A combined camera calibration device and method in three-dimensional space

Abstract

The application discloses a three-dimensional space combined camera calibration device and method, which comprises a rectangular calibration board, a calibration module arranged on the calibration board and a camera system arranged in equiangular symmetry on the circumference of a same circle around the periphery of the calibration board; a cylindrical calibration object is arranged in a common shooting space area of the camera system, and the three-dimensional calibration object is synchronously shot by the multi-camera system in a coplanar manner; and the shot image is subjected to operations of laser line extraction, point set clustering, ellipse fitting and matrix matching in sequence, so that the camera calibration is completed. The three-dimensional space combined camera calibration device and method can establish a high-precision, high-robustness and high-efficiency camera calibration.

Description

Technical Field

[0001] This invention belongs to the field of three-dimensional topography measurement and three-dimensional defect detection technology, specifically relating to a three-dimensional combined camera calibration device and method. Background Technology

[0002] In image measurement and machine vision applications, to determine the 3D geometric position of a point on the surface of a spatial object and its corresponding point in the image, a geometric model of camera imaging must be established. These geometric model parameters are the camera parameters. Under most conditions, these parameters (intrinsic parameters, extrinsic parameters, distortion parameters) must be obtained through experimentation and calculation; this process of solving for these parameters is called camera calibration. Whether in image measurement or machine vision applications, camera parameter calibration is a crucial step. The accuracy of the calibration results and the stability of the algorithm directly affect the accuracy of the results produced by the camera. Therefore, proper camera calibration is a prerequisite for subsequent work, and improving calibration accuracy is key to high-precision measurement and inspection.

[0003] The purpose of calibration is mainly to solve two problems: a) to determine the transformation relationship (internal and external parameters) between three-dimensional spatial points and pixel points in the pixel plane under the world coordinate system; b) to determine the distortion coefficients in the camera imaging process for image correction.

[0004] Invention application CN200710062825.6 discloses "a calibration method and apparatus for a multi-camera system." The calibration method includes: a calibration component consisting of three or more marker points with known geometric information performing rigid body motion within the working area of ​​the multi-camera system; acquiring multiple images of the calibration component and inputting them into a computer; extracting the image coordinates of the marker points; treating image points of the same marker point on the calibration component in a certain pose captured by different cameras as a set of corresponding points; linearly reconstructing the projection transformation matrix of each camera in the multi-camera system based on the geometric information contained in the corresponding image points and marker points; and optimizing the matrix using a nonlinear algorithm. The apparatus includes: a calibration component, a handle, an operating system, a working area, and a multi-camera system.

[0005] The invention application with application number CN201710564438.6 discloses "a multi-camera intrinsic and extrinsic parameter calibration method based on three-point calibration objects", which includes the following steps: image point data acquisition; pairwise camera intrinsic and extrinsic parameter calibration; multi-camera intrinsic and extrinsic parameter calibration under a unified reference camera coordinate system; and multi-camera intrinsic and extrinsic parameter calibration under a specified world coordinate system. Summary of the Invention

[0006] To address the above problems, this invention provides a three-dimensional combined camera calibration device and method, the specific technical solution of which is as follows:

[0007] A three-dimensional space combined camera calibration device, characterized in that:

[0008] Includes: rectangular calibration plate,

[0009] Calibration module on calibration board

[0010] and

[0011] A camera system arranged symmetrically in equal arcs around the perimeter of a calibration plate on the same circle.

[0012] A three-dimensional spatial combined camera calibration device according to the present invention is characterized in that:

[0013] The calibration plate is square in shape.

[0014] The camera system is configured for coplanar shooting.

[0015] The calibration module is located in the common field of view of the camera.

[0016] A three-dimensional spatial combined camera calibration device according to the present invention is characterized in that:

[0017] The calibration plate is made of aluminum and has a black coating on its surface.

[0018] A three-dimensional spatial combined camera calibration device according to the present invention is characterized in that:

[0019] The calibration module is cylindrical and is arranged on the calibration plate as follows:

[0020] A calibration module is placed at the intersection of the horizontal and vertical axes of symmetry of the calibration plate.

[0021] Based on this calibration module, two calibration modules are set up symmetrically along the horizontal axis of symmetry, and two sets of calibration modules are set up symmetrically along the vertical axis of symmetry.

[0022] Each of the two sets of calibration modules consists of calibration modules evenly distributed on the circumference of a circle formed by a point on the axis of symmetry at a set distance from the intersection of the two axes of symmetry.

[0023] A three-dimensional spatial combined camera calibration device according to the present invention is characterized in that:

[0024] Each camera system takes pictures according to the set structured light, and the spectrum of the structured light formed by each camera system is different.

