KR102318196B1 - A method for auto white balance of image and an electronic device to process auto white balance method - Google Patents

Abstract

The present invention comprises the steps of determining at least one candidate pixel having a gray color in natural light for each frame, and R (red), G (green), B (blue) for each frame so that the at least one candidate pixel becomes a gray point pixel. A frame including a plurality of frames to which the gain adjustment algorithm is applied, comprising applying a gain adjustment algorithm for adjusting a gain value of a channel, wherein the gain adjustment algorithm is applied to each frame a number of times less than or equal to a predetermined threshold value There is provided an image correction method, characterized in that the parameter value for the channel gain of the sequence converges to a specific value or less.

Description

A METHOD FOR AUTO WHITE BALANCE OF IMAGE AND AN ELECTRONIC DEVICE TO PROCESS AUTO WHITE BALANCE METHOD

The present invention relates to a method for automatic white correction of an image and an electronic device for the same.

An image obtained by the imaging device for a three-dimensional object may be determined according to intrinsic characteristics (reflectance) of the object, direction, shape, position and spectrum of a light source, light-sensitive properties of the imaging device, and the like. The human eye has 'color constancy' ability to recognize the color of the same object in different lighting environments by adjusting the light spectrum reflected from the object and perceived by the eye.

However, the image sensor is not like the human eye. Therefore, the color of an object that appears white in natural light may be photographed in blue under lighting with a high color temperature (eg, fluorescent lamp) or may be photographed in red under lighting with a low color temperature (eg, incandescent lamp). That is, in the image sensor, depending on the light source, white distortion may occur in which pure white is photographed with different color information instead of pure white.

Therefore, the color of the object in the image sensor needs to be calibrated so that it is captured as natural as the color perceived by the human eye. When pixels that are originally white under a standard light source (usually D65) are corrected to be photographed with the original white color regardless of the light source, pixels of other colors can also be corrected to the color under the standard light source. This automatic correction process is called automatic white correction or color uniformity, and automatic white balance (hereinafter referred to as AWB) can be an important function to obtain a high-quality image from a camera.

On the other hand, the video camera may include a CCTV camera, an IPTV camera, etc., these cameras are fixed type and can be widely used for video surveillance purposes. When the camera is stationary, the background is fixed and the lighting does not change significantly during the day. Of course, the change of the foreground object occurs frequently, but generally, the scene on the screen takes up more of the background, so the sudden change of the scene is not frequent.

In general, full frame (30fps) video streaming processing is usually required for IP network cameras. At this time, image compression, image transmission, video stabilization, and motion detection must be achieved within a given interframe time (1/30 second). Real-time processing of automatic white correction of a video stream with Full-HD resolution (1920x1080p) should be done in much less than 1/30 second.

As mentioned above, in order to obtain a high-quality image from the camera, an automatic white correction procedure is essential. However, in an IP network camera that requires Full-HD Full-frame video streaming and other image processing requirements, it may be difficult to perform automatic white correction in real time due to a load required for data processing. Accordingly, an object of the present invention is to provide a method for automatic white correction processing of a high-quality image obtained from a camera.

The present invention comprises the steps of determining at least one candidate pixel having a gray color in natural light for each frame, and R (red), G (green), B (blue) for each frame so that the at least one candidate pixel becomes a gray point pixel. A frame including a plurality of frames to which the gain adjustment algorithm is applied, comprising applying a gain adjustment algorithm for adjusting a gain value of a channel, wherein the gain adjustment algorithm is applied to each frame a number of times less than or equal to a predetermined threshold value It provides an image correction method in which the parameter value for the channel gain of the sequence converges to a specific value or less.

According to an embodiment, the determining of the at least one candidate pixel may include identifying pixels having the same R value, G value, and B value for each frame, and the R value, the G value, and the B value are all and determining, among the same pixels, a pixel having a U value or a V value less than a preset threshold as a candidate pixel, wherein the U value and the V value are the R value, the G value, and the B value. V) may be a value obtained by converting to a color system.

According to an embodiment, each frame is divided into a plurality of windows, each window is divided into a plurality of pixel blocks, and the step of determining the candidate pixel and the step of applying the gain adjustment algorithm are performed on some pixel blocks among the plurality of pixel blocks. can only be performed for

According to an embodiment, in the case of an image having a resolution of 1920*1080, each frame is divided into 63 windows in a horizontal direction and 35 windows in a vertical direction, and the candidate pixel determination step and the gain adjustment algorithm within each window The number of pixel blocks in which the application step is performed may be four.

According to an embodiment of the present disclosure, the threshold value may increase as the arithmetic processing capability of the data processing processor constituting the electronic device increases.

According to an embodiment, the image correction method may further include determining whether a background scene of the image is changed, and initializing the gain adjustment algorithm when the background scene of the image is changed.

The present invention provides a candidate pixel determiner that determines at least one candidate pixel having a gray color in natural light for each frame, and R (red), G (green), and B for each frame so that the at least one candidate pixel becomes a gray point pixel. (blue) comprising an algorithm application unit for applying a gain adjustment algorithm for adjusting the gain value of the channel, wherein the algorithm application unit applies the gain adjustment algorithm to each frame by the number of times less than or equal to a predetermined threshold value, the gain adjustment algorithm is Provided is an electronic device in which a parameter value for a channel gain of a frame sequence including a plurality of applied frames converges to a specific value or less.

According to an embodiment, the candidate pixel determiner identifies a pixel having the same R value, G value, and B value for each frame, and a U value or V value among pixels having the same R value, the G value, and the B value. A pixel whose value is less than a preset threshold is determined as a candidate pixel, and the U value and the V value may be values obtained by converting the R value, the G value, and the B value into a YUV color system. .

According to an embodiment, each frame is divided into a plurality of windows, each window is divided into a plurality of pixel blocks, and the candidate pixel determiner determines at least one candidate pixel only for some pixel blocks among the plurality of pixel blocks. and the algorithm applying unit may apply the gain adjustment algorithm to only some pixel blocks among the plurality of pixel blocks.

According to an embodiment, the threshold value may increase as the amount of computation processing available by the algorithm application unit increases.

According to an embodiment, the algorithm application unit may determine whether a background scene of the image is changed, and when the background scene of the image is changed, initialize the gain adjustment algorithm.

According to an embodiment disclosed in the present invention, it is possible to efficiently support white processing correction even for high-quality images by reducing the amount of computation required for white correction processing.

In addition, the image correction method disclosed in the present invention can be applied not only to CCTV cameras and IP network cameras requiring real-time white correction, but also to systems requiring acquisition and display of all video images.

1 is a flowchart of an image correction method according to an embodiment disclosed in the present invention.
2 is a view for explaining a gain adjustment process according to an embodiment disclosed in the present invention.
3 is a view for explaining the structure of an image frame according to an embodiment disclosed in the present invention.
4 is a view for explaining a white correction method according to an embodiment disclosed in the present invention.
5 is a diagram for explaining the effect of the white correction method disclosed in the present invention.
6 is a block diagram of an electronic device for correcting an image according to an embodiment disclosed in the present invention.

Since the present invention can have various changes and can have various embodiments, specific embodiments will be described in detail with reference to the drawings. However, this is not intended to limit the present invention to specific embodiments, and it should be understood to include all modifications, equivalents and substitutes included in the spirit and scope of the present invention. In describing each figure, like reference numerals have been used for like elements.

Terms such as first, second, A, and B may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. and/or includes a combination of a plurality of related description items or any of a plurality of related description items.

When a component is referred to as being “connected” or “connected” to another component, it may be directly connected or connected to the other component, but it should be understood that other components may exist in between. something to do. On the other hand, when it is said that a certain element is "directly connected" or "directly connected" to another element, it should be understood that no other element is present in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. does not

Throughout the specification and claims, when a part includes a certain element, it means that other elements may be further included, rather than excluding other elements, unless otherwise stated.

A color space model is a mathematical model that expresses a color in terms of three or four color components, and when associated with an accurate description of how the color components can be interpreted, the resulting set of colors It is called color space. Standardized color space models include RGB, sRGB, HSV, YUV (YCbCr), CMYK, CIE XYZ, CIE RGB, CIE LUV, CIE LAB, and LMS models. sRGB (standard Red Green Blue) is an RGB color space model agreed upon by HP and Microsoft for use in monitors, printers, and the Internet, and is usually used as the default color space when image pixels are expressed in 8 bits for each channel. do.

