JP2003259117A - Image processor and processing method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a superior image processor and its processing method which are low-cost and produce a print of high quality by efficiently reducing error memory capacity while avoiding deterioration in image quality. <P>SOLUTION: The image processor which performs multi-valued error diffusion processing wherein error data generated when the value of a pixel of interest of multi-valued image data is quantized to an N value are stored in an error memory and diffused to circumferential pixels according to a specified diffusion coefficient and the value of the pixel of interest to which the error data are diffused is quantized to an N value, limits the number of bits of the error data generated when the value of the pixel of interest of the multi-valued image data is quantized by the multi-valued error diffusion processing according to a recording mode made to correspond to the multi-valued image data and stores it in the error memory, and extends the number of bits of the error data of which the number of bits read out of the error memory is limited according to the recording mode. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention stores error data generated when quantizing a value of a pixel of interest of multivalued image data into an N value, and stores the stored error data in a predetermined diffusion coefficient. The present invention relates to an image processing apparatus and a processing method for performing multi-valued error diffusion processing in which the value of a pixel of interest diffused to peripheral pixels according to the above is quantized into an N value.

[0002]

2. Description of the Related Art In recent years, personal computers (P
C), copiers, word processors, and other OA devices have become widespread, and image formation (recording) of these OA devices
2. Description of the Related Art As a type of apparatus, an apparatus that performs digital image recording by an inkjet method is rapidly developing and becoming popular.
In particular, the colorization of OA equipment is progressing with the sophistication of OA equipment,
Along with these, various color inkjet recording apparatuses have been developed.

Generally, an ink jet recording apparatus is provided with a recording means (print head), a carriage on which an ink tank is mounted, a conveying means for conveying recording paper, and a control means for controlling these. Then, a direction (main scanning direction) orthogonal to the recording paper conveyance direction (sub scanning direction) for the print head that ejects ink droplets from the plurality of ejection ports.
The serial scanning is performed on the one hand, while the non-printing is intermittently conveyed by an amount equal to the print width. Furthermore, in the case of a color-compatible inkjet recording apparatus, a color image is formed by superposing ink droplets ejected from print heads of a plurality of colors.

In an ink jet recording apparatus, as a method of ejecting ink, a heating element (electrical / thermal energy converter) is provided in the vicinity of the ejection port, and an electric signal is applied to the heating element to locally heat the ink. Then, a thermal method of causing a pressure change and ejecting the ink from an ejection port, or a piezo method of applying a mechanical pressure to the ink to eject the ink by using an electric / pressure conversion means such as a piezo element is used. ing. Generally, in the former thermal method, it is easy to increase the density of nozzles and the head can be constructed at low cost, but on the other hand, heat is used, so that ink or the head is easily deteriorated. On the other hand, the latter piezo method is characterized by excellent ejection controllability, high ink flexibility, and semi-permanent head life.

In these recording methods, characters and figures are recorded by ejecting ink as minute droplets from a discharge port onto a recording medium in accordance with a recording signal.
Since it is non-impact, it has low noise, low running cost, easy downsizing of the device, and easy colorization.
It is widely used as an image forming (recording) means in a recording device such as a copying machine, a printer, a facsimile, etc., which is used together with a computer, a word processor or the like, or used alone.

16 and 17 are block diagrams showing a schematic configuration of a controller unit and an engine unit of the ink jet recording apparatus.

First, the function and general operation of the controller section will be described. In FIG. 16, the CPU 1601 is a USB interface 1604 or IEEE 1394.
Host PC 1607 via interface 1605
A ROM 1609 connected to (1607) and storing a control program, and an EEPROM storing an updatable control program, a processing program, and various constant data.
The RAM 1608 for storing command signals and image information received from the M1610 and the host PC 1606 is accessed, and the recording operation is controlled based on the information stored in these memories. The instruction information input from the keys of the operation panel 1612 is the operation panel interface 161.
1 is transmitted to the CPU 1601 and the CPU 16
Similarly, in response to a command from 01, the LED of the operation panel 1612 is turned on or the L
CD display is controlled. The expansion interface 1615 is an interface for expanding functions by connecting an expansion card such as a LAN controller or HDD. The image information is converted into dot data of each ink color by the image data processing unit 1613 and output to the engine unit.
In addition, the transmission and reception of various commands and status information between the controller unit and the engine unit are similarly performed by the image data processing unit 1.
Via 613.

Next, the function and operation outline of the engine section will be described. In FIG. 17, the engine unit is connected to the controller unit via a band memory control block 1712. A CPU 1701 is a ROM 1703 storing a control program, an EEPRO storing an updatable control program, a processing program, various constant data, and the like.
The M1704 and the RAM 1702 for storing the command signal and the image information received from the controller unit are accessed, and the recording operation is controlled based on the information stored in these memories. The carriage motor 17 is output via the output port 1705 and the carriage motor control circuit 1707.
By operating 09, the carriage 1711 is moved, and by operating the paper transport motor 1708 via the output port 1705 and the paper transport motor control circuit 1706, the paper transport mechanism 1710 such as transport rollers is operated. Further, the CPU 1701 uses the band memory control unit 171 based on various information stored in the RAM 1702.
2 and the print head controller 1714 are controlled to drive the print head 1715, whereby a desired image can be recorded on the recording medium.

A logic drive voltage Vcc for operating the CPU and various control circuits is supplied from a power supply circuit (not shown).
(Eg 3.3V), various motor drive voltages Vm (eg 24V), heat voltage Vh (eg 12V) for driving the print head, etc. are output.

In the conventional ink jet recording method,
It was necessary to use a special coated paper with an ink absorption layer to obtain a high-color color image without ink bleeding, but in recent years, due to improvements in ink, etc., it is commonly used in large quantities in printers and copiers. A method that has printability on paper has also been put into practical use. Furthermore, it is desired to support various recording media such as OHP sheets, cloths, plastics, and sheets, and in order to meet such demands, recording media (recording media) having different ink absorption characteristics are required as necessary. The development and commercialization of a recording device capable of performing the best recording regardless of the type of the recording medium when the selection is made are underway. Regarding the size of the recording medium, a large size is required for woven cloth such as posters and clothes for advertisement. It can be said that such an ink jet recording apparatus is in high demand in a wide range of fields as an excellent recording means, is required to be provided with an image of higher quality, and is required to be further increased in speed.

Generally, the color ink jet recording method is cyan (Cy), magenta (Mg), yellow (Ye).
3 color inks are used and black (B
Color recording is realized by using the four color inks added with k). In such a color inkjet recording apparatus, different from a monochrome inkjet recording apparatus that prints only characters, various elements such as color developability, gradation and uniformity are required for recording a color image image.

Further, in the ink jet recording apparatus, in order to form a higher quality natural image with more gradations, conventional (Cy), magenta (Mg), yellow (Ye),
In addition to 4 colors of black (Bk), 3 colors of light Cy, light Mg, and light Ye with low ink density are used, and many colors are realized by using 7-color ink. Has been done.

However, the quality of the recorded image largely depends on the performance of the print head alone. That is, slight differences between the nozzles that occur during the printhead manufacturing process, such as the shape of the discharge ports of the printhead and variations in the electrical / thermal converter (discharge heater), cause the amount of ink discharged and the direction of the discharge direction. Affect
This causes deterioration in image quality as density unevenness of the finally formed recorded image. As a result, there are "white" portions that cannot periodically satisfy the area factor of 100% with respect to the head main scanning direction, conversely dots are overlapped more than necessary, or white streaks occur. Becomes These phenomena are usually perceived by the human eye as uneven density.

Therefore, a method called a multi-pass printing method has been proposed as a countermeasure against these density unevenness. For simplification, a case where a single ink color head having 8 nozzles is used will be described as an example.

First, a description will be given of a case where two-pass printing is realized by using a fixed mask method in which pass data is generated by thinning out print data using an even column / odd column pattern or a zigzag / inverse zigzag pattern. Printing is completed by printing a staggered pattern in the first scan, feeding the paper by half the printing width (4 dot width), and then printing an inverse zigzag pattern in the second scan. That is, the paper feed in units of 4 dots and the recording of the staggered / reverse zigzag pattern are alternately performed in order to complete the recording area in units of 4 dots for each scan.

