KR101026049B1 - White balance adjusting apparatus and method considering effect of single tone image - Google Patents

Abstract

PURPOSE: A white balance adjusting apparatus and a method for considering effect of a single tone image are provided to solve a problem of white balance defects due to an effect of an excessive single tone image. CONSTITUTION: A single tone image determining unit(14) respectively compares the maximum value of the histogram of the Cb value or the Cr value with a preset first threshold value. The single tone image determining unit decides an input image frame which has bigger than the first threshold value for a single tone image. An auto white balance gain computing unit(17) computes the automatic white balance gain using a color information of the input image frame excluding the decided input image frame.

Description

WHITE BALANCE ADJUSTING APPARATUS AND METHOD CONSIDERING EFFECT OF SINGLE TONE IMAGE}

The present invention relates to the field of digital image processing, and more particularly, to prevent the automatic white balance from falling down in consideration of the effect of the single color on the gain calculation for automatic white balance adjustment in a monochromatic image having a large distribution of single colors. The present invention relates to an automatic white balance adjustment apparatus and method considering the influence of a monochrome image.

In general, the human vision is characterized by always recognizing the same color with respect to the same subject even if the light source is changed. That is, even when the light source changes to a light source having a different color temperature such as sunlight, fluorescent lamps, or incandescent lamps, the human vision always recognizes white as the same white color. On the contrary, since the image sensor to which the solid-state image sensor is applied is a device that converts light into an electrical signal, the color of the subject that changes according to the light source is detected as it is. Therefore, in order to be able to express the color detected from such an image sensor similar to the human vision, most digital image processing devices (e.g., digital cameras) have automatic white balance adjustment for automatically adjusting the white balance. The algorithm is applied.

Typically, automatic white balance adjustment made in the digital image processing apparatus is performed by causing a white subject having an intermediate brightness in the image to be represented by a color signal (RGB signal) having the same output levels. The most common algorithm used for the automatic white balance adjustment is to calculate the color average value by assuming the color average value of the screen as white regardless of the variation of the light source, and then obtain a gain of the color signal level that can shift the color average value to white. Applied to the whole image. For example, a conventional automatic white balance adjustment method integrates Cb and Cr values, which are color difference signals of pixels belonging to a white estimation region in an input image, and calculates average values thereof to obtain gains of red (R) and blue (B). The method of adjustment was adopted.

However, these conventional automatic white balance adjustment techniques can calculate the gain for white balance in a monochromatic image in which one color is very widely distributed. Therefore, when the calculated gain is applied, the white balance is broken. do. Therefore, a technique for determining whether an input image is a monochrome image is required for an automatic white balance adjustment algorithm.

As a conventional method of determining a monochrome image, a method of determining a monochrome image by comparing values with adjacent pixels in consideration of all data of RGB of an input image and accumulating data on a difference in value between adjacent pixels is known. have. This conventional method can reduce the efficiency of the system and determine the ideal monochromatic image by comparing with neighboring pixels, but the monochromatic image input from the camera sensor is not a case where all of the monochromatic image is concentrated in one specific area. Because it is composed in the form of concentration around the area, if the comparison error between neighboring pixels continues to accumulate, there is a high probability that an error of judgment will eventually occur. In addition, the comparison with neighboring pixels can be considered as a monochromatic image when considering only one frame to prevent color distortion, but in a composite image where various colors are distributed relatively uniformly, a single color is widely distributed in one color. There is a problem in that the progress to the image cannot be determined.

As another method of determining a monochrome image, it is assumed that the center of the screen is the area of the subject for camera shooting, and the contour area determined as the subject in the area is judged, and the determined contour area is divided into blocks of a predetermined size and the subject area in the center of the screen. A method of determining a monochromatic image by comparing an average of a luminance value and a scatter diagram of a is known. In this method, the load on the system may be increased in determining the contour area of the center of the screen and dividing the area into blocks of a certain size, and the desired result may not be obtained if the assumption about the subject area is out of order. There are disadvantages. In addition, this method also has a disadvantage in that an incorrect result can be obtained when the above case occurs because there is no consideration of the process from the composite image to the monochrome image.

The present invention is to solve the problem to provide an automatic white balance adjustment device and method that can prevent the white balance is broken even if the automatic white balance adjustment when there is an excessively monochromatic region in the generation of still images or movies It is a technical problem.