The LMS color space model is a light ("S", 420 nm - 440 nm), middle ("M", 530 nm - 540 nm), and long ("L", 560 nm - 580 nm) long wavelength ("L", 560 nm - 580 nm), medium-wavelength ("M", 530 nm - 540 nm), short-wavelength ("S", 420 nm - 440 nm) spectra of the human eye cone optic nerve L cells, M cells, S It is a color space model representing the peak reactivity of cells. The CIE XYZ color space model is the standard color space model of the International Lighting Institute (CIE), which aims to encompass all color sensitivities seen by people with average vision. CIE XYZ is a device-independent color space representation and is a standard reference model for other color space models. In CIE XYZ, Y is the luminance, and Z is the reactivity of the S optic nerve cell, which indicates the approximate blue color value. X may represent a red color value as a reactivity of L optic nerve cells. In CIE XYZ, color expression is represented by three normalized real numbers (x, y, z) values as shown in Equation 1 below.

[Formula 1]

Meanwhile, the color of an object may appear differently depending on an illuminant. An illuminant may include characteristics such as a color temperature of the light source, a relative spectral power distribution of the light source, and a white point of the light source. The color temperature of the light source is expressed in absolute temperature (Kevin; °K) and can be defined as black body radiation. Solar light sources usually appear between 5,000°K and less than 6500°K.

A light source with a high color temperature (over 5,000°K) is called a cold color (blue-white) light source, and fluorescent lamps are an example. An example is an incandescent lamp. The spectral distribution of an illumination light source is determined by the color temperature of the light source. The white point of an illumination light source is the chromaticity of a white (white) object point under the illumination light source, and CIE 1931 or CIE 1931 or It is specified in chromaticity coordinates, such as those in the CIE 1964 chromaticity diagram, etc. The white point relates only to the chromaticity and is not affected by the intensity of the illuminating light source. Although it is uniquely determined, there may be several illumination sources corresponding to a given white point In case of the standard illumination light source D series, it is known that the spectral distribution can be mathematically derived from the chromaticity coordinates of the corresponding white point.

According to one embodiment, white calibration is usually accomplished in two steps: 1) Estimation of scene illumination light source color (illuminant light source color temperature or white point) based on several assumptions from the input image (Illuminant Color Estimation), 2) Using the estimated scene illumination light source color, the input image is re-expressed in color under standard light source Chromatic Adaptation, which color-corrects the input image as much as possible. According to various embodiments, the white point refers to a pixel in which R=G=B=1 when each color channel value is scaled to (0,1) among gray color points satisfying R=G=B.

로 나타나고, 표준광원 (예; D65) 하에서 화이트 포인트 LMS 좌표 가

라고 하면, LMS 색 공간에서 색순응 변환, M_LMS는 다음의 행렬 수식 2로 표현할 수 있다.According to an embodiment, the color adaptive transformation is mathematically expressed as a 3x3 matrix. For example, the LMS coordinates of the white point of an object at that illuminating source are

, and the white point LMS coordinates under a standard light source (eg D65) are

, the color adaptive transformation, M _LMS , in the LMS color space can be expressed by the following matrix Equation (2).

[Formula 2]

은 아래의 수식 3에 의해 구해질 수 있다.According to an embodiment, if the conversion from the CIE XYZ color space to the LMS color space is M _A , then in the CIE XYZ color space model, for the coordinate values (X,Y,Z) of a given object pixel, the color in the LMS is spatial coordinate values

can be obtained by Equation 3 below.

[Equation 3]

If the coordinate values in the sRGB (RGB) color space of the object pixel are (R, G, B), the coordinate values in the CIE XYZ color space are (X, Y, Z), and the conversion between them is M _X , Equation 4 can be established.

[Equation 4]

According to an embodiment, M _X in Equation 4 may be as follows.

Meanwhile, by summarizing the related transformations so far, the color adaptive transformation M _CAT in the sRGB (RGB) color space can be obtained through Equation 5 below.

[Equation 5]

Therefore, the sRGB(RGB) color space coordinate values (R _s , G _s , B _s ) of the object point in the corresponding light source are obtained by color adaptive transformation, as shown in Equation 6 It can be obtained as a coordinate value (R _d , G _d , B _d ).

[Equation 6]

According to an embodiment, the automatic white correction algorithms may be divided into three types. One may be a static method, the other may be a color gamut mapping method, and the last may be a learning-based method. According to various embodiments, static methods are methods applied to an image with fixed parameter settings, and may include 'gray world', 'white patch', and 'gray edge' methods.

According to an embodiment, the 'Gray World' algorithm starts under the assumption that the average color of a scene image obtained under white light (natural light) is an achromatic color (R=G=B; white, gray, and black). According to various embodiments, the extent to which the average color of a scene deviates from gray may be implied to be an effect of an illumination light source other than white light. Under this assumption, the calculated average value for each color channel of the acquired image frame is estimated as a gray color point in the corresponding illumination light source. That is, it can be calculated based on Equation 7 below.

