JPH01314389A - Image processor - Google Patents

Abstract

PURPOSE:To obtain an illustration processing image of a natural image by decreasing the number of gradations of hue, lightness and saturation data which have been brought to color conversion, bringing them to reverse conversion to R, G and B data, and putting a black line into a turning point of colors of the hue, lightness and saturation data whose number of gradations has been decreased or the R, G and B data which have been brought to reverse conversion. CONSTITUTION:A color converting means 5 brings R, G and B data of a natural image to color conversion to hue, lightness and saturation data, and a gradation number decreasing means 8 decreases the number of gradations of the hue, lightness and saturation data which have been brought to color conversion. A reverse converting means 6 brings the hue, lightness and saturation data whose number of gradation has been decreased to reverse conversion to the R, G and B data. A black picture drawing means 9 puts a black line into a turning point of colors of the hue, lightness and saturation data whose number of gradations has been decreased or the R, G and B data which have been brought to reverse conversion. In such a way, from a natural image, anybody can form easily an illustration image by a simple operation.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing apparatus, and particularly to an image processing apparatus that forms an illustrated processed image from a natural image.

[Prior Art] Traditionally, illustrations of natural images have been done manually. For example, a photographic image is projected onto paper, and the positions where its outline and gradation change are traced using a line diagram. And Roma-
An illustration image with a different feel from the original was created by filling in parts of the same color and gradation with paint or pasting colored paper called Pantone. However, ■ Representing an almost infinite number of colors with a limited number of colors requires considerable sense and experience.

■ Tracing, coloring, and pasting on colored paper are detailed manual tasks that require a high level of skill and long hours.

■ Once an illustration image is created, it is difficult to modify it later.
Therefore, the work requires extreme caution.

It has the following disadvantages.

On the other hand, as computer image conversion technology, electronic plate making technology in the printing industry and professional laboratory technology that applies computer photo image processing in the photo lab industry are known.
For example, a document image is read with a high-precision scanner such as a drum scanner, and density expression correction (
γ correction), gradation settings, color correction, cropping and compositing, etc., and the resulting signals operate an image output device such as a laser printer or film recorder to obtain a reproduced image.

This type of device can do the following:

■ Restoration of fading color film ■ Adjusting the gradation of highlights and shadows to exaggerate color expression ■ Relief from equipment failures, projection errors, development errors, etc. ■ Removal of unnecessary objects (wires, dust, scratches, etc.) in the screen Erasure, correction ■ Express your creative image, expand the image area, and create new designs.

Such special effects can be obtained by a regularly arranged mosaic process, a solarization process that makes one curve unrealistic, a bossarization process that makes one curve discontinuous, or the like.

By the way, this bossarization processing seems to be relatively close to illustration processing.

A characteristic of illustrations is that the number of colors is limited, and in that sense, the bossarization process that reduces the number of gradations satisfies this condition.

When handling color images on a computer, one pixel is represented by RlG, B data of the three primary colors red (R), green (G), and blue (B), and in bossarization processing, R, G
, B data is reduced from, for example, 256 gradations to 5 gradations. As a result, the number of colors, which is more than 16 million colors in the original picture, is reduced to 125 (5') colors.

Figure 11 (A) shows R, G data (
FIG. Conventionally, this R and G data was subjected to bossarization processing using the input/output relationship as shown in FIG. 11(B). FIG. 11(C) shows output R9G data as a result of processing. If you look at this, the left side~
The color of a white body is yellow at point A, orange between points A and B, reddish orange between points B and C, orange between points C and D, and approaching red again after point D. Therefore, conventional bossarization processing is not suitable for illustration processing.

