Method for determining calibration precision of binocular deflection system, medium and equipment

Abstract

The invention relates to a binocular deflection system calibration precision determination method, a medium and equipment, and the method comprises the steps: setting a plurality of coding patterns in the overlapping sight range of each collection device, and converting the coordinates of any angular point of the coding patterns into a same coordinate system, so as to calculate the distance error or / and the relative error between the coordinates in the same coordinate system, thereby achieving the calibration precision of a binocular deflection system. Determining a dual-target calibration error, acquiring a coordinate and a normal direction of a center point of the encoding pattern under any acquisition equipment coordinate system as a first three-dimensional parameter, then displaying a set stripe pattern to acquire a deformed stripe pattern reflected by the surface of the encoding pattern, and reconstructing the deformed stripe pattern to obtain a first three-dimensional parameter; the coordinates and the normal direction of the center point of the coding pattern under the coordinate system of the equipment are obtained and serve as second three-dimensional parameters, so that the deflection calibration error is determined, and the calibration precision of the binocular deflection system is determined according to the binocular calibration error and the deflection calibration error. The problem that the calibration result cannot be evaluated from the global perspective in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of deflection measurement technology, and in particular to a method, medium and device for determining the calibration accuracy of a binocular deflection system. Background Art

[0002] Phase measurement deflectometry is mainly used for reconstruction and measurement of high-brightness surfaces, such as car paint, mobile phone screens, etc. The binocular deflectometry system consists of a set of binocular acquisition devices and a screen. The relative positions of the binocular acquisition devices and the screen are fixed, and the screen is used to project the stripe pattern.

[0003] Phase measurement deflectometry uses the law of light reflection. The screen projects a fringe pattern onto a bright surface. After mirror reflection, the acquisition device captures the virtual image of the fringe pattern projected on the screen. The acquired fringe pattern is then decoded, reconstructed, integrated, and other operations are performed to restore the three-dimensional surface shape of the bright object being measured.

[0004] The accurate calibration of system parameters has a decisive influence on the accuracy of surface measurement. Therefore, after the calibration is completed, the calibration results need to be evaluated to verify their accuracy.

[0005] After the deflectometry system is calibrated, the existing method generally evaluates the calibration results through reprojection error. The reprojection evaluation method refers to projecting the three-dimensional feature points onto the pixel plane according to the calibration results. It is mainly used to verify the internal parameters and distortion information in the calibration results of the acquisition equipment. It cannot evaluate the quality of the calibration results from a global perspective.

[0006] Therefore, how to improve it is a technical problem that needs to be solved urgently in this field. Summary of the invention

[0007] Based on this, the purpose of the present application is to provide a method, medium and device for determining the calibration accuracy of a binocular deflection system to solve at least one of the technical problems mentioned in the above background technology.

[0008] In a first aspect, the present application provides a method for determining the calibration accuracy of a binocular deflection system, comprising:

[0009] Setting a number of coded patterns within the overlapping line of sight of each acquisition device;

[0010] Obtain the coordinates of any corner point of the coding pattern in the coordinate system of each acquisition device, and convert the coordinates of the corner point to the same coordinate system to calculate the distance error and / or relative error between the coordinates in the same coordinate system to determine the binocular positioning error;

[0011] Taking any acquisition device as the main device, the coordinates and normal direction of the center point of the coding pattern in the main device coordinate system are obtained as the first three-dimensional parameters;

[0012] Displaying a set stripe pattern, reflecting the coding pattern surface to collect a deformed stripe pattern reflected by the coding pattern surface, and reconstructing the deformed stripe pattern to obtain the coordinates and normal of the center point of the coding pattern in the main device coordinate system, which are the second three-dimensional parameters;

[0013] determining a deflection calibration error based on the first three-dimensional parameter and the second three-dimensional parameter;

[0014] The calibration accuracy of the binocular deflection system is determined based on the binocular calibration error and the deflection calibration error.

[0015] Furthermore, the step of calculating the distance error also includes:

[0016] Traverse the corner points of the coded pattern and obtain the coordinates of the selected corner points in the coordinate system of each acquisition device;

[0017] The coordinates of the selected corner points in each acquisition device coordinate system are converted to the coordinate systems of other acquisition devices, and the average value of the distance errors between the coordinates of the selected corner points in each acquisition device coordinate system is calculated as the final distance error.

[0018] Furthermore, the step of calculating the relative error also includes:

[0019] The basic matrix is ​​obtained according to the prior calibration external parameters and the prior acquisition device internal parameters; the prior calibration external parameters include the rotation transformation matrix and the translation transformation matrix between the acquisition devices; the prior acquisition device internal parameters include any one or more of the focal length, the principal point, the distortion coefficient and the acquisition device matrix;

[0020] Obtaining the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device;

[0021] Traverse the coordinates of the corner points in the coordinate system of each acquisition device, calculate the straight line where the coordinates of each corner point are located in the image coordinate system of another acquisition device according to the basic matrix, and calculate the distance value between the coordinates of each corner point and the corresponding straight line in the same coordinate system, and obtain the average value of all distance values, which is the relative error.

