JPH02166968A - Method and device for picture processing and transmission - Google Patents

Abstract

PURPOSE:To form the recorded picture of satisfactory gradient and tone by specifying correlative relation between the density value of an original picture and the area ratio of a picture element recorded in the recorded picture. CONSTITUTION:Conversion processing is executed by an expression I from the density value (x) of a sampling point on the original picture to the ratio (y) of the number of the unit picture element recorded to the number of the unit picture element which constitutes a picture element block corresponding to the sampling point in the recorded picture to be formed. In the expression I, (x) is the density value of the sampling point (x) on the original picture and (y) is the ratio of the number of the unit picture element ot be recorded to the number of the unit picture element which constitutes a picture element block Y corresponding to an X on the recorded picture. Then, yH is the ratio of the number of the unit picture elements recorded to the number of the unit picture elements, which constitute the picture element block in a brightest part H of the recorded picture, and yS is ratio of the unit picture elements recorded to the number of the unit picture elements which constitute the picture element block in a brightest part S in the recorded picture. Then, alpha is reflection factor of a recording sheet, beta is 10<-gamma>, (k) is a gamma/original picture density area and gamma is an arbitrary coefficient. Thus, the recorded picture of satisfactory quality can be formed.

Description

[Detailed Description of the Invention] (Field of Industrial Application) The present invention provides a novel method for converting image information signals obtained by photoelectric scanning of original images such as continuous tone images in facsimiles and the like that process and transmit image signals. Relates to a method and apparatus for performing conversion processing using a gradation conversion method and forming a recorded image having a pixel distribution corresponding to the density of a yK document on a recording sheet on a receiving side based on the gradation-converted output signal. It is something.

(Prior art) A document image having continuous gradation such as a photograph is converted into an image signal on the sending side of an image transmission means such as a facsimile, and then transmitted.
When recording this on a recording sheet on the receiving side, it is difficult to reproduce the density gradation, especially in the case of color facsimiles.

It is not easy to adjust the color tone (color balance) as well as the density gradation described above.

For this reason, recently, similar to the method of converting continuous originals into halftone dots in photolithography in printing, image signals obtained by photoelectric scanning of original images with continuous gradations such as photographs are processed. Many improvements have been made to image processing techniques that use signals to form an image on a recording sheet using a pixel distribution having gradations and tones corresponding to the original image.

However, in the currently used image processing technology, when the gradation characteristics of the original image are reflected in the final recorded image based on the pixel distribution, what kind of characteristics should the pixel distribution have?

Since no scientific studies have been conducted on how to obtain such a pixel distribution, images with satisfactory gradation and color tone reproduction cannot be obtained. is the current situation.

That is, for the density value of a predetermined sample point on the original image, the pixel on the recorded image corresponding to the sample point.

No scientific correlation formula has been developed to determine what kind of concentration values should be correlated, and currently these device manufacturers are making decisions in advance based on experience, intuition, or a limited number of fixed conditions. I have no choice but to depend on what I do.

As a result, originals with image quality that equipment manufacturers did not expect,
For example, in the case of non-standard color film originals (overexposed originals that are too bright, underexposed originals that are too dark, etc.), it is difficult to obtain a desired recorded image with excellent one gradation or color tone. Therefore, recorded images of the desired quality can be obtained not only from originals with standard image quality but also from non-standard originals mentioned above, and the image quality of the original can be arbitrarily changed or modified (changes in gradation or color tone, corrections etc.). ), it has not been possible to develop a flexible image processing and transmission device that can perform This means that it has not been scientifically elucidated what pixel density value should correspond to each value on the corresponding recorded image.

(Problem M to be Solved by the Invention) The cause of the above-mentioned problems in conventional image processing is that when forming a recorded image based on the final pixel distribution from both fM original images such as the continuous technique 1 I Jrs image, The idea lies in the process of image gradation conversion, which is the first step and plays an important role. That is, the density value of a predetermined sample point on a document image is expressed as the ratio of the number of recorded unit pixels to the number of unit pixels constituting the pixel block (hereinafter referred to as one pixel source) in a pixel block on the corresponding recorded image. When converting to gradation values (sometimes referred to as degree values), the conventional way of thinking about gradation conversion is that it must be done based on scientifically rational gradation conversion means, but it must be done solely based on experience. The reason is that it relied on intuition.

The inventors of the present invention have focused on this situation, and have established a rational image gradation conversion technique in order to ultimately streamline the image processing and formation process and to form recorded images of excellent quality. We conducted extensive research based on the basic understanding that this must be the case.

(Means for Solving the Problems) To summarize the present invention, 1. The present invention processes an image information signal obtained by photoelectrically scanning an original image on a transmitting side using a gradation adjustment mechanism, compresses the processed signal, etc. The receiving side modulates the received signal and performs restoration processing to create an image information signal. Based on this signal, the pixel distribution corresponding to the original is printed on the recording sheet. In an image processing transmission device such as a facsimile for forming a recorded image by The difference between the density value and the density value at the brightest part on the same image) is calculated as the number of unit pixels to be recorded relative to the number of unit pixels constituting the pixel block corresponding to the sample point in the recorded image to be formed. To the ratio (y),
The present invention relates to an image processing and transmission device that performs conversion processing according to the following relational expression 〇〉.

Relational expression ...■ Hereinafter, the configuration of the present invention will be explained in detail.

In image processing and transmission such as facsimile, a document image such as a continuous tone photograph is converted into an image information signal, one image processing and transmission is performed, and an image corresponding to the document is recorded from the received image information signal on the receiving side. In the process of forming on a sheet, one image information signal is processed so that image elements at arbitrary sample points of the original image correspond to dots (pixels) or aggregates thereof, and an image is formed on the recording sheet based on this. be done. As is well known, in image forming devices on recording sheets used in facsimiles, etc., and output printers, in order to reproduce halftones, each pixel (halftone dot) is created in accordance with the density of the original image, as in the case of photolithography in printing. Because it is difficult to change the size of a pixel block, which is a matrix-like collection of unit pixels of a fixed size, the number of unit pixels that make up the pixel block and the number of unit pixels that are recorded as an image are It is often possible to correspond to the size of the halftone dot by changing the ratio.

