System and method of extracting camera parameter

Abstract

The present invention relates to a camera variable extraction system and method, and more specifically, to a camera variable extraction system and method that detects pixel coordinates of corner points from a center image captured so that a chart is positioned at the center of an image and at least one outer image captured so that the chart is positioned at the outer edge of an image by changing the camera orientation, and derives camera distortion coefficients by optimizing to minimize the reprojection error between the pixel coordinates of the corner points predicted according to a camera model and the pixel coordinates of the corner points actually detected.

Description

Technology Field

[0001] The present invention relates to a camera variable extraction system and method, and more specifically, to a camera variable extraction system and method that detects pixel coordinates of corner points from a center image captured so that a chart is positioned at the center of an image and at least one outer image captured so that the chart is positioned at the outer edge of an image by changing the camera orientation, and derives camera distortion coefficients by optimizing to minimize the reprojection error between the pixel coordinates of the corner points predicted according to a camera model and the pixel coordinates of the corner points actually detected. Background Technology

[0002] Since images captured using a camera contain various lens distortions, a camera calibration process is required to correct them.

[0003] For camera calibration, the relationship between the pixel coordinates of the image and the 3D coordinates of the real world must be determined, and to do this, a checkerboard with known 3D coordinates of the real world is used.

[0004] Meanwhile, the camera's intrinsic parameters consist of linear elements such as focal length and principal point, and non-linear elements such as the lens distortion coefficient.

[0005] Due to the distortion coefficient of the lens, the outer edges of an image captured by a camera are significantly distorted, unlike the center. Even when the field of view (FOV) is small, around 20 degrees, some distortion occurs at the edges of the image, and in the case of a fisheye lens, the distortion is even more severe. Due to this distortion, if a chart is located at the edge of the image, it may become difficult to detect corner points on a checkerboard chart.

[0006] To correct such distortion, distortion coefficients must be accurately derived. While the focal length and principal point among the camera's intrinsic parameters can be easily determined, accurately deriving distortion coefficients is difficult because they are affected by lens zoom-in / zoom-out movements or changes in camera orientation. Therefore, a technology capable of precisely estimating lens distortion coefficients is required. Prior art literature

[0007] Korean Published Patent No. 10-2023-0040458 (March 23, 2023) The problem to be solved

[0008] The problem to be solved by the present invention is to provide a camera variable extraction system and method that detects pixel coordinates of corner points from a center image captured so that a chart is positioned at the center of the image and at least one outer image captured so that the chart is positioned at the outer edge of the image by changing the camera orientation, and derives camera distortion coefficients by optimizing to minimize the reprojection error between the pixel coordinates of the corner points predicted according to a camera model and the pixel coordinates of the corner points actually detected. means of solving the problem

[0009] To solve the above-mentioned problem, the present invention provides a camera variable extraction system comprising: an image acquisition unit that collects a center image captured so that a chart is positioned at the center of the image and at least one outer image captured so that the chart is positioned at the outer edge of the image by changing the camera position; a corner detection unit that detects pixel coordinates of corner points, which are grid intersection points of the chart, in the center image and the outer image; and an optimization unit that derives a camera distortion coefficient by optimizing to minimize the sum of squares of reprojection errors, wherein the reprojection error is the error between the pixel coordinates predicted by projecting the 3D coordinates of the corner points into 2D according to the camera position and the detected pixel coordinates.

[0010] Preferably, the optimization unit applies a perspective camera model, projects the 3D coordinates of the corner point into 2D according to the camera pose and pinhole camera model, predicts pixel coordinates with a distortion coefficient applied, and determines the error between the predicted pixel coordinates and the detected pixel coordinates as a re-projection error.

[0011] Preferably, the optimization unit applies a generic camera model for a fisheye lens, applies a distortion coefficient to the 3D coordinates of the corner point, projects it into 2D according to the generic camera model to predict pixel coordinates, and determines the error between the predicted pixel coordinates and the detected pixel coordinates as the reprojection error.

[0012] Preferably, the outer images are four images obtained by changing the camera position so that the chart is located at the top left, top right, bottom left, and bottom right of the image, respectively, and the optimization unit can derive a camera distortion coefficient by optimizing to minimize the sum of the squares of the reprojection errors for all corner points of the center image and the four outer images.

[0013] Preferably, the corner detection unit can extract the pixel coordinates of the corner point closest to the center in the image and extract the pixel coordinates of the corner point by extending from the center to the surrounding outer edge through quadruple fitting.

