Distortion correction method and apparatus for fisheye lens / panoramic camera

Abstract

A method and apparatus for distortion correction of a fisheye lens / panoramic camera are provided. The coordinates of each pixel are obtained from a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera, and the coordinates of each pixel are converted into a three-dimensional curved coordinate system based on a preset first transformation formula. In this case, values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula are obtained from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​resulting from the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor. Subsequently, the coordinates of the three-dimensional curved coordinate system corresponding to each pixel are converted into a planar coordinate system to obtain a distortion-corrected image.

Description

Technology Field

[0001] The present invention relates to distortion correction of a camera, and more specifically, to a method and apparatus for distortion correction of a fisheye lens / panoramic camera. Background Technology

[0002] Fisheye and panoramic cameras can have a wider surveillance range compared to standard cameras, but they have the disadvantage that subjects are distorted, making it difficult to gauge distances or accurately determine the shape of subjects.

[0003] Accordingly, technologies are being used to correct distortion; however, the computational load on the computing devices required for this correction is so high that it is difficult to apply these technologies in actual use. Consequently, research is being conducted to reduce the computational load and increase processing speed to enable practical application.

[0004] Related technology includes “image processing device, image capture system, image processing method and storage medium” disclosed in U.S. Patent Publication No. US 2019-0347775. The problem to be solved

[0005] The problem that the present invention aims to solve is to provide a distortion correction method and apparatus capable of accelerating the computational speed and reducing the computational load when correcting distortion in a fisheye lens / panoramic camera. means of solving the problem

[0006] According to one embodiment, a distortion correction method is provided. The distortion correction method comprises: a distortion correction device acquiring a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera; a distortion correction device acquiring coordinates of each pixel from the distorted image; a distortion correction device converting the coordinates of each pixel into a three-dimensional curved coordinate system based on a preset first transformation formula - acquiring values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​corresponding to the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor -; and a distortion correction device converting the coordinates of the three-dimensional curved coordinate system corresponding to each pixel into a planar coordinate system to acquire a distortion-corrected image.

[0007] In one implementation, the first input factor and the second input factor in the complex trigonometric function product LUT are In the range The angles are units, and can be configured in a form where a value resulting from the multiplication of the two trigonometric functions corresponding to the angles is stored.

[0008] In one implementation, the two trigonometric functions in the complex trigonometric function product LUT may be identical or different trigonometric functions. When obtaining a value corresponding to the multiplication of trigonometric functions from the complex trigonometric function product LUT, the two trigonometric functions may be combined to generate a first hash key to select one of a plurality of three-dimensional blocks, and the two input parameters may be combined to generate a second hash key to select a corresponding value within the selected block.

[0009] In one implementation, the step of converting to the three-dimensional curved surface coordinate system may correspond to projecting the distorted image onto a unit sphere with a radius of 1 in a three-dimensional space without distortion.

[0010] In one implementation, the above first transformation formula is as follows:

[0011]

[0012] It can be like that.

[0013] In one implementation, when converting the coordinates of the three-dimensional curved surface coordinate system to a planar coordinate system, the coordinates of each pixel are converted to a planar coordinate system based on a preset second transformation formula, and values ​​corresponding to the multiplication of trigonometric functions included in the second transformation formula can be obtained from the composite trigonometric function product LUT.

[0014] According to another embodiment, a distortion correction device is provided. The distortion correction device comprises: an interface device configured to receive a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera; and a processor configured to perform correction processing on the distorted image input through the interface device to obtain a distortion-corrected image, wherein the processor is configured to perform the operation of obtaining the coordinates of each pixel from the distorted image obtained through the interface device; the operation of converting the coordinates of each pixel into a three-dimensional curved surface coordinate system based on a preset first transformation formula - obtaining values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​corresponding to the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor -; and the operation of converting the coordinates of the three-dimensional curved surface coordinate system corresponding to each pixel into a planar coordinate system to obtain a distortion-corrected image.

[0015] In one implementation, the first input factor and the second input factor in the complex trigonometric function product LUT are In the range The angles are units, and can be configured in a form where a value resulting from the multiplication of the two trigonometric functions corresponding to the angles is stored.

[0016] In one implementation, the operation of converting to the three-dimensional curved surface coordinate system may correspond to projecting the distorted image onto a unit sphere with a radius of 1 in a three-dimensional space without distortion.