By the way, what can be considered as a basic component for expressing an image is the pixel density value (the number of recorded pixels out of the number of unit pixels constituting a given pixel block) expressed by the pixel distribution described above. ) and the surface reflection density of the image recording material (toner, etc.), of which human vision can detect, for example, 1% of the size of the halftone dot area in a printed image. The pixel density value expressed by the pixel distribution, which has the same relationship as the size of the halftone dot area, is extremely important as a means of forming images. play a role. That is, in the recorded pixels of a certain predetermined pixel block.

When we examine the effects that changes in the amount of toner applied to them and changes in pixel size have on the gradation, we find that the latter is much larger, and it is extremely important how the density value of each pixel should be set. This is a serious problem.

In addition, in connection with the above, when trying to form one image, it is important to note that the quality content of the original image varies widely, that the image forming process has various characteristics, and also that the image quality evaluation standard is These complex and unstable factors must be overcome.

For this reason, when converting a document image such as a continuous tone image into a recorded image based on the pixel distribution, the density ratio (νH) of the pixel block in the brightest area (H) in the recorded image based on the pixel distribution to be created and the density ratio (ys) of the pixel block in the darkest part (S) can be arbitrarily selected, and the gradation of the image from the brightest part (H) to the darkest part (S) can be rationally and easily adjusted. It is absolutely necessary to have a means of coordinating and managing this.

The gradation adjustment method of the present invention was devised based on this idea, specifically, the gradation adjustment method defined by the above-mentioned relational expression.

First, the process of deriving the above <Relational Expression 0> will be explained.

The present inventors have proposed that when creating a halftone print image from a continuous tone color film original, 1. the tone can be rationally converted (conversion from continuous tone to halftone tone). For this reason, we previously proposed a gradation conversion formula that is the predecessor of the above-mentioned relational formula (2). (Patent Application No. 148912-1982, Patent Application No. 148912-
(See No. 2590).

The gradation conversion formula (hereinafter referred to as relational formula 〇〉) proposed by the present inventors can be used not only to create printed images, but also to create recorded images using various printers, or to transmit and reproduce images by facsimile etc. Although it can be used sometimes,
The explanation will be limited to the creation of a print image as shown below.

<Relational expression 〇> ...■ Comparing the relational expression 〇〉 and the relational expression 〇〉, the meanings of β value and i value are different, and 〈Relational expression ■〉 does not specify the γ value. These differences will be described later, but in order to help understand the present invention, the process of deriving the relational expression (2) will be explained.

The relational expression (2) for determining the numerical value (ν) of the halftone dot area percentage used when creating the above-mentioned printed image is derived from the generally accepted density formula (photographic density, optical density).

When this general formula regarding density is applied to plate making and printing, it becomes as follows.

O Density (D') in plate making/printing = log - ■ Relational formula ■〉 is the density formula (D') related to plate making/printing
In addition, it is possible to arbitrarily set halftone dots of a desired size on the H part and the 8th part of the printed image.

The basic density value (1
) and the numerical value (-t) of the halftone area percentage of the halftone dot at the corresponding sample point on the halftone gradation image. It was induced to match.

When applying the above relational expression 〇〉 to the image gradation conversion method when creating a printed image, the reflectance (α) of the printing paper,
Surface reflectance (β) of printing ink and printed image density range/
Based on the value of the original image density range ratio (嬬), determine the size of halftone dots (P) that you want to place on the H and 8 parts of the printed image
While arbitrarily selecting ut 'ts), calculate the value of the halftone dot area percentage of the halftone dot at the corresponding sample point (Y) on the printed image from the basic density value (1) of an arbitrary sample point (X) on the original image. It is operated to find (y). Thereby, the density gradation of the original image (continuous gradation image) can be faithfully reproduced on the printed image (halftone gradation image) with 1=1.

In addition, multicolor plate making (generally cyan (C), magenta (M),
In the case of four plates (yellow (Y) and black (BL), which are considered to be one set), the standard plate (in the case of multicolor plate making, as is well known, the cyan plate (C) is the standard plate. ), i.e., the halftone gradation characteristic curve that serves as the standard for converting the density information value of the original image into the halftone dot area value of the printed image (obtained by graphing the work value and V value described above) (This is a curve that serves as the basis for converting continuous gradation to halftone gradation.)
Once determined, the working standard characteristic curves for other color plates can always be determined by multiplying the value of ν of the standard plate by an appropriate adjustment value based on the gray balance ratio of each printing ink color.
It goes without saying that the working standard characteristic curves for each color plate determined in a manner that can be determined rationally are each reasonable characteristic curve, and furthermore, the gradation and color tone between those characteristic curves are The mutual relationship with respect to this is also reasonable and appropriate.

In other words, when creating a halftone print image from a continuous tone original image, if the tone conversion is performed based on the above relational expression (■), then the tone of the image can be calculated by relying on conventional experience and intuition. It is possible to break away from the conversion method and convert the image gradation arbitrarily and rationally.

Therefore, it is also possible to rationally adjust the color tone, which is closely and inseparably related to the gradation. The above-mentioned methods previously proposed by the present inventors make it possible to obtain a printed image having a density gradient and color tone that are natural to the human visual sense.

However, in subsequent research, the above-mentioned relational formula ■〉
It has been found that there are certain limitations in the operation of

Its limitations are: - It is extremely effective if the original image is of standard quality, but it is extremely effective if the original image is of non-standard quality, especially if it is of extremely poor quality (for example, overexposure when taking a photograph). or under).

To explain this from the point of view of the operation of the above relational formula ■, ・In the case of a standard quality document (standard manuscript), (the molecule that determines the value is printed with a solid yellow ink with a large stimulation value) Printing density value (the typical density value is 0.9 to 1.0
It is. ) to perform gradation conversion (in the case of multi-color plate making, the 0 plate is produced using this value), it is extremely effective, but it is especially useful for non-standard originals with poor quality content as mentioned above.・When dealing with non-standard originals, the surface reflectance of the printing ink (as mentioned above, yellow ink is the standard) and other values may not be fully satisfied with the β value. For example, even if a company is selected and adopted, it may not be fully satisfactory.

In order to overcome the above-mentioned limitations, it is necessary to adapt the halftone gradation characteristic curve, which serves as the work standard for gradation conversion, to not only standard originals but also non-standard originals.

The shape of the curve must be able to be changed arbitrarily and rationally. As a result of study, the present inventors have found that satisfactory results can be obtained when gradation conversion is performed under first-order conditions.

