JP3129421B2 - Image gradation recording device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image recording apparatus that performs halftone recording of an image signal representing shading in a limited range of recording gradation per pixel.

[Conventional technology]

When reproducing a grayscale image (halftone image) in a digital copying machine, it is general to binarize the grayscale image by the systematic dither method and output it from a printer. However, the systematic dither method cannot express the part where the concentration change is large,
There is a disadvantage that edges of characters and line drawings are blurred. Further, when a halftone image used for a printed matter or the like is reproduced, moire occurs due to interference between a periodic pattern of the halftone dots and a dither pattern.

In addition, there is a method using an error diffusion method (minimum average error method) as one having relatively excellent resolution (for example, Japanese Patent Application Laid-Open No. 61-5677). The error diffusion method is a conditional decision method that determines a threshold value for a certain pixel of an input image in consideration of a pixel level in the cycle. However, among the binary recording methods, the resolution and gradation characteristics are relatively good visually. The present invention relates to an improvement in a recording apparatus for realizing the error diffusion method.

In the error diffusion method, a possible value of a pixel level Xij (i, j indicates a secondary coordinate in an image) of one pixel of an input image is normalized to a range from 0 to 1, while an output for the image is normalized. Assuming that Yij is 0 or 1, that is, only on / off, an error eij between the input and output of the pixel is expressed by the following equation.

eij = Xij−Yij (1) However, the possible value of Yij is a discrete value compared to Xij. That is, the output gradation is smaller than the input gradation.

Further, when the target pixel signal Xij is input, a correction operation is performed by the following equation using the error of the peripheral pixel whose output value has already been determined.

Xij ′ = Xij + 1 / (Σαkl) Σ (αkl · ei + kj + 1) (2) αkl is a weight coefficient and is a matrix element. The following is an example of α.

The * mark indicates the position of the target pixel Xij. In the equation (2), the output value Yij is determined by comparing Xij 'with predetermined m threshold groups. Assuming that m thresholds are T _{1 to} Tm, Is determined. The number of thresholds Ti (i = 1, 2,..., M) is determined by the number of output gradations. When the number of gradations is m + 1, m thresholds are required. For example, to determine the octal output, 7
It will be compared with the threshold values.

FIG. 11 is a schematic block diagram of a conventional apparatus for performing the above processing. When the target pixel data Xij is input,
Xij 'is calculated by the correction operation circuit in accordance with the above equation (2). The correction operation circuit includes a multiplier, an adder, and a latch for calculating the product-sum operation of the error e i + k j + 1 of the peripheral pixels and the weight coefficient αkl and calculating the sum of the result and Xij. If the coefficient and 式 αkl are set to be a power of 2 as shown in the following equation (5) as the matrix α, multiplication and division can be performed only by bit shift, so that an expensive multiplier is not used. There is a merit that can be done.

Xij 'is compared with a plurality of thresholds Ti in the next multi-level circuit to determine an output value Yij. The multi-level circuit is composed of a plurality of comparators, and the error operation circuit is a subtractor, which calculates a difference between Xij 'and Yij, that is, an error eij. eij is stored in an error memory, and is used for an error correction operation for determining an output value from the error memory. As shown in FIG. 12, the multi-level conversion circuit and the error operation circuit can have a simpler configuration if a ROM (or RAM) table is used. Xij 'is input as a ROM address signal. As a result of comparing the predetermined Ti and Xij 'with the address addressed by Xij' in the ROM, the Yij and the error eij are stored, so that high-level multiplication and error arithmetic processing can be performed in a table reference formula. .

Generally, an input image signal is often a signal linear in density value close to a human psychological stimulus value.
However, in the error diffusion method, addition and subtraction of an image signal are performed when calculating an error and a correction value. Since the density value is the logarithm of the reciprocal of the reflectance (or transmittance), the physical meaning of the sum or difference of the density values is unclear. On the other hand, since the reflectivity is substantially linear with the area ratio of the region to be blackened, it has an effective meaning for the sum and difference.

Therefore, in order to more strictly control the reproducibility of the output value with respect to the input value in the error diffusion method, it is preferable to use a linear reflectance image signal. However, when it is not necessary to control the absolute value of the output value and only consider the relative value between pixels, an image signal having a linear density or other characteristics may be used. That is, when it is sufficient that the change in the density of the image is reproduced smoothly, it is not necessary to stick to the characteristics of the image signal.

[Problems to be solved by the invention]

In the error diffusion method, the value of an output dot is evaluated to a predetermined value irrespective of binarization or multi-valued conversion, regardless of the output of surrounding pixels (for example, Japanese Patent Application Laid-Open No. Sho 61-5677). The evaluation of the pixel output will be described below, but for simplicity, the case of binarization processing will be described.

6a is assigned to each of the nine pixels arranged in three columns and three rows.
As shown in FIG. These are the target pixels to be binarized. The eight pixels and 〜 are eight adjacent peripheral pixels. The main scanning direction and the sub-scanning direction are as shown in FIG. 6a.

In the error diffusion method, since the quantization error of a pixel whose output has already been determined is sequentially corrected, the output values of four pixels up to may be considered as the influence of peripheral pixels.

FIG. 6b shows a configuration of a pixel and a shape of a recording dot. In a laser printer or an inkjet printer, the shape of a dot can be approximately regarded as a circle. Assuming that the pixel pitch is P, the dot diameter (radius r) is set to a size including a rectangular pixel area of one side P so that no gap occurs in the solid portion. Therefore, overlapping occurs between adjacent recording dots. FIGS. 7a, 7b, and 7c show this overlapping state.
This is shown in the figure and FIG. 7d.

FIG. 7a shows one dot (black solid portion). This dot area is defined as A.

FIG. 7b shows an overlapping portion (black solid portion) of two adjacent dots in the sub-scanning direction (up and down). The area of this overlapping portion is B. Since the dots are circular, the area of the overlapping portion of two adjacent dots in the main scanning direction is also B.

FIG. 7c shows an overlapping portion (black solid portion) of two obliquely adjacent dots. The area of this overlapping portion is defined as C.

FIG. 7d shows an overlapping portion (black solid portion) of the four dots when the four dots are adjacent to each other. This area is defined as D.

FIG. 8 shows an output pattern of four pixels (〜) around the recording pixel (). As shown in FIG. 8, there are 16 output patterns. Hereinafter, the dot gain (the area to be newly blackened) of the target pixel in each output pattern will be described. The outline circle drawn in the peripheral pixel area indicates that the pixel is recorded (blackened), and the solid black area is newly blackened by the recording in the target pixel. Indicates the region to be used. White circles indicate recording dots of pixels adjacent to the target pixel.

The dot gain of the pixel of interest in each of the 16 patterns a to p shown in FIG. 8 can be expressed as shown in Table 1 using the above-described overlapping areas A to D. In Table 1, if P ² , which is the area of the pixel region on one side P, is P ² = 1, the dot gain is as shown in Table 2.

Since B> C ≧ D, the dot gain becomes larger than 1 except for the pattern o or p. Conventionally, the dot gain was evaluated to be 1 irrespective of the output pattern of the peripheral pixels, so that the overall output image was dark. Further, when the dot gain evaluation is corrected to a uniform rate, for example, to the average value of each pattern, a dark portion of the image is output brightly, and a problem that black solid cannot be reproduced in particular has occurred.

Therefore, the invention of the present application provides, in addition to the image data conversion error correction by the error diffusion method, correction according to the overlap area between a pixel whose print output gradation level has already been determined and a plurality of pixels adjacent to the pixel, A first object is to improve halftone reproduction characteristics.

By the way, the output state of the printer (for example, one dot diameter or one dot recording density) changes with the lapse of time of the recording apparatus, environmental changes,
Alternatively, it may change due to a difference in recording paper. When the output state of the printer changes, it is necessary to reset or adjust the above-described error data. The invention of the present application is to arbitrarily reset or adjust the error data,
A second object is to obtain an output image which is always constant and has good reproducibility.