[0017] In one implementation, the two trigonometric functions in the complex trigonometric function product LUT may be identical or different trigonometric functions. When obtaining a value corresponding to the multiplication of trigonometric functions from the complex trigonometric function product LUT, the two trigonometric functions may be combined to generate a first hash key to select one of a plurality of three-dimensional blocks, and the two input parameters may be combined to generate a second hash key to select a corresponding value within the selected block.

[0018] In one implementation, the processor may be configured to convert the coordinates of each pixel into a planar coordinate system based on a preset second transformation formula when converting the coordinates of the three-dimensional curved surface coordinate system into a planar coordinate system, and to obtain values ​​corresponding to the multiplication of trigonometric functions included in the second transformation formula from the composite trigonometric function product LUT. Effects of the invention

[0019] According to an embodiment, when performing an operation to correct a distorted image, instead of a direct calculation based on a transformation formula, the value of the operation can be obtained by using a pre-configured line calculation table for the product of two input factors and different trigonometric functions.

[0020] Therefore, computation speed can be improved and computational load can be significantly reduced. In particular, since simultaneous computation per pixel can be performed for each pixel of the image through a parallel processing unit, processing time can also be significantly reduced.

[0021] By accelerating image distortion correction and reducing computational load, distortion-corrected images of the same quality can be viewed even on lower-performance image processing devices, thereby reducing production and delivery costs. Brief explanation of the drawing

[0022] Figure 1 is an example diagram showing the degree of distortion of an image obtained from a fisheye lens / panoramic camera. FIG. 2 is a schematic diagram showing the process of acquiring a distortion-corrected image according to an embodiment of the present invention. FIG. 3 is an exemplary diagram showing the actual imaging plane and planar projection plane of a fisheye lens / panoramic camera according to an embodiment of the present invention. FIG. 4 is an exemplary diagram showing the geometric relationship between the distorted space coordinates and the three-dimensional space coordinates according to an embodiment of the present invention. FIG. 5 is an exemplary diagram showing a projection tangent surface on a three-dimensional curved surface according to an embodiment of the present invention. FIG. 6 is an exemplary diagram showing a complex trigonometric function product LUT according to an embodiment of the present invention. FIG. 7 is a flowchart of a distortion correction method according to an embodiment of the present invention. FIG. 8 is a structural diagram illustrating a computing device for implementing a method according to an embodiment of the present invention. Specific details for implementing the invention

[0023] Embodiments of the present invention are described below with reference to the attached drawings so that those skilled in the art can easily implement them. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present invention in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals.

[0024] Throughout the specification, when a part is described as "including" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0025] Expressions described in the singular in this specification may be interpreted as singular or plural unless explicit expressions such as "one" or "single" are used.

[0026] Additionally, terms including ordinal numbers, such as first, second, etc., used in embodiments of the present invention may be used to describe components, but the components should not be limited by these terms. The 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.

[0027] Hereinafter, a distortion correction method and apparatus for a fisheye lens / panoramic camera according to an embodiment of the present invention will be described.

[0028] Figure 1 is an example diagram showing the degree of distortion of an image obtained from a fisheye lens / panoramic camera.

[0029] In the case of images captured by cameras using wide-angle fisheye lenses or panoramic cameras, the part of the image formed near the center of the lens is high-resolution and poses no significant problem; however, as the image is formed further from the center, the resolution drops, and thus the impact of lens distortion becomes greater.

[0030] That is, as exemplified in Fig. 1, as the angle of view widens, radiation distortion becomes more severe as it moves toward the outer edges away from the optical axis.

[0031] In an embodiment of the present invention, an image acquired by a camera using a fisheye lens or a panoramic camera is projected onto a curved surface in three-dimensional space, and then projected again onto a plane in three-dimensional space without distortion to obtain a distortion-corrected image. Here, the image acquired by the camera or panoramic camera may be a still image or a video.

[0032] FIG. 2 is a schematic diagram showing the process of acquiring a distortion-corrected image according to an embodiment of the present invention. FIG. 3 is an exemplary diagram showing the actual imaging plane and the planar projection plane of a fisheye lens / panoramic camera according to an embodiment of the present invention, FIG. 4 is an exemplary diagram showing the geometric relationship between the distortion space coordinates and the three-dimensional space coordinates according to an embodiment of the present invention, and FIG. 5 is an exemplary diagram showing the projection tangent plane on a three-dimensional curved surface according to an embodiment of the present invention.

