Large cabin pose measurement method and system based on cooperative target and binocular vision

Abstract

The invention provides a large cabin pose measurement method and system based on a cooperative target and binocular vision, and the method specifically comprises the following steps: S1, designing a cooperative target which comprises a feature base circle, checkerboard diagonal corners located in the feature base circle, and a dot coding ring; s2, cooperative targets with different serial number codes are installed in positioning holes in the same side of the large cabin section and an end effector of a posture adjusting mechanism respectively, so that the large cabin section and the posture adjusting mechanism are distinguished and recognized; and S3, carrying out image acquisition by using a binocular camera of which the view field range covers the large cabin section and the attitude adjusting mechanism, carrying out cooperative target identification according to the acquired image, and finally obtaining the six-degree-of-freedom pose of the two cabin sections under the coordinate system of the attitude adjusting mechanism. The pose measurement method and system provided by the invention have the characteristics of high operation efficiency, strong light complexity resistance, excellent measurement precision and the like.

Description

Technical Field

[0001] This invention relates to the field of computer vision technology, specifically to a method and system for measuring the pose of large cabin sections based on cooperative targets and binocular vision. Background Technology

[0002] In key national engineering projects such as aerospace, defense industry, and heavy-duty machinery, high-precision pose measurement of large cylindrical components (such as rocket shells, oil tanks, and pipelines) is a crucial step in assembly, docking, and inspection. Traditional measurement methods (such as total stations and laser trackers) rely on manual operation, which is inefficient and difficult to use for real-time dynamic tracking. In contrast, vision-based pose measurement technology offers advantages such as automation, high precision, and strong real-time performance, making it a research hotspot in industrial measurement.

[0003] Vision-based pose measurement techniques can be divided into two categories based on the number of cameras: The first is monocular vision pose measurement, which has advantages such as simple design structure and a large measurement field of view, but cannot directly measure the depth information of the target object. The second is multi-view vision pose measurement, with binocular vision playing a crucial role. It can directly measure the depth information of the target object, has a stronger ability to handle complex scenes, and also possesses good accuracy and robustness.

[0004] In binocular vision pose measurement schemes, cooperative and non-cooperative methods are generally used. The cooperative method involves pre-installing a known cooperative target on the target spacecraft, tracking the spacecraft's onboard camera to capture images of the cooperative target, identifying the target, and tracking the relative pose relationship between the spacecraft and the target spacecraft to complete the on-orbit mission. The non-cooperative method involves identifying distinctive devices on the target spacecraft to assist in measuring the pose of the space target. Compared to the non-cooperative method, the cooperative method has better robustness and measurement accuracy, making it more suitable for high-reliability spacecraft modules. However, this pose measurement method based on cooperative targets and binocular vision still has problems such as low camera calibration efficiency, target recognition being greatly affected by the scene, and insufficient calculation accuracy within a large field of view. Summary of the Invention

[0005] The purpose of this invention is to propose a method and system for measuring the pose of large-scale cabin sections based on cooperative targets and binocular vision, which features high operating efficiency, strong resistance to complex lighting conditions, and excellent measurement accuracy.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows:

[0007] The method for measuring the pose of large modules based on cooperative targets and binocular vision includes the following steps:

[0008] S1. Design a cooperative target, which includes a feature base circle, a checkerboard diagonal within the feature base circle, and a dot coding ring. The feature base circle and the pattern within the feature base circle use different colors.

[0009] The checkerboard grid includes two diagonally distributed right-angled sectors; the dot coding ring includes several dots arranged in a ring around the outer perimeter of the checkerboard grid, so as to encode the sequence number of the cooperative target by the number and orientation of the dots.

[0010] S2. Cooperative targets with different serial numbers are installed on the positioning holes on the same side of the large module and the end effector of the attitude adjustment mechanism to distinguish and identify the large module and the attitude adjustment mechanism.

[0011] The large module includes a left module and a right module, and each module is equipped with at least three different numbered cooperative targets along the circumference.

[0012] The end effector of the attitude adjustment mechanism is used to adjust the attitude of the large compartment to achieve the installation alignment of the left and right compartments.

[0013] S3. Use a binocular camera with a field of view covering the large cabin section and attitude adjustment mechanism to acquire images, and use the acquired images to identify cooperative targets. Combine the binocular camera's extrinsic parameter matrix to obtain the three-dimensional coordinates of the cooperative target's center corner point in the camera coordinate system.

[0014] The end effector of the attitude adjustment mechanism is made to move along the three motion axes of the attitude adjustment mechanism. Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system are established.

[0015] Based on the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system, the three-dimensional coordinates of the three cooperative target center corners on the left and right sections in the camera coordinate system are converted into three-dimensional coordinates in the attitude adjustment mechanism coordinate system. This is used to establish the coordinate systems of the left and right sections, and to obtain the rotation and translation matrices of the left and right sections with respect to the attitude adjustment mechanism. The rotation and translation matrices of the right section with respect to the left section in the attitude adjustment mechanism coordinate system are then solved, and finally, the six-degree-of-freedom pose of the two sections in the attitude adjustment mechanism coordinate system is obtained.

[0016] A large cabin section attitude measurement system based on cooperative target and binocular vision, wherein the system is implemented using any of the above-mentioned large cabin section attitude measurement methods based on cooperative target and binocular vision, including a cooperative target, a binocular camera, a large cabin section, an attitude adjustment mechanism, and an external parameter calibration scale.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] The pose measurement method and system of the present invention have the characteristics of high operating efficiency, strong resistance to complex light conditions, and excellent measurement accuracy. In addition, the system has a large field of view and strong adaptability to the attitude adjustment mechanism of the cabin section. It provides a new solution for the high-precision assembly of large cylindrical components in the construction fields of national key projects such as aerospace, defense industry, and heavy-duty mechanisms. Attached Figure Description

[0019] Figure 1 This is a structural diagram of the large-section pose measurement system based on cooperative targets and binocular vision according to the present invention;

[0020] Figure 2 This is a schematic diagram of the calibration process of the posture adjustment mechanism of the present invention;

[0021] Figure 3 This is a schematic diagram of the cooperative target of the present invention;

[0022] Figure 4 This is a flowchart of the cooperative target identification method of the present invention;

[0023] Figure 5 This is a schematic flowchart illustrating the extrinsic parameter calibration of the binocular camera according to the present invention;

[0024] Figure 6 This is a schematic flowchart illustrating the kinematic calibration of the posture adjustment mechanism of the present invention;

[0025] Figure 7 This is a schematic diagram of the six-degree-of-freedom pose calculation of the module section of the present invention. Detailed Implementation

[0026] The following is in conjunction with the appendix Figure 1-7 The technical solution of the present invention will be described in detail below.