[Problems to be Solved by the Invention] The present invention eliminates the above-mentioned drawbacks of the prior art, and its purpose is to obtain illustration processed images of natural images without requiring special skills. An object of the present invention is to provide an image processing device that can perform

[Means for Solving the Problems] In order to achieve the above object, the image processing device of the present invention has the following features:
a color conversion means for color converting R, G, and B data of a natural image into hue, brightness, and saturation data; and a gradation number reduction means for reducing the number of gradations of the color-converted hue, brightness, and chroma data;
The hue, brightness and saturation data with the number of gradations reduced are R
, G, B data, and hue, brightness, and saturation data with the number of gradations reduced or Ro after the inverse conversion.
The outline of the system is to include a black drawing means for drawing a black line at the transition point between the colors of G and B data.

[Operation] In this configuration, the color conversion means converts R, G,
The 0 gradation number reduction means for color converting the B data into hue, brightness, and saturation data reduces the number of gradations of the color-converted hue, brightness, and saturation data.

The inverse conversion means inversely converts the hue, brightness, and saturation data with the reduced number of gradations into R, G, and B data.

The black drawing means inserts a black line at the color transition of the hue, brightness, and saturation data with the number of gradations reduced or the R, G, and B data after inverse conversion.

[Description of Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram of an image processing apparatus according to an embodiment. In FIG. 0, 1 is a control processor unit (
2 is a CPU memory, which stores the control program, for example, shown in FIG. 2, executed by the CPU 1, and various other parameters for control. 3 is a command I10, and an image processing command is inputted from a keyboard (not shown) or the like. 4 is a noise removal device, which removes noise data from image data. 5 is a color conversion arithmetic unit, which converts the R, G, and B data of the original image into hue H1 according to the HSL bihexagonal pyramid color model, for example.
Convert to H, L, and S data with brightness L1 and saturation S. 6 is a color conversion calculation device, which is the same <) (H, L according to the SL bihexagonal force color model).

Convert S data back to R, G, B data. Reference numeral 7 denotes a gradation conversion device, which in the embodiment converts the gradation characteristics of the chroma S data so as to increase the chroma (make the color tone more vivid). 8
is a gradation reduction device, which in this embodiment reduces each H, L, and S data to an appropriate number of gradations. Reference numeral 9 denotes a black point extraction device, which detects whether each pixel of interest has a different color from its surrounding pixels, and enables black line drawing. 10-12
is an image memory, R, G, B or HIL, S
Stores image data at each stage.

For example, each has 8 bits (256 gradations), and memory 10 to
If you add up 12 bits, you get 24 bits per pixel. A video controller 13 reads out the contents of the image memories 10 to 12 and outputs them to the outside as a video signal. A video monitor, video printer, etc. (not shown) can be connected to the outside.

Reference numeral 14 denotes image data I10, and the image data is input from an image inserting device (not shown) to the image memories 10 to 12 or outputted to an image output device (not shown).

FIG. 2 is a flowchart of the illustration processing procedure of the embodiment. When an image processing command is given via command l103, it is input to this processing.

In step Sl, the noise removal device 4 performs noise removal processing on the R, G, and B data of the original image in the image memories 10 to 12.The noise removal is performed using a convolution filter as shown in FIG. (
3×3) is determined and used as the pixel data of the pixel of interest (for example, the center pixel).

In step S2, one pixel's worth of R, G, and B data (8 bits each) is read out from the image memory 10-12.

In step S3, the color conversion calculation device 5 converts the RlG, B data into HlL, S data of hue H11 degrees L and saturation S.

FIG. 3 is a flowchart of the R, G, B-H, L, S color conversion processing procedure of the embodiment.
1, the maximum value among the R, G, and B data is set as MAX. In step S32, the smallest value among the R, G, and B data is set as MIN. In step S33, the brightness data L is
MAX+M I N L = (1).

In step S34, it is determined whether MAX=MIN.

If MAX=MIN, the hue H does not exist (it is an achromatic color), so the process advances to step S35, and the saturation S is set to 0. Further, although the hue H is undefined, it is defined as H=O. If MAX=MIN is not determined in step S34, it is determined in step S36 whether MAX=R. If MAX=R, it contains a large amount of red component R, so in step S37 the hue H is set as
56 gradations), constant 85 is 256/3, constant 4
3 means 256/6, meaning that hue data containing a large amount of red component R is encoded within a range of ±43 with 85 as the center. If this is considered as 360 degrees of a general hue circle, it means that hue data containing a large amount of red component is encoded around 120 degrees.

