KR102412275B1 - Image Distortion Correction Method and Apparatus in Measurement of Three Dimensional Shape Information using Stereo Method - Google Patents

Abstract

A method and apparatus for correcting image distortion in 3D shape information measurement using a stereo method are presented. An image distortion correction method performed by a computer device according to an embodiment includes: generating a plurality of PCG patterns including a first pattern set and a second pattern set; estimating a distortion center point by photographing a first set of patterns among the PCG patterns, detecting lattice points, and linearly regressing lattice points constituting one row or column among the detected lattice points; After photographing the second set of patterns among the PCG patterns, lattice points are detected, an area within a predetermined range from the distortion center point in the image is defined as a micro-distortion area, and lattice points within the micro-distortion area are detected in the row direction or estimating a vanishing point in a column direction; obtaining distortion amount data by performing distortion correction on all the grid points of the first pattern set based on the estimated vanishing point; and measuring the distortion amount map of the camera lens by curve fitting the distortion amount data scattered over the entire image area.

Description

Image Distortion Correction Method and Apparatus in Measurement of Three Dimensional Shape Information using Stereo Method}

The following embodiments relate to a method and apparatus for correcting image distortion in 3D shape information measurement using a stereo method, and more particularly, a method and apparatus for estimating a distortion amount map of a camera in advance for 3D stereo measurement is about

The field of application using a camera is diverse throughout the industry, and the field of application is becoming increasingly vast. For example, a camera such as robot control of an assembly line, product quality inspection, diagnosis in the medical field, security system, recognition of an image system, etc., is being used in industrial automation instead of the human eye.

In general, to obtain a high-precision image of an object, a narrow-angle lens is installed in front of the camera, and a wide-angle lens is used when the object is large and it is necessary to observe a wide area. device and use it. In the case of using a wide-angle lens, it has the advantage of securing a wide field of view, but there is a disadvantage in that the central portion of the lens has high resolution and the resolution gradually decreases toward the outer angle portion of the lens.

In addition, the wide-angle lens exhibits remarkable distortion, ie, radial distortion, in which the image is curved toward the outer part along with a decrease in the resolution of the outer part. Such radiation distortion is also a major factor in the degradation of resolution.

Korean Patent No. 10-1014572 describes a technique for such an image distortion correction method and an image processing apparatus employing the image distortion correction method.

Korean Patent Registration No. 10-1014572

Tan, L.; Wang, Y.; Yu, H. S.; Zhu, J. Automatic camera calibration using active displays of a virtual Pattern. Sensors 2017, 17, 685.

The embodiments describe a method and apparatus for correcting image distortion in 3D shape information measurement using a stereo method, and more specifically, provide a technique for estimating a distortion amount map of a camera in advance for 3D stereo measurement.

The embodiments provide image distortion correction in 3D shape information measurement using a stereo method that can solve the problem that the calibration area and the 3D precision measurement area of the image are separated in a large-scale situation by estimating the distortion amount map in advance at a short distance. To provide a method and apparatus.

An image distortion correction method performed by a computer device according to an embodiment includes: generating a plurality of PCG patterns including a first pattern set and a second pattern set; estimating a distortion center point by photographing a first set of patterns among the PCG patterns, detecting lattice points, and linearly regressing lattice points constituting one row or column among the detected lattice points; After photographing the second set of patterns among the PCG patterns, lattice points are detected, an area within a predetermined range from the distortion center point in the image is defined as a micro-distortion area, and lattice points within the micro-distortion area are detected in the row direction or estimating a vanishing point in a column direction; obtaining distortion amount data by performing distortion correction on all the grid points of the first pattern set based on the estimated vanishing point; and measuring the distortion amount map of the camera lens by curve fitting the distortion amount data scattered over the entire image area.

The first pattern set may be existing PCG patterns, and the second pattern set may be new PCG patterns.

The generating of the PCG patterns may include: generating a virtual PCG pattern; displaying an image projected by a virtual camera of the virtual PCG pattern on a monitor; and acquiring an image by photographing the PCG pattern displayed on the monitor with an actual camera.

The generating of the PCG patterns may include infinitely generating virtual PCG patterns having a specific rotation matrix and a specific translation matrix based on the virtual camera through a pattern generation method using a virtual camera.

The first pattern set may be a pattern that allows the location of grid points to be detected even in an image in which out-of-focus occurs.

In the estimating of the distortion center point, it is assumed that the straight line of the grid points closest to the linearly regressed straight line passes through the distortion center point, and the intersection of the straightest lines among the row and column direction straight lines is estimated as the distortion center point. can do.

In the step of obtaining distortion amount data by performing distortion correction, distortion correction may be performed using perspective projection invariance for all the grid points of the first pattern set based on the estimated vanishing point.

As the PCG patterns photographed with the camera fixed are generated at different positions in the monitor using computer software, distortion correction can be automatically measured.

a PCG pattern generator for generating a plurality of PCG patterns including a first pattern set and a second pattern set according to another embodiment; a distortion center point estimator that detects grid points after photographing a first set of patterns among the PCG patterns, and linearly regresses grid points constituting one row or column among the detected grid points to estimate a distortion center point; After photographing the second set of patterns among the PCG patterns, lattice points are detected, an area within a predetermined range from the distortion center point in the image is defined as a micro-distortion area, and lattice points within the micro-distortion area are detected in the row direction or a vanishing point estimator for estimating a vanishing point in a column direction; a distortion correction unit for obtaining distortion amount data by performing distortion correction on all the grid points of the first pattern set based on the estimated vanishing point; and a distortion amount map measurer configured to measure the distortion amount map of the camera lens by curve fitting the distortion amount data scattered over the entire image area.

The first pattern set may be existing PCG patterns, and the second pattern set may be new PCG patterns.

The PCG pattern generator may generate a virtual PCG pattern, display an image obtained by projecting the virtual PCG pattern with a virtual camera on a monitor, and acquire an image by photographing the PCG pattern displayed on the monitor with a real camera.

The PCG pattern generator may indefinitely generate a virtual PCG pattern having a specific rotation matrix and a specific translation matrix based on the virtual camera through a pattern generation method using a virtual camera.

The first pattern set may be a pattern that allows the location of grid points to be detected even in an image in which out-of-focus occurs.

The distortion center point estimator may assume that the straight line of the grid points closest to the linearly regressed straight line passes through the distortion center point, and the point of intersection of the straightest straight lines among the row and column direction straight lines can be estimated as the distortion center point. .

The distortion correction unit may perform distortion correction using perspective projection invariance for all the grid points of the first pattern set based on the estimated vanishing point.

According to embodiments, by estimating the distortion amount map in advance at a short distance, an image in 3D shape information measurement using a stereo method that can solve the problem that the calibration area and the 3D precision measurement area of the image are separated in a large-scale situation A distortion correction method and apparatus may be provided.

1 is a diagram schematically illustrating an image distortion correction method according to an exemplary embodiment.
2 is a diagram schematically illustrating a method for correcting image distortion of a plurality of virtual PCG patterns according to an exemplary embodiment.
3 is a diagram illustrating a method for measuring a distortion amount map according to an exemplary embodiment.
4 is a flowchart illustrating an image distortion correction method according to an exemplary embodiment.
5 is a block diagram illustrating an apparatus for correcting image distortion according to an exemplary embodiment.
6 is a diagram illustrating an example of an existing grid point set and a new grid point set according to an embodiment.
7 is a diagram illustrating an example of estimating a distortion center point according to an embodiment.
8 is a diagram illustrating an example of estimating a micro-distortion region and a vanishing point according to an exemplary embodiment.
9 is a diagram for describing a method for estimating a micro-distortion region and a vanishing point according to an exemplary embodiment.
10 is a diagram for explaining a distortion correction calculation by perspective projection invariance according to an exemplary embodiment.
11 is a diagram for explaining a method of fitting a surface with distortion amount data according to an exemplary embodiment.

Hereinafter, embodiments will be described with reference to the accompanying drawings. However, the described embodiments may be modified in various other forms, and the scope of the present invention is not limited by the embodiments described below. In addition, various embodiments are provided in order to more completely explain the present invention to those of ordinary skill in the art. The shapes and sizes of elements in the drawings may be exaggerated for clearer description.

The following embodiments relate to an image distortion correction method and apparatus in 3D shape information measurement using a stereo method, and a distortion amount map of a camera can be estimated in advance for 3D stereo measurement.

The embodiments are specialized for large-scale situations, and in order to solve the problem of separation of the calibration area and the three-dimensional precision measurement area of the image in the large-scale situation, the distortion map is estimated in advance at a short distance.