In addition, an object of the present invention is to provide an automatic white balance adjustment apparatus and method capable of reducing the amount of computation and hardware size for automatic white balance adjustment.

According to an aspect of the present invention,

A sampling interval setting unit for determining a pixel sampled in the input image frame according to a size of an input image frame;

A histogram generator for generating a histogram of Cb and Cr values for the pixels to be sampled determined by the sampling interval setting unit;

The maximum value of the histogram of the Cb value or the maximum value of the histogram of the Cr value is compared with a first preset threshold value, and an input image frame having at least one of the maximum values greater than the first threshold value is monochrome. Monochromatic image determination unit judging as; And

Automatic in consideration of the influence of the monochrome image including an automatic white balance gain calculator which calculates and outputs an automatic white balance gain using color information of the input image frame except the input image frame determined as the monochrome image by the monochrome image determination unit. Provides a white balance adjustment device.

In one embodiment of the present invention, the automatic white balance gain calculator is calculated for the previous input image frame without calculating the automatic white balance gain for the input image frame determined as the monochrome image by the monochrome image determiner. Auto white balance gain can be output.

In one embodiment of the present invention, the histogram generator may create a histogram of one of the values Cb and Cr for one input image frame.

In one embodiment of the present invention, the Cb value and Cr adjacent to each other with a frequency greater than a second predetermined threshold value in the histogram of the Cb value and Cr value of the input image frame not determined as the monochrome image by the monochrome image determination unit. The display apparatus may further include a removal target pixel determination unit configured to group values, and determine pixels belonging to a group whose Cb value and Cr value range is narrower than a preset reference range as pixels not used for automatic white balance adjustment. In this embodiment, the automatic white balance gain calculator may calculate the automatic white balance gain using the remaining pixels except the pixels determined as pixels not used for the automatic white balance adjustment from the input image frame.

An embodiment of the present invention may further include a removal amount determination unit for comparing the number of pixels determined to be not used for automatic white balance adjustment by the removal target pixel determination unit with a third preset threshold value. If the number of pixels determined not to be used for white balance adjustment is larger than the third threshold value, the automatic white balance gain calculation unit does not calculate the automatic white balance gain for the corresponding input image frame, but for the previous input image frame. The calculated auto white balance gain can be output.

As another means for solving the above technical problem, the present invention,

Determining a pixel sampled in the input image frame according to the size of an input image frame;

Creating a histogram of Cb and Cr values for the pixels determined to be sampled in the step of determining the sampled pixels;

Comparing the maximum value of the histogram of the Cb value or the maximum value of the histogram of the Cr value with a first predetermined threshold value, respectively; And

Determine an input image frame in which at least one of the maximum values is larger than the first threshold value as a monochrome image, and use an automatic white balance gain calculated for the input image frame except for the input image frame determined as the monochrome image. The present invention provides an automatic white balance adjustment method in consideration of the influence of a monochrome image including the step of performing automatic white balance adjustment.

In one embodiment of the present invention, the performing of the automatic white balance adjustment may include: calculating the automatic white balance calculated for the previous input image frame without calculating the automatic white balance gain for the input image frame determined as the monochrome image; The automatic white balance adjustment may be performed using the gain.

In an embodiment of the present disclosure, generating the histogram may include generating a histogram of one of a Cb value and a Cr value for one input image frame.

According to one embodiment of the present invention, when all of the maximum values are not greater than the first threshold value, Cb adjacent to each other with a frequency greater than a second predetermined threshold value in the histogram of the Cb value and the Cr value of the input image frame. Grouping values and Cr values, respectively, and comparing the ranges of the Cb and Cr values of the groups with a preset reference range;

Determining pixels belonging to a group in which the ranges of the Cb and Cr values are narrower than a preset reference range as pixels not used for automatic white balance adjustment; And

The method may further include calculating an automatic white balance gain using the remaining pixels except the pixels determined as pixels not used for the automatic white balance adjustment from the input image frame.

In one embodiment of the present invention, the method further comprises: comparing the number of pixels determined not to be used for the automatic white balance adjustment with a third preset threshold value, and the number of pixels determined not to be used for the automatic white balance adjustment. If greater than the third threshold value, performing the automatic white balance using the automatic white balance gain calculated on the previous input image frame without calculating the automatic white balance gain on the corresponding input image frame; Can be.

