CN111383277B - Wide-interval double-camera module AA method and system - Google Patents
Wide-interval double-camera module AA method and system Download PDFInfo
- Publication number
- CN111383277B CN111383277B CN201811631687.3A CN201811631687A CN111383277B CN 111383277 B CN111383277 B CN 111383277B CN 201811631687 A CN201811631687 A CN 201811631687A CN 111383277 B CN111383277 B CN 111383277B
- Authority
- CN
- China
- Prior art keywords
- module
- corner
- camera
- utilizing
- coordinates
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/50—Constructional details
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/57—Mechanical or electrical details of cameras or camera modules specially adapted for being embedded in other devices
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Studio Devices (AREA)
Abstract
The invention relates to the technical field of double-shot module assembly processes, and discloses a wide-interval double-shot module AA method and a system, wherein the double-shot module comprises a first module and a second module, and is characterized in that the method comprises the following steps: s10: acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module; s20: calculating a homography matrix between the first module and the second module according to a preset calculation model; the preset calculation model is formed by utilizing a camera calibration principle; s30: and moving the second module by utilizing a six-axis platform according to the homography matrix so as to adjust the spatial position of the second module relative to the first module. The invention can meet the high-precision control of the wide-space double-shot module AA and improve the imaging quality and the processing capability of the wide-space double-shot module.
Description
Technical Field
The invention relates to the technical field of double-camera module assembly processes, in particular to a wide-interval double-camera module AA method and a system.
Background
Along with the increasing requirements of customers on optical axis precision, tilt precision, optical center precision and the like of the double-shot module, the common double-shot assembly process cannot meet the requirements on high precision adjustment of the optical axis, tilt and the like, particularly the wide-interval double-shot module, the interval between the two modules is about 120mm-200mm, the overlapping degree of acquired images is little, even no overlapping is caused, if two image imaging parameters are acquired through shooting a plane-like target image, the position relation between the two imaging parameters is calculated through the overlapping condition of the imaging parameters to guide the precision adjustment of the optical axis, tilt and the like, and larger errors tend to appear, so that the requirements of customers can be better met, the enterprise competitiveness is enhanced, and the development of a high-precision double-shot module AA scheme is imperative.
AA, active alignment, is interpreted as chinese, active alignment, a technique that determines the relative position of parts during assembly. The AA difficulty in the existing dual-camera module AA process is that the relative positions of two camera modules are adjusted, generally, in the process scheme, the two camera modules are firstly grabbed into an AA device, then a picture is respectively shot through the two camera modules, a relation coordinate axis is established by the optical axis of one camera module, an offset angle between the two camera modules is obtained through an algorithm, then the camera modules with the own optical axes establishing the relation coordinate axis are fixed, the other camera module is adjusted to rotate around the three coordinate axes of the established coordinate axis according to the offset angle, and finally the high-precision dual-camera module is obtained by adjustment;
for example, in the patent document "a method and apparatus for dual-camera automatic AA" with publication number CN106060399a, a picture is taken by using dual cameras, where the taken picture is a table chart, and the table chart can be beneficial to selecting a target point and determining coordinates of the target point on the table chart, so as to facilitate the implementation of a subsequent algorithm. The pictures shot by the double cameras are combined with the established coordinate axes, the offset angles of the main camera and the auxiliary camera can be obtained through an algorithm, then the main camera is fixed, and the auxiliary camera rotates around the three coordinate axes of the established coordinate axes respectively according to the calculated angles, so that the double cameras with parallel optical axes can be obtained. However, when the imaging device is applied to wide-space double shooting, errors exist, and imaging quality is affected.
Disclosure of Invention
The invention aims to solve the technical problem of providing a method and a system for controlling the wide-space double-shot module AA with high precision, which can meet the requirement of the wide-space double-shot module AA with high imaging quality and processing capacity.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
a wide-pitch dual-camera module AA method, the dual-camera module comprising a first module and a second module, the method comprising:
s10: acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
s20: calculating a homography matrix between the first module and the second module according to a preset calculation model; the preset calculation model is formed by utilizing a camera calibration principle;
s30: and moving the second module by utilizing a six-axis platform according to the homography matrix so as to adjust the spatial position of the second module relative to the first module.
In the scheme, the relative spatial position between the first module and the second module is calculated through the relation conversion between the basic coordinate systems mainly according to the principle of camera calibration, then one module is kept still, and six-axis parameters of the other module are moved to realize position calibration. The method is simple and easy to operate, and is beneficial to improving the process efficiency; specifically, the images shot by the first module and the second module are images of the same three-dimensional checkerboard, compared with the plane target in the prior art, the three-dimensional checkerboard is applied to the two-shot module capable of helping wide intervals to fully acquire three-dimensional coordinate parameters of characteristic points of the checkerboard, sufficient data are obtained for calculation of a homography matrix, the homography matrix calculation error is reduced, and finally the six-axis platform is guided to accurately and rapidly adjust relative space positions between the two modules according to the calculated homography matrix.