[Means for Solving the Problem 1] In order to achieve the first object, the first invention of the present application provides a recording output level (Yij) of a target pixel of an input image,
Intermediate determined based on the input gray level (Xij) of the pixel and error data (e i + k j + 1) corresponding to the gray level error between input and output for the pixel whose print output gray level has already been determined. Tone reproduction processing means (100, 110, 120, 130 /
100, 110, 160), the halftone reproduction processing means outputs the recording output gradation level (Yi + kj +) of the pixel whose recording output gradation level has already been determined.
l) and the recording output level (Yij) of the target pixel,
Error data (ei) corresponding to the gray level error between the input and output of the target pixel according to the gray level (Y'ij) corresponding to the overlapping area of the target pixel and a plurality of adjacent pixels.
j) an error output means for generating (150, 140, 130/150, 160),
It is characterized by including.

Symbols in parentheses indicate corresponding elements in the embodiment shown in the drawings and described later.

[Operation 1] According to this, in addition to the image data conversion error correction by the error diffusion method, the error output means (150, 140, 130/150, 160)
Generates error data (eij) according to the gradation level (Y′ij) corresponding to the overlapping area with a plurality of adjacent pixels. The disturbance of the recording density is corrected, and the halftone reproduction characteristics are improved.

[Means for Solving the Problem 2] In order to achieve the first and second objects, the second invention of the present application is a method for determining the recording output level (Yij) of a pixel of interest in an input image by changing the input output level of that pixel. Error data (e i + k j +) corresponding to the difference between the gradation level (Xij) and the gradation level between the input and output of the pixel whose print output gradation level has already been determined.
l), the halftone reproduction processing means (100,
110, 120, 130/100, 110, 160), wherein the halftone reproduction processing means has a recording output gradation level (Yi + kj +) of a pixel whose recording output gradation level has already been determined.
l) and the recording output level (Yij) of the target pixel,
Error data (ei) corresponding to the gray level error between the input and output of the target pixel according to the gray level (Y'ij) corresponding to the overlapping area of the target pixel and a plurality of adjacent pixels.
j) an error output means for generating (150, 140, 130/150, 160),
The apparatus further includes a test pattern forming unit (3) for forming a test pattern of the recording pixel, and an image reading unit (1) for reading the test pattern of the recording pixel formed by the test pattern forming unit (3).
And an evaluation parameter setting means (30) for setting or updating error data in the error output means based on the image data of the test pattern read by the image reading means (1).

[Action 2] According to this, in addition to the image data conversion error correction by the error diffusion method, the error output means (150, 140, 130/150, 160)
Generates error data in accordance with the gradation level corresponding to the overlapping area with a plurality of adjacent pixels, so that the recording density disturbance around the target pixel due to recording / non-recording (output pattern) of the adjacent pixels is corrected, Halftone reproduction characteristics are improved.

Further, since the error data is set or adjusted based on the actual printing based on the test pattern, the change in the dot printing density (for example, dot diameter, 1 dot density) due to the aging of the printing apparatus, environmental changes, etc. However, high gradation reproducibility is maintained.

Other objects and features of the present invention will become apparent from the following description of embodiments with reference to the drawings.

〔Example〕

FIG. 1 shows a gradation information processing circuit 33P for correcting image information in accordance with the printing characteristics of a printer according to an embodiment of the present invention. The processing circuit 33P includes an error memory circuit 100, a correction operation circuit 110, a multi-level conversion circuit 120, and an error operation circuit 130.
, A dot gain evaluation circuit 140, and a data memory 150. Except for the dot gain circuit 140 and the data memory 150, it is the same as the conventional circuit shown in FIG.

The output data memory 150 stores the output data Yi + kj + 1 of the peripheral four pixels or two pixels and simultaneously refers to the output data. The output data memory 150 includes a line memory and a latch.

The dot gain evaluation circuit 140 can be easily realized by using a ROM or a RAM table. That is, by storing the dot gain (Yij ′) in advance at the address specified by the output data of the peripheral pixel and the output data of the target pixel, the dot gain evaluation value can be obtained in a table reference formula. The capacity of this ROM (or RAM) is 2 bits × (4 + 1) when the output data is a 2-bit signal.
Assuming that an 8-bit output is required for an evaluation value for an input of = 10 bits, 1k × 8 bits are required. In addition, if a RAM is used, there is an advantage that when the dot gain evaluation value changes due to a change over time in the recording apparatus, an environmental change, or the like, data can be easily changed to cope with the change.

The error calculation circuit 130 outputs the error-corrected input data (Xi
j ′) and the dot gain evaluation value (Yij ′) are output as an error (eij) at the target pixel position.

FIG. 2 shows a modification of the processing circuit 33P shown in FIG. 1. The processing circuit 33P shown in FIG.
The multi-value conversion circuit 120, the dot gain evaluation circuit 140, and the error calculation circuit 130 are converted to a multi-value / error calculation circuit 160 (ROM or RA
M). This multi-level / error calculation circuit 1
Numeral 60 can perform high-speed processing in a table reference manner by using input data after error correction and output data of peripheral pixels as input (address) data, and using the output value and error of the target pixel as output values. If the output data is a 2-bit signal, the input data is an 8-bit signal, and the error is an 8-bit signal, the address space is 2 bits × 4 + 8 bits = 16 bits, and the output is 2 bits + 8 bits = 10 bits. Commercially available products have 1kbit RAM and S of 64k x 16bit configuration.
-RAM, D-RAM or the like can be used. In addition, by combining the multi-value conversion circuit and the dot gain evaluation circuit into one, the output value (Yij) can be selected from the output value of the dot gain that minimizes the error with respect to the input value (Xij ′). It is possible.

FIG. 3 shows the mechanism of one embodiment of the present invention. This is a digital color copier. This copier includes a scanner unit 1 for reading a document, an image processing unit 2 for electrically processing an image signal output as a digital signal from the scanner unit 1, and image recording information of each color from the image processing unit 2. And a printer unit 3 for forming an image on copy paper based on the image data. The scanner section 1 has a lamp 5 for scanning and illuminating the original on the original placing table 4, for example, a fluorescent lamp. The reflected light from the original when illuminated by the fluorescent lamp 5 is reflected by the mirrors 6, 7, 8 and is incident on the imaging lens 9.
The image light is converted to a dichroic prism by the imaging lens 9.
Imaged on 10, for example 3 of red R, green G, blue B
It is split into light of different wavelengths, and a light receiver for each wavelength light, for example, CCD11R for red, CCD11G for green,
Light is incident on the CCD 11B. Each of the CCD 11R, CCD 11G, and CCD 11B converts incident light into a digital signal and outputs the digital signal. The output is subjected to necessary processing in the image processing unit 2, and recording color information of each color, for example, black (hereinafter abbreviated as Bk) , Yellow (Y
), Magenta (abbreviated as Μ), cyan (abbreviated as C)
Is converted into a signal for recording formation of each color.

FIG. 3 shows an example in which four colors of Bk, Y, M and C are formed.
A color image can be formed using only colors. In this case, the number of recording devices can be reduced by one set compared to the example shown in FIG.

Signals from the image processing unit 2 are input to the printer unit 3 and the laser light emitting devices 12Bk, 12C, 12M, and 12Y of the respective colors.
Sent to

In the example of the figure, the printer unit 3 has four sets of recording devices 13Y and 13Y.
M, 13C and 13Bk are arranged side by side. Since the recording devices 13 are made of the same components, the recording device for C will be described for simplicity of description, and other colors will be omitted. In addition, for each color, the same portions are denoted by the same reference numerals, and in order to distinguish the configuration of each color, a reference numeral indicating each color is added to the reference numeral.

The recording device 13C has a photoconductor 14 outside the laser beam emitting device 12C.
C, for example, having a photosensitive drum.

The photoreceptor 14C is provided similarly to a known copying apparatus such as a charging charger 15C, an exposure by a laser beam emitting device 12C, a developing device 16C, and a transfer charger 17C.

