KR100738179B1 - Auto white balance cotrol device, and method for controling the same - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to semiconductor manufacturing technology, and more particularly, to a method for adjusting an automatic white balance for determining a monochrome image and adjusting an automatic white balance. To this end, the present invention, an image input unit for receiving an image signal from the image sensor, an interpolation unit for converting the image signal into a high resolution image signal, a monochrome image determination unit for determining whether the converted image signal is a monochrome image and the determination There is provided an automatic white balance adjusting device including an automatic white balance processing unit for compensating an image for color distortion of an image signal.

The method may further include receiving an image signal from an image sensor, converting the image signal into a high resolution image signal, determining whether the converted image signal is a monochrome image, and converting the determined image signal into color distortion. An automatic white balance adjustment method is provided that includes compensating for an image.

.

Auto White Balance, Compensator, Frame, CMOS Image Sensor, Block

Description

AUTO WHITE BALANCE COTROL DEVICE, AND METHOD FOR CONTROLING THE SAME}

1 is a flow chart showing an automatic white balance adjustment process according to the prior art.

Figure 2 is a flow chart showing the automatic white balance adjustment process of the CMOS image sensor according to the present invention.

Figure 3 is a flow chart showing a monochrome image discrimination method according to the present invention.

4 is a diagram illustrating dividing a frame of an image signal into blocks.

Explanation of symbols on the main parts of the drawings

101: image input unit 102: interpolation unit

103: monochrome image discrimination unit 104: automatic white balance processing unit

TECHNICAL FIELD The present invention relates to semiconductor manufacturing technology, and more particularly, to an automatic white balance adjusting device.

Automatic white balance is a method for compensating an image due to color distortion caused by a light source. The function of expressing the color of an object that is reflected differently according to the light source and input to an image sensor using the automatic white balance adjustment method Has

With the recent proliferation of cellular phones and PCS embedded with CMOS image sensors, Auto White Balance has become an essential signal processing process for CMOS image sensors.

1 is a flowchart illustrating an automatic white balance adjustment process according to the prior art.

Referring to FIG. 1, an image input unit 101, an interpolator 102, and an automatic white balance processor 103 are configured.

In more detail, first, the image input unit 101 performs a role of receiving an image signal from the CMOS image sensor, and the interpolation unit 102 determines the R of the pixel from unit color values of neighboring pixels. By inferring and outputting values of G, B, the low resolution image data of the image sensor is converted into high resolution image data.

The automatic white balance processor 103 adjusts R, G, and B values by comparing the value output from the interpolator 102 with a value set by the user.

The automatic white balance adjustment method of the CMOS image sensor according to the related art uses the assumption that the average of one frame is achromatic, and when the average value of one frame is out of the white area, the white balance is adjusted by adjusting the gain of R and B. Done.

By the way, the automatic white balance adjustment method according to the prior art severely distorts the color when applied to a single color image rather than white. That is, there is a problem in that it is out of the area of the original color by performing the automatic white balance of the prior art to adjust the gain of R and B to white for monochromatic images other than white.

The present invention has been proposed to solve the above-mentioned problems of the prior art, and an object thereof is to provide an automatic white balance adjustment method for determining whether a monochrome image is used to adjust automatic white balance.

According to an aspect of the present invention for achieving the above object, an image input unit for receiving an image signal in the image sensor, an interpolation unit for converting the image signal into a high resolution image signal, whether the converted image signal is a monochrome image There is provided an automatic white balance adjusting apparatus including a monochromatic image discriminating unit for discriminating and an automatic white balance processing unit for compensating an image for color distortion of the determined image signal.

The method may further include receiving an image signal from an image sensor, converting the image signal into a high resolution image signal, determining whether the converted image signal is a monochrome image, and converting the determined image signal into color distortion. An automatic white balance adjustment method is provided that includes compensating for an image.

Hereinafter, the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the technical idea of the present invention. .

2 is a flowchart illustrating an automatic white balance adjustment process of the CMOS image sensor according to the present invention.

Referring to FIG. 2, an image input unit 201, an interpolation unit 202, a monochrome image determination unit 203, and an automatic white balance processing unit 204 are configured.

In more detail, first, the image input unit 201 serves to receive an image signal from the CMOS image sensor, and the interpolation unit 202 receives the R of the pixel from unit color values of neighboring pixels. By inferring and outputting values of G, B, the low resolution image data of the image sensor is converted into high resolution image data.

Subsequently, the monochrome image determining unit 203 determines whether the received image signal is a monochrome image instead of white, and the automatic white balance processing unit 204 determines the image determined by the monochrome image determining unit 203. The auto white balance process is performed when the signal is a white or monochrome image.

3 is a flowchart illustrating a method of determining a monochrome image according to the present invention.

