JPS62281069A - Color image processing device - Google Patents

Abstract

PURPOSE:To correct color balance with a constant algorithm by making the color difference quantity of the picture element of a prescribed luminance into the dislocation of the color balance, and executing the color correction in accordance with the luminance from the dislocation. CONSTITUTION:Since it can be considered that color differences R-Y and B-Y of the picture element, in which luminance Y is maximum, means the 'dislocation' of a color balance as it does, when the color difference quantity is considered to be a correcting quantity DELTAE(DELTARY and DELTABY) and the color correction is executed, a correcting image can be obtained. At the time of calculating the correcting quantity, the correcting quantity to the color difference of an optional picture element in the image is made into the quantity to distribute DELTAE in proportion by a luminance/maximum luminance (Y/Ymax). Namely, a CPU 8 reads the contents of an image memory 2, detects the correcting quantity DELTAE, and inputs respectively the correcting quantities DELTARY and DELTABY to correction ROM 3 and 4. Thus, since the table conversion by three inputs of the correcting quantity DELTAE, the luminance and the color difference quantity is used, a high speed and effective color balance can be corrected by a simple constitution.

Description

Detailed Description of the Invention 3. Detailed Description of the Invention [Field of Industrial Application] The present invention relates to a color image processing device that corrects a shift in color balance with respect to an input image.

[Prior Art] Conventionally, two methods have been used to adjust the color balance of an image: (1) Adjusting the color balance before taking a picture; (2) Correcting the image after it has been taken.

(2) corresponds to the "white balance switch" of a video camera, which takes a picture of a white piece of paper, etc. before starting shooting, and uses that "white" as a reference to set the white balance. ■ is widely practiced in the field of printing, etc., but much of it depends on the craftsman's intuition and experience. As described above, the problem is that a fixed algorithm for adjusting color balance has not yet been developed.

[Problems to be Solved by the Invention] The present invention has been made in view of the above-mentioned problems, and its purpose is to propose a color image processing device capable of correcting color balance using a certain algorithm. It's about doing.

[Means for Solving the Problems] The configuration of the color image processing device of the present invention for realizing the above-mentioned problems includes a detection means for detecting pixels of a predetermined luminance from a color image signal consisting of luminance and color difference; It consists of a correction amount calculation means for calculating a correction amount, and a correction means for correcting the color difference of each constituent pixel of an image.

[Operation] In the color image processing device configured as described above, the detection means detects pixels of a predetermined luminance from a color image signal consisting of luminance and color difference, and the correction amount calculation means detects a pixel of an image from the color difference of the pixels of the predetermined luminance. Calculate the amount of correction for the color difference of each constituent pixel,
The correction means corrects the color difference of each constituent pixel of the image by the correction amount.

[Examples] Examples according to the present invention will be described in detail below with reference to the accompanying drawings.

<Processing Concept> The color balance correction in this embodiment is performed by focusing on highlight points (pixels with the highest brightness) in the image. That is, in an image with correct color balance, there is a high probability that the color difference of the highlight point is 0'', that is, rwhite. Therefore, the outline of the color balance correction of the color image processing device of this embodiment is as follows: (1) The color difference amount of the highlight point pixel in the obtained image is considered to be a "shift" in the color balance, so the color difference amount is Let ΔE be the value to correct the balance.

■: For all pixels, from the amount of color difference of each pixel, ΔE × brightness /
The color balance is adjusted by subtracting the maximum brightness and using the difference as a corrected color image signal.

FIG. 2 shows a color solid of an image with a shifted color balance, and FIG. 2(b) shows a color solid of an image after color balance correction. The color difference (R-Y, B-Y) of the pixel with the maximum brightness (Y)
) is considered to directly mean a "shift" in color balance, so these color differences can be expressed as correction amounts ΔE (ΔRv, ΔB
If color correction is performed assuming that the color is y), a corrected image as shown in FIG. 2(b) is obtained. When calculating this amount of correction, the amount of correction for the color difference of any pixel in the image is the amount obtained by proportionally distributing ΔE by brightness/maximum brightness (Y/Y, , ,) as shown in Figure 2 (a). I will make it happen. In addition, in the following examples, ΔE
There are two examples of methods for detecting (ΔRY+ΔBY).