[0022] Further, the step of determining the binocular positioning error according to the distance error and / or the relative error includes:

[0023] According to the size of distance error and relative error, the larger error is taken as the error of binocular calibration;

[0024] Or, according to the distance error and the relative error, the final error value is calculated according to the weight, which is the binocular positioning error.

[0025] Furthermore, the step of obtaining the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device includes:

[0026] Acquire a 2D image of the coding pattern within the acquisition range of each acquisition device;

[0027] Establish a plane coordinate system with any corner point of the 2D image as the origin, and obtain the coordinates of each corner point in the plane coordinate system;

[0028] The rotation and translation matrix between the plane coordinate system and each acquisition device is obtained, so as to transform the coordinates of each corner point in the plane coordinate system to the coordinate system of each acquisition device according to the rotation and translation matrix.

[0029] Furthermore, the step of obtaining the first three-dimensional parameter includes:

[0030] A plane coordinate system is established with any point of the coding pattern as the origin; preferably, the plane coordinate system is established with the center point of the coding pattern as the origin.

[0031] According to the coordinates of the corner points of each coding pattern in the plane coordinate system and the main device coordinate system, a rotation matrix and a translation matrix between the plane coordinate system and the main device coordinate system are obtained;

[0032] Fitting the plane equation according to the coordinates of the corner points of the coding pattern to obtain the normal direction of the center point of the coding pattern;

[0033] The coordinates and normal of the center point of the coding pattern in the plane coordinate system are converted to the main device coordinate system according to the rotation matrix and the translation matrix, so as to obtain the coordinates and normal of the center point of the coding pattern in the main device coordinate system, which are the first three-dimensional parameters.

[0034] Further, the step of determining the deflection calibration error according to the first three-dimensional parameter and the second three-dimensional parameter includes:

[0035] Get the distance between the coordinates of each center point in the main device coordinate system, which is the coordinate distance difference;

[0036] Get the angle between the normals of each center point in the main device coordinate system, which is the angle distance difference.

[0037] Furthermore, the step of determining the calibration accuracy of the binocular deflection system according to the binocular calibration error and the deflection calibration error includes:

[0038] Obtaining the ratios of the binocular positioning error and the deflection calibration error to the corresponding error thresholds, and determining the binocular positioning accuracy and the deflection calibration accuracy;

[0039] According to the size of binocular positioning accuracy and deflection calibration accuracy, the larger accuracy is taken as the binocular deflection system calibration accuracy.

[0040] In a second aspect, the present application further provides a computer storage medium storing an executable program code; the executable program code is used to execute the method for determining the calibration accuracy of the binocular deflection system described in any one of the first aspects.

[0041] In a third aspect, the present application also provides a terminal device comprising a memory and a processor; the memory stores a program code that can be executed by the processor; the program code is used to execute the method for determining the calibration accuracy of the binocular deflection system as described in any one of the first aspects.

[0042] The present invention provides a method, medium and device for determining the calibration accuracy of a binocular deflection system. A plurality of coded patterns are set within the overlapping sight lines of each acquisition device, the coordinates of any corner point of the coded pattern in the coordinate system of each acquisition device are obtained, and the coordinates of the corner point are converted to the same coordinate system to calculate the distance error or / and relative error between the coordinates in the same coordinate system, and the binocular calibration error is determined. The binocular calibration error is determined by obtaining the error of the three-dimensional coordinates of the corner points of the coded pattern in each acquisition device, so that the calibration result can be evaluated from a global perspective. Then, any acquisition device is used as a main device, the coordinates and normal of the center point of the coded pattern in the coordinate system of the main device are obtained as the first three-dimensional parameters, and then a set stripe pattern is displayed, and the deformed stripe pattern reflected by the surface of the coded pattern is collected, and the deformed stripe pattern is reconstructed to obtain the coordinates and normal of the center point of the coded pattern in the coordinate system of the main device, which are the second three-dimensional parameters. The deflection calibration error can be determined according to the first three-dimensional parameters and the second three-dimensional parameters. Finally, the calibration accuracy of the binocular deflection system is determined according to the binocular calibration error and the deflection calibration error. This solves the problem that the existing technology cannot evaluate the quality of the calibration results from a global perspective. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of a method for determining calibration accuracy of a binocular deflection system according to an embodiment of the present invention;

[0044] Figure 2 FIG. 4 is a schematic diagram of a dual-target positioning system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] It should be noted that if the embodiments of the present invention involve directional indications, such as up, down, left, right, front, back, etc., then the directional indication is only used to explain the relative position relationship, movement status, etc. between the components in a certain specific posture. If the specific posture changes, the directional indication will also change accordingly. In addition, if the embodiments of the present invention involve descriptions of "first, second", "S1, S2", "step one, step two", etc., then such descriptions are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of indicated technical features or indicating the execution order of the method, etc. Those skilled in the art can understand that anything that does not violate the gist of the invention under the technical concept of the invention should be included in the protection scope of the present invention.