[0033] An image obtained by a camera using a fisheye lens or a panoramic camera is a distorted image, i.e., a distorted still image / distorted video, as shown in the attached FIG. 2 (a), and is referred to as a "distorted image" for convenience of explanation below. This distorted image is projected onto a curved surface in three-dimensional space as shown in FIG. 2 (b), and then re-projected onto a plane in three-dimensional space as shown in FIG. 2 (c).

[0034] Since the actual imaging surface of a fisheye lens / panoramic camera is composed of a spherical / curved surface as shown in FIG. 3 (a), if it is distorted into a flat plane and projected, it can be displayed as a general image (still image / moving image) as shown in FIG. 3 (b).

[0035] Based on this, a first process of projecting the distorted image onto a curved surface in a preset three-dimensional space and a second process of projecting it back onto a plane from the curved surface are performed.

[0036] In the first process of projecting a distorted image onto a curved surface, each pixel of the distorted image is converted from distorted space coordinates, i.e., the distorted coordinate system, to coordinates in three-dimensional space, i.e., the three-dimensional curved surface coordinate system (also called the non-distorted coordinate system). The distorted space coordinates and the coordinates in three-dimensional space have a geometric relationship as exemplified in FIG. 4, and the actual coordinates in three-dimensional space and the distorted coordinates based on the distorted image are as exemplified in FIG. 5. The distorted image is projected onto a virtual unit sphere as exemplified on the left side of FIG. 5, and in the embodiment of the present invention, when converting the coordinate system of the distorted image to the curved surface coordinate system in the first process, the unit sphere is set to a radius of 1.

[0037] In an embodiment of the present invention, to minimize computational load, a distorted image is projected onto a curved surface in a three-dimensional space without distortion, specifically a unit sphere with a radius of 1. When projecting, the formula for converting each pixel of the distorted image from the distortion coordinate system to the three-dimensional curved surface coordinate system is as follows.

[0038] [Mathematical Formula 1]

[0039]

[0040] Here, x, y, and z represent the coordinates of the pixel in the 3D curved coordinate system.

[0041] In this mathematical equation 1, the sine function (sin θ) and the cosine function (cos Trigonometric functions that are ) are used. In an embodiment of the present invention, in order to significantly reduce the computational load, a specific angle (e.g., θ, A line calculation table, i.e., a Loot Up Table (LUT), is used to pre-calculate values ​​such as sine and cosine functions for ) and store them in a table format. In particular, the present invention uses a complex trigonometric function product LUT.

[0042] FIG. 6 is an exemplary diagram showing a complex trigonometric function product LUT according to an embodiment of the present invention.

[0043] Based on a formula (e.g., Equation 1) that transforms each pixel of a distorted image from a distorted coordinate system to a three-dimensional surface coordinate system according to an embodiment of the present invention, two input parameters (θ, ) and two different trigonometric functions, the sine function (sin θ) and the cosine function (cos The multiplication of ) is performed frequently. Accordingly, as shown in the attached FIG. 6, a complex trigonometric function product LUT comprising a plurality of 3D blocks is implemented. Each 3D block of the complex trigonometric function product LUT is configured to include values ​​resulting from the multiplication of two trigonometric functions, and in particular, is configured to include values ​​resulting from the multiplication of two trigonometric functions according to two input values, namely, a first input factor (input angle 1: A) and a second input factor (input angle 2: B). Specifically, the first block (B1) is configured to include values ​​resulting from the multiplication of a first trigonometric function and a second trigonometric function (e.g., Sin(A) * Cos(B)) corresponding to a first input argument and a second input argument, the second block (B2) is configured to include values ​​resulting from the multiplication of a first trigonometric function and a second trigonometric function (e.g., Sin(A) * Sin(B)) corresponding to a first input argument and a second input argument, and the third block (B3) is configured to include values ​​resulting from the multiplication of a third trigonometric function and a second trigonometric function (e.g., Cos(A) * Cos(B)) corresponding to a first input argument and a second input argument.

[0044] In a complex trigonometric function product LUT containing these three-dimensional blocks, two trigonometric functions are combined to generate a hash key (also called the first hash key) to select one of the multiple three-dimensional blocks, and two input arguments are combined to generate a hash key (also called the second hash key) to select a value within the selected block. In this way, a value corresponding to the multiplication of two trigonometric functions according to two input arguments can be quickly obtained from the pre-configured complex trigonometric function product LUT.

