Large-view-field high-precision measurement method and system

Abstract

The invention relates to the technical field of binocular vision measurement, solves the technical problems of low measurement precision and incapability of realizing large-range measurement and real-time measurement in a large view field in the prior art, and particularly relates to a large-view-field high-precision measurement method and system. A plurality of cooperative targets used for correcting external parameters of the camera are arranged in the public view field; the measurement method comprises the following steps: calculating initial external parameters of the binocular camera according to the position and attitude data of the follow-up turntable; identifying the coordinate of the center of the cooperative target in the common field of view in the image, and correcting the initial external parameters to obtain external parameters of the binocular camera; and measuring the space coordinates of the measured target by using the corrected external parameters of the binocular camera. According to the invention, the measurement effects of low cost, large field of view and high precision can be realized, the position and attitude data of the binocular measurement system are corrected in real time through the cooperation target to provide accurate external parameter data, and thus the measurement precision of the system is improved.

Description

Technical Field

[0001] The present invention relates to the field of binocular vision measurement technology, and in particular to a large-field-of-view high-precision measurement method and system. Background Art

[0002] Visual measurement is a non-contact measurement system with the advantages of high measurement accuracy, no impact on the measured system, and low system cost. It has broad application prospects in the field of measurement and control. The resolution of the image acquisition system is an important factor affecting the measurement accuracy of the system. The higher the resolution, the higher the measurement accuracy.

[0003] For visual measurement systems with a large field of view, if higher measurement accuracy is expected, the conventional solution is to use a high-resolution image acquisition system or a high-precision turntable to perform follow-up measurement of the target. However, both high-resolution image acquisition systems and high-precision turntables will significantly increase the cost of the entire system. Therefore, how to obtain high measurement accuracy in a large field of view with a lower-resolution camera and an ordinary turntable is a valuable research direction. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a large-field-of-view high-precision measurement method and system, which solves the technical problems of the prior art in which the measurement accuracy is low under a large field of view and the measurement over a large range and in real time cannot be achieved.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: a large field of view and high-precision measurement system, comprising two follow-up turntables with cameras installed at their free ends, which are relatively far away from the target to be measured and distributed on both sides thereof to form a binocular measurement system. The follow-up turntable adjusts the posture of the binocular measurement system in real time according to the moving trajectory of the target to be measured, so that the target to be measured is always in the common field of view of the binocular measurement system, and a number of cooperative targets for correcting the camera extrinsic parameters are arranged in the common field of view.

[0006] Furthermore, the measurement system also includes a data processing workstation for calculating the initial external parameters of the camera according to the position and posture data of the servo turntable itself.

[0007] Furthermore, the cooperative target is composed of a four-quadrant pattern and numbers. The center of the cooperative target is the intersection of two diagonals in the four-quadrant pattern. The numbers in each cooperative target are different and are the ID of the cooperative target, which is used by the target recognition algorithm to automatically distinguish the cooperative targets.

[0008] The technical solution also provides a measurement method applied to the above-mentioned large field of view high-precision measurement system, which comprises the following steps:

[0009] S1. Obtain the position and attitude data of the follow-up turntable, and solve the initial external parameters of the binocular camera according to the position and attitude data of the follow-up turntable;

[0010] S2, identifying the coordinates of the center of the cooperative target in the common field of view in the image, and correcting the initial external parameters to obtain the external parameters of the binocular camera;

[0011] S3. Use the corrected binocular camera external parameters to measure the spatial coordinates of the target according to the binocular intersection measurement principle.

[0012] Further, in step S1, the position of the follower turntable 1 is assumed to be (x1, y1, z1), the yaw angle is A1, and the pitch angle is E1; the position of the follower turntable 2 is assumed to be (x2, y2, z2), the yaw angle is A2, and the pitch angle is E2;

[0013] The external parameters of the two cameras are obtained according to the position and posture data of the follow-up turntables 1 and 2.