Next, two-pass printing is realized by adopting a table reference method in which pass data is generated by thinning out print data using a random mask pattern in which print dots and non-print dots are randomly arranged. The case will be described.

FIG. 14 is a diagram showing a configuration example of a mask table for each printing scan. Table areas A and B shown in FIG. 14 are complementary mask tables used in the first pass and the second pass, respectively. Table is 1bit / dot
Here, 0 indicates that it is a mask target, and 1 indicates that it is a non-mask target. The mask tables A and B are tables each having a size corresponding to 12 pixels in the main scanning direction × 4 pixels in the sub scanning direction, and these are repeatedly developed in each direction to be used as mask data. The number of nozzles included in the print head is 8, and the number of pixels corresponding to the paper transport amount in 2-pass printing is 8/2 = 4.
And B in the sub-scanning direction.

FIG. 15 is a diagram for explaining the state of the print scan using the mask table shown in FIG. 8
With respect to the data of 8 lines corresponding to each nozzle, A and B are applied as mask patterns every 4 lines. In each printing scan, mask processing (replacement of printing dots with non-printing dots) of image data is executed using the stored mask table, and pass data is generated and output. Specifically, the logical product of the image data and the mask data is used to output the image data as it is when the mask data is 1, and replace the image data with 0 when the mask data is 0. It is realized by All image areas are always masked in the order of A and B by scanning twice to generate print data. Here, the mask OFF (1) ratios of A and B are equal to 5 each.
It is about 0%.

In this way, one line is divided into two different lines.
By recording using one nozzle, it is possible to form a high-quality image with suppressed density unevenness. In addition, the multi-pass printing method has the effect of suppressing bleeding (bleeding) by printing while drying the ink, and it also suppresses the temperature rise of the print head that causes ejection failure because it reduces the number of print dots for each scan. The effect of doing, etc. can be achieved at the same time. Although the main scanning direction has been described here, it is possible to further improve the image quality by thinning out and recording dots that are continuous in the sub scanning direction. Further, when it is desired to realize image formation in the sub-scanning direction at a resolution higher than the nozzle resolution, this thinning-out recording in the sub-scanning direction is an essential process.

As a method of generating pass data for each scan, pass data is generated by thinning out print data using a random mask pattern in which print dots and non-print dots are randomly arranged as described above. In addition to the method (called the table reference method) and the method of generating pass data by thinning the print data using even-numbered / odd-numbered patterns and staggered / inverse-staggered patterns (called the fixed mask method), There is known a method of generating pass data by performing thinning processing by focusing on dots (referred to as a data mask method), a method of using these in combination, or the like.

Next, the image data processing in the image data processing section 1613 in the controller section shown in FIG. 16 will be described.

FIG. 18 is a diagram showing the flow of image data processing in the image data processing unit 1613. For example, RGB multi-valued image data received from the host PC 1606 is subjected to an ink color (for example, C
y, Mg, Ye, Bk) multivalued image data,
Then, it is converted into binary data for each ink color by a quantization process 1802. In this way, the multi-valued image data is converted into a level (binary here) that can be output by the engine unit (print head).

The error diffusion method and the dither method are widely known as the methods of this quantization processing. In the error diffusion method, a diffusion coefficient is assigned to peripheral pixels for a target pixel, and the quantization error generated in the target pixel is distributed to the peripheral pixels according to the diffusion coefficient. As a result, the density of the entire image is preserved, and good pseudo gradation expression is possible. On the other hand, in the dither method, a dither matrix made up of matrix-like threshold values is prepared, and one-to-one pixel comparison between each threshold value and each pixel of input data is performed to determine ON / OFF. Generally, in the dither method, the image quality tends to be lower than that in an image to which the error diffusion method is applied, but the error diffusion method cannot proceed to the processing of the next pixel until the error propagates, and thus high-speed processing is difficult.

A basic binarization processing method of the error diffusion method will be specifically described. FIG. 5 is an example of a diffusion coefficient matrix when processing from the left to the right of the raster as shown by the arrow.
This diffusion coefficient matrix indicates the rate of error propagation for each pixel. In the error diffusion method, a threshold value is compared with a pixel value to determine ON / OFF of a dot, and at the same time, the error is calculated and propagated to surrounding pixel values.
Here, the error is represented by the difference between the pixel value and the evaluation value. If the evaluation value is ON for a 10-bit (1024 gradation) pixel value, it is 1023 if it is ON, 0 if it is OFF, and if the threshold value is fixed at 512, the determination result is dot ON if the pixel value is 722,
The error is 301. In this example, the following two pixels A and B on the same line and C, D, E, F, and G on the lower line
The error is diffused to the 5 pixels. By thus propagating the error, the density of the image data is saved.

Further, in a high-resolution binary output device such as a recent ink jet recording device, an error diffusion method (multi-valued error diffusion method) of expressing with multiple gradations is used in order to express a smoother halftone image. It's being used. The quantization method combining the multi-valued error diffusion method and the density pattern method will be described below with reference to specific examples.

First, the multi-valued image data is subjected to a quinarization process by the multi-valued error diffusion method. Multi-valued error diffusion method 5
In the binarization, four threshold values are provided, and the output value of 0 to 4 is determined by comparison with the pixel value. The error generated there is propagated to surrounding pixel values according to the diffusion matrix in the same manner as in binarization. Further, the 5-valued data is developed into binary data by using the density pattern method. In particular,
This is a halftone dot expansion process for five-valued data according to a 2 × 2 halftone dot matrix. Here, the resolution (input resolution) of the multi-valued image data is 600 ppi × 600 ppi (ppi is p
ixel / inch) and the final recording resolution (output resolution) by the print head is 1200 dpi x 1200 dpi.
i (dpi is dot / inch). With this method, it is possible to express 5 gradations by expanding each pixel that can be expressed only with 2 gradations of dot ON / OFF, and the error diffusion processing, which is difficult to speed up, is lower than the output resolution.
By carrying out at 0 ppi × 600 ppi, the error diffusion processing time can be reduced to ¼.

In an actual ink jet recording apparatus, the relationship between the input resolution and the output resolution is set in association with the print mode and the like, and the optimum output gradation by the multi-value error diffusion method is set according to the halftone dot matrix size according to the relationship. A method has been proposed in which a number N is selected and a quantization process is executed. For example, in high speed mode, input resolution is 300 ppi x
Output resolution 1 using a 4 × 4 halftone dot matrix after performing octalization by the multi-valued error diffusion method for 300 ppi
Print dot data of 200 dpi x 1200 dpi is generated, and input resolution is 600 ppi x 60 in high definition mode.
After 0-ppi is quaternarized by the multi-valued error diffusion method, an output resolution of 120 is obtained by using a 2 × 2 halftone dot matrix.
Print dot data of 0 dpi × 1200 dpi is generated. As a result, it becomes possible to perform a quantization process having an excellent balance between high speed processing and high quality processing depending on the purpose.

Further, an ink jet recording apparatus for accelerating monochrome image formation is also put into practical use by forming only Bk so that the amount of ejected ink droplets is larger than that of other ink colors and performing image formation at a lower resolution only for Bk. ing. In such a device, for example, if the input resolution is 300 ppi × 300 ppi for all ink colors, only Bk has an output resolution of 600 dpi × 600 dpi, and the other ink colors have an output resolution of 1200 dpi × 1200 dpi,
In Bk, after performing 4 value conversion by the multi-value error diffusion method, 2 ×
It is developed using a halftone dot matrix of 2 and is octalized by the multivalued error diffusion method for other ink colors, and then 4 × 4.
It is necessary to perform processing such as expansion using the halftone dot matrix of.

[0029]

However, there are the following problems in the quantization processing using the error diffusion method.
Hereinafter, problems in quantization by the error diffusion method will be described in detail with reference to specific examples.