If MAX=R is not determined in step S36, it is determined in step S38 whether MAX=G.

If MAX=G, the hue H is set in step S3A.The constant 170 means that the hue data containing a large amount of green component G is 170.
This means that the angle is encoded around 240 degrees. Also, if MAX=G is not determined in step 338, the hue is determined to be H in step S39. means to be made into In step S3B, the data obtained in step S33 is compared with 127, and if L≦127, the saturation S is set as follows in step S3C. If L≦127, then in step S3D, the saturation S is set to 511-MAX-M I N .

Returning to step S4 in FIG. 2, of the obtained H9L and S data, the S data is sent to the gradation conversion device 7, where its gradation characteristics are converted. Here, the S data has vivid colors (
The color saturation S is high). Assuming that the S data is 8 bits, the conversion formula is, for example, as follows.

In step S5, the obtained H, L, and S data are sent to the gradation reduction device 8 to perform gradation reduction processing.

FIG. 9 is a conceptual diagram explaining the gradation reduction process of the embodiment. In FIG.
For example, the hue H is reduced to 122 gradations, the brightness is reduced to 4 gradations, and the saturation S is reduced to 3 gradations. . The reason why the number of gradations of hue H is large is to cover a wide range of basic colors during illustration.If there are 12 levels of 0 hues, the three basic primary colors, their complementary colors, and their intermediate colors are included. Furthermore, if there are four levels of brightness, for example, colors such as dark orange, slightly darker orange, slightly brighter orange, and brighter orange can be expressed. Also, since there are three levels of saturation S, you can express colors such as dull orange, normal orange, and bright orange. In this case, the total number of colors is 144.

In step S6, the color conversion calculation device 6 outputs H and S.

Perform L→R, G, B conversion processing.

FIG. 4 is a flowchart of the H, S, LR, G, and B color conversion processing procedures of the embodiment. In the figure, step S6
1, it is determined whether L>127 or not, and if L>127, the parameter M2 is set as M2=L+SL*S (8) in step S63.
This corresponds to the inverse transformation of equation (6), and the parameter M2 corresponds to MAX determined by the inverse transformation.

Moreover, if L>127, the parameter M2 is set to M2=L (1+s) (9) in step S62, which corresponds to the inverse transformation of equation (5).

In step S64, the parameter M1 is set as M1=2*L−
M2 (10), which corresponds to the inverse transformation of equation (1), and the parameter M1 is the ME obtained by the inverse transformation.
Corresponds to N. In step S65, Ml (MIN),
M2(MAX), R using H(7) value F(Ml.

M2. H), and G is F (Ml, M2.H+170
) and find B as F (Ml, M2.H+85).

FIG. 5 shows the function F (X
, Y, Z).

In the figure, in step S66, the third parameter Z
The value of (hueH) is compared with 43, and if Z<43, in step S69, F (X, Y, Z) is set. This corresponds to M2 in equation (9). If z<43, Z is compared with 128 in step S67. Z<128
If so, proceed to step S6G and set F (X, Y, Z)=Y
(12), and if Z<128, Z is compared with 170 in step 868. If Z<170, proceed to step S6B and set F (X, Y, Z); if Z<170, proceed to step S6A, and set F (X, Y, Z)=X (14).

Returning to Figure 2, R and G are thus inversely transformed.

The B data is written into the image memories 10 to 12 by the CPU 1 in step S7.

In step S8, the above calculation operations are performed on all pixels. Eventually, when all pixels are completed, the process moves to black line addition processing. This is achieved by blackening the boundary pixels between colors in the image.

In step S9, the CPUI selects image memories 10 to 12.
The pixel of interest and its adjacent pixels are read out from the pixel of interest and sent to the black line extraction device 9.