[0027] This invention proposes a method for measuring the pose of large cabin sections based on cooperative targets and binocular vision, specifically including the following steps:

[0028] S1. Design a cooperative target. The target is a square pattern. The cooperative target includes a feature base circle, a checkerboard diagonal within the feature base circle, and a dot coding ring. The feature base circle and the pattern within the feature base circle are different colors. The white area is filled with directional high-reflective paint.

[0029] The checkerboard grid includes two diagonally distributed right-angled sectors; the dot coding ring includes several dots arranged in a ring around the outer perimeter of the checkerboard grid, so as to encode the sequence number of the cooperative target by the number and orientation of the dots.

[0030] S2. Cooperative targets with different serial numbers are installed on the positioning holes on the same side of the large module and the end effector of the attitude adjustment mechanism to distinguish and identify the large module and the attitude adjustment mechanism.

[0031] The large module includes a left module and a right module, and each module is equipped with at least three different numbered cooperative targets along the circumference.

[0032] The end effector of the attitude adjustment mechanism is used to adjust the attitude of the large compartment to achieve the installation alignment of the left and right compartments.

[0033] S3. Use a binocular camera with a field of view covering the large cabin section and attitude adjustment mechanism to acquire images, and use the acquired images to identify cooperative targets. Combine the binocular camera's extrinsic parameter matrix to obtain the three-dimensional coordinates of the cooperative target's center corner point in the camera coordinate system.

[0034] The end effector of the attitude adjustment mechanism is made to move along the three motion axes of the attitude adjustment mechanism. Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system are established.

[0035] Based on the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system, the three-dimensional coordinates of the three cooperative target center corners on the left and right sections in the camera coordinate system are converted into three-dimensional coordinates in the attitude adjustment mechanism coordinate system. This is used to establish the coordinate systems of the left and right sections, and to obtain the rotation and translation matrices of the left and right sections with respect to the attitude adjustment mechanism. The rotation and translation matrices of the right section with respect to the left section in the attitude adjustment mechanism coordinate system are then solved, and finally, the six-degree-of-freedom pose of the two sections in the attitude adjustment mechanism coordinate system is obtained.

[0036] In this embodiment, the cooperative target identification based on the acquired images is specifically as follows:

[0037] Step A1: Determine the optimal threshold for the image and convert the input image into a binary image based on the optimal threshold; based on the binary image, use an edge detection algorithm to filter out the parts that conform to the feature base circle contour;

[0038] Step A2: Convert the selected feature base circle contour from the binarized image into a polar coordinate image. If it meets the frequency and energy conditions of the Fourier function, it is identified as a diagonal pattern.

[0039] Step A3: Using the center corner points of the opposite corners of the chessboard as feature points, locate the feature points by fitting the center corner points with a non-orthogonal quadratic parabola;

[0040] Step A4: Decode the dot-coded ring to determine the cooperative target sequence number.

[0041] In this embodiment, step A1 is specifically as follows:

[0042] Step A11: Determine the optimal threshold for the image and convert the input image into a binary image based on the optimal threshold:

[0043]

[0044]

[0045]

[0046] Where: L is the total number of gray levels in the image (256), and i is the number of gray levels. ; Let i be the probability of gray level i. , Let N be the number of pixels with gray level i, and N be the total number of pixels in the image. ;l is a possible threshold, and ; and These are the cumulative probabilities for the background and foreground classes, respectively. and These represent the expected grayscale values ​​for the background and foreground categories, respectively.

[0047] Step A12: Extract the fitted ellipse based on the binarized image:

[0048]

[0049] in: The fitted ellipse is obtained by minimizing the geometric error from the candidate set. Let be the ellipse parameter, and c is the center point of the ellipse in the image coordinate system, and , The x and y coordinates of the center point of the ellipse, and the axis length are respectively. a and b represent the major and minor axes of the ellipse, respectively. Rotation angle Let be the counterclockwise rotation angle of the major axis relative to the minor axis of the ellipse. ; To filter out the set of vertices for the contour to be fitted, and , For the selected contours that have sub-contours but whose sub-contours are not nested, p represents the vertex of the contour to be fitted;

[0050] Step A13: Eliminate unfitted contours that deviate from the fitted ellipse using the area ratio criterion:

[0051]

[0052] in: To fit the area of ​​the ellipse, and ; Let be the area of ​​the contour to be fitted, and , , These are the x and y coordinates of the nth vertex of the contour to be fitted in the image coordinate system; The tolerance for area deviation between the profile to be fitted and the fitted ellipse can be selected based on empirical values. ;

[0053] Step A14: Obtain the characteristic base circle that meets the conditions by judging the intersection-union (IoU) ratio.

[0054] ,

[0055] in: The area of ​​the polygon is calculated using Green's formula. The set of points transformed from the vertices of the contour to be fitted. , Indicates the total number of points in the set; To create a discretized polygon that fits the ellipse. , Rotation matrix, Let k'' be the counterclockwise angle of the discretized vertex relative to the center point, with the starting point of the counterclockwise angle coinciding with the X-axis of the image coordinate system. k'' is the index of the discretized vertex. The angular interval between adjacent discretized vertices , This represents the total number of vertices of the fitted contour of the fitted ellipse.

[0056] In this embodiment, step A2 is specifically as follows:

[0057] Step A21: Based on the image of the selected feature base circle contour and the center point of the fitted ellipse Perform polar coordinate transformation:

[0058]

[0059] Where: I is the graph function described in polar coordinates; Represented in polar coordinates The grayscale value of the image at that location; This is the distance from the center point of the fitted ellipse to the discretized vertex (the discretized vertex is the discretized result of the fitted ellipse). , , The ratio of the radius of the diagonal of the chessboard of the cooperative target to the radius of the characteristic base circle; it can be selected as 0.8 based on experience.

[0060] Step A21: Take the average signal value of the polar coordinate image along the angular direction. :

[0061]

[0062] Then perform a Fourier transform to obtain the main frequency after the Fourier transform:

[0063] ,

[0064]

[0065] in: Let be the complex value of the k-th frequency component in the frequency domain. for Discretized form ,Right now , express The One sampling point, , The total number of samples of the signal. j is the imaginary unit; Let be a complex basis function, representing the frequency as The sinusoidal component; The main frequency after Fourier transform;

[0066] Determine the ratio of the main frequency to the target frequency after Fourier transform, and the ratio of the main frequency energy to the total frequency energy. If the following conditions are met, it is determined to be a diagonal pattern of a checkerboard:

[0067]

[0068]

[0069] Wherein: target frequency For the black and white corner frequencies of the cooperative target, ; To allow for an upper limit of the relative error between the main frequency and the target frequency, an empirical value of 0.1 can be used. This represents the complex-valued spectral components of the Fourier transform at the dominant frequency. The minimum proportion of the main frequency energy to the total frequency energy can be taken as 0.7 to ensure that the periodicity of the signal is sufficiently significant.

[0070] In this embodiment, step A3 is specifically as follows:

[0071] A31. The set of points on the two intersecting lines of the diagonal pattern of the chessboard of the cooperative target is obtained by the Canny edge detection algorithm. and :

[0072] ,

[0073] in: and Let x and y be the x and y coordinates of the k'-th point on the intersection line of the two lines on the image of the feature base circle contour, respectively.

[0074] A32, Point Set and Fitting a hyperbola:

[0075]

[0076] in, For point set The coefficients of the quadratic form of the fitted curve. For point set The coefficients of the first term of the fitted curve, For point set Coefficients of the constant term in the fitted curve; For point set The coefficients of the quadratic form of the fitted curve. For point set The coefficients of the first term of the fitted curve, For point set Coefficients of the constant term in the fitted curve;

[0077] A33. Minimize the intersection of two lines by SVD decomposition. and The coefficient vectors of the two fitted curves are obtained:

[0078] ,

[0079] ,

[0080] in: The coefficients are k'×6 quadratic forms, derived from the point set. The coordinates of the points in the middle constitute the structure. For point set The coefficient vector of the fitted curve to be determined; The coefficients are k'×6 quadratic forms, derived from the point set. The coordinates of the points in the middle constitute the structure. For point set The coefficient vector of the fitted curve to be determined;

[0081] A34. After obtaining the coefficient vectors of the two fitted curves, solve the equations of the two curves simultaneously to obtain the coordinates of the intersection point of the two lines on the image, which are the coordinates of the central corner point.

[0082] In this embodiment, step A4 is specifically as follows:

[0083] A41. Images of the selected feature base circle contours The Sobel operator is used to calculate the gradient of the image along the x-axis. and gradient along the y-axis :

[0084] ,

[0085] A42. Calculate the gradient direction for each pixel. and gradient magnitude ;

[0086] Set a gradient magnitude threshold (which can be set to 20), and cluster points with high gradient magnitudes that are higher than the gradient magnitude threshold. The clustered points are used to find the edges of the diagonal pattern of the cooperative target (i.e., the four lines on the cooperative target starting from the center point).

[0087] Each gradient direction Points mapped onto the unit circle Next, the K-Means algorithm is used for clustering, with the number of clusters set to 4 (because the diagonal pattern of the cooperative target has 4 sides). After clustering, four cluster centers that meet the requirements are obtained. , Let x and y represent the gradients of the J-th cluster center point along the x-axis and y-axis, respectively; then calculate the gradient direction angles corresponding to the cluster center points:

[0088]

[0089] Obtain four angles These are the four main gradient directions;

[0090] A43. Since the gradient direction is periodic (0~360°), taking the median value of the four principal gradient directions yields... Four additional angles; then sort all angles and add 0° and 360° as boundaries to obtain the sector boundaries. ,in, Indicates the boundary of sector j'. , , , , , , , , , ;

[0091] A44. Identify the encoded points within the feature base circle contour, and use the center of each circle of the dot encoding ring as the position of the encoded point in the image coordinate system to obtain the encoded point set. , Let the x and y coordinates of the j''-th encoding point be respectively located in the image coordinate system. , This represents the number of coding points currently identified in the cooperative target.

[0092] Then input the center point of the feature base circle. Calculate the polar angle of the coding point :

[0093]

[0094] A45. The sector starting position is selected as the boundary line that changes from black to white in a counterclockwise direction along the diagonal pattern as the starting position for decoding. The sector where the encoding point is located is determined based on the polar angle of the encoding point. To calculate the binary code of the dot-coded ring. :

[0095] , ,

[0096] in, This is an intermediate quantity;

[0097] Since there are two boundary lines in the diagonal pattern, the decoding result is optimized by cyclic shift, and the minimum value is taken as the decoding result of the encoding loop. :

[0098]

[0099] in This indicates the sector where the encoding point is located.

[0100] In this embodiment, the cooperative target identification is performed based on the acquired images, and the three-dimensional coordinates of the center corner point of the cooperative target in the camera coordinate system are obtained by combining the extrinsic parameter matrix of the stereo camera, as follows:

[0101] Step B1: Obtain relative position information using a calibration ruler made from a cooperative target, and calculate the essential matrix E of the binocular camera system based on the SVD decomposition method.