Referring to FIG. 3, first, a frame of an image signal passed through the interpolation unit (see FIG. 2) is divided into blocks (see FIG. 4), and each of R, G, and B values of each block of the frame is divided. Find the representative value.

Subsequently, a subtraction value (Diff_R, Diff_G, Diff_B) of one block horizontally neighboring the current block is obtained with respect to the representative value by a method of determining a single color in the horizontal direction. In this case, when the deduction value of R is larger than the limit value (Diff_Limit) set by the user, the deduction value of G is larger than the limit value, or the deduction value of B is larger than the limit value, 1 Is assigned to the current block. In the same way, 0 is assigned to the current block if the subtraction of R is less than the threshold, or the subtraction of G is less than the threshold, or the subtraction of B is less than the threshold.

The above description in a programming language is equivalent to program 1 (301).

(Program 1)

If (Diff_R> Diff_Limit) or (Diff_G> Diff_Limit) or (Diff_B> Diff_Limit) return = 1;

Else return = 0;

Subsequently, the value assigned to the current block is determined. When the allocation value is 0, the horizontal zero count (horizontal_zero_count) is increased by 1, and the representative values of R, G, and B of the current block are accumulated. When the allocation value is 1, the accumulated representative values of R, G, and B are divided by horizontal zero counts, and the average is taken. The value of the horizontal zero count increased so far is compared with the value of the previous horizontal zero count stored. To store large values. If the previous horizontal zero count is stored again, the average value of the R, G, and B representative values is the same as the previous value, and when the new horizontal zero count is saved, the average value of the currently calculated R, G, and B representative values is used. Save the new file.

Finally, the horizontal zero count, the R average (AvgR), the G average (AvgG), and the B average (AvgB) are all initialized to zero.

Expressed in programming language, the description is the same as program 2 (302).

(Program 2)

If (return == 0) {

Horizontal_zero_count ++;

R ++; G ++; B ++; }

Else (return == 0) {

[Saved zero_count, Index] = Max [Horizontal_zero_count, Saved zero_count];

if (Index == 1) {

AvgR = R / horizontal_zero_count;

AvgG = G / horizontal_zero_count;

AvgB = B / horizontal_zero_count;

}

Subsequently, the above sequence is repeated to the last column, in which each row performs the sequence independently. In the last column, if the last stored horizontal zero count value is larger than the one color row reference value (horizon_Tr1), the row is determined as a single color row (horizon_uni = 1). If smaller, it is determined that it is not a monochrome row (horizon_uni = 0).

The above description in programming language is equivalent to program 3 (303).

(Program 3)

If (max. Horizontal_zero_count> horizon_Tr1) horizon_uni = 1;

Else horizon_uni = 0;

Subsequently, when the sum (sum.Horizon_uni) of the sum of the number of rows determined to be a monochromatic row is larger than the horizontal reference value (horizon_Tr2), the image is determined to be a monochrome image in the horizontal direction, and when smaller than the horizontal reference value (horizon_Tr2). It is determined that the image is not a single color in the horizontal direction.

 Expressed in programming language, the description is the same as program 4 (304).

(Program 4)

If (sum.Horizon_uni> horizon_Tr2) horizontally monochromatic image;

Else horizontal non-monochrome image;

Subsequently, the monochromatic discrimination method in the vertical direction is as follows.

The last stored R average (AvgR), G average (AvgG), and B average (AvgB) of each row obtained due to the horizontal monochrome determination are used.

A frame divided into several blocks is obtained by subtracting the R average or the G average or the B average in the vertical direction (Diff_AvgR, Diff_AvgG, Diff_AvgB), and the subtraction value of the R average or the G average or the B average is set by the user. If (Diff_Limit) is large, 1 is assigned. If the subtraction values of the R, G, and B averages are all smaller than the threshold, 0 is assigned.

Expressed in a programming language, the description is the same as program 5 (305).

(Program 5)

If (Diff_AvgR> Diff_Limit) or (Diff_AvgG> Diff_Limit) or (Diff_AvgB> Diff_Limit) return = 1;

Else return = 0;

Subsequently, when the assigned value is 0, the vertical zero count (vertical_zero_count) is increased by 1, and the R average, the G average, and the B average are accumulated.

When the assigned value is 1, the accumulated R average, G average, and B average are divided by the vertical zero count to take an average, and the value of the vertical zero count increased so far is compared with the previous vertical zero count value stored. To store large values.

In this case, when the previous vertical zero count is stored again, the R average, the G average, and the B average use the old value as it is, and when the new vertical zero count is stored, the currently calculated R average, the G average, and the B average Save the new file. Finally, we initialize all of the vertical zero counts, the R mean, the G mean, and the B mean to zero.

Expressed in programming language, the description is the same as program 6 (306).