According to the present invention, when performing the automatic white balance, it is possible to solve the problem that the white balance is broken by the influence of the excessive monochromatic region included in the image frame to implement a natural color image.

In addition, according to the present invention, a natural color can be realized in a video by excluding a monochromatic region that increases when performing automatic white balance in the process of a composite image from a monochrome image to an automatic white balance gain calculation.

In addition, according to the present invention, since only some pixels, not all pixels of the input image frame are used, and only one histogram of Cr or Cb is created for one image frame, it is possible to reduce the amount of computation and hardware required for determining a monochrome image. It can be effective.

Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, embodiments of the present invention may be modified in various other forms, and the scope of the present invention is not limited to the embodiments described below. Embodiment of this invention is provided in order to demonstrate this invention more completely to the person skilled in the art to which this invention belongs. Therefore, it should be noted that the shape and size of the components shown in the drawings may be exaggerated for more clear explanation.

1 is a block diagram of a white balance adjustment device according to an embodiment of the present invention.

As shown in FIG. 1, the white balance adjusting apparatus according to the exemplary embodiment of the present invention includes a color coordinate converting unit 11, a sampling interval setting unit 12, a histogram generating unit 13, and a monochrome image determining unit 14. ), A removal target pixel determination unit 15, a removal amount determination unit 16, an automatic white balance (AWB) gain calculator 17, and an automatic white balance application unit 18.

The color coordinate converter 11 converts an image frame expressed in RGB color coordinates into a YCbCr color coordinate. In general, a signal captured by an image sensor and outputting a scene represents a Bayer pattern in which each pixel is represented by one of the colors of RGB, and each pixel is represented as an RGB value by applying a color interpolation algorithm to an image frame output in the Bayer pattern. Is converted to RGB color coordinates. The color coordinate conversion unit 11 of the present invention converts an image frame expressed by this RGB color coordinate into a YCbCr color coordinate.

The sampling interval setting unit 12 determines a pixel sampled from an input image frame. That is, the sampling interval setting unit 12 sets the sampling interval in the horizontal or / and vertical direction to select some pixels from all the pixels included in the input image frame, and includes only pixels corresponding to the sampling interval. To generate down-sampled image frames. The sampling interval setting unit 12 is for generating an image frame having an appropriate size for determining a monochrome image to be performed later, and may set the sampling interval appropriately according to the size of the input image frame.

The histogram generator 13 generates a histogram for Cb or Cr for an image frame including pixels determined to be sampled by the sampling interval setting unit 12 using the converted YCbCr color coordinates. That is, the Cb or Cr value of each pixel of the input image frame is grasped, and the histogram or the x-axis represents the Cr value and the y-axis represents the frequency and the y-axis represents the Cb value, and the y-axis represents the frequency. Create a histogram of each. Preferably, the histogram generator 13 may generate a histogram of one of a Cb value and a Cr value with respect to one sampled image frame. Selection of the Cb value or Cr value for generating such a histogram may be performed alternately according to the order of the input frames.

According to an exemplary embodiment of the present invention, an operation amount of a single color image is determined by sampling a portion of pixels of an image frame by the sampling interval setting unit 12 and calculating a histogram with respect to one color value for one image frame. And hardware size can be reduced.

The monochrome image determination unit 14 compares a maximum value of each of the histograms of Cb and Cr generated by the histogram generator 13, that is, a maximum frequency and a preset first threshold value, so that at least one of the maximum values is determined. If larger than the first threshold value, the corresponding image frame is determined as a monochrome image.

The removal target pixel determination unit 15 analyzes the histograms of Cb and Cr to determine whether the corresponding image frame is changed from the composite image to the monochrome image, and to prevent the monochrome region from affecting the auto white balance. The pixels of the monochrome region are determined to remove the monochrome region in the balance gain operation. The removal target pixel determination unit 15 groups adjacent values among the values of Cb and Cr having a frequency greater than a second preset threshold in the histograms of Cb and Cr, and the ranges of Cb and Cr of these groups are If it is narrower than the preset reference range, the pixels belonging to this group are determined to be pixels included in the monochrome region, and the automatic white balance gain can be calculated using the remaining pixels except for the pixels belonging to the group in the corresponding image.

The removal amount determination unit 16 determines the number of pixels determined to be excluded from the auto white balance gain calculation by the removal target pixel determination unit 15, and determines that an excessive number of pixels should be excluded from the gain calculation. In this case, the automatic white balance gain calculation is blocked from performing a new operation on the determined image frame.