[0045] Here, in the embodiments of the present invention, for each trigonometric function Regarding the range A complex trigonometric product LUT can be constructed in units. In each 3D block of the complex trigonometric product LUT, the first input argument and the second input argument Set interval in range ( It corresponds to the angle of ), and for each angle, a value resulting from the multiplication of two trigonometric functions may be included. The angle interval according to the embodiment of the present invention is not limited thereto. If the angle corresponding to the first input factor and / or the second input factor does not exist within the block, interpolation may be used to estimate the value resulting from the multiplication of two trigonometric functions corresponding to the angle of the first input factor and / or the second input factor.

[0046] In this way, when performing calculations on the formula (e.g., Equation 1) used in the process of converting pixels of a distorted image from a distorted coordinate system to a 3D curved coordinate system, by constructing a composite trigonometric function product LUT for each and performing the calculation, the amount of calculation can be significantly reduced and the calculation speed can be accelerated.

[0047] Afterwards, the second process is performed to convert the 3D curved coordinate system into a planar coordinate system to finally obtain a distortion-corrected image.

[0048] The 3D spatial coordinate system (3D surface coordinate system) of a distorted image is expressed based on the yaw-pitch-roll (quaternion rotation)-translation (3D spatial translation) of general 3D space. However, when expressing the coordinate system of a spherical / distorted image based on the 2D polar coordinate system, the values ​​of longitude and latitude are converted through translation, and the angles of yaw, pitch, and roll are used as rotation transformation formulas, which can be used as polar coordinate translation formulas for projection transformation between the surface and the tangent surface.

[0049] Longitude and latitude values ​​can be moved through translation as shown in Equation 2 below.

[0050] [Mathematical Formula 2]

[0051]

[0052]

[0053] Here, θ represents longitude, represents latitude.

[0054] In the above mathematical formula 3, for x, y, and z, the operation is performed using products of trigonometric functions, and accordingly, in the embodiment of the present invention, as described above, corresponding values ​​can be obtained quickly by using a complex trigonometric function product LUT to obtain x, y, and z.

[0055] Subsequently, as the longitude and latitude values ​​are converted into Translation (3D spatial movement), the distorted source point This distortion correction reference point It moves to, and based on this, projection transformation between the surface and the tangent surface can be performed. Distortion correction reference point It can be calculated according to the following formula.

[0056] [Mathematical Formula 3]

[0057]

[0058]

[0059] Here, represents the use of yaw-pitch-roll rotation in three-dimensional space as Translation in polar coordinates. S represents a scaling factor. Here, the value of S can be fixed to 1, and a value of S of 1 indicates that no scaling is performed. The present invention is not necessarily limited thereto.

[0060] Based on these distortion correction reference points, the 3D curved coordinate system of the distorted image can be converted into a planar coordinate system to finally obtain a distorted image.

[0061] Next, a distortion correction method according to an embodiment of the present invention will be described based on the above.

[0062] FIG. 7 is a flowchart of a distortion correction method according to an embodiment of the present invention.

[0063] A distortion correction device according to an embodiment of the present invention acquires a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera, as shown in the attached FIG. 7 (S100).

[0064] The distortion correction device obtains the coordinates of each pixel from the distorted image (S110).

[0065] Then, a first process is performed to convert the coordinates of each pixel of the distorted image into a three-dimensional curved surface coordinate system (S120). At this time, the distortion correction device projects the coordinates of the pixels of the distorted image onto a unit sphere with a radius of 1 of the three-dimensional curved surface, and converts the pixel coordinates into coordinates of the three-dimensional curved surface coordinate system based on a preset first transformation formula. In particular, values ​​corresponding to the multiplication of trigonometric functions used in the first transformation formula are obtained from a preset composite trigonometric function product LUT, and the pixel coordinates are converted into coordinates of the three-dimensional curved surface coordinate system.

[0066] After performing the first processing, the distortion correction device performs a second processing to convert the coordinates of the pixel in the 3D curved surface coordinate system into the coordinates of the 2D planar coordinate system, thereby obtaining a distortion-corrected image (S130). In one implementation example, the distortion correction device may obtain values ​​corresponding to the multiplication of trigonometric functions used in the second conversion formula (e.g., Equations 3 and 4) from a pre-configured composite trigonometric function product LUT to convert the coordinates of the pixel in the 3D curved surface coordinate system into the coordinates of the 2D planar coordinate system.