[0014] Furthermore, in step S2, the specific process includes the following steps:

[0015] S21. Obtain known parameters of the camera, including the camera's intrinsic parameter K, distortion coefficient D, rotation matrix R, translation matrix T, and the world coordinates of the i-th cooperative target. The projection coordinates of the i-th cooperative target in the image

[0016] S22. Calculate the reprojection error E of the i-th target according to the known parameters i , the calculation formula is:

[0017]

[0018] In the formula, Indicates that there is no error in the camera intrinsic parameter K, distortion coefficient D, rotation matrix R, and translation matrix T. Imaging coordinates on the image; f(·) represents the camera projection equation;

[0019] S23, with the reprojection error E i The objective function is constructed with the minimum as the goal to determine the optimization problem. The expression of the objective function is:

[0020]

[0021] The optimization problem is:

[0022] minE(x)

[0023] Where x represents the parameters to be solved, that is, the optimized external parameters, including the rotation matrix R′ and the translation matrix T′; E represents all reprojection errors E iThe matrix formed is:

[0024]

[0025] In the formula, n is the number of cooperative targets;

[0026] S24, calculating the increment Δ according to the Jacobian matrix J of the objective function, the calculation formula is:

[0027] Δ=-(J T J+λI) -1 J T E

[0028]

[0029] In the formula, x0, x m represents the variable to be solved, that is, the parameter to be solved x; λ is the damping factor; I is the unit matrix; the superscript T represents the transpose;

[0030] S25, iterate the parameter x to be solved according to the increment Δ to obtain the corrected external parameter of the binocular camera, the expression is:

[0031] x ω+1 =x ω +Δ

[0032] In the formula, x ω represents the ω-th iteration of the parameter x to be solved;

[0033] When the increment Δ is less than the increment threshold, or the number of iteration steps is greater than the iteration threshold, the iteration is stopped to obtain the corrected external parameters.

[0034] Further, in step S25, the increment threshold is ≤0.05, and the iteration threshold is ≤10.

[0035] By means of the above technical solution, the present invention provides a large field of view high-precision measurement method and system, which has at least the following beneficial effects:

[0036] 1. The measurement system proposed by the present invention uses a low-cost ordinary turntable to track the target to be measured, and then uses a cooperative target to correct the external parameters of the binocular measurement system to achieve a large field of view and high-precision measurement effect. The cooperative target in the field of view can be used to correct its own external parameters, which improves the measurement accuracy. It does not need to maintain a certain fixed distance and posture relationship with the target to be measured, so it is convenient to implement on site. At the same time, the binocular station layout plan can be adjusted according to the measurement field range, so that the target can be measured under a large field of view.

[0037] 2. The measurement method proposed in the present invention can achieve low-cost, large field of view, and high-precision measurement effects. It can correct the position and posture data of the binocular measurement system in real time through cooperative targets, provide accurate external parameter data for the binocular measurement system, and thus improve the measurement accuracy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0039] Figure 1 is a schematic diagram of a measurement system in the present invention;

[0040] Figure 2 is a schematic diagram of the cooperative target in the present invention;

[0041] Figure 3 is a flow chart of the measurement method in the present invention;

[0042] Figure 4 It is a schematic diagram of the position and posture data of the follow-up turntable in the present invention.

[0043] In the figure: 1. Follow-up turntable; 2. Binocular camera; 3. Cooperative target; 4. Target to be measured; 5. Data processing workstation. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0045] In the prior art, there is a visual measurement method for the deformation of a moving structure, such as the publication number CN113503825A, which provides a large-field-of-view, high-precision, and universal method for the deformation of a moving structure based on dynamic tracking and dynamic visual measurement technology. It is suitable for visual measurement of structural deformation under moving conditions such as translation, lifting, and low-speed rotation, and is also suitable for visual measurement of deformation of different structures under other similar or static conditions. It mainly solves the problem of large-field-of-view deformation of moving structures during the development or service of large and complex equipment in the aerospace, aviation, shipbuilding, heavy machinery and other manufacturing industries, and effectively solves the problem that traditional visual measurement cannot follow the target body, the field of view is very limited, and it is difficult to capture large-scale moving targets. However, the above method has the following substantial defects, as follows:

[0046] High system cost. The existing technology requires that the relative distance and relative attitude angle between the high-stable baseline structure and the measured moving structure be visually measured, and the error between the initial distance and relative attitude angle be kept within ±2%. This requirement places high demands on the accuracy of the high-stable baseline structure for visual measurement, which will increase the cost of the entire system.