[0102]

[0103] in: as well as The normalized image coordinates of the center corner point identified for the cooperative target in the left and right image coordinate systems, i.e. , , , These are the intrinsic parameter matrices for the left and right cameras, respectively. , These are the left and right center corner points of the image identified through cooperative target identification;

[0104] Step B2: Preliminary calculation of the extrinsic parameter matrix by decomposing the essential matrix:

[0105] ,

[0106]

[0107]

[0108] in: It is a diagonal matrix, and its diagonal elements are the singular values ​​of E. (Because the rank of E is 2); U is an orthogonal matrix, the first column is the epipolar direction in the left camera coordinate system, the second column is the direction orthogonal to the first column, and the third column is the projection of the left camera center onto the left image; The third column of U; V is also an orthogonal matrix, with the first column being the epipolar direction in the right camera coordinate system, the second column being the direction orthogonal to the first column, and the third column being the projection of the right camera center onto the left image; , and , These represent two possible solutions for the extrinsic parameters, where R and t are the rotation and translation matrices of the stereo camera's extrinsic parameters, respectively. The intermediate matrix is ​​used to convert the decomposition result into a rotation matrix;

[0109] Step B3: Based on each set of external parameters obtained, calculate the calibration ruler spacing through triangulation, that is, the actual three-dimensional distance between the center corner points of the two cooperative targets on the calibration ruler:

[0110]

[0111] Where: X1 and X2 are the three-dimensional coordinates of the center corner of the cooperative target in the camera coordinate system, and the binocular camera uses the left camera coordinate system as the camera coordinate system of the binocular system; if The prior length L of the distance from the calibration ruler d If the values ​​are similar (i.e., the difference between the two is less than the preset value), then the extrinsic parameters of the current group are selected as the initial values ​​for subsequent nonlinear optimization.

[0112] Step B4: Capture multiple sets of left and right images with calibration rulers. Optimize the extrinsic parameter matrix of the multiple sets of images by minimizing the residuals. The optimization objective function is as follows:

[0113]

[0114] in: Represents the projection error of three-dimensional coordinates. , For the projection function of the left camera, ; For the projection function of the right camera, ; This represents the three-dimensional coordinates of the center corner of the cooperative target in the camera coordinate system. and Let these represent the identity matrix and the extrinsic parameter matrix, respectively:

[0115] ,

[0116] and The three-dimensional coordinates of the center corner points of the two cooperative targets on the calibration ruler within the n'th image group are represented; N' represents the number of left and right image groups used for calibration. , and , Indicates and and Corresponding 2D image coordinates; The reprojection error weight can be set to 0.2. and Let be the three-dimensional coordinates of the center corner points of the two cooperative targets on the calibration ruler within the n'th image group;

[0117] When the objective function residual reaches its minimum stable value, the final optimized extrinsic rotation matrix of the left camera relative to the right camera in the stereo camera system is obtained. Translation matrix

[0118] Step B5: Input the binocular point pairs identified through cooperative target recognition. , And the fundamental matrix F obtained from camera calibration;

[0119] Binocular point pair , This represents the coordinates of the center corner of the cooperative target for left and right image recognition. , , These represent the x and y coordinates of the center corner of the cooperative target in the left image recognition. These represent the horizontal and vertical coordinates of the center corner of the cooperative target in the right image recognition, respectively.

[0120] The fundamental matrix F is calculated as follows:

[0121]

[0122] In the formula: Let be the antisymmetric matrix formed, and let but:

[0123]

[0124] Then, based on epipolar constraints, the geometric error cost function is minimized for the stereo point pair. , Perform matching estimation optimization:

[0125]

[0126]

[0127] in The Euclidean distance between two points. and To satisfy the epipolar constraint point pairs;

[0128] When the analytical global optimum of the geometric error cost function under the Gaussian noise assumption is obtained, the optimized point pair is returned. , ;

[0129] After removing mismatched points using RANSAC homography estimation, the normalized camera coordinates are calculated:

[0130]

[0131] and The normalized image coordinates of the center corner of the cooperative target in the left and right image coordinate systems; This represents the coordinates of the image center in the left camera intrinsic parameter matrix. This represents the coordinates of the image center in the right camera intrinsic parameter matrix; Let be the focal length of the left camera in the x and y directions in the left camera intrinsic parameter matrix. is the focal length of the right camera in the x and y directions in the right camera intrinsic parameter matrix;

[0132] Then, using triangulation, decompose the sample using SVD:

[0133]

[0134] in, These are rows 1, 2, and 3 of the left camera projection matrix. This refers to rows 1, 2, and 3 of the right camera projection matrix; The solution is a homogeneous coordinate solution;

[0135] The homogeneous coordinate solution of the center corner point of the cooperative target is obtained, that is:

[0136]

[0137] The three-dimensional coordinates of the center corner of the cooperative target in the normalized camera coordinate system are as follows:

[0138]

[0139] Where W' is the projection scaling factor of the 3D point;

[0140] The normalized 3D coordinates of the center corner of the cooperative target in the camera coordinate system are compared with the optimized binocular point pair for iterative optimization based on LM:

[0141]

[0142] in, This is the normalized projection function from three-dimensional coordinates to two-dimensional coordinates. Under the condition of minimizing reprojection error, the three-dimensional coordinates of the center corner of the cooperative target in the camera coordinate system are obtained.

[0143] In this embodiment, the end effector of the attitude adjustment mechanism moves along the three motion axes of the attitude adjustment mechanism. Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system are established, as follows:

[0144] Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the movement vectors of each coordinate axis are determined, thereby determining the orientation of each axis of the attitude adjustment mechanism coordinate system in the camera coordinate system:

[0145]

[0146] in, The initial coordinates of the center corner of the cooperative target on the end effector of the attitude adjustment mechanism. , , The coordinates of the center corner of the cooperative target after it has moved along the three axes of the attitude adjustment mechanism coordinate system in the camera coordinate system. , , That is, the calculated directions of motion of the attitude adjustment mechanism along the three axes in the camera coordinate system;

[0147] The origin of the coordinate system for the attitude adjustment mechanism is calculated as follows:

[0148]

[0149] in, Let the initial coordinates of the center corner of the cooperative target be on the attitude adjustment mechanism axis. , , The coordinates of the cooperative target after it has rotated along the three axes of the attitude adjustment mechanism. This represents the origin of the attitude adjustment mechanism's coordinate system in the camera coordinate system. , The x, y, and z coordinates of the origin of the attitude adjustment mechanism's coordinate system in the camera coordinate system;

[0150] The rotation matrix from the camera coordinate system to the attitude adjustment mechanism coordinate system is obtained based on the motion directions of the attitude adjustment mechanism along the three axes in the camera coordinate system. , ;

[0151] The translation matrix from the camera coordinate system to the attitude adjustment mechanism coordinate system is obtained based on the x, y, and z axis coordinates of the origin of the attitude adjustment mechanism coordinate system in the camera coordinate system. , .

