Light source-camera joint calibration and correction method for a fixed visual test box

Abstract

The present application relates to a kind of light source-camera joint calibration and correction method for fixed visual test box, method includes: step S1: the built-in light, supplementary light, the three light sources additive physical model of ambient light in the imaging process in fixed visual test box;Step S2: close supplementary light, and the parameter of ambient light in the three light sources additive physical model is estimated;Step S3: solving the supplementary light parameter in the three light sources additive physical model, and supplementary light parameter is real-time issued to supplementary light, realize the optical compensation of fixed visual test box;Step S4: after optical compensation, on the basis of initial checkerboard calibration, the inherent recognition area of fixed visual test box is realized distortion parameter continuous optimization.The present application can effectively realize the light source-camera joint calibration and correction of fixed visual test box.

Description

Technical Field

[0001] This invention relates to the field of calibration of fixed vision test chambers, and in particular to a method for joint calibration and correction of light source and camera for fixed vision test chambers. Background Technology

[0002] Fixed vision test chambers are highly integrated and structurally robust vision systems primarily used in university teaching and research, and industrial training. Camera calibration within a fixed vision test chamber is a core step in acquiring intrinsic, extrinsic, and distortion coefficients for machine vision systems; its accuracy directly determines the reliability of subsequent vision tasks. In fixed vision test chambers used in university teaching and industrial training, the camera and main structure are relatively fixed, but traditional calibration methods still face the following challenges:

[0003] First, the interference from mixed lighting is difficult to eliminate: the light source in the test chamber has limited intensity, and the actual illumination consists of built-in light, active supplementary light, and uncontrollable external ambient light. This time-varying mixed lighting causes severe distortion in the color imaging of objects. Traditional methods only correct this through image post-processing algorithms such as white balance and Retinex, without modeling and separating the contributions of different light sources at the physical level. This makes it impossible to fundamentally solve the color distortion problem, and the accuracy of color recognition is difficult to guarantee.

[0004] Second, calibration parameters are prone to drift and correction is cumbersome: The traditional Zhang's calibration method uses a checkerboard pattern for one-time calibration, but camera lenses can slowly deform due to factors such as temperature and stress, causing the initial calibration parameters to gradually become invalid and resulting in deviations in geometric measurements. Re-performing the checkerboard calibration operation is complex and cannot achieve real-time monitoring and correction, seriously affecting the continuity and reliability of teaching experiments.

[0005] In summary, existing calibration methods for fixed visual test chambers cannot solve the problem of color distortion in objects, and the calibration methods are relatively cumbersome. Summary of the Invention

[0006] Therefore, the technical problem to be solved by the present invention is to overcome the problems in the existing calibration methods for fixed visual test boxes that cannot solve the problem of object color distortion and the complexity of re-performing the checkerboard calibration operation.

[0007] To address the aforementioned technical problems, this invention provides a method for joint calibration and correction of a light source and camera in a fixed vision testing chamber, comprising:

[0008] Step S1: Construct an additive physical model of the three light sources—built-in light, supplementary light, and ambient light—for the imaging process inside the fixed vision test chamber;

[0009] Step S2: Turn off the fill light and estimate the parameters of the ambient light in the three-source additive physical model;

[0010] Step S3: Solve for the parameters of the supplementary light lamp in the three-source additive physical model, and send the supplementary light lamp parameters to the supplementary light lamp in real time to realize optical compensation for the fixed vision test box;

[0011] Step S4: After optical compensation, based on the initial checkerboard calibration, the distortion parameters of the inherent recognition area in the fixed vision test chamber are continuously optimized.

[0012] In one embodiment of the present invention, step S1 constructs a three-source additive physical model of the imaging process within a fixed visual test chamber, consisting of built-in light, supplementary light, and ambient light, as follows:

[0013] ;

[0014] in, This represents the final image pixel value; Pixel contribution for unknown ambient light; Contribute to built-in light pixels with known characteristics; Contribute to the pixels of the fill light; , , The intensity coefficients of ambient light, built-in light, and supplementary light are respectively... It is a fixed value.