[Equation 7]

로 근사화 될 수 있다. 즉, 그레이 월드 자동 백색 보정은 해당 프레임 픽셀의 R_s, G_s, B_s 값을 다음의 식에 따라서 백색 보정된 픽셀 값, R_d, G_d, B_d 각각은 다음 수식 8과 같을 수 있다.According to one embodiment, after this, Equation 6 may be applied to every pixel in the scene to achieve automatic white correction. According to various embodiments, the RGB color space model may be generally regarded as an approximation of the LMS color space model. Also, the gray point is

can be approximated as That is, in gray world automatic white correction _{, the R s} , G _s , B _s values of the corresponding frame pixels are white-corrected according _{to the following equation, and each of R d} , G _d , B _d may be the same as the following Equation 8 .

[Equation 8]

1 is a flowchart of an image correction method according to an embodiment disclosed in the present invention. According to an embodiment, the flowchart illustrated in FIG. 1 may be performed by the electronic device illustrated in FIG. 6 .

According to an embodiment, in operation S110 , at least one candidate pixel having a gray color in natural light may be determined for each frame. According to various embodiments, a pixel having the same R value, G value, and B value is identified for each frame, and a U value or V value is a preset threshold among pixels having the same R value, the G value, and the B value. A pixel less than the value may be determined as a candidate pixel. Meanwhile, the U value and the V value may be values obtained by converting the R value, the G value, and the B value into a YUV color system.

According to an embodiment, R=G=B may be established for the candidate pixel, and in this case, YUV conversion as shown in Equation 9 may be considered.

[Equation 9]

According to various embodiments, it can be confirmed that U=V=0 when R=G=B in Equation 9. That is, it can be confirmed that the candidate pixels satisfy the condition U=V=0.

According to an embodiment, under illumination other than natural light, candidate pixels may not satisfy U=V=0 and may have a color. For example, the candidate pixels may have very small U and V values compared to their Y values even under illumination with a color temperature different from natural light. Accordingly, the present invention discloses a method of determining a pixel that is less than a specific threshold value T as a candidate pixel based on Equation 10 below.

[Equation 10]

According to an embodiment, in step S120 , a gain adjustment algorithm for adjusting the gain values of the R, G, and B channels for each frame so that the candidate pixel becomes a gray point pixel may be applied. According to various embodiments, the gain control algorithm may be applied to each frame as many times as the number of times less than or equal to a predetermined threshold value, and a parameter value for a channel gain of a frame sequence including a plurality of frames to which the gain control algorithm is applied is a specific value. may converge to the following.

According to an embodiment, it may be determined whether the number of times of application of the gain adjustment algorithm exceeds a specific threshold in step S130. For example, when the threshold value is 5, the gain adjustment algorithm may be applied up to 5 times for each frame. According to various embodiments, the threshold value may be determined based on an arithmetic processing capability of a data processing processor constituting the electronic device. For example, the threshold value may increase as the arithmetic processing capability of the data processing processor increases.

Meanwhile, although not shown, the method may further include a step of checking whether a background scene of the image has been changed before step S110, and initializing a gain adjustment algorithm when the background scene of the image is changed. According to various embodiments, the gain adjustment algorithm may be initialized not only when the background scene is changed but also when the background scene is the start scene.

2 is a view for explaining a gain adjustment process according to an embodiment disclosed in the present invention. The gain adjustment process shown in FIG. 2 may be a procedure performed by step S120 of FIG. 1 .

로 정의할 수 있다. 따라서 i번째 반복 때의 I_i 및 O_i의 R, G, B 채널 값을 각각

라 하면,

의 관계식이 성립할 수 있다.According to an exemplary embodiment, I _i may mean an input pixel at the i-th iteration, and O _i may mean an output pixel at the i-th iteration of which the gain is adjusted. According to various embodiments, G _i may mean the gain of each R, G, and B color channel at the i-th iteration, and thereafter, the gains of each R, G, and B color channel at the i-th iteration, respectively.

can be defined as Therefore, the R, G, and B channel values _{of I i} and O _{i at} the i-th iteration are respectively

If you say

relationship can be established.

According to an embodiment, Φi may be defined as in Equation 11 below.

