KR20070004202A - Method for correcting lens distortion in digital camera - Google Patents

Abstract

A method for correcting the distortion of screen through a lens in a digital camera is provided to promptly correct the distortion of screen by correcting pixel coordinates through pixel coordinate values of a predetermined grid pattern and displacement values for the pixel coordinate values. An image, which is photographed by a lens, is temporally stored. Respective distortion pixel coordinate values are corrected into ideal pixel coordinate values based on displacement values stored in a displacement value storing unit(S15). The distortion pixel coordinate values form a grid pattern in the photographed image. The pixels around the ideal pixel coordinate values are interpolated by pixel data of the corrected ideal pixel coordinate values(S17).

Description

Method for correcting lens distortion in digital camera

1 is an electrical block diagram of a digital camera to which a distortion correction method according to the lens of the present invention is applied;

2 is a flowchart of a method for correcting image distortion by a lens in a digital camera that can be applied to the present invention;

3 is an electrical block diagram illustrating a process of generating a distortion conversion table in a coordinate conversion system applicable to the present invention;

4 is a flowchart illustrating a process of generating a distortion transformation table in a coordinate transformation system applicable to the present invention;

5A through 5C are exemplary views illustrating the process of generating the distortion conversion table of FIGS. 3 and 4;

6A to 6C are exemplary views for explaining a process of obtaining coordinate displacement values stored in the distortion conversion table of FIG. 4.

*** Explanation of symbols for the main parts of the drawing ***

100: digital camera 110: distortion correction

120: detector 130: memory

131: data storage unit 133: distortion conversion table

135: program storage unit 140: image sensor unit

150: noise removing unit 160: signal conversion unit

170: image processing unit 180: display unit

190: lens 200: subject

300: coordinate transformation system 310: coordinate transformation unit

320: filter unit 330: signal conversion unit

340: image sensor unit 350: lens

360: Distortion Conversion Table 400: Standard Image

500: distorted image 600: edge image

The present invention relates to a method for correcting image distortion by a lens in a digital camera, and in particular, to correct pixel coordinates of a distorted image by a distortion conversion table in which respective pixel coordinate values forming a preset grid pattern and displacement values thereof are stored. The present invention relates to a method of correcting picture distortion by a lens in a digital camera.

 When the reflected light of a subject coming through the lens of the camera forms on a recording medium such as a film or an image sensor, the light does not collect at one point, and the image is blurred or skewed. This phenomenon is called aberration. This aberration is largely due to chromatic aberration that causes color blurring while rays of different wavelengths focus on different image planes, and when the lens is spherical, the marginal ray of the lens is a paraxial ray ( The focal length is shorter than that of the paraxial ray, and the spherical aberration differs depending on the height incident on the optical system. It may be classified into a field curvature formed on a curved surface rather than a plane, and a distortion aberration that generates an uneven image.

On the other hand, the distortion characteristics include barrel distiortion caused by the short focal length of the lens and the outer part of the image being bent toward the center, and, conversely, failure type distortion caused by the central part of the image due to the long focal length of the lens. It can be classified as (Pincushion Distiortion), and cylindrical distortion is more prominent than failure type distortion. This is because the size of the image sensor that records the image coming through the lens, such as a CCD or CMOS sensor that converts the optical signal into an electrical signal is generally smaller than the film of the film camera, so the focal length is short. Therefore, because of the distortion characteristics of the lens, the image size and perspective is different from the original image, so that the camera user is provided with an uneven image. For example, in the case of using a digital camera as a rear view device of an automobile, the distortion characteristic of the lens makes it impossible to effectively perform the function as a rear view device provided for driving or parking.

In order to compensate for such distortion, there is an aspherical lens effective for reducing spherical aberrations in which the focal position is shifted and becomes an unclear image or distortion in which the image is distorted, but the lens itself is expensive. Further, the pixels extracted from the distorted image the X-axis and Y-axis coordinate value of each of X _d and Y _d, respectively, X _u of X-axis and Y-axis coordinate value of the pixel extracted from the image after the distortion correction and the Y _u If the lens distortion parameter obtained through the camera calibration is k, it can be assumed that there exists a distortion conversion equation such as the following equations 1 and 2 due to lens distortion between the two coordinate systems. Algorithms for obtaining a corrected image close to the original image may be implemented in software.

X _u = X _d (1 + KR _d ² )

Y _u = Y _d (1 + KR _d ² )

Here, R _d = (X _d ² + Y _d ² ) ^1/2 means the distance from the center of the image to (X _d , Y _d ), which is the pixel coordinate extracted from the distorted image.

However, since the ECU (Electronic Control Unit) of the digital camera implemented by the algorithm described above repeats the distortion correction process for all the pixels in real time, there is a problem that the execution speed decreases as the size of the image increases.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problem, and the digital coordinates of the distorted image can be corrected by a distortion conversion table in which each pixel coordinate value and a displacement value which form a predetermined grid pattern are stored. An object of the present invention is to provide a method for correcting distortion of a screen by a lens in a camera.