[0014] In addition, to achieve the above-mentioned purpose, the present invention provides a method for extracting camera variables, comprising the steps of: (a) collecting a center image captured so that a chart is positioned at the center of an image, and at least one outer image captured so that the chart is positioned at the outer edge of an image by changing the camera position; (b) detecting the pixel coordinates of corner points, which are grid intersection points of the chart, in the center image and the outer image; and (c) deriving camera distortion coefficients by optimizing to minimize the sum of squares of reprojection errors, wherein the reprojection error is the error between the pixel coordinates predicted by projecting the 3D coordinates of the corner points into 2D according to the camera position and the detected pixel coordinates.

[0015] Meanwhile, to achieve the above-mentioned objective, the present invention provides a computer-readable recording medium that records a program for implementing the above-mentioned camera variable extraction method. Effects of the invention

[0016] According to the present invention, pixel coordinates of corner points are detected from a center image captured so that a chart is positioned at the center of the image and at least one outer image captured so that the chart is positioned at the outer edge of the image by changing the camera position, and by optimizing to minimize the reprojection error between the pixel coordinates of the corner points predicted according to the camera model and the pixel coordinates of the corner points actually detected, camera distortion coefficients can be accurately derived.

[0017] In addition, according to the present invention, the coordinates of the corner points can be accurately derived through quadratic fitting even when the corner points of the chart are located at the outer edge of the image. Brief explanation of the drawing

[0018] FIG. 1 is a block diagram showing a camera variable extraction system according to the present invention. FIG. 2 is a flowchart illustrating a camera variable extraction method according to the present invention. Figure 3 is a diagram illustrating the center image and the outer image in the present invention. FIG. 4 is a block diagram showing the corner detection unit of the present invention. FIG. 5 is a block diagram showing the optimization part of the present invention. Figure 6 is a diagram illustrating the reprojection error according to the perspective camera model and the generic camera model. Specific details for implementing the invention

[0019] The present invention is capable of various modifications and may take various forms, and specific embodiments are illustrated in the drawings and described in detail in the text. However, this is not intended to limit the invention to the specific disclosed forms, and it should be understood that the invention includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention.

[0020] Terms such as "first," "second," etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component.

[0021] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, terms such as "comprising" or "having" are intended to indicate the presence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0022] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which this invention pertains.

[0023] Terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0024] In addition, the term "at least one" is defined as a term including both singular and plural forms, and it is self-evident that even if the term "at least one" does not exist, each component may exist in a singular or plural form and may mean either singular or plural. Furthermore, whether each component is provided in a singular or plural form may be changed according to the embodiment.

[0025] Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings. Identical reference numerals in each drawing indicate identical components.

[0026] FIG. 1 is a block diagram showing a camera variable extraction system according to the present invention, and FIG. 2 is a flowchart showing a camera variable extraction method according to the present invention. Since the camera variable extraction method according to the present invention is essentially the same as the camera variable extraction system according to the present invention, they will be described together below.

[0027] In the present invention, a chart (10) displaying a checkerboard and a camera (20) capturing the chart (10) may, of course, constitute part of a camera variable extraction system (100) according to the present invention.

[0028] Referring to FIG. 1, the camera variable extraction system (100) according to the present invention includes an image acquisition unit (110), a corner detection unit (120), and an optimization unit (130).

[0029] The image collection unit (110) is a component that collects a center image and an outer image, and receives and stores from the camera (20) a center image taken so that the chart (10) is positioned at the center of the image and at least one outer image taken so that the chart (10) is positioned at the outer edge of the image by changing the camera position (S100).

[0030] Although there is no significant limit to the number of outer images, it is preferable to use four outer images by changing the camera position so that the chart (10) is located at the top left, top right, bottom left, and bottom right of each image.

[0031] Figure 3 is a diagram illustrating the center image and the outer image in the present invention.

[0032] Referring to FIG. 3, the first image in the upper left is an outer image taken with the chart (10) positioned at the lower right, the second image in the upper right is an outer image taken with the chart (10) positioned at the lower left, the third image in the center is a center image taken with the chart (10) positioned at the center of the image, the fourth image in the lower right is an outer image taken with the chart (10) positioned at the upper left, and the fifth image in the lower left is an outer image taken with the chart (10) positioned at the upper right. As can be seen in FIG. 3, it can be seen that the distortion of the lens becomes more severe as it moves toward the outer edges.

[0033] The corner detection unit (120) detects pixel coordinates in the image of a corner point, which is a grid intersection point within the chart, from the center image and the outer image (S200).

[0034] On the other hand, in the center and each outer image, distortion is minimal in the central part of the image, so there is no problem in detecting the pixel coordinates of corner points; however, as one moves toward the outer parts of the image, distortion becomes severe and resolution drops, making it difficult to accurately detect the pixel coordinates of corner points even when the pixels are magnified.