[0047] like Figure 1 As shown, the present invention provides a method for determining the calibration accuracy of a binocular deflection system:

[0048] S1: Set a number of coding patterns within the overlapping line of sight of each acquisition device;

[0049] Specifically, it is optional but not limited to obtaining the sight range of all acquisition devices, that is, the acquisition range, to set several coding patterns in the overlapping area of ​​each acquisition range; the acquisition device may include digital acquisition devices, cameras, mobile phones and other acquisition devices that can be used to capture static images or continuous videos. More specifically, the coding pattern includes but is not limited to structured light coding patterns that can be used for pixel-by-pixel coding, such as QR codes, stepping stripe patterns, Gray code stripe patterns, sinusoidal stripe patterns, binary stripe patterns, and phase-shifted stripe patterns.

[0050] Preferably, the material carrying the QR code preferably has reflective properties and can clearly reflect the patterns projected by other light sources, and the size of the coding pattern is known.

[0051] S2: Obtain the coordinates of any corner point of the coding pattern in the coordinate system of each acquisition device, and convert the coordinates of the corner point to the same coordinate system to calculate the distance error between the coordinates in the same coordinate system and determine the binocular positioning error;

[0052] Specifically, it is optional but not limited to obtaining a 2D image of the coding pattern through an acquisition device, and then selecting any corner point in the coding pattern as the current corner point, and optionally obtaining the coordinates of the current corner point in the coordinate system of each acquisition device according to a computer vision algorithm, and then according to the prior conversion matrix between the coordinate systems of each acquisition device, taking a certain acquisition device as the main acquisition device, converting the coordinates of the current corner point in the coordinate system of other acquisition devices to the main coordinate system, and converting the corner point coordinates to the same coordinate system, and obtaining the distance error between the coordinates of each current corner point, which is the error of binocular positioning. Specifically, the computer vision algorithm may optionally include common algorithms such as a QR code corner point recognition algorithm, a PnP algorithm, a feature point detection and description algorithm, and a three-dimensional reconstruction algorithm; the prior conversion matrix may be optionally, but not limited to, pre-calibrated by a technician in this field or directly obtained from engineering documents / technical manuals.

[0053] For example, Figure 2 As shown, it is optional but not limited to provide a binocular positioning system, including a first acquisition device 1, a second acquisition device 2 and a coding pattern 3, two straight lines diffracted from the acquisition device represent the boundaries of the acquisition range of the corresponding acquisition device, and the coding pattern 3 is set in the common acquisition range of the first acquisition device 1 and the second acquisition device 2, so that each acquisition device respectively acquires a 2D image of the coding pattern, and then preferably uses a two-dimensional code corner point recognition algorithm and a PnP algorithm to calculate the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device, and any corner point is regarded as the current corner point, and then the coordinates of the corner point in the coordinate system of the first acquisition device are converted to the coordinate system of the second acquisition device, and the coordinates of the corner point in the coordinate system of the second acquisition device can also be converted to the coordinate system of the first acquisition device, and finally the distance error err between the two points in the same coordinate system is calculated, so as to determine the binocular positioning accuracy according to the specific value of the distance error err. Preferably, the specific steps for calculating the distance error err can be selected as shown in formulas 2-1, 2-2, and 2-3:

[0054] P l =[X l Y l Z l ] T 2-1

[0055] P r =[X r Y r Z r ] T 2-2

[0056] err=RP l +tP r 2-3

[0057] Among them, P l and (X l, Y l , Z l ) is the corner point coordinate in the coordinate system of the first acquisition device, P r and (X r , Y r , Z r ) is the corner point coordinate in the coordinate system of the second acquisition device, T is the transpose operator, which is used to convert the row vector (X l , Y l , Z l ) and (X r , Y r , Z r ) is converted to a column vector P l and P r , R is the rotation transformation matrix between each acquisition device, t is the translation transformation matrix between each acquisition device, and err is the distance value between two points after the corner point coordinates in the first acquisition device coordinate system are converted to the second acquisition device coordinate system.

[0058] More specifically, since the differences between the current corner points in other coordinate systems are ignored and the errors of other corner points of the coding pattern are not taken into account, the calibration accuracy may be reduced. Therefore, for a multi-acquisition device, step S2 is preferably:

[0059] S21: traverse the corner points of the coding pattern to obtain the coordinates of the selected corner points in the coordinate system of each acquisition device;

[0060] S22: converting the coordinates of the selected corner points in the coordinate system of each acquisition device to the coordinate system of another acquisition device, and calculating the average value of the distance errors between the coordinates of the selected corner points in the coordinate system of each acquisition device, which is the binocular positioning error;

[0061] Specifically, it is optional but not limited to obtaining a 2D image of the coded pattern through an acquisition device, then traversing all corner points in the coded pattern, taking the current corner point as the selected corner point, and optionally obtaining the coordinates of the selected corner point in each acquisition device coordinate system according to a computer vision algorithm, and then converting the coordinates of the selected corner point in each acquisition device coordinate system to other coordinate systems according to a priori transformation matrices between the coordinate systems of each acquisition device, and obtaining the average value of the distance error between the coordinates of each selected corner point in all acquisition device coordinate systems, which is the error of binocular positioning.

[0062] In this step, the error of binocular positioning is determined mainly by calculating the distance between the collected corner point coordinates and the calculated corner point coordinates. However, this step still has many problems. For example, in the coordinate conversion process, errors may be introduced due to multiple transformations. Especially when using matrix operations, the loss of numerical precision may accumulate, resulting in inaccurate final results. In addition, the acquisition accuracy of the acquisition equipment is affected by environmental factors, resulting in errors. Therefore, in order to improve the accuracy of determining the binocular positioning accuracy, step S2 preferably also includes:

[0063] S23: Obtaining a basic matrix according to a priori calibration external parameters and a priori acquisition device internal parameters; the a priori calibration external parameters include the rotation transformation matrix R and the translation transformation matrix t between the acquisition devices; the a priori acquisition device internal parameters K include any one or more of the focal length, the principal point, the distortion coefficient and the acquisition device matrix;

[0064] Specifically, taking the dual-target calibration system provided in step S2 as an example, it is optional but not limited to substituting the prior calibration external parameters and the acquisition device internal parameters into equation 2-4 to calculate the basic matrix:

[0065]

[0066] Among them, F is the basic matrix, K r is the internal parameter of the second acquisition device, t^ is the antisymmetric matrix of the translation transformation matrix between the acquisition devices, R is the rotation transformation matrix between the acquisition devices, K l is the internal parameter of the first acquisition device, (t x , t y , t z ) are the components of the translation transformation matrix t in the x, y, and z directions respectively.

[0067] S24: Obtaining the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device;

[0068] Specifically, taking the binocular positioning system provided in step S2 as an example, it is optional but not limited to collecting 2D images of the coding pattern through each acquisition device, and then extracting corner points of the 2D image according to a computer vision algorithm to obtain the corner point coordinates of the coding pattern in the coordinate system of each acquisition device.

[0069] Preferably, the step of obtaining the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device may optionally include but is not limited to:

[0070] S241: Acquire a 2D image of a coding pattern within a collection range of each collection device;

[0071] S242: Establishing a plane coordinate system with any corner point of the 2D image as the origin, and obtaining the coordinates of each corner point in the plane coordinate system;

[0072] S243: Obtaining a rotation and translation matrix between the plane coordinate system and each acquisition device, so as to transform the coordinates of each corner point in the plane coordinate system into the coordinate system of each acquisition device according to the rotation and translation matrix;

[0073] Specifically, it is optional but not limited to taking any corner point of the 2D image of the coding pattern as the origin, and the technical personnel in this field arbitrarily set the x-axis and y-axis according to the specific shape of the coding pattern to establish a plane coordinate system, and then preferably obtain the coordinates of each corner point of the coding pattern in the plane coordinate system according to the corner point detection algorithm, and then calculate the rotation and translation matrix between the plane coordinate system and the coordinate system of each acquisition device according to the computer vision algorithm, so as to convert the coordinates of each corner point in the plane coordinate system to the coordinate system of each acquisition device according to the rotation and translation matrix; the computer vision algorithm may optionally include commonly used algorithms such as the PnP algorithm, the least squares method, and the random sampling consistency algorithm.

[0074] More specifically, during the acquisition process, the acquisition device may be affected by various factors, such as a wide-angle lens causing points near the center of the image to be stretched or compressed, or an imperfect alignment between the lens and the image sensor causing the image to be tilted or offset, etc., so that the coordinates of each point in the acquired image are inconsistent with the actual position. Therefore, after obtaining the coded pattern 2D image within the acquisition range of each acquisition device in step S241, it is preferred that the following further be included:

[0075] S241': Dedistort the coded pattern 2D image to obtain corrected corner point coordinates;

[0076] Specifically, it is optional but not limited to using an image dedistortion method to dedistort the 2D image of the encoded image based on parameters such as focal length, principal point and distortion coefficient in the prior acquisition device internal parameters to restore the true geometric shape of the 2D image, thereby obtaining the corrected corner point coordinates; the image dedistortion method may include common algorithms such as OpenCV dedistortion function, remapping method, deep learning method and geometric correction method.

[0077] S25: traverse the coordinates of the corner points in the coordinate system of each acquisition device, calculate the straight line where the coordinates of each corner point are located in the image coordinate system of another acquisition device according to the basic matrix, and calculate the distance value between the coordinates of each corner point and the corresponding straight line in the same coordinate system, and obtain the average value of all distance values, which is the relative error;

[0078] Specifically, taking the binocular positioning system provided in step S2 as an example, assume a corner point P, whose coordinates in the coordinate system of the first acquisition device are p l =[u l , v l ] T , whose coordinates in the coordinate system of the second acquisition device are p r =[u r, v r ] T , assuming that the equation of the straight line is ax+by+c=0, the corner point coordinates p in the coordinate system of the first acquisition device can be calculated according to equations 2-8 and 2-9. l The equation of the line in the coordinate system of the second acquisition device is l r And the corner point coordinates p in the coordinate system of the second acquisition device r With straight line l r The distance d:

[0079] l r =[abc] T =Fp l

[0080]

[0081] Taking the above steps as an example, by traversing the coordinates of all corner points, we can get the distance between all corner points in the same coordinate system and the straight line where the corresponding corner points in another coordinate system are located in this coordinate system, and then calculate the average value of these distances according to formula 2-10, which is the relative error d err :

[0082]

[0083] S26: Determine a binocular positioning error according to the distance error and the relative error.

[0084] Specifically, it is optional but not limited to selecting one of the distance error and the relative error to represent the binocular positioning error based on the size of the distance error and the relative error, or assigning different weights to the distance error and the relative error based on their accuracy, and taking the value after the weighted combination of the two as the binocular positioning error.

[0085] Preferably, step S26 may optionally include but is not limited to:

[0086] S261: according to the size of the distance error and the relative error, the larger error is used as the error of the binary target positioning; for example, it is determined whether the distance error is greater than the relative error, if so, the distance error is used as the error of the binary target positioning, if not, the relative error is used as the error of the binary target positioning.

[0087] Specifically, the larger error between the distance error and the relative error can be selected to represent the binocular positioning error, thereby ensuring that the binocular positioning result is acceptable under the most unfavorable external environmental influence, thereby improving the reliability and robustness of the system by providing a more conservative and comprehensive evaluation method.

[0088] Further preferably, since the precisions of the distance error and the relative error are different, their importance to determining the binocular positioning error is also different. Therefore, step S26 may optionally include but is not limited to:

[0089] S261': Calculate the final error value according to the distance error and the relative error, which is the binocular positioning error.

[0090] Preferably, the final error value=distance error×first weight+relative error×second weight.

[0091] Specifically, since the precision of the distance error and the relative error are different in the process of determining the binocular positioning error, their importance to determining the binocular positioning error is also different. Therefore, their respective weights are also different. In order to improve the precision of determining the binocular positioning error, weights can be assigned to the distance error and the relative error respectively, and then the final binocular positioning error is calculated according to their respective weights.

[0092] S3: Taking any acquisition device as the main device, obtaining the coordinates and normal direction of the center point of the coding pattern in the coordinate system of the main device as the first three-dimensional parameters;

[0093] Specifically, it is optional but not limited to establishing a plane coordinate system with any point of the coding pattern as the origin, and then optionally selecting any acquisition device as the main device, converting the coordinates and normal of the center point of the coding pattern in the plane coordinate system to the main device coordinate system, and obtaining the coordinates and normal of the center point in the main device coordinate system as the first three-dimensional parameters.

[0094] Preferably, the step of obtaining the coordinates and normal of the center point of the coding pattern in the main device coordinate system may optionally include but is not limited to:

[0095] S31: Establishing a plane coordinate system with any point of the coding pattern as the origin; preferably establishing a plane coordinate system with the center point of the coding pattern as the origin;

[0096] S32: Obtain a rotation matrix and a translation matrix between the plane coordinate system and the main device coordinate system according to the coordinates of each corner point of the coding pattern in the plane coordinate system and the main device coordinate system;

[0097] Specifically, it is optional but not limited to obtaining the coordinates of each corner point of the coding pattern in the plane coordinate system based on the known size of the coding pattern in step S1, and based on the coordinates of each corner point of the coding pattern in the plane coordinate system and the main device coordinate system, preferably using the PnP algorithm to calculate the rotation matrix and translation matrix between the plane coordinate system and the main device coordinate system.

[0098] S33: fitting a plane equation according to the coordinates of the corner points of the coding pattern to obtain the normal direction of the center point of the coding pattern;

[0099] Specifically, it is optional but not limited to setting the plane model according to the plane features and the equations that need to be satisfied, and obtaining the plane normal vector, that is, the normal of the center point of the coding pattern. For example, the plane model is set with the plane normal vector and the constant term as parameters; more specifically, the plane model parameters can be optionally but not limited to be defined as: (A, B, C, D), where A, B, C are the components of the plane normal vector, D is the constant term of the plane equation, and the plane equation can be optionally but not limited to be expressed as: Ax+By+Cz+D=0; it is worth noting that the equations that need to be satisfied by the plane model and the parameters used are examples and are not limited to this.

[0100] Preferably, it is optional but not limited to constructing a homogeneous equation using the three-dimensional coordinates of each corner point, and performing singular value decomposition on the matrix on the left side of the equation to obtain the equation coefficients of the current plane equation.

[0101] For example, suppose there are four corner points P1, P2, P3, P4, where P1 is (X1, Y1, Z1), and so on, the matrix M 3-1 can be constructed:

[0102]

[0103] Then perform singular value decomposition on the matrix M according to formula 3-2:

[0104] M=UΣV T 3-2

[0105] Among them, U and V are orthogonal matrices, and Σ is a diagonal matrix.

[0106] Then, through SVD, we can get the solution of v. Usually, the last column of V corresponds to the minimum singular value, which represents the solution of the equation, and we can get the coefficients of the plane equation: A, B, C, D.

[0107] S34: Convert the coordinates and normal of the center point of the coding pattern in the plane coordinate system to the main device coordinate system according to the rotation matrix and the translation matrix, and obtain the coordinates and normal of the center point of the coding pattern in the main device coordinate system, which are the first three-dimensional parameters.

[0108] Specifically, it is optional but not limited to converting the coordinates of the center point of the encoded pattern in the plane coordinate system to the main device coordinate system according to the rotation matrix and the translation matrix, and converting the normal of the center point of the encoded pattern in the plane coordinate system to the main device coordinate system according to the rotation matrix, to obtain the coordinates and normal of the center point of the encoded pattern in the main device coordinate system, which are the first three-dimensional parameters.

[0109] S4: Displaying a set stripe pattern, reflecting the coding pattern surface to collect a deformed stripe pattern reflected by the coding pattern surface, and reconstructing the deformed stripe pattern to obtain the coordinates and normal direction of the center point of the coding pattern in the main device coordinate system, which are the second three-dimensional parameters;

[0110] Specifically, it is optional but not limited to displaying a set stripe pattern on the screen of a display device, mapping the stripe pattern to the surface of the coding pattern, and then obtaining the deformed stripe pattern reflected by the surface of the coding pattern through an acquisition device, preferably using a multi-frequency heterodyne method to decode the stripe pattern to obtain an absolute phase map, and reconstructing it using a stereo deflection method, so as to obtain the coordinates and normal of the center point of the coding pattern in the coordinate system of the main device, which is the second three-dimensional parameter; the display device, which may optionally include a liquid crystal display (LCD), an organic light emitting diode display (OLED), and other display devices that can display images and videos on the screen; the set stripe pattern includes several stripe patterns with different frequencies and quantities.

[0111] S5: determining a deflection calibration error according to the first three-dimensional parameter and the second three-dimensional parameter;

[0112] Specifically, it is optional but not limited to obtaining a coordinate distance difference and a normal angle difference between the first three-dimensional parameter and the second three-dimensional parameter, and using the coordinate distance difference and the normal angle difference to represent the deflection calibration error.

[0113] Preferably, the step of determining the deflection calibration error according to the first three-dimensional parameter and the second three-dimensional parameter may optionally include:

[0114] S51: Obtain the distance between the coordinates of each center point in the main device coordinate system, which is the coordinate distance difference;

[0115] Taking the determination of the deflection calibration error according to the coordinate distance difference and the normal angle difference as an example, assuming that the first three-dimensional parameters include coordinates P1 (3, 4, 5) and normal N1 (0, 0, 1), and the second three-dimensional parameters include coordinates P2 (2, 5, 6) and normal N2 (0.1, 0.1, 0.99), the coordinate distance difference d between coordinates P1 and P2 can be calculated according to formula 5-1:

[0116]

[0117] S52: Obtain the angle between the normal directions of each center point in the main device coordinate system, which is the angle distance difference.

[0118] Specifically, the angle between the normal N1 and the normal N2 may be obtained according to the angle difference calculation formula, which is the normal angle difference θ. The angle difference calculation formula may be expressed as, but is not limited to, formula 5-2:

[0119]

[0120] A=N1×N2 5-3

[0121] Among them, θ is the angle between normal N1 and normal N2, |N1|, |N2| are the moduli of normal N1 and normal N2 respectively, and A is the dot product between normal N1 and normal N2.

[0122] The distance difference and angle difference obtained by combining the above steps are the deflection calibration error.

[0123] S6: Determine the calibration accuracy of the binocular deflection system according to the binocular calibration error and the deflection calibration error.

[0124] Specifically, it is optional but not limited to setting a binocular positioning error threshold and a deflection calibration error threshold, so as to obtain the binocular positioning accuracy and the deflection calibration accuracy according to the ratio between the binocular positioning error and the binocular positioning error threshold and the ratio between the deflection calibration error and the deflection calibration error threshold, so as to select one of the binocular positioning accuracy and the deflection calibration accuracy to represent the binocular deflection system calibration accuracy, or assign different weights to the binocular positioning accuracy and the deflection calibration accuracy, and the value after the weighted combination of the two is the binocular deflection system calibration accuracy.

[0125] Preferably, the step of determining the calibration accuracy of the binocular deflection system according to the binocular calibration error and the deflection calibration error comprises:

[0126] S61: Obtain the ratios of the binocular positioning error and the deflection calibration error to the corresponding error thresholds to determine the binocular positioning accuracy and the deflection calibration accuracy;

[0127] S62: according to the binocular positioning accuracy and the deflection calibration accuracy, the accuracy with the larger absolute value is the binocular deflection system calibration accuracy;

[0128] For example, assuming that the binocular positioning error is 5, the deflection calibration error is 3, the binocular positioning error threshold is 4, and the deflection calibration error threshold is 6, we can obtain binocular positioning accuracy = 5 / 4 = 1.25, deflection calibration accuracy = 3 / 6 = 0.5. Since the absolute value of the binocular positioning accuracy is greater than the deflection calibration accuracy, the binocular positioning accuracy is taken as the binocular deflection system calibration accuracy.

[0129] Further preferably, step S62 may be:

[0130] S62': Calculate the final calibration accuracy according to the binocular calibration accuracy and the deflection calibration accuracy, which is the binocular deflection system calibration accuracy.

[0131] Preferably, final calibration accuracy=binocular calibration accuracy×binocular weight+deflection calibration accuracy×deflection weight.

[0132] Specifically, since the binocular positioning accuracy and the deflection calibration accuracy have different accuracies in the process of determining the binocular deflection system calibration accuracy, their importance to determining the binocular deflection system calibration accuracy is also different. Therefore, their respective weights are also different. In order to improve the accuracy of determining the binocular deflection system calibration accuracy, it is possible to assign weights to the binocular positioning accuracy and the deflection calibration accuracy respectively, and then calculate the final binocular deflection system calibration accuracy according to their respective weights.

[0133] In this embodiment, a rigid body calibration method of a multi-scanning device under a large viewing angle of the present invention is provided. By setting a number of coded patterns within the overlapping line of sight of each acquisition device, the coordinates of any corner point of the coded pattern in the coordinate system of each acquisition device are obtained, and the coordinates of the corner point are converted to the same coordinate system to calculate the distance error or / and relative error between the coordinates in the same coordinate system, and determine the binocular positioning error. By obtaining the error of the three-dimensional coordinates of the corner points of the coded pattern in each acquisition device, the binocular positioning error is determined, and the calibration result can be evaluated from a global perspective. Then, any acquisition device is used as the main device, and the coordinates and normal of the center point of the coded pattern in the coordinate system of the main device are obtained as the first three-dimensional parameters. Then, the set stripe pattern is displayed, and the deformed stripe pattern reflected by the surface of the coded pattern is collected, and the deformed stripe pattern is reconstructed to obtain the coordinates and normal of the center point of the coded pattern in the coordinate system of the main device, which are the second three-dimensional parameters. The deflection calibration error can be determined according to the first three-dimensional parameters and the second three-dimensional parameters. Finally, the calibration accuracy of the binocular deflection system can be determined according to the binocular positioning error and the deflection calibration error. This solves the problem that the existing technology cannot evaluate the quality of the calibration results from a global perspective.

[0134] On the other hand, the present invention further provides a computer storage medium storing an executable program code; the executable program code is used to execute any of the above-mentioned methods for determining the calibration accuracy of a binocular deflection system.

[0135] On the other hand, the present invention also provides a terminal device, including a memory and a processor; the memory stores a program code that can be executed by the processor; the program code is used to execute any of the above-mentioned methods for determining the calibration accuracy of the binocular deflection system.

[0136] Exemplarily, the program code may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the program code in the terminal device.

[0137] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the terminal device may also include an input / output device, a network access device, a bus, etc.

[0138] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0139] The memory may be an internal storage unit of the terminal device, such as a hard disk or a memory. The memory may also be an external storage device of the terminal device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the terminal device. Further, the memory may also include both an internal storage unit of the terminal device and an external storage device. The memory is used to store the program code and other programs and data required by the terminal device. The memory may also be used to temporarily store data that has been output or is to be output.