[Equation 11]

및

는 수식 10에서 구한 후보 픽셀들의 U, V 들의 평균값을 의미할 수 있다. 일 실시예에 따르면, 희망 신호 값 di는 0이 될 수 있으며,

일 수 있다. 다양한 실시예에 따르면, 도 1에서 게인 조절 알고리즘은 다음 수식 12와 같을 수 있다.in formula 11

and

may mean an average value of U and V of the candidate pixels obtained in Equation 10. According to one embodiment, the desired signal value di may be zero,

can be According to various embodiments, the gain adjustment algorithm in FIG. 1 may be as shown in Equation 12 below.

[Equation 12]

는 다음 수식 13과 같을 수 있다.According to an embodiment, μ is a constant and may be a parameter that determines the convergence speed of the algorithm. According to various embodiments,

may be as in Equation 13 below.

[Equation 13]

가 충분히 작아질 때까지 반복적으로 수행(예를 들어 수식 10, 11, 12의 과정이 반복적으로 수행)될 수 있다. According to one embodiment, a=0.8, b=0.15 may be, and R, G, B gain value adjustment for automatic white correction is

It may be iteratively performed (for example, the processes of Equations 10, 11, and 12 are repeatedly performed) until is sufficiently small.

3 is a view for explaining the structure of an image frame according to an embodiment disclosed in the present invention.

According to an embodiment, the input image frame may be divided into windows as shown in FIG. 3 . According to various embodiments, the input image frame may be divided into various sizes and numbers of windows based on parameters such as window sizes (WINW, WINH), number (WINHC*WINVC), and start positions (WINSH, WINSV).

According to an embodiment, each window may be sub-sampled and divided into 2*2 pixel blocks. According to various embodiments, AEWINCH and AEWINCV may be parameters indicating a horizontal length and a vertical length between 2*2 pixel blocks, respectively. On the other hand, in the case of 1920*1080 size video, if the video frame is divided into WINHC=63, WINVC=35 windows in the horizontal and vertical direction and the distance between blocks is sampled with AEWINCH=15 and AEWINCV=15, respectively, 8820 pixel blocks ( 63*35*4) may be formed.

According to an embodiment, the load required for the automatic whiteness correction algorithm may be reduced by applying the automatic whiteness correction algorithm to only some pixel blocks in each frame.

4 is a view for explaining a white correction method according to an embodiment disclosed in the present invention.

According to an embodiment, step S410 may be an initialization step for automatic white correction. According to various embodiments, application of the gain adjustment algorithm may be initialized at the time of a start scene or a scene change in step S410 .

및 이들 합인

를 구한 후에, 모든 픽셀에 대해 수식 8을 적용할 수 있으며, 샘플링된 각 픽셀 블록의 각 채널 R, G, B값의 합으로부터 이를 각각 Y, U, V로 변환한후에 수식 9를 적용하여 점검한 후에 이를 만족하면 해당 픽셀 블록을 그레이 컬러 포인트 후보 픽셀로 결정할 수 있다.According to an embodiment, step S420 may be a graypoint candidate detection step. According to various embodiments, a gray point candidate may be detected for every frame of an image. Average of each channel R, G, B of pixel blocks sampled, e.g. upon initialization or scene change detection

and their sum

Equation 8 can be applied to all pixels after obtaining If this is satisfied later, the corresponding pixel block may be determined as a gray color point candidate pixel.

를 이용하여 수식 11을 적용해 φ를 구할 수 있으며, φ를 이용하여 프레임의 모든 픽셀에 대해 수식 12를 적용할 수 있다.According to an embodiment, step S430 may be a graypoint-based white correction adjustment step. According to various embodiments, gray point-based automatic white adjustment may be performed for every frame. For example, average values of U and V of gray color point candidate pixel blocks through step S420

ϕ can be obtained by applying Equation 11 using

5 is a diagram for explaining the effect of the white correction method disclosed in the present invention.

Fig. 5 (a) is a view showing an image before the white correction method is applied, and Fig. 5 (b) is a view showing an image to which the white correction method proposed in the present invention is applied. On the other hand, Figure 5 (c) is a view showing an image seen with the naked eye. That is, according to the embodiment disclosed in the present invention, it can be confirmed that the image is obtained similarly to the image seen with the naked eye. In addition, according to an embodiment of the present invention, it can be confirmed that the white correction process is efficiently performed at a high speed even on a high-resolution image.

6 is a block diagram of an electronic device for correcting an image according to an embodiment disclosed in the present invention.