FIG. 6 is a diagram showing the relationship between the pixel of interest X and adjacent pixels A-D.
), Do (X), and Da (X). The condition for determining the pixel of interest
(X) #Da (A) Da (X)≠D (1(B) DB (x)≠D, when one or more of (A) DB (X) ≠ DB (B) is satisfied The pixel of interest X is determined to be a pixel to be converted into a black dot.At this stage, the number of colors has decreased, so changes between adjacent pixels can be easily determined.

Returning to FIG. 2, in step Sll it is determined whether the pixel of interest is a pixel that should be a black point, and if it is a black point pixel, step S12
Then black data (0) is written to the R, G, and B values of the relevant pixel in the image memories 10 to 12. If it is not a black point pixel, skip step S12 and proceed to step S13, and continue to step 89 until the inspection of all pixels is completed. - Repeat the process of S13.

In the explanation of FIG. 6, the reason why only pixel A on the right and pixel B below were selected as the pixels to be compared with the pixel of interest X is that the order of image processing is as shown in FIG. This is because the black dot processing is performed in the direction of the arrow from the upper left of the image, so it is necessary to compare the black dot processing with the unprocessed portion. If the left pixel C
Comparing with the pixels above 1, for example, X, C,
Even if D is the same color, if pixel C becomes a black dot, pixel X and pixel C will have different colors, and pixel X will also become a black dot, resulting in the entire image being completely black. This is because

Further, in the above embodiment, the tone conversion device 7 and the tone reduction device 8 are provided independently, but if the conversion of the characteristics as shown in FIG. 10 is performed in the tone reduction device 8, for example, both conversions can be performed at once.

In addition, in the above embodiment, R, G, B data is converted to HlL, S data, but it can also be realized using a color coordinate system that expresses hue, brightness, and saturation, such as an H, V, and S coordinate system. .

Further, in the above embodiment, each block was processed by a dedicated device, but all the processing may be programmed by a general-purpose computer.

Further, in the above embodiment, black line insertion processing was performed after inverse conversion from H, L, S data to RlG, B data, but H, L
, black line insertion processing may be performed at the stage of S data.

[Effects of the Invention] As described above, according to the present invention, anyone can easily create an illustration image from a natural image by simple operations.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of the image processing device of the embodiment, Fig. 2 is a flowchart of the illustration processing procedure of the embodiment, and Fig. 3 is the R, G, B-H, L, S color conversion processing of the embodiment. Flow chart of the procedure, FIGS. 4 and 5 are H, S, and LR of the embodiment. A flowchart of the G and B color conversion processing procedure, Figure 6 is a diagram showing the relationship between the pixel of interest X and adjacent pixels A-D, Figure 7 is a diagram showing an example of a convolution filter, and Figure 8 is an image of the example. A diagram showing the processing direction, FIG. 9 is a conceptual diagram explaining the gradation reduction process of the embodiment, and FIG. 10 is a gradation conversion characteristic when the gradation characteristic conversion process and the gradation reduction process of the embodiment are performed at once. 11A to 11C are diagrams illustrating conventional bossarization processing. In the figure, 1... CPU, 2... CPU memory, 3...
・-Command I10.4...Noise removal device, 5.6
... Color conversion calculation device, 7... Gradation conversion device, 8...
・Gradation reduction device, 9...Sunspot extraction device, 10-12・
...Image memory, 13...Video controller,
14...Image data I10. Figure 4 Figure 6 Figure 7 Figure 8 Figure 9 Figure 1Q

Claims (1)

[Scope of Claims] An image processing device that forms an illustrated processed image from a natural image, comprising: a color conversion unit that converts R, G, and B data of the natural image into hue, brightness, and saturation data; and the color conversion unit. gradation number reducing means for reducing the number of gradations of the hue, brightness and saturation data; and R, G
, an inverse conversion means for inversely converting into B data, and a black drawing means for inserting black lines at color transitions of the hue, brightness, and saturation data with the number of gradations reduced or the R, G, and B data after inverse conversion. An image processing device comprising: