JPS62149258A - Picture processing system - Google Patents

Abstract

PURPOSE:To reduce the skipping of gradation being a problem at a dot intermediate section by narrowering an input density range giving an intermediate density when number of reproduced gradation of an output density is decreased and widening the input density range when the number is increased. CONSTITUTION:A dither circuit 6 makes a dither size and an input density versus output density characteristic curve different depending on the kind of an input picture. For example, the range of an input density giving an output density region is made lease as to, e.g., a character/line drawing part to apply the expression of four gradation. When the resolution of an operation switch 9 is selected high, the intermediate density of the output density is made arisen slightly than other parts as to the dot/intermediate tone so as to make the range of the input density an intermediate width. In this case, the expression of eight gradation is applied. In throwing the switch to the position of 'low', as to the dot/intermediate tone part in this case, the expression of 32 gradation is applied based on a curve where the input versus output is nearly proportional, then the range of the input density to give the output density is provided by the largest range.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an image processing method that performs image processing by varying the number of reproduction gradations depending on the type of input image or by arbitrary selection.

[Technical background of the invention and its problems]

Area gradation methods such as the textured dither method are well known, but in order to express gradation using this method, a certain degree of resolution must be sacrificed.

In other words, in the organized dither method, the square root of the number of gradations and the resolution (
(1g+number of halftone dots per unit) has a certain relationship. Therefore, when a certain level of resolution is required, such as when an input image contains characters, the number of gradations must be sacrificed.

On the other hand, conventionally, when displaying a halftone image using a printer, the input density is divided at approximately equal intervals from the lowest level to the highest level, and a predetermined gradation is provided corresponding to each input level.

However, when assigning gradations in this manner, the following problem occurs.

That is, the halftone/medium IJI portion generally has the largest amount of information in the intermediate density. Therefore, when the number of gradations is reduced, information □concentrates in the middle of the gradations. Furthermore, in normal printers, the characteristics are not necessarily linear at intermediate densities, and the input density regions assigned to each level are non-uniform. Therefore, with the conventional method of allocating gradations at equal intervals from the lowest level to the highest level of input density, when the number of gradations is reduced as described above, a considerable number of the intermediate densities, which contain the most information, are Parts are output at the same density level. As a result, there is a problem in that gradation jumps in halftone areas are noticeable, resulting in significant deterioration in image quality.

[Purpose of the invention]

The present invention was made in consideration of such problems, and
Even when the number of gradations is reduced, the intermediate density level is 11i.
The purpose of the present invention is to provide an image processing method that does not cause noticeable tonality jumps and provides comparable image quality.

(Summary of the Invention) When changing the number of reproduction gradations of an output image as necessary, the present invention provides a method for changing the input density range that gives an intermediate density to the output image when the number of reproduction gradations of the output density is decreased. When the number of output density reproduction gradations is increased, the range of input density that gives an intermediate density to the output image is widened.

〔Effect of the invention〕

According to the present invention, even when the reproduction Wi key of the output image is reduced, the range of input densities that give the output image an intermediate 31 degrees is narrow, so that in the intermediate densities where the most information j is present among the input densities, there is only a small Even changes in gradation can be expressed at different gradation levels.

Therefore, according to this invention, the number of tones and! Even when using a gradation expression method that has limited image intensity combinations, it is possible to prevent intermediate density levels with a large amount of information from being output at the same level, making jumps at intermediate density levels less noticeable. image quality can be improved.

In addition, by expressing the intermediate density level with relatively fine gradations and expressing the vicinity of black or white level with the same -WAvA, it is possible to reduce the problems that may occur during scanning in the case of images with many solid areas, such as illustrations. This makes it possible to solve the problem of stains on solid areas due to the influence of noise.

[Embodiments of the invention]

Examples of the present invention will be described below.

FIG. 1 is a diagram showing an embodiment in which the present invention is applied to a digital color copying machine.

The scanner 1 scans an original (not shown), photoelectrically converts the light reflected from the original through three color filters, and obtains a digital image signal Si' by A/D converting these signals. be. Note that here, i represents the color of the color filter, i-(red (r), green (g)
, blue (b)).

(yellow (y), green (g), cyan (C)), etc. are usually used. This digital image signal of 3 Hz is input to the shading correction circuit 2. In the shading correction circuit 2, the signal value obtained by reading the white reference plate is, for example, “1°°
, the signal value obtained by reading the black reference plate is normalized so that it is output as, for example, “O11.If the reflectance of the white reference plate is Rwi and the reflectance of the black reference plate is Rbi, then the reflectance of the original is The signal value Si when is R1 is expressed by the following equation.