[0035] To this end, the corner detection unit (120) can extract the pixel coordinates of the corner point closest to the center in the image and extract the pixel coordinates of the corner point by extending from the center of the image to the surrounding outer edge through quadratic fitting.

[0036] FIG. 4 is a block diagram showing the corner detection unit of the present invention.

[0037] Referring to FIG. 4, the corner detection unit (120) includes a center extraction unit (121), a fitting window generation unit (122), a quadrature fitting unit (123), and an expansion extraction unit (124).

[0038] The center extraction unit (120) extracts the center from the image of the chart (10) and extracts the coordinates of the corner point closest to the extracted center.

[0039] The center of an image can be easily extracted, and corner points, which are the intersections of a checkerboard, can be extracted through various algorithms such as Harris corner detection, edge and corner detection through the convolution of a derivative filter, and finding the peak value of a corner. Since extracting the coordinates of a corner point is common and can be done in any way, a detailed explanation will be omitted. What is important in this invention is to find the center of an image capturing a chart, extract the corner point closest to that center, and extract the pixel coordinates of that corner point.

[0040] The fitting window generation unit (122) extracts multiple corner points around the center and generates a fitting window having the extracted multiple corner points (S200).

[0041] Here, the fitting window can be created to have at least 9 corner points. The fitting window can be created in the shape of a square, rectangle, or rhombus.

[0042] In the case of a square, it can be implemented as 3 X 3, 4 X 4, or 5 X 5 depending on the number of horizontal or vertical corner points.

[0043] The quadrature fitting unit (123) performs quadrature fitting by calculating the pixel coordinates of corner points around the fitting window generated by the fitting window generation unit (122) through a quadrature function.

[0044] The following [Mathematical Equation 1] is an example of a quadrature function.

[0045]

[0046] In the quadrature function used in the present invention, A to F are coefficients, and approximate solutions for A to F are obtained using the coordinates of corner points around a center whose coordinates are already known, and the coordinate positions of the surroundings are predicted using this.

[0047] Due to distortion of the camera lens, the distortion becomes parabolic as it moves from the center to the periphery, and the grid becomes slightly distorted from the center to the periphery. Therefore, by using a quadrilateral function to predict and calculate the coordinates of the corner points in the periphery and expanding from the center, it is possible to accurately obtain the coordinate positions even in cases where it is difficult to accurately obtain the coordinates of the corner points due to reduced resolution, as shown in Fig. 1.

[0048] The expansion extraction unit (124) can calculate the coordinates of all corner points by moving the fitting window by the corner point interval.

[0049] As described above, the coordinate positions of corner points around the fitting window can be predicted and obtained one by one through a quadrature function. When the pixel coordinate positions of corner points corresponding to one column around the fitting window are predicted, the fitting window can be moved one column at a time, and the coordinate positions of corner points can be predicted and calculated while continuously expanding the surroundings.

[0050] At this time, it goes without saying that the fitting window can be moved in one direction by the distance of the corner points, in a diagonal direction, or in a zigzag direction.

[0051] The optimization unit (130) derives a camera distortion coefficient by optimizing to minimize the sum of the squares of the re-projection error (S300). Here, the re-projection error is the error between the pixel coordinates predicted by projecting the 3D coordinates of the corner point into 2D according to the camera position and the pixel coordinates detected in the actual image.

[0052] Meanwhile, the optimization process of this optimization unit (130) is different for a model applied to a lens with a narrow field of view (FOV) of 20 degrees or less and a model applied to a lens with a wide field of view, such as a fisheye lens.

[0053] For lenses with a narrow field of view, a perspective camera model with a pinhole camera is applied, and for lenses with a wide field of view, such as fisheye lenses or wide-angle lenses, a generic camera model is applied.

[0054] FIG. 5 is a block diagram showing the optimization part of the present invention.

[0055] Referring to FIG. 5, the optimization unit (130) includes a perspective application unit (131) and a generic application unit (132).

[0056] The perspective application unit (131) applies a perspective camera model, projects the 3D coordinates of the corner point into 2D according to the camera position and pinhole camera model, predicts pixel coordinates with a distortion coefficient applied, and determines the error between the predicted pixel coordinates and the detected pixel coordinates as the re-projection error.

[0057] The generic application unit (132) applies a generic camera model for a fisheye lens, applies a distortion coefficient to the 3D coordinates of the corner point, and then projects it into 2D according to the generic camera model to predict pixel coordinates, and determines the error between the predicted pixel coordinates and the detected pixel coordinates as the re-projection error.