Further, the step S20 includes the steps of:
s21: detecting corner points of checkerboard in the two images respectively to obtain corner point coordinates, and sequencing the corner points;
s22: converting the corner coordinates into three-dimensional space coordinates to obtain corner camera coordinates;
s23: establishing a corner homography relationship by utilizing the corner camera coordinates and pre-stored corner coordinate plates, wherein the corner coordinate plates are coordinates of the corner points on the three-dimensional checkerboard coordinate plates;
s24: and calculating a homography matrix between the first module and the second module according to the angular point homography relation of the first module and the angular point homography relation of the second module.
According to the method, by using the principle of double-shot calibration, angular points of a checkerboard in an image are detected to serve as characteristic points for relation conversion of a coordinate system, then a homography matrix is obtained through homography relation between two shooting modules and the same target, a three-dimensional scene can be reconstructed accurately according to a calculation model established by the method, and the homography matrix with high accuracy is obtained to represent the relative position relation between two shooting modules of the double-shot module.
Further, the step S30 includes the steps of:
S31A: splitting the homography matrix into a translation matrix and three Euler angles;
S32A: and according to the translation matrix, utilizing a six-axis platform to adjust Shift parameters of the second module relative to the first module, and utilizing the six-axis platform to adjust Tilt and Rotation parameters of the second module relative to the first module according to the three Euler angles.
The homography matrix can be split into a translation matrix and three Euler angles according to a conversion formula, and the control of Shift, tilt and Rotation parameters of the double-camera module can be realized according to the parameters, so that the method is simple and has good practicability.
Further, the step S30 includes the steps of:
S31B: minimizing ghost errors of the homography matrix to obtain an optimized homography matrix;
S32B: splitting the optimized homography matrix into a translation matrix and three Euler angles;
S33B: and according to the translation matrix, utilizing a six-axis platform to adjust Shift parameters of the second module relative to the first module, and utilizing the six-axis platform to adjust Tilt and Rotation parameters of the second module relative to the first module according to the three Euler angles.
The method introduces the step of error reduction, performs parameter optimization on the homography matrix, and adopts a mode of iterative optimization of the reprojection error, so that the calculation error of the homography matrix can be improved, and the measurement error of image points (namely corner points) can be improved, so that the accuracy of the result is higher.
Further, the step S31B includes iteratively optimizing the ghost error value by using a gradient descent method and a Gauss-Newton method until the ghost error value is less than or equal to a preset standard value.
Further, the step S10 includes the following steps:
the double-shooting module support, which is used for grabbing and fixing the first module, is placed at the geometric center of the three-dimensional checkerboard target;
grabbing the second module and placing the second module into a through hole of the double-camera module bracket, wherein the bottom of the second module is abutted with a six-axis platform;
and controlling the first module and the second module to shoot the images of the three-dimensional checkerboard target.
In the scheme, the position of the first module is emphasized to be fixed all the time, and the situation that the calibrated parameter acquisition is influenced by shaking caused by placing or moving the second module is avoided.
A wide-pitch dual camera module AA system, the system comprising:
the acquisition module is used for acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
the computing module is used for computing a homography matrix between the first module and the second module according to a preset computing model; the preset calculation model is formed by utilizing a camera calibration principle;
and the adjusting module is used for moving the second module by utilizing the six-axis platform according to the homography matrix so as to adjust the spatial position of the second module relative to the first module.
Further, the computing module includes:
the detection ordering unit is used for respectively detecting the angular points of the checkerboard in the two images to obtain angular point coordinates and ordering the angular points;
the conversion unit is used for converting the angular point coordinates into three-dimensional space coordinates so as to obtain angular point camera coordinates;
the relationship establishing unit is used for establishing a corner homography relationship by utilizing the corner camera coordinates and pre-stored corner target coordinates, wherein the corner target coordinates are coordinates of the corner on the three-dimensional checkerboard target;
and the matrix calculation unit is used for calculating the homography matrix between the first module and the second module according to the angular point homography relation of the first module and the angular point homography relation of the second module.
Further, the adjustment module includes:
the splitting unit is used for splitting the homography matrix into a translation matrix and three Euler angles;
and the adjusting unit is used for adjusting the Shift parameters of the second module relative to the first module by utilizing the six-axis platform according to the translation matrix, and adjusting the Tilt and Rotation parameters of the second module relative to the first module by utilizing the six-axis platform according to the three Euler angles.