Si − (Ri − Rbi) / (Rwi − Rbi)
-=(1) The color signal S1 standardized using the white and black reference plates is a signal in which sensitivity unevenness of the sensor and light unevenness due to illumination light, color filters, etc. are corrected.

Furthermore, even when the sensor is a CCD sensor with multiple chips arranged on the same substrate, a signal can be obtained in which unevenness such as variation in gain of multiple amplifiers connected to each sensor chip is corrected.

The color signal Si is input to the matrix circuit 3,
The signal is separated into a luminance signal (2) and color difference signals C1 and C2. Separated luminance signal ■ and color difference signal Cr.

C2 is input to the filtering processing circuit 4 and converted into I', Cs', and C2', respectively. These m
11 signals 111 color difference signals CI' and C2' are input to the color conversion circuit 5. This color conversion circuit 5' is composed of, for example, a read-only memory (ROM), and the luminance signal I
', color difference signals CI', and C2', which are converted into ink amounts required when outputting the color expressed by the combination of the values of the color difference signals CI' and C2' in a table. When the signals 1', Cr', and C2' are inputted to the color conversion circuit 5, an output converted into the amount of ink required for printing is obtained. This color conversion table may be created using a method described using a masking equation, a Neugebauer equation, or the like. Each signal converted into the ink amount, for example, each signal for yellow, magenta, cyan, and black in a four-color printer, is input to the dither circuit 6. This dither circuit 6 is a circuit that performs binarization using a dither method so that even a color printer capable of only binary recording can perform halftone recording.

The dither circuit 6 selects a dither size for basic dither and a dither correction curve in accordance with the result of determining the type of input image as described later, and performs dither processing based on the dither size and dither correction curve. The signal binarized by the dither circuit 6 is output to the digital converter 7. The printer 7 prints ink of each color according to an input signal to generate a color copy.

The luminance signal I separated by the matrix circuit 3 is also input to the character discrimination circuit 8, and the Laplacian is binarized to extract data characterizing the type of input image such as a character image or illustration. It is determined whether the part of the image currently being processed is a character/line drawing or a halftone/halftone part, and the dither circuit 6 is switched. In addition, the filtering processing circuit 4 applies a high-frequency emphasis filter to the portions of the luminance signal I determined to be characters/line drawings by the character determining circuit 8, and applies a high-frequency emphasis filter to the portions determined to be halftones/halftones. A low-frequency emphasis filter is applied to this area, and no filter is applied to the other areas. On the other hand, a low-pass emphasis filter is unconditionally applied to the color difference signals C1 and C2. therefore,
The dither processing described below is controlled depending on the image type.

Further, the dither circuit 6 is also given information from the operation switch 9. The operation switch 9 is used by the operator to arbitrarily select the resolution of the output image. That is,
The operation switch 9 can select either "high" or "low" resolution.

The dither circuit 6 switches the content of processing for the ink amount signal from the color conversion circuit 5 based on the determination result of the filtering processing circuit 4 and the selection information of the operation switch 9.

For example, for characters/line drawings in the input image,
Regardless of the resolution selection information on the operation switch 9,
As shown in FIG. 2(a), a dither matrix with the minimum size of 2×2 is selected.

The number of gradations at this time is 4, which is the smallest value among the selectable numbers of gradations. In addition, when printing characters more clearly, it is also possible to use two gradations.

On the other hand, for a portion of the input image that is determined to be a halftone dot or halftone, one of two dither sizes is selected by selecting the operation switch 9.

That is, when the resolution is selected with the operation switch 9, 2fl is displayed as shown in FIG. 2(b).
A dither matrix of intermediate size of X2fl is selected. The number of gradations at this time is eight, which is the middle number among the number of selectable gradations.

In addition, if you select “low” resolution, please refer to Figure 2 (b).
), the dither matrix with the maximum size of 4-r"Z .

In this way, the number of dither tones used for text/line drawing parts is M+, the number of dither tones used for halftone dots/halftone parts when a high-resolution pattern is selected is M2, and the resolution is " If the number of dither tones used for the halftone/halftone part when low °' is selected is Ml, then Mr, M2
, Ml, the following magnitude relationship is established.

Mt <M2 <Ml −(
2) Ml is the smallest because there is no need for intermediate density information for characters and line drawings, and it is appropriate for the number of gradations to be about four gradations at most. In this way, in the I-weave dither method used in this embodiment, the product of the square root of the number of gradations and the resolution is constant, so in order to increase the resolution, it is necessary to reduce the number of gradations. Therefore, when the resolution is set to "high", the number of halftones and halftones is greater than when the resolution is set to "low".

On the other hand, the dither circuit 6 varies the dither size and the input density versus output density (here, density refers to the value normalized with the solid ink being 1) characteristic curve as well as the dither size depending on the type of input image.