In the digital camera of the present invention for achieving the above object, a distortion correction method by a lens includes a respective distortion pixel coordinate value and a respective distortion pixel coordinate forming a predetermined grid pattern to represent distortion by a lens on a screen. Temporarily storing the image taken by the lens in a digital camera comprising a displacement value storage means for storing a displacement value for correcting a value to an abnormal pixel coordinate value corresponding to each of the distortion pixel coordinate values. ; Correcting each distorted pixel coordinate value forming the grid pattern in the photographed image to the abnormal pixel coordinate value based on the displacement value stored in the displacement value storing means and surrounding the corrected abnormal pixel coordinate value. A method of correcting image distortion by a lens in a digital camera, the method comprising: interpolating a color by pixel data of the corrected abnormal pixel coordinate value.

Hereinafter, a method of correcting screen distortion by a lens in a digital camera according to an exemplary embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is an electrical block diagram of a method for correcting image distortion by a lens in a digital camera that can be applied to the present invention.

As shown in FIG. 1, the configuration of the digital camera to which the present invention can be applied includes a lens 190 that receives the reflected light of the subject 200, and color, luminance, or brightness of an optical signal incident through the lens 190. The image sensor unit 140 and the image sensor unit 140 converting the optical signal passing through the detection unit 120 into an electrical signal, that is, generating a charge in proportion to the intensity of light. Noise removal unit 150 for removing unnecessary electrical signals in addition to the image information among the electrical signals converted through), the signal converter 160 for converting the analog signal filtered by the noise removal unit 150 into a digital signal, a preset grid Coordinate values for each pixel constituting the pattern (minimum object unit of the image sensor unit 140 that recognizes the light intensity) and displacements thereof are stored, such as a distortion conversion table 131 and a color interpolation algorithm. A memory 130 including a program storage unit 133 for storing an application program, a digital image converted by the image processing unit 170 and the signal conversion unit 160 for driving the display unit 180 to output an image signal In the signal, the coordinate value of the pixel corresponding to the coordinate value stored in the distortion conversion table 131 is changed, that is, after correcting the pixel coordinate value of the distorted image due to the distortion aberration characteristic of the lens, the program storage unit 133 It may include a distortion correction 110 for interpolating the color for the pixel according to the stored color interpolation algorithm.

In the above-described configuration, the detection unit 120 combines a color image by combining a UV filter that blocks ultraviolet rays, a polarization filter that removes reflection of light, a color correction filter that can correct a specific light source among various light sources, or color information of reflected light. It can be implemented as an RGB filter for transmitting. The image sensor unit 140 may be implemented as a CCD sensor composed of a photoelectric conversion semiconductor and a charge coupled device or a CMOS sensor composed of a photoelectric conversion semiconductor and a complementary metal oxide semi-conductor (CMOS) switch. The larger the number of arranged pixels, the more accurate the image is generated. The distortion conversion table 131 may be implemented as a nonvolatile memory (ROM), preferably an EPROM. When a lens is replaced or added, the distortion conversion table 131 may be changed to a new coordinate displacement value corresponding to the distortion ratio of the lens. Can be. To this end, an external connection terminal such as USB may be provided outside the digital camera 100. The program storage unit 133 may also be embodied as a nonvolatile memory (ROM), preferably an EPROM. The neighbor neighbor replication, bilinear interpolation, and pattern may be implemented. Color interpolation algorithms such as pattern matching based interpolation or edge sensing interpolation may be stored.

In addition, the noise removing unit 150 may be implemented as a correlated double sampling circuit (CDS), which is a noise reduction circuit, and the signal converter 160 may be implemented as an A / D converter. In addition, the distortion correction unit 110 may be implemented as a microcomputer, the image processor 170 may be implemented as a DSP, and the display unit 180 may be implemented as an LCD.

2 is a flowchart of a method for correcting distortion of a screen by a lens in a digital camera that can be applied to the present invention, and the subject is a distortion correction unit 110.

First, in step S11, after receiving the digital signal from the signal converter 160, the flow advances to step S13 to extract a digital signal corresponding to the pixel coordinate value of the grid pattern. Next, in step S15, the pixel coordinate value of the extracted digital signal is changed based on the coordinate displacement value stored in the distortion conversion table 131, and then the flow advances to step S17 to color the pixel around the pixel whose coordinate value is changed. For example, pixel values whose coordinates are changed according to neighboring pixel interpolation algorithms are assigned to the nearest pixels, respectively. Finally, in operation S19, the image processor 170 processes the digital signal received from the distortion correction unit 110 and outputs the digital signal to the display unit 180.

3 is an electrical block diagram illustrating a process of generating a distortion conversion table in a coordinate transformation system applicable to the present invention, and FIGS. 5A to 5C are screen views for explaining FIG. 3.