[0152] In this embodiment, based on the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system, the three-dimensional coordinates of the three cooperative target center corner points on the left and right sections are converted from the camera coordinate system to the attitude adjustment mechanism coordinate system. This is used to establish the left and right section coordinate systems, obtaining the rotation and translation matrices of the left and right sections relative to the attitude adjustment mechanism. The rotation and translation matrices of the right section relative to the left section in the attitude adjustment mechanism coordinate system are then solved, ultimately obtaining the six-DOF pose of the two sections in the attitude adjustment mechanism coordinate system, as detailed below:

[0153] Identify the coordinates of the center corners of the three cooperative targets on the left side in the camera coordinate system. , , And the coordinates of the center corners of the three cooperative targets on the right section in the camera coordinate system. , , ,in, , , The three-dimensional coordinates of the center corner points of the three cooperative targets identified on the left section, in camera coordinates. , , The three-dimensional coordinates of the center corner points of the three cooperative targets identified on the right section in camera coordinates; and These are the three-dimensional coordinates of the center corner of the first cooperative target identified on the left and right sections, respectively, in camera coordinates. and These are the three-dimensional coordinates of the center corner of the second cooperative target identified on the left and right sections, respectively, in camera coordinates. and These are the three-dimensional coordinates of the center corner point of the third cooperative target identified on the left and right sections, respectively, in camera coordinates.

[0154] Then, the coordinates of the cooperative target's three-dimensional points in the camera coordinate system are transformed into the coordinates of the pose adjustment mechanism's three-dimensional points:

[0155]

[0156] in, , The coordinates of the three center corners of the three cooperative targets on the left and right modules in the attitude adjustment mechanism coordinate system are given. and These are the three-dimensional coordinates of the center corner points of the three cooperative targets on the left and right sections in the camera coordinate system. , ;

[0157] Calculate the coordinates of the outer circles of the three points at the center corners of the three cooperative targets on the left section in the coordinate system of the attitude adjustment mechanism. From the coordinates of the center of the circle Establish the coordinate system of the left section The coordinates of the origin of the coordinate system for the left section are: The directions of the three axes of the left section coordinate system , , They are respectively:

[0158]

[0159] Obtain the rotation matrix of the left section with respect to the attitude adjustment mechanism. and translation matrix for:

[0160]

[0161] Calculate the coordinates of the outer circles of the three points at the center corners of the three cooperative targets on the right section in the coordinate system of the attitude adjustment mechanism. From the coordinates of the center of the circle Establish the coordinate system of the right section The coordinates of the origin of the right-side section coordinate system are: The directions of the three axes of the right-side section coordinate system , , They are respectively:

[0162]

[0163] Obtain the rotation matrix of the left section with respect to the attitude adjustment mechanism. and translation matrix for:

[0164] ;

[0165] The rotation and translation matrices of the right section with respect to the left section in the attitude adjustment mechanism coordinate system are obtained by solving the coordinate system of the two sections:

[0166]

[0167] Then rotate the matrix The ZYX coordinate system is decomposed into Euler angles Rx, Ry, and Rz, where Rx, Ry, and Rz represent the rotation angles around the X, Y, and Z axes of the attitude adjustment mechanism coordinate system, respectively. The translation matrix is ​​then... The coordinates are decomposed into x, y, and z, where x, y, and z represent the movement along the X-axis, Y-axis, and Z-axis of the attitude adjustment mechanism coordinate system, respectively. Finally, the six-degree-of-freedom pose of the two modules in the attitude adjustment mechanism coordinate system is x, y, z, Rx, Ry, and Rz.

[0168] The present invention also proposes a large cabin section pose measurement system based on cooperative target and binocular vision. The system is implemented by any of the above-mentioned large cabin section pose measurement methods based on cooperative target and binocular vision, including a cooperative target, a binocular camera, a large cabin section, an attitude adjustment mechanism, and an external parameter calibration scale.

[0169] The main technical specifications of this measurement system are as follows:

[0170] ① The measurement frequency is better than 5Hz;

[0171] ②The measurement range is not less than a three-dimensional space of 2m×2m×3m;

[0172] ③ The measurement accuracy is better than 0.2mm;

[0173] ④ When CPU and memory usage are below 50%, the system can guarantee uninterrupted operation 24 / 7.

[0174] ⑤ The optimal working field of view depth is 1.5m to 2.5m.

[0175] The above are preferred embodiments of the present invention. Any changes made to the technical solution of the present invention that do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for measuring the pose of large cabin sections based on cooperative targets and binocular vision, characterized in that, Specifically, the following steps are included: S1. Design a cooperative target, which includes a feature base circle, a checkerboard diagonal within the feature base circle, and a dot coding ring. The feature base circle and the pattern within the feature base circle use different colors. The checkerboard grid includes two diagonally distributed right-angled sectors; the dot coding ring includes several dots arranged in a ring around the outer perimeter of the checkerboard grid, so as to encode the sequence number of the cooperative target by the number and orientation of the dots. S2. Cooperative targets with different serial numbers are installed on the positioning holes on the same side of the large module and the end effector of the attitude adjustment mechanism to distinguish and identify the large module and the attitude adjustment mechanism. The large module includes a left module and a right module, and each module is equipped with at least three different numbered cooperative targets along the circumference. The end effector of the attitude adjustment mechanism is used to adjust the attitude of the large compartment to achieve the installation alignment of the left and right compartments. S3. Use a binocular camera with a field of view covering the large cabin section and attitude adjustment mechanism to acquire images, and use the acquired images to identify cooperative targets. Combine the binocular camera's extrinsic parameter matrix to obtain the three-dimensional coordinates of the cooperative target's center corner point in the camera coordinate system. The end effector of the attitude adjustment mechanism is made to move along the three motion axes of the attitude adjustment mechanism. Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system are established. Based on the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system, the three-dimensional coordinates of the three cooperative target center corners on the left and right sections in the camera coordinate system are converted into three-dimensional coordinates in the attitude adjustment mechanism coordinate system. This is used to establish the coordinate systems of the left and right sections, and to obtain the rotation and translation matrices of the left and right sections with respect to the attitude adjustment mechanism. The rotation and translation matrices of the right section with respect to the left section in the attitude adjustment mechanism coordinate system are then solved, and finally, the six-degree-of-freedom pose of the two sections in the attitude adjustment mechanism coordinate system is obtained.

2. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 1, characterized in that, The cooperative target identification based on the acquired images is performed as follows: Step A1: Determine the optimal threshold for the image and convert the input image into a binary image based on the optimal threshold; based on the binary image, use an edge detection algorithm to filter out the parts that conform to the feature base circle contour; Step A2: Convert the selected feature base circle contour from the binarized image into a polar coordinate image. If it meets the frequency and energy conditions of the Fourier function, it is identified as a diagonal pattern. Step A3: Using the center corner points of the opposite corners of the chessboard as feature points, locate the feature points by fitting the center corner points with a non-orthogonal quadratic parabola; Step A4: Decode the dot-coded ring to determine the cooperative target sequence number.

3. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 2, characterized in that, Step A1 is as follows: Step A11: Determine the optimal threshold for the image and apply it according to the optimal threshold. Convert the input image into a binary image: Where: L is the total number of gray levels in the image (256), and i is the number of gray levels. ; Let i be the probability of gray level i. , Let N be the number of pixels with gray level i, and N be the total number of pixels in the image. ;l is a possible threshold, and ; and These are the cumulative probabilities for the background and foreground classes, respectively. and These represent the expected grayscale values ​​for the background and foreground categories, respectively. Step A12: Extract the fitted ellipse based on the binarized image: in: The fitted ellipse is obtained by minimizing the geometric error from the candidate set. Let be the ellipse parameter, and c is the center point of the ellipse in the image coordinate system, and , The x and y coordinates of the center point of the ellipse, and the axis length are respectively. a and b represent the major and minor axes of the ellipse, respectively. Rotation angle Let be the counterclockwise rotation angle of the major axis relative to the minor axis of the ellipse. ; To filter out the set of vertices for the contour to be fitted, and , For the selected contours that have sub-contours but whose sub-contours are not nested, p represents the vertex of the contour to be fitted; Step A13: Eliminate unfitted contours that deviate from the fitted ellipse using the area ratio criterion: in: To fit the area of ​​the ellipse, and ; Let be the area of ​​the contour to be fitted, and , , These are the x and y coordinates of the nth vertex of the contour to be fitted in the image coordinate system; The tolerance for the area deviation between the contour to be fitted and the fitted ellipse; Step A14: Obtain the characteristic base circle that meets the conditions by judging the intersection-union (IoU) ratio. , in: The area of ​​the polygon is calculated using Green's formula. The set of points transformed from the vertices of the contour to be fitted. , Indicates the total number of points in the set; To create a discretized polygon that fits the ellipse. , Rotation matrix, Let k'' be the counterclockwise angle of the discretized vertex relative to the center point, with the starting point of the counterclockwise angle coinciding with the X-axis of the image coordinate system. k'' is the index of the discretized vertex. The angular interval between adjacent discretized vertices , This represents the total number of vertices of the fitted contour of the fitted ellipse.

4. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 3, characterized in that, Step A2 is described in detail below: Step A21: Based on the image of the selected feature base circle contour and the center point of the fitted ellipse Perform polar coordinate transformation: Where: I is the graph function described in polar coordinates; Represented in polar coordinates Image grayscale value at the location; To fit the distance from the center point of the ellipse to the discretized vertices, , , The ratio of the radius of the diagonal of the checkerboard grid of the cooperative target to the radius of the characteristic base circle; Step A21: Take the average signal value of the polar coordinate image along the angular direction. : Then perform a Fourier transform to obtain the main frequency after the Fourier transform: ， in: Let be the complex value of the k-th frequency component in the frequency domain. for Discretization form ,Right now , express The One sampling point, , The total number of samples of the signal. j is the imaginary unit; Let be a complex basis function, representing the frequency as The sinusoidal component; The main frequency after Fourier transform; Determine the ratio of the main frequency to the target frequency after Fourier transform, and the ratio of the main frequency energy to the total frequency energy. If the following conditions are met, it is determined to be a diagonal pattern of a checkerboard: Wherein: target frequency For the black and white corner frequencies of the cooperative target, ; To set an upper limit for the relative error between the main frequency and the target frequency; This represents the complex-valued spectral components of the Fourier transform at the dominant frequency. The minimum proportion of the main frequency energy to the total frequency energy.

5. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 3, characterized in that, Step A3 is as follows: A31. The set of points on the two intersecting lines of the diagonal pattern of the chessboard of the cooperative target is obtained by the Canny edge detection algorithm. and : ， in: and Let x and y be the x and y coordinates of the k'-th point on the intersection line of the two lines on the image of the feature base circle contour, respectively. A32, Point Set and Fitting a hyperbola: in, For point set The coefficients of the quadratic form of the fitted curve. For point set The coefficients of the first term of the fitted curve, For point set Coefficients of the constant term in the fitted curve; For point set The coefficients of the quadratic form of the fitted curve. For point set The coefficients of the first term of the fitted curve, For point set Coefficients of the constant term in the fitted curve; A33. Minimize the intersection of two lines by SVD decomposition. and The coefficient vectors of the two fitted curves are obtained: , , in: The coefficients are k'×6 quadratic forms, derived from the point set. The coordinates of the points in the middle constitute the structure. For point set The coefficient vector of the fitted curve to be determined; The coefficients are k'×6 quadratic forms, derived from the point set. The coordinates of the points in the middle constitute the structure. For point set The coefficient vector of the fitted curve to be determined; A34. After obtaining the coefficient vectors of the two fitted curves, solve the equations of the two curves simultaneously to obtain the coordinates of the intersection point of the two lines on the image, which are the coordinates of the central corner point.

6. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 2, characterized in that, Step A4 is described in detail below: A41. Images of the selected feature base circle contours The Sobel operator is used to calculate the gradient of the image along the x-axis. and gradient along the y-axis : ， A42. Calculate the gradient direction for each pixel. and gradient magnitude ; Set a gradient magnitude threshold and cluster points with high gradient magnitudes that exceed the threshold. These clustered points are used to find the edges of the diagonal pattern of the cooperative target in the future. Each gradient direction Points mapped onto the unit circle Then, the K-Means algorithm was used for clustering, with the number of clusters set to 4. After clustering, four cluster centers that met the requirements were obtained. , Let x and y represent the gradients of the J-th cluster center point along the x-axis and y-axis, respectively; then calculate the gradient direction angles corresponding to the cluster center points: Obtain four angles These are the four main gradient directions; A43. Since the gradient direction is periodic, taking the median value of the four principal gradient directions yields... Four additional angles; then sort all angles and add 0° and 360° as boundaries to obtain the sector boundaries. ,in, Indicates the boundary of sector j'. , , , , , , , , , ; A44. Identify the encoded points within the feature base circle contour, and use the center of each circle of the dot encoding ring as the position of the encoded point in the image coordinate system to obtain the encoded point set. , Let the x and y coordinates of the j''-th encoding point be respectively located in the image coordinate system. , This represents the number of coding points currently identified in the cooperative target. Then input the center point of the feature base circle. Calculate the polar angle of the coding point : A45. The sector starting position is selected as the boundary line that changes from black to white in a counterclockwise direction along the diagonal pattern as the starting position for decoding. The sector where the encoding point is located is determined based on the polar angle of the encoding point. To calculate the binary code of the dot-coded ring. : ， ， in, This is an intermediate quantity; Since there are two boundary lines in the diagonal pattern, the decoding result is optimized by cyclic shift, and the minimum value is taken as the decoding result of the encoding loop. : in This indicates the sector where the encoding point is located.

7. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 1, characterized in that, The process involves identifying cooperative targets based on the acquired images and obtaining the three-dimensional coordinates of the center corner of the cooperative target in the camera coordinate system by combining the extrinsic parameter matrix of the binocular camera, as detailed below: Step B1: Obtain relative position information using a calibration ruler made from a cooperative target, and calculate the essential matrix E of the binocular camera system based on the SVD decomposition method. in: as well as The normalized image coordinates of the center corner point identified for the cooperative target in the left and right image coordinate systems, i.e. , , , These are the intrinsic parameter matrices for the left and right cameras, respectively. , These are the left and right center corner points of the image identified through cooperative target identification; Step B2: Preliminary calculation of the extrinsic parameter matrix by decomposing the essential matrix: ， in: It is a diagonal matrix, and its diagonal elements are the singular values ​​of E. U is an orthogonal matrix. The first column is the epipolar direction in the left camera coordinate system, the second column is the direction orthogonal to the first column, and the third column is the projection of the left camera center onto the left image. The third column of U; V is also an orthogonal matrix, with the first column being the epipolar direction in the right camera coordinate system, the second column being the direction orthogonal to the first column, and the third column being the projection of the right camera center onto the left image; , and , These represent two possible solutions for the extrinsic parameters, where R and t are the rotation and translation matrices of the stereo camera's extrinsic parameters, respectively. The intermediate matrix is ​​used to convert the decomposition result into a rotation matrix; Step B3: Based on each set of external parameters obtained, calculate the calibration ruler spacing through triangulation, that is, the actual three-dimensional distance between the center corner points of the two cooperative targets on the calibration ruler: Where: X1 and X2 are the three-dimensional coordinates of the center corner of the cooperative target in the camera coordinate system, and the binocular camera uses the left camera coordinate system as the camera coordinate system of the binocular system; if The prior length L of the distance from the calibration ruler d If the difference is less than the preset value, then the extrinsic parameters of the current group are selected as the initial values ​​for subsequent nonlinear optimization. Step B4: Capture multiple sets of left and right images with calibration rulers. Optimize the extrinsic parameter matrix of the multiple sets of images by minimizing the residuals. The optimization objective function is as follows: in: Represents the projection error of three-dimensional coordinates. , For the projection function of the left camera, ; For the projection function of the right camera, ; This represents the three-dimensional coordinates of the center corner of the cooperative target in the camera coordinate system. and Let these represent the identity matrix and the extrinsic parameter matrix, respectively: ， The three-dimensional coordinates of the center corner points of the two cooperative targets on the calibration ruler within the n'th image group are represented; j represents the sequence number of the two cooperative targets on the calibration ruler; N' represents the number of left and right image groups taken for calibration. and Indicates and Corresponding 2D image coordinates; Weights for reprojection errors; and Let be the three-dimensional coordinates of the center corner points of the two cooperative targets on the calibration ruler within the n'th image group; When the objective function residual reaches its minimum stable value, the final optimized extrinsic rotation matrix of the left camera relative to the right camera in the stereo camera system is obtained. Translation matrix Step B5: Input the binocular point pairs identified through cooperative target recognition. , And the fundamental matrix F obtained from camera calibration; Binocular point pair , This represents the coordinates of the center corner of the cooperative target for left and right image recognition. , , These represent the x and y coordinates of the center corner of the cooperative target in the left image recognition. These represent the horizontal and vertical coordinates of the center corner of the cooperative target in the right image recognition, respectively. The fundamental matrix F is calculated as follows: In the formula: Let be the antisymmetric matrix formed, and let but: Then, based on epipolar constraints, the geometric error cost function is minimized for the stereo point pair. , Perform matching estimation optimization: in The Euclidean distance between two points. and To satisfy the epipolar constraint point pairs; When the analytical global optimum of the geometric error cost function under the Gaussian noise assumption is obtained, the optimized point pair is returned. , ; After removing mismatched points using RANSAC homography estimation, the normalized camera coordinates are calculated: and The normalized image coordinates of the center corner of the cooperative target in the left and right image coordinate systems; This represents the coordinates of the image center in the left camera intrinsic parameter matrix. This represents the coordinates of the image center in the right camera intrinsic parameter matrix; Let be the focal length of the left camera in the x and y directions in the left camera intrinsic parameter matrix. is the focal length of the right camera in the x and y directions in the right camera intrinsic parameter matrix; Then, using triangulation, decompose the sample using SVD: in, These are rows 1, 2, and 3 of the left camera projection matrix. This refers to rows 1, 2, and 3 of the right camera projection matrix; The solution is a homogeneous coordinate solution; The homogeneous coordinate solution of the center corner point of the cooperative target is obtained, that is: The three-dimensional coordinates of the center corner of the cooperative target in the normalized camera coordinate system are as follows: Where W' is the projection scaling factor of the 3D point; The normalized 3D coordinates of the center corner of the cooperative target in the camera coordinate system are compared with the optimized binocular point pair for iterative optimization based on LM: in, This is a normalized projection function from 3D coordinates to 2D coordinates; under the condition of minimizing reprojection error, the 3D coordinates of the center corner of the cooperative target in camera coordinates are obtained.