[0047] Low measurement accuracy. The existing technology uses two cameras for measurement. The measurement principle should be the binocular intersection measurement principle, but it does not introduce how to correct the jitter of the binocular camera posture in real time during dynamic measurement. It only relies on the mechanical structure to keep the binocular camera structure stable, especially in a vibrating environment, and the system measurement accuracy is low.

[0048] Large-scale measurement cannot be achieved. The binocular camera used in the prior art is installed on the baseline structure, which reduces the flexibility of the system and cannot truly achieve large-scale measurement due to the size limitation of the baseline structure.

[0049] Real-time measurement cannot be achieved. The existing technology requires multiple measurements of the target to be measured each time, and then the average value of the measurement results is taken, so real-time measurement cannot be achieved.

[0050] In view of the technical defects existing in the prior art, this embodiment provides two specific implementation methods, namely, Embodiment 1 and Embodiment 2, wherein Embodiment 1 proposes a new large-field-of-view high-precision measurement system, and Embodiment 2 proposes a measurement method applied to the measurement system of Embodiment 1, so as to solve the technical problems of low measurement accuracy, inability to achieve large-range measurement and real-time measurement under a large field of view in the prior art.

[0051] Embodiment 1

[0052] Please refer to Figure 1-Figure 2 This embodiment proposes a high-precision measurement system with a large field of view, including two servo turntables 1 with cameras 2 installed at their free ends, which are relatively far away from the target 4 and distributed on both sides thereof to form a binocular measurement system. The servo turntable 1 adjusts the posture of the binocular measurement system in real time according to the moving trajectory of the target 4, so that the target 4 is always in the common field of view of the binocular measurement system. Several cooperative targets 3 for correcting the external parameters of the camera 2 are arranged in the common field of view. The measurement system also includes a data processing workstation 5 for solving the initial external parameters of the camera 2 according to the position and posture data of the servo turntable 1 itself.

[0053] In this embodiment, the measurement system uses the binocular intersection measurement principle to measure the target 4. Two cameras 2 are respectively installed on two ordinary follow-up turntables 1 to form a binocular measurement system. The posture of the binocular measurement system can be adjusted in real time according to the moving trajectory of the target 4, so that the target 4 is always in the public field of view of the binocular measurement system. Therefore, there is no need for a high-resolution camera to cover the entire field of view. Considering that the position and posture data provided by the ordinary turntable are of poor accuracy, a number of cooperative targets 3 are set in the field of view. The spatial coordinates of the cooperative targets 3 have been measured in advance. The position and posture data of the binocular measurement system can be corrected in real time through the cooperative targets 3, providing accurate external parameter data for the binocular measurement system, thereby improving the measurement accuracy of the system.

[0054] Specifically, the cooperative targets used to correct the initial external parameters of the camera are as follows: Figure 2 As shown, the cooperative target 3 is composed of a four-quadrant pattern and numbers. The center of the cooperative target 3 is the intersection of two diagonal lines in the four-quadrant pattern, which is easy to identify. The numbers in each cooperative target 3 are different, which is the ID of the cooperative target 3, and is used for the target recognition algorithm to automatically distinguish the cooperative target 3. Compared with the traditional annular coded target, the cooperative target 3 has the advantages of low ID recognition error rate and high target center positioning accuracy.

[0055] This embodiment designs a new large-field-of-view high-precision measurement system, uses a low-cost ordinary turntable to track the target, and then uses a cooperative target to correct the external parameters of the binocular measurement system to achieve a large-field-of-view, high-precision measurement effect.

[0056] It can use the cooperative target in the field of view to correct its own external parameters, improve the measurement accuracy, and does not need to maintain a fixed distance and posture relationship with the target, so it is convenient to implement on site. At the same time, the binocular station layout can be adjusted according to the measurement field of view, so that the target can be measured in a large field of view.