Further, the adjustment module includes:
an optimizing unit, configured to minimize a ghost error of the homography matrix to obtain an optimized homography matrix;
the splitting unit is used for splitting the optimized homography matrix into a translation matrix and three Euler angles;
and the adjusting unit is used for adjusting the Shift parameters of the second module relative to the first module by utilizing the six-axis platform according to the translation matrix, and adjusting the Tilt and Rotation parameters of the second module relative to the first module by utilizing the six-axis platform according to the three Euler angles.
After the technical scheme is adopted, the invention has the beneficial effects that: the standard board is a three-dimensional checkerboard standard board, so that the first module and the second module can be fully assisted in acquiring calibration reference image data, distortion errors are reduced, and the sufficiency and accuracy of the reference data are ensured; the homography matrix is calculated and obtained by utilizing the camera calibration principle to represent the relative spatial position relation between the first module and the second module, and the ghost error of the homography matrix is minimized and optimized, so that the accuracy of data is further guaranteed; the optimized homography matrix is used as a moving basis, the spatial position of the second module relative to the first module is adjusted by moving the second module only, the uncertain factor interference during assembly is reduced, the assembly precision and the imaging quality are ensured, the assembly efficiency is improved, the capability of a manufacturing process is improved, and the enterprise competitiveness is enhanced.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the accompanying drawings are as follows:
FIG. 1 is a flowchart of a method for forming a wide-pitch dual-camera module AA according to embodiment 1 of the present invention;
FIG. 2 is a flowchart of a method for forming a wide-pitch dual-camera module AA according to embodiment 2 of the present invention;
FIG. 3 is a flowchart of a preferred step S30 method provided in embodiment 3 of the present invention;
fig. 4 is a block diagram of a preferred wide-pitch dual-camera module AA system according to embodiment 4 of the present invention.
Detailed Description
The invention solves the problem that the high-precision AA control of a wide-interval double-shot module cannot be realized in the prior art, provides a wide-interval double-shot module AA method and a system, and utilizes the principle of double-shot calibration to set a three-dimensional checkerboard mark plate to help the calculation of the relative spatial positions of two camera modules, and then fixes the position of one camera by fixing the position of the camera, namely, firstly, the camera is subjected to the processes used when a series of single modules such as glue drawing, exposure, test and the like are fixed on a module support, then the position of the other camera is moved to calibrate, and finally, the other camera is fixed after the calibration of the relative positions of the two modules is completed to finish the AA of a dual-mode product.
The following are specific embodiments of the present invention and the technical solutions of the present invention will be further described with reference to the accompanying drawings, but the present invention is not limited to these embodiments.
Example 1
As shown in fig. 1, the present embodiment provides a wide-pitch dual-camera module AA method, where the dual-camera module includes a first module and a second module, and the method includes:
s10: acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
specifically, when the first module and the second module shoot images of the three-dimensional checkerboard target, the placement positions of the first module and the second module are preferably arranged at the geometric center of the three-dimensional checkerboard target, namely, the geometric center of a double-shot module formed by the first module and the second module coincides with the geometric center of the three-dimensional checkerboard target, wherein the three-dimensional checkerboard target can be a three-dimensional annular target or a multi-piece planar target distributed in a rectangular mode, and the three-dimensional checkerboard target is required to be provided with patterns with clear black and white contrast; the application of the three-dimensional checkerboard mark is beneficial to the first module and the second module to acquire sufficient image parameters for calculating the relative position relationship of the first module and the second module;
during the AA process, the first module and the second module execute image shooting after receiving control signals, and the setting of the step is to acquire two images shot by the first module and the second module;
s20: calculating a homography matrix between the first module and the second module according to a preset calculation model; the preset calculation model is formed by utilizing a camera calibration principle; presetting a processing process of setting parameters of an image in a calculation model; generally, to obtain the homography matrix between the first module and the second module, the homography relationship between the first module and the target and the homography relationship between the second module and the target are obtained first, so that the homography matrix of the first module and the homography matrix of the second module are calculated according to the relationship between the first module and the second module and the same target, and the obtaining of the homography relationship between the first module and the second module and the target needs to use the conversion between the coordinates of the image feature points in the imaging plane coordinate system and the coordinates of the image feature points in the camera coordinate system in the camera calibration principle.
S30: and moving the second module by utilizing a six-axis platform according to the homography matrix so as to adjust the spatial position of the second module relative to the first module.
