JPH01124071A - Color image processing system - Google Patents

Abstract

PURPOSE:To comparatively simply execute image emphasis prevented from changing a color difference by separating lightness information from a picture element value expressed by an RGB space consisting of a picture element for emphasizing an image and four picture elements adjacent to said picture element and executing correction processing. CONSTITUTION:The picture element value expressed by the RGB color space consisting of the picture element for emphasizing the image and four picture elements adjacent to said picture element is inputted to a 1st converter 1 for executing color space conversion in order to separate lightness information. Color shear is removed from the lightness information separated by the 1st converter 1 and correction for removing the difference of emphasis intensity between a bright part and a dark part in the color space is executed by a correction processing part 2. Image emphasis is executed by a 2nd converter 3 for receiving the outputs of the 1st converter 1 and correction processing part 2 and outputting RGB color space information. Consequently, color shear which may be generated at the time of emphasizing an image can be suppressed.

Description

[Detailed Description of the Invention] [Summary] The present invention relates to a color image processing method that enhances a color image, and an object of the present invention is to enable image enhancement that does not cause color differences when viewing a color image. A first converter that receives pixel values expressed in RGB color space of a pixel for image enhancement and four pixels in the vicinity of the pixel, and performs color space conversion to separate brightness information. a correction processing unit that performs correction processing on the brightness information output from the first converter; and a correction processing unit that receives the output of the first converter and the output of the correction processing unit and outputs RGB color space information. 2 converters, and is configured to enhance the image.

[Industrial application field]

The present invention relates to a color image processing method for enhancing color images.

A color image processing device performs image enhancement by multiplying pixel values by Laplacian in a color space such as RGB, YI[l, ^ma, etc.] in order to clarify the outline of an image.

There is a need for a color image processing method that can easily enhance color images with high accuracy.

[Conventional technology]

Conventionally, when applying a Laplacian to Lab space to emphasize the outline of an image, if the value obtained by enhancement goes outside the color space that can actually be expressed, the obtained value will be The color will be completely different from that of Furthermore, when the RGB color space is multiplied by the Laplacian, color shift occurs regardless of whether it is inside or outside the space.

On the other hand, when the YIQ space is multiplied by the Laplacian, no color shift occurs even outside the space. However, there is a drawback that the strength of emphasis is different when visually observing dark and bright areas on the Y axis. FIG. 4 fan, (bl shows the sensitivity characteristics of the eye.

[Problem that the invention seeks to solve]

As described above, when a color image is conventionally enhanced using Laplacia Pyzun, problems occur in each of the RGB, YI, and Lab spaces.

SUMMARY OF THE INVENTION In view of the above-mentioned conventional situation, it is an object of the present invention to provide a color image processing method that can relatively easily perform image enhancement without causing color differences when viewing a color image.

[Means for solving problems]

As shown in the principle diagram of FIG. 1, this invention receives pixel values expressed in RGB space of a pixel for image enhancement and four pixels near the pixel, and uses a color space to separate brightness information. a first converter 1 that performs conversion; a correction processing unit 2 that performs correction processing on brightness information output from the first converter;
A second converter 3 receives the output of the first converter and the output of the correction processing section 2 and outputs RGG color space information.

[For production]

A correction processing unit performs correction for eliminating color shift for the brightness information separated by the first converter 1 and also for eliminating differences in emphasis strength between bright and dark areas within the color space. 2, color shift and differences in the strength of emphasis between bright and dark areas within the color space are eliminated.

〔Example〕

FIG. 2 is a block diagram showing the first embodiment of the present invention, and FIG. 3 is a schematic diagram showing the relationship between the emphasized pixel and four neighboring pixels in this embodiment. As shown in FIG. 3, registers 4-5, 4-1 to 4-4, and registers 4-5 and 4-4 store the values of pixels u and d above and below the pixel C to be emphasized, and pixels z and r to the left and right of the pixel C to be emphasized, respectively. An RGB/Q converter (first converter) 1-1 that inputs pixel values of 1 to 4-5, and an RGB/Q converter 1
registers 4-6 to 4-10 that store the output value of -1;
A Laplacian calculator (correction processing unit) 2-1 that performs the Laplacian of one value (brightness information) (He, Lld, City, 4-Q) stored in registers 4-6 to 4-10, and a Laplacian calculator 2-1. +, 'rq/Rc input the output value of 1 and Ic, Qc of the emphasized pixel stored in the register 4-10.
B converter (second converter) 3-1.

The pixel to be emphasized and the four neighboring pixels are RGB/L'I
[] It is converted into one piece by the converter 1-1, and the Laplacian is applied to the converted piece by the Laplacian calculator 2-1. Therefore, one value of 5 pixels (I''!u, tl, I!'
((,L'r, 1.'c,) is subjected to Laplacian processing to eliminate color shift due to emphasis.

FIG. 5 is a block diagram showing a second embodiment of the present invention.

The pixels u above and below the pixel C to be emphasized shown in FIG.
Registers 5-5. 1 converter) 1-2, registers 5-6 to 5-10 that store the output values of RGB/YTQ converter 1-2, respectively, and Y values ( Yu, Yd, Y6
．． Yr.

