JP3788669B2 - Color image processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a color image processing apparatus that reproduces a color image with high image quality, and more particularly to a color image processing apparatus that reproduces black characters with high image quality and that does not cause image quality deterioration in other areas. Applicable.
[0002]
[Prior art]
As a conventional color image reproduction apparatus, there is an apparatus that reproduces an image with high image quality by performing processing different from that of other areas on characters, particularly black character areas (here, assuming that the reading resolution is approximately 400 dpi). I will explain below).
[0003]
For example, in the method described in Japanese Patent Publication No. 7-8007, a black character region is detected from an image (a recognition signal is a binary signal of black characters / non-black characters), and the UCR rate is switched according to the result. is there. As a result, black characters are reproduced almost in black, so that color blur due to misregistration that can occur when images are reproduced in four colors of CMYK does not occur, and as a result, high quality reproduction of black characters is possible. .
[0004]
[Problems to be solved by the invention]
However, in the above-described method, determination of black characters / non-black characters is performed by local image processing. Therefore, erroneous determination cannot be avoided at the current technical level. In particular, the tendency is strong for a low-line number halftone dot region of 65 lines to 85 lines or a halftone dot region enlarged and scaled. Moreover, since the above determination is a binary determination of whether the character is black or not, the UCR rate is switched dynamically in this image reproduction process, and it is erroneously determined that the character is a black character in the design. In this case, the image quality is significantly deteriorated.
[0005]
On the other hand, there has also been proposed an apparatus for controlling the ink at the edge portion by performing so-called multi-step determination in terms of edge amount (see, for example, Japanese Patent Publication No. 7-108019). In this apparatus, the under color removal amount is controlled in multiple stages according to the edge amount and the black component, and since the process is continuously switched, there is no significant image quality deterioration due to erroneous determination.
[0006]
However, in the above-described method for calculating the edge amount of the device, the Laplacian filter is used, so there is a response to the characters, but the edge amount reacts more than necessary to the entire halftone area, and the edge amount from the halftone area is increased. Will be detected. In particular, the edge amount reacts greatly even with respect to halftone dot highlights (halftone dot area ratio of 10% to 50%), and as a result, high UCR (high ink) processing is performed on the halftone dot highlight area. As a result, a rough image with poor graininess is reproduced.
[0007]
The present invention has been made in consideration of the above-mentioned background,
The object of the present invention is to control the black ink according to the conventional edge amount by calculating a feature amount that responds only to characters, or in other words, does not respond as much as possible to the halftone area. An object of the present invention is to provide a color image processing apparatus that reproduces an image without causing image quality problems.
[0008]
[Means for Solving the Problems]
In order to achieve the above object, according to the first aspect of the present invention, means for reading a document as a digital signal and calculating the fine line degree of the pixel of interest from the digital signal, and the higher the calculated fine line degree, A color image processing apparatus having means for increasing the black generation amount of pixels, wherein the means for calculating the fine line degree calculates the steepness degree of the character ridge .
[0009]
According to the second aspect of the present invention, means for reading an original as a digital signal and generating an achromatic signal from the digital signal, means for calculating the fine line degree of the pixel of interest from the achromatic color signal, and the calculated fine line degree Is a color image processing apparatus comprising means for increasing the black generation amount of the pixel of interest, the means for calculating the fine line degree being characterized by calculating the steepness degree of the character ridge .
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.
In the present invention, as means for replacing the edge amount, means for generating a fine line degree indicating the probability that the target pixel is a line portion is provided, and black ink control is performed based on the magnitude of the thin line degree.
[0011]
<Example 1>
FIG. 1 shows the configuration of Embodiment 1 of the present invention. The color scanner 1 outputs a multi-value digital signal (in this case, a substantially linear 400 dpi, 1-pixel 8-bit signal, where the white of the original is 255 and black is 0) obtained by reading an unillustrated original. .
[0012]
The Log converter 2 includes a table for converting the signal characteristics from the reflectance space to the density space, and outputs a signal representing the ink amount.
[0013]
The color correction circuit 3 removes the turbid component of the color separation film at the scanner and the turbid component of the ink. In general, color correction is performed by a masking method, a memory map method (interpolation method), or the like, but here, the former masking method will be used for explanation. The conversion is performed in the following line format, for example.
[0014]
c = kll * R + k12 * G + k13 * B + k14
m = k21 * R + k22 * G + k23 * B + k24
y = k31 * R + k32 * G + k33 * B + k34
Here, kll to k34 are constants and are determined based on experiments.
[0015]
The black generating circuit 4 calculates a K (black) signal from the cmy signal. This is to replace a part of the CMY recording material with a black recording material. Generally, a black character is desirable to have a high UCR rate with a large replacement amount. However, when a picture, particularly a highlight portion, is reproduced with a high UCR, it is rough. Becomes noticeable and is not preferable.