[0058] Figure 6 is a diagram illustrating the reprojection error according to the perspective camera model and the generic camera model.

[0059] (A) is based on the perspective camera model, and (B) is based on the generic camera model.

[0060] (A) can be used when the field of view is small, about 20 degrees.

[0061] The 3D coordinates (X,Y,Z) of the chart's corner points are 2D coordinates (x) according to the pinhole camera model u , y u Projected as ), and depending on the lens distortion coefficient, 2D coordinates (x d , y d Ideally, it should be distorted into ) . However, the corresponding corner point is pixel coordinates (x in the actual image c , y c It forms in ).

[0062] The difference between the ideal predicted coordinates and the pixel coordinates in the actual image is the reprojection error, and in the optimization process, the distortion coefficient of the camera lens can be derived through an optimization process that minimizes the sum of the squares of the reprojection errors of countless corner points collected and analyzed from the image.

[0063] [Equation 2] below is the conversion equation between the world coordinate system and the camera coordinate system according to the pinhole camera model.

[0064]

[0065] (X,Y,Z) are 3D coordinates in the world coordinate system, and (x u ,y u ) is an ideal 2D projected pixel coordinate that is undistorted, that is, without any distortion coefficients applied.

[0066] And, f is the focal length of the lens, and (pp x , pp y ) is the principal point coordinate of the lens, R is the rotation matrix, and t is the transformation vector.

[0067] Meanwhile, (x d , y d ) is the ideal 2D projected pixel coordinate with distortion factors applied, and (x c , y c If ) represents the actually detected pixel coordinates, then the distortion coefficients k1~k n It has the same relationship as the following [Mathematical Equation 3].

[0068]

[0069] At this time, And, is. Here, r u is (x u ,y u It is the distance from ) to the origin, and r d is (x d , y d It is the distance from ) to the origin.

[0070] (B) is based on a generic camera model applied to lenses with a wide field of view, such as wide-angle lenses or fisheye lenses. In the case of a generic camera model, lens distortion is first applied to the 3D coordinates of the corner points, and then projected according to the generic camera module to obtain predicted pixel coordinates.

[0071] In a generic camera model, the world coordinate system and the normalized world coordinate system have the following relationship [Equation 4].

[0072]

[0073] Here, (X,Y,Z) is the original world coordinate system, and (X u ,Y u ,Z u ) is the normalized world coordinate.

[0074] is the angle that (X,Y,Z) makes with the z-axis, and is (X u ,Y u ,Z u It is the angle that ) makes with the z-axis, and is (X d ,Y d ,Z d If ) is the angle formed with the z-axis, each angle is as follows [Equation 5].

[0075]

[0076] At this time, (X d ,Y d ,Z d ) and (x d ,y d ) has a relationship as shown in the following [Equation 6].

[0077]

[0078] Meanwhile, in the case of (A) and (B), the reprojection error ( ) is as follows.

[0079]

[0080] Ultimately, by inputting initial values ​​and repeating the process of minimizing the sum of the squares of the reprojection errors using [Equation 3] and [Equation 5], the optimized distortion coefficients k1~k n It can derive.

[0081] The present invention may be implemented as a method, device, system, etc. The method of the present invention may also be implemented as a computer-readable program, code, application, software, etc. on a computer-readable recording medium.

[0082] When implemented and executed as software or an application, the constituent means of the present invention are programs or code segments that inevitably perform necessary tasks. For example, the image acquisition unit (110), corner detection unit (120), and optimization unit (130) in the present invention may be composed of program modules, code segments, and / or data that implement the functions of each component, and each of these components is formed to be stored in a memory device and executed by one or more processors.

[0083] Additionally, programs or code segments may be stored on a computer-readable recording medium or transmitted by computer data signals combined with a carrier wave on a transmission medium or communication network.

[0084] Computer-readable recording media include all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Additionally, computer-readable recording media may be distributed across networked computer systems, allowing computer-readable code to be stored and executed in a distributed manner.

[0085] In addition, the camera variable extraction system and method according to an embodiment of the present invention may be implemented with a hardware component, a software component, and / or a combination of the hardware component and the software component.

[0086] The above devices and components may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions.

[0087] The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible.