[0067] In this way, in the process of converting each pixel of a distorted image from a distorted coordinate system to a 3D surface coordinate system and / or from a 3D surface coordinate system to a planar coordinate system, instead of direct calculation based on a transformation formula, processing is performed by referencing a composite trigonometric function product LUT composed of multiple 3D blocks containing values ​​resulting from the multiplication of two trigonometric functions for each of two input parameters, thereby obtaining the value of the operation, so the operation speed can be improved and the operation load can be significantly reduced. In particular, since simultaneous operation per pixel can be performed for each pixel of the image through a parallel processing unit, the processing time can also be significantly reduced.

[0068] FIG. 8 is a structural diagram illustrating a computing device for implementing a distortion correction method according to an embodiment of the present invention.

[0069] As illustrated in the attached FIG. 8, a distortion correction method according to one embodiment of the present invention can be implemented using a computing device (100).

[0070] The computing device (100) may include at least one of a processor (110), memory (120), input interface device (130), output interface device (140), storage device (160), and network interface device (150). Each component may be connected by a bus (170) to communicate with one another. Additionally, each component may be connected via individual interfaces or individual buses centered around the processor (110), rather than via a common bus (170).

[0071] The processor (110) can be implemented in various types such as an Application Processor (AP), a Central Processing Unit (CPU), a Graphic Processing Unit (GPU), etc., and may be any semiconductor device that executes instructions stored in memory (120) or a storage device (160). The processor (110) can execute program instructions stored in at least one of memory (120) and the storage device (160). Such a processor (110) can be configured to implement the functions and methods described above based on FIGS. 1 to 7. For example, the processor (110) can pre-configure a complex trigonometric function product LUT consisting of multiple three-dimensional blocks containing values ​​resulting from the multiplication of two trigonometric functions for each of two input factors. The pre-configured complex trigonometric function product LUT can be stored in memory (120) or a storage device (160) and used. When the processor (110) performs a first process of converting the coordinates of each pixel of a distorted image into a three-dimensional curved coordinate system and / or a second process of converting from a three-dimensional curved coordinate system to a planar coordinate system, it can obtain the coordinates of the three-dimensional curved coordinate system or the planar coordinate system corresponding to the coordinates of each pixel by using a composite trigonometric function product LUT stored in the memory (120) or storage device (160). Here, the composite trigonometric function product LUT Set interval in range ( It may include values ​​resulting from the multiplication of two trigonometric functions for each of the two input factors having an angle of ). Such a processor (110) may be configured to include a parallel processing unit that obtains a corresponding value from each three-dimensional block of the complex trigonometric function product LUT.

[0072] The memory (120) and storage device (160) may include various forms of volatile or non-volatile storage media. For example, the memory may include ROM (read-only memory) (121) and RAM (random access memory) (122). In an embodiment of the present invention, the memory (120) may be located inside or outside the processor (110), and the memory (120) may be connected to the processor (110) through various known means.

[0073] The input interface device (130) is configured to provide data (e.g., a distorted image) to the processor (110), and the output interface device (140) is configured to output data from the processor (110).

[0074] The network interface device (150) can transmit or receive signals to or from another device via a wired network or a wireless network, and may optionally be included.

[0075] A computing device (100) having such a structure is named a distortion correction device and can implement a distortion correction method according to one embodiment of the present invention.

[0076] In addition, at least some of the distortion correction method according to one embodiment of the present invention may be implemented as a program or software executed on a computing device (100), and the program or software may be stored on a computer-readable medium.

[0077] In addition, at least some of the distortion correction methods according to one embodiment of the present invention may be implemented in hardware that can be electrically connected to a computing device (100).

[0078] Embodiments of the present invention are not limited to being implemented only through the apparatus and / or methods described above, but may also be implemented through a program for realizing a function corresponding to the configuration of an embodiment of the present invention, a recording medium on which such a program is recorded, etc. Such implementation can be easily achieved by a person skilled in the art to which the present invention pertains from the description of the embodiments described above.

[0079] Although embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention as defined in the following claims also fall within the scope of the present invention.