The automatic white balance gain calculator 17 calculates an automatic white balance gain that can change a color value by using various techniques known in the art, using the color information of each pixel of the input image frame. The automatic white balance gain calculator 17 calculates the automatic white balance gain by using the color information of the input image frame, and when the input image frame is determined to be a monochrome image by the monochrome image determiner 14, the corresponding input image. It is possible to output the automatic white balance gain calculated for the previous input image frame without calculating the automatic white balance gain using the frame. Also, the automatic white balance gain calculator 17 excludes the pixels determined to be excluded from the automatic white balance gain calculation by the removal target pixel determination unit 15 from the input image frame and uses the color information of the remaining pixels. Automatic white balance gain can be calculated. In addition, when the number of pixels determined by the removal amount determining unit 16 to be excluded from the automatic white balance gain calculation by the removal amount determining unit 16 is excessive, the automatic white balance gain calculating unit 17 is excessively large. Instead of calculating the automatic white balance gain using the corresponding input image frame, the automatic white balance gain calculated for the previous input image frame may be output as it is.

The automatic white balance gain applying unit 18 applies the automatic white balance gain calculated and output according to the color state of the input image frame by the automatic white balance gain calculating unit 17 to the input image frame to automatically adjust the white balance. Create an image frame.

The operation and effects of the automatic white balance adjustment device in consideration of the influence of the monochrome image according to the exemplary embodiment of the present invention configured as described above will be described in more detail with reference to the accompanying drawings.

2 is a flowchart illustrating an automatic white balance adjustment method considering the influence of a monochrome image according to an embodiment of the present invention.

As shown in FIG. 1 and FIG. 2, in one embodiment of the present invention, an input image frame in which each pixel is expressed in RGB color coordinates is converted by the color coordinate converting unit 11 into YCbCr color coordinates (S21) and the input image frame. In operation S22, a sampling interval for performing sampling for selecting some pixels may be set.

Subsequently, the histogram generator 13 generates a histogram of Cb or Cr in order to consider the color information of the image frame including the pixel selected according to the sampling interval (S23). In this case, a histogram of one of a Cb value and a Cr value may be generated for one sampled image frame. The X-axis of the histogram has a value from 0 to 255, which represents the range of values that Cb and Cr can have in one pixel. Each histogram X-axis can generate a histogram by dividing it into 256 groups by considering all values from 0 to 255 one by one, but the burden on the system can be increased. Therefore, 256 groups generated from 0 to 255 are considered by grouping a range of values into each other so that the efficiency of the system can be improved by reducing the group of data for analyzing the histogram to less than 256. At this time, if the value of one group is smaller, the histogram accuracy is increased, but the complexity of the system is increased. If the value of one group is larger, the number of groups to be considered is reduced, which reduces the burden on the system. However, the accuracy of histograms can be lowered. For example, if you specify a group size of 4 from 0 to 255, the group is divided into 0 ~ 3, 4 ~ 7, ..., 248 ~ 251, 252 ~ 255 to 64 groups in total. You can create a histogram. The Y-axis of the histogram represents the frequency at which the Cb and Cr values of the X-axis appear in consideration of the entire pixels of the input image frame. In this case, the frequency expressed on the Y-axis may represent the number of pixels representing the corresponding Cb and Cr values as they are, but the actual number of pixels may be divided by the total number of pixels and normalized to 0-1. The histograms of Cb and Cr generated as described above can know how many pixels are distributed in the respective values of Cb and Cr compared to the entire image.

Subsequently, the monochrome image determination unit 14 analyzes the histograms of Cb and Cr created on the sampled input image frame to determine whether the input image frame is a monochrome image (S24).

3 and 4 illustrate an example of a histogram of Cb and Cr generated by an embodiment of the present invention. FIG. 3 illustrates a histogram of a composite image in which regions of one color are not widely distributed. 4 shows a histogram of a monochrome image in which one color region is widely distributed.