8. The method for measuring the pose of large cabin sections based on cooperative targets and binocular vision according to claim 1, characterized in that, The end effector of the attitude adjustment mechanism moves along the three motion axes of the attitude adjustment mechanism. Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system are established, as follows: Based on the initial and final positions of the cooperative target on the end effector identified in the camera coordinate system, the movement vectors of each coordinate axis are determined, thereby determining the orientation of each axis of the attitude adjustment mechanism coordinate system in the camera coordinate system: in, The initial coordinates of the center corner of the cooperative target on the end effector of the attitude adjustment mechanism. , , The coordinates of the center corner of the cooperative target after it has moved along the three axes of the attitude adjustment mechanism coordinate system in the camera coordinate system. , , That is, the calculated directions of motion of the attitude adjustment mechanism along the three axes in the camera coordinate system; The origin of the coordinate system for the attitude adjustment mechanism is calculated as follows: in, Let the initial coordinates of the center corner of the cooperative target be on the attitude adjustment mechanism axis. , , The coordinates of the cooperative target after it has rotated along the three axes of the attitude adjustment mechanism. This represents the origin of the attitude adjustment mechanism's coordinate system in the camera coordinate system. , The x, y, and z coordinates of the origin of the attitude adjustment mechanism's coordinate system in the camera coordinate system; The rotation matrix from the camera coordinate system to the attitude adjustment mechanism coordinate system is obtained based on the motion directions of the attitude adjustment mechanism along the three axes in the camera coordinate system. , ; The translation matrix from the camera coordinate system to the attitude adjustment mechanism coordinate system is obtained based on the x, y, and z axis coordinates of the origin of the attitude adjustment mechanism coordinate system in the camera coordinate system. , .

9. The method for measuring the pose of a large module based on cooperative targets and binocular vision according to claim 8, characterized in that, The rotation and translation matrices from the camera coordinate system to the attitude adjustment mechanism coordinate system are used to convert the three-dimensional coordinates of the three cooperative target center corners on the left and right sections from the camera coordinate system to the attitude adjustment mechanism coordinate system. This is used to establish the left and right section coordinate systems and obtain the rotation and translation matrices of the left and right sections relative to the attitude adjustment mechanism. The rotation and translation matrices of the right section relative to the left section in the attitude adjustment mechanism coordinate system are then solved, and finally, the six-DOF pose of the two sections in the attitude adjustment mechanism coordinate system is obtained, as detailed below: Identify the coordinates of the center corners of the three cooperative targets on the left side in the camera coordinate system. , , And the coordinates of the center corners of the three cooperative targets on the right section in the camera coordinate system. , , ,in, , , The three-dimensional coordinates of the center corner points of the three cooperative targets identified on the left section, in camera coordinates. , , The three-dimensional coordinates of the center corner points of the three cooperative targets identified on the right section in camera coordinates; and These are the three-dimensional coordinates of the center corner of the first cooperative target identified on the left and right sections, respectively, in camera coordinates. and These are the three-dimensional coordinates of the center corner of the second cooperative target identified on the left and right sections, respectively, in camera coordinates. and These are the three-dimensional coordinates of the center corner point of the third cooperative target identified on the left and right sections, respectively, in camera coordinates. Then, the coordinates of the cooperative target's three-dimensional points in the camera coordinate system are transformed into the coordinates of the pose adjustment mechanism's three-dimensional points: in, , The coordinates of the three center corners of the three cooperative targets on the left and right modules in the attitude adjustment mechanism coordinate system are given. and These are the three-dimensional coordinates of the center corner points of the three cooperative targets on the left and right sections in the camera coordinate system. , ; Calculate the coordinates of the outer circles of the three points at the center corners of the three cooperative targets on the left section in the coordinate system of the attitude adjustment mechanism. From the coordinates of the center of the circle Establish the coordinate system of the left section The coordinates of the origin of the coordinate system for the left section are: The directions of the three axes of the left section coordinate system , , They are respectively: Obtain the rotation matrix of the left section with respect to the attitude adjustment mechanism. and translation matrix for: Calculate the coordinates of the outer circles of the three points at the center corners of the three cooperative targets on the right section in the coordinate system of the attitude adjustment mechanism. From the coordinates of the center of the circle Establish the coordinate system of the right section The coordinates of the origin of the right-side section coordinate system are: The directions of the three axes of the right-side section coordinate system , , They are respectively: Obtain the rotation matrix of the left section with respect to the attitude adjustment mechanism. and translation matrix for: ； The rotation and translation matrices of the right section with respect to the left section in the attitude adjustment mechanism coordinate system are obtained by solving the coordinate system of the two sections: Then rotate the matrix The ZYX coordinate system is decomposed into Euler angles Rx, Ry, and Rz, where Rx, Ry, and Rz represent the rotation angles around the X, Y, and Z axes of the attitude adjustment mechanism coordinate system, respectively. The translation matrix is ​​then... The coordinates are decomposed into x, y, and z, where x, y, and z represent the movement along the X-axis, Y-axis, and Z-axis of the attitude adjustment mechanism coordinate system, respectively. Finally, the six-degree-of-freedom pose of the two modules in the attitude adjustment mechanism coordinate system is x, y, z, Rx, Ry, and Rz.

10. A large cabin segment pose measurement system based on cooperative target and binocular vision, wherein the system is implemented using the large cabin segment pose measurement method based on cooperative target and binocular vision as described in any one of claims 1-9, and includes a cooperative target, a binocular camera, a large cabin segment, an attitude adjustment mechanism, and an external parameter calibration scale.

Images