Claims

Claim 1 As a distortion correction method, the distortion correction device acquires a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera; the distortion correction device acquires the coordinates of each pixel from the distorted image; the distortion correction device converts the coordinates of each pixel into a three-dimensional surface coordinate system based on a preset first transformation formula - acquiring values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​resulting from the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor -; The distortion correction device comprises the step of converting the coordinates of the three-dimensional curved surface coordinate system corresponding to each pixel into a planar coordinate system to obtain a distortion-corrected image, wherein in the composite trigonometric function product LUT, the two trigonometric functions are identical or different from each other, and when obtaining a value corresponding to the multiplication of trigonometric functions from the composite trigonometric function product LUT, the two trigonometric functions are combined to generate a first hash key to select one of a plurality of three-dimensional blocks, and the two input factors are combined to generate a second hash key to select a corresponding value within the selected block. Claim 2 A distortion correction method comprises: a distortion correction device acquiring a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera; a distortion correction device acquiring the coordinates of each pixel from the distorted image; a distortion correction device converting the coordinates of each pixel into a three-dimensional curved coordinate system based on a preset first transformation formula - acquiring values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​resulting from the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor -; and a distortion correction device acquiring a distortion-corrected image by converting the coordinates of the three-dimensional curved coordinate system corresponding to each pixel into a planar coordinate system, wherein the first input factor and the second input factor in the composite trigonometric function product LUT In the range A distortion correction method configured such that the angles are units, and the value resulting from the multiplication of the two trigonometric functions corresponding to the angles is stored. Claim 3 delete Claim 4 A distortion correction method according to claim 1, wherein the step of converting to a three-dimensional curved surface coordinate system corresponds to projecting the distorted image onto a unit spherical surface with a radius of 1 in a three-dimensional space without distortion. Claim 5 In paragraph 1, the above first transformation formula is as follows: Distortion correction methods such as Claim 6 A distortion correction method according to claim 1, wherein when converting the coordinates of the three-dimensional curved surface coordinate system to a planar coordinate system, the coordinates of each pixel are converted to a planar coordinate system based on a preset second transformation formula, and values ​​corresponding to the multiplication of trigonometric functions included in the second transformation formula are obtained from the composite trigonometric function product LUT. Claim 7 As a distortion correction device, the interface device is configured to receive a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera; and the processor is configured to perform correction processing on the distorted image input through the interface device to obtain a distortion-corrected image, wherein the processor comprises: an operation of obtaining the coordinates of each pixel from the distorted image obtained through the interface device; and an operation of converting the coordinates of each pixel into a three-dimensional surface coordinate system based on a preset first transformation formula - obtaining values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​resulting from the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor -; A distortion correction device configured to perform an operation of converting the coordinates of the three-dimensional curved surface coordinate system corresponding to each pixel into a planar coordinate system to obtain a distortion-corrected image, wherein in the composite trigonometric function product LUT, the two trigonometric functions are identical or different from each other, and when the processor obtains a value corresponding to the multiplication of trigonometric functions from the composite trigonometric function product LUT, the processor is configured to perform an operation of combining the two trigonometric functions to generate a first hash key to select one of a plurality of three-dimensional blocks, and combining the two input parameters to generate a second hash key to select a corresponding value within the selected block. Claim 8 As a distortion correction device, the interface device is configured to receive a distorted image, which is an image captured by a camera using a fisheye lens or a panoramic camera; and the processor is configured to perform correction processing on the distorted image input through the interface device to obtain a distortion-corrected image, wherein the processor comprises: an operation of obtaining the coordinates of each pixel from the distorted image obtained through the interface device; and an operation of converting the coordinates of each pixel into a three-dimensional surface coordinate system based on a preset first transformation formula - obtaining values ​​corresponding to the multiplication of trigonometric functions included in the first transformation formula from a preset composite trigonometric function product LUT (Look Up Table), wherein the composite trigonometric function product LUT includes a plurality of three-dimensional blocks, and each three-dimensional block is configured to include values ​​resulting from the multiplication of two trigonometric functions corresponding to a first input factor and a second input factor -; and is configured to perform an operation of obtaining a distortion-corrected image by converting the coordinates of the 3D curved surface coordinate system corresponding to each pixel into a planar coordinate system, and in the composite trigonometric function product LUT, the first input parameter and the second input parameter are In the range A distortion correction device configured such that the angles are units, and the value resulting from the multiplication of the two trigonometric functions corresponding to the angles is stored. Claim 9 A distortion correction device according to claim 7, wherein the operation of converting to the three-dimensional curved surface coordinate system corresponds to projecting the distorted image onto a unit spherical surface with a radius of 1 in a three-dimensional space without distortion. Claim 10 delete Claim 11 A distortion correction device according to claim 7, wherein the processor, when converting the coordinates of the three-dimensional curved surface coordinate system to a planar coordinate system, converts the coordinates of each pixel to a planar coordinate system based on a preset second transformation formula, and is configured to obtain values ​​corresponding to the multiplication of trigonometric functions included in the second transformation formula from the composite trigonometric function product LUT.

Images