As shown in FIG. 3, the histogram distribution of Cb and Cr of the composite image shows a form in which the distribution of Cb and Cr data is widely spread due to the distribution of various colors in the entire image. Thus, the Y-axis, That is, it can be seen that the maximum value of the frequency does not appear large. In contrast, the distribution of the Cb and Cr histograms of the monochrome image of FIG. 4 is concentrated in a portion corresponding to a single color, thereby increasing the number of cumulative counting pixels. Thus, the maximum value of the Cb and Cr histograms is also the histogram of the composite image. It can be seen that the value is relatively larger than the maximum value of. In this case, the definition of a monochrome image is not a monochromatic image in which ideally the RGB values of the image are completely concentrated in one value. Instead, a monochromatic image may be judged as a monochrome image in a variety of images input from an image sensor, that is, a natural environment. It includes wall surfaces, monochromatic fibers, and the like. Therefore, due to the influence of the characteristics of the image sensor used or the reflected light of the illumination, the RGB values of the image are not formed intensively all at one value, but rather are formed around the one value. Therefore, in the present invention, it is possible to determine a monochrome image in consideration of this characteristic of the monochrome image.

In one embodiment of the present invention, by using the characteristics of the histogram of the composite image and the monochromatic image, a predetermined threshold value Th1 obtained through experiments using various images is predetermined, and the maximum value of the Cb and Cr histograms is determined. If the threshold value is smaller than the threshold Th1, the image is judged as a composite image, and auto white balance is normally performed. If the maximum value of the Cb Cr histogram is larger than the threshold value, the monochrome image is judged as a monochrome image, and a new automatic white balance gain is not generated. Instead, by applying the automatic white balance gain generated by the previous input image frame (S27), it is possible to prevent the discoloration of the monochrome image. In addition, in the case of a general monochrome image, the Cb histogram and the Cr histogram have similar shapes, but in some monochrome images, the histograms of Cb and Cr are not similar, and the histogram of one of the histograms of Cb or Cr is a composite image. The histogram may be similar to that of the other, and the other one may show the histogram of the monochrome image. In consideration of such a case, in the present invention, in order to determine a monochrome image, it is preferable to determine that the monochrome image is the maximum value of the Cb histogram or the maximum value of the Cr histogram.

By determining whether the input image frame is a monochromatic image as described above, the performance of automatic white balance can be improved in obtaining a fixed still image. However, if a video is to be provided as a result, several frames are performed in one second. Even if a single frame is judged as a monochrome image, color distortion may occur due to the influence of the monochrome image occupying a part in the composite image during the process from the composite image to the monochrome image or from the monochrome image to the composite image. In such a case, there is no problem in the process of progressing from a monochrome image to a composite image anyway because a new automatic white balance gain is generated for the automatic white balance in the composite image anyway, but it is judged as a monochrome image when proceeding from the composite image to the monochrome image. Until then, the automatic white balance gain is calculated, resulting in color distortion due to the influence of the monochromatic region occupying a portion of the composite image. That is, after completing the process of the monochrome image, the automatic monochrome balance gain is no longer changed due to the above-described monochrome image determination process (S24), and the distortion is maintained because the automatic white balance gain before the progress to the monochrome image is maintained. Although the monochromatic image is not discolored because the discolored color is continuously maintained, a problem occurs that the distorted color is maintained. In order to solve this problem, an embodiment of the present invention includes a technique for preventing distortion of colors occurring in the process of moving from a composite image to a monochromatic image when generating a moving image result.

FIG. 5 illustrates an example image of a process of proceeding from a composite image to a monochrome image in units of frames at a predetermined interval, and FIG. 6 illustrates Cb and Cr histograms of some image frames of FIG. 5. As shown in FIG. 5 and FIG. 6, in the process of proceeding from the composite image to the monochrome image, it may be seen that the frequency of the partial histogram is gradually increased due to the influence on the monochromatic region while gradually exhibiting the characteristics of the composite image. . That is, after the monochromatic region is increased, it can be seen that the histogram data of the first composite image is almost disappeared and is composed of the data generated by the monochromatic region. Therefore, it can be judged that the data of the histogram, in which the frequency of some areas of the histogram increases significantly, is the influence of the monochromatic color in the process from the composite image to the monochrome image. According to the above determination, it is preferable to remove the influence of the monochrome image proceeding in the composite image and to perform the automatic white balance gain calculation by using the data of the remaining composite image before it is completely determined as the monochrome image.