Problems that occur in stereo measurement on a large scale can be classified into two categories. First, the separation of the calibration area and the precision measurement area occurs. Second, a wide-angle lens must be used for large-scale photography, and accordingly, a measurement error due to lens distortion is greatly generated.

Embodiments measure the amount of distortion of the camera in advance in order to solve these two problems, so that when performing stereo measurement on a large scale, 3D measurement can be performed more accurately than conventional methods.

1 is a diagram schematically illustrating an image distortion correction method according to an exemplary embodiment. Also, FIG. 2 is a diagram schematically illustrating a method for correcting image distortion of a plurality of virtual PCG patterns according to an exemplary embodiment.

1 and 2 , the image distortion correction method according to an embodiment may generate a virtual PCG pattern 130 and display an image projected by the virtual camera 110 on the monitor 120 . The image 150 may be obtained by photographing the PCG pattern 130 displayed on the monitor 120 with the actual camera 140 .

Here, PCG pattern 1 is a pattern that allows the location of grid points to be detected even in an out-focused image. Out-of-focus occurs when the camera is focused on the hull block at a distance of about 10 m and needs to see an object at a short distance (60 cm). Therefore, it does not recognize the grid points of a general checkboard. Therefore, the PCG pattern can be used.

For high-precision distortion correction, the perspective projection invariance method, which is a relatively precise method among distortion correction methods, may be used. This can correct the position of the distorted grid points in the image using linearity, cross-ratio, and vanishing point, which are properties that the grid points of the checkboard must satisfy in the image.

In the present embodiment, instead of the check board, the grid points of the PCG pattern may be detected and distortion correction of the grid points may be performed. The difference between the distorted point and the corrected point corresponds to the amount of distortion for one grid point. The amount of distortion data is generated as many as the number of grid points per pattern. This corresponds to 54 /1 patterns in the case of FIG. 1 and is a fairly small number considering that the image pixels are 3036 X 4024. Therefore, the number of distortion amounts to obtain the distortion amount of the entire image area is too small for one pattern.

In order to generate an indefinite pattern, a pattern generation method using a virtual camera (Non-Patent Document 1) can be combined. Based on the virtual camera, a virtual pattern having a specific rotation matrix and a specific translation matrix can be created indefinitely. Since the specific rotation matrix and the specific translation matrix are values provided by the user through computer software, they can be set arbitrarily. That is, the position and posture of the virtual PCG pattern in the monitor can be freely changed to fit the entire image area.

In the existing method, the PCG pattern was displayed on the monitor and the camera was moved directly. On the other hand, in the present embodiments, the distortion correction can be automatically measured as PCG patterns are generated at different positions in the monitor by computer software while the camera is fixed.

In addition, the existing method of estimating the amount of distortion map using the grid points of the checkboard had a limitation in the number of patterns because it was performed manually. The number of is markedly insufficient. On the other hand, in the embodiments, virtual patterns are created indefinitely in arbitrary postures and positions with computer software, and the number of distortion data is infinitely increased. Accordingly, it is possible to more accurately estimate the distortion amount map of the entire image area.

3 is a diagram illustrating a method for measuring a distortion amount map according to an exemplary embodiment.

Referring to FIG. 3 , in order to measure the amount of distortion map, a virtual PCG pattern is first generated 310 , and an image projected by a virtual camera is displayed on the monitor 320 , and then the PCG pattern displayed on the monitor An image may be obtained by photographing 330 with a real camera. In this case, distortion amount correction may be performed ( 340 ), and the distortion amount may be calculated ( 350 ) and reflected in the virtual PCG pattern, thereby more accurately estimating the distortion amount map of the entire image area.

4 is a flowchart illustrating an image distortion correction method according to an exemplary embodiment.

Referring to FIG. 4 , the image distortion correction method performed by a computer device according to an embodiment includes generating a plurality of PCG patterns including a first pattern set and a second pattern set ( 410 ), the PCG patterns Detecting grid points after photographing the first pattern set, linearly regressing grid points constituting one row or column among the detected grid points to estimate a distortion center point ( 420 ), a second pattern among PCG patterns Detecting grid points after photographing the set, defining an area within a preset range from the distortion center point in the image as a micro-distortion area, and estimating the vanishing point in the row or column direction by detecting the grid points within the micro-distortion area ( 430), obtaining distortion amount data by performing distortion correction on all lattice points of the first pattern set based on the estimated vanishing point (440), and curve fitting the distortion amount data scattered over the entire image area The method may include measuring a distortion amount map of the camera lens ( 450 ).

Below, each step of the image distortion correction method according to an embodiment will be described in more detail.

The image distortion correction method according to an embodiment may be described using the image distortion correction apparatus according to the embodiment.

5 is a block diagram illustrating an apparatus for correcting image distortion according to an exemplary embodiment.

Referring to FIG. 5 , an image distortion correction apparatus 500 according to an embodiment includes a PCG pattern generator 510 , a distortion center point estimator 520 , a vanishing point estimator 530 , a distortion corrector 540 , and distortion. It may include an amount map measuring unit 550 .

In operation 410 , the PCG pattern generator 510 may generate a plurality of PCG patterns including the first pattern set and the second pattern set. Here, the first pattern set may be existing PCG patterns, and the second pattern set may be new PCG patterns. For example, the first pattern set may be a pattern that enables the location of grid points to be detected even in an image in which out-of-focus occurs.

The PCG pattern generator 510 may indefinitely generate a virtual PCG pattern having a specific rotation matrix and a specific translation matrix based on the virtual camera through a pattern generation method using a virtual camera. In addition, the PCG pattern generator 510 generates a virtual PCG pattern, displays an image projected by the virtual camera with the virtual PCG pattern, on the monitor, and then captures the PCG pattern displayed on the monitor with a real camera to obtain an image. can

In step 420, the distortion center point estimator 520 detects lattice points after photographing the first pattern set among the PCG patterns, and linearly regresses the lattice points constituting one row or column among the detected lattice points. The distortion center point can be estimated.

Here, the distortion center point estimator 520 assumes that the straight line of the grid points closest to the linearly regressed straight line passes the distortion center point, and estimates the intersection of the straight lines with the most linearity among the row and column direction straight lines as the distortion center point. can do.

In step 430, the vanishing point estimator 530 detects lattice points after photographing the second pattern set among the PCG patterns, defines an area within a preset range from the distortion center point in the image as a micro-distortion area, and micro-distortion A vanishing point in a row direction or a column direction can be estimated by detecting grid points in the region.

In operation 440 , the distortion correction unit 540 may obtain distortion amount data by performing distortion correction on all grid points of the first pattern set based on the estimated vanishing point. The distortion correction unit 540 may perform distortion correction on all lattice points of the first pattern set based on the estimated vanishing point using perspective projection invariance.

In operation 450 , the distortion amount map measuring unit 550 may measure the distortion amount map of the camera lens by curve fitting the distortion amount data scattered over the entire image area.

In this way, as PCG patterns photographed while the camera is fixed are generated at different positions in the monitor using computer software, distortion correction can be automatically measured.

Hereinafter, an image distortion correction method according to an exemplary embodiment will be described in more detail with reference to the drawings.

Step 410 of the detailed distortion correction process first generates a myriad of PCG patterns. At this time, the patterns may be composed of a total of two sets. For example, 250 sheets, co-rotation matrix, various translation matrices, pattern set A (first pattern set) of the existing grid point set and 50 sheets of co-rotation matrix, various translation matrices, pattern set B of the new grid point set (second pattern set). pattern set).

6 is a diagram illustrating an example of an existing grid point set and a new grid point set according to an embodiment.

As shown in FIG. 6 , the new set of grid points may add points that divide the interval between each grid point of the existing grid point into quarters.

7 is a diagram illustrating an example of estimating a distortion center point according to an embodiment.

Step 420 is a step of estimating a distortion center point. First, after all pattern set A is photographed, grid points are detected. Among the detected grid points, the grid points constituting one row (or column) are linearly regressed. The closer to the distortion center point, the smaller the distortion effect, so a straight line shows a straight line, and the farther away it shows a curved shape. Therefore, as shown in FIG. 7 , it can be assumed that the straight line of the grid points closest to the linearly regressed straight line passes through the distortion center point, and the intersection of the straight lines with the most linearity among the row and column direction straight lines is the distortion center point. It can be assumed that

8 is a diagram illustrating an example of estimating a micro-distortion region and a vanishing point according to an exemplary embodiment. Also, FIG. 9 is a diagram for explaining a method of estimating a micro-distortion region and a vanishing point according to an exemplary embodiment.