[0140] The above-mentioned computer storage medium and terminal device are created based on the method for determining the calibration accuracy of the above-mentioned binocular deflection system. Their technical functions and beneficial effects are no longer repeated here. The various technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the various technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0141] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for determining the calibration accuracy of a binocular deflection system, characterized in that: include: Setting a number of coded patterns within the overlapping line of sight of each acquisition device; Obtain the coordinates of any corner point of the coding pattern in the coordinate system of each acquisition device, and convert the coordinates of the corner point to the same coordinate system to calculate the distance error and / or relative error between the coordinates in the same coordinate system to determine the binocular positioning error; Taking any acquisition device as the main device, the coordinates and normal direction of the center point of the coding pattern in the main device coordinate system are obtained as the first three-dimensional parameters; Displaying a set stripe pattern, reflecting the coding pattern surface to collect a deformed stripe pattern reflected by the coding pattern surface, and reconstructing the deformed stripe pattern to obtain the coordinates and normal of the center point of the coding pattern in the main device coordinate system, which are the second three-dimensional parameters; determining a deflection calibration error based on the first three-dimensional parameter and the second three-dimensional parameter; The calibration accuracy of the binocular deflection system is determined based on the binocular calibration error and the deflection calibration error.

2. The method according to claim 1, characterized in that The step of calculating the distance error also includes: Traverse the corner points of the coded pattern and obtain the coordinates of the selected corner points in the coordinate system of each acquisition device; The coordinates of the selected corner points in each acquisition device coordinate system are converted to the coordinate systems of other acquisition devices, and the average value of the distance errors between the coordinates of the selected corner points in each acquisition device coordinate system is calculated as the final distance error.

3. The method according to claim 1, characterized in that The step of calculating the relative error also includes: The basic matrix is ​​obtained according to the prior calibration external parameters and the prior acquisition device internal parameters; the prior calibration external parameters include the rotation transformation matrix and the translation transformation matrix between the acquisition devices; the prior acquisition device internal parameters include any one or more of the focal length, the principal point, the distortion coefficient and the acquisition device matrix; Obtaining the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device; Traverse the coordinates of the corner points in the coordinate system of each acquisition device, calculate the straight line where the coordinates of each corner point are located in the image coordinate system of another acquisition device according to the basic matrix, and calculate the distance value between the coordinates of each corner point and the corresponding straight line in the same coordinate system, and obtain the average value of all distance values, which is the relative error.

4. The method according to claim 1, characterized in that The steps of determining the binocular positioning error according to the distance error and / or the relative error include: According to the size of distance error and relative error, the larger error is taken as the error of binocular calibration; Or, according to the distance error and the relative error, the final error value is calculated according to the weight, which is the binocular positioning error.

5. The method according to claim 3, characterized in that: The step of obtaining the coordinates of the corner points of the coding pattern in the coordinate system of each acquisition device includes: Acquire a 2D image of the coding pattern within the acquisition range of each acquisition device; Establish a plane coordinate system with any corner point of the 2D image as the origin, and obtain the coordinates of each corner point in the plane coordinate system; The rotation and translation matrix between the plane coordinate system and each acquisition device is obtained, so as to transform the coordinates of each corner point in the plane coordinate system to the coordinate system of each acquisition device according to the rotation and translation matrix.

6. The method according to claim 1, characterized in that The step of obtaining the first three-dimensional parameter comprises: Establishing a plane coordinate system with any point of the coding pattern as the origin; preferably establishing a plane coordinate system with the center point of the coding pattern as the origin; According to the coordinates of the corner points of each coding pattern in the plane coordinate system and the main device coordinate system, a rotation matrix and a translation matrix between the plane coordinate system and the main device coordinate system are obtained; Fitting the plane equation according to the coordinates of the corner points of the coding pattern to obtain the normal direction of the center point of the coding pattern; The coordinates and normal of the center point of the coding pattern in the plane coordinate system are converted to the main device coordinate system according to the rotation matrix and the translation matrix, so as to obtain the coordinates and normal of the center point of the coding pattern in the main device coordinate system, which are the first three-dimensional parameters.

7. The method according to claim 6, characterized in that The step of determining the deflection calibration error according to the first three-dimensional parameter and the second three-dimensional parameter comprises: Get the distance between the coordinates of each center point in the main device coordinate system, which is the coordinate distance difference; Get the angle between the normals of each center point in the main device coordinate system, which is the angle distance difference.

8. The method according to claim 1, characterized in that The steps of determining the calibration accuracy of the binocular deflection system according to the binocular calibration error and the deflection calibration error include: Obtaining the ratios of the binocular positioning error and the deflection calibration error to the corresponding error thresholds, and determining the binocular positioning accuracy and the deflection calibration accuracy; According to the size of binocular positioning accuracy and deflection calibration accuracy, the larger accuracy is taken as the binocular deflection system calibration accuracy.

9. A computer storage medium, characterized in that An executable program code is stored; the executable program code is used to execute the method for determining the calibration accuracy of the binocular deflection system according to any one of claims 1 to 8.

10. A terminal device, characterized in that: It comprises a memory and a processor; the memory stores a program code executable by the processor; the program code is used to execute the method for determining the calibration accuracy of a binocular deflection system as described in any one of claims 1-8.

Images