In other words, for the text/line drawing part of the input image,
As shown in Figure 3 (a), the input density range that provides the middle 11 degree region of the output mKJ is minimized.Following this correction curve, the four gradations are as shown in Figure 4 (a). The width of the intermediate density region D1 is the narrowest.

In addition, when the resolution of the operation switch 9 is set to ``high degree'', the halftone and halftone areas are set so that the intermediate density area of the output density is slightly higher than the other areas, as shown in Figure 3(b). Based on this curve, e and 11m are expressed as shown in FIG. It's sato.

Furthermore, when the operating switch 9 is set to ``low'', the input and output are approximately proportional to each other in the halftone dot and halftone areas, as shown in FIG. 3(C). Therefore, when 322 gradations are expressed based on this curve, the input density range D3 that provides the intermediate density of the output density is the widest range, as shown in FIG. 4(C).

With this setting, by narrowing the range that is output as a uniform level in the intermediate density area of the input image's density area in the text/line drawing area, it is possible to reduce the problem gradation in the halftone/halftone area. It is possible to reduce the flying of.

Figure 5 shows an example of using 8-gradation dither for halftone dots and halftones when the resolution is selected as "high".As a result of experiments, the recording device used this time has It was possible to express the input density vs. output density characteristic curve obtained when no correction was applied to the digital recording device.
It was as shown in FIG. 5(a). As can be seen from this figure, even when the input density is changed in a constant width, the output density does not necessarily change in a constant width.

FIG. 5(b) shows the characteristic obtained when such a characteristic is corrected so as to obtain a linear characteristic with a slope of 1. As is clear from this figure, a fairly wide density area of the input image is expressed at the same level. It is necessary to increase the expressiveness of the image at intermediate densities by allocating the largest number of gradation levels to the portion concentrated on one level. Therefore, the present inventors conducted an experiment with subjects to determine how much of the input density region can be allocated to one gradation level when using 8-gradation dithering. As a result, it was found that on average, up to 0.08 looks natural. Also, the number of gradations used in this experiment is 6.
Since it is a gradation, in this case, the expressible region of the input image has a value of 0.08x6=0.48. The characteristics obtained when correction is performed in this manner are shown in FIG. 5(C). As is clear from this figure, (1)
) is corrected to a curve characteristic as shown in (C), the maximum gradation jump in the input density is reduced from 0.34 to 0.19. In the case of (b), a fairly wide portion (approximately 1/3 of the total) of the input density region is output at the same level, which is not very good. On the other hand, in the case of (C), the output at the same level is narrower to just under 115 overall, so the load on this level is considerably reduced. As in this example, when performing correction, it is more effective to use a curve that has a steep slope at intermediate densities and becomes gentler as it approaches the white and black levels, rather than necessarily having a slope of 1 or linear. It's also possible.

The halftone dots and
Halftone areas are effective when used for input documents such as illustrations because dither halftone dots are not noticeable and solid filling is less affected by noise and is easier. be.

Note that the present invention is not limited to the above embodiments.

In the above embodiment, 2111 dither was used for gradation expression, but the present invention can also be applied to an arrangement using multi-value dither of three or more values. In this case, even if the number of gradations does not change the dither size, it can be done by switching the dither from, for example, binary to ternary. In other words, it is possible to obtain a high-resolution image even if the number of gradations is large.

Although the above has mainly described an example in which the image processing device of the present invention is applied to a digital copying machine, the present invention is not limited to copying machines, but can be applied to various devices such as color television cameras, electronic still cameras, etc. Needless to say, it can be applied to other uses.

[Brief explanation of drawings]

Fig. 1 is a block diagram of a digital copying machine according to an embodiment of the present invention, Fig. 2 is a diagram showing each dither pattern corresponding to the resolution in the embodiment, and Fig. 3 is a diagram showing each dither pattern corresponding to the resolution in the embodiment. A diagram showing the corresponding density correction curve, FIG. 4 is a diagram showing a density correction curve obtained by combining the above dither pattern and the above density correction curve, and FIG. FIG. 2 is a diagram showing an example of density correction. 1...Scanner, 2...Shading correction circuit,
3... Matrix circuit, 4... Filtering processing circuit, 5... Color conversion circuit, 6... Dither circuit, 7...
...Printer, 8...Character discrimination circuit, 9...Operation switch. Applicant's agent Patent attorney Takehiko Suzue (a) (b) Figure 5 (C)

Claims (1)

[Claims] In an image processing method that changes the number of reproduction gradations of an output image variously as necessary, when the number of reproduction gradations of an output image is decreased, the input density range that gives an intermediate density to the output image is narrowed, and the number of reproduction gradations of the output image is reduced. An image processing method characterized by widening the input density range that gives an intermediate density to an output image when the number of reproduced gradations is increased.