[0025] A method for calibrating a combined camera in three-dimensional space, characterized in that:

[0026] The cylindrical marker was placed in the common shooting space area of ​​the multi-camera system.

[0027] Simultaneous imaging of the 3D calibration object using a multi-camera system in a coplanar synchronous manner;

[0028] The captured images are then subjected to sequential laser line extraction, point clustering, ellipse fitting, and matrix matching operations to complete camera calibration.

[0029] According to the present invention, a method for calibrating a combined camera in three-dimensional space is characterized in that:

[0030] The cylindrical calibration object is mounted on a calibration plate;

[0031] The calibration plate is rectangular in shape.

[0032] The multi-camera system is arranged in an equal arc symmetrical pattern around the periphery of the calibration plate, on the circumference of the same circle.

[0033] According to the present invention, a method for calibrating a combined camera in three-dimensional space is characterized in that:

[0034] The calibration objects are arranged on the calibration plate in a manner that allows for spatial differentiation.

[0035] According to the present invention, a method for calibrating a combined camera in three-dimensional space is characterized in that:

[0036] The calibration objects are arranged on the calibration plate in the following manner:

[0037] A calibration module is placed at the intersection of the horizontal and vertical axes of symmetry of the calibration plate.

[0038] Based on this calibration module, two calibration modules are set up symmetrically along the horizontal axis of symmetry, and two sets of calibration modules are set up symmetrically along the vertical axis of symmetry.

[0039] Each of the two sets of calibration modules consists of calibration modules evenly distributed on the circumference of a circle formed by a point on the axis of symmetry at a set distance from the intersection of the two axes of symmetry.

[0040] According to the present invention, a method for calibrating a combined camera in three-dimensional space is characterized in that:

[0041] The calibration plate is made of aluminum and has a black coating on its surface.

[0042] According to the present invention, a method for calibrating a combined camera in three-dimensional space is characterized in that:

[0043] Each camera system takes pictures according to the set structured light, and the spectrum of the structured light formed by each camera system is different.

[0044] According to the present invention, a method for calibrating a combined camera in three-dimensional space is characterized in that:

[0045] The aforementioned point set clustering is specifically as follows:

[0046] S1: Perform Euclidean distance-based clustering algorithm on the point set formed by structured light extracted from laser lines to complete the point set clustering operation and obtain the contour curve accordingly;

[0047] S2: Establish a length threshold according to the set length, calculate the curve length based on calculus for lengths greater than or equal to the threshold, and determine the curve used for ellipse fitting based on the calculation.

[0048] This invention discloses a three-dimensional combined camera calibration device and method. The invention examines and sets corresponding parameters from the perspectives of: the selection of calibration materials; the impact of the calibration material's interface on the captured image; the influence of the calibration material's spatial structure on subsequent calculations; the influence of the temporal accuracy of the captured image on the final calibration accuracy; and the influence of the integrity of the structured light of each camera in a multi-camera system on the final calibration accuracy. These corresponding settings establish their own and mutual technical matching settings from both structural and computational perspectives. Through these settings, a high-precision, highly robust, and efficient camera calibration scheme is ultimately established. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the device structure in this invention;

[0050] Figure 2 This is a front view of the calibration module structure in this invention;

[0051] Figure 3 for Figure 2 Side view;

[0052] Figure 4 This is a schematic diagram of ellipse matching in an embodiment of the present invention;

[0053] Figure 5 This is a schematic diagram of the image processing flow in an embodiment of the present invention. Detailed Implementation

[0054] The following is a further detailed description of a three-dimensional spatial combined camera calibration device and method according to the present invention, based on the accompanying drawings and specific embodiments.

[0055] like Figure 1 , 2 The three-dimensional space combined camera calibration device shown in Figure 3 is a type of device.

[0056] Includes: rectangular calibration plate,

[0057] Calibration module on calibration board

[0058] and

[0059] A camera system arranged symmetrically in equal arcs around the perimeter of a calibration plate on the same circle.

[0060] in,

[0061] The calibration plate is square in shape.

[0062] The camera system is configured for coplanar shooting.

[0063] The calibration module is located in the common field of view of the camera.

[0064] in,

[0065] The calibration plate is made of aluminum and has a black coating on its surface.

[0066] in,

[0067] The calibration module is cylindrical and is arranged on the calibration plate as follows:

[0068] A calibration module is placed at the intersection of the horizontal and vertical axes of symmetry of the calibration plate.

[0069] Based on this calibration module, two calibration modules are set up symmetrically along the horizontal axis of symmetry, and two sets of calibration modules are set up symmetrically along the vertical axis of symmetry.

[0070] Each of the two sets of calibration modules consists of calibration modules evenly distributed on the circumference of a circle formed by a point on the axis of symmetry at a set distance from the intersection of the two axes of symmetry.

[0071] in,

[0072] Each camera system takes pictures according to the set structured light, and the spectrum of the structured light formed by each camera system is different.

[0073] A method for calibrating a combined camera in three-dimensional space.

[0074] The cylindrical marker was placed in the common shooting space area of ​​the multi-camera system.

[0075] Simultaneous imaging of the 3D calibration object using a multi-camera system in a coplanar synchronous manner;

[0076] The captured images are then subjected to sequential laser line extraction, point clustering, ellipse fitting, and matrix matching operations to complete camera calibration.

[0077] in,

[0078] The cylindrical calibration object is mounted on a calibration plate;

[0079] The calibration plate is rectangular in shape.

[0080] The multi-camera system is arranged in an equal arc symmetrical pattern around the periphery of the calibration plate, on the circumference of the same circle.

[0081] in,

[0082] The calibration objects are arranged on the calibration plate in a manner that allows for spatial differentiation.

[0083] in,

[0084] The calibration objects are arranged on the calibration plate in the following manner:

[0085] A calibration module is placed at the intersection of the horizontal and vertical axes of symmetry of the calibration plate.

[0086] Based on this calibration module, two calibration modules are set up symmetrically along the horizontal axis of symmetry, and two sets of calibration modules are set up symmetrically along the vertical axis of symmetry.

[0087] Each of the two sets of calibration modules consists of calibration modules evenly distributed on the circumference of a circle formed by a point on the axis of symmetry at a set distance from the intersection of the two axes of symmetry.

[0088] in,

[0089] The calibration plate is made of aluminum and has a black coating on its surface.

[0090] in,

[0091] Each camera system takes pictures according to the set structured light, and the spectrum of the structured light formed by each camera system is different.

[0092] in,

[0093] The aforementioned point set clustering is specifically as follows:

[0094] S1: Perform Euclidean distance-based clustering algorithm on the point set formed by structured light extracted from laser lines to complete the point set clustering operation and obtain the contour curve accordingly;

[0095] S2: Establish a length threshold according to the set length, calculate the curve length based on calculus for lengths greater than or equal to the threshold, and determine the curve used for ellipse fitting based on the calculation.

[0096] Working principle, working process and implementation examples

[0097] like Figure 1 As shown, the calibration module is placed within the common space of the combined camera's field of view. Each camera and its associated structured light imaging system captures images of the calibration module, obtaining images containing structured light information. This example uses four camera systems, but six or other systems are also possible. Each imaging system captures images simultaneously to save time and ensure high precision. Figure 2 As shown, first draw three sets of circles according to the horizontal and vertical axes of symmetry, dividing the space into symmetrical arrangements in all directions; on each circle, evenly distribute some small cylinders of the same diameter, i.e. Figure 2 The black cylinders shown are not limited in number, as long as they are evenly distributed on the circle. A pair is symmetrically arranged about the intersection of the horizontal axis of symmetry, and a group is symmetrically arranged along the vertical axis of symmetry. The calibration plate used to set the calibration module is made of lightweight aluminum, and is painted black to prevent laser line reflection. To improve efficiency and achieve one-time image calibration, the structured light is set to have a distinguishable spectrum to prevent interference between structured light beams. Correspondingly, the camera only accepts input with the corresponding spectrum, such as... Figure 1 As shown.

[0098] Image acquisition completed by the calibration module with this structure is used in subsequent image processing. Figure 5 The processing flow is as follows: First, a structured light image is acquired, and the skeleton contour point set of the structured light is extracted. The discrete point set is then clustered using an Euclidean distance clustering algorithm to obtain partial contour curves. The curve length is calculated by counting points, and curves with lengths below a threshold are deleted. The remaining contour curves are used to calculate the best-fit ellipse. The calculated ellipse is then matched with a set standard model, such as... Figure 4 As shown, the external calibration parameters of each camera are calculated. The calibration module is structured such that a single calibration module is placed at the intersection of the horizontal and vertical axes of symmetry of the calibration plate. Based on this module, two additional calibration modules are symmetrically arranged along the horizontal axis of symmetry, and two sets of calibration modules are symmetrically arranged along the vertical axis of symmetry. Each set of calibration modules consists of uniformly spaced calibration modules arranged on the circumference of a circle centered at a point a predetermined distance from the intersection of the two axes of symmetry. This arrangement allows for spatial differentiation of the calibration objects on the calibration plate, providing spatial markers for image calculations and reducing computational complexity. Furthermore, the symmetrical arrangement in both the horizontal and vertical directions ensures spatial robustness in practical applications. (See also...) Figure 4Point E is relatively outlier and far from the group points, as shown in Figures A, B, C, and D. During imaging, point E can be quickly determined by calculating its distance from the center of the circle, thus creating a spatial distinction and improving the accuracy of the center position matching. This reduction in computational load is specifically reflected in the fact that to calculate the rigid transformation matrix, at least n (n≥2) contours need to be arbitrarily selected from the contours to calculate the correspondence matrix with the cylinders on the template. The required number of calculations is C(N,n)×P(M,n), where N is the number of contours in the acquired image, M is the number of cylinders in the calibration module, and n is a number less than N. However, the number of valid correspondence pairs is very small, with a probability of D=C(N,n) / (C(N,n)×P(M,n)). To obtain a high-probability T reliable pairing relationship, at least k attempts are needed, k=log(1-T) / log(1-D). This requires a lot of computation; however, in this embodiment, two calibration modules set on the left and right sides of the horizontal axis of symmetry provide effective spatial indication, thereby reducing the amount of computation. Figure 5 The calculations following ellipse fitting employ conventional matrix matching operations, involving transformations between the world coordinate system, camera coordinate system, and image coordinate system typically used in camera calibration.

[0099] This invention discloses a three-dimensional combined camera calibration device and method. The invention examines and sets corresponding parameters from the perspectives of: the selection of calibration materials; the impact of the calibration material's interface on the captured image; the influence of the calibration material's spatial structure on subsequent calculations; the influence of the temporal accuracy of the captured image on the final calibration accuracy; and the influence of the integrity of the structured light of each camera in a multi-camera system on the final calibration accuracy. These corresponding settings establish their own and mutual technical matching settings from both structural and computational perspectives. Through these settings, a high-precision, highly robust, and efficient camera calibration scheme is ultimately established.

Claims

1. A three-dimensional space combined camera calibration device, characterized in that: comprising a rectangular calibration board, a calibration module arranged on the calibration board, and a camera system arranged on the circumference of the same circle around the periphery of the calibration board in equiangular symmetry. The calibration board is arranged in a square shape. The camera system is arranged in a coplanar shooting mode. The calibration module is arranged in the common field of view space of the cameras. The calibration board is made of aluminum material and has a black coating on the surface. The calibration module is arranged on the calibration board in the following manner: a calibration module is arranged at the intersection of the horizontal and vertical symmetry axes of the calibration board, and two calibration modules are symmetrically arranged along the horizontal symmetry axis and two groups of calibration modules are symmetrically arranged along the vertical symmetry axis based on the calibration module. Each group of the two groups of calibration modules is composed of calibration modules arranged on the circumference of a circle with a center at a point on the symmetry axis at a distance from the intersection of the two symmetry axes and a radius.

2. The three-dimensional space combined camera calibration device according to claim 1, characterized in that: each camera system shoots according to a set of structured light, and the structured light formed by each camera system has different spectra.

3. A three-dimensional space combined camera calibration method using the three-dimensional space combined camera calibration device according to claim 1, characterized in that: the cylindrical calibration module is placed in the common shooting space of the multi-camera system, the three-dimensional calibration module is simultaneously shot by the multi-camera system in a coplanar and synchronous manner, and the camera calibration is completed by sequentially performing laser line extraction, point clustering, ellipse fitting, and matrix matching operations on the shot images. The cylindrical calibration module is arranged on a calibration board. The multi-camera system is arranged on the circumference of the same circle around the periphery of the calibration board in equiangular symmetry. The calibration modules are arranged on the calibration board in a manner of spatial position division. A calibration module is arranged at the intersection of the horizontal and vertical symmetry axes of the calibration board, and two calibration modules are symmetrically arranged along the horizontal symmetry axis and two groups of calibration modules are symmetrically arranged along the vertical symmetry axis based on the calibration module. Each group of the two groups of calibration modules is composed of calibration modules arranged on the circumference of a circle with a center at a point on the symmetry axis at a distance from the intersection of the two symmetry axes and a radius.

4. The three-dimensional space combined camera calibration method according to claim 3, characterized in that: the calibration board is made of aluminum material and has a black coating on the surface.

5. The three-dimensional space combined camera calibration method according to claim 3, characterized in that: each camera system shoots according to a set of structured light, and the structured light formed by each camera system has different spectra.

6. The three-dimensional space combined camera calibration method according to claim 3, characterized in that: the point clustering is specifically: ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ S1: the point set formed by the structured light based on the laser line extraction is subjected to the clustering algorithm operation based on the Euclidean distance, so as to complete the point set clustering operation, and the contour curve is obtained accordingly; S2: a length threshold is established according to the set length, the length greater than or equal to the threshold is subjected to the curve length calculation based on the calculus, and the curve used for the ellipse fitting is determined according to the calculation.

Images