[0057] Embodiment 2

[0058] Based on the first embodiment, due to the low accuracy of the parameters of the servo turntable itself, the external parameters of the camera also have errors. Figure 3-Figure 4 This embodiment proposes a high-precision measurement method with a large field of view, which can use a cooperative target to correct the position and posture data of the binocular measurement system in real time and provide accurate external parameter data for the binocular measurement system. The method includes the following steps:

[0059] S1. Obtain the position and attitude data of the servo turntable, and solve the initial external parameters of the binocular camera according to the position and attitude data of the servo turntable; in this embodiment, the camera is installed on the servo turntable, and the servo turntable can dynamically adjust its own attitude according to the target image captured by the camera, so that the measured target is always located in the camera field of view, thereby realizing large field of view measurement. The servo turntable can output its own position and attitude data, and the camera can solve its own external parameters according to the position and attitude data of the servo turntable. Since the parameter accuracy of the servo turntable itself is low, the external parameters of the camera also have errors. In this embodiment, the external parameters of the camera are corrected by the cooperative target, thereby realizing high-precision measurement.

[0060] Assume that the position of the follower turntable 1 is (x1, y1, z1), the yaw angle is A1, and the pitch angle is E1; the position of the follower turntable 2 is (x2, y2, z2), the yaw angle is A2, and the pitch angle is E2. Figure 4 The external parameters of the two cameras can be obtained through the position and attitude data of the follow-up turntables 1 and 2.

[0061] S2, identify the coordinates of the center of the cooperative target in the public field of view in the image, and correct the initial external parameters to obtain the external parameters of the binocular camera; since the spatial coordinates of the center of the cooperative target have been measured in advance, the initial external parameters of the camera are corrected according to the spatial coordinates of the center of the cooperative target and the coordinates of the target center in the image, thereby improving the measurement accuracy of the system. The specific process includes the following steps:

[0062] S21. Obtain known parameters of the camera. Specifically, let the camera's internal parameter be K, the distortion coefficient be D, the rotation matrix be R, the translation matrix be T, and the world coordinates of the i-th cooperative target be The projection coordinates of the i-th cooperative target in the image are:

[0063] S22. Calculate the reprojection error E of the i-th target according to the known parameters i , the calculation formula is:

[0064]

[0065] In the formula, Indicates that there is no error in the camera intrinsic parameter K, distortion coefficient D, rotation matrix R, and translation matrix T. Imaging coordinates on the image; f(·) represents the camera projection equation;

[0066] S23, with the reprojection error E i The objective function is constructed with the minimum as the goal to determine the optimization problem. The expression of the objective function is:

[0067]

[0068] The optimization problem is:

[0069] minE(x)

[0070] Where x represents the parameters to be solved, that is, the optimized external parameters, including the rotation matrix R′ and the translation matrix T′; E represents all reprojection errors E i The matrix formed is:

[0071]

[0072] In the formula, n is the number of cooperative targets;

[0073] S24, calculating the increment Δ according to the Jacobian matrix J of the objective function, the calculation formula is:

[0074] Δ=-(J T J+λI) -1 J T E

[0075]

[0076] In the formula, x0, x m represents the variable to be solved, that is, the parameter x to be solved; λ is the damping factor; I is the unit matrix; the superscript T represents the transpose;

[0077] S25, iterate the parameter x to be solved according to the increment Δ to obtain the corrected external parameter of the binocular camera, the expression is:

[0078] x ω+1 =x ω +Δ

[0079] In the formula, x ω represents the ω-th iteration of the parameter x to be solved;

[0080] When the increment Δ is less than the increment threshold, or the number of iteration steps is greater than the iteration threshold, the iteration is stopped to obtain the corrected external parameter. In this embodiment, the increment threshold is ≤0.05, and the iteration threshold is ≤10.

[0081] S3. Use the corrected binocular camera external parameters to measure the spatial coordinates of the target under test according to the binocular intersection measurement principle; the binocular intersection measurement principle is also called the binocular intersection positioning principle or the binocular vision measurement principle, which is a technology that uses a binocular vision system for ranging and positioning. This technology is based on the principle that two cameras simulate the stereoscopic vision of the human eye. By acquiring image information at different angles and processing the image in combination with a computer vision algorithm, the positioning and ranging of the target object can be achieved. By calculating the parallax information and utilizing the coordinate transformation principle, accurate measurement of the spatial coordinates of the target under test can be achieved. As a prior art means, this embodiment will not elaborate on the implementation process here.