As shown in FIG. 5A, the coordinate transformation system 300 to which the present invention can be applied includes a standard image 400 having a horizontal and vertical length equal to R 410 and a uniform difference in contrast. The analog signal converted by the lens 350, the image sensor unit 340 for converting the optical signal incident through the lens 350 into an electrical signal, and the image sensor unit 340 is converted into a digital signal. Coordinates of the edge signal extracted from the filter unit 330 and the filter unit 320 to extract a predetermined edge signal from the digital signal received through the signal converter 330 And a distortion conversion table 360 that stores the displacement value of the coordinates changed by the coordinate transformation generator 310 and the coordinate transformation generator 310 to change the value into a predetermined pattern.

In the above-described configuration, the standard image 400 of FIG. 5A is input to the image sensor unit 340 while being distorted as shown in FIG. 5B according to the distortion aberration characteristic of the lens. Meanwhile, the filter unit 320 may be implemented as a Laplacian image filter. As shown in FIG. 5C, the edge filter 600 may be obtained by extracting only an edge from the distorted image 500. The coordinate transformation generator 310 may be implemented by a microcomputer or a personal computer. In addition, since the image sensor unit 340 and the signal converter 330 perform the same functions as the image sensor unit 140 and the signal converter 160 which are components of the digital camera 100 described above, the repeated description will be repeated. Omit.

 4 is a flowchart illustrating a process of generating a distortion transformation table in a coordinate transformation system that may be applied to the present invention. The subject is a coordinate transformation generator 310. 6A to 6C are exemplary diagrams for describing FIG. 4.

First, in step S21, as shown in FIG. 5C, after receiving the edge signal filtered by the filter unit 320, the flow proceeds to step S23. As shown in FIG. 6A, in step S23, the pixel coordinate values of the outermost edge line A1 of the horizontal axis are first extracted from the edge image 600, and as shown in FIG. 6B, the process proceeds to step S25 and the extracted maximum Select the point closest to the center of the camera lens among the outer lines A1. Next, in step S27, a straight line A2 is formed at right angles to the center line of the camera lens at the selected point, and then each pixel coordinate value forming the straight line is extracted, and then the process proceeds to step S29 to correspond to each of A1 to A2. The vertical displacement (dxa, 0) for the pixel of is stored in the distortion conversion table 360.

Subsequently, in step S31, the pixel coordinate values forming the horizontal axis line B1 separated by R from A1 are extracted, and the flow proceeds to step S33 to generate a straight line B2 horizontal to A2 separated by R from A2 as shown in Fig. 6C. And extract the pixel coordinate values. Thereafter, in step S35, the vertical displacements dxb and 0 for each pixel corresponding to B1 to B2 are stored in the distortion conversion table 360. In this way, when the storage of displacement values for all horizontal axis edge lines is completed, the displacement values for vertical axis edge lines are obtained and stored in the same manner. So, finally, in step S37, the coordinate displacement values between the test image 300 and the distorted image 400 for all edges are obtained as described above, and then stored in the distortion conversion table 360.

In the digital camera of the present invention, the method for correcting the distortion of a screen by a lens is not limited to the above-described embodiment, and may be modified in various ways within the scope of the technical idea of the present invention.

According to the method for correcting image distortion by a lens in the digital camera of the present invention as described above, a fast correction process is performed by correcting a distorted coordinate from a distortion conversion table in which pixel coordinate values of a predetermined grid pattern and displacement values thereof are stored. You can see the natural video without the effect is effective.

Claims (2)

Displacement for correcting each of the distorted pixel coordinate values and the respective distorted pixel coordinate values forming a predetermined grid pattern to represent the distortion by the lens on the screen to the abnormal pixel coordinate values corresponding to the respective distorted pixel coordinate values. In the digital camera comprising a displacement value storage means is stored, Temporarily storing an image captured by the lens; Correcting the distorted pixel coordinate values forming the grid pattern in the photographed image to the abnormal pixel coordinate values based on the displacement values stored in the displacement value storing means; And color interpolating the pixel around the corrected abnormal pixel coordinate value by pixel data of the corrected abnormal pixel coordinate value. The method of claim 1, Generation of the displacement value stored in the displacement value storage means, (a) receiving an edge image obtained by passing through a filter after photographing a standard image having a checkered pattern having the same horizontal and vertical lengths and having uniform differences in contrast; (b) extracting pixel coordinate values of an arbitrary outermost edge line A1 among edge lines included in the input edge image; (c) selecting a point closest to the center of the camera lens among the outermost edge lines A1; (d) generating a straight line A2 perpendicular to the center line of the camera lens at the selected point and extracting pixel coordinate values; (e) storing the vertical displacement value corresponding to the straight line A2 at the outermost edge line A1 in the displacement value storing means; (f) extracting pixel coordinate values of the edge line B1 after the outermost edge line A1; (g) generating a straight line B2 parallel to the straight line A2 at a distance apart from the straight line A2 by a predetermined value and extracting pixel coordinate values; (h) storing the vertical displacement value corresponding to the straight line B2 at the edge line B1 in the displacement value storing means; and and (i) storing the displacement values for all the edges in the displacement value storing means in the same manner as in the above steps (a) to (h). Calibration method.

Images