First, the cumulative error data generated for each pixel will be described. Input image data is 10 bits (0 to 1
023). The threshold is fixed at 512 and the judgment result is O
If a diffusion matrix in which N is 1023, OFF is 0, and a fixed sum of coefficients is 1, the generated error data is -511 to +511, and the accumulated diffusion error data is -511 as well. To +511, corrected pixel data added with accumulated error data is -511 to +153.
It is represented by 4.

In the error diffusion method, in order to propagate the error to pixels other than the same line as the processed pixel, it is necessary to provide an error memory for storing the diffusion error to the next line for each line processing. Specifically, a memory having a capacity (bit number) obtained by multiplying the number of line pixels by 10 bits is required for each ink color. For example, when the error is diffused to the following one line in addition to the processing line, 6 colors × 60 inches × 1200 dpi × 10 bits × 1 line are required to support 6 ink colors, a resolution of 1200 dpi, and a maximum print width of 60 inches. = A large capacity memory of 4.1 Mbits is required.

Therefore, the increase of the error memory capacity in the large-format ink jet recording apparatus corresponding to the wide print width such as 42 inch or more than 60 inch is one of the major factors causing the cost increase.

On the other hand, the following methods for reducing the error memory have been proposed and implemented. That is, the cumulative error bit number for each pixel is reduced by limiting the diffusion error of the processed pixel to the next line. For example,
By limiting the diffusion error to the next line to a multiple of 16, the diffusion error can be reduced to 6 bits, and by limiting it to a multiple of 64, the diffusion error can be reduced to 4 bits. 40%, 60% of the reduction effect.

However, the above measures have the following problems. As described above, if the restrictions placed on the diffusion method in the quantization processing by the error diffusion method are increased, the diffusion of error data that occurs for each pixel is limited, and a unique unpleasant texture occurs. This may cause deterioration in image quality such as deterioration of image smoothness. In particular, the number of output gradations N of the multi-valued error diffusion method is associated with the recording mode as described above.
In a system that selectively controls, when N is large and the generated error is relatively small, the constraint of the error diffusion state is apparently further increased, and the problem of deterioration of image quality is particularly remarkable. There was a problem such as.

The present invention has been made in view of the above problems, and an object thereof is to efficiently reduce an error memory capacity while avoiding deterioration of image quality, and to achieve both low cost and high quality printing. An object of the present invention is to provide an excellent image processing device and its processing method.

[0036]

In order to achieve the above object, the present invention stores error data generated when quantizing a value of a pixel of interest of multivalued image data into an N value in an error memory, In the image processing device, which performs a multi-valued error diffusion process of diffusing the stored error data to peripheral pixels according to a predetermined diffusion coefficient and quantizing the value of the pixel of interest to which the error data is diffused into an N value, In the error memory, the number of bits of error data generated when quantizing the value of the pixel of interest of the multivalued image data by the multivalued error diffusion process is limited according to the recording mode associated with the multivalued image data. An error bit number limiting means for storing and an error bit number expanding means for expanding the bit number of the error data whose number of bits read from the error memory is limited according to the recording mode are characterized.

In order to achieve the above object, the present invention stores error data generated when quantizing a value of a pixel of interest of multivalued image data into an N value in an error memory and stores the stored error. In the image processing apparatus, which performs multi-valued error diffusion processing of diffusing data into peripheral pixels according to a predetermined diffusion coefficient and quantizing the value of the pixel of interest into which the error data has been diffused into an N value, An error bit number limiting means for limiting the number of bits of error data generated when quantizing the value of the pixel of interest of the multi-valued image data by processing according to the ink color information to be processed and storing it in an error memory, An error bit number expansion means for expanding the number of bits of the error data read from the error memory and having a limited number of bits according to the ink color information.

Further, in order to achieve the above object, the present invention stores error data generated when quantizing a value of a pixel of interest of multi-valued image data into an N value, and stores the stored error data. A processing method of an image processing apparatus for performing multi-valued error diffusion processing of diffusing data to peripheral pixels according to a predetermined diffusion coefficient, and quantizing the value of a pixel of interest to which the error data is diffused into an N value. An error memory that limits the number of bits of error data generated when quantizing a value of a pixel of interest of multivalued image data by the multivalued error diffusion processing according to a recording mode associated with the multivalued image data And an error bit number expanding step of expanding the bit number of the error data whose number of bits read from the error memory is limited according to the recording mode.

Further, in order to achieve the above object, the present invention stores error data generated when quantizing a value of a pixel of interest of multi-valued image data into an N value in an error memory and stores the stored error. A processing method of an image processing apparatus for performing multi-valued error diffusion processing of diffusing data to peripheral pixels according to a predetermined diffusion coefficient, and quantizing the value of a pixel of interest to which the error data is diffused into an N value. An error bit for limiting the number of bits of error data generated when quantizing a value of a pixel of interest of multivalued image data by the multivalued error diffusion processing according to ink color information to be processed and storing it in an error memory It is characterized by including a number limiting step and an error bit number expanding step of expanding the bit number of the error data whose number of bits read from the error memory is limited according to the ink color information.

[0040]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings.

[First Embodiment] In the first embodiment, multi-valued image data input from a host PC is quantized into N values by a multi-valued error diffusion method, and the quantized image data is density pattern method. Limits the number of bits of the quantization error generated by the multi-value error diffusion processing when binarizing (halftone processing) and outputting to the engine in accordance with the recording mode associated with the input multi-valued image data. The capacity of the error memory is reduced without degrading the image quality.

FIG. 3 is a diagram showing the construction of the recording section in the ink jet recording apparatus. In FIG. 3, reference numeral 301 denotes a print head, which includes black (Bk) and cyan (C
y), magenta (Mg), and yellow (Ye), four color inks are filled in each of the ink tanks, and four multi-heads are formed corresponding to each of the ink tanks. The number of nozzles for each color is 1280
It is a nozzle. A carriage 302 supports the print head 301 and moves these together with recording.
Incidentally, the carriage 302 is at the home position position HP during standby such as in a non-recording state.

Reference numeral 303 denotes a paper feed roller, which rotates together with an auxiliary roller (not shown) while pressing the recording paper 305, and feeds the recording paper 305 in the Y direction at any time. A paper feed roller 304 serves to feed the recording paper 305 and hold the recording paper 305 similarly to the paper feed roller 303 and the auxiliary roller. Here, the print head 301 is B
Each of the four colors of k, Cy, Mg, and Ye has 1280 nozzles arranged in the paper feed direction.

Here, a basic recording operation in the recording section having the above-mentioned structure will be described. The carriage 3 waiting at the home position HP by the recording start command
Reference numeral 02 denotes a plurality of nozzles of the print head 301 while moving in the X direction to eject ink onto the recording paper 305 according to recording data to perform recording. After that, when printing of the print data up to the edge of the print sheet 305 is completed, the carriage 302
Returns to the original home position HP. Then, the paper feed roller 303 rotates in the arrow direction to feed the paper in the Y direction by a predetermined width, and the carriage 302 again ejects ink while moving in the X direction to start recording. Data recording is realized by repeating such a scanning operation and a paper feeding operation.

The ink jet recording apparatus of this embodiment converts the input image information into an interface for exchanging image information and various control information with the host PC, and ON / OFF data of dots for each ink color. A controller including an image data processing unit (see FIG. 1).
6) and an engine (FIG. 17) that forms the image by controlling the print head 301 while carrying the recording paper and driving the carriage.

Next, the image data processing unit in the controller will be described in detail.

FIG. 4 is a block diagram showing the arrangement of the image data processing unit in the controller. The image data processing unit includes a color conversion unit 401, a multi-valued error diffusion processing unit 402, and a halftone dot processing unit 403, and converts the image information input by the color conversion unit 401 into multi-valued image data for each ink color. After that, the multi-valued error diffusion processing unit 402 quantizes the multi-valued image data into the number of output gradations according to the recording mode by the multi-valued error diffusion method, and the halftone dot processing unit 403 uses the density pattern method for each ink. Generate color dot data. The number of output gradations N by the multi-valued error diffusion method is variable, and these are adaptively combined to develop into a dot pattern. The generated dot data is output to the engine.