[0016]
Therefore, in the present embodiment, the black generation amount is controlled according to the following equation by using a thin line degree s1 (characteristic region, which is a character region that becomes larger as it gets thinner), which will be described later.
[0017]
K = (0.6 + 0.4 × s1 / 255) × min (c, m, y)
In the black generating circuit 4, the c, m, and y signals are through processing and are not particularly converted.
[0018]
The subtracting circuit 5 calculates C, M, and Y signals for driving the color printer according to the following equation. As a result of this processing, the color signal around the thin line character is reduced.
[0019]
C = c-K
M = m−K
Y = y-K
The color printer 6 receives the printer drive signals C, M, Y, and K and reproduces an image. Here, for simplicity of explanation, a color printer capable of outputting 256 gradations per dot is assumed. In the case of a color printer having a gradation level lower than this, gradation processing is performed by multi-value dither processing or multi-value error diffusion processing.
[0020]
Next, the thin line degree calculation means that is the recognition system of the present embodiment will be described.
As preprocessing, the min (R, G, B) calculation circuit 7 obtains a black signal. This is a process performed to increase the black rate only for black characters. However, in the case where all characters including color characters are reproduced with a high black rate, the following processing may be performed using, for example, a G signal instead of the black signal. A characteristic of reproducing color characters at a high black rate is that the color becomes somewhat cloudy due to an increase in black, but there is a tendency for the character to be cut better because the total amount of color material is reduced.
[0021]
FIG. 2 shows a configuration example of the fine line degree calculation circuit. Here, first, means for calculating the character width by pattern matching will be described. The pattern example described below is for detecting a portion where a character extends in the vertical direction, and it is sufficient to prepare a pattern for detecting a portion where the character extends in the horizontal direction or oblique direction based on the same idea. . The important points are to reject the dot area where dots are not continuous by introducing a black bar pattern as much as possible, and the portion that you want to reproduce at a high black rate is generally a thin black line that is weak against printer printing. Therefore, the final output of the thin line degree calculation circuit is that the thinner the line, the larger the response.
[0022]
The fine line degree calculation circuit includes an MTF correction filter 21, binarization circuits 22 and 23, a black bar pattern detection circuit 24, a line width calculation circuit 25, a logic circuit 26, and a table conversion circuit 27.
[0023]
First, as preprocessing for binarization, an MTF correction fill 21 for emphasizing high-frequency spatial frequency components is applied. FIG. 3 shows an example of the MTF correction fill. The binarization circuit 22 detects pixel values equal to or higher than the threshold thl as black pixels, and pixel values less than the threshold thl as non-black pixels. The binarization circuit 23 detects pixel values equal to or lower than the threshold th2 as white pixels and threshold th2 A pixel value exceeding 1 is detected as a non-white pixel.
[0024]
The black bar pattern detection circuit 24 detects a black bar pattern when the pixel of interest is a black pixel and pixels on a plurality of lines above and below it are black pixels. This detection result is 1 bit (whether or not it is a black bar pattern). FIG. 4 shows an example of a black bar pattern. In the line width calculation circuit 25, the target pixel is a non-white pixel, and the line width is calculated according to the number of left and right non-white pixels. A value from 0 to 12 (multi-value, 4 bits) is output as the calculation result. FIG. 5 shows a pattern example when calculating the line width. When the line width is the thinnest, that is, when the number of non-white pixels is the smallest, the value 12 is output as the line width, and the line width is the thickest. That is, if the pattern has the largest number of non-white pixels, the value 1 is output as the line width.
[0025]
The logic circuit 26 outputs the value of the line width calculation circuit 25 when the black bar pattern detection circuit 24 outputs “1”, and outputs the value 0 when the black bar pattern detection circuit 24 outputs “0”. To do. The table conversion circuit 27 performs primary conversion (normalization) on the values 0 to 12, which are the outputs from the logic circuit 26, from the values 0 to 255, as shown in the table example of FIG.
[0026]
As another example of the fine line degree calculation circuit, means for utilizing the steepness of the character ridge will be described. This utilizes the property that a thin character has a large level difference between a pixel corresponding to the ridge and a neighboring pixel, and conversely a thick character has a small level difference. Here too, the important point is that by introducing character ridge detection (ridge pixel plus its continuous attribute), the dot area where dots are not continuous is rejected as much as possible, and the part that you want to reproduce at a high black rate is In general, since the thin black line is weak against printing of a printer, the final output of the thin line degree calculation circuit is that the thinner the line, the larger the response.
[0027]
FIG. 7 shows another configuration example of the fine line degree calculation circuit, which includes a ridge pixel detection circuit 31, a steepness degree calculation circuit 32, a continuity detection circuit 33, a logic circuit 34, and a table conversion circuit 35.