In step S30, the relative spatial position between the two modules can be represented according to the homography matrix obtained in the previous step, and after comparing and calculating the homography matrix with the accuracy (such as Rotation, tilt, optical center, optical axis accuracy, etc.) which is pre-stored in the database and theoretically required to be achieved between the two modules, the six-axis adjustment parameter (X, Y, Z, tx, ty, tz) of the second module can be further obtained, and the ideal distance between the two modules can be finally obtained by moving the position change of the second module according to the six-axis adjustment parameter, so that the high-accuracy control of the wide-distance two-camera module AA is realized.
In summary, the embodiment realizes high-precision control of the wide-interval dual-camera module AA, which is beneficial to improving the imaging quality and the production line processing capacity of the wide-interval dual-camera module. Specifically, the relative spatial position between the first module and the second module is calculated through the relation conversion between basic coordinate systems mainly according to the principle of camera calibration, then one module is kept still, and six-axis parameters of the other module are moved to realize position calibration. The method is simple and easy to operate, and is beneficial to improving the process efficiency; specifically, the images shot by the first module and the second module are images of the same three-dimensional checkerboard, compared with the plane target in the prior art, the three-dimensional checkerboard is applied to the two-shot module capable of helping wide intervals to fully acquire three-dimensional coordinate parameters of characteristic points of the checkerboard, sufficient data are obtained for calculation of a homography matrix, distortion errors are reduced, calculation errors of the homography matrix are reduced, and finally accurate and rapid guiding of the six-axis platform to adjust out the relative spatial positions meeting requirements between the two modules is facilitated according to the calculated homography matrix.
Example 2
As shown in fig. 2, this embodiment further optimizes the solution of embodiment 1, simplifies the method steps and improves the calculation accuracy, and the step S20 includes the following steps:
s21: detecting corner points of checkerboard in the two images respectively to obtain corner point coordinates, and sequencing the corner points; the corner coordinates are coordinates of the corner in an imaging plane coordinate system, the method for detecting the corner can be used for automatically extracting the corner based on a straight line detection algorithm (LSD) in the prior art, and specifically, the algorithm mainly comprises the steps of detecting crossed straight lines, pseudo-sorting and analyzing straight line length and angle, removing pseudo-straight lines, merging adjacent straight line endpoints and optimizing coordinates, recovering a checkerboard structure and sorting the corner. The coordinates of the corner points can be accurately grasped and correctly ordered.
S22: converting the corner coordinates into three-dimensional space coordinates to obtain corner camera coordinates; the method is characterized in that the angular point coordinates are converted into three-dimensional space coordinates under a camera coordinate system by utilizing the principle of optical imaging to obtain angular point camera coordinates, and the angular point camera coordinates are a two-dimensional point-to-three-dimensional point conversion process which is applied to internal reference data of a camera module, and particularly refer to the camera calibration principle in the prior art.
S23: establishing a corner homography relationship by utilizing the corner camera coordinates and pre-stored corner coordinate plates, wherein the corner coordinate plates are coordinates of the corner points on the three-dimensional checkerboard, namely coordinates under a world coordinate system; the establishment of the relationship immediately establishes a projection mapping relationship between coordinates of the corner point under a camera coordinate system and coordinates of the corner point under a world coordinate system; angular point homography relation H 1 The method can be obtained by the following formula:
wherein X is the corner camera coordinates, and Xp is the pre-stored corner target coordinates.
S24: and calculating a homography matrix between the first module and the second module according to the angular point homography relation of the first module and the angular point homography relation of the second module. The target is identical, so that the homography matrix between the first module and the second module can be conveniently obtained according to the homography relation of the respective angular points of the first module and the second module, and the homography matrix is 3*3, which represents the coordinate relation of the first module and the second module on XYZ three axes in the world coordinate axis.
According to the method, by using the principle of double-shot calibration, angular points of a checkerboard in an image are detected to serve as characteristic points for relation conversion of a coordinate system, then a homography matrix is obtained through homography relation between two shooting modules and the same target, a three-dimensional scene can be reconstructed accurately according to a calculation model established by the method, and the homography matrix with high accuracy is obtained to represent the relative position relation between two shooting modules of the double-shot module.
The step S30 includes the steps of:
S31A: splitting the homography matrix into a translation matrix and three Euler angles; the homography matrix can be split into a translation matrix and three Euler angles according to a conversion formula (Rodrigos formula), wherein the translation matrix represents the Shift parameter of the second module relative to the first module, and the three Euler angles represent Rotation parameters rotating around three axes of XYZ respectively, namely the Tilt and Rotation parameters of the second module relative to the first module. The specific conversion process is as follows:
by means of
Expressing homography matrix to obtain +.>
And t, wherein->
(n=1, 2, 3) is a column vector, t is a translation vector; the translation matrix is obtained according to t;
by means of
，/>
And r=r (yaw) R (pitch) R (roll) to obtain three euler angles (yaw, picture, roll).
