringent space reference information. Furthermore, stars are widely distributed in the universe, generate a large amount of data, and possess stable and reliable spatial positions, making them suitable as high-precision three-dimensional spatial calibration objects to constrain the camera attitude calculation process.

[0044] Therefore, this invention proposes a method for high-precision positioning and orientation of an on-orbit visual measurement system based on star points and a reference length scale, to solve the problem of the lack of artificial reference objects to assist camera orientation in current on-orbit multi-camera visual measurement systems. The on-orbit autonomous orientation method, system, medium, and computing device of this invention for a multi-camera system includes: acquiring the image plane coordinates of star points and the marker points at both ends of the reference scale based on star point images acquired by multiple cameras, performing star map recognition and matching to find the correspondence between each star image point and the stars in the star map, and obtaining the right ascension and declination angles of each star image point; performing multi-source data fusion of star image points, reference length image points, and the reference length based on the image plane coordinates of the star points and the marker points at both ends of the reference scale and the right ascension and declination angles of each star image point, achieving high-precision positioning and orientation of the multiple cameras through bundle adjustment, and obtaining the relative exterior orientation parameters between the cameras; reconstructing the spatial coordinates of the reference point based on the calibration results of the relative exterior orientation parameters, and using the coordinate values ​​of the reference point in the reference coordinate system to calculate the coordinate system transformation parameters, compensating them into the relative exterior orientation parameters of each camera, and obtaining the absolute exterior orientation parameters of each camera in the reference coordinate system. This invention enables autonomous positioning and orientation of a multi-camera 3D vision measurement system through simple on-orbit operation, minimal weight and volume, and minimal environmental requirements.

[0045] In this embodiment, as shown in Figure 1a, a pre-set orientation and measurement device is used to measure star 1. The measurement device includes four cameras 2, a controller, and a computer. The cameras 2 are mounted on camera mounts 3, and the camera mounts 3 are mounted on a rotating mechanism 4. The rotating mechanism 4 is connected to a synchronization trigger 11 via a data transmission line 12, and the synchronization trigger 11 is connected to the computer 10. The cameras capture images of marker points on the object being measured, the controller controls the cameras to acquire images synchronously, and the computer is equipped with the multi-camera system on-orbit autonomous orientation method of this invention. The computer receives and analyzes the four images, and uses the interior and exterior orientation parameters of each camera to calculate the spatial coordinates of the target point. It should be noted that this measurement device is not the protected content of this invention, and its form can be replaced with other camera systems of different numbers, intersection angles, and layouts.

[0046] As shown in Figure 1b, the orientation device includes a rotating mechanism 4 and four length reference rulers 5 for visual measurement. The length reference rulers 5 are set on a stable structure of the object being measured, on which a connecting piece 8 is mounted, and the object being measured 7 is installed on the connecting piece 8. Each length reference ruler 5 also has a reflection point 6 at both ends. The rotating mechanism generates a pitch rotation, causing the visual measurement system to point towards the starry sky or towards the object being measured. The lengths of the four length reference rulers 5 are determined by the dimensions of the measured space, and are not less than 1 / 3 of the diagonal length of the measured space. They are placed at the edge of the object being measured, and the spatial distance between the reflection points at both ends of each length reference ruler is precisely measured on the ground. This embodiment does not limit the rotation angle of the rotating mechanism or the spatial position and attitude accuracy of the reference rulers. Reference marker points 9 are set on the stable structure of the object being measured to solve the coordinate system transformation relationship, realizing the transformation of the three-dimensional coordinate measurement results from the measurement system coordinate system to the reference coordinate system, thereby obtaining the required deformation or error field data.

[0047] Before the measurement begins, the controller issues a orientation calibration command and controls the rotating mechanism to point towards the zenith. Each camera captures images of star points. To reduce the influence of random errors such as image noise and space radiation, multiple frames can be acquired and averaged. After capturing star points, the control system instructs the rotating mechanism to point towards the object being measured. Each camera captures images of four reference length rulers, and multiple frames can also be acquired and averaged.