First, a basic error diffusion method will be described. Here, for the sake of simplicity, the description will be made assuming that the number of output gradations is 2. FIG. 5 is a diagram showing a diffusion coefficient matrix in the error diffusion method. In this example, as indicated by the arrow in the figure,
An example of the diffusion coefficient matrix when processing from the left to the right of the raster is shown. In the error diffusion method, a threshold value and a pixel value are compared to determine whether a dot is ON or OFF, and at the same time,
The error is calculated and propagated to surrounding pixel values. Here, the error is represented by the difference between the pixel value and the evaluation value. Evaluation value is 1
If the 0-bit pixel value is ON, it is 1023, and if it is OFF, it is 0. The threshold value is 512 (fixed). For example,
When the pixel value is 722, the determination result is dot ON and the error is 301. In this example, the following A on the same line,
The error is diffused into 2 pixels of B and 5 pixels of C, D, E, F and G on the lower line. Further, the diffusion error to the line number (L + 1) generated by the quantization processing of the pixel of the line number L is temporarily stored in the error memory as the diffusion error for each pixel.

Next, the quantization processing in the image data processing section will be described in detail. The quantization processing in this embodiment is realized by combining the N-valued processing based on the multi-valued error diffusion method and the binarized processing (halftone dot processing) using the density pattern method. In this embodiment, adaptive quantization processing is selectively executed according to the three recording modes.

FIG. 6 is a diagram showing the relationship between the recording mode and the quantization processing in this embodiment. As shown in FIG. 6, in the “high speed mode”, the input resolution is 300 ppi × 300.
4 × 4 after converting the ppi to eight by the error diffusion method
Output using the halftone dot matrix of
Dot data of 200 dpi × 1200 dpi is obtained.
In "standard mode", input resolution 600 ppi x 600
2 × 2 after quaternarization of ppi by error diffusion method
Output using the halftone dot matrix of
Dot data of 200 dpi × 1200 dpi is obtained.
Input resolution 1200 ppi x 1 in "High Definition Mode"
Binarization processing is performed on 200 ppi only by the error diffusion method, and the output resolution is 1200 dpi × 1200 dpi.
Get the dot data of. In this way, by gradually selecting the processing resolution of error diffusion with a heavy processing load for each recording mode, it is possible to perform high-speed processing and high-quality processing in a well-balanced manner. It should be noted that here, the threshold value and the evaluation value with equal intervals are always used.

Next, the bit deletion / insertion control of the diffusion error in the multi-valued error diffusion processing characteristic of the present embodiment will be described in detail.

FIG. 2 is a block diagram showing the arrangement of the next-line error diffusion processing unit for executing the error diffusion processing to the next-line pixels in the multi-valued error diffusion processing unit. The error is diffused to each of the five pixels C, D, E, F, and G located immediately below the processing line, and the accumulated error data from the upper line of each pixel is sent to an error memory (not shown). . In FIG. 2, reference numeral 201 denotes an error bit limiting unit, which selectively cuts out some bits of the input error data, limits the number of bits, and outputs the bits to the subsequent stage. 202-206
Is a C diffusion error calculation unit, a D diffusion error calculation unit, an E diffusion error calculation unit, an F diffusion error calculation unit, and a G diffusion error calculation unit, respectively, and calculates an error diffused to pixels C, D, E, F, and G. And output. Reference numerals 207 to 210 denote a C adder, a D adder, an E adder, and an F adder, which add a diffusion error to the accumulated error up to immediately before and output it. A delay unit 211 delays data by a pixel processing cycle which is a processing time unit. Reference numeral 212 denotes a recording mode corresponding bit limit control unit, which executes a bit number limit control of error data corresponding to each recording mode based on a control signal from the outside.

Next, the basic operation of calculating the accumulated error from the upper line according to the above configuration will be described. Error data corresponding to each pixel is sequentially input for each pixel processing cycle,
Further, in each unit, the pixel processing cycle is processed as a delay unit. Here, the input error is signed 10-bit data and can take values from -511 to +511.

The error bit limiting unit 201 selectively cuts out 7 bits out of the remaining 9 bits while holding the sign of the input error data based on the instruction from the recording mode corresponding bit limiting unit 212. Specifically, the following three ways can be selected. The first is the remaining upper 7 bits after discarding the lower 2 bits. The second is the remaining 7 bits after discarding the most significant 1 bit and the least significant 1 bit. And 3
The first is the remaining lower 7 bits after discarding the upper 2 bits.

The C diffusion error calculation unit 202, the D diffusion error calculation unit 203, the E diffusion error calculation unit 204, the F diffusion error calculation unit 205, and the G diffusion error calculation unit 206 are based on the input 8-bit signed error data. Pixels C, D, E,
8 bits of diffusion error to F and G are calculated. Here, using 7 bits of the absolute value of the error regardless of whether the error is positive or negative,
The diffusion error is calculated by referring to the table of 7-bit input and 7-bit output. First, the diffusion error to the pixel G, which is the output of the G diffusion error calculation unit 206, is determined by the delay unit 211.
After being held (delayed) by the next pixel processing cycle, the F addition unit 210 performs addition processing with the diffusion error to the pixel F which is the output of the F diffusion error calculation unit 205. In the next pixel processing cycle, the E addition unit 209 performs the addition process of the diffusion error to the pixel E, which is the output of the E diffusion error calculation unit 204, and the output of the F addition unit 210. In the next pixel processing cycle, the D addition unit 208 adds the diffusion error to the pixel D, which is the output of the D diffusion error calculation unit 203, and the output of the E addition unit 209. Further, in the next pixel processing cycle, the C addition unit 207 outputs the C diffusion error calculation unit 202 output diffusion error to the pixel C and D.
The addition processing with the output of the addition unit 208 is performed, and the upper line accumulated error that is the accumulated error from the upper line of the pixel of the line located immediately below the processed pixel is obtained.

FIG. 8 is a block diagram showing the arrangement of a processing pixel line error diffusion processing unit for executing error diffusion processing on processing line pixels in the multi-valued error diffusion processing unit. The accumulated error data from the upper line of each pixel, which is generated by the previous line processing and is temporarily stored in an error memory (not shown), is input, and is further positioned on the right side of the same line as the processed pixel. The error is diffused to each of the two pixels A and B to generate final accumulated error data.

In FIG. 8, reference numeral 801 denotes a cumulative error bit expansion unit, which selectively shifts (bit inserts) the cumulative error data from the upper line read from the error memory to expand the number of bits and expand it. The accumulated error data thus generated is output to the subsequent stage. Reference numerals 802 and 803 denote an A diffusion error calculation unit and a B diffusion error calculation unit, respectively, which calculate and output the error diffused to the pixels A and B based on the input error data of the processed pixel. Reference numerals 804 and 805 denote an A addition section and a B addition section, respectively, which add a diffusion error to the accumulated error up to immediately before and output it. Reference numeral 806 denotes a recording mode corresponding bit expansion control unit, which executes the bit number expansion control of the accumulated error data corresponding to the recording mode based on a control signal from the outside.

Next, the basic operation of calculating the final accumulated error with the above configuration will be described. Error data corresponding to each pixel is sequentially input for each pixel processing cycle, and each unit processes the pixel processing cycle as a delay unit. The input accumulated error data is accumulated information of the error diffused from the upper line temporarily stored in the error memory (not shown), and is a signed 8-bit data and takes a value of -127 to +127. On the other hand, the error data of the input processed pixel is 10-bit data with a sign and can take values of -511 to +511.

The accumulated error bit expansion unit 801 selectively expands the remaining 7 bits to 9 bits while keeping the sign of the input accumulated error data based on the instruction from the recording mode corresponding bit expansion unit 806. Specifically, the following three ways can be selected. First, 7 bits excluding the sign of the accumulated error data are regarded as the upper 7 bits and 0 is inserted in the lower 2 bits. The second is to insert 0 into the most significant 1 bit and the least significant 1 bit with 7 bits excluding the sign of the accumulated error data as the middle 7 bits. And
The third is 7 bits excluding the sign of the accumulated error data, the lower 7 bits.
0 is inserted in the upper 2 bits as a bit.