The control of Shift, tilt and Rotation parameters of the double-camera module can be easily and conveniently realized by combining the pre-stored precision of OC (optical center), optical axis, tilt, rotation and the like by taking the parameters as the basis, as described in the following step S32A;
S32A: according to the translation matrix and the precision parameters pre-stored in the database, obtaining translation parameters, moving the second module to adjust the Shift parameters of the second module relative to the first module according to the translation parameters by using a six-axis platform, obtaining Rotation parameters according to the three Euler angles and the precision parameters pre-stored in the database, and moving the second module to adjust the Tilt and Rotation parameters of the second module relative to the first module according to the Rotation parameters by using the six-axis platform.
In summary, the calculation method of the embodiment is simple, has good practicability, and can be used for controlling the three position relation parameters of Shift, tilt and Rotation of two modules in the double-camera module with high precision.
Example 3
As shown in fig. 3, the difference between the present embodiment and the previous embodiment is that in order to reduce the accuracy effect caused by the calculation error and the image measurement error, the homography matrix is optimized, which is specifically as follows:
the step S30 includes the steps of:
S31B: minimizing ghost errors of the homography matrix to obtain an optimized homography matrix;
S32B: splitting the optimized homography matrix into a translation matrix and three Euler angles;
S33B: and according to the translation matrix, utilizing a six-axis platform to adjust Shift parameters of the second module relative to the first module, and utilizing the six-axis platform to adjust Tilt and Rotation parameters of the second module relative to the first module according to the three Euler angles.
In the above method, the error reduction step S31B is introduced, parameter optimization is performed on the homography matrix, and the re-projection error iterative optimization method is used, so that not only can the calculation error of the homography matrix be improved, but also the measurement error of the image point (i.e. the corner point) can be improved, so that the accuracy of the result is higher.
Preferably, the step S31B includes iteratively optimizing the ghost error value using a gradient descent method and a Gauss-Newton method until the ghost error value is equal to or less than a preset standard value. Specifically, the gradient descent method ensures that each iteration function is descended, the descent speed is very fast just when the initial point is far away from the optimal point, but convergence is very slow, and the combined application of Gauss-Newton method ensures the convergence speed.
The step S10 includes the following steps:
the double-shooting module support, which is used for grabbing and fixing the first module, is placed at the geometric center of the three-dimensional checkerboard target; that is, the first module is in a fixed state before the relative position AA between the first module and the second module is performed.
Grabbing the second module into a through hole of the double-camera module bracket, wherein the bottom of the second module is abutted with a six-axis platform and is used for facilitating the six-axis platform to move the second module to change positions;
and controlling the first module and the second module to shoot the images of the three-dimensional checkerboard target.
In the design of the step, the position of the first module is emphasized to be always fixed, and only the second module is adjusted to complete the calibration process, so that the influence of shaking caused by placing or moving the second module on accurate acquisition of calibrated parameters can be avoided, interference items are reduced, and high matching of parameter acquisition, theoretical calculation and actual conditions is ensured.
The embodiment can better optimize the homography matrix, so that the six-axis adjustment parameters obtained according to the homography matrix are more accurate.
Example 4
As shown in fig. 4, the present embodiment provides a wide-pitch dual-camera module AA system, which includes:
the acquiring module 100 is configured to acquire images of the same three-dimensional checkerboard target captured by the first module and the second module;
the calculating module 200 is configured to calculate a homography matrix between the first module and the second module according to a preset calculating model; the preset calculation model is formed by utilizing a camera calibration principle;
the adjusting module 300 is configured to move the second module by using a six-axis platform according to the homography matrix to adjust a spatial position of the second module relative to the first module.
Preferably, the calculation module 200 includes:
a detection ordering unit 210, configured to detect corner points of the checkerboard in the two images respectively to obtain corner point coordinates, and order the corner points; the detection and sorting unit 210 performs corner detection and accurate sorting on the two images captured by the first module and the second module, where the corner detection process is a process of accurately acquiring the coordinates of the corner, and the algorithm used may be described in embodiment 2.
A conversion unit 220, configured to convert the corner coordinates into three-dimensional space coordinates to obtain corner camera coordinates; the corner camera coordinates are the coordinates under the camera coordinate system, in the first module camera coordinate system, the first module center point is the center point of the camera coordinate system, the first module main axis is the Z axis of the camera coordinate system, and the first module is the same.
A relationship establishing unit 230, configured to establish a corner homography relationship using the corner camera coordinates and pre-stored corner coordinate values, where the corner coordinate values are coordinates of the corner points on the three-dimensional checkerboard coordinate values;
and a matrix calculating unit 240, configured to calculate a homography matrix between the first module and the second module according to the angular point homography relationship of the first module and the angular point homography relationship of the second module.