[0088] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or command the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0089] A method according to an embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the embodiment, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

[0090] Although the present invention has been described above with reference to specific embodiments illustrated in the drawings, this is merely illustrative. Therefore, those skilled in the art to which the present invention pertains will be able to make various substitutions, additions, and modifications within the scope of the technical concept described above without departing from it. Accordingly, the scope of protection of the present invention should be interpreted by the claims set forth below, and all technical concepts within the equivalent and equivalent scope should be interpreted as being included within the scope of protection of the present invention. Explanation of the symbols

[0091] 10: Chart 20: Camera 110: Video Acquisition Unit 120: Corner detection unit 121: Center extraction unit 122: Fitting Window Generation Unit 123: Quadruple Fitting Unit 124: Extended Extraction Section 130: Optimization Section 131: Perspective Application Section 132: Generic Application Section

Claims

Claim 1 An image acquisition unit that collects a center image captured so that the chart is positioned at the center of the image, and at least one outer image captured by changing the camera position so that the chart is positioned at the outer edge of the image; a corner detection unit that detects pixel coordinates of corner points, which are grid intersection points of the chart, in the center image and the outer image; The method includes an optimization unit that derives camera distortion coefficients by optimizing to minimize the sum of squares of reprojection errors, wherein the reprojection error is the error between the pixel coordinates predicted by projecting the 3D coordinates of the corner point into 2D according to the camera orientation and the detected pixel coordinates, and the outer images are four images obtained by changing the camera orientation so that the chart is located at the top-left, top-right, bottom-left, and bottom-right of the image, respectively, and the optimization unit derives camera distortion coefficients by optimizing to minimize the sum of squares of reprojection errors for all corner points of the center image and the four outer images, wherein the corner detection unit does not independently detect corner points in the outer images, but detects the pixel coordinates of the corner points that are grid intersection points of the chart in the center image, performs quadruple fitting on the detected corner points to extract corner points at the subpixel level, predicts the next corner point in the outer direction based on the extracted corner points, and repeats the process of performing quadruple fitting again on the predicted next corner points. A camera variable extraction system characterized by extending and extracting the pixel coordinates of corner points to the outer edge of the above outer image. Claim 2 A camera variable extraction system according to claim 1, wherein the optimization unit applies a perspective camera model, projects the 3D coordinates of the corner point into 2D according to the camera pose and pinhole camera model, predicts pixel coordinates with a distortion coefficient applied, and determines the error between the predicted pixel coordinates and the detected pixel coordinates as a re-projection error. Claim 3 A camera variable extraction system according to claim 1, wherein the optimization unit applies a generic camera model for a fisheye lens, predicts pixel coordinates by applying a distortion coefficient to the 3D coordinates of the corner point and projecting it into 2D according to the generic camera model, and determines the error between the predicted pixel coordinates and the detected pixel coordinates as a re-projection error. Claim 4 delete Claim 5 delete Claim 6 A method performed by a camera variable extraction system, comprising: (a) collecting a center image captured so that a chart is positioned at the center of the image, and at least one outer image captured so that the chart is positioned at the outer edge of the image by changing the camera position; (b) detecting the pixel coordinates of corner points, which are grid intersection points of the chart, in the center image and the outer image; and (c) a step of deriving camera distortion coefficients by optimizing to minimize the sum of squares of reprojection errors, wherein the reprojection error is the error between the pixel coordinates predicted by projecting the 3D coordinates of the corner point into 2D according to the camera pose and the detected pixel coordinates, and the outer images are four images obtained by changing the camera pose so that the chart is located at the top-left, top-right, bottom-left, and bottom-right of the image, respectively, and the step (c) derives camera distortion coefficients by optimizing to minimize the sum of squares of reprojection errors for all corner points of the center image and the four outer images, wherein the step (b) does not independently detect corner points in the outer images, but detects the pixel coordinates of the corner points that are grid intersection points of the chart in the center image, performs quadruple fitting on the detected corner points to extract corner points at the subpixel level, predicts the next corner point in the outer direction based on the extracted corner points, and performs quadruple fitting again on the predicted next corner point A camera variable extraction method characterized by repeating the process to extend and extract pixel coordinates of corner points to the outer edge of the outer image. Claim 7 A method for extracting camera variables according to claim 6, wherein a perspective camera model is applied to step (c), the 3D coordinates of the corner point are projected into 2D according to the camera pose and pinhole camera model, the pixel coordinates with a distortion coefficient applied are predicted, and the error between the predicted pixel coordinates and the detected pixel coordinates is determined as the re-projection error. Claim 8 A method for extracting camera variables according to claim 6, wherein a generic camera model for a fisheye lens is applied in step (c), and a distortion coefficient is applied to the 3D coordinates of the corner point, and then projected into 2D according to the generic camera model to predict pixel coordinates, and the error between the predicted pixel coordinates and the detected pixel coordinates is determined as a re-projection error. Claim 9 delete Claim 10 delete

Images