The A diffusion error calculation unit 802 and the B diffusion error calculation unit 803 calculate the diffusion error of 10 bits for the pixels A and B based on the input error data of signed 10 bits. The diffusion error is calculated by referring to the 9-bit input / 9-bit output table using the absolute value of the error of 9 bits regardless of whether the error is positive or negative. The signed 10-bit accumulated error data output from the accumulated error bit expansion unit 801 is converted by the B addition unit 805 into the B diffusion error calculation unit 80.
Addition processing is performed with the diffusion error to the pixel B having three outputs. Further, in the next pixel processing cycle, the addition process of the diffusion error to the pixel A which is the output of the A diffusion error calculation unit 802 and the output of the B addition unit 805 is performed by the A addition unit 804 to obtain the final accumulated diffusion error. be able to.

The most significant 7 bits (or the middle 7 bits) out of the 9 bits in the error bit limiting unit 201.
Is cut out, the error component corresponding to the least significant 2 bits (or the least significant 1 bit) to be deleted is diffused to the pixels A and B in the same line as the processed pixel. Further, in the present embodiment, the configuration using the table as the diffusion error calculating means has been described as an example, but it may be realized by an arithmetic process or a bit shift process, or a combination thereof.

Next, the bit deleting process and the bit inserting process associated with the recording mode performed by the error bit limiting unit 201 and the cumulative error bit expanding unit 801 will be described in detail.

First, the bit deletion (bit number restriction) control in the error bit restriction unit 201 will be described. In this control, the error bit limiting unit 201 is controlled by the recording mode corresponding bit limiting control unit 212.

As described above, the number of output gradations in the multi-valued error diffusion processing differs according to the three recording modes. The "high speed mode" has eight values (3 bits) and the "standard mode" has four values (2 bits). , "High-definition mode" is binary (1 bit).

FIG. 1 is a flow chart showing a process of deleting a bit of error data according to a recording mode. First, the recording mode corresponding bit limitation control unit 212 acquires information about the recording mode (step S101), and performs bit cutout control according to the recording mode. When the recording mode is the "high definition mode" (Yes in step S102)
s), since the number of output gradations is binary, the error generated as a result of quantizing each pixel takes a value of -511 to +511. Therefore, in the 10-bit error data input to the error bit limiting unit 201, all the 9 bits excluding the sign are superior, so the error bit limiting unit 201 selects the upper 7 bits of the 9 bits excluding the input sign. It is cut out and output with a sign (step S106).

In the case of "standard mode" (step S
(Yes at 103), the error data generated as a result of quantizing each pixel is −2 because the number of output gradations is four.
It takes a value of 55 to +255. Therefore, the 10-bit error data input to the error bit limiting unit 201 can take only 0 for the upper 1 bit out of the 9 bits excluding the sign, so the error bit limiting unit 201 excludes the input code.
Of the bits, 7 bits excluding the highest and the lowest are cut out and output with a sign (step S105).

Further, in the case of "high speed mode" (step S
No. 103), the error data generated as a result of quantizing each pixel is −12 because the number of output gradations is eight.
It takes a value of 7 to +127. Therefore, the error bit limiting unit 2
In the 10-bit error data input to 01, only the upper 2 bits of 9 bits excluding the sign can be 0. Therefore, the error bit limiting unit 201 excludes the input code to 9 bits.
Of the bits, the lower 7 bits are cut out and output with a sign (step S104). Next, the bit insertion (bit number expansion) control in the cumulative error bit expansion unit 801 will be described. This control is performed by the cumulative error bit extension unit 80.
1 is controlled by the recording mode compatible bit expansion controller 806.

FIG. 7 is a flow chart showing the processing for inserting the bits of the accumulated error data according to the recording mode. First, the recording mode corresponding bit expansion control unit 806 acquires information regarding the recording mode (step S701),
Bit insertion control according to the recording mode is performed in conformity with the bit deletion control in the recording mode corresponding bit limitation control unit 212 described above. That is, the cumulative error bit extension unit 8
When the recording mode is the “high quality mode” for the signed 8-bit accumulated error data input to 01 (Yes in step S702), 2 is added to the lower 9 bits excluding the sign.
Bits are added and output as signed 10 bits (step S706). In the case of the “standard mode” (Yes in step S703), 1 bit is added to each of the upper and lower bits of 9 bits excluding the sign and the signed 1 is added.
It is output as 0 bit (step S705). Furthermore,
In the case of the "high speed mode" (No in step S703), 2 bits are added to the higher order of 9 bits excluding the sign and output as a signed 10 bits (step S704).

In this way, the error information is shifted (bit limited) according to the number of output gradations associated with the recording mode, and is supplied to the diffusion error calculation unit, so that the range of the generated error is met. It is possible to appropriately control the diffusion of the error, and it is possible to efficiently reduce the error memory capacity and simplify the diffusion error calculating means.

As described in detail above, in the quantization (N-valued) processing using the error diffusion method, the bit of error data is determined according to the error occurrence range in the output gradation number N associated with the recording mode. It is possible to reduce the total capacity of the error memory while ensuring the optimum error diffusion control by performing the cutout control of the data and executing the calculation process of the diffusion error to the next line based on the error data whose bit number is limited. It will be possible.

As a result, the deterioration of the image quality due to the generation of an unpleasant texture is efficiently suppressed for different output gradation numbers N, and a high-quality image is formed at low cost without providing a large-scale error memory. Can be realized.

[Second Embodiment] Next, a second embodiment of the present invention will be described in detail with reference to the drawings.

In the first embodiment, the case where images are formed with the same resolution for all ink colors has been described, but the present invention is not limited to this, and different quantization control is performed for each color element. Can also be applied to a system for forming images at different resolutions. In the second embodiment, a detailed description will be given of quantization control when image formation is performed on Bk at an output resolution different from that of other ink colors (Cy, Mg, Ye). That is, it is possible to form a monochrome image at a high speed by configuring Bk so that the amount of ejected ink droplets is larger than that of the other ink colors and performing image formation with a lower resolution only for Bk.

The recording section of the ink jet recording apparatus in the second embodiment is the same as that in the first embodiment (FIG. 3). Further, the basic configuration of the engine and the controller that configure the inside is the same as that of the first embodiment.

The image data processing unit in the controller in the second embodiment is the same as that in the first embodiment (FIG. 4). In the image data processing unit, after converting the input image information into multi-valued image data for each ink color, by quantizing the multi-valued image data using the multi-valued error diffusion method and the density pattern method, Ink color dot data is generated. The generated dot data is output to the engine.

In the second embodiment, adaptive quantization processing is selectively executed according to the color element. FIG. 11 shows
It is a figure which shows the quantization process performed according to a color element.
As shown in the figure, for Bk data, the input resolution is 3
4 by error diffusion method for 00 ppi x 300 ppi
After digitizing, develop it using a 2 × 2 halftone dot matrix,
The dot data of 600 dpi × 600 dpi, which is the output resolution, is obtained. On the other hand, for ink color data other than Bk, the input resolution of 300 ppi × 300 ppi is binarized by the error diffusion method, and then expanded using a 4 × 4 halftone dot matrix to obtain an output resolution of 1200 dpi ×.
Dot data of 1200 dpi is obtained. In this way, it is possible to generate and output dot data having different output resolutions for each ink color, using the same input resolution data for all ink colors.

Next, the bit deletion / insertion control of the diffusion error in the characteristic multilevel error diffusion processing of the second embodiment will be described in detail.

FIG. 10 is a block diagram showing the configuration of the next line error diffusion processing section for executing the error diffusion processing to the next line pixel in the multi-valued error diffusion processing block. The error is diffused to each of the five pixels C, D, E, F, and G located immediately below the processing line, and the accumulated error data from the upper line of each pixel is sent to an error memory (not shown). . The basic structure is the same as that of the first embodiment (FIG. 2), and those having the same functions are given the same numbers.