Photoconductor 14C uniformly charged by charging charger 15C
Forms a latent image of a cyan light image by exposure with a laser beam emitting device 12C, and develops the latent image with a developing device 16C to form a visible image. The transfer paper supplied from the paper supply unit 19, for example, one of the two paper supply cassettes, by the paper supply roller 18, is sent to the transfer belt 21 with the leading ends thereof aligned by the registration rollers 20 and at the same timing. The transfer paper conveyed by the transfer belt 21 is a photosensitive member 14Bk, 14
C, 14M, and 14Y sequentially transferred to the transfer chargers 17Bk, 17C, 17M,
A visible image is transferred under the action of 17Y. The transferred transfer paper is fixed by a fixing roller 22 and discharged by a discharge roller 23.

The transfer paper can be accurately conveyed at the speed of the transfer belt by being electrostatically attracted to the transfer belt 21.

The transfer belt 21 is supported by a belt drive roller 24 and a driven roller 25, moves in the direction A to convey the transfer paper, and the cleaning unit 26 removes the toner adhering to the belt.

FIG. 4 is a block diagram schematically showing a control system of the digital copying machine shown in FIG.

The system controller 30 controls each module of the scanner 1, the image processing unit 2, and the printer 3 in FIG. The control contents include display control of the operation panel 31,
And key input processing, a start signal to the scanner 1 and the printer 3 according to the mode set on the operation panel 31,
Transmission of a scaling ratio designation signal, transmission of an image processing mode designation signal (color conversion, masking, tricking, mirroring, etc.) to the image processing unit 2, an abnormal signal from each module, an operation status signal (Wait, Ready, Busy, Stop etc.) to control the whole system.

The scanner 32 in FIG. 3 is a system controller.
The original is scanned at a scanning speed corresponding to the magnification specified by the start signal from 30, the original image is read by a reading element such as a CCD, and sent to the image processor 33 as R, G, B image data.

The image processor 33 sends R, G,
The B image data is converted to Bk, Y, M, and C recording color density data, and corrections such as γ correction, UCR (under color removal), color correction, error diffusion correction, and dot gain correction are performed. The completed recording color density data is applied to a delay memory circuit 34 for outputting the data with a shift of the photosensitive drum interval of the printer 3. The correction processing circuit 33P shown in FIG. 1 or FIG. 2 is included in the image processor 33, and Bk, Y,
The above-described error diffusion correction and dot gain correction are performed on the M and C recording color density data.

The printer control circuit 36 modulates the laser beam emitting device according to the Bk, Y, M and C image data read from the delay memory circuit 34, and obtains a copy image on a transfer paper by an electrophotographic process.

Next, the setting of the dot gain by the system controller 30 and the image processor 33 will be described. In this embodiment, a predetermined test pattern is registered in the image processor 33, and its record is created by a copying machine.
The image is read by a copying machine and the dot gain is set.

9a and 9b show examples of test patterns. No.
In FIGS. 9A and 9B, each rectangular area indicates one pixel area, and a pixel painted out black is blackened (dot recording). The patterns are a1 and a2, b1 and b
As shown in FIG. 2, two patterns of the same letter of the alphabet form one set. Each pattern is a pixel pattern shown in FIG. 8 and corresponds to the same as the alphabet. One with a number (a1 to p1)
Is a pattern in which the target pixel is not blackened, and a pattern having a number of 2 (a2 to p2) is a pattern in which the target pixel is blackened.

Therefore, for example, when the output pattern of the peripheral pixels is the pattern c shown in FIG. 8, the dot gain is c1 in FIG. 9a.
It can be obtained as the difference in the average reflectance of the output pattern of c2. The same applies to other patterns a to p.

The patterns shown in FIGS. 9a and 9b
This is a periodic pattern in which 4 × 3 pixels of 3 pixels in the sub-scanning direction are used as a unit. This is because only the area of 3 × 2 pixels is blackened in the pattern indicated by p2 which is the maximum of the pattern, and a blank area is formed between adjacent units, so that the area is not affected by the adjacent area. It is. For this reason, an appropriate evaluation of the dot gain can be performed. Dot gain Gx of output pattern X (X is a to p) of peripheral element
Is given by the following equation, where the average reflectance of the patterns X1 and X2 is Rx ₁ and Rx ₂ respectively.

Gx = (Rx ₂ −Rx ₁ ) × (4 × 3) (4) 4 × 3 on the right side is the number of pixels constituting the unit of the test pattern. Also, it is assumed that the reflectance is standardized to be 1 at the background portion of the recording paper and 0 at the solid black portion.

As described above, the dot diameter r is Since the influence of the adjacent pixels in the oblique direction can be neglected when it is close to, only the dot gain of the patterns a, b, c, and d of the pattern shown in FIG. Further, when the dots can always be regarded as circular with the same diameter, it is possible to evaluate the dot gain in another pattern from only the dot gain of the pattern a shown in FIG.

If the dot gain of the pattern a is Ga, the dot diameter r
The relationship is Once the dot diameter is determined, FIGS. 7a, 7b, 7c and
Since the overlapping area of the dots shown in FIG. 7D is determined, the parameters A, B, C, and D shown in Table 1 can be calculated.

Therefore, the dot gain for each output pattern can be obtained from these parameter values.

FIG. 5 shows a dot gain setting processing operation executed by the system controller 30 and the image processor 33. In the following, reference is made to FIG.

The mode for setting the dot gain is selected by the operator through the operation panel 31 (step 1: hereinafter, the word "step" is omitted in parentheses). At this time, when the print key of the operation panel is pressed (2), the system The image processor 33 outputs the test data via the controller 30, and the printer 3 prints out the test pattern image (3). At this time, the scanner 1 is controlled by the system controller 30 so as not to perform the reading operation.

As the test patterns, 16 patterns shown in FIGS. 9a and 9b are output in a distribution as shown in FIG. 10 and printed out. The recorded image of each pattern such as a ₁ and a ₂ is read by the scanner 1 later, but it is easy to match the pattern area with the reading position by the scanner 1.
Also, in order to average out the variation in the sensitivity of the output system and the reading system, consideration is given to allocating an area of 1 cm × 1 cm or more per pattern. At this time, if the test pattern image is not output normally due to a paper jam or the like, a warning is issued to the operator, and the abnormality is processed and re-executed (4, 5).

If the test pattern data is generated and there is no error in the recording, the system controller 30 enters the reading and analyzing mode of the test pattern recorded image (6). In addition,
At this time, if the dot gain setting mode is turned off, the normal copy state is returned (7).

Next, when the operator places the test pattern recorded image on the document table 4 of the scanner 1 and presses the print key (8), the CPU in the system controller 30 reads this and drives the scanner 1 to drive the test pattern. Each pattern image of the recorded image is read, and the dot gain of each pattern is calculated by the above equations (6) and (7) (9, 10). At this time, the printer 3 is controlled by the system controller 30 so as not to perform image formation.

The read data of the test pattern recorded image is calculated as a representative value of each pattern, the average value of a plurality of pixels in each area is calculated, stored, and used for the next analysis. At the time of analysis, input data is also checked. The representative value of each pattern is compared with a predetermined value, and when it is within a predetermined range, the dot data is calculated assuming that the input data is reasonable. For example, when the input data is unreasonable data and the representative value of each pattern is all 0, it is conceivable that the test pattern image output is placed upside down on the document table.
If the representative value is out of the predetermined range in many cases even if not all 0s, the output of the test pattern image may be erroneous in the mounting direction such as up and down, length and width, or may be greatly deviated from the reference position.

If a failure is detected by input data check in this way, a warning is issued to the operator as an analysis error,
Re-read the test pattern image (11, 12). Here, when the dot gain setting mode is turned off, the dot gain setting process is stopped, and the process returns to the normal copy standby mode (7).

On the other hand, if the calculation of the dot gain is completed normally in step 11, the dot gain evaluation circuit 140 (first
(FIG. 2) or the multi-value / error calculation circuit 160 (FIG. 2) is set with the calculated dot gain, the dot gain setting mode is released, and the copy standby mode is restored (13, 14). The dot gain is stored in a non-volatile memory, and is used as initial data when the system is started next time.