<Overview of the apparatus of the embodiment> FIG. 1 is a block diagram of a color image processing apparatus of the embodiment. The outline of the image processing apparatus shown in FIG. 1 is as follows.

The image memory 2 contains an A/D converted color image signal Y1.
Color differences (R-Y, B-Y) are stored. The CPU 8 reads the contents of the image memory 2 and displays the contents in FIG. 3 or 4 (a).
), (b) to detect the correction amount ΔE. This correction amount (ΔRY, ΔBY) is the correction R
Manufactured by OM3.4. Other manual inputs of the correction ROM 3.4 are brightness (Y) and R-Y and B-Y, respectively. The output of the correction ROM 3.4 and the brightness (Y) are input to the masking circuit 5, converted into C, M, Y, D/A converted, and printed by the head driver 7. Note that FIG. 1 shows color balance correction using a correction ROM, and a detailed explanation will be given later, but there are also other methods such as color balance correction for each constituent pixel of an image, which is performed by the CPU 8 using software. Examples will also be explained later. The RAM 9 stores a program for procedures such as those shown in FIG.

<Determination of ΔE> The method for determining ΔE shown in Figure 3 is based on the maximum brightness (Y□8
) is defined as ΔE. First, in step S1, a highlight point pixel is found by comparing the brightness Y of each pixel in the image memory,
In step S2, the color difference amounts R-Y and B-Y of the pixel are considered as "shifts" in color balance, and the correction amounts ΔE(ΔRY, ΔB
Y).

FIG. 4(a) shows the concept of another method for determining ΔE, and FIG. 4(b) shows the procedure. This is because in the above-mentioned method of calculating ΔE, the correction amount may become unstable because it is calculated from the single point with the highest brightness.
Therefore, the correction amount is obtained by collecting about 10 samples of the pixels with the highest brightness and averaging the color difference amounts of those pixels, thereby performing more stable and effective color balance correction.

FIG. 4(h) is a flowchart of the process for determining the correction amounts ΔRY and ΔBY. First, 10 highlight points are searched from within the image, but searching from every pixel takes a very long time. Therefore, Fig. 4 (a
), when the input image is an image of 640 x 480 pixels, the inner (121 to 520) x (86 to 2
95), sample every 4 pixels vertically/horizontally and search for 10 highlight points. Even with this thinning sampling, it is possible to obtain almost the same results as when searching from all pixels.

Color difference 1R-Y, B-Y of 10 highlights is (R-y
)+, (FL-y)2. −(RY)lOb and (
B Y) I, (B Y)2.・= (B −Y)
At the end of +o, ΔE (ΔRy, ΔBY) is ΔRy = ((RY)++ (RY)2+...+(
R-y),. )/10 ΔBY = ((B Y)++ (B Y)2+...
・+(B-Y)+o)/10. This makes it possible to obtain a more effective correction amount than determining the correction amount from a single highlight point.

<Color Balance Correction> Now that the maximum correction amount has been obtained in this way, an example will be described in which the CPU 8 performs the color balance correction using software. This color balance correction is a variable correction proportional to the brightness of each constituent pixel of the image, with the correction amount ΔRY+ΔBY being the maximum value. This is because it was noted that if ΔRY and ΔB7 were simply subtracted from RY and BY of each constituent pixel of the image, the lower the brightness of the pixel, the more the result would be overcorrection. Therefore, Figure 2 (
As explained in a), for all pixels, R-Y' ← (RY) -ΔR7XY/YI1
．． 1lXB-Y' -(B-Y) -ΔB y
By performing the correction as shown in XY/Ya+ax, a smaller amount of correction is subtracted from a pixel with lower luminance.

Here, Y.

RY, B-Y are image signals of each constituent pixel of the image, Yl,
8 indicates maximum brightness. The procedure is shown in FIG. By the above method, the image whose color balance is shifted as shown in FIG. 2(a) in the image memory is corrected as shown in FIG. 2(b).

As explained above, by subtracting the correction amount obtained from the highlight point multiplied by the brightness/maximum brightness from the color difference of each constituent pixel of the image, you can quickly and effectively correct color balance with simple calculations. This is a color correction method that does this. still,
Although the above color balance correction amount is a correction amount that changes in accordance with luminance or in a near manner, it may be a non-linear correction amount by adding a weight according to luminance.

(Color Balance Correction Table) Next, the correction ROM 3.4 that constitutes a correction table for color balance correction will be described. This ROM is the fifth
The calculation performed in step s21 in the figure is performed by hardware instead of the CPU 8, thereby achieving high-speed processing.

As shown in the figure in the first part, the input of this ROM for R-Y is composed of three manual inputs: Y, R-Y, and correction amount ΔRY, and as a result, the corrected R-Y' is output. . B-Y
The same applies to The correction amounts ΔR7 and ΔBY are calculated and stored in advance based on the following equations.

RY' = (RY) -A Ry X y/ Y
mallB-Y'+ (B-Y)-ABYXY/Ym
ax In this way, by using table conversion, high-speed processing becomes possible. FIG. 1 is a block diagram of a color video printer using this table conversion. here,
The input of ROM3.4 shown in FIG. 1 is, for example, R-Y
Regarding, only RY, ΔR7 and Y are input, and Y□8 is not input.

This is done in order to reduce the number of input bits of the ROM.
This is because □8 was fixed. For example, if the brightness (Y) is 8 bits, fix Y as "255",
The value calculated by Y□,=255 according to the above equation is stored in the ROM. Even in this case, the correction accuracy does not decrease, and the size of the ROM is reduced accordingly.

In this way, by using the three-man manual table conversion of correction ff1(ΔE), luminance, and color difference, it is possible to realize a processing device that performs high-speed and effective color balance correction with a simple configuration.

FIG. 6 shows a modification of the correction circuit. This modification is a ROM
In order to reduce the scale of ΔRYxY/Y in the above formula,
l11. ROM only the part of X, ΔBYxY/Ymax
The subtraction part is performed by subtracters 22 and 23.

[Effects of the Invention] As explained above, according to the present invention, the amount of color difference of a pixel with a predetermined brightness is taken as a color balance deviation, and color correction is performed based on this deviation according to the brightness, thereby correcting the color balance. A color image processing device that can perform processing using a certain algorithm can be provided. Further, it becomes more effective if the predetermined brightness is set to the maximum brightness.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a color image processing apparatus according to an embodiment of the present invention, and FIGS. 2(a) and (b) are diagrams before and after correction for explaining the principle of color balance correction in the embodiment. 3 is a flowchart related to a control procedure for detecting a correction amount from a pixel with maximum brightness], FIG. 4(a) is a diagram explaining another embodiment for detecting a correction amount, 4(b) is a flowchart 1 related to the control procedure of the embodiment of FIG. 4(a), FIG. 5 is a flowchart of the control procedure when color balance correction is performed by software, and FIG. FIG. 7 is a circuit diagram of another embodiment of color balance correction. In the figure, 1... A/D converter, 2... Image memory, 3.4°20.21... Correction ROM, 5... Masking circuit, 6.10... D/A conversion 7-head driver, 8-CPU, 9-RAM, 20.21--subtractor.

Claims (4)

[Claims] (1) A detection means for detecting a pixel of a predetermined luminance from a color image signal consisting of luminance and color difference, and a correction amount calculation for calculating a correction amount for the color difference of each constituent pixel of the image from the color difference of the pixel of the predetermined luminance. and a correction means for correcting the color difference of each constituent pixel of the image by the correction amount. (2) The color image processing device according to claim 1, wherein the predetermined brightness is a maximum brightness. (3) The correction amount is obtained by proportionally distributing the color difference for the pixel with the maximum brightness according to the ratio of the brightness of each component pixel of the image to the maximum brightness. The color image processing device described in Section 1. (4) The color image processing apparatus according to claim 2, wherein the correction means subtracts the correction amount from the color difference of each constituent pixel of the image.