In FIG. 10, an error bit limiter 1001 selectively cuts out some bits of the input error data, limits the number of bits, and outputs the bits to the subsequent stage. Reference numerals 202 to 206 denote C diffusion error calculation unit, D diffusion error calculation unit, E diffusion error calculation unit, F diffusion error calculation unit, and
The G diffusion error calculation unit calculates and outputs an error diffused to the pixels C, D, E, F, and G. Reference numerals 207 to 210 denote a C addition unit, a D addition unit, an E addition unit, and an F addition unit, respectively.
Diffusion error is added to the accumulated error up to immediately before and output. A delay unit 211 delays data by a pixel processing cycle which is a processing time unit. Reference numeral 1012 denotes a color element-corresponding bit limit control unit, which executes bit number limitation control of error data according to a color element to be processed based on a control signal from the outside.

Next, the basic operation of calculating the accumulated error from the upper line with the above-mentioned configuration will be described. Error data corresponding to each pixel is sequentially input for each pixel processing cycle,
Further, in each unit, the pixel processing cycle is processed as a delay unit. Here, the input error is signed 10-bit data and can take values from -511 to +511.

The error bit limiting section 1001 selectively cuts out 7 bits from the remaining 9 bits while keeping the sign of the input error based on the instruction from the color element corresponding bit limiting section 1012. Specifically, the following two ways can be selected. The first is the remaining 7 bits after discarding the most significant 1 bit and the least significant 1 bit. And the second is the top two
It is the lower 7 bits remaining after discarding the bits.

The C diffusion error calculation unit 202, the D diffusion error calculation unit 203, the E diffusion error calculation unit 204, the F diffusion error calculation unit 205, and the G diffusion error calculation unit 206 are based on the input signed 8-bit error data. Pixels C, D, E,
8 bits of diffusion error to F and G are calculated. Here, using 7 bits of the absolute value of the error regardless of whether the error is positive or negative,
The diffusion error is calculated by referring to the table of 7-bit input and 7-bit output. First, the diffusion error to the pixel G, which is the output of the G diffusion error calculation unit 206, is determined by the delay unit 211.
After being held (delayed) by the next pixel processing cycle, the F addition unit 210 performs addition processing with the diffusion error to the pixel F which is the output of the F diffusion error calculation unit 205. In the next pixel processing cycle, the E addition unit 209 performs the addition process of the diffusion error to the pixel E, which is the output of the E diffusion error calculation unit 204, and the output of the F addition unit 210. In the next pixel processing cycle, the D addition unit 208 adds the diffusion error to the pixel D, which is the output of the D diffusion error calculation unit 203, and the output of the E addition unit 209. Further, in the next pixel processing cycle, the C addition unit 207 outputs the C diffusion error calculation unit 202 output diffusion error to the pixel C and D.
The addition processing with the output of the addition unit 208 is performed, and the upper line accumulated error that is the accumulated error from the upper line of the pixel of the line located immediately below the processed pixel is obtained.

FIG. 13 is a block diagram showing the arrangement of a processing pixel line error diffusion processing unit for executing error diffusion processing on processing line pixels in the multi-valued error diffusion processing unit. The accumulated error data from the upper line of each pixel, which is generated by the previous line processing and is temporarily stored in an error memory (not shown), is input, and is further positioned on the right side of the same line as the processed pixel. The error is diffused to each of the two pixels A and B to generate final accumulated error data. The basic structure is the same as that of the first embodiment (FIG. 8), and the same numbers are given to those having the same functions.

In FIG. 13, reference numeral 1301 denotes a cumulative error bit expansion unit, which selectively shifts (bit inserts) the cumulative error data from the upper line read from the error memory to expand the number of bits and expand the data. The accumulated error data thus generated is output to the subsequent stage. Reference numerals 802 and 803 denote an A diffusion error calculation unit and a B diffusion error calculation unit, respectively, which calculate and output the error diffused to the pixels A and B based on the input error data of the processed pixel. Reference numerals 804 and 805 denote an A addition section and a B addition section, respectively, which add a diffusion error to the accumulated error up to immediately before and output it. A color element corresponding bit expansion control unit 1306 executes a bit number expansion control of accumulated error data according to a color element to be processed based on a control signal from the outside.

Next, the basic operation of calculating the final accumulated error with the above configuration will be described. Error data corresponding to each pixel is sequentially input for each pixel processing cycle, and each unit processes the pixel processing cycle as a delay unit. The input accumulated error data is accumulated information of the error diffused from the upper line temporarily stored in the error memory (not shown), and is a signed 8-bit data and takes a value of -127 to +127. On the other hand, the error data of the input processed pixel is 10-bit data with a sign and can take values of -511 to +511.

The cumulative error bit expanding unit 1301 selectively expands the remaining 7 bits to 9 bits while holding the sign of the input cumulative error based on the instruction from the color element corresponding bit expanding unit 1306. Specifically, the following two ways can be selected. First, 7 bits excluding the sign of the accumulated error data are set to 7 bits in the middle and 0 is inserted in the most significant 1 bit and the least significant 1 bit. The second is that 7 bits excluding the sign of the accumulated error data are lower 7 bits and 0 is inserted in the upper 2 bits.

The A diffusion error calculation unit 802 and the B diffusion error calculation unit 803 calculate the diffusion error of 10 bits for the pixels A and B based on the input signed error data of 10 bits. The diffusion error is calculated by referring to the 9-bit input / 9-bit output table using the absolute value of the error of 9 bits regardless of whether the error is positive or negative. The signed 10-bit accumulated error data output from the accumulated error bit expansion unit 801 is converted by the B addition unit 805 into the B diffusion error calculation unit 80.
Addition processing is performed with the diffusion error to the pixel B having three outputs. Further, in the next pixel processing cycle, the addition process of the diffusion error to the pixel A which is the output of the A diffusion error calculation unit 802 and the output of the B addition unit 805 is performed by the A addition unit 804 to obtain the final accumulated diffusion error. be able to.

In the error bit limiting unit 1001, 9
When the most significant 7 bits (or the middle 7 bits) are cut out, the error components corresponding to the least significant 2 bits (or the least significant 1 bit) to be deleted are the pixels A and B on the same line as the processing pixel. Is spread to. Also,
In the present embodiment, the configuration using the table as the diffusion error calculating means has been described as an example, but it may be realized by an arithmetic process or a bit shift process, or a combination thereof.

Next, the bit deleting process and the bit inserting process associated with the ink color elements in the process performed by the error bit limiting unit 1001 and the cumulative error bit expanding unit 1301 will be described in detail.

First, the bit deletion (bit number restriction) control in the error bit restriction unit 1001 will be described.
In this control, the error bit limiting unit 1001 is controlled by the color element corresponding bit limiting control unit 1012.

As described above, the number of output gradations in the multi-valued error diffusion processing differs according to the ink color, and is 4 values (2 bits) for Bk data and 8 values (2 bits) for other ink colors. .

FIG. 9 is a flow chart showing the processing for deleting the bits of the error data according to the ink color element.
First, the color element corresponding bit limit control unit 1012 acquires information about the target ink color (step S90).
1) The bit cutout control according to the color element is performed. When the ink color element is Bk data (Yes in step S902)
s), since the number of output gray levels is four, the error data generated as a result of quantizing each pixel is -255 to +25.
Takes a value of 5. Therefore, in the 10-bit error data input to the error bit limiting unit 1001, only the upper 1 bit of the 9 bits excluding the sign can be 0. Therefore, the error bit limiting unit 1001 inputs the 9-bit error data excluding the input sign. Of these bits, 7 bits excluding the highest and the lowest are cut out and output with a sign (step S903).

In addition, the ink color elements are Cy, M other than Bk,
In the case of g and Ye (No in step S902), since the number of output gray levels is eight, the error data generated as a result of quantizing each pixel takes a value of -127 to +127.
Therefore, the 10-bit error data input to the error bit limiting unit 1001 is the upper 2 bits of the 9 bits excluding the sign.
Since only 0 bits can be taken, the error bit limiting unit 10
For 01, the lower 7 bits of the 9 bits excluding the input code are cut out and output with a sign (step S90).
4).