Step 430 is a vanishing point estimation step. After all pattern set B is photographed, grid points are detected. In the image, a region 5% from the distortion center point can be assumed to be relatively accurate positions of grid points because the amount of distortion is small. On the other hand, here, although it is limited to a region 5% from the center of the distortion, the range is adjustable depending on the embodiment. Accordingly, as shown in FIGS. 8 and 9 , this is defined as a micro-distortion region, and vanishing points in the row and column directions can be estimated by detecting lattice points in the micro-distortion region.

In step 440, distortion correction of all grid points of the pattern set A is performed using perspective projection invariance based on the estimated vanishing points. Distortion data can be obtained for all 250 grid points, which fills the entire image area.

10 is a diagram for explaining a distortion correction calculation by perspective projection invariance according to an exemplary embodiment. As shown in FIG. 10 , the amount of distortion (correction amount) may be calculated using perspective projection invariance.

In operation 450 , the distortion amount map of the camera lens may be measured by curve fitting the distortion amount data scattered over the entire image area.

11 is a diagram for explaining a method of fitting a surface with distortion amount data according to an exemplary embodiment. As shown in FIG. 11 , surface fitting can be performed with all distortion amount (correction amount) data.

The embodiments can be applied to manufacturing automation in manufacturing, and in particular, can be directly applied to three-dimensional coordinate measurement technology in situations where large-scale manufacturing is required, such as shipbuilding and construction. Also, it is possible to produce a camera in which the distortion amount map is pre-measured through the method according to the embodiments. In addition, a program (SW) such as PCG pattern generation, grid point detection, distortion correction, and distortion map for measuring the distortion amount map of the camera, which is a method according to embodiments, may be developed.

According to embodiments, more accurate distortion correction may be performed than a conventional perspective projection invariance method.

A new set of grid points is available for this purpose.

As shown in FIG. 6 , in the vanishing point estimation step, the new set of grid points increases the number of grid points that fall into the micro-distortion region than the existing set of grid points. Therefore, it is robust to noise according to the location of the grid points during linear regression, so that it is possible to estimate the vanishing point more accurately. And, the reason that a new set of grid points cannot be created in the existing method is that the interval between grid points is too narrow, and concentric PCG patterns overlap, making it impossible to detect grid points. However, in this embodiment, since it is generated by computer software, it is possible to generate the patterns while showing them in parallel. That is, the red, blue, green, and pink grid points of the new grid point set in FIG. 6 are displayed sequentially on the monitor instead of being generated all at once.

Also, the same rotation matrix can be used.

Both pattern set A 250 and pattern set B 50 are patterns of the same rotation matrix for the virtual camera. That is, virtual patterns with the same posture differ only in their positions. Therefore, straight lines passing through each row (or column) of all these patterns must pass through the same vanishing point. Therefore, more data (lattice points) can be used for estimating one vanishing point, and since the vanishing point estimated from the pattern set B is shared with the pattern set A, it can be used for distortion correction of the lattice points of the pattern set A.

Also, it is possible to perform distortion center point estimation.

In this embodiment, taking advantage of the fact that the distortion center point estimation can fill the entire image area with grid points, the linearity of all grid points is investigated, and the distortion center point is estimated using the characteristics of the distortion center point. This leads to the setting of the micro-distortion area, and more accurate estimation of the vanishing point is possible using the grid points in the micro-distortion area.

Embodiments may be used in all fields requiring measurement of three-dimensional coordinates or shapes using a camera on a large scale. For example, it can be used for 3D measurement of large-scale products and parts in the shipping industry and manufacturing industry. It can also be used to estimate distortion parameters of RGB cameras of autonomous vehicles. It can be used to estimate the distortion parameters of an out-of-focus camera.

Through the present embodiments, it is possible to perform three-dimensional measurement of large-scale products and parts, such as hull blocks, using a stereo camera. This is not only industrial application, but also has a strong ripple effect due to its strong practicality. In addition, a general camera can be used as equipment to replace the existing optical waveguide. As a result, it is possible not only to reduce the price, but also to reduce the difficulty and time of the workers, thereby contributing to the improvement of the work efficiency of the workers.