- 1 value = γ/(density range value of the original image) - γ value = any positive or negative value - β value: A value determined from the γ value that defines the above A value by β = 10-.

By applying the above-mentioned relational formula (■) under the above conditions, it is possible to create printed images with excellent reproducibility of density gradation and inseparable color tones from standard and non-standard originals. can.

The above explanation has centered on the creation of printed images with halftone gradations, but the theory that supports the gradation conversion work described above can also be applied to image processing in facsimile machines and the creation of images in transmission devices. Yes, it goes without saying.

It goes without saying that the gradation conversion formula suitable for image processing and image creation in the transmission device of the present invention becomes the following relational formula (2) when the above-mentioned study results are incorporated and organized.

Next, the meaning of each term of the above-mentioned relational expression 〇〉 of the present invention, operational characteristics, etc. will be explained.

In the operation of the above-mentioned relational expression (2) of the present invention, the basic density value (1) must be determined from the image information signal of the original image. The density information value may be any value as long as it reflects the physical quantity related to the density of each pixel of the original image, and should be interpreted in the broadest sense. Synonymous terms include reflection density, transmission density, brightness, light amount, current/voltage value, etc. These density information values may be extracted as density information signals by photoelectrically scanning the original image.

In the above relational expression (g) of the present invention, the basic concentration value (1) is measured using a densitometer (for example).

0.2 to 2.70 for portraits on positive color film
There are some that have a concentration value of . ), also yn[
Pixel density value set for the @search block in the brightest part (Dan)
The number. ), y [pixel density value recorded in a pixel block (Y) corresponding to an arbitrary sample point (X) on the original image] is calculated as a percentage value.

In the operation of the above-mentioned Kukusha-style of the present invention, it can be used by deforming it as shown in the following, but it can also be used by any processing, deformation,
You are also free to use it as a guide.

ν"?o+E (110-'")(ys?)
I) In the above modification, α=1. This is based on the surface reflectance of the recording paper (base material) being 100%. In practice, the value of α may be set to 1.0.

Moreover, according to the above modification (α==1.0), the brightest part H on the recorded image in the image processing and transmission equipment of facsimile etc. o and νS at the darkest part S can be set as planned. This is the brightest part H on the recorded image.
In this case, engineering = O.

= γ), therefore, it is clear that one filter = -γ.

In the operation of the above relational expression of the present invention, α.

β, γ (as mentioned above, β = 10, so β
Define the value. ) takes various values. In the present invention, by appropriately selecting these values, it is possible to rationally perform image gradation conversion processing regardless of the quality characteristics of the original image. can.

That is, the image gradation conversion processing method of the present invention based on the above-mentioned relational expression (2) reproduces the gradation and color tone of the original image, that is, reproduces the tone of the original image in a recorded image with a 1=1 ratio. However, its usefulness is not limited to this. In addition to faithfully reproducing the characteristics of the original image, the above-mentioned relational expression (■) of the present invention can also be applied to α, β, γ values, and even ? This is extremely useful for rationally changing or modifying the characteristics of a document image by appropriately selecting the H+rS value.

To explain this in detail, when applying the above <relational expression ω>,
It should be noted that users (workers) have the following degrees of freedom.

Bad point 1: Use the relational expression ■ for the purpose of forming an image that is faithful to the original image, that is, to obtain an image that visually and perceptually looks exactly the same when observed with the human eye. In the present invention, the attitude of gradation adjustment, which involves applying the relational expression 〇〉, is explained using the term ``conversion of gradation (of an image)''.

Bad point 2: Using the relational expression ■ to change or modify the original striped image due to the technical needs of image formation, artistic requirements, or the needs of the ordering party. The primary objective is to obtain a modified or altered visual and sensory image itself when observed with the eyes.

To operate the relational expression 〇〉, this Inventive tone adjustment attitude like ``(image) floor m() correction (also) (changed)” ing.

When forming a multicolor image using the above <Relational Expression 〇> 1
For example, when a color original is reproduced using an image processing and transmission device such as a facsimile according to the present invention, color separation, which is well known in the field of printing, is used to separate light reflected from the color original into blue (B) and green (G). , red (R) to obtain a density information signal for each color, process this with a gradation adjustment mechanism using the above relational expression 〇〉, and form an image based on this processed information. In that case, is it related to the standard color version (for example, 0 version)? value, that is, the gradation characteristic curve (γ value) of the standard color plate, and
By plotting the values, a gradation characteristic curve similar to the halftone gradation characteristic curve in printing technology is obtained. ), and the gradation characteristic curves for other color versions (M version, Y version) are determined by multiplying the γ value of the reference color version by an appropriate adjustment value based on the gray balance ratio of each ink. Since it can be determined rationally, an image can be formed using this image information.

The γ value for each color plate determined as described above, that is, the gradation characteristic curve for each color plate, is defined by the relational expression The correlation between the curves in terms of gradation and color tone is also reasonable and appropriate.

As explained above, when a recorded image is formed by a device for image processing and transmission such as a facsimile, the gradation adjustment mechanism section is equipped with hardware that performs gradation conversion based on the above-mentioned <Relational Expression Or by incorporating software,
It is possible to obtain a recorded image with excellent reproduction of tone as well as gradation, or a recorded image in which the image quality of the original image is arbitrarily modified or changed.

The present invention will be explained in more detail below based on examples, but the present invention is not limited to these examples unless it goes beyond the gist of the present invention.

As described above, the present invention has the greatest feature in that the gradation is converted in the gradation adjustment mechanism of a facsimile or the like using the relational expression (2). Therefore, we will explain how to use the relational expression 〇〉 with ease, especially the handling of the γ value.

(Example) ■About the method for determining the γ value used in the relational expression 〇〉.

When a recorded image is formed by an image processing and transmission device such as a facsimile, the present invention performs the core gradation conversion work in the recorded image creation process based on the above relational expression 〇〉. Its greatest feature is that it allows people to do things.

In this case, it is desirable to create an image of the same quality as a recorded image formed from a standard original whose quality content is standard, even from original images that vary in quality such as brightness or darkness. Of course.

To achieve this, we must be independent of the quality of the original image.

γ that gives the same quality as the recorded image obtained from the standard manuscript
It is necessary to obtain a gradation characteristic curve that defines the relationship between the γ value and the γ value.
It is a value.

Below, we will explain how to determine the γ value, which has extremely important significance in the operation of the relational expression (■). By determining this, it is possible to form a recorded image with excellent gradation and color tone reproducibility.

Table 1 shows how the γ value (that is, pixel density value) changes with respect to various γ values. Table 1 shows that while changing the γ value (γ
Value = 2.00 to -〇, 20 is adopted).

The above 〈Relational Expression〇〉, 'ts=3%+? 5=95%,
Each density step (1st Table 1 shows the γ values in which the density range of the original image is divided into nine steps.

(The following is a blank space) According to Table 1, when the γ value is changed, individual gradation characteristic curves corresponding to each change can be obtained.

Therefore, it is sufficient to perform gradation conversion by setting the optimum quality based on the quality content of a given original image. The results in Table 1 are illustrated in FIG.

Therefore, when a document image with predetermined quality content is given, the optimal γ value to be adopted in the relational expression 〇〉 is determined as follows:
The problem is how to make a rational decision.

We used monochrome and color films, which are said to have particularly faithful reproduction of gradations and tones, as original images, and established a method for determining the γ value to be adopted according to the image quality of the original. do. I mean.

This is because if a method for determining the γ value that is effective for monochrome and color film originals with high gradation reproducibility is established, it will be useful for other original images as well.

A detailed analysis of the image quality of the original color film revealed that
The image quality, such as high-key (photographed with overexposure) and low-key (photographed with underexposure), is vastly different compared to a standard color film original photographed with standard exposure. be. However, the difference in image quality of color film originals is because the difference in exposure directly affects the brightest density value Hn of the original.
By focusing on this point, it is possible to define it objectively.

1. As previously proposed by the present inventors, the standard manuscript (
In case of proper exposure,), the γ value is 0.9~
Considering that it takes a value of 1.0, it is sufficient to find the correlation between Hn and γ. Note that Hn was selected because the density region near the brightest part is important in gradation reproduction. Theoretically, the darkest part density value S of the original
It goes without saying that you may choose n.

Therefore, we conducted an experiment to form recorded images of excellent image quality using various color film originals and to determine the relationship between H and the γ value. The experimental data are shown in Table 2. In Table 2, experiment Nα2 is from the m quasi-manuscript, and the γ value is 0.
9 was adopted.

Table 2 (Note) Hn and Sn indicate the brightest and darkest density values of a given individual color film original.

From these experiments, the γ value can be reasonably determined by the following formula.

(i) γ in Table 2. When the relationship between and Hn is graphed as shown in Figure 2 (total logarithm graph), γ. is determined by the formula below.

γ. =γ. ±IDnltanoc (ii) In addition, standard manuscripts (density range 0.20-2.
80) as γ. =1. An experiment was conducted in which recorded images were formed using OO, and recorded images of the same quality were obtained from various color film originals. As a result, γ. The relationship between and Hn could be defined as linear.

(a) γn=1.70 2.2961(logHn
+1) (Relational expression obtained when both γ□ and Hn are displayed on a logarithmic scale) % formula %) (Relational expression obtained when γ is expressed on a normal scale and Hn is expressed on a logarithmic scale) From the above In order to reproduce recorded images with excellent gradation reproducibility using image processing and transmission equipment such as facsimiles from original images with widely varying quality contents, first, the γ value is calculated from the Hn value of the original image. .

It is only necessary to determine this, then use it as the γ value of <Relational Expression 0>, and perform the conversion process for one gradation.

In order to operate the relational expression It is sufficient to add a mechanism for measuring Hn of various original images and a mechanism for calculating γ value from Ho to the processing and transmission device, or it is also possible to leave these measurements and calculations to the operator.

■ Fixation (constantization) of the γ value adopted in the relational expression■〉
Regarding the method for determining the above-mentioned relational expression (g) of the present invention, the above-mentioned method for determining the γ value is complicated, and the recorded image created by this method is, strictly speaking, obtained from a standard manuscript. The image differs from the recorded image. This is natural since the brightest density value (Hn) of a color film original is different from the brightest density value (Ho) of a standard original.

As explained earlier, in the relational formula (■) useful for gradation conversion of standard originals, a value of γ = 0.9 (or a value between 0.9 and 1.0) allows reproduction of not only one gradation but also color tones. It is possible to create duplicate images with excellent quality. Therefore, the relational expression〇〉
In this operation, in order to set the value of γ to a constant such as γ=0.9, the density gradation of the original image must be adjusted (corrected) to the density gradation of the standard image. Below, γ
We will explain how to convert a value into a constant.

In the case of a color film original, the density gradation adjustment described above can be carried out very rationally. This will be explained with reference to FIG.

As is well known, the relationship between the exposure amount (X) of the color film sensitive material (note that it is different from X of the sample point X mentioned above) and the color film density (D) at that time is shown in Figure 3. This is as shown in the basic concentration characteristic curve.

The standard document and the non-standard document depend on whether the exposure amount is appropriate or not, and each density characteristic curve is shown as having a specific range on the basic density characteristic curve. In Figure 3, the former is shown as a standard concentration characteristic curve, and the latter is shown as an individual concentration characteristic curve (note that
FIG. 3 shows an underexposed original as a non-standard original. ) Therefore, adjusting the density characteristics of a non-standard original to the density characteristics of a standard original can be done extremely easily by converting the basic density characteristic curve into a function.

The basic concentration characteristic curve described above is defined by the function D=fo(X), as shown in Table 3 below (Table 3 also shows the inverse function).

Note that the method of converting the basic concentration characteristic curve into a function in Table 3 should be understood as an example, and a more simplified formula may be used.

(Leaving space below) Table 3 (Example of function representation of basic concentration characteristic curve) (Note) In the basic concentration characteristic curve shown in Figure 3, the function fo < A function fX(D) for determining the function fX(D) is shown.

(Note) In order to faithfully define the basic concentration characteristic curve, it is expressed mathematically for each domain of X or D.

To match the individual density characteristic curve of a color original to the standard density characteristic curve, the first step should be followed (see Figure 3).
.

(i) Density values of H and S of a color original image and the basic density characteristic curve of the color film sensitive material of the color original (D=fo
(X)), define the individual density characteristic curve of the color original image, and (ii) determine the density value D Hn M of the color original.
Substituting j D sn into X=fx(D), find the value range of the color original image on the X axis, X, n-X, n, and (
) The value range on the X axis of the concentration characteristic curve based on this,
Match HO to XSO. (Story) Next, the value range of the D axis of the reference concentration characteristic curve, 000 to I)so, is determined.

It goes without saying that if the individual density characteristic curve of a color document matches the reference density characteristic curve, matching of the two is unnecessary.

It is also possible to set an arbitrary tolerance range for the reference density characteristic curve, and perform image processing by assuming that the curve is the same as the reference density characteristic curve when it is within the tolerance range.

In the matching procedure of the individual and reference concentration characteristic curves described above,
Since it is normal that XRo (exposure range for standard originals) and XRn (exposure range for non-standard color-based originals) do not match, XRn is replaced by XR.

It is necessary to match the above (■) and (ii)
Refer to step i)). To match XRn to XRo, there are two methods: simple matching (matching the brightest density values to the same value and not worrying about matching the darkest part) and proportional matching (the brightest density matching). There is an attitude of matching both the darkest part density value and the darkest part density value.
In the figure, the case of mathematically proportional matching is shown.

As shown in FIG. 3, the density information value of the individual density characteristic curve (density information value 5on between Do. and Den) is substituted into the basic density characteristic curve D=fo(X), and XR1 is obtained.
The density information value (density information value between DHQ to D3o, Do) of the color original is obtained by adjusting the X value adjusted to XR, but the X value after adjusting XRn to XRo is obtained. The relational expression for determining the value can be obtained by simple calculation as follows.

■ For simple alignment: X = fx (Dn) ± 1m ■ For proportional alignment: m: Required amount of translation Standard image quality (DH0=
0.20.0s. = 2.80), high key (overexposed) one (Don=0.10.D5n=2.70)
, and low-key (underexposed) ones (DHn=0.
60°Dsn=3.20), and matching the individual concentration characteristic curves to the standard concentration characteristic curve based on the basic concentration characteristic curve shown in Table 3 is shown in Table 4 below. .

(Space below) Table 5 (Tone characteristic curve setting data) In the matching experiment described above, the density range (DR) of the three color film originals used was all 2.60, so one was simple. matching, and proportional matching for others.

In the density values of Dn and Do in Table 4 above, the ν value (pixel density value 9%
) was sought. The results are shown in Table 5, and the correlation between the ν values and Dn values in Table 5 is shown in FIG.

The curve shown in FIG. 4 makes it possible to create a recorded image with excellent gradation reproducibility from a non-standard quality original image.

This is a gradation characteristic curve that defines the correlation between the engineering value and the ν value.

(Left below) In order to operate the gradation adjustment mechanism of the facsimile and other image processing and transmission devices of the present invention by fixing (constantizing) the γ value of <Relational Expression 〇>, It is necessary to incorporate into the device a mechanism for measuring the density of the original image (measuring the density of H, S, and H-3), and software and hardware that matches the individual density characteristic curve of the original image with the standard density characteristic curve. This makes it possible to create recorded images with excellent gradations and color tones from originals of any quality content.

■About the device for image processing and transmission of the present invention.

(i) Overview of gradation adjustment During image processing and transmission such as by facsimile, the image formed on the receiving side is represented by minute dots (pixels). When the original image has continuous gradation, it is difficult to change the size of pixels (halftone dots) like halftone dots in printing, so as mentioned above, it is difficult to change the size of pixels (halftone dots). This method, which often expresses gradation by the distribution of recorded pixels (number of recorded pixels, distribution form) within the containing pixel block, is already well known, but this method will be briefly explained below. It is shown in FIG.

The example in FIG. 5(a) shows changes in the distribution of pixels within a pixel block, and one pixel block is composed of 4×4=16 unit pixels in this example. The number of unit pixels formed as a constituent part of an image within the image density represents the image density corresponding to the sum of the areas of the pixels. In other words, if the number of pixels is the same, the image density will be the same regardless of the distribution position of the five pixels, but in the image formation process, the pattern of this pixel distribution is determined and the pixels are sequentially formed by performing calculations according to that pattern. Just keep doing it.

For example, as shown in FIG. 5, in the case of the column (a) in FIG. 5, the distribution of recorded pixels becomes more dispersed within the pixel block as the number of recorded pixels increases. Alternatively, for example, it may be possible to make the dots spread out in a spiral shape from the center of the pixel block, in which case the dots will be similar to halftone dots in photolithography. Also, the column (b) shows halftone dots having an area corresponding to the number of pixels in the column (a).

Although the pixel block has been described here as a 4×4 matrix type, 17 levels of gradation can be expressed by this.

-n2 matrix type pixel blocks for boat fishing
+1 step of gradation (0 to 100%) is expressed.

This method of expressing the density of a continuous tone image by the distribution of pixels formed in matrix-type pixel blocks is commonly known as the dither matrix method.

(For example, JP-A-58-85434, JP-A No. 58-1145)
No. 69, No. 58-52969, No. 60-141585
No. 62-186663, etc.).

(it) Apparatus for image processing and transmission The apparatus for image processing and transmission of the present invention will be explained based on FIGS. 6 to 8.

Figure 6 shows image processing using the gradation adjustment method of the present invention.

FIG. 2 is a block diagram of a device that performs transmission and image formation on a recording sheet. There are various possible devices for forming images on recording sheets on the receiving side, but here we will use an electrophotographic device that forms a latent image on a photoconductive image forming body by scanning a laser beam. There is. Sending side equipment MA
The detection section 2 reads the original 1 by photoelectric scanning, etc., and the correction processing section 3 performs necessary processing such as shading correction on the output signal of the detection section 2, and then the gradation adjustment section 4 processes the original according to the relational expression The distribution state of pixels to be formed on the recording sheet is determined in accordance with the density of . When transmitting the image information signal obtained by the gradation adjustment section 4, the compression section 5 performs compression processing to remove redundancy of the image information in order to improve transmission efficiency and speed.

As for this compression processing method, CCITT's G3. There are MH method, MR method, etc. based on the G4 standard. The compressed image information signal is modulated into a carrier signal for transmission in the modulator 6, and then transmitted via a line, data network, etc. In device B on the receiving side, the received signal is first modulated in the modulation section 7, and further restored to an uncompressed image information signal in the restoration section 8. The restored signal is converted by the output section 9 into an image forming signal for the image forming section 10 used when forming an image on a recording sheet.

In the case of an electrophotographic image forming method, a uniformly charged photoconductive image forming body surface is scanned with a laser beam modulated by a signal from the output section 9 to form a charge distribution corresponding to an image. A latent image is formed, and this latent image is developed with toner in a developing device. The toner is charged as necessary by frictional charging or the like so that it can easily adhere to the latent image. The toner image formed on the image forming surface is transferred to a recording sheet in a transfer section. Normally, in this transfer section, transfer is performed while a corona discharger discharges the toner from the back side of the recording sheet to transfer the toner to the recording sheet side. The recording sheet onto which the toner image has been transferred is transferred to a fixing section, where it is fixed by heating, pressure, etc., and image formation is completed.

This is a process in which one type of toner is used, but next, an example in which a multicolor image is formed will be explained with reference to FIG.

In FIG. 7, a color original 1 is read by a detection unit 2, and the output signal is subjected to color separation in a processing unit 3.
The signal is color separated into G, B, and USM component signals, and necessary processing such as shading correction is performed. The output signal of the processing section 3 is subjected to N tone adjustment in the tone adjustment section 4 according to the relational expression (2).

The signal is converted into a signal indicating the distribution of pixels to be formed on the recording sheet in accordance with the density of the original. The output signal of the gradation adjustment section 4 is compressed in a compression section 5 for transmission, modulated in a modulation section 6, and transmitted. The signal received on the receiving side B is demodulated by the demodulation section 7 and restored to the uncompressed image information signal by the restoration section 8. The restored image information signal is converted into a signal for forming an image in the image forming section 10 at the output section 9 for each color component. Output section 9
The image forming section 10 forms a latent image by scanning the laser beam in the same manner as in the case of one color, using signals for each color from the image forming section 10.

The steps of development and transfer are performed, and the steps for each color are repeated as many times as the number of colors, and after alignment and transfer to the recording sheet, fixing is performed, and the image forming step is completed.

After one image formation is completed, there is toner remaining on the photoconductive image forming body without being transferred to the recording sheet, so this is removed with a cleaning blade.

It is removed with a brush or the like, and the remaining charges are removed by light irradiation or corona discharge to prepare for the next image formation.

If the toner is transferred to the recording sheet after each color is developed, this toner should be removed after each color has been transferred, or if the toner is to be transferred at once after all colors have been developed onto the image forming body. The remaining toner is removed after one transfer.

In FIGS. 6 and 7, the image forming section 10 has been described as an electrophotographic type that forms an electrostatic latent image on a photoconductive image forming body by scanning a laser beam, but a recorded image is formed by the distribution of one pixel. As a forming method, other methods such as electrostatic recording and magnetic recording heads can be adopted.

In electrostatic recording, each electrode of a recording head has a large number of electrodes arranged in a direction perpendicular to the moving direction of the image forming body in the vicinity of or in contact with a moving image forming body made of a dielectric material in the form of a sheet or a rotating drum. A voltage is applied to form an electrostatic latent image. The steps after the step of adding toner to this latent image and developing it are the same as in the case of electrophotography. A voltage as an output signal corresponding to the image to be recorded is applied from the dot control section of the output section 9 to the recording head as an assembly of electrodes. In addition, in the magnetic recording type, the image forming body is, for example, a drum whose surface is uniformly coated with a magnetic material, and a voltage as an image information signal is applied to the magnetic recording head in contact with the surface. A magnetic latent image is formed on the recording surface by relatively moving the recording surface and the recording surface. A toner made of magnetic material is used to develop this magnetic latent image, but other processing is performed in the same manner as in the case of electrophotography. In the case of a magnetic latent image, it takes longer to form a latent image than in the case of an electrostatic latent image, but since the persistence of the formed latent image is good, it is possible to form a latent image once with one type of toner. An image is formed by developing with toner many times. Therefore, it is suitable for receiving the same original image signal of one color and forming images on multiple sheets. In the case of color images, a plurality of magnetic image forming bodies should be provided for each color, and each should be provided with a developing device or the like.

FIG. 8 shows the gradation adjustment method according to the present invention in more detail. Reflected light or transmitted light from a color original is detected by a photoelectric conversion element such as a photomultiply or a solid-state image sensor (CCD) in the detection unit 2. R, G, B, U as current
Each SM signal is output, and the A/V converter 21 converts this signal into a voltage signal.

In the color separation unit 3, the voltage signals of R, G, B, and USM from the detection unit 2 are logarithmically operated in the log amplifier 31 and converted into density, and the basic masking unit (BM) 32 converts this density into gray (K). In the zero-order color correction (CC) 33 that separates the components and further separates each component of Y, M, and C, the Y component, two control components, and The C component is controlled, and the gray component of the original is processed by UCR (under control removal) in the UCR/UCA section 34.
), or UCA (under control a)
ddit4on), determine the ratio expressed by the three components Y, M, and C. These Y.

Conventionally, after the M and C components are obtained, a gradation adjustment section (IM
Area ratio occupied by pixels of each component in the gradation control section in C) Ye'+me, ca'
, I used inverse log transformation to find ke', but
In this embodiment, an adjustment section 41 is used in place of the gradation control section and the inverse log conversion section, and here, from Y, M, C, K, ya', ll1a', ce
'.

Conversion to ke' is being performed. The adjustment unit 41 has an algorithm of <Relational Expression (G)> inside, and Y, M, C.

Apply 〈Relational Expression〉〉 to each of K, ye.

Find ms', ce', and ke'.

The gradation adjustment unit 41 has the algorithm of the relational formula
A general-purpose computer with F (interface), an electric circuit realized by a general-purpose IC with the algorithm as internal logic, and an RO that holds the calculation results of the algorithm.
It can take four different forms, such as an electric circuit including M, a PAL that embodies the algorithm as internal logic, a gate array, and a custom IC.

The pixel area ratio obtained by the adjustment unit 41 is input to the color channel selector 42, and the color channel selector 42 inputs ye', ms', as', k
e' is sequentially and selectively output. This output is A/D converted by the A/D converter 43 and then input to the compression processor 5.

The device configuration for this gradation adjustment is merely an example, and may be modified, omitted, simplified, etc. as necessary.

Regarding the usefulness of the relational expression 〇〉, the usefulness of the relational expression 〉〉 applied to the gradation adjustment mechanism of the image processing and transmission device of the present invention will be supplementarily explained.

This is a supplementary explanation to facilitate understanding of the present invention, and mainly describes the operation of the relational formula 〇〉 applied to the gradation adjustment mechanism of the image processing and transmission device of the present invention and the significance of the result. I will explain in detail as follows.

(a) Operational experiment of the relational expression 〇〉 The following two experiments were conducted as basic experiments for incorporating the relational expression 〇〉 into the gradation adjustment mechanism of the image processing and transmission device of the present invention6 a. ) First of all, using a normal simple calculator, namely Sharp Pythagoras EL509A (manufactured by Sharp Corporation),
By operating the simple calculator while applying desired values to "Zero Adjustment Method", image gradation adjustment parts shown in Tables 6, 7, and 8 below were created.

As a result, the times required for these tasks, including the time required to check the calculated results, were 3 hours, 2 hours, and 2 hours, respectively.

b) The following experiment was also conducted.

Simple personal computer (NEC PC-98
01-M2), the desired software obtained separately is input as function data, and the basic am value (1) of the original image (continuous tone image) is recorded in the pixel block on the recorded image according to the pixel distribution corresponding to it. Ratio of number of pixels (
y) (hereinafter referred to as the area ratio of recorded pixels).

As a matter of course, the result was the same value as the result of manual calculation using the above-mentioned simple calculator.

Moreover, in this experiment, there was no need to use any special software to create the above software used to adjust the gradation of the image input to the personal computer, and it was created using the N88-BASIC that came with the personal computer. The work took only one hour to complete.

In addition, software that allows you to directly input the measured values from a densitometer ranging from highlights (H) to shadows (S) of the original image instead of the basic density value of the original image can also be used to convert or correct the gradation of the target image. It was confirmed that it can be done.

Using these software, set the desired density interval (for example, 0.05 to 0.00 to 1.00) on the original image.
Carving, 1. (0.10 increments from OO to 3.00)
By setting the value and inputting the value to the personal computer, it was possible to obtain the target pixel area ratio (v).

Furthermore, by inputting the density values of a plurality of locations from highlights to shadows on the original image, it was possible to obtain desired area ratios (v) of recorded pixels corresponding thereto.

The area ratio of pixels recorded by the software described above (1F
) allows output of both positive and negative images individually or at the same time.

(b) Calculation results obtained from <relational expression (g)> and their usefulness Next, the usefulness of Table 6, Table 7, and Table 8 will be explained. (In each section, the ν value is the pixel density value.) [Regarding Table 6 ■■■] Table 6 shows the image recording materials such as toner when forming a black and white image from an original image using an image forming device. density (in the table, it is indicated as the recorded image density range, which corresponds to the γ value of the relational formula The value is displayed, 0~
Three cases are shown: 100%, 0-98%, and 0-95%. ) changes, in order to obtain an ideal tone characteristic curve, the pixel area ratio (ν) at each sample point is changed.
This is a list of how (pixel density values) should be set.

Also, from this table, even if the density of toner, etc. is the same, when the usage range of the area ratio of recorded pixels is changed (that is, when the γ value is changed), the ideal gradation characteristic curve is You can see how things can and should change.

In Table 6, the β value that determines the E value is determined by β=10−y. Incidentally, when the recorded image density range=γ value=1.0, ε=1/(1−β)=1.1111.

Also, the value of one pixel density value (%) is β=O(t=
1.0) is the theoretical value.

Furthermore, it is necessary to acquire techniques and techniques for forming a black and white image with a 1:1 pixel distribution from a document image such as a continuous Fl tone image, and for arbitrarily adjusting the gradation characteristics of a black and white image. It is also the basis of multicolor image formation.

C About Table 7 J Table 7, like Table 6, shows that when a black and white image is formed from a yXX tone image by an image forming apparatus, the density of the image forming material (toner, etc.) changes (i.e. the recorded image density. range = when the γ value changes), the usage range of the maximum pixel density value is from 0% to
The area ratio (ν) of pixels at each sample point required to form an image with the same image tone and similar image quality to the human visual sense, excluding the contrast of the entire image, assuming 100%. This is a list of (pixel density values).

In other words, when the conditions are ideal, the area ratio (y
) is a list.

[About Table 8] Table 8 has the same basic conditions as Table 7.

When using the maximum pixel density value usage range (5% to 95%), what is the area ratio (y) of recorded pixels at each sample point on the ideal gradation characteristic curve (pixel density value)? This is a table showing what should be set.

Until today, color separation work in the creation of printed images, etc.
Color correction using masking technology
tion) is primarily emphasized, and image gradation adjustment work basically remains solely dependent on human experience and intuition, or on a limited number of fixed, given data. For this reason, it is necessary to develop a scientific gradation conversion technique when creating a duplicate image, based on the side of the image to be duplicated, such as a printed image.

The relational expression 〇〉 of the present invention performs gradation conversion in a rational method when creating a duplicate image.

The data in Tables 6 to 8 show the correlation between the basic degree value (1) of the original image obtained by the relational expression 〇〉 and the area ratio (ν) of the recorded pixels.

It serves as a useful basic material for scientific examination of various basic matters in color separation work during image formation.

From these parts, what is the essence and principle that exists between the original image and color separation work, and what should be paid attention to and considered in order to rationally match that essence and principle with practice? You can extract what needs to be done.

(c) Regarding the application of the relational expression 〉 to correction (or change) of the gradation, the following relational expression It is effective not only for converting to a gradation image based on the distribution of gradations, but also for modifying (or changing) the gradation of the original image itself.

This correction (or change) of the gradation of the image is based on the intention of the orderer, the type of target image in the color original, the purpose of use of the image to be formed, etc.
There may be cases where this is necessary due to the whiteness of the recording paper, the density of the image recording material (ink), etc., but in any case, it can be handled rationally by applying the relational formula 〇〉, and various Color separation work can be standardized.

Furthermore, according to the present invention, it is possible to modify (or change) the gradation of an image in a highlight portion or a shadow portion in the same manner. As shown in Figure 1, this is a gradation characteristic curve (a curve that defines the correlation between the engineering value and the ? value) depending on the γ value adopted.
This is clear from the fact that the shape of can be changed arbitrarily. Furthermore, it has been confirmed that the gradation conversion according to the relational expression (g) of the present invention can automatically remove color fog in the highlight portions of a color original without taking any special countermeasures.