9a and 9b are given as examples of test patterns, but as can be seen from these drawings, patterns a2, b1,
c1, d1, and e1 are substantially equivalent patterns. Also, d2 and
j1, c2, h1, etc. are also equivalent patterns. The test pattern may be omitted such that any one of the equivalent patterns is used.

In FIGS. 9a and 9b, there are 16 sets and 32 sets of patterns because of the example of the binarized pattern. However, in the case of the multi-valued pattern, for example, up to ⁴
= 256 sets of patterns are required.

In the case of quaternary conversion, three types of dot gains are required for one set, so that the number of patterns to be analyzed is 256 × 3 = 768. When there are many patterns as described above, the system may be controlled so that the test pattern image is divided into a plurality of sheets and output, and the plurality of test pattern images are sequentially read to calculate the dot gain.

In the above embodiment, the test pattern image was input using the scanner 1 for reading the original. However, if a dedicated sensor is provided, the operator does not need to perform the operation of reading the test pattern image. There is an advantage that the resetting process can be performed as soon as the time of setting on the table, and the setting error of the test pattern image does not occur. The position of the dedicated sensor is preferably provided between the fixing roller 22 and the paper discharge roller 23 shown in FIG. 3 in order to read the test pattern image after the image forming process is completed.

In addition, although the accuracy of dot gain correction is reduced, a configuration in which a test pattern image is formed and read on the transfer belt 21 without feeding paper can be used without using paper. The position of the sensor at this time may be provided between the transfer charger 17Y and the cleaning unit 26 along the transfer belt.

Further, if the configuration is such that the potential of the electrostatic latent image on the photoconductor drums 15Y, 15M, 15C, and 15Bk is measured and analyzed, the toner is not consumed.

As described above, the binarization of the test pattern has been described as an example, but the present invention can be applied to the case of multi-value processing such as ternary and quaternary. Although a color laser printer using an electrophotographic process has been described as an example of the recording apparatus, a monochrome printer may be used. The image forming process can be applied to an ink jet system, a thermal transfer system, or the like.

In this embodiment, when determining the image recording level, the evaluation is performed based on the output patterns of four adjacent pixels, but an output pattern other than four pixels may be used. In this case, only the number of output patterns changes.

However, in an actual printer, the influence of adjacent four pixels is particularly strong, and a sufficient effect for improving halftone reproduction characteristics can be obtained only by considering these pixels.

The overlapping areas B, C, and D of the dots required for the evaluation when determining the recording pixel level can be calculated if the dot diameter r is determined. In particular, when the dots circumscribe the rectangular area of the pixel, Table 3 shows the dot gain at this time.

In this case, since there is no overlap with the adjacent pixels in the oblique direction, only the adjacent pixels in the main scanning direction and the sub-scanning direction need to be considered. That is, the dot gain can be evaluated from the output patterns of the two pixels shown in FIGS. 6B and 6B. The dot diameter r is Is also B >> C ≧ D, and
A sufficient effect for improving the halftone reproduction characteristic can be obtained only by considering the output pattern of the pixels.

[Effects of the Invention] As described above, according to the present invention, in addition to the image data conversion error correction by the error diffusion method, the overlap area between a pixel whose print output gradation level has already been determined and a plurality of pixels adjacent to the pixel is also determined. By performing correction in accordance therewith, halftone reproduction characteristics can be improved.