The device described above may be implemented as a hardware component, a software component, and/or a combination of the hardware component and the software component. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

Software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or apparatus, to be interpreted by or to provide instructions or data to the processing device. may be embodied in The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

As described above, although the embodiments have been described with reference to the limited embodiments and drawings, various modifications and variations are possible from the above description by those skilled in the art. For example, the described techniques are performed in an order different from the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (15)

An image distortion correction method performed by a computer device, comprising:
generating a plurality of PCG patterns composed of a first pattern set and a second pattern set;
estimating a distortion center point by photographing a first pattern set among the PCG patterns, detecting grid points, and linearly regressing grid points constituting one row or column among the detected grid points;
After photographing the second pattern set among the PCG patterns, lattice points are detected, an area within a predetermined range from the distortion center point in the image is defined as a micro-distortion area, and lattice points within the micro-distortion area are detected in the row direction or estimating a vanishing point in a column direction;
obtaining distortion amount data by performing distortion correction on all the grid points of the first pattern set based on the estimated vanishing point; and
measuring the distortion amount map of the camera lens by curve fitting the distortion amount data scattered over the entire image area
Including, image distortion correction method. According to claim 1,
wherein the first pattern set is existing PCG patterns, and the second pattern set is new PCG patterns.
characterized in that, image distortion correction method. According to claim 1,
The generating of the PCG patterns comprises:
creating a virtual PCG pattern;
displaying an image projected by a virtual camera of the virtual PCG pattern on a monitor; and
acquiring an image by photographing the PCG pattern displayed on the monitor with a real camera
Including, image distortion correction method. According to claim 1,
The generating of the PCG patterns comprises:
Infinitely generating a virtual PCG pattern having a specific rotation matrix and a specific translation matrix based on a virtual camera through a pattern generation method using a virtual camera.
characterized in that, image distortion correction method. According to claim 1,
The first pattern set is a pattern that allows the location of grid points to be detected even in an out-focused image
characterized in that, image distortion correction method. According to claim 1,
The step of estimating the distortion center point,
Assume that the straight line of the grid points closest to the linearly regressed line passes through the distortion center point, and the point of intersection of the straightest lines among the row and column direction straight lines is estimated as the distortion center point
characterized in that, image distortion correction method. According to claim 1,
The step of obtaining distortion amount data by performing the distortion correction,
Performing distortion correction using perspective projection invariance for all the grid points of the first pattern set based on the estimated vanishing point
characterized in that, image distortion correction method. According to claim 1,
Automatic measurement of distortion correction by generating the PCG patterns photographed with the camera fixed at different positions in the monitor using computer software
characterized in that, image distortion correction method. a PCG pattern generator for generating a plurality of PCG patterns including the first pattern set and the second pattern set;
a distortion center point estimator that detects grid points after photographing a first set of patterns among the PCG patterns, and linearly regresses grid points constituting one row or column among the detected grid points to estimate a distortion center point;
After photographing the second pattern set among the PCG patterns, lattice points are detected, an area within a predetermined range from the distortion center point in the image is defined as a micro-distortion area, and lattice points within the micro-distortion area are detected in the row direction or a vanishing point estimator for estimating a vanishing point in a column direction;
a distortion correction unit for obtaining distortion amount data by performing distortion correction on all the grid points of the first pattern set based on the estimated vanishing point; and
A distortion map measuring unit that measures the distortion map of the camera lens by curve fitting the distortion data scattered over the entire image area.
Including, image distortion correction device. 10. The method of claim 9,
wherein the first pattern set is existing PCG patterns, and the second pattern set is new PCG patterns.
Characterized in, image distortion correction device. 10. The method of claim 9,
The PCG pattern generation unit,
generating a virtual PCG pattern, displaying an image projected by a virtual camera of the virtual PCG pattern on a monitor, and acquiring an image by photographing the PCG pattern displayed on the monitor with a real camera
Characterized in, image distortion correction device. 10. The method of claim 9,
The PCG pattern generation unit,
Infinitely generating a virtual PCG pattern having a specific rotation matrix and a specific translation matrix based on a virtual camera through a pattern generation method using a virtual camera.
Characterized in, image distortion correction device. 10. The method of claim 9,
The first pattern set is a pattern that allows the location of grid points to be detected even in an out-focused image
Characterized in, image distortion correction device. 10. The method of claim 9,
The distortion center point estimation unit,
Assume that the straight line of the grid points closest to the linearly regressed line passes through the distortion center point, and the point of intersection of the straightest lines among the row and column direction straight lines is estimated as the distortion center point
Characterized in, image distortion correction device. 10. The method of claim 9,
The distortion correction unit,
Performing distortion correction using perspective projection invariance for all the grid points of the first pattern set based on the estimated vanishing point
Characterized in, image distortion correction device.

Images