Optionally, the adjustment module 300 includes:
a splitting unit 310, configured to split the homography matrix into a translation matrix and three euler angles;
and the adjusting unit 320 is configured to adjust Shift parameters of the second module relative to the first module by using a six-axis platform according to the translation matrix, and adjust Tilt and Rotation parameters of the second module relative to the first module by using the six-axis platform according to the three euler angles.
Optionally, the adjustment module 300 includes:
the optimizing unit 330: for minimizing ghost errors of the homography matrix to obtain an optimized homography matrix; minimizing all calculation methods can be described with reference to example 3, but is not limited thereto.
A splitting unit 310, configured to split the optimized homography matrix into a translation matrix and three euler angles;
and the adjusting unit 320 is configured to adjust Shift parameters of the second module relative to the first module by using a six-axis platform according to the translation matrix, and adjust Tilt and Rotation parameters of the second module relative to the first module by using the six-axis platform according to the three euler angles.
In conclusion, the method and the device have the advantages that the targets are arranged to be the three-dimensional checkerboard targets, so that the first module and the second module can be fully helped to acquire calibration reference image data, distortion errors are reduced, and the sufficiency and accuracy of the reference data are ensured; the homography matrix is calculated and obtained by utilizing the camera calibration principle to represent the relative spatial position relation between the first module and the second module, and the ghost error of the homography matrix is minimized and optimized, so that the accuracy of data is further guaranteed; the optimized homography matrix is used as a moving basis, the spatial position of the second module relative to the first module is adjusted by moving the second module only, the uncertain factor interference during assembly is reduced, the assembly precision and the imaging quality are ensured, the assembly efficiency is improved, the capability of a manufacturing process is improved, and the enterprise competitiveness is enhanced.
The specific embodiments described herein are offered by way of example only to illustrate the spirit of the invention. Those skilled in the art may make various modifications or additions to the described embodiments or substitutions thereof without departing from the spirit of the invention or exceeding the scope of the invention as defined in the accompanying claims.
Claims (10)
1. A wide-pitch dual-camera module AA method, the dual-camera module comprising a first module and a second module, the method comprising:
s10: acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
s20: calculating a homography matrix between the first module and the second module according to a preset calculation model; the preset calculation model is formed by utilizing a camera calibration principle;
s30: according to the homography matrix, moving the second module by utilizing a six-axis platform to adjust the spatial position of the second module relative to the first module; the step S30 includes the steps of:
S31A: splitting the homography matrix into a translation matrix and three Euler angles;
S32A: and according to the translation matrix, utilizing a six-axis platform to adjust Shift parameters of the second module relative to the first module, and utilizing the six-axis platform to adjust Tilt and Rotation parameters of the second module relative to the first module according to the three Euler angles.
2. The method of claim 1, wherein the step S20 includes the steps of:
s21: detecting corner points of checkerboard in the two images respectively to obtain corner point coordinates, and sequencing the corner points;
s22: converting the corner coordinates into three-dimensional space coordinates to obtain corner camera coordinates;
s23: establishing a corner homography relationship by utilizing the corner camera coordinates and pre-stored corner coordinate plates, wherein the corner coordinate plates are coordinates of the corner points on the three-dimensional checkerboard coordinate plates;
s24: and calculating a homography matrix between the first module and the second module according to the angular point homography relation of the first module and the angular point homography relation of the second module.
3. The method of claim 1, wherein the step S10 is preceded by the steps of:
the double-shooting module support, which is used for grabbing and fixing the first module, is placed at the geometric center of the three-dimensional checkerboard target;
grabbing the second module and placing the second module into a through hole of the double-camera module bracket, wherein the bottom of the second module is abutted with a six-axis platform; and controlling the first module and the second module to shoot the images of the three-dimensional checkerboard target.
4. A wide-pitch dual-camera module AA method, the dual-camera module comprising a first module and a second module, the method comprising:
s10: acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
s20: calculating a homography matrix between the first module and the second module according to a preset calculation model; the preset calculation model is formed by utilizing a camera calibration principle;
s30: according to the homography matrix, moving the second module by utilizing a six-axis platform to adjust the spatial position of the second module relative to the first module; the step S30 includes the steps of:
S31B: minimizing ghost errors of the homography matrix to obtain an optimized homography matrix;
S32B: splitting the optimized homography matrix into a translation matrix and three Euler angles;
S33B: and according to the translation matrix, utilizing a six-axis platform to adjust Shift parameters of the second module relative to the first module, and utilizing the six-axis platform to adjust Tilt and Rotation parameters of the second module relative to the first module according to the three Euler angles.