(The following is a margin) [Effects of the Invention] The image gradation adjustment method according to the present invention, the image processing using the same, and the transmission method have the following excellent effects.

l) In determining the correlation between the density value of the original image and the surface area ratio of the recorded pixels in the recorded image formed on the receiving side, which is a fundamental problem in processing and transmitting images with continuous gradation, Rather than something irrational like
The following relational expression (■) can be replaced with a very rational and simple determination method. In addition, when converting a continuous tone image to a recorded image based on pixel distribution, tone management (gradation conversion,
Corrections or changes) are not limited to just the gradation of the image.

Since it is directly related to the tone of the image, by adopting the relational expression 〉, it is possible to rationally manage and systematize the gradation and tone of the image.

2) By incorporating the algorithm of <Relational Expression 〇> into a device for processing and transmitting images, equipment can be rationalized and simplified, and manufacturing costs can be reduced. In addition, the operation of these devices is simplified and clarified, and the number of reworks can be extremely reduced. In particular, it has the great advantage of being able to form recorded images with excellent gradation and color tone, no matter what the quality of the original image being handled.

3) The gradation adjustment mechanism that employs the algorithm of <relational expression ω> scientifically adjusts the gradation characteristic curve for forming a recorded image based on the density characteristics of the original image up to H, S, and H-5. Since it can be defined objectively, it is possible to rationalize the gradation adjustment mechanism of the current complicated and unscientific image processing and transmission apparatus.

4) Evaluation standards for the quality of recorded images to be formed can be rationally and easily defined independently of the content of the original image. Therefore, customer needs can be met rationally.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the relationship between the γ value and the shape change of the gradation characteristic curve. Figure 2 shows γ. It is a correlation diagram of Hn and Hn. FIG. 3 is a diagram illustrating the principle of matching between the individual density characteristic curve of a color film original image and the reference density characteristic curve. FIG. 4 is a diagram showing a gradation characteristic curve set for a non-standard original. FIG. 5(a) shows an example where a continuous tone original image is expressed by the distribution of unit pixels within a pixel block, and FIG. 5(b) corresponds to the case of (a). FIG. 2 is a diagram showing a case where halftone dot sizes are expressed in photoengraving. FIG. 6 is a block diagram of an apparatus that performs image processing, transmission, and recording using the gradation adjustment method of the present invention. FIG. 7 is a block diagram of an apparatus that performs image processing, transmission, and recording of a color original using the gradation adjustment method of the present invention. FIG. 8 is a block diagram showing an example of the configuration of a gradation adjustment section incorporating the gradation adjustment method of the present invention. Patent applicant Yamatoya Shokai Co., Ltd. Agent Yoshio Mizuno, patent attorney Figure 1 Relationship between r value and change in shape of gradation characteristic curve C”r knee 0.
20~2.00) 23456789 Continuous tone density step diagram (tone characteristic curve) Fig.

Claims (1)

[Claims] 1. On the transmitting side, the image information signal obtained from the original image is processed by a gradation adjustment mechanism, then the processed signal is compressed and modulated, and transmitted, and the received signal is demodulated on the receiving side. In a method for processing and transmitting an image in which an image is formed using an output signal obtained by performing a restoration process, the gradation adjustment mechanism adjusts the basis of an arbitrary sample point on a document image based on an image information signal. The density value (x) (the difference between the density value at the sample point and the density value at the brightest part on the same image) is calculated as a unit that constitutes a pixel block corresponding to the sample point in the recorded image formed on the receiving side. Image processing and transmission characterized in that the ratio (y) of the number of recorded unit pixels to the number of pixels is subjected to conversion processing according to the following <relational expression (1)> to perform gradation adjustment. method for. Relational expression> y=y_H+{α(1-10^-^k_x)/(α-β
)}(y_S−y_H)...(1) [However, x: Basic density value of an arbitrary sample point X on the original image. That is, the density value is obtained by subtracting the density value at the brightest part H of the same image from the density value at an arbitrary sample point X of the same image. y: Ratio of the number of recorded unit pixels to the number of unit pixels constituting the pixel block Y corresponding to the X on the recorded image formed on the receiving side. y_H: Ratio of the number of recorded unit pixels to the number of unit pixels constituting the pixel block set for the pixel block of the brightest part H of the recorded image formed on the receiving side. y_S: Ratio of the number of recorded unit pixels to the number of unit pixels constituting the pixel block, set for the pixel block of the darkest part S of the image formed on the receiving side α: Reflectance of the recording paper . β: Numerical value determined by β=10^-^γ. k: A value determined by γ/(density range of original image). γ: arbitrary coefficient. respectively. ] 2. A detection unit that reads a document by photoelectric scanning or the like and converts it into an image information signal, and a distribution of pixels to be formed on a recording sheet by performing correction processing on the output signal of the detection unit and further performing gradation adjustment processing. an image information transmitting side device having a processing section that determines a state, a compression processing section that applies compression processing to an output signal of the processing section, and a transmission control section that transmits after modulating the output signal of the compression processing section; A demodulating section receives and demodulates the signal from the transmitting side device, an output section decompresses the output signal from the demodulating section and outputs it as an uncompressed image information signal, and the signal from the output section is used to output the signal to the recording sheet. A receiving side device includes a recording image forming section that forms an image based on pixel distribution. In the apparatus for image processing and transmission, the gradation adjustment process of the image information signal is performed as defined by <Relational Expression (1)> according to claim 1. A device for processing and transmitting images, characterized by: 3. After the image forming section forms a latent image representing the distribution of the pixels by scanning a laser beam on an image forming body having a uniformly charged photoconductive layer, and develops the latent image with toner; 3. The apparatus for image processing and transmission according to claim 2, wherein the image is transferred onto a recording sheet and further fixed. 4. The image forming section applies a voltage to a large number of recording electrodes arranged in a direction perpendicular to the moving direction of the moving electrostatic recording type image forming body to create an electrostatic latent in the electrostatic recording body. 3. The apparatus for image processing and transmission according to claim 2, wherein an image is formed, the latent image is developed with toner, and then transferred to a recording sheet and further fixed. 5. A series of operations, or a part thereof, of forming a latent image on the image forming body, developing it with toner, and transferring it to a recording sheet,
A claim in which the same operation is performed using toner of a specific color, the same operation is performed using toner of a different color, and the same operation is repeated for the required number of colors, aligned and transferred onto the same recording sheet, and then fixed. An apparatus for processing and transmitting images according to item 3 or 4.