[0082] The measurement method proposed in this embodiment can achieve low-cost, large field of view, and high-precision measurement effects based on the measurement system of Example 1. It can correct the position and posture data of the binocular measurement system in real time through the cooperative target, provide accurate external parameter data for the binocular measurement system, and thus improve the measurement accuracy of the system.

[0083] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, so the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0084] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the above embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0085] The above implementation methods have been described in detail. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A high-precision measurement system with a large field of view, comprising two follow-up turntables (1) with cameras (2) installed at their free ends, which are relatively far away from a measured object (4) and distributed on both sides thereof to form a binocular measurement system, characterized in that: The follow-up turntable (1) adjusts the posture of the binocular measurement system in real time according to the moving trajectory of the measured target (4), so that the measured target (4) is always in the common field of view of the binocular measurement system, and a plurality of cooperative targets (3) for correcting the external parameters of the camera (2) are arranged in the common field of view.

2. The large field of view high precision measurement system according to claim 1, characterized in that: The measurement system also includes a data processing workstation (5) for calculating the initial external parameters of the camera (2) based on the position and posture data of the servo turntable (1) itself.

3. The large field of view high precision measurement system according to claim 1, characterized in that: The cooperative target (3) is composed of a four-quadrant pattern and numbers. The center of the cooperative target (3) is the intersection of two diagonal lines in the four-quadrant pattern. The numbers in each cooperative target (3) are different and are the ID of the cooperative target (3), which is used for the target recognition algorithm to automatically distinguish the cooperative target (3).

4. A measurement method applied to the large field of view high precision measurement system according to any one of claims 1 to 3, characterized in that: The measurement method includes the following steps: S1, obtain the position and posture data of the follow-up turntable, and solve the initial external parameters of the binocular camera according to the position and posture data of the follow-up turntable; S2, identifying the coordinates of the center of the cooperative target in the common field of view in the image, and correcting the initial external parameters to obtain the external parameters of the binocular camera; S3. Use the corrected binocular camera external parameters to measure the spatial coordinates of the target according to the binocular intersection measurement principle.

5. The measuring method according to claim 4, characterized in that: In step S1, the position of the follower turntable 1 is (x1, y1, z1), the yaw angle is A1, and the pitch angle is E1; the position of the follower turntable 2 is (x2, y2, z2), the yaw angle is A2, and the pitch angle is E2; The external parameters of the two cameras are obtained according to the position and posture data of the follow-up turntables 1 and 2.

6. The measuring method according to claim 4, characterized in that: In step S2, the specific process includes the following steps: S21. Obtain known parameters of the camera, including the camera's intrinsic parameter K, distortion coefficient D, rotation matrix R, translation matrix T, and the world coordinates of the i-th cooperative target. The projection coordinates of the i-th cooperative target in the image S22. Calculate the reprojection error E of the i-th target according to the known parameters i , the calculation formula is: In the formula, Indicates that there is no error in the camera intrinsic parameter K, distortion coefficient D, rotation matrix R, and translation matrix T. Imaging coordinates on the image; f(·) represents the camera projection equation; S23, with the reprojection error E i The objective function is constructed with the minimum as the goal to determine the optimization problem. The expression of the objective function is: The optimization problem is: minE(x) Where x represents the parameters to be solved, that is, the optimized external parameters, including the rotation matrix R′ and the translation matrix T′; E represents all reprojection errors E i The matrix formed is: In the formula, n is the number of cooperative targets; S24, calculating the increment Δ according to the Jacobian matrix J of the objective function, the calculation formula is: Δ=-(J T J+λI) -1 J T AND In the formula, x0, x m represents the variable to be solved, that is, the parameter to be solved x; λ is the damping factor; I is the unit matrix; the superscript T represents the transpose; S25, iterate the parameter x to be solved according to the increment Δ to obtain the corrected external parameter of the binocular camera, the expression is: x ω+1 =x ω +Δ In the formula, x ω represents the ω-th iteration of the parameter x to be solved; When the increment Δ is less than the increment threshold, or the number of iteration steps is greater than the iteration threshold, the iteration is stopped to obtain the corrected external parameters.

7. The measuring method according to claim 6, characterized in that: In step S25, the increment threshold is ≤0.05, and the iteration threshold is ≤10.

Images