[0048] The controller transmits the acquired image data to the computer. It obtains the image plane coordinates of the star points and the markers at both ends of the reference scale (referencing the gray-scale centroid method in photogrammetry); then it identifies the correspondence between each star image point and the stars in the star image, thereby obtaining the right ascension and declination angles of each star image point (referencing star sensor-related identification and matching methods); next, it obtains the matching relationships between reference scale markers on different images, the correspondence between reference scale markers and reference scale points in space, the length information between point pairs, and the matching relationships between reference markers (referencing the epipolar matching method in photogrammetry); then, it runs the orientation module to achieve multi-source (star image points, reference length image points, reference length) data fusion, and uses bundle adjustment to achieve high-precision positioning and orientation of multiple cameras, obtaining the relative exterior orientation parameters between each camera; finally, it reconstructs the spatial coordinates of the reference points using the calibration results of the exterior orientation parameters, and uses the coordinate values ​​of the reference points in the reference coordinate system to calculate the coordinate system transformation parameters, compensating them into the relative exterior orientation parameters of each camera, obtaining their absolute exterior orientation parameters in the reference coordinate system, and the orientation process ends.

[0049] In one embodiment of the present invention, an on-orbit autonomous orientation method for a multi-camera system is provided. This method is an autonomous, high-precision, convenient, and effective positioning and orientation method suitable for on-orbit visual measurement systems in space, providing accurate and reliable exterior orientation parameters for the visual measurement system to solve for three-dimensional spatial coordinates. In this embodiment, the method utilizes stars and a reference length ruler to calibrate the exterior orientation parameters of the multi-camera visual measurement system, as shown in Figure 2. Specifically, the method includes:

[0050] 1) Based on the star point images acquired by multiple cameras, obtain the image plane coordinates of the star points and the marker points at both ends of the reference scale, and perform star map recognition and matching to find the correspondence between each star image point and the stars in the star map, and obtain the right ascension and declination angle of each star image point;

[0051] 2) Based on the image plane coordinates of the star points and the markers at both ends of the reference scale, as well as the right ascension and declination angles of each star image point, multi-source data fusion of star image points, reference length image points, and reference length is performed. High-precision positioning and orientation of multiple cameras is achieved through beam adjustment, and the relative exterior orientation parameters between each camera are obtained.

[0052] 3) Reconstruct the spatial coordinates of the reference point based on the calibration results of the relative exterior orientation parameters, and use the coordinate values ​​of the reference point in the reference coordinate system to calculate the coordinate system transformation parameters, compensate them into the relative exterior orientation parameters of each camera, and obtain the absolute exterior orientation parameters of each camera in the reference coordinate system.

[0053] In step 1) above, the image plane coordinates of the star point are the star point imaging equation for the camera's star imaging process:

[0054]

[0055] In the formula, f1 is the implicit star imaging equation; m = 1, 2, 3…M represents the m-th camera; s = 1, 2, 3…S represents the s-th star; the subscript sm represents the s-th star within the field of view of the m-th camera; p is the principal distance of the camera; α and β are the right ascension and declination of the star, respectively; xy sm This represents the coordinate vector of the star point image plane after distortion correction.

[0056] In this embodiment, the celestial coordinate system has the center of the celestial sphere as its origin, the x-axis pointing to the vernal equinox as its x-axis, and the z-axis pointing to the north celestial pole as its z-axis, as shown in Figure 3. α and β represent the right ascension and declination of a star in the celestial coordinate system, respectively, and the star catalog provides information on the right ascension and declination of stars.

[0057] Then the unit vector u of this star in the celestial coordinate system c for:

[0058]

[0059] Let the coordinates of the star's imaging point be (x, y), and the camera's principal distance be p. Then, what is the unit vector w of the star in the camera coordinate system? s for:

[0060]

[0061] Let R be the rotation moment of the camera coordinate system relative to the celestial coordinate system. Then the unit vector of the same star in both the celestial and camera coordinate systems satisfies the following relationship:

[0062] w s =Ru c

[0063] Where R is the camera relative to the celestial coordinate system y c x c , z c Rotation angle of the shaft The calculated coordinates of the star's image plane are represented by the following formula:

[0064]

[0065]

[0066] R ij This represents an element in the rotation matrix R. The star vector remains constant over long periods, and therefore can serve as a spatial reference to aid in solving for camera attitude angles.

[0067] Simplifying the above equation yields the star-point imaging equation for the camera's star imaging process.