Next, the bit insertion (bit number expansion) control in the cumulative error bit expansion unit 1301 will be described. In this control, the cumulative error bit extension unit 1301 is controlled by the color element corresponding bit extension control unit 1306.

FIG. 12 is a flow chart showing a process of inserting a bit of accumulated error data according to an ink color element. First, the color element corresponding bit expansion control unit 1306 acquires information about the target ink color (step S1).
201), and the above-described color element corresponding bit limit control unit 101
The bit insertion control according to the color element is performed in conformity with the bit deletion control in 2. That is, when the ink color element is Bk with respect to the signed 8-bit accumulated error data input to the accumulated error bit expansion unit 1301 (step S
(Yes in 1202), 1 bit is added to each of the upper and lower bits of the 9 bits excluding the sign and output as the signed 10 bits (step S1203). When the ink color elements are Cy, Mg, and Ye other than Bk (step S
No of 1202), 2 bits are added to the lower order of 9 bits excluding the sign, and output as a signed 10 bits (step S1204).

In this way, the error information is shifted (bit limited) in accordance with the number of output gradations associated with the ink color to be processed and supplied to the diffusion error calculation unit, so that the error that occurs is generated. It is possible to appropriately control the diffusion of the error according to the range, and it is possible to efficiently reduce the error memory capacity and simplify the diffusion error calculating means.

As described above in detail, in the quantization (N-valued) processing using the error diffusion method, the bit of error data is determined according to the error occurrence range in the output gradation number N associated with the color element. It is possible to reduce the total capacity of the error memory while ensuring the optimum error diffusion control by performing the cutout control of the data and executing the calculation process of the diffusion error to the next line based on the error data whose bit number is limited. It will be possible.

As a result, the deterioration of the image quality due to the generation of an unpleasant texture is efficiently suppressed for different output gradation numbers N, and a high-quality image is formed at low cost without providing a large-scale error memory. Can be realized.

[Modification] In the first embodiment, the case where the evaluation values in the multi-valued error diffusion process are set as fixed values with equal intervals, and the determination threshold is set as a fixed value located at the center of the evaluation values will be described. However, the present invention is not limited to this.

Here, in the quantization method using the error diffusion method, a phenomenon of image quality deterioration due to a delay in dot generation occurs with respect to a rapid density change at an edge portion of an image. This is caused by the fact that the error is not easily accumulated in the region where the pixel value is very close to the evaluation value, and further, the large amount of quantization error accumulated in the past low-concentration region or high-concentration region is different. It occurs remarkably when it is diffused into the density region. The binarization result in different density regions may be unnaturally distorted by a large amount of errors from the past regions, and in some cases, a part of the peripheral line image may be lost.

A method of adding a noise component to the determination threshold is effective for the problem in the quantization processing using such an error diffusion method, and the error diffusion using the threshold matrix composed of each threshold component is effective. Many processing methods are adopted. As a result, it is possible not only to avoid or suppress the delay of dot generation, but also to suppress the generation of an unpleasant texture peculiar to the error diffusion method and improve the image quality. In addition to a method of adding a noise component, a method of adaptively varying a threshold value according to an input pixel value has been proposed and implemented.

Furthermore, not only the threshold value but also the evaluation value is not always set to a uniform fixed value. Many methods have been proposed, such as a method of giving a predetermined characteristic to the evaluation value and a method of adaptively changing the evaluation value.

In this way, when the adaptive control is applied to the evaluation value and the judgment threshold value, the maximum error that can be inevitably expanded or changed, so that the maximum error amount for each output gradation number N is changed. Bit limit control.

In the first and second embodiments, when a 2 × 2 halftone dot matrix is used as the output gray scale number of the multi-valued error diffusion process, a four-valued or 4 × 4 halftone dot matrix is used. In the case of use, eight values have been described as an example, but the number of output gradations with respect to the matrix size should be optimally set according to the print mode, recording medium, print head characteristics, and the like. Further, the number of output gradations is not limited to 4 or 16 and can be applied to all the output gradation numbers of 2 or more.

Further, in the first and second embodiments, the case where the error is diffused to the line immediately below in addition to the pixel on the same line as the processing pixel has been described as an example, but the present invention is not limited to this. However, the present invention can be applied to a method of diffusing an error for two or more line pixels below. Further, there is no limitation on the number of pixels for which the error is diffused.

Further, in the first and second embodiments, the method of always processing pixels in a predetermined direction for a line has been described, but the processing direction may be alternately switched for each line. The processing direction may be randomly selected.

Further, in the first and second embodiments, description has been made of the case where an image is finally formed by using dots of a single size based on binary image data (binary recording). It is also possible to selectively form dots of different sizes on the basis of multi-valued image data to complete an image (multi-valued recording) or to perform overprinting of the same ink.

Furthermore, in the first and second embodiments, black (Bk), cyan (Cy), magenta (M
Although the inkjet type image forming system using the four color inks of g) and yellow (Ye) has been described, the number of colors and color types are not limited thereto. 3 excluding Bk
The color ink may be used, another special color may be added, or a low density light color ink such as light Cy or light Mg may be used in combination. Further, the number of print heads to be mounted is not limited to one set (one for each color), and the present invention can be applied to an image forming system or the like having a plurality of sets (two or more for each color) of print heads to realize high-speed processing.

Further, in the first and second embodiments, the configuration in which the quantization processing and the pass data generation processing are performed inside the ink jet recording apparatus has been described, but a part or all of these are connected to the host PC. It is obvious that the configuration may be realized by the driver on the side or other external device.

Furthermore, the first and second embodiments are not limited by the operating principle and configuration of the print head. That is, the print head is provided with a heating element (electrical / thermal energy conversion element) in the vicinity of the ejection port, and by applying an electric signal to this heating element, the ink is locally heated to cause a pressure change, and the ink is ejected. A thermal method of ejecting from an ejection port or a piezo method of ejecting the ink by applying a mechanical pressure to the ink by using an electric / pressure conversion means such as a piezo element may be used.

Further, in the first and second embodiments, the ink jet type image forming system has been described as an example, but the applicable image forming system is not limited to the ink jet type and may be a fusion type. It is also possible to apply it to a thermal transfer system or a laser system.

Further, the form of the image forming system according to the present invention is not limited to one provided integrally or separately as an image output device of an information processing device such as a computer or a word processor, but a copying device combined with a reading device. It may be a facsimile machine having a communication function.

Even if the present invention is applied to a system composed of a plurality of devices (eg, host computer, interface device, reader, printer, etc.), a device composed of one device (eg, copying machine, facsimile device). Etc.) may be applied.

Further, the object of the present invention is to supply a recording medium having a program code of software for realizing the functions of the above-described embodiments to a system or apparatus, and to supply the computer (CPU or MP) of the system or apparatus.
It goes without saying that U) is also achieved by reading and executing the program code stored in the recording medium.

In this case, the program code itself read from the recording medium realizes the functions of the above-described embodiments, and the recording medium storing the program code constitutes the present invention.

As a recording medium for supplying the program code, for example, a floppy (registered trademark) disk,
Hard disk, optical disk, magneto-optical disk, CD-
ROM, CD-R, magnetic tape, non-volatile memory card, ROM, etc. can be used.

Moreover, not only the functions of the above-described embodiments are realized by executing the program code read by the computer, but also the OS (operating system) running on the computer based on the instructions of the program code. It is needless to say that this also includes a case where the above) performs a part or all of the actual processing and the processing realizes the functions of the above-described embodiments.

Further, after the program code read from the recording medium is written in the memory provided in the function expansion board inserted into the computer or the function expansion unit connected to the computer, based on the instruction of the program code, It goes without saying that a case where the CPU provided in the function expansion board or the function expansion unit performs a part or all of the actual processing and the processing realizes the functions of the above-described embodiments is also included.