In order to remove the histogram data generated by the influence of the monochromatic region, the form in which the data is formed in the histogram for the input image frame is compared to determine whether the group is a composite image or a group based on the ongoing monochrome image. To this end, in the present invention, in the histogram of Cb, Cr, Cb, Cr values of adjacent X axes having a value larger than the preset threshold value Th2 are formed as a group, and the ranges of the Cb and Cr values of each group are preset and set. By comparing the reference range, it is determined that the group whose Cb and Cr values are smaller than the preset reference range proceeds to the monochromatic region (S25). Pixels belonging to the group determined to proceed to the monochrome region are excluded from the automatic white balance gain calculation. 7 shows an example of a histogram form that reproduces the process from a composite image to a monochrome image. In FIG. 7, the second group G2 represents the shape of the histogram data of the composite image, and the first group G1 represents the shape of the histogram data of the monochrome region that occurs in the process from the composite image to the monochrome image. . In general, even when the process from the composite image to the monochromatic image is performed, the group is not formed as completely independent data, but is divided into two groups as shown in FIG. 7, but overall, a small amount of data remains between the two groups. Done. Therefore, in order to clearly distinguish the two groups of the histogram data of the composite image and the histogram data of the monochromatic image, a frequency larger than a predetermined threshold (Th2) to remove the influence of very small data formed between the two groups By considering only the case of, the group by the composite image and the group by the monochromatic region can be clearly distinguished. In order to determine the date based on the monochromatic region, each group compares the X-axis, that is, the range of Cb and Cr values of each group. In general, the histogram data of a composite image is composed of various color information, and is spread in a wide form. The histogram data of a monochromatic image is composed of various color information and is concentrated in one place. Therefore, in the histogram having a value from 0 to 255 of the X axis, the composite image data has a group area whose size of the group area occupied on the X axis is larger than the size of the group area occupied by the monochrome image on the X axis. Therefore, the reference range that can be judged as a monochromatic region is set in advance, and among the groups shown in the histogram, a group having an x-axis range smaller than this reference range, that is, a range of Cb and Cr, is determined as a group based on a monochrome image. The pixels belonging to the group are excluded from the gain calculation for the automatic white balance (S28), and then the automatic white balance gain is calculated (S29).

On the other hand, when the pixels belonging to the monochromatic region are removed in the automatic white balance gain calculation as described above, a problem may occur in which a sufficient number of pixels are not used for the white balance gain calculation due to excessive pixel removal. For example, as shown in FIG. 8, one or more types of monochrome images may be input as the input image frame. In this case, the histogram data is not determined as a monochrome image because the maximum value of the Y-axis of the histogram is smaller than the first threshold value in the process of determining a monochrome image as shown in FIG. 9, and the pixels of the group included in the monochrome region are not determined. The condition in the process of removing from the automatic white balance gain operation S25 is satisfied, so that most pixels may be excluded from the white balance gain operation. In this case, when the automatic white balance gain is calculated and applied, the number of pixels used for the white balance gain calculation may be too small, resulting in a severe color distortion or an undesirable result. In order to prevent such a case, when the histogram data is removed due to the influence of the monochromatic region, the amount of pixels excluded from the automatic white balance gain calculation is determined (S26). This is because it is determined that more than one monochrome image is input and does not perform a gain operation for automatic white balance. Therefore, by applying the automatic white balance gain calculated by the previous input image frame in the same way as when the monochrome image is determined, This solves a problem that may occur because a large number of pixels are excluded from the operation.

In the detailed description of the present invention, specific embodiments have been described, but various modifications may be made without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the following claims and their equivalents.

1 is a block diagram of a white balance adjustment device according to an embodiment of the present invention.

2 is a flowchart illustrating an automatic white balance adjustment method considering the influence of a monochrome image according to an embodiment of the present invention.

3 and 4 are examples of histograms of Cb and Cr produced by one embodiment of the present invention.

FIG. 5 is a diagram illustrating an image of a process of proceeding from a composite image to a monochrome image in units of frames at a predetermined interval, and FIG. 6 is a Cb and Cr histogram of some image frames of FIG. 5.

7 is a histogram showing the progress of a composite image to a monochrome image.

8 is a diagram illustrating an example of a monochrome image composed of two colors.

9 is a histogram of the monochrome image illustrated in FIG. 8.

* Description of the symbols for the main parts of the drawings *

11: color coordinate conversion unit 12: sampling interval setting unit

13: histogram generator 14: monochrome image determination unit

15: removal target pixel determination unit 16: removal amount determination unit

17: automatic white balance gain calculator

18: Automatic white balance gain application

Claims (10)

A sampling interval setting unit for determining a pixel sampled in the input image frame according to a size of an input image frame; A histogram generator for generating a histogram of Cb and Cr values for the pixels to be sampled determined by the sampling interval setting unit; The maximum value of the histogram of the Cb value or the maximum value of the histogram of the Cr value is compared with a first preset threshold value, and an input image frame having at least one of the maximum values greater than the first threshold value is monochrome. Monochromatic image determination unit judging as; And An automatic white balance gain calculator for calculating and outputting an automatic white balance gain using color information of an input image frame except for an input image frame determined as a monochrome image by the monochrome image determination unit. Automatic white balance adjustment device considering the influence of the monochrome image including a. The method of claim 1, The automatic white balance gain calculator is The monochrome image determining unit outputs the automatic white balance gain calculated for the previous input image frame without calculating the automatic white balance gain for the input image frame determined as the monochrome image. Considered automatic white balance adjustment device. The method of claim 1, wherein the histogram generator, An automatic white balance adjustment apparatus considering the influence of a monochrome image, characterized by generating a histogram of one of a Cb value and a Cr value for one input image frame. The method of claim 1, The Cb and Cr values adjacent to each other are grouped with a frequency greater than a second predetermined threshold value in the histogram of the Cb and Cr values of the input image frame not determined as the monochrome image by the monochrome image determining unit, and the Cb values And a removal target pixel determination unit determining pixels belonging to a group whose Cr range is narrower than a preset reference range as pixels not used for automatic white balance adjustment. The automatic white balance gain calculator may calculate the automatic white balance gain using the remaining pixels except for the pixels determined as pixels not used for the automatic white balance adjustment from the input image frame. Considered automatic white balance adjustment device. The method of claim 4, wherein The removal target pixel determination unit may further include a removal amount determination unit comparing the number of pixels determined not to be used for automatic white balance adjustment with a preset third threshold value. If the number of pixels determined not to be used for the automatic white balance adjustment is greater than the third threshold value, the automatic white balance gain calculation unit does not calculate the automatic white balance gain for the corresponding input image frame, but the previous input image frame. And an automatic white balance gain calculated for the automatic white balance gain. Determining a pixel sampled in the input image frame according to the size of an input image frame; Creating a histogram of Cb and Cr values for the pixels determined to be sampled in the step of determining the sampled pixels; Comparing the maximum value of the histogram of the Cb value or the maximum value of the histogram of the Cr value with a first predetermined threshold value, respectively; And Determine an input image frame having at least one of the maximum values greater than the first threshold value as a monochrome image, and use the automatic white balance gain calculated for the input image frame except for the input image frame determined as the monochrome image. Steps to Perform Automatic White Balance Adjustment Automatic white balance adjustment method considering the influence of the monochrome image including a. The method of claim 6, The step of performing the automatic white balance adjustment, Performing automatic white balance adjustment using the automatic white balance gain calculated on the previous input image frame without calculating the automatic white balance gain on the input image frame determined as the monochrome image. Automatic white balance adjustment method considering the influence of monochrome image. The method of claim 6, wherein the creating the histogram comprises: A method of adjusting the automatic white balance in consideration of the influence of a monochrome image, characterized in that the step of creating a histogram of one of the Cb value and Cr value for one input image frame. The method of claim 6, When the maximum values are not all greater than the first threshold value, the adjacent Cb and Cr values are grouped with frequencies greater than a second predetermined threshold value in the histograms of the Cb and Cr values of the input image frame, respectively. Comparing the ranges of the Cb and Cr values of the groups with a preset reference range; Determining pixels belonging to a group in which the ranges of the Cb and Cr values are narrower than a preset reference range as pixels not used for automatic white balance adjustment; And And calculating an automatic white balance gain using the remaining pixels other than the pixels determined as pixels not used for the automatic white balance adjustment from the input image frame. How to adjust the balance. 10. The method of claim 9, Comparing the number of pixels determined not to be used for the automatic white balance adjustment with a preset third threshold value; And If the number of pixels determined not to be used for the automatic white balance adjustment is greater than the third threshold, the automatic white balance calculated for the previous input image frame without calculating the automatic white balance gain for the corresponding input image frame. A method for automatically adjusting the white balance in consideration of the influence of the monochrome image, further comprising the step of performing automatic white balance using the gain.

Images