According to an embodiment, the electronic device 600 includes a candidate pixel determiner 610 that determines at least one candidate pixel having a gray color in natural light for each frame, and each frame so that the at least one candidate pixel becomes a gray point pixel. Each R (red), G (green), B (blue) may include an algorithm applying unit 620 for applying a gain adjustment algorithm for adjusting the gain value of the channel. According to various embodiments, the algorithm application unit 620 applies the gain adjustment algorithm to each frame by the number of times less than or equal to a preset threshold, and a channel gain of a frame sequence including a plurality of frames to which the gain adjustment algorithm is applied. A parameter value for ? may converge to a specific value or less.

According to an embodiment, the candidate pixel determiner 610 identifies pixels having the same R value, G value, and B value for each frame, and among pixels having the same R value, G value, and B value A pixel having a U value or a V value less than a preset threshold is determined as a candidate pixel, and the U and V values are obtained by converting the R value, the G value, and the B value into a YUV color system. may be a given value.

The above description is merely illustrative of the technical idea of the present invention, and various modifications and variations will be possible without departing from the essential characteristics of the present invention by those skilled in the art to which the present invention pertains. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the technical spirit of the present invention is not limited by these embodiments. The protection scope of the present invention should be construed by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present invention.

Claims (11)

An image correction method performed by an electronic device, comprising:
checking pixels having the same R value, G value, and B value for each frame;
determining, as a candidate pixel, a pixel having a U value or a V value less than a preset threshold among pixels having the same R value, the G value, and the B value; and
and applying a gain adjustment algorithm for adjusting the gain values of R (red), G (green), and B (blue) channels for each frame so that at least one of the candidate pixels becomes a gray point pixel,
The gain control algorithm is applied to each frame by a number of times less than or equal to a preset threshold, and the parameter value for the channel gain of a frame sequence including a plurality of frames to which the gain control algorithm is applied is converged below a specific value. doing,
How to edit video. According to claim 1,
The U value and the V value are values obtained by converting the R value, the G value, and the B value into a YUV color system,
How to edit video. According to claim 1,
Each frame is divided into a plurality of windows, and each window is divided into a plurality of pixel blocks, and the step of determining the candidate pixel and the step of applying the gain adjustment algorithm are performed only for some pixel blocks among the plurality of pixel blocks. doing,
How to edit video. 4. The method of claim 3,
In the case of a resolution 1920*1080 size image, each frame is divided into 63 windows in the horizontal direction and 35 windows in the vertical direction, and pixels in which the candidate pixel determination step and the gain adjustment algorithm application step are performed within each window. The block is characterized in that there are 4,
How to edit video. According to claim 1,
The threshold value is characterized in that the larger the arithmetic processing capability of the data processing processor constituting the electronic device, the greater,
How to edit video. According to claim 1,
determining whether a background scene of the image has been changed; and
When the background scene of the image is changed, further comprising the step of initializing the gain adjustment algorithm,
How to edit video. An electronic device for correcting an image, comprising:
For each frame, pixels having the same R value, G value, and B value are identified, and among pixels having the same R value, the G value, and the B value, a pixel having a U value or a V value less than a preset threshold is selected as a candidate pixel a candidate pixel determiner that determines and
An algorithm application unit for applying a gain adjustment algorithm for adjusting the gain values of R (red), G (green), and B (blue) channels for each frame so that at least one of the candidate pixels becomes a gray point pixel,
The algorithm applying unit applies the gain adjustment algorithm to each frame by a number of times less than or equal to a predetermined threshold, and the parameter value for the channel gain of a frame sequence including a plurality of frames to which the gain adjustment algorithm is applied converges to a specific value or less. characterized by being
electronic device. 8. The method of claim 7,
The U value and the V value are values obtained by converting the R value, the G value, and the B value into a YUV color system,
electronic device. 8. The method of claim 7,
Each frame is divided into a plurality of windows, each window is divided into a plurality of pixel blocks, the candidate pixel determining unit determines at least one candidate pixel for only some pixel blocks among the plurality of pixel blocks, and the algorithm applying unit characterized in that the gain adjustment algorithm is applied only to some pixel blocks among the plurality of pixel blocks,
electronic device. 8. The method of claim 7,
The threshold value is characterized in that the larger the amount of arithmetic processing possible by the algorithm application unit, the larger,
electronic device. 8. The method of claim 7,
The algorithm application unit checks whether the background scene of the image is changed, characterized in that when the background scene of the image is changed, to initialize the gain adjustment algorithm,
electronic device.

Images