[0068] In step 1) above, the image plane coordinates of the marker points at both ends of the reference scale are the imaging equations of the reference scale marker points in the relative exterior orientation parameter imaging model:

[0069]

[0070] Where f2 represents the implicit collinearity equation based on relative exterior orientation parameters; m = 1, 2, 3…M represents the m-th camera; n = 1, 2, 3…N represents the n-th length ruler; k = 1, 2 represents the k-th endpoint of the length ruler; xy mnk Represents the vector of coordinate points on the image plane of the length scale; X nk T represents the spatial coordinates of the k-th endpoint of the n-th length ruler; m E represents the camera's translation vector. m Represents the camera's attitude angle vector When the first camera is selected as the reference camera for shooting, T1 is [0,0,0] and E1 is [0,0,0].

[0071] In this embodiment, the principles of traditional single-camera star point imaging and marker point imaging based on collinearity equations are relatively mature. However, in the on-orbit environment of space, there is no known spatial position information of any target in the celestial coordinate system to assist in calculating the absolute spatial position of all cameras. When using a spacecraft's camera system to photograph stars and length reference scales in space, since the stars are infinitely far away, the overall positional movement of the measurement network (including the camera system, the reference length scale, and spatial measurement points) does not affect the imaging effect. This means that the entire measurement network can be arbitrarily translated in space, and there is no unique solution to the camera positioning problem. Mathematically, this manifests as a rank-deficient phenomenon in the normal equations during the nonlinear iterative calculation of the camera position using the spatial information of the reference length.

[0072] To address the coupling phenomenon among the exterior orientation parameters of multiple cameras, this invention fixes the measurement coordinate system to one of the cameras (called the reference camera) by using the relative exterior orientation of multiple cameras. This allows the position and attitude of each camera to be uniquely determined by the relative exterior orientation parameters, reducing six unknowns and fundamentally eliminating the problem of non-unique solutions caused by rank deficiency. As shown in Figure 4, P is a point in three-dimensional space, O-XYZ is the left camera coordinate system, O′-X′Y′Z′ is the right camera coordinate system, O and O′ are the optical centers of the left and right cameras respectively, and OP and O′P′ intersect the image planes of the two cameras with points p and p′ respectively, which are the imaging points of point P on the two cameras. The world coordinate system is defined at the left camera, and the right camera has six degrees of freedom in spatial position and attitude relative to the left camera. (where [X0,Y0,Z0] represents spatial location, (for attitude), where X, Y, and Z represent the translation of the right camera coordinate system relative to the left camera coordinate system along the X, Y, and Z axes, and the rotation angles around the Y, X, and Z axes, respectively. These are the relative exterior orientation parameters of the right camera relative to the left camera.

[0073] The relative external orientation parameter imaging equation is established with the three-dimensional coordinates of the endpoints of each length ruler in space and the external orientation parameters of other cameras besides the reference camera as unknowns. The implicit expression of the collinearity equation is shown in equation (2).

[0074] In step 2) above, if the coordinates of several star points on the camera's image plane are known (image processing, target point localization, which can be obtained using existing methods), and the right ascension and declination angles of the corresponding star points (star map recognition and matching, which can be obtained using existing methods), then the three angular degrees of freedom in the exterior orientation parameters can be solved. To obtain the other three translational degrees of freedom, it is necessary to combine spatial three-dimensional information and solve [X0, Y0, Z0] using Equation (2). However, the spatial coordinates of the two endpoints of the length reference scale are unknown, and only the length between the endpoints is known. This invention further illustrates how to use star point imaging and combine spatial length information as constraints to solve all six degrees of freedom in the relative exterior orientation parameters.

[0075] Specifically, obtaining the relative exterior orientation parameters between each camera includes the following steps:

[0076] 2.1) Based on the image plane coordinates xy of several star points sm By matching star charts, the right ascension and declination angles of stars corresponding to these star images are obtained. The linear approximation equations of all star imaging equations on all cameras are combined to obtain the error equation of star imaging.

[0077] In this embodiment, after multiple cameras capture images of the top star, the coordinates (x, y) of several star points on the image plane can be obtained. sm .

[0078] 2.2) For any reference ruler mark point, after a first-order Taylor expansion, a linear approximation equation for the image equation of the reference ruler mark point is obtained; and based on the known length of the reference ruler, a linear approximation equation for the preset spatial length information constraint condition is obtained.

[0079] The preset spatial length information constraint condition is as follows:

[0080]

[0081] Where n1 and n2 represent the two target points of the nth length ruler, Ln represents the distance between the two ends of the length ruler, and (X, Y, Z) are the spatial coordinates.