(Program 6)

If (return == 0) {

Vertical_zero_count ++;

R ++; G ++; B ++; }

Else (return == 0) {

[Saved zero_count, Index] = Max [Vertical_zero_count, Saved zero_count];

if (Index == 1) {

AvgR = AvgR / Vertical_zero_count;

AvgG = AvgG / Vertical_zero_count;

AvgB = AvgB / Vertical_zero_count;

}

In the vertical direction, this operation is performed only once. In the vertical direction, the value of the finally stored vertical zero count is compared with the reference value (vertical_Tr) set by the user. When the vertical zero count is larger than the reference value, it is determined to be a solid color in the vertical direction. It is determined that it is not a single color.

 Expressed in a programming language, the description is the same as program 7 (307).

(Program 7)

If (max. Vertical_zero_count> vertical_Tr) monochrome image in the vertical direction;

Else non-monochrome image in vertical direction;

Subsequently, if it is determined that the color is monochromatic in the horizontal or vertical direction, the automatic white balance is applied with the monochromatic correction gain. If it is determined that the color is not monochromatic in the horizontal and vertical directions, the existing automatic white balance operation is performed. .

As described above, the present invention divides one frame of an image signal input from the CMOS image sensor into several blocks to obtain a difference from the reference block in the vertical or horizontal direction, and continuously adds a value smaller than the reference value. If the value occupies a large portion of the entire image, it is determined as a single color.

At this time, when it is determined that a monochrome image other than white is input, the monochrome automatic white balance method is applied so that the monochrome image does not lose its original color.

The present invention described above is not limited to the above-described embodiments and the accompanying drawings, and various substitutions, modifications, and changes can be made in the art without departing from the technical spirit of the present invention. It will be clear to those of ordinary knowledge.

As described above, the present invention considers the image of a single color rather than white, and in the case of a single color, when the image of a single color other than white is input by performing a single color automatic white balance, the original color does not lose and distortion does not occur. Have the effect of preventing.

Claims (6)

An image input unit which receives an image signal from an image sensor; An interpolator for converting the video signal into a high resolution video signal; A monochromatic image discriminating unit for discriminating whether the converted image signal is a monochromatic image and discriminating in the horizontal direction and the vertical direction, and determining the horizontal direction before the vertical direction; And Automatic white balance processing unit for compensating for the image distortion of the determined image signal color Automatic white balance adjustment device having a. Receiving an image signal at an image sensor; Converting the video signal into a high resolution video signal; Determining whether the converted video signal is a monochrome image; And Compensating for the distortion of the color of the determined image signal; The determining of whether the converted image signal is a monochrome image is performed in the horizontal direction and the vertical direction, wherein the horizontal direction is determined before the vertical direction. delete The method of claim 2, The step of determining whether the image is a monochrome image in the horizontal direction Dividing a frame of the video signal into blocks; Obtaining representative values of R, G, and B of the various blocks; Obtaining a subtraction value of the R, G, and B representative values from a reference block to a right neighboring block; Forming an allocation value by allocating 1 to a corresponding block when the deduction value is large and 0 when the deduction value is large compared to a preset threshold value of the R, G, and B subtraction values; If the allocation value is 0, the horizontal zero count is increased by 1 and the R, G, and B representative values of the corresponding block are accumulated. If the allocation value is 1, the R, G, and B representative values are divided by the horizontal zero counts and the average value is accumulated. Forming a value and storing a large value among the horizontal zero counts; Initializing the horizontal zero count, the R, G, and B mean values; After operation up to the last column of the block, comparing the finally stored horizontal zero count with a preset reference value and determining that the row of the block is monochromatic if the horizontal zero count is large, and not monochromatic if it is small. The automatic white balance adjustment method characterized by the above-mentioned. The method of claim 2, The step of determining whether the image is a monochrome image in the vertical direction Obtaining a subtraction value from neighboring blocks in the vertical direction from the reference block with respect to the R, G, and B average values determined by the vertical direction determination; Forming an allocation value by allocating 1 to a corresponding block when the deduction value is large and 0 when the deduction value is large compared to a preset threshold value of the R, G, and B subtraction values; If the allocation value is 0, the vertical zero count is increased by 1 and the R, G, and B average values of the corresponding blocks are accumulated. If the allocation value is 1, the average value is divided by the vertical zero count to form an average value. And storing a large value of the vertical zero counts; Initializing the vertical zero count and the R, G, and B mean values; After operation up to the first row of the block, comparing the stored vertical zero count with a preset reference value and determining that the row of the block is monochromatic if the vertical zero count is large and not monochromatic if it is small. How to adjust the white balance automatically. The method of claim 2, The monochromatic discrimination in the vertical direction and the horizontal direction performs automatic white balance with monochromatic correction gain when it is determined that the horizontal or vertical direction is monochromatic, and performs normal automatic white balance when it is determined that the horizontal and vertical directions are not monochromatic. Auto white balance adjustment method, characterized in that performed.

Images