A Laplacian calculator 2-2 that converts Yc) into a Laplacian, a ROM 2-3 that is searched by the Yc value stored in a register 5-10 and operates as a coefficient conversion unit, and a ROM 2-3 that operates as a coefficient conversion unit, and the output of the Laplacian calculator 2 and the memory of the register 5-10. I do
c, a YIQ/RGB converter (second converter) 3-2 that converts the Qc values into RGB, and a YIQ/RGB converter 3.
-2 output register 5-1).

The Laplacian calculator 2-2 and the Ill0M 2-3 constitute the correction processing section 2.

Note that the RGB/Yltl converter 1-2 is configured as shown in FIG. For example, the Ru, Go, and Ru values stored in the register 5-1 are stored in the arithmetic unit 12-bl2-2.
12-3 and are converted into X, Y, and Q values.

The computing units 12-1, 12-2, and 12-3 each compute the following equations.

Yu=0.299Ru+0.587Gu+0.
1) 4BuIu=0.596Ru -0,274Gu
-0,322BuQu=0.21)Ru -0,523
Gu + 0.312Bu other registers 5-2 to 5-5
The above-mentioned calculation is also performed on the value of .

The Laplacian calculator 2-2 is provided with multipliers 12-1) to 12-15 to which respective Y values inputted from registers 5-6 to 5-10 shown in FIG. Figure 7). Coefficients 1 to 2 are input to the other input terminals of the multipliers 12-1) to 12-15. Therefore, for example, the multiplier 12-1) calculates YuXkl. Each of the other multipliers 12-12 to 12-15 also performs similar multiplication.

The multiplication results are added by adders 12-16.

That is, Yu−kl +Yd−kl +Y1− kl +
1 is calculated for Yr- and 2 is calculated for +Yc-. However, k2
=-(4Kl +1) The Laplacian calculator 2-2 performs Laplacian processing on the addition result obtained as described above to obtain an emphasized Y value.

The coefficients used in this Laplacian calculation are stored in advance in the ROM 2-3 that constitutes the coefficient conversion section.

The YIQ/RGB converter 3-2 has the same configuration as that shown in FIG. 5, and each arithmetic unit performs the following calculation.

R=Y+〇, 956I+0.6210 G,,Y-0,2721~0.647[]]B=Y-1
．． 1061+1.703 The enhanced image data obtained by the IQ/RGB converter 3-2 is stored in the register 5-1).

〔Effect of the invention〕

As is clear from the above description, according to the present invention, non-linearity in bright and dark areas of color shift that occurs when an image is emphasized is eliminated, and the present invention is extremely effective when enhancing a color image.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the principle of the present invention. FIG. 2 is a block diagram showing the first embodiment of the present invention. FIG. 3 is a schematic diagram showing the relationship between the emphasized pixel and four neighboring pixels of this embodiment. Figure 4 (al, (bl is a sensitivity characteristic diagram of the eye, Figure 5 is a block diagram showing the second embodiment of the present invention, Figure 6 is a block diagram of the configuration of the RGB/YIQ converter of the present invention, 1' 0 Fig. 7 is a schematic diagram for explaining the Laplacian calculator of the present invention. In the figure, 1 is the first converter, 2 is the correction processing section, 3 is the second converter, and -1 is the RGB/l, 'IQ converter, 1
-2 is RGB/YIQ converter, 2-1 and 2-2 are Laplacian calculators, 2-3 is ROM, 3-1 is LIQ/l
? GB conversion shoulder F! '7...Kuchi, Schiff diagram Figure 6 Explanation υ 1000~Repeat diagram Figure 7

Claims (3)

[Claims] (1) A first converter (1 )
, a correction processing unit (2) that performs correction processing on the brightness information output from the first converter (1), and an output of the first converter (1) and the correction processing unit (2). A color image processing method characterized by comprising a second converter (3) which receives an output and outputs RGB color space information, and enhances an image. (2) The first converter (1) is composed of an RGB/L^*IQ converter, and the correction processing section (2) is configured to convert the RGB/L^*IQ converter.
It is composed of a Laplacian calculator that inputs each L^* value output from the L^*IQ converter, and
) is the L value output from the Laplacian calculator and the RG
This is an L^*IQ/RGB converter that inputs the I value and Q value of the emphasized pixel output from the B/L^*IQ converter, and
2. The color image processing method according to claim 1, wherein the image is enhanced by applying a Laplacian of pixel values L^*. (3) The first converter (1) is composed of an RGB/YIQ converter, and the correction processing section (2) is configured with the RGB/YIQ converter.
a Laplacian calculator to which the output of the Q converter is input; and a coefficient conversion section that converts the coefficients input to the Laplacian calculator according to the pixel value to be emphasized;
Based on the Y value output from the YIQ converter, the coefficients are converted to perform Laplacian calculation, and the second converter converts the output of the Laplacian calculator and the I value and Q value of the pixel value to be emphasized. 2. The color image processing method according to claim 1, further comprising: a YIQ/RGB converter that receives input signals; and a YIQ/RGB converter.