[0119]

As described above, according to the present invention,
Shift (bit limitation) according to the number of output gradations N determined in association with the image forming mode and the color element to be processed
By generating a diffusion error based on the error information obtained and storing it in the error memory with a reduced number of bits, it is possible to reduce the error memory capacity while performing effective error diffusion processing. Therefore, it is possible to efficiently suppress the deterioration of the image quality due to the generation of an unpleasant texture for different output gradation numbers N, and it is possible to form a high-quality image at low cost without providing a large-scale error memory. It has a remarkable effect that it can provide the quantization processing.

[Brief description of drawings]

FIG. 1 is a flowchart showing a process of deleting a bit of error data according to a recording mode.

FIG. 2 is a block diagram showing a configuration of a next-line error diffusion processing unit that executes error diffusion processing to a next-line pixel in the multi-valued error diffusion processing unit.

FIG. 3 is a diagram showing a configuration of a recording unit in the inkjet recording apparatus.

FIG. 4 is a block diagram showing a configuration of an image data processing unit in the controller.

FIG. 5 is a diagram showing a diffusion coefficient matrix in the error diffusion method.

FIG. 6 is a diagram showing a relationship between a recording mode and a quantization process in the present embodiment.

FIG. 7 is a flowchart showing a process of inserting a bit of accumulated error data according to a recording mode.

FIG. 8 is a block diagram showing a configuration of a processing pixel line error diffusion processing unit that performs error diffusion processing on processing line pixels in a multi-valued error diffusion processing unit.

FIG. 9 is a flowchart showing a process of deleting a bit of error data according to an ink color element.

FIG. 10 is a block diagram showing a configuration of a next-line error diffusion processing unit that executes error diffusion processing to a next-line pixel in the multi-valued error diffusion processing block.

FIG. 11 is a diagram showing a quantization process executed according to a color element.

FIG. 12 is a flowchart showing a process of inserting a bit of accumulated error data according to an ink color element.

FIG. 13 is a block diagram showing a configuration of a processing pixel line error diffusion processing unit that performs error diffusion processing on processing line pixels in a multi-valued error diffusion processing unit.

FIG. 14 is a diagram showing a configuration example of a mask table for each print scan.

FIG. 15 is a diagram for explaining a state of print scanning using the mask table shown in FIG.

FIG. 16 is a diagram showing a configuration of a controller unit of the inkjet recording apparatus.

FIG. 17 is a diagram showing a configuration of an engine unit of the inkjet recording apparatus.

FIG. 18 is a diagram showing a flow of image data processing in the image data processing unit 1613.

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 2C057 AF39 AL32 AM28 CA05 CA10                 2C262 AA02 AA24 AB20 BB03 BB08                       EA18 EA19 GA09 GA14                 5B057 CA01 CA08 CA12 CA16 CB01                       CB07 CB12 CB16 CE14 CE17                       CE18 CH09 CH11                 5C077 LL17 MP01 NN11 PQ22 RR02                       TT05

Claims (14)

[Claims] 1. Error data generated when a value of a pixel of interest of multi-valued image data is quantized into an N value is stored in an error memory, and the stored error data is stored in peripheral pixels according to a predetermined diffusion coefficient. In an image processing device that performs a multi-valued error diffusion process that diffuses and quantizes a value of a target pixel to which the error data is diffused into an N value, a value of a target pixel of multi-valued image data Error bit number limiting means for limiting the number of bits of error data generated at the time of quantization according to the recording mode associated with the multi-valued image data and storing the error bit in the error memory, and the bit read from the error memory. An image processing apparatus, comprising: an error bit number expansion means for expanding the number of bits of error data whose number is limited according to the recording mode. 2. The error bit number limiting means limits the number of bits to a predetermined lower bit out of the total number of bits when the recording mode is a high speed mode in which high speed processing is required, and also requires high quality processing. The image processing apparatus according to claim 1, wherein in the high-quality mode, the upper bits are limited to a predetermined upper bit, and in the standard mode, the intermediate bits are limited. 3. The error bit number expanding means adds an upper bit to the predetermined lower bit and expands to all bits when the recording mode is a high speed mode in which high speed processing is required, and a high quality process. In the case of the high-quality mode required, the lower bits are added to the predetermined upper bits to extend all bits, and in the standard mode, the upper bits and the lower bits are added to the intermediate bits to add all bits. The image processing apparatus according to claim 2, wherein the image processing apparatus is extended to bits. 4. Error data generated when a value of a pixel of interest of multi-valued image data is quantized into an N value is stored in an error memory, and the stored error data is stored in peripheral pixels according to a predetermined diffusion coefficient. In an image processing device that performs a multi-valued error diffusion process that diffuses and quantizes a value of a target pixel to which the error data is diffused into an N value, a value of a target pixel of multi-valued image data An error bit number limiting means for limiting the number of bits of error data generated at the time of quantization according to the ink color information to be processed and storing it in the error memory; and the number of bits read from the error memory is limited. An image processing apparatus, comprising: an error bit number expanding means for expanding the bit number of error data according to the ink color information. 5. The error bit number limiting means limits the number of bits to a predetermined lower bit of the total number of bits when the ink color information is not black, and limits it to an intermediate bit in the case of black. The image processing device according to claim 4. 6. The error bit number expanding means adds an upper bit to the predetermined lower bit and expands to all bits when the ink color information is not black, and expands to an intermediate bit in the case of black. The image processing apparatus according to claim 5, wherein the upper and lower bits are added and expanded to all bits. 7. Error data generated when a value of a pixel of interest of multi-valued image data is quantized into an N value is stored in an error memory, and the stored error data is stored in peripheral pixels according to a predetermined diffusion coefficient. A processing method of an image processing device for performing multi-valued error diffusion processing, wherein the multi-valued error data is diffused and the value of a pixel of interest to which the error data is diffused is quantized into an N value. An error bit number limiting step of limiting the number of bits of error data generated when quantizing the value of the pixel of interest in accordance with a recording mode associated with the multi-valued image data and storing the error bit number in the error memory; An error bit number expanding step of expanding the number of bits of error data whose number of bits read from a memory is limited according to the recording mode. 8. In the error bit number limiting step, when the recording mode is a high speed mode in which high speed processing is required, the error bit number limiting step limits the number of bits to a predetermined lower bit and requests high quality processing. The processing method of the image processing apparatus according to claim 7, wherein in the high-quality mode, the predetermined upper bits are limited, and in the standard mode, the intermediate bits are limited. 9. In the error bit number expanding step, when the recording mode is a high speed mode requiring high speed processing, an upper bit is added to the predetermined lower bit to expand to all bits, and a high quality processing is performed. In the case of the high-quality mode required, the lower bits are added to the predetermined upper bits to extend all bits, and in the standard mode, the upper bits and the lower bits are added to the intermediate bits to add all bits. 9. The processing method of the image processing apparatus according to claim 8, wherein the processing method is extended to bits. 10. Error data generated when a value of a pixel of interest of multi-valued image data is quantized into an N value is stored in an error memory, and the stored error data is stored in peripheral pixels according to a predetermined diffusion coefficient. A processing method of an image processing device for performing multi-valued error diffusion processing, wherein the multi-valued error data is diffused and the value of a pixel of interest to which the error data is diffused is quantized into an N value. An error bit number limiting step of limiting the number of bits of error data generated when quantizing the value of the pixel of interest according to the ink color information to be processed and storing it in the error memory, and the bit read from the error memory An error bit number expanding step of expanding the number of bits of the error data of which the number is limited according to the ink color information. 11. The error bit number limiting step includes a step of: out of all bit numbers when the ink color information is not black.
11. The processing method for an image processing apparatus according to claim 10, wherein the predetermined lower bits are limited, and in the case of black, the intermediate bits are limited. 12. The error bit number expanding step adds an upper bit to the predetermined lower bit to expand to all bits when the ink color information is not black, and to an intermediate bit in the case of black. The high-order bit and the low-order bit are added to extend all bits.
The processing method of the image processing device according to. 13. A program for causing a computer to function as the means according to any one of claims 1 to 6. 14. A computer-readable recording medium in which the program according to claim 13 is recorded.