[0028]
First, the ridge pixel detection circuit 31 detects a pixel corresponding to a ridge of a character using a mask as shown in FIG. 8, for example. That is, for example, when the pixel value of the pixel of interest is e, the pixel value diagonally upper left of the pixel of interest is a, and the pixel value diagonally lower right of the pixel of interest is i, e−a is equal to or greater than the predetermined threshold th3. When ei is equal to or greater than a predetermined threshold th3, the target pixel is detected as a ridge pixel.
[0029]
Furthermore, in the character portion there is a high possibility that the ridge pixels are continuous, the continuity detection circuit 33 detects the continuity with the pattern matching as shown in FIG. 10 for example. Then, the detection result from the continuity detection circuit 33 is output as to whether or not it is 1 bit, that is, continuity.
[0030]
On the other hand, the steepness degree calculation circuit 32 calculates the steepness degree by obtaining the contrast between the target pixel and its surrounding pixels. The calculation result is a multi-value output. FIG. 9 shows an example of the mask for calculating the urgency level. The logic circuit 34 outputs the degree of suddenness when continuity is detected, and the output value is first converted by the table conversion circuit 35. This table conversion is not particularly shown here, as long as the output value is appropriately normalized to 0 to 255.
[0031]
<Example 2>
Embodiment 2 of the present invention will be described below.
[0032]
FIG. 11 shows the configuration of the second embodiment of the present invention. The difference from the first embodiment is that a chromatic / achromatic determination circuit 9 is provided in parallel with the fine line degree calculation circuit, and further, a logic circuit 10 for calculating outputs of the chromatic / achromatic determination circuit 9 and the fine line degree calculation circuit 8. This is the point.
[0033]
The chromatic / achromatic determination circuit 9 will be described. Conventionally, methods for determining chromatic pixels / achromatic pixels with high accuracy have been proposed. Here, an embodiment will be described in which the black character fine line degree is calculated using them.
[0034]
As the chromatic / achromatic determination method, for example, the applicant previously proposed as Japanese Patent Laid-Open No. 3-260878, and the chromatic / achromatic determination circuit 9 and the fine line degree calculation circuit 8 are processed in parallel. As a result, it is possible to reduce the processing delay on the image reproduction side as compared with the case of processing them serially. That is, in the black generation circuit, it is necessary to delay until the recognition result of the target pixel is output. Therefore, when there are a plurality of recognition processes that require a line buffer, they are performed in parallel processing, and the maximum delay amount among the plurality of recognition processes becomes the delay amount in the plurality of integrated processes. On the other hand, if serial processing is performed, the delay is simply added.
[0035]
The logic circuit 10 outputs (s1) the fine line degree of the pixel determined to be achromatic by the chromatic / achromatic determination circuit 9. Note that the chromatic / achromatic determination may be processing that outputs an intermediate level between chromatic and achromatic, instead of the binary determination as described above.
[0036]
【The invention's effect】
As described above, according to the present invention, UCR processing suitable for each of the black character thin line portion, the color character thin line portion, and the halftone dot pattern region can be performed. In particular, low UCR processing can be realized over the entire halftone dot region.
[Brief description of the drawings]
FIG. 1 shows a configuration of Embodiment 1 of the present invention.
FIG. 2 shows an example of a fine line degree calculation circuit.
FIG. 3 shows an example of an MTF correction filter.
FIG. 4 shows an example of a black bar pattern.
FIG. 5 shows a pattern example when calculating the line width.
FIG. 6 shows an example of table conversion for normalizing line widths.
FIG. 7 shows another example of the fine line degree calculation circuit.
FIG. 8 is a diagram for explaining detection conditions for ridge pixels;
FIG. 9 shows an example of a steepness degree detection mask.
FIG. 10 shows an example pattern when detecting continuity.
FIG. 11 shows a configuration of Example 2 of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Color scanner 2 Log conversion part 3 Color correction circuit 4 Black generating circuit 5 Subtraction circuit 6 Color printer 7 min (R, G, B) calculation circuit 8 Fine line degree calculation circuit

Claims (2)

A color image processing apparatus comprising means for reading a document as a digital signal, calculating a fine line degree of the target pixel from the digital signal, and means for increasing the black generation amount of the target pixel as the calculated fine line degree is higher The color image processing apparatus is characterized in that the means for calculating the fine line degree calculates a steep degree of the character ridge. Means for reading an original as a digital signal and generating an achromatic signal from the digital signal; means for calculating the fine line degree of the target pixel from the achromatic color signal; and the higher the calculated fine line degree, A color image processing apparatus comprising means for increasing the amount of black generation, wherein the means for calculating the fine line degree calculates a steep degree of a character ridge.

Images