[0015] In one embodiment of the present invention, the method for step S2, which involves turning off the supplementary lighting and estimating the parameters of the ambient light in the three-source additive physical model, includes:

[0016] Turn off the fill light, that is Only the built-in light of the fixed vision test chamber is retained;

[0017] Photograph a reference board with a preset grayscale value placed in the recognition area of ​​the test chamber;

[0018] Based on known and the actual measurement The ambient light intensity coefficient is obtained by inverse solving based on the additive physical model of three light sources. With color temperature parameters:

[0019] ;

[0020] in, The grayscale mean of the reference image; The standard grayscale average value of the reference board when illuminated by the built-in light alone; This is the standard reference luminance for ambient light.

[0021] In one embodiment of the present invention, step S3, which solves for the fill light parameters in the three-source additive physical model and sends the fill light parameters to the fill light in real time to achieve optical compensation, includes the following method:

[0022] In normal operating mode, with the goal of ensuring that the spectrum of the recognition area in the fixed vision test chamber is close to that of standard white light D65, the optimal supplementary light intensity coefficient is solved in real time. :

[0023] Objective function:

[0024] ;

[0025] Constraints:

[0026] ;

[0027] ;

[0028] ;

[0029] in, This is the identification area inside the test chamber, used to place the object to be identified. The target point on the object; , , The RGB channel intensity ratio of a standard D65 light source; , , These represent the pixel contributions for the three RGB channels of the target point, respectively. , , These represent the ambient light pixel contributions corresponding to the three RGB channels of the target point; , , These represent the contributions of the built-in light source pixels corresponding to the three RGB channels of the target point; , , These represent the pixel contributions of the fill light corresponding to the three RGB channels of the target point;

[0030] The obtained supplementary lighting parameters Data is sent to the fill light in real time, achieving millisecond-level optical compensation.

[0031] In one embodiment of the present invention, step S4, based on the initial checkerboard calibration, continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber, including:

[0032] Establish a four-coordinate system adapted to the fixed test chamber and clarify the transformation relationships, specifically as follows:

[0033] Establish a world coordinate system The origin is the inner corner of the lower left corner of the stage in the identification area of ​​the test chamber. The platform plane is flat, The physical dimensions of the stage are known fixed values. , , These are the three directional axes of the world coordinate system;

[0034] Establish camera coordinate system With the camera's optical center as the origin The optical axis is If the axes are defined, then the rigid body transformation from world coordinates to camera coordinates can be expressed as: ,in, for Any one of them, for Any one of them, It is a 3×3 rotation matrix. It is a 3×1 translation vector. , , These are the three orientation axes of the camera coordinate system;

[0035] Establish the physical coordinate system of the image The origin is the center of the imaging plane. The perspective projection from the camera coordinates to the image physical coordinates is: , ,in, , These are the two direction axes of the image's physical coordinate system. The physical focal length of the camera;

[0036] Establish pixel coordinate system The conversion formula is: Taking the top-left corner of the image as the origin. , ,in, , These are the two direction axes of the pixel coordinate system. , For pixel physical size, , Let be the pixel coordinates of any point in the image.

[0037] In one embodiment of the present invention, step S4, based on the initial checkerboard calibration, continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber, and further includes:

[0038] Image coordinate correction is performed on the image in the recognition area within the fixed test chamber, specifically as follows:

[0039] Normalized image coordinates: , ,in, , The focal length is the pixel value.

[0040] Construct a radial distortion correction term while retaining the radial distortion parameters from the initial checkerboard calibration. , , :

[0041] ;

[0042] ;

[0043] in , The distance from the target point to the origin;

[0044] Construct a tangential distortion correction term while retaining the tangential distortion parameters from the initial checkerboard calibration. , :

[0045] ;

[0046] ;

[0047] The image coordinates after distortion correction are represented as follows:

[0048] ;

[0049] ;

[0050] in, , The x and y coordinates are the image after distortion correction.

[0051] In one embodiment of the invention, the radial distortion parameter is further included. , , and tangential distortion parameters , The update will be performed as follows:

[0052] Assuming the distortion parameters of the initial chessboard grid calibration It follows a Gaussian distribution;

[0053] Each time an object is successfully identified, the corner points of the image in the identified area are detected, and the reprojection error between the observed and theoretical values ​​is calculated.

[0054] ;

[0055] in, For reprojection error, , The x and y coordinates of the corner observations are: , The x and y coordinates represent the theoretical values ​​of the corner points;

[0056] Based on the reprojection error Construct the likelihood function:

[0057] ;

[0058] in, To preset lens distortion parameters as Under the premise of observing the true corner coordinates ( How likely is it that...? This refers to the error variance;

[0059] Posterior probability optimization: based on Bayes' theorem Maximizing the posterior probability is equivalent to minimizing the objective function:

[0060] ;

[0061] in, The regularization coefficient is used.

[0062] The distortion parameters of the initial chessboard grid are determined by minimizing the objective function. Update.

[0063] To address the aforementioned technical problems, this invention provides a light source-camera joint calibration and correction system for a fixed vision testing chamber, comprising:

[0064] Building Module: Used to build an additive physical model of the three light sources—built-in light, supplementary light, and ambient light—for the imaging process inside a fixed vision test chamber;

[0065] Parameter estimation module: used to turn off the fill light and estimate the parameters of ambient light in the three-source additive physical model;

[0066] Compensation module: used to solve the parameters of the supplementary light lamp in the three-source additive physical model, and send the supplementary light lamp parameters to the supplementary light lamp in real time to realize optical compensation for the fixed vision test box;

[0067] Optimization module: After optical compensation, based on the initial checkerboard calibration, it continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber.

[0068] To address the aforementioned technical problems, the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the light source-camera joint calibration and correction method for a fixed vision test chamber as described above.

[0069] To address the aforementioned technical problems, the present invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method for joint calibration and correction of a light source and camera for a fixed vision test chamber.

[0070] Compared with the prior art, the above-described technical solution of the present invention has the following advantages:

[0071] The light source-camera joint calibration and correction method for fixed vision test chamber described in this invention achieves color reproduction close to the imaging effect under standard D65 light source through physical modeling of three light sources and active supplementary lighting compensation, and significantly enhances the ability to resist mixed lighting interference, thus meeting the high-precision requirements of color recognition in teaching experiments.

[0072] The geometric accuracy of this invention is stable over a long period of time: the "one-time calibration" is changed to "continuous self-optimization", which effectively offsets the parameter drift of the lens caused by temperature and stress, and improves the long-term repeatability of geometric measurement results by more than 30%.

[0073] This invention is highly practical, requires no special hardware, and is easy to promote and apply in various fixed vision test chambers. Attached Figure Description

[0074] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0075] Figure 1 This is a flowchart of the method of the present invention;

[0076] Figure 2 This is a schematic diagram of the four-coordinate transformation relationship in an embodiment of the present invention. Detailed Implementation

[0077] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0078] Example 1

[0079] Reference Figure 1 As shown, this invention relates to a method for joint calibration and correction of a light source and camera for a fixed vision test chamber, comprising:

[0080] Step S1: Construct an additive physical model of the three light sources—built-in light, supplementary light, and ambient light—for the imaging process inside the fixed vision test chamber;

[0081] Step S2: Turn off the fill light and estimate the parameters of the ambient light in the three-source additive physical model;

[0082] Step S3: Solve for the parameters of the supplementary light lamp in the three-source additive physical model, and send the supplementary light lamp parameters to the supplementary light lamp in real time to realize optical compensation for the fixed vision test box;

[0083] Step S4: After optical compensation, based on the initial checkerboard calibration, the distortion parameters of the inherent recognition area in the fixed vision test chamber are continuously optimized.

[0084] The following is a detailed description of this embodiment:

[0085] 1. Hybrid Light Source Modeling and Co-calibration Module

[0086] The imaging process inside the test chamber is modeled as a linear superposition of three light sources to achieve active illumination compensation, as shown in the following formula:

[0087] (1) Physical model of additive imaging with three light sources

[0088]

[0089] in, This represents the final image pixel value; Pixel contribution for unknown ambient light; Contribute to built-in light source pixels with known characteristics (intensity, color temperature); Pixel contribution for controllable supplemental lighting (RGBW); , , The intensity coefficients of ambient light, built-in light, and supplementary light, respectively. (This is a fixed value, determined by the equipment parameters).

[0090] (2) Light source calibration mode (online estimation of ambient light parameters)

[0091] The system enters this mode periodically (e.g., upon daily startup) or manually, following the procedure:

[0092] Turn off the fill light. Only the built-in light source is used;

[0093] Photograph a standard neutral color reference plate (grayscale value 255) placed in the identification area (flat stage) of the test chamber;

[0094] Based on known and the actual measurement Inverse solution of ambient light intensity coefficient With color temperature parameters:

[0095]

[0096] in: The grayscale mean of the reference image (i.e., the mean of all pixels in an image); The standard grayscale average value of the reference board when illuminated by the built-in light source alone (pre-calibration); The standard reference luminance for ambient light is set based on the D65 standard light setting.

[0097] (3) Adaptive lighting strategy (active light compensation)

[0098] In normal operating mode, with the goal of identifying regions whose spectra are close to standard white light D65, the optimal supplementary light intensity coefficient is solved in real time. :

[0099] Objective function:

[0100]

[0101] Constraints:

[0102]

[0103]

[0104]

[0105] in, This is the identification area inside the test chamber, used to place the object to be identified. The target point on the object; , , The RGB channel intensity ratio of a standard D65 light source; , , These represent the pixel contributions for the three RGB channels of the target point, respectively. , , These represent the ambient light pixel contributions corresponding to the three RGB channels of the target point; , , These represent the contributions of the built-in light source pixels corresponding to the three RGB channels of the target point; , , These represent the pixel contributions of the fill light corresponding to the three RGB channels of the target point.

[0106] The solution obtained Data is sent to the fill light in real time, achieving millisecond-level optical compensation.

[0107] 2. Online fine-tuning module for geometric parameters based on scene priors

[0108] Based on the initial chessboard calibration, the distortion parameters are continuously optimized using the inherent recognition area of ​​the test chamber:

[0109] (1) Four-coordinate system linkage and simplified distortion model

[0110] Please see Figure 2 Establish a four-coordinate system adapted to the fixed test chamber and clarify the transformation relationships:

[0111] World coordinate system The origin is the inner corner of the lower left corner of the recognition area (stage). The platform plane is flat( Its physical dimensions are known fixed values ​​(such as length). ,Width ), , , These are the three directional axes of the world coordinate system;

[0112] Camera coordinate system With the camera's optical center as the origin The optical axis is Rigid body transformation from world coordinates to camera coordinates: ,in, for Any one of them, for Any one of them, It is a 3×3 rotation matrix. It is a 3×1 translation vector. , , These are the three orientation axes of the camera coordinate system;

[0113] Image physical coordinate system The origin is the center of the imaging plane. Perspective projection from camera coordinates to image physical coordinates: , ,in, , These are the two direction axes of the image's physical coordinate system. The physical focal length of the camera;

[0114] Pixel coordinate system The origin is the top left corner of the image. Conversion formula: , ,in, , These are the two direction axes of the pixel coordinate system. , For pixel physical size, , Let be the pixel coordinates of any point in the image.

[0115] Based on the initial checkerboard calibration, the distortion parameters of the inherent recognition area within the fixed vision test chamber are continuously optimized. This also includes image coordinate correction of the image in the recognition area within the fixed test chamber.

[0116] Normalized image coordinates: , ,in, , The focal length is the pixel value.

[0117] Construct a radial distortion correction term while retaining the radial distortion parameters from the initial checkerboard calibration. , , :

[0118]

[0119]

[0120] in , The distance from the target point to the origin;

[0121] Construct a tangential distortion correction term while retaining the tangential distortion parameters from the initial checkerboard calibration. , :

[0122]

[0123]

[0124] The image coordinates after distortion correction are represented as follows:

[0125]

[0126]

[0127] (2) This embodiment also includes radial distortion parameters , , and tangential distortion parameters , The update is as follows:

[0128] 1. The system pre-stores the precise physical dimensions (e.g., length) of the recognition area (stage). ,Width ) and world coordinates .

[0129] 2. Prior probability setting: distortion parameters based on the initial chessboard grid. Assuming the prior mean, the distortion parameter follows a Gaussian distribution: , This is the covariance matrix based on the initial calibration error. It should be noted that... Equivalent to the default value, if No adjustments are needed; it will be used directly. The value is taken as The purpose of this embodiment is to improve accuracy through subsequent iterations.

[0130] 3. Likelihood function construction: Each time an object is successfully identified, detect the corner points of the image in the recognition area (a total of 4), and calculate the reprojection error between the observed and theoretical values:

[0131]

[0132] in, For reprojection error, , The x and y coordinates of the corner observations are: , The x and y coordinates represent the theoretical values ​​(equivalent to the true values) of the corner points.

[0133] Based on the reprojection error Construct the likelihood function:

[0134]

[0135] in, To preset lens distortion parameters as Under the premise of observing the true corner coordinates ( How likely is it that...? This refers to the error variance;

[0136] 4. Posterior probability optimization: based on Bayes' theorem Maximizing the posterior probability is equivalent to minimizing the objective function:

[0137]

[0138] in, is the regularization coefficient.

[0139] The distortion parameters of the initial chessboard calibration are obtained by minimizing the objective function using the gradient descent algorithm. Update.

[0140] Example 2

[0141] This embodiment provides a light source-camera joint calibration and correction system for a fixed vision test chamber, including:

[0142] Building Module: Used to build an additive physical model of the three light sources—built-in light, supplementary light, and ambient light—for the imaging process inside a fixed vision test chamber;

[0143] Parameter estimation module: used to turn off the fill light and estimate the parameters of ambient light in the three-source additive physical model;

[0144] Compensation module: used to solve the parameters of the supplementary light lamp in the three-source additive physical model, and send the supplementary light lamp parameters to the supplementary light lamp in real time to realize optical compensation for the fixed vision test box;

[0145] Optimization module: After optical compensation, based on the initial checkerboard calibration, it continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber.

[0146] Example 3

[0147] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the light source-camera joint calibration and correction method for a fixed vision test chamber as described in Embodiment 1.

[0148] Example 4

[0149] This embodiment provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the light source-camera joint calibration and correction method for a fixed vision test chamber as described in Embodiment 1.

[0150] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0151] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0152] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0153] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0154] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for joint calibration and correction of a light source and camera for a fixed vision test chamber, characterized in that, include: Step S1: Construct an additive physical model of the three light sources—built-in light, supplementary light, and ambient light—for the imaging process inside the fixed vision test chamber; Step S2: Turn off the fill light and estimate the parameters of the ambient light in the three-source additive physical model; Step S3: Solve for the parameters of the supplementary light lamp in the three-source additive physical model, and send the supplementary light lamp parameters to the supplementary light lamp in real time to realize optical compensation for the fixed vision test box; Step S4: After optical compensation, based on the initial checkerboard calibration, the distortion parameters of the inherent recognition area in the fixed vision test chamber are continuously optimized.

2. The light source-camera joint calibration and correction method for a fixed vision test chamber according to claim 1, characterized in that: Step S1 constructs a three-source additive physical model of the imaging process inside the fixed vision test chamber, consisting of built-in light, supplementary light, and ambient light, which is represented as: ； in, This represents the final image pixel value; Pixel contribution for unknown ambient light; Contribute to built-in light pixels with known characteristics; Contribute to the pixels of the fill light; , , The intensity coefficients of ambient light, built-in light, and supplementary light are respectively... It is a fixed value.

3. The light source-camera joint calibration and correction method for a fixed vision test chamber according to claim 2, characterized in that: The method for turning off the fill light in step S2 and estimating the parameters of ambient light in the three-source additive physical model includes: Turn off the fill light, that is Only the built-in light of the fixed vision test chamber is retained; Photograph a reference board with a preset grayscale value placed in the recognition area of ​​the test chamber; Based on known and the actual measurement The ambient light intensity coefficient is obtained by inverse solving based on the additive physical model of three light sources. With color temperature parameters: ； in, The grayscale mean of the reference image; The standard grayscale average value of the reference board when illuminated by the built-in light alone; This is the standard reference luminance for ambient light.

4. The light source-camera joint calibration and correction method for a fixed vision test chamber according to claim 1, characterized in that: The method for solving the supplementary lighting parameters in the three-source additive physical model in step S3, and sending the supplementary lighting parameters to the supplementary lighting in real time to achieve optical compensation, includes: In normal operating mode, with the goal of ensuring that the spectrum of the recognition area in the fixed vision test chamber is close to that of standard white light D65, the optimal supplementary light intensity coefficient is solved in real time. : Objective function: ； Constraints: ； ； ； in, This is the identification area inside the test chamber, used to place the object to be identified. The target point on the object; , , The RGB channel intensity ratio of a standard D65 light source; , , These represent the pixel contributions for the three RGB channels of the target point, respectively. , , These represent the ambient light pixel contributions corresponding to the three RGB channels of the target point; , , These represent the contributions of the built-in light source pixels corresponding to the three RGB channels of the target point; , , These represent the pixel contributions of the fill light corresponding to the three RGB channels of the target point; The obtained supplementary lighting parameters Data is sent to the fill light in real time, achieving millisecond-level optical compensation.

5. The light source-camera joint calibration and correction method for a fixed vision test chamber according to claim 1, characterized in that: Step S4, based on the initial checkerboard calibration, continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber, including: Establish a four-coordinate system adapted to the fixed test chamber and clarify the transformation relationships, specifically as follows: Establish a world coordinate system The origin is the inner corner of the lower left corner of the stage in the identification area of ​​the test chamber. The platform plane is flat, The physical dimensions of the stage are known fixed values. , , These are the three directional axes of the world coordinate system; Establish camera coordinate system With the camera's optical center as the origin The optical axis is If the axes are defined, then the rigid body transformation from world coordinates to camera coordinates can be expressed as: ,in, for Any one of them, for Any one of them, It is a 3×3 rotation matrix. It is a 3×1 translation vector. , , These are the three orientation axes of the camera coordinate system; Establish the physical coordinate system of the image The origin is the center of the imaging plane. The perspective projection from the camera coordinates to the image physical coordinates is: , ,in, , These are the two directional axes of the image's physical coordinate system. The physical focal length of the camera; Establish pixel coordinate system The conversion formula is: Taking the top-left corner of the image as the origin. , ,in, , These are the two direction axes of the pixel coordinate system. , The physical size of a pixel. , Let be the pixel coordinates of any point in the image.

6. The light source-camera joint calibration and correction method for a fixed vision test chamber according to claim 5, characterized in that: Step S4, based on the initial checkerboard calibration, continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber, and also includes: Image coordinate correction is performed on the image in the recognition area within the fixed test chamber, specifically as follows: Normalized image coordinates: , ,in, , The focal length is the pixel value. Construct a radial distortion correction term while retaining the radial distortion parameters from the initial checkerboard calibration. , , : ； ； in , The distance from the target point to the origin; Construct a tangential distortion correction term while retaining the tangential distortion parameters from the initial checkerboard calibration. , : ； ； The image coordinates after distortion correction are represented as follows: ； ； in, , The x and y coordinates are the image after distortion correction.

7. The light source-camera joint calibration and correction method for a fixed vision test chamber according to claim 6, characterized in that: It also includes radial distortion parameters , , and tangential distortion parameters , The update will be performed as follows: Assuming the distortion parameters of the initial chessboard grid calibration It follows a Gaussian distribution; Each time an object is successfully identified, the corner points of the image in the identified area are detected, and the reprojection error between the observed and theoretical values ​​is calculated. ； in, For reprojection error, , The x and y coordinates of the corner observations are: , The x and y coordinates represent the theoretical values ​​of the corner points; Based on the reprojection error Construct the likelihood function: ； in, To preset lens distortion parameters as Under the premise of observing the true corner coordinates ( How likely is it that...? This refers to the error variance; Posterior probability optimization: based on Bayes' theorem Maximizing the posterior probability is equivalent to minimizing the objective function: ； in, The regularization coefficient is used. The distortion parameters of the initial chessboard grid calibration are obtained by minimizing the objective function. Update.

8. A light source-camera joint calibration and correction system for a fixed vision test chamber, characterized in that, include: Building Module: Used to build an additive physical model of the three light sources—built-in light, supplementary light, and ambient light—for the imaging process inside a fixed vision test chamber; Parameter estimation module: used to turn off the fill light and estimate the parameters of ambient light in the three-source additive physical model; Compensation module: used to solve the parameters of the supplementary light lamp in the three-source additive physical model, and send the supplementary light lamp parameters to the supplementary light lamp in real time to realize optical compensation for the fixed vision test box; Optimization module: After optical compensation, based on the initial checkerboard calibration, it continuously optimizes the distortion parameters of the inherent recognition area within the fixed vision test chamber.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the steps of the light source-camera joint calibration and correction method for a fixed vision test chamber as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the light source-camera joint calibration and correction method for a fixed vision test chamber as described in any one of claims 1 to 7.

Images