5. The method according to claim 4, wherein the step S31B includes iteratively optimizing the ghost error value by using a gradient descent method and a Gauss-Newton method until the ghost error value is less than or equal to a predetermined standard value.
6. The method of claim 4, wherein the step S10 is preceded by the steps of:
the double-shooting module support, which is used for grabbing and fixing the first module, is placed at the geometric center of the three-dimensional checkerboard target;
grabbing the second module and placing the second module into a through hole of the double-camera module bracket, wherein the bottom of the second module is abutted with a six-axis platform; and controlling the first module and the second module to shoot the images of the three-dimensional checkerboard target.
7. The method of claim 4, wherein the step S20 includes the steps of:
s21: detecting corner points of checkerboard in the two images respectively to obtain corner point coordinates, and sequencing the corner points;
s22: converting the corner coordinates into three-dimensional space coordinates to obtain corner camera coordinates;
s23: establishing a corner homography relationship by utilizing the corner camera coordinates and pre-stored corner coordinate plates, wherein the corner coordinate plates are coordinates of the corner points on the three-dimensional checkerboard coordinate plates;
s24: and calculating a homography matrix between the first module and the second module according to the angular point homography relation of the first module and the angular point homography relation of the second module.
8. A wide-pitch dual-camera module AA system, the dual-camera module comprising a first module and a second module, the system comprising:
the acquisition module is used for acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
the computing module is used for computing a homography matrix between the first module and the second module according to a preset computing model; the preset calculation model is formed by utilizing a camera calibration principle;
the adjusting module is used for moving the second module by utilizing a six-axis platform according to the homography matrix so as to adjust the spatial position of the second module relative to the first module; the adjustment module includes:
the splitting unit is used for splitting the homography matrix into a translation matrix and three Euler angles;
and the adjusting unit is used for adjusting the Shift parameters of the second module relative to the first module by utilizing the six-axis platform according to the translation matrix, and adjusting the Tilt and Rotation parameters of the second module relative to the first module by utilizing the six-axis platform according to the three Euler angles.
9. The wide-pitch dual camera module AA system of claim 8, wherein said computing module comprises: the detection ordering unit is used for respectively detecting the angular points of the checkerboard in the two images to obtain angular point coordinates and ordering the angular points;
the conversion unit is used for converting the angular point coordinates into three-dimensional space coordinates so as to obtain angular point camera coordinates;
the relationship establishing unit is used for establishing a corner homography relationship by utilizing the corner camera coordinates and pre-stored corner target coordinates, wherein the corner target coordinates are coordinates of the corner on the three-dimensional checkerboard target;
and the matrix calculation unit is used for calculating the homography matrix between the first module and the second module according to the angular point homography relation of the first module and the angular point homography relation of the second module.
10. A wide-pitch dual-camera module AA system, the dual-camera module comprising a first module and a second module, the system comprising:
the acquisition module is used for acquiring images of the same three-dimensional checkerboard target shot by the first module and the second module;
the computing module is used for computing a homography matrix between the first module and the second module according to a preset computing model; the preset calculation model is formed by utilizing a camera calibration principle;
the adjusting module is used for moving the second module by utilizing a six-axis platform according to the homography matrix so as to adjust the spatial position of the second module relative to the first module; the adjustment module includes:
an optimizing unit, configured to minimize a ghost error of the homography matrix to obtain an optimized homography matrix;
the splitting unit is used for splitting the optimized homography matrix into a translation matrix and three Euler angles;
and the adjusting unit is used for adjusting the Shift parameters of the second module relative to the first module by utilizing the six-axis platform according to the translation matrix, and adjusting the Tilt and Rotation parameters of the second module relative to the first module by utilizing the six-axis platform according to the three Euler angles.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811631687.3A CN111383277B (en) | 2018-12-29 | 2018-12-29 | Wide-interval double-camera module AA method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811631687.3A CN111383277B (en) | 2018-12-29 | 2018-12-29 | Wide-interval double-camera module AA method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111383277A CN111383277A (en) | 2020-07-07 |