In addition, by arbitrarily resetting or adjusting the error data, it is possible to obtain an output image that is always constant and has good reproducibility.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram showing a configuration of a gradation recording data correction processing circuit 33P which is a main part of one embodiment of the present invention. FIG. 2 is a block diagram showing a modification of the gradation recording data correction processing circuit 33P. FIG. 3 is a side view showing a mechanism of a digital color copying machine according to one embodiment of the present invention. FIG. 4 is a block diagram showing an outline of the configuration of a recording control electric system of the digital color copying machine shown in FIG. 3. The correction processing circuit 33P shown in FIG. 1 or 2 is an image processor shown in FIG. Included in 33. FIG. 5 is a flowchart showing a dot gain setting processing operation of the system controller 30 and the image processor 33 shown in FIG. FIG. 6a is a plan view showing the pixel division on the image. FIG. 6b is a plan view showing the correlation between the pixel division on the image and the recording dots. FIG. 7a is a plan view showing an isolated recording mode of only one pixel. FIG. 7b is a plan view showing a recording mode of two vertically adjacent pixels. FIG. 7c is a plan view showing a recording mode of two pixels adjacent in the oblique direction. FIG. 7d is a plan view showing a recording mode of four pixels adjacent to each other. FIG. 8 is a plan view showing a recording pattern of a total of five pixels including a target pixel and four pixels adjacent thereto. 9a and 9b are plan views showing a test pattern image for dot gain setting generated by the image processor 33 shown in FIG. FIG. 10 is a plan view showing a recording position distribution of the test pattern image shown in FIGS. 9a and 9b. FIG. 11 is a block diagram showing a configuration of a conventional gradation processing data correction processing circuit by an error diffusion method. FIG. 12 is a plan view showing a ROM or a RAM in place of the multi-level circuit and the error calculation circuit shown in FIG. 1: Scanner, 2: Image processing unit 3: Printer, 4: Document table 5: Lamp, 6-8: Mirror 9: Image forming lens, 10: Dichroic prism 11R, 11B, 11G: Receiver 12Y, 12M, 12C , 12Bk: Laser beam emitting devices 13Y, 13M, 13C, 13Bk: Recording devices 14Y, 14M, 14C, 14Bk: Photoconductors 15Y, 15M, 15C, 15Bk: Charger 16Y, 16M, 16C, 16Bk: Developing devices 17Y, 17M , 17C, 17Bk: transfer charger 18: paper feed roller, 19: paper feed unit 20: registration roller, 21: transfer belt 22: fixing roller, 23: paper discharge roller 24: belt drive roller, 25: driven roller 26: cleaning Unit, 30: System controller 31: Operation panel, 33: Image processor 34: Delay memory circuit, 35: Writing system control circuit 36: Printer control circuit, 100: Error memory circuit 110: Correction operation circuit, 120: Multi-value Circuit 130: error calculation circuit, 140: dot gain evaluation circuit 150: output data memory, 160: multi-value / error calculation circuit

Continuation of the front page (56) References JP-A-3-34680 (JP, A) "MULTIPLE ERROR CORRECTION ALGORITHM FOR HALFTONE, CON TINUOS TONE AND TE XT REPRODUCTION, IBM Technology Corporation. , March 1981, Vol. 23, No. 10, p. 4433-4435, "HALFTONING TECHN IQUESING USING ERROR CORRECTION", Processe ding of the SID, 1986, Vol. 24/4, p. 305-308 (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N 1/40-1/409 H04N 1/46 H04N 1/60

Claims (2)

(57) [Claims] 1. A recording output level of a pixel of interest of an input image,
An image having halftone reproduction processing means for determining based on the input gradation level of the pixel and error data corresponding to the gradation level error between input and output for the pixel for which the recording output gradation level has already been determined. The halftone reproduction processing means, wherein the target pixel and a plurality of pixels adjacent thereto based on the print output gray level of the pixel whose print output gray level has already been determined and the print output level of the target pixel An error output means for generating error data corresponding to an error of a gradation level between input and output with respect to a pixel of interest in accordance with a gradation level corresponding to an overlapping area of the image. Recording device. 2. A recording output level of a target pixel of an input image,
An image having halftone reproduction processing means for determining based on the input gradation level of the pixel and error data corresponding to the gradation level error between input and output for the pixel for which the recording output gradation level has already been determined. The halftone reproduction processing means, wherein the target pixel and a plurality of pixels adjacent thereto based on the print output gray level of the pixel whose print output gray level has already been determined and the print output level of the target pixel Error output means for generating error data corresponding to an error between the input and output of the pixel of interest in accordance with a gray level corresponding to an overlapping area with the target pixel. Forming a test pattern; reading means for reading a test pattern of a recording pixel formed by the test pattern forming means; and reading by the reading means. An image gradation recording apparatus, comprising: error setting means for setting or updating error data in the error output means based on image data of a test pattern.