[0082] After the rotating mechanism drives the visual measurement system to capture images of zenith stars, it obtains the xy coordinates of several star points on the image plane. sm And by matching the star map, the corresponding right ascension and declination angles of the stars are obtained. A first-order Taylor expansion of equation (1) yields a linear approximation equation for the star point imaging equation:

[0083]

[0084] Where v sm It is the residual error of the m-th camera imaging the s-th star point, l sm This represents the difference between the image plane coordinates of the star's image point and the coordinates obtained using the exterior orientation parameter estimates. J smThe Jacobian matrix represents the derivative of the star-point imaging equation (1) with respect to the external parameters of the m-camera. This represents the correction amount for the camera's extrinsic parameters. By simultaneously solving the linear approximation equations of all star imaging equations for all M cameras, we obtain the error equation of the measurement system for star imaging as follows:

[0085]

[0086] Where J represents the Jacobian matrix of the stellar point imaging equation (1) with respect to the camera's exterior orientation parameters.

[0087] For the imaging expression (2) of any reference scale marker point, after expansion using the first-order Taylor formula, the linear approximation equation of the collinearity equation is obtained as follows:

[0088]

[0089] A and B represent the Jacobian matrices obtained by differentiating the collinear equation (2) with respect to the camera's exterior orientation parameters and the spatial coordinates of the length scale endpoints, respectively. and These represent the corrections for the camera's exterior orientation parameters and the spatial coordinates of the length scale endpoints, respectively. mnk and l mnk These represent the residual error of the image plane and the error in calculating the coordinates of the imaging point for the m-th camera relative to the k-th endpoint of the n-th reference length ruler, respectively.

[0090] Furthermore, since the length of the reference length ruler is known, the linear approximate equation for the distance constraint can also be obtained according to formula (3):

[0091]

[0092] Where C represents the Jacobian matrix of the distance constraint equation (3) with respect to the coordinates of the endpoint spatial points, n1 and n2 represent the two target points of the nth length ruler, and L n It indicates the distance between the two ends of the length ruler.

[0093] 2.3) Based on the linear approximation equation of the imaging equation of the reference scale marker point and the linear approximation equation of the spatial length information constraint, the extended error equation containing the image plane coordinates of the reference length scale endpoints and the spatial distance constraint is obtained.

[0094] The extended error equation is:

[0095]

[0096] 2.4) Based on the error equation of star point imaging and the extended error equation containing the image plane coordinates and spatial distance constraints of the reference length scale endpoint, a multi-camera extrinsic parameter joint adjustment model containing starlight vector, length scale endpoint and reference length constraints is established.

[0097] Since relative exterior orientation parameters are used and the coordinate system is defined at camera 1, formula (5) does not optimize or update the exterior orientation parameters of camera 1. By combining error equations (4) and (5), a multi-camera exterior parameter joint adjustment model is established, which includes starlight vectors, visual measurement markers (endpoints of the length ruler), and reference length constraints:

[0098]

[0099] in, and These represent the corrections for the camera's external orientation parameters and the spatial coordinates of the length scale endpoints, respectively.

[0100] 2.5) Solve for the optimal solution of the relative external orientation parameters and the spatial coordinates of the scale endpoints through multiple iterations of least squares.

[0101] In step 2) above, before performing multi-source data fusion of stellar image points, reference length image points, and reference length, the weights of stellar image points, reference length image points, and reference length are determined.

[0102] Because stellar imaging, length scale marker imaging, and length constraint data are different types, their measurement accuracies also differ. Therefore, it is necessary to determine the weights of the three types of observation data during the fusion and adjustment process. Let the prior standard deviation of the length scale endpoint observation error be s. pr The prior standard deviation of the spatial distance length error is s. l The standard deviation of the observation error of the star point image plane is s ps Using the standard deviation of spatial distance as the unit weight error, the weight matrix participating in the joint adjustment optimization can be represented as follows:

[0103]

[0104] Among them, P1, P2, and P3 are all diagonal matrices, with diagonal elements s and s respectively. l 2 / s pr 2 s l 2 / s l 2 and s l 2 / s ps 2 Then the optimal solution for the unknown parameter correction is:

[0105]

[0106] The estimated values ​​of the camera's exterior orientation parameters and the coordinates of the length scale endpoints are continuously corrected through iteration until the optimal result is obtained. A weight matrix is ​​introduced to fuse multiple types of data, which solves the problems of ill-conditioned adjustment calculation models and inaccurate calibration results caused by the different types and error magnitudes of stellar and length scale imaging and length-constrained observation data.

[0107] This invention employs a nonlinear least-squares iterative process constrained by stellar vectors and spatial length, which allows for very low initial accuracy requirements on the unknown exterior orientation parameters (e.g., the rotation angles of each camera can be set so that the camera's optical axis points to the zenith, and the translation amounts can be set to zero vectors), thus ensuring the convergence of the iterative optimization process. At this point, the relative exterior orientation parameters between the cameras in the measurement system are solved, obtaining the six-degree-of-freedom pose of any m-th camera relative to the reference camera coordinate system.

[0108] In step 3) above, although the positioning and orientation results between the cameras have been obtained, these results are relative to the reference camera coordinate system. This means that subsequent spatial coordinate measurements are also defined in the reference camera coordinate system. However, large spacecraft structures have their own design or reference coordinate system, and all deformation or geometric analysis results should be converted to the reference coordinate system. Therefore, in this embodiment, it is necessary to convert the exterior orientation parameters of each camera to the reference coordinate system to achieve absolute positioning and orientation of each camera.

[0109] This invention also establishes several reference points on the reference stable structure of the object under test, whose precise three-dimensional spatial coordinates [X] in the reference coordinate system are... r Y r Z r The coordinates of the reference point are determined in advance on the ground. After the spacecraft structure is launched, the coordinates of the reference point do not change or the change is so small as to be negligible.

[0110] Specifically, the absolute exterior orientation parameters of each camera in the reference coordinate system include the following steps:

[0111] 3.1) Using the relative exterior orientation parameters of each camera and the coordinates of the reference point on the image plane of each camera [x rm y rm The three-dimensional coordinates [X] of the reference point in the measurement coordinate system (reference camera coordinate system) are obtained by forward intersection of the light beams. c Y c Z c ];

[0112] 3.2) Using the coordinates of the reference point in the measurement coordinate system and the reference coordinate system, the six-degree-of-freedom transformation parameters from the reference coordinate system to the measurement coordinate system are solved by nonlinear least squares method.

[0113] 3.3) The obtained six-degree-of-freedom transformation parameters are compensated into the relative exterior orientation parameters to obtain the absolute exterior orientation parameters of each camera relative to the reference coordinate system:

[0114]

[0115] Example: To verify the feasibility and accuracy of the proposed on-orbit calibration method for exterior orientation parameters, four industrial cameras (MV-CH120-10UM / UC) were used, equipped with four lenses (12FA2524-25MP). The camera sensor resolution was 4096×3000 pixels, the camera field of view was 30°, the pixel size was 3.45um×3.45um, and the camera baseline distance was 500mm. In this example, the aforementioned function was achieved by manually controlling the pointing of the measurement system through the rotation structure of the tripod.

[0116] To increase the common field of view and measurement range, the camera intersection angle was adjusted to 10°. Illumination was provided using four flashes. In this embodiment, a carbon fiber reference length ruler was used, with a distance of 1096.0372 mm between the two end markers and a marker diameter of 12 mm.

[0117] An MVM-B-0008 synchronous trigger box was also used to ensure that the four cameras and flash units simultaneously exposed and acquired images of the length ruler. Outdoor calibration experiments were conducted using the aforementioned multi-camera visual measurement system. First, five sky regions were photographed, with 20 star images taken for each region, to complete the multi-camera attitude angle calibration. Then, a length ruler was used to take 100 sets of images at four different positions and attitudes to complete the camera position calibration. Finally, the external parameters of each camera were jointly optimized.

[0118] Among them, camera 1 was selected as the reference camera, and the results of the calibrated relative exterior orientation parameters are shown in Table 1.

[0119] Table 1. Calibration results of relative external orientation parameters

[0120]

[0121] The image plane error of the star point and the endpoint of the length ruler, as well as the length error of the length ruler, were statistically analyzed, and the results are shown in Table 2.

[0122] Table 2 shows the image plane errors of star points, length scale endpoints, and scale length errors.

[0123]

[0124] As can be seen, the method of this invention can successfully calibrate the camera's exterior orientation parameters. Simultaneously, to verify the measurement accuracy of the calibrated system, the control field was measured using both this invention and the V-STARS multi-image photogrammetry system. Twelve coded points, 321 ordinary points, and one crosshair target were arranged on a 2.5m × 1.4m wall. Absolute orientation and coordinate system transformation were performed using the reference point on the crosshair target. The measurement error of the method of this invention was statistically analyzed using the coordinate measurement results of the V-STARS system as the true value, and 300 measurement experiments were conducted. Table 3 shows the root mean square value and maximum value of the spatial coordinate measurement error given by the V-STARS system, demonstrating its high measurement accuracy, which can be used as a true value to evaluate the measurement accuracy of the system.

[0125] Table 3 V-STARS System Coordinate Measurement Errors

[0126]

[0127] The measurement errors of the method of the present invention were statistically analyzed, and the error distribution is shown in Table 4.

[0128] Table 4 Measurement errors of this system

[0129]

[0130] As shown in Figures 5a to 5c, the spatial point reconstruction error histogram shows that the spatial point reconstruction error exhibits a clear normal distribution. The standard deviations of the X, Y, and Z spatial coordinate measurement errors of the method of the present invention are approximately 0.15 mm, 0.04 mm, and 0.05 mm, respectively, and the maximum errors are 0.36 mm, 0.15 mm, and 0.23 mm, respectively. The relative measurement accuracy of the system reaches 1 / 17000 (1σ).

[0131] Therefore, the method of calibrating the exterior orientation parameters of an on-orbit camera by establishing a joint adjustment model through photographing stars and using a reference length ruler has high reliability and accuracy, and can provide a method and data reference for the system parameter calibration problem faced in on-orbit applications of visual measurement.

[0132] In one embodiment of the present invention, a multi-camera system on-orbit autonomous orientation system is provided, comprising:

[0133] The first processing module, based on the star point images acquired by multiple cameras, obtains the image plane coordinates of the star points and the marker points at both ends of the reference scale, performs star map recognition and matching, finds the correspondence between each star image point and the stars in the star map, and obtains the right ascension and declination angle of each star image point.

[0134] The second processing module performs multi-source data fusion of star image points, reference length image points, and reference length based on the image plane coordinates of the star points and the markers at both ends of the reference scale, as well as the right ascension and declination angles of each star image point. It then achieves high-precision positioning and orientation of multiple cameras through beam adjustment and obtains the relative exterior orientation parameters between each camera.

[0135] The third processing module reconstructs the spatial coordinates of the reference point based on the calibration results of the relative exterior orientation parameters, and uses the coordinate values ​​of the reference point in the reference coordinate system to calculate the coordinate system transformation parameters, which are then compensated into the relative exterior orientation parameters of each camera to obtain the absolute exterior orientation parameters of each camera in the reference coordinate system.

[0136] In the above embodiments, the image plane coordinates of the star points are represented by the star point imaging equation for the camera's star imaging process:

[0137]

[0138] In the formula, f1 is the implicit star imaging equation; m = 1, 2, 3…M represents the m-th camera; s = 1, 2, 3…S represents the s-th star; the subscript sm represents the s-th star within the field of view of the m-th camera; p is the principal distance of the camera; α and β are the right ascension and declination of the star, respectively; xy sm This represents the coordinate vector of the star point image plane after distortion correction.

[0139] In the above embodiment, the image plane coordinates of the marker points at both ends of the reference scale are the imaging equations of the reference scale marker points in the relative exterior orientation parameter imaging model:

[0140]

[0141] Where f2 represents the implicit collinearity equation based on relative exterior orientation parameters; m = 1, 2, 3…M represents the m-th camera; n = 1, 2, 3…N represents the n-th length ruler; k = 1, 2 represents the k-th endpoint of the length ruler; xy mnk Represents the vector of coordinate points on the image plane of the length scale; X nk T represents the spatial coordinates of the k-th endpoint of the n-th length ruler; m E represents the camera's translation vector. m This represents the camera's attitude angle vector.

[0142] In the above embodiments, obtaining the relative exterior orientation parameters between each camera includes:

[0143] Based on the image plane coordinates xy of several star points smBy matching star charts, the right ascension and declination angles of stars corresponding to each star image point are obtained. The linear approximation equations of all star image equations on all cameras are combined to obtain the error equation of star image.

[0144] For any reference ruler mark point, after a first-order Taylor expansion, a linear approximation equation for the image equation of the reference ruler mark point is obtained; and based on the known length of the reference ruler, a linear approximation equation for the preset spatial length information constraint condition is obtained.

[0145] Based on the linear approximation equation of the imaging equation of the reference scale marker point and the linear approximation equation of the spatial length information constraint, the extended error equation containing the image plane coordinates of the reference length scale endpoint and the spatial distance constraint is obtained.

[0146] Based on the error equation of star point imaging and the extended error equation containing the image plane coordinates and spatial distance constraints of the reference length scale endpoint, a multi-camera extrinsic parameter joint adjustment model containing starlight vector, length scale endpoint and reference length constraints is established.

[0147] The optimal solution for the relative external orientation parameters and the spatial coordinates of the scale endpoints is obtained through multiple iterations using least squares.

[0148] In the above embodiments, the preset spatial length information constraint condition is:

[0149]

[0150] Where n1 and n2 represent the two target points of the nth length ruler, Ln represents the distance between the two ends of the length ruler, and (X, Y, Z) are the spatial coordinates.

[0151] In the above embodiments, before performing multi-source data fusion of stellar image points, reference length image points, and reference length, the weights of stellar image points, reference length image points, and reference length are determined.

[0152] In the above embodiments, the absolute exterior orientation parameters of each camera in the reference coordinate system include:

[0153] The three-dimensional coordinates of the reference point in the measurement coordinate system are obtained by using the relative exterior orientation parameters of each camera and the image plane coordinates of the reference point on each camera through the forward intersection of the light beams.

[0154] Using the coordinates of the reference point in the measurement coordinate system and the reference coordinate system, the six-degree-of-freedom transformation parameters from the reference coordinate system to the measurement coordinate system are solved by nonlinear least squares method;

[0155] The obtained six-degree-of-freedom transformation parameters are compensated into the relative exterior orientation parameters to obtain the absolute exterior orientation parameters of each camera relative to the reference coordinate system.

[0156] The system provided in this embodiment is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0157] A computing device provided in one embodiment of the present invention can be a terminal, which may include: a processor, a communication interface, memory, a display screen, and an input device. The processor, communication interface, and memory communicate with each other via a communication bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. When the computer program is executed by the processor, it implements an on-orbit autonomous orientation method for a multi-camera system. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals. Wireless communication can be achieved through Wi-Fi, a management network, NFC (Near Field Communication), or other technologies. The display screen can be a liquid crystal display or an e-ink display. The input device can be a touch layer covering the display screen, or buttons, a trackball, or a touchpad mounted on the casing of the computing device, or an external keyboard, touchpad, or mouse. The processor can call logical instructions stored in the memory.

[0158] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0159] In one embodiment of the present invention, a computer program product is provided, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, and when the program instructions are executed by a computer, the computer is able to perform the methods provided in the above-described method embodiments.

[0160] In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided, which stores server instructions that cause a computer to perform the methods provided in the above embodiments.

[0161] The computer-readable storage medium provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0162] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0163] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0164] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0165] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for on-orbit autonomous orientation of a multi-camera system, characterized in that, include: Based on star point images acquired by multiple cameras, the image plane coordinates of the star points and the markers at both ends of the reference scale are obtained. Star map recognition and matching are then performed to find the correspondence between each star image point and the stars in the star map, obtaining the right ascension and declination angles of each star image point. Based on the image plane coordinates of the star points and the markers at both ends of the reference scale, and the right ascension and declination angles of each star image point, multi-source data fusion of star image points, reference length image points, and the reference length is performed. High-precision positioning and orientation of the multiple cameras is achieved through bundle adjustment, obtaining the relative exterior orientation parameters between the cameras, including: based on the image plane coordinates of several star points. By matching star charts to obtain the right ascension and declination angles of stars corresponding to these star image points, and simultaneously solving the linear approximation equations of all star image equations from all cameras, the error equation for star image is obtained. For the image equation of any reference scale marker point, after a first-order Taylor expansion, the linear approximation equation for the image equation of the reference scale marker point is obtained. Based on the known length of the reference scale, the linear approximation equation for the preset spatial length information constraint is obtained. Based on the linear approximation equations of the image equations of the reference scale marker points and the linear approximation equations for the spatial length information constraint, the extended error equation containing the image plane coordinates of the reference scale endpoint and the spatial distance constraint is obtained. Based on the error equation for star image and the extended error equation containing the image plane coordinates of the reference scale endpoint and the spatial distance constraint, a multi-camera extrinsic parameter joint adjustment model containing starlight vectors, scale endpoints, and reference length constraints is established. The optimal solutions for the relative exterior orientation parameters and the spatial coordinates of the scale endpoints are obtained through multiple iterations of least squares. Based on the calibration results of the relative exterior orientation parameters, the spatial coordinates of the reference point are reconstructed, and the coordinate transformation parameters are calculated using the coordinate values ​​of the reference point in the reference coordinate system. These parameters are then compensated into the relative exterior orientation parameters of each camera to obtain the absolute exterior orientation parameters of each camera in the reference coordinate system. This includes: obtaining the three-dimensional coordinates of the reference point in the measurement coordinate system through forward intersection of the ray beams, using the relative exterior orientation parameters of each camera and the image plane coordinates of the reference reference point in each camera; using the coordinate values ​​of the reference reference point in both the measurement and reference coordinate systems, the six-degree-of-freedom transformation parameters from the reference coordinate system to the measurement coordinate system are solved using a nonlinear least squares method; and the obtained six-degree-of-freedom transformation parameters are compensated into the relative exterior orientation parameters to obtain the absolute exterior orientation parameters of each camera relative to the reference reference coordinate system.

2. The on-orbit autonomous orientation method for a multi-camera system as described in claim 1, characterized in that, The image plane coordinates of the star point are the star point imaging equation for the camera's star imaging process: In the formula, f1 is the implicit star imaging equation; m=1,2,3…M represents the m-th camera; s=1,2,3…S represents the s-th star; the subscript sm represents the s-th star in the field of view of the m-th camera; p is the principal distance of the camera; α and β are the right ascension and declination of the star, respectively. xy sm This represents the coordinate vector of the star point image plane after distortion correction.

3. The on-orbit autonomous orientation method for a multi-camera system as described in claim 1, characterized in that, The image plane coordinates of the marker points at both ends of the reference scale are the imaging equations of the reference scale marker points in the relative exterior orientation parameter imaging model: Where f2 represents the implicit collinearity equation based on relative exterior orientation parameters; m=1,2,3…M represents the m-th camera; n=1,2,3…N represents the n-th length ruler; k=1,2 represents the k-th endpoint of the length ruler; xy mnk Represents the vector of coordinate points on the image plane of the length scale; X nk T represents the spatial coordinates of the k-th endpoint of the n-th length ruler; m E represents the camera's translation vector. m This represents the camera's attitude angle vector.

4. The on-orbit autonomous orientation method for a multi-camera system as described in claim 1, characterized in that, The preset spatial length information constraint is as follows: Where n1 and n2 represent the two target points of the nth length ruler, Ln represents the distance between the two ends of the length ruler, and (X, Y, Z) are the spatial coordinates.

5. The on-orbit autonomous orientation method for a multi-camera system as described in claim 1, characterized in that, Before fusing multi-source data of stellar image points, reference length image points, and reference length, the weights of stellar image points, reference length image points, and reference length are determined.

6. A multi-camera system on-orbit autonomous orientation system, used to implement the multi-camera system on-orbit autonomous orientation method as described in any one of claims 1 to 5, characterized in that, include: The first processing module, based on the star point images acquired by multiple cameras, obtains the image plane coordinates of the star points and the marker points at both ends of the reference scale, performs star map recognition and matching, finds the correspondence between each star image point and the stars in the star map, and obtains the right ascension and declination angle of each star image point. The second processing module performs multi-source data fusion of star image points, reference length image points, and reference length based on the image plane coordinates of the star points and the markers at both ends of the reference scale, as well as the right ascension and declination angles of each star image point. It then achieves high-precision positioning and orientation of multiple cameras through beam adjustment and obtains the relative exterior orientation parameters between each camera. The third processing module reconstructs the spatial coordinates of the reference point based on the calibration results of the relative exterior orientation parameters, and uses the coordinate values ​​of the reference point in the reference coordinate system to calculate the coordinate system transformation parameters, which are then compensated into the relative exterior orientation parameters of each camera to obtain the absolute exterior orientation parameters of each camera in the reference coordinate system.

7. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods described in claims 1 to 5.

8. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing any of the methods described in claims 1 to 5.

Images