| CN111383277B true CN111383277B (en) | 2023-05-19 |
Family
ID=71217752
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811631687.3A Active CN111383277B (en) | 2018-12-29 | 2018-12-29 | Wide-interval double-camera module AA method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111383277B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111890255B (en) * | 2020-07-29 | 2021-10-01 | 苏州华兴源创科技股份有限公司 | Crimping method and crimping system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106060399A (en) * | 2016-07-01 | 2016-10-26 | 信利光电股份有限公司 | Automatic AA method and device for double cameras |
| CN206524916U (en) * | 2017-02-28 | 2017-09-26 | 西安中科光电精密工程有限公司 | A kind of self-adapting detecting and the double devices for taking the photograph optical axis of adjustment |
| CN107578450A (en) * | 2017-09-14 | 2018-01-12 | 长沙全度影像科技有限公司 | A kind of method and system for the demarcation of panorama camera rigging error |
| CN107644442A (en) * | 2016-07-21 | 2018-01-30 | 宁波舜宇光电信息有限公司 | Double space position calibration methods for taking the photograph module |
| CN108089275A (en) * | 2016-11-21 | 2018-05-29 | 宁波舜宇光电信息有限公司 | For camera module concentricity adjusting apparatus |
| CN207820117U (en) * | 2017-12-26 | 2018-09-04 | 余姚舜宇智能光学技术有限公司 | A kind of adjustable in pitch pair is taken the photograph module |
| CN109064513A (en) * | 2018-08-14 | 2018-12-21 | 深圳中科精工科技有限公司 | The automatic calibration algorithm of six degree of freedom in a kind of encapsulation of camera |
-
2018
- 2018-12-29 CN CN201811631687.3A patent/CN111383277B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106060399A (en) * | 2016-07-01 | 2016-10-26 | 信利光电股份有限公司 | Automatic AA method and device for double cameras |
| CN107644442A (en) * | 2016-07-21 | 2018-01-30 | 宁波舜宇光电信息有限公司 | Double space position calibration methods for taking the photograph module |
| CN108089275A (en) * | 2016-11-21 | 2018-05-29 | 宁波舜宇光电信息有限公司 | For camera module concentricity adjusting apparatus |
| CN206524916U (en) * | 2017-02-28 | 2017-09-26 | 西安中科光电精密工程有限公司 | A kind of self-adapting detecting and the double devices for taking the photograph optical axis of adjustment |
| CN107578450A (en) * | 2017-09-14 | 2018-01-12 | 长沙全度影像科技有限公司 | A kind of method and system for the demarcation of panorama camera rigging error |
| CN207820117U (en) * | 2017-12-26 | 2018-09-04 | 余姚舜宇智能光学技术有限公司 | A kind of adjustable in pitch pair is taken the photograph module |
| CN109064513A (en) * | 2018-08-14 | 2018-12-21 | 深圳中科精工科技有限公司 | The automatic calibration algorithm of six degree of freedom in a kind of encapsulation of camera |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111383277A (en) | 2020-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2018209968A1 (en) | Camera calibration method and system | |
| CN106485758B (en) | Unmanned aerial vehicle camera calibration device, calibration method and assembly line calibration implementation method | |
| CN107270810B (en) | The projector calibrating method and device of multi-faceted projection | |
| WO2012053521A1 (en) | Optical information processing device, optical information processing method, optical information processing system, and optical information processing program | |
| CN108326850B (en) | Method and system for robot to accurately move mechanical arm to reach specified position | |
| CN106910221B (en) | Global calibration method and device | |
| CN110827360B (en) | Photometric stereo measurement system and method for calibrating light source direction thereof | |
| CN113724337B (en) | Camera dynamic external parameter calibration method and device without depending on tripod head angle | |
| WO2020073940A1 (en) | Method and device for calibrating print positioning platform of crystalline silicon photovoltaic solar cell on the basis of machine vision | |
| CN112767338A (en) | Assembled bridge prefabricated part hoisting and positioning system and method based on binocular vision | |
| CN110044262B (en) | Non-contact precision measuring instrument based on image super-resolution reconstruction and measuring method | |
| CN113781579B (en) | Geometric calibration method for panoramic infrared camera | |
| CN109978960A (en) | Based on photogrammetric high-precision screen-camera pose scaling method | |
| CN112489137A (en) | RGBD camera calibration method and system | |
| JP2014060324A (en) | Calculation device, transfer robot system, and calculation method | |
| CN109003312A (en) | A kind of camera calibration method based on nonlinear optimization | |
| CN104167001B (en) | Large-visual-field camera calibration method based on orthogonal compensation | |
| CN112598747A (en) | Combined calibration method for monocular camera and projector | |
| CN111445537A (en) | Calibration method and system of camera | |
| CN113658270B (en) | Method, device, medium and system for multi-vision calibration based on workpiece hole center | |
| CN110044266B (en) | Photogrammetry system based on speckle projection | |
| CN111383277B (en) | Wide-interval double-camera module AA method and system | |
| CN118014832A (en) | Image stitching method and related device based on linear feature invariance | |
| CN208478309U (en) | A kind of optical imaging device and wafer alignment system for wafer alignment | |
| CN108898585A (en) | A kind of axial workpiece detection method and its device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |