JP2015112773A - Image processing method, image processing device, recording method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing device capable of reducing bleeding in which different colors of inks are mixed when an image part of image and a character image part are adjacent, suppressing a contour of character from being viewed as white color, and performing image recording with high quality.SOLUTION: An edge of a character image part is detected, and a recording dot of an image part of image adjacent to the edge, is moved so as to overlap the edge of the character image part. Therefore, occurrence of bleeding and void on a boundary part between the image part of image and the character image part, can be suppressed.

Description

  The present invention relates to an image processing method, an image processing apparatus, a recording method, and a recording apparatus for recording a character image and an image image on a recording medium.

  An ink jet recording apparatus that performs recording by discharging ink onto a recording medium is capable of high-density and high-speed recording. This ink jet recording apparatus has many advantages such as low running cost and low noise generated during recording, and is widely used as various image output apparatuses such as printers and portable printers. As a recording method of the ink jet recording apparatus, a serial method for scanning a carriage on which a recording head is mounted with respect to the recording medium, or a recording medium with respect to a recording head in which ejection openings are arranged in a range corresponding to the width of the recording medium. A full line system for conveying is known.

  As the ink jet recording apparatus as described above, many products capable of color recording using a plurality of colors of ink are provided. In such a color ink jet recording apparatus, since black ink is generally used for recording characters and the like, an ink capable of realizing sharpness, sharpness, and high density of recorded characters is required. With respect to this point, a method using a black ink having low penetrability with respect to a recording medium is known in order to prevent a coloring material such as a dye or pigment in the black ink from penetrating into the recording medium. By using such an ink, it is possible to increase the amount of the coloring material remaining and fixed near the surface of the recording medium, and to realize the sharpness and high density of the recorded image.

  On the other hand, for color inks, when inks of different colors are applied to adjacent areas on the recording medium, these inks mix at the boundary and cause a phenomenon (bleed) that degrades the quality of the color image. There is. In order to prevent this, it is known to use ink having high permeability to the recording medium. Accordingly, it is possible to prevent a large amount of ink applied to each region from penetrating into the recording medium and mixing with ink of other colors beyond the boundary.

  When the above black ink and color ink are used, the following problems may occur. In an image where the black and color areas are adjacent, the black ink with low penetrability does not penetrate quickly into the recording medium, but mixes with the color ink across the boundary of the area so that it is pulled by the color ink with high penetrability. May fit. That is, bleeding occurs at the boundary between the black area and the color area.

  For such a problem, Patent Document 1 describes that bleeding at a boundary is suppressed by thinning out color pixels adjacent to black pixels.

JP-A-6-135015

  However, in the conventional bleed reduction technique, there is a possibility that white spots occur at the boundary between the character portion and the color portion by thinning out the color ink to be applied to the color region adjacent to the black region. Although the bleeding between the black ink and the color ink can be suppressed, the outline of the character part is visually recognized as white, which may deteriorate the character quality. In particular, bleeding is likely to occur when the amount of shot at the boundary between the color image portion and the black character portion is greatly different, or when highly penetrable ink is in contact with low penetrable ink. In addition, as the printing speed of the printer is improved, the difference in driving time between adjacent different color areas is shortened, so that bleeding is likely to occur.

  The object of the present invention is to suppress the so-called bleed that inks of different colors are mixed when the image image portion and the character image portion are adjacent to each other, and to suppress the white outline of the character from being visually recognized. Provided are an image processing apparatus and an image processing method capable of performing accurate image recording.

  In order to solve the above-described problems, the present invention provides an image processing method for generating data for recording an image on a recording medium using a recording head that ejects ink droplets, based on the image data to be recorded. Character data for recording a character object and indicating ink droplet ejection or non-ejection; image data for recording an image object and image data indicating ink droplet ejection or non-ejection; A detection step of detecting an edge pixel corresponding to an edge of the character object in the character data, an ink droplet ejection in the image data, and the edge pixel of the character data. Data indicating non-ejection of ink droplets is set for pixels corresponding to adjacent pixels, and the set image is set. Characterized in that it has a relative to the pixel corresponding to the edge pixels of adjacent and the character data, a setting step of setting data indicating the ejection of ink droplets.

  According to the above configuration, when an image including an image image portion and a character image portion is recorded, the occurrence of bleeding at the boundary between the image image and the character image is reduced, and the outline of the character or line drawing is reduced. Occurrence of omission is suppressed, and a high-quality image can be recorded.

It is a schematic diagram of an inkjet printer and a recording head. It is a block diagram which shows a recording system. It is a figure explaining the subject of this invention. It is a figure explaining the process in 1st Embodiment. 3 is a flowchart of image processing according to the first embodiment. 3 is a flowchart of image processing according to the first embodiment. It is a figure which shows an input image page buffer. 3 is a flowchart of image processing according to the first embodiment. It is a figure which shows the page buffer for image image parts, and the page buffer for character image parts. It is a figure which shows the page buffer for image image parts, and the page buffer for character image parts. It is a dot arrangement pattern in the first embodiment. It is a figure explaining the overlap control process which concerns on 1st Embodiment. It is a figure explaining the edge detection method of the character image part which concerns on 1st Embodiment. It is a figure explaining the edge detection result which concerns on 1st Embodiment. It is a flowchart of the image processing which concerns on 2nd Embodiment. It is a figure which shows the page buffer in 2nd Embodiment. It is a figure which shows the page buffer in 2nd Embodiment. 10 is a flowchart of image processing according to the third embodiment. It is a figure explaining the overlap control process which concerns on 3rd Embodiment. It is a quantization pattern in the third embodiment. It is a figure for selecting the quantization pattern in 3rd Embodiment.

(First embodiment)
FIG. 1A is a diagram schematically illustrating a recording apparatus according to the present embodiment. The recording apparatus of the present embodiment is a full-line type ink jet printer 100 and includes recording heads 101, 102, 103, and 104. The recording heads 101 to 104 are recording heads that can eject black (K), cyan (C), magenta (M), and yellow (Y) ink droplets. In each recording head, a nozzle row in which a plurality of nozzles (ejection ports) that eject ink droplets are arranged along the x direction in the drawing is arranged at an interval of 1200 dpi. In this way, the recording heads 101 to 104 that discharge ink droplets of different colors are arranged along the y direction in the figure and apply ink droplets of a plurality of colors to the recording medium 106 that is conveyed in the y direction in the figure. Thus, a color image is formed.

  FIG. 1B is a diagram illustrating the nozzle arrangement on the face surface (discharge port surface) of the recording head 101. As shown in the drawing, a plurality of ejection substrates 1011, 1012, and 1013 are arranged in the recording head 101 along the x direction in the figure. On each discharge substrate, nozzles are arranged in the x direction in the figure, and four nozzle rows are arranged in the y direction.

  Returning to FIG. 1A, the recording medium 106 is conveyed in the y direction in the figure by rotating the conveying roller 105 (and other unillustrated rollers) by a motor (not shown). While the recording medium 106 is conveyed, an operation of ejecting ink droplets (ink dots) is performed from the nozzles provided in the recording heads 101 to 104 at a frequency corresponding to the conveying speed of the recording medium 106. By controlling the conveyance speed of the conveyance roller 105 and the frequency for ejecting ink droplets from the recording head, an image for one page of the recording medium 106 is recorded.

  In the figure, a scanner 107 is arranged at a position downstream of the recording heads 101 to 104 in parallel with the recording heads 101 to 104 in the y direction. Reading elements are arranged at a predetermined pitch in the scanner 107, read an image recorded by the recording heads 101 to 104, and outputs multi-value RGB data.

  FIG. 2 is a block diagram illustrating the recording system according to the present embodiment. This recording system includes the printer 100 shown in FIG. 1 and a host PC 200 as its host device. The host PC 200 mainly has the following elements. The CPU 201 executes processing according to a program held in the HDD 203 or the RAM 202 which is a storage unit. The RAM 202 is a volatile storage and temporarily stores programs and data. The HDD 203 is a non-volatile storage and similarly holds programs and data. A quantization mask described later is also stored in the HDD 203. A data transfer I / F (interface) 204 controls data transmission / reception with the printer 100. As a connection method for this data transmission / reception, USB, IEEE 1394, LAN, or the like can be used. A keyboard / mouse I / F 205 is an I / F that controls an HID (Human Interface Device) such as a keyboard and a mouse, and a user can input via the I / F. A display I / F 206 controls display on a display (not shown).

  On the other hand, the printer 100 mainly includes the following elements. The CPU 211 executes each process described later according to programs stored in the ROM 213 and the RAM 212. The RAM 212 is a volatile storage and temporarily stores programs and data. The ROM 213 is a non-volatile storage, and holds table data and programs used in processing to be described later. The data transfer I / F 214 controls data transmission / reception with the host PC 200. The head controller 215 supplies print data to the print heads 101 to 104 and controls the discharge operation of the print head. Specifically, the head controller 215 reads control parameters and print data from a predetermined address in the RAM 212. Then, when the CPU 211 writes the control parameter and the recording data at the predetermined address in the RAM 212, the processing is started by the head controller 215, and ink is ejected from the recording head.

  The image processing accelerator 216 is hardware that can execute image processing faster than the CPU 211. Specifically, the image processing accelerator 216 reads parameters and data necessary for image processing from a predetermined address in the RAM 212. When the CPU 211 writes the parameters and data at the predetermined address in the RAM 212, the image processing accelerator 216 is activated and predetermined image processing is performed on the data. In this embodiment, a quantization mask determination process, which will be described later, is performed by software by the CPU 211, and image processing at the time of recording including processing of the quantization processing unit is performed by hardware by the image processing accelerator 216. The image processing accelerator 216 is not an indispensable component, and the above table parameter creation processing and image processing may be executed only by processing by the CPU 211 in accordance with the specifications of the printer.

  Here, the deterioration of character quality, which is a problem in the present embodiment, and the characteristic configuration of the present embodiment will be described with reference to FIGS. FIG. 3 is a diagram in which recording dots in an image portion indicating an image object and recording dots in a character portion indicating a character object are added to adjacent pixels on the recording medium 106. In this figure, the color image portion is provided with C ink recording dots, and the black character portion is provided with K ink recording dots. Reference numeral 2123C denotes a C ink color density buffer of the color image portion, and each pixel indicates a color density value of C ink indicated by 256 gradations. Reference numeral 2124C denotes a C ink print data buffer for the color image portion. The signal value of each pixel is the result of quantization processing of the ink color density value of 2123C into binary, and indicates recording (On: 1) or non-recording (Off: 0) of the C ink recording dots. Reference numeral 2123K2 denotes a K ink color density buffer for a character portion, and each pixel indicates a color density value of K ink indicated by 256 gradations. 2124K2 is a K ink print data buffer for the character portion, and the signal value of each pixel is the result of quantizing the ink color density value of 2123K2 into binary values, and recording of K ink recording dots (On: 1) Or, non-recording (Off: 0) is indicated.

  In the figure, the C ink recording dots are applied to the columns 1401 and 1402, and the K ink recording dots are applied to the columns 1403 and 1404. At this time, since the C ink column 1402 of the color image portion and the K ink column 1403 of the character portion are adjacent to each other, bleeding occurs at this boundary portion.

  Here, the water-based ink used in the ink jet recording apparatus of the present embodiment will be described. The water-based ink contains a coloring material and water as a main solvent, a water-soluble organic solvent, and other components, for example, a viscosity modifier, a pH adjuster, an antiseptic, a surfactant, an antioxidant, and the like as necessary. Composed. Since the main solvent is water, it is used in many ink jet recording apparatuses. There are two types of coloring materials, dyes and pigments, which are used properly according to the printing application.

  As the water-based ink, there are a highly permeable ink having high permeability to a recording medium and a low permeable ink having low permeability to a recording medium. Highly penetrating inks are less likely to bleed due to the high ink permeation speed, but tend to decrease in density when the colorant penetrates deeply into the recording medium. There is a problem that feathering is likely to occur because it easily penetrates in the direction. On the other hand, low penetrability inks tend to bleed due to the slow ink permeation rate, but the colorant tends to remain on the surface of the recording medium, so the concentration tends to increase, and feathering is unlikely to occur. There is an advantage that reproducibility is good. Even if the color material is a dye or a pigment, bleeding may occur. The penetrability of the water-based ink into the recording medium can be controlled by using a surfactant or the like. In order to adjust the penetrability in order to achieve both bleed and feathering for each ink color, it penetrates for each ink color. It is common for gender to be different.

  Next, a comparative example and the present invention will be described with reference to FIG. 4A shows a comparative example, and FIG. 4B shows a dot arrangement of the present invention. FIG. 4A shows an example in which bleeding is avoided by thinning out the recording dots of the C ink in the column 1402. Reference numeral 2125C shows a state after thinning out the C ink applied to the boundary portion by turning off the pixels in the column 1402 adjacent to the character portion in the C ink print data buffer 2124C of the color image portion. In this case, the bleeding at the boundary between the color image portion and the character portion can be suppressed, but the whiteness of the width T occurs, and the character quality deteriorates.

  On the other hand, FIG. 4B is a diagram for explaining a method of moving the C ink recording dots in the column 1402 adjacent to the character portion in the color image portion to the boundary of the character portion in the present embodiment. Reference numeral 2126C shows a state in which the pixels in the column 1402 are turned off and the pixels in the column 1403 are turned on with respect to the C ink print data buffer 2124C of the color image portion. That is, the recording dot of C ink to be added to the column 1402 is moved by one pixel to the character portion side and applied to the column 1403. At this time, since there is data for applying the K ink in the column 1403, the K ink and the C ink are applied in an overlapping manner. Thereby, the recording dots in the color image portion overlap with the recording dots in the black character portion, so that not only the occurrence of bleed can be suppressed, but also mechanical dot gain occurs. The mechanical dot gain is a phenomenon in which the recording dots spread and the coverage area of the recording dots on the recording medium increases. As a result, since the width of the boundary between the color image portion and the black character portion is t (t <T), it is possible to achieve both suppression of bleeding at the boundary between the color image portion and the black character portion and reduction of white spots.

  FIG. 5 is a flowchart showing an image processing flow executed by the ink jet recording system according to the present embodiment. Here, the host PC and the inkjet printer are the host PC 200 and the printer 100 shown in FIG.

  The host PC 200 performs the following steps s2001 to s2006. In step s2001, an input image including a character image and an image is drawn (developed) on a page buffer provided in the RAM 202 based on PDL (page description language). In step s2002, the character portion and the image image portion are separated from the drawn input image, and character image data and image image data are generated. In step s2004, the image data is irreversibly compressed, and in step s2005, the character image data is reversibly compressed. In step s2006, the compressed image image data and character image data are sent to the printer 100. In addition, there is no restriction | limiting in the order which performs the branched process about the process of step s2004, and the process of step s2005.

  Next, the printer 100 performs the following processing in steps s1001 to s1010. In step s1001, the compressed image image data and character image data are received from the host PC 200. In step s1002, the irreversible compressed image is expanded to generate image data. In step s1003, the image image data is converted into multivalued CMYK image ink color data corresponding to the respective CMYK ink colors. In step s1004, multivalued CMYK image data is quantized by error diffusion processing to generate image ink data indicating ejection or non-ejection of ink dots for each of CMYK. In step s1005, the character image data subjected to lossless compression is expanded to generate character image data. In step s1006, the character image data is converted into multi-value K character ink color data corresponding to K ink. In step s1007, the K character ink color data is quantized to generate character ink data indicating ejection or non-ejection of K ink dots. In step s1008, as described above, the image data is controlled so that the dots of the image ink color data at the boundary between the character portion and the image image portion are recorded so as to overlap the dots of the character ink color data. The detailed control method is as described with reference to FIG. In step s1009, the image ink data and the character ink data are combined to generate combined ink data. In step s1010, ink ejection from the recording head is controlled based on the synthetic ink data, and an image is recorded on the recording medium. Note that the processing in steps s1002 to s1004 and the processing in steps s1005 to s1007 may be performed in parallel after branching from step s1001, or the other processing may be performed after one processing is completed. There is no limit to the order.

Next, processing performed in each step will be described in detail. FIG. 6 is a processing example when drawing an input image including a character image and an image image based on PDL (page description language) in step s2001. Although PDL is generally described here, there are the following PDL entities, but any existing PDL may be used.
・ PostScript, PDF (Portable Document Format) (trade name, manufactured by Adobe Systems)
-XPS (XML Paper Specification) (trade name, manufactured by MicroSoft)
In step s20011, data described in PDL is received. In step s20012, the PDL is interpreted to create a list in which object drawing commands are arranged in the drawing order. In step s20013, if there are no more object drawing commands in the list, the process ends. If there is still an object drawing command, the process proceeds to step s20014. In step s20014, the object to be drawn this time is drawn in the input image page buffer of the RAM 202. Here, the drawing resolution is drawn at a resolution that is finally desired as the character quality. In this embodiment, image data to be drawn is 8-bit data with 256 values for each of RGB. However, it is not necessary to be limited to 8 bits, and higher gradation expression may be performed by setting a larger number of bits, for example, 16 bits.

In step s20015, it is determined whether the object drawn this time is a character attribute object. In this embodiment, a black character is determined as a character attribute object. Specifically, a pixel that satisfies all of the following conditions is determined to be a character attribute object, and a pixel that does not satisfy the condition is determined to be an image object.
Determination condition 1: The object to be drawn is not a bitmap image Determination condition 2: If the RGB value of the object to be drawn is not a character attribute R = G = B = 0, the process proceeds to step s20066. Advances to step s20017. In general, in PDL, an image image such as a photograph is described to be drawn as a bitmap image, and a character object is described in the form of a vector drawing command or a character code + font information. Therefore, attribute information (attribute data) of each pixel can be acquired based on data described in PDL by the above method. In step s20066, the attribute value of each pixel drawn this time is set to “image”, and the flow advances to step s2001 to proceed to processing of the next drawing object. In step s20017, the attribute value of each pixel drawn this time is set to “character”, and the flow advances to step s20013 to proceed to processing of the next drawing object. In this embodiment, the attribute value of the pixel corresponding to the character object is set to “character”, but the attribute value of the pixel corresponding to the line drawing object may be set to “character”. Further, it may be determined that a character object is determined when only one of the two conditions is satisfied, and an image object is determined when neither is satisfied.

  FIG. 7A shows an input image page buffer 2021 set in the RAM 202 of the host PC 200. The input image page buffer 2021 has an area corresponding to a predetermined number of horizontal pixels × vertical pixels, and the head pixel is 20211. Each pixel includes an RGB (Red, Green, Blue) value and an attribute value A (Attribute). The RGB value is data indicating color coordinates (R, G, B) in color space coordinates such as sRGB, which is an expression color of the monitor. In this case, in step s20014, color information is set in the RGB area of each pixel, and the attribute value A area is set in step s20066 or s20017.

  FIG. 7B is another example of the input image page buffer. Each of the input image page buffer 2022 and the attribute value buffer 2022a has an area corresponding to a predetermined number of horizontal pixels × vertical pixels, and the top pixels are 20221 and 2022a1. Each pixel of the input image page buffer 2022 is composed of RGB values, and each pixel of the attribute value buffer 2022a is composed of attribute values A. In this case, color information is set in the RGB area of each pixel in the input image page buffer 2022 in step s20014, and the attribute value A area in the attribute value buffer is set in step s20066 or step s20017. In addition, various embodiments such as holding R, G, B, A, and separate buffers are conceivable, but any method may be used. In the case of the buffer configuration shown in FIG. 7A, since only one address management is required, the control can be simplified and the program size can be reduced. 7B, when the number of bits representing the attribute value is smaller than the number of bits of the RGB value, the attribute value buffer size is reduced to reduce the amount of memory used. be able to.

  FIG. 8 shows a process of separating the black character image portion and the image image portion from the input image data in step s2002 to generate character image data and image image data. First, in step s20021, a process start pixel and an end pixel of the input image page buffer are set. In step s20022, it is determined whether the end pixel has been reached and there are no more pixels to process. If there are no more pixels to be processed, the process ends, and if there are, the process proceeds to step s20023. In step s20023, it is determined whether the pixel to be processed has a character attribute. If the determination result is not a character attribute, that is, an image attribute, the process proceeds to step s20024. If the determination result is a character attribute, the process proceeds to step s20025. In step s20024, the image pixel value in the input image page buffer is copied to the image image page buffer, and the process proceeds to step s20022 to process the next pixel. In step s20025, the character pixel value in the input image page buffer is copied to the character image page buffer, and the process proceeds to step s20022 to process the next pixel.

  FIG. 9 is a diagram showing the image image page buffer and the character image page buffer set in the RAM 202 in the host PC 200 in step s2002. The image image page buffer 2023 and the character image page buffer 2023T have the same number of horizontal pixels × vertical number of pixels as the input image page buffer, and the top pixels are 20231 and 2023T1. Each pixel of the image image page buffer 2023 is composed of RGB values, and each pixel of the character image page buffer 2023T is composed of character information T. At this time, in step s20024, color information is set in the RGB area of each pixel in the image image page buffer 2023, and character information T in the character image page buffer 2023T is set in step s20025.

  In the present embodiment, the character information T in the character image page buffer 2023T can have a smaller number of bits than the RGB value. The RGB value in the input image page buffer of the pixel in which the character information T is set in the character image page buffer 2023T is “R = G = B” according to “determination condition 2” when determining whether or not the object is a character object in step s20015. This is because “= 0”. Therefore, it can be expressed in 1 bit.

  Returning to FIG. 4, processing for compressing data and transferring it to the printer 100 and processing performed on the printer 100 side will be described.

  In step s2004, the image data generated in s2002 is irreversibly compressed. In this embodiment, the compression method defined by JPEG (Joint Picture Expert Group) is used, but any compression method may be used. The reason why lossy compression is selected is that the deterioration of details is not as conspicuous compared to the original characters. It should be noted that the compression processing is not an essential component for implementing the present invention.

  In step s2005, the character image data generated in s2002 is reversibly compressed. In the present embodiment, the RL (Run Length) compression method is used, but any compression method may be used. The reason why the lossless compression is used is that the character image is required to have detail reproducibility as compared with the image image. Further, as in step s2004, the compression processing is not an essential configuration. In this embodiment, a black character is determined as a character object, and the character information T of each pixel constituting the character image page buffer 2023T is 1 bit. Therefore, at least the data amount is 1/24 compared to RGB data. . Therefore, compression may not be required depending on the balance between the print speed and the transfer speed.

  In step s2006, the compressed image image data and character image data are sent to the printer 100. As a compression method, if lossless compression is applied to an image and irreversible compression is applied to a black character portion, the amount of transferred data can be further reduced while maintaining black character quality. Thus, since the image image and the character image are separate data, it is possible to apply a suitable compression method to each.

  In step s1001, the compressed image image data and character image data are received from the host PC 200. The received image data is expanded in step s1002, and the compressed character image data is expanded in step s1005. If the host PC 200 did not perform compression processing, the image image data skips step s1002 and proceeds to step s1003, and the character image data skips step s1005 and proceeds to step s1006.

  In step s1002, the irreversible compressed image is developed. The decompression method here corresponds to the lossy compression method performed in step s2004 of the host PC 200. In this embodiment, JPEG compressed data is decoded. The decoded data is held in an image image buffer (not shown) in the RAM 212 in the printer 100 in the same format as the image image buffer 2024 described in FIG.

  In step s1003, the image data developed in step s1002 is converted into multivalued CMYK image ink color data corresponding to the ink color of the printer 100 (ink color separation processing). Each pixel value (RGB value) of the image image data is data indicating color coordinates (R, G, B) in a color space coordinate such as sRGB that is an expression color of the monitor. This is converted into multi-value density data corresponding to the ink color (CMYK) of the printer by a known method such as matrix calculation processing or processing using a three-dimensional LUT. Since the printer 100 of this embodiment uses cyan (C), magenta (M), yellow (Y), and black (K) ink, the RGB signal image data is 8 bits 256 each of C, M, Y, and K. It is converted into image data consisting of color signals of values. In order to increase the color reproduction range and realize a high-quality color image, the value is normally set so that dots can be recorded up to the maximum acceptable amount that the recording medium can accept. It is possible to optimize so as not to cause image adverse effects within the maximum acceptance amount. In addition, although four inks of CMYK are given as an example, in order to improve the image quality, for example, light cyan (Lc), light magenta (Lm), gray (Gy) and other light inks, red (R), Special color inks such as green (G) and blue (B) may be additionally used.

  FIG. 10A shows the CMYK ink color density buffer for the image portion and the K ink color buffer for the character portion set in the RAM 212 in the printer 100. The CMYK ink color density buffer for the image image portion includes a C ink color density buffer 2121C, an M ink color density buffer 2121M, a Y ink color density buffer 2121Y, and a K ink color density buffer 2121K. The first pixel of the C ink color density buffer 2121C is 2121C1, and the density value of C ink is set for each pixel. The same applies to M ink, Y ink, and K ink. Each of the four ink color density buffers has the same number of horizontal pixels × vertical number of pixels as the image image page buffer 2023, and each pixel is 8-bit 256-value multi-value information. The RGB data of the first pixel 20231 of 2023 in FIG. 9 is converted into CMYK ink color data in step s1003, and the conversion result is 2121C1, 2121M1, 2121Y1, FIG. 2121K11. The same applies to other pixels.

  In step s1004, the CMYK image ink color data corresponding to the image portion is quantized. As a result, the gradation is reduced, and binary image ink data is generated for each of CMYK. As a quantization method, not only an error diffusion method but also any means such as a dither method may be used.

  FIG. 10B shows a CMYK ink data buffer for the image portion and a K ink data buffer for the character portion set in the RAM 212 in the printer 100. The CMYK ink data buffer for the image image portion includes a C ink data buffer 2122C, an M ink data buffer 2122M, a Y ink data buffer 2122Y, and a K ink data buffer 2122K1. The leading pixel of the C ink data buffer 2122C is 2122C1, and information indicating ejection or non-ejection of C ink is set in each pixel. The same applies to M ink, Y ink, and K ink. All of the four ink data buffers have the same size as the number of horizontal pixels and the number of vertical pixels of the input image page buffer 2021. One pixel of the ink data buffer corresponds to each nozzle of the recording head.

  FIG. 11 shows an example of quantization in the present embodiment. FIG. 11A shows an example of a quantization pattern when each ink color data is quantized in step s1004.

  In step s1005, the character image data is generated by expanding the losslessly compressed character image. The decompression method here corresponds to the lossless compression method performed in step s2005 of the host PC 200. In the present embodiment, RL compressed data is decoded. The decoded data is held in a character image page buffer (not shown) in the RAM 212 in the printer 100.

  In step s1006, the character image data is converted into K multi-value character data. In the present embodiment, the character image data is data indicating R = G = B = 0 in color space coordinates such as sRGB which is an expression color of the monitor. This is converted into K ink color density data (K) of the printer by a known method such as matrix calculation processing or processing using a one-dimensional LUT. In order to improve the density of black characters, the value is usually set so that dots are recorded up to the maximum acceptance amount that the recording medium can accept. It is possible to optimize so as not to cause image adverse effects within the maximum acceptance amount. Further, although the number of inks is one color K, other inks such as a light gray (Gy) ink may be added to improve image quality.

  Also, the K ink color buffer 2121K2 for the character portion in FIG. 10A has the same area as the number of horizontal pixels × the number of vertical pixels as the input image page buffer, and the top pixel is 2121K21. The character information T of the head pixel 2023T1 of the character image page buffer 2023T shown in FIG. 9 is converted into multi-value K ink color data in step s1006, and the conversion result is stored in 2121K21 of FIG. The character information T is 0 or 255, and the K ink color data after conversion is multi-value data of 8 bits 256 values. In this way, by controlling the density of the K ink in the character portion with multiple values, it is possible to control an appropriate ink recording amount according to the recording medium.

  In step s1007, the K ink color data is quantized to generate K character ink data. FIG. 10B shows a K ink data buffer 2122K2 for the character image portion. The first pixel is 2122K2, and information indicating ejection or non-ejection of K ink is set in each pixel. This ink data buffer has the same size as the number of horizontal pixels and the number of vertical pixels of the input image page buffer 2021. Note that when character image data is separated (generated) with binary 1-bit data in step s2002, the binary 1-bit data is held as it is until s1006, and thus binarization processing is performed to reduce the number of gradations. This step s1007 can be omitted. As will be described in detail later, when separating an object having a character attribute other than black (R = G = B = 0) as a character part, CMYK ink data of 2 bits or more having a gradation number of 3 or more is obtained in step s1006. Since it is generated, quantization here is required. In this case, at least one of the image image portion in step s1004 and the character image portion in step s1007 is quantized by error diffusion processing, thereby causing overlapping of recording dots. Usually, a higher quality image can be recorded by using error diffusion than dithering. FIG. 11B shows an example of a quantization pattern when the K ink color data is quantized in step s1007.

  In step s1008, dot overlap is controlled based on the image ink data of the image portion generated in step s1004 and the character ink data of the character portion generated in step s1007. Specifically, as described with reference to FIG. 4B, the pixel that becomes the recording dot On at the boundary with the character portion in the image ink data is overlapped with the K ink data in the character image portion on the recording medium. The data shifted by one pixel is generated.

  FIG. 12 is a flowchart showing a method of controlling the image data so that the recording dots overlap at the boundary between the image image portion and the character image portion in step s1008. In step s121, it is determined whether character edge detection has been completed for all pixels for character ink data in the character image portion. In the following, steps s122 to s124 are examples in which data for C ink is focused on among the image ink data. The same process is performed for inks of colors other than C. In step s122, edge pixels of the character image are detected in the character ink data.

  Here, an example of applying a Sobel filter will be described as a method for detecting a character edge. FIG. 13 is a coefficient matrix of a Sobel filter, FIG. 13A is a horizontal coefficient matrix, and FIG. 13B is a vertical coefficient matrix. The color density value of nine pixels, which is the sum of the target pixel having K quantization data in the character image portion and the surrounding eight pixels centering on the target pixel, is multiplied by the horizontal coefficient matrix, and the results are totaled. Similarly, the color density value of 9 pixels is multiplied by the coefficient matrix in the vertical direction, and the results are summed.

  When the total value in the horizontal direction is Fh and the total value in the vertical direction is Fv, the total value of a certain target pixel is Fs, and the Fs value can be obtained by the following equation (1).

  When the threshold Th for determining as an edge portion is used, it is determined that the target pixel is an edge pixel when Th <Fs. As an example, the Sobel filter is applied to detect edge pixels that are the edges of the character, but the detection of the edge portion of the character is not limited to this method.

  Returning to FIG. 12, in step s123, it is determined whether or not the recording dot is On in the pixel in the image portion corresponding to the same position on the recording medium as the pixel adjacent to the pixel detected as the edge portion of the character. FIG. 14 shows the result of determining whether the recording dot of the image portion adjacent to the pixel detected as the edge portion of the character is On. 14A and 14B are data buffers for recording at the same position on the recording medium. FIG. 14A is a part of the K ink data buffer 2122K2 for the character image portion, and each pixel is On / Off of the recording dot. Further, the pixels painted with diagonal lines are pixels detected as the edge portion of the character in step s122. FIG. 14B is a part of the C ink data buffer 2122C of the image portion, and each pixel is On / Off of the recording dot. Pixel value 1 is recording dot On, and pixel value 0 is recording dot Off. Pixels painted with diagonal lines are pixels that are adjacent to the edge of the character image portion and whose recording dots in the image image portion are On.

  Returning to FIG. 12, in step s124, the pixel corresponding to the same position on the recording medium as the pixel adjacent to the pixel determined to be the edge of the character for which the recording dot is On for the image image portion data. Set the recording dot to Off. Then, processing is performed to set the recording dot of the pixel adjacent to the pixel set to Off, which is determined to be the edge of the character, to the pixel at a position overlapping on the recording medium to On. FIG. 14C shows data of the image portion after this processing is performed. This is a pixel in which the recording dots are turned on so that the pixels painted with diagonal lines overlap the edge of the character.

  Returning to FIG. 5, in step s1009, the K ink data of the character image portion and the CMYK ink data of the image image portion generated in step s1008 are combined to generate combined CMYK ink data. The combined CMYK ink data buffer has the same configuration as the CMYK ink data buffer for the image portion in FIG. As a specific processing example, the logical sum of the K ink data in the K ink data buffer for the image image and the K ink data in the K ink data buffer for the character image is taken, and the K ink data of the combined CMYK ink data Is generated. Note that the C, M, and Y ink data is the same as the data in the CMYK ink data generated in step s1008.

  In step s1010, the composite CMYK ink data is output, the recording head is controlled based on the output data, and an image is recorded on the recording medium. In the present embodiment, an image is recorded by controlling a recording mechanism including the recording heads 101 to 104 shown in FIG.

  As described above, in the present embodiment, among the pixels to which ink is applied as the image image portion, ink is not applied to the pixels adjacent to the character image portion, and one pixel adjacent to the edge of the character image portion. The process of moving data is performed. With such a configuration, it is possible to reduce white spots while avoiding bleeding at the boundary between the image image and the character image.

  In addition, when black ink is used at the boundary between the image image portion and the character image portion, it is difficult to visually recognize the bleed. For this reason, the above-described processing may not be performed for the black ink in the image image portion. In other words, the above processing is not necessary for C, M, and Y ink data, and the above processing is not necessary for black ink. Further, in a form using a gray ink having the same color as that of the black ink and having a low color material concentration, it is preferable to carry out the above-described processing because bleeding may be visually recognized. In addition, when the bleed can be avoided by the material contained in the ink, the above-described processing may be performed only for the ink that is likely to bleed.

  Further, when the recording dots are turned off in all ink colors of the image image portion adjacent to the edge portion of the character and the recording dot of the adjacent pixel is turned on so as to overlap the edge portion of the character, the amount of shots is locally increased. There is a possibility of overflowing. For this reason, the above-described processing may be performed within a range in which the driving amount does not overflow. For example, the number of dots hitting the edge portion is counted, and the above processing is performed preferentially from ink colors that are likely to cause bleeding based on the counted result. At this time, if the count value reaches the threshold value, the subsequent ink color processing is not performed as a determination that ink may overflow.

  In addition, since the difference in driving time between adjacent pixels differs between the image image portion and the character image portion according to the conveyance speed of the recording medium, the degree of occurrence of bleeding also changes. For this reason, the above-described processing may be performed when the conveyance speed of the recording medium at the time of image recording is equal to or higher than a predetermined speed, and the above-described processing may not be performed when it is lower than the predetermined speed.

The determination of whether the character is an image or an image may be made after converting RGB data into converted CMYK data or data corresponding to the ink color to be used. Further, whether or not the object to be drawn is a bitmap image is used as one of the character determination conditions, but the following conditions may be used. When the following conditions are satisfied, the character object is determined.
The drawing object is a character object. The drawing object is a vector drawing object. The drawing object is not a 1-bit binary bitmap object. The rendering intent in the PDF format is not Colorimetric. Although treated as a “character image”, it may be performed for colored characters. In this case, the process is the same. In addition, by processing the data of “single color ink pixel” and “multiple ink color pixel” at different resolutions, it is possible to form an image with high resolution that is less likely to cause color misregistration for characters formed with a single color of ink. . In this case, an example of the determination condition is specifically:
Judgment condition 1: The object to be drawn is not a bitmap image. Judgment condition 2: One of the colors of the drawn object is R, G, and B is 0, and the remaining two are 255.

  As preconditions here, when R = 0 and G = B = 255, only cyan ink is used, and when G = 0 and R = B = 255, only magenta ink is used, B = 0, R = G = In the case of 255, an image is formed only with yellow ink. In this case, the same processing as that for black characters is applied to each single color ink character data, and the combined image data is output. Thereby, the same high-definition characters can be realized not only for black characters but also for single-color ink characters. The determination condition may be determined to be a character when only one of the conditions is satisfied.

  In this embodiment, the character image and the image image are separated depending on whether or not the pixel indicating the character attribute is a black character image of R = G = B = 0. However, the present invention can also be applied to character images other than black characters. It is. A case where color characters are separated as character objects will be described with reference to the flowchart of FIG. The flow for performing the same processing as that for black characters is omitted.

  First, in step s2001 in FIG. 5, the determination condition in s20015 in FIG. 6 is set to “determination condition 1: the object to be drawn is not a bitmap image” as the determination condition for separating the character portion and the image portion. That is, “determination condition 2: the RGB value of the object to be drawn is R = G = B = 0” is not adopted as the determination condition. Then, the attribute value of a pixel of a character other than black is also “character” and is set in the attribute value buffer. Next, when separating the character portion and the image image portion in step s2002, the pixel values are copied to the image image buffer and the character image buffer, respectively. In the above-described method for determining a black character as “character”, a pixel of R = G = B = 0 is used as a character, so the character image buffer can be set to 1 bit. However, here, a color character is also “character”. Since the determination is made, the RGB value of each pixel is copied in 8 bits. In step s1006, when converting into multi-valued CMYK ink color data, data of each CMYK is generated. That is, in the character image buffer of FIG. 10, in addition to the K ink color buffer 2121K2, multi-value data of 8-bit 256 values is stored in the C ink color buffer 2121C2, the M ink color buffer 2121M2, and the Y ink color buffer 2121Y2. In step s1007, 256-value CMYK ink color data is quantized, and the gradation is reduced to generate CMYK ink data for the image portion and CMYK ink data for the character portion. That is, an ink data buffer for the color image portion and an ink data buffer for the character portion are prepared for each of CMYK, and data indicating ink ejection or non-ejection is stored.

  Next, in step s1008, processing for controlling the overlap of the recording dots of the CMYK ink data in the character portion and the CMYK ink data in the image portion is performed. In FIG. 12 described above, character edge detection is performed on quantized K character ink data, but here edge detection is also performed on CMY quantized ink data. Next, it is determined whether the recording dot of the CMYK ink data buffer of the image image portion adjacent to the character edge portion of the character portion C ink data buffer is On. When comparing with the character C ink data buffer, the recording dot of the MYK print data buffer of the image image portion is turned off, and the recording dot of the pixel of the image image portion corresponding to the edge of the character image portion is turned on. The C ink data buffer in the image image portion turns the recording dot off. This is to prevent the occurrence of bleed and ink with recording dots in the color image portion turned on so as to overlap the character portion.

  The same processing is performed for the M ink data buffer, Y ink data buffer, and K ink data buffer in the character portion. As a result, control is performed such that the recording dots are ejected so that the M ink, Y ink, and K ink overlap in the image image data with respect to the pixels from which the C ink at the edge portion of the character image portion is ejected. Degradation can be suppressed.

  However, when comparing the Y ink data buffer of the character portion and the CMK ink data buffer of the image portion, if the recording dot of the pixel of the image portion that overlaps the edge of the character portion is set to On, the CMK ink is applied to the yellow character. Will be printed in layers. If the yellow character quality is deteriorated by performing such control, the CMK recording dots may not be moved, that is, may not be ejected in an overlapping manner.

(Second Embodiment)
In the above-described embodiment, the method for separating the input PDL data into the image image portion and the character image portion and controlling the CMYK ink data so that the recording dots in the image image portion and the recording dots in the character image portion overlap is described. . In recent years, as the printing speed of a printer increases and the resolution of an input image increases, the data transfer time or the data processing time on the printer side may become longer than the time for printing the input image. . Therefore, in this embodiment, a method for reducing the transfer capacity at the time of data transfer between the printer driver and the printer by reducing the resolution of the image portion and the amount of processing after data expansion on the printer side will be described. .

  In business documents and the like, the quality of black characters is very important, and printing of black characters with higher resolution is required. Therefore, the amount of data is reduced by reducing the resolution of the image portion.

  FIG. 15 is a flowchart illustrating an example of a configuration in which the resolution of the image portion is reduced in the present embodiment. The difference in the flowchart of FIG. 5 of the first embodiment is that step s2003 is added. First, the host PC 200 performs the processes of steps s2001 to s2006. Since the processes other than step s2003, step s2004, and step s2006 are the same as those in the first embodiment, they are omitted here.

  In step s2003, the resolution reduction processing is performed on the separated multi-valued image image data to generate low-resolution image image data. In step s2004, the low resolution image is irreversibly compressed. In step s2006, the compressed low-resolution image data and the character image data reversibly compressed in step s2005 are sent to the printer 100.

  Next, in the printer 100, the processes in steps s1001 to s1010 are performed. Differences in processing contents from the first embodiment will be described below. In step s1001, the compressed low-resolution image data and character image data are received from the host PC 200. In step s1002, the irreversibly compressed low resolution image data is developed. In step s1003, the developed multi-value low-resolution image data is converted into multi-value low-resolution CMYK ink color data. In step s1004, the multi-valued low-resolution CMYK ink color data is quantized to reduce the gradation, and binary CMYK ink data is generated. The character image portion uses CMYK ink, but in the following description, a black character using K ink will be described as an example.

  Next, a preferred processing example for each step will be described in detail. In step s2003, resolution reduction processing is performed on the separated image data.

  FIG. 16 shows an example of a low resolution image page buffer set in the RAM 202 in the host PC 200. The low-resolution image image page buffer 2024 is an area corresponding to half the number of horizontal pixels × the number of vertical pixels of the image image page buffer 2023, and the top pixel is 20241. The RGB values corresponding to the respective pixels of the low resolution image image page buffer 2024 are stored.

  In step s2003, the image in the image image page buffer 2023 is reduced, and a low resolution image is generated in the low resolution image page buffer. The solid line in the low resolution image image page buffer 2024 represents the boundary of the reduced image pixel, and the broken line represents the boundary of the image pixel in the corresponding image image page buffer of FIG. In the present embodiment, the reduction ratio is 1/2 and 1/2, and 2 × 2 pixels in the image page buffer correspond to one pixel in the low resolution image page buffer.

  As a specific reduction processing method, the average value of the pixel values of the first pixel 20231 and the right, lower, and lower right pixels of the image image page buffer in FIG. 9 is calculated, and the low resolution image image page buffer of FIG. Is the pixel value of the first pixel 20241. Similarly, reduction processing is performed on the pixels of all the low-resolution image image page buffers using the corresponding image image page buffer pixels. In the reduction process, the reduction rate is preferably set to 1 / integer. This is because, when the reduction processing is not a fraction of an integer, periodic noise such as moire occurs. Furthermore, it can be realized by an enlargement process of an integral multiple at the enlargement process performed in step s1004 in the printer 100 to be described later, the process is simple, and finally a desirable image quality can be obtained. More preferably, when the reduction ratio is set to a power of 1, the calculation can be performed at high speed by calculation by bit shift processing without using division when calculating the average value.

  Returning to the flow, in step s2004, the low resolution image is irreversibly compressed. In the present embodiment, the number of pixels has already been reduced to 1/2 in both the vertical and horizontal directions, and at least the data amount is 1/4. Therefore, depending on the balance between the print speed and the transfer speed, no further compression may be required. Further, when the data is reduced to 1/3 in the vertical and horizontal directions, the data amount is further reduced to 1/9, and when the data is reduced to 1/4 in the vertical and horizontal directions, the data amount becomes 1/16. In step s2006, the compressed image image data and character image data are sent to the printer 100.

  As described above, by using the low resolution image and the character portion that is not reduced, the amount of transfer data can be significantly reduced while maintaining the quality of black characters. Further, by applying reversible compression to the image and irreversible compression to the character part, it is possible to further reduce the amount of transferred data while maintaining the black character quality. Thus, since the image data and the character portion are separate data, it is possible to apply a suitable compression method.

  In step s1001, the compressed low-resolution image image data and character image data are received from the host PC 200. The compressed low-resolution image data is processed in step s1002, and the compressed character image data is processed in step s1005. If the host PC 200 has not performed compression processing, the low resolution image is skipped to step s1003 and the black character image is skipped to step s1005 and sent to s1006.

  In step s1002, the irreversible compressed low resolution image data is expanded to generate an image. The decompression method here corresponds to the lossy compression method performed in step s2004 of the host PC 200. In this example, JPEG compressed data is decoded. The decoded data is held in a low resolution image image page buffer (not shown) in the RAM 212 in the printer 100 in the same format as the low resolution image image page buffer 2024 shown in FIG.

  In step s1003, the low resolution image is converted into low resolution CMYK ink color data. RGB which is a low-resolution image image pixel value is data indicating color coordinates (R, G, B) in color space coordinates such as sRGB which is an expression color of the monitor. This is converted into ink color density data (CMYK) of the printer by a known method such as matrix calculation processing or processing using a three-dimensional LUT. Since the printer 100 of this embodiment uses black (K), cyan (C), magenta (M), and yellow (Y) ink, the RGB signal image data is 8 for each of K, C, M, and Y. It is converted into image data consisting of bit color signals. In addition, the number of inks has been exemplified by four colors K, C, M, and Y, but light cyan (Lc), light magenta (Lm), and gray (Gy) inks with low density are used to improve image quality. Other inks may be added.

  FIG. 17 shows the low-resolution CMYK ink color density buffer for the image portion and the K ink color buffer for the character portion set in the RAM 212 in the printer 100. The low-resolution CMYK ink color density buffer for the image image portion is a low-resolution C ink color density buffer 2127C, a low-resolution M ink color density buffer 2127M, a low-resolution Y ink color density buffer 2127Y, and a low-resolution K ink color density buffer 2127K. is there. The first pixel of the low-resolution C ink color density buffer 2127C is 2127C1, and stores the C ink color density value of each pixel. The same applies to M ink, Y ink, and K ink. Each of the four ink color buffers has the same number of horizontal pixels × vertical pixels as the low-resolution image image page buffer 2024, and each pixel is 8-bit multi-value information. The RGB values corresponding to the pixels one-to-one and corresponding to the head pixel 20241 of 2024 shown in FIG. 16 are converted into data corresponding to the ink color in step s1004, and 2127C1, 2127M1, 2127Y1, 2127K11 in FIG. Stored in

  In step s1004, the low-resolution CMYK ink color data is quantized to generate CMYK ink data. Any method such as error diffusion or dither may be used as the quantization method. The ink data after the quantization processing is stored in the ink data buffer in FIG. 10B, as in the first embodiment. Each of the CMYK ink data buffers in the image image portion of FIG. 10B has the same area as the number of horizontal pixels × the number of vertical pixels as the input image page buffer 2021. Compared to the low-resolution CMYK ink color data buffer, the number of pixels in both the horizontal and vertical directions is doubled. This is because each ink color data is quantized to 5 levels of 2 × 2 pixels in step s1004 of this embodiment. This magnification is determined in accordance with the reduction ratio at the time of the resolution reduction processing in step s2003 processed in the host PC 200.

  In step s1008, control is performed so that the recording dots do not overlap in the ink data of the character portion and the ink data of the image image portion. As in the first embodiment, for the pixels indicating ejection in the ink data of the character portion, if the ink data of the character portion and the image image portion is generated so as to indicate non-ejection in the ink data of the image image portion. Good.

  In this embodiment, the resolution reduction in step s2003 in the host PC 200 is 1/2 × 1/2, the enlargement ratio of the image portion in step s1004 in the printer 100 is 2 × 2, and the character portion in step s1006. The enlargement ratio is 1 × 1. The reduction ratio and the enlargement ratio do not necessarily have a reciprocal relationship. For example, the resolution reduction is 1/2 × 1/2, and the enlargement ratio of the image portion in step s1004 in the printer 100 is the same. 4 × 4, and the enlargement ratio of the character portion in step s1007 may be 2 × 2. That is, it is only necessary that the reduction ratio of the image portion and the character portion at the time of reduction and the enlargement ratio of the image portion and the character portion at the time of enlargement have an inverse relationship.

  As described above, in the present embodiment, the overlap between the recording dots based on the character image data and the recording dots based on the image image data is controlled by the difference in resolution between the character image and the image image. With the above configuration, it is possible to simultaneously reduce the amount of transferred data and the amount of processing on the printer side while maintaining character recording quality. Further, by moving the dots in the image image portion so as to overlap with the recording dots in the character portion, it becomes possible to suppress white spots in the boundary portion while reducing bleeding.

(Third embodiment)
In the present embodiment, a method for controlling the overlap of recording dots by referring to the ink data of the image portion and the ink data of the character portion will be described.

  FIG. 18 is a flowchart showing the configuration of the present embodiment. Step s1011 is added to FIG. 15 instead of step s1008. In step s1011, the overlap of the recording dots in the image portion and the character portion is controlled. Since the description other than step s1011 is the same as that of the above-described embodiment, the description thereof is omitted.

  FIG. 19 is a preferred processing example of step s1011. In step s191, it is determined whether quantization has been completed for all pixels for the CMYK ink color data in the image portion and the K ink color data in the character portion. In step s192, character edge detection is performed. The character edge detection method is the same as the method described in step s122 of FIG. However, in step s192, the filtering process is performed on the binary data of the recording dot On / Off of the K quantized data of the character portion. However, since the K ink page buffer is 8-bit multivalued data, step s122 The threshold value Th is different from the case of. In step s193, it is determined whether quantization has been completed for all pixels for the CMYK ink color data in the color image portion and the K ink color data in the character portion.

In step s194, the CMYK ink color data of the 1 × 1 pixel image portion is quantized to 5 levels. An example of ink color data and quantization level is shown below.
Ink color data: 0 or more and less than 32 → quantization level 0
Ink color data: 32 or more and less than 96 → quantization level 1
Ink color data: 96 or more and less than 160 → quantization level 2
Ink color data: 160 to less than 224 → quantization level 3
Ink color data: 224 or more and 255 or less → quantization level 4
FIG. 20 shows an example of the quantization pattern corresponding to the quantization level.
Quantization level 0 → quantization pattern 0
Quantization level 1 → quantization pattern 10-13
Quantization level 2 → quantization pattern 20-25
Quantization level 3 → quantization pattern 30 to 33
Quantization level 4 → quantization pattern 40
In step s195, quantization processing is performed so that the K ink color data of the character portion of 2 × 2 pixels is quantized to two levels. In the present embodiment, when the resolution of the image portion is reduced, the reduction ratio is set to 1/2 in both the vertical and horizontal directions. Therefore, the pixel size of the character portion corresponding to the image image portion 1 × 1 pixel is 2 × 2, and the character portion performs quantization processing on a total of 4 pixels of 2 × 2 pixels. Quantization level 0 is recording dot Off, and quantization level 1 is recording dot On.

In step s196, a quantization pattern is selected according to the CMYK quantization level of the image portion. FIG. 21 shows an example of selecting a quantization pattern. 2126C is a part of the C ink color density buffer 2121C of the image portion, and each pixel value is a color density value. Reference numerals 2126C11 to 2126C13 denote pixels. The ink color density value quantization level has the following relationship.
Ink color density 64 → quantization level 1
Ink color density 150 → quantization level 2
Ink color density 200 → quantization level 3
2126K2 is a part of the K ink color density buffer 2121K2 of the black character portion, and each pixel value is a color density value. Reference numerals 2126K11 to 2126K13 denote 2 × 2 pixel units in the black character portion corresponding to one pixel in the image portion. Pixels painted with diagonal lines are pixels that are determined to be the edge portion of the character in step s192. The ink density value 255 is a quantization level 1 pixel, and the ink density value is a quantization level 0 pixel. 2126K2 and 2126C are ink data buffers for recording at the same position on the recording medium.

  In step s196, the 1 × 1 pixel CMYK quantization pattern in the image image portion is compared with the 2 × 2 pixel recording dot pattern in the character portion K, and the recording dot in the image image portion and the recording dot in the character portion are compared. Are adjacent to each other.

  2126C11 has an ink density value of 64, a quantization level of 1, and 10 to 13 quantization patterns shown in FIG. 20 can be selected. The character portion 2126K11 corresponding to the recording position of the image image portion 2126C11 has an ink density value of 0, so the recording dot is Off. In the adjacent 2126K12, the lower left and lower right pixels are the recording dots On. In 2126C11, when the quantization pattern 13 is selected, bleeding occurs because the recording dots in the image portion and the recording dots in the character portion are adjacent to each other. In order to avoid bleeding, the quantization patterns 10 to 12 are selected. When there are a plurality of selectable quantization patterns, a random number is applied to select which quantization pattern is applied. When only the same quantization pattern is applied, a pattern such as a texture is seen and image degradation occurs. Therefore, it is preferable to select a quantization pattern at random by applying a random number or the like.

  Next, since the ink density value of 2126C12 is 150 and the quantization level is 2, 20 to 25 quantization patterns shown in FIG. 20 can be selected. In the character portion 2126K12 corresponding to the recording position of 2126C12 in the image image portion, the lower left and lower right pixels are recording dots On. In 2126C12, when 20 to 22 and 24 to 25 are selected as the quantization pattern, bleeding occurs because there are recording dots in which the recording dots in the image image portion and the recording dots in the character portion are adjacent to each other. In order to avoid bleeding, the quantization pattern 23 is selected. When the quantization pattern 23 is selected, the recording dots in the image image portion and the recording dots in the character portion overlap each other, so that a mechanical dot gain is generated, and white spots at the boundary between the image image portion and the character portion can be reduced. By selecting the quantization pattern in this way, it is possible to avoid bleeding and reduce white spots at the boundary between the image portion and the character portion.

  Next, 2126C13 has an ink density value of 200, a quantization level of 3, and a quantization pattern of 30 to 33 shown in FIG. 20 can be selected. In the black character portion 2126K13 corresponding to the recording position of 2126C13 in the image image portion, the lower left and lower right pixels are recording dots On. In 2126C13, when any of the quantization patterns 30 to 33 is selected, the recording dot in the image image portion and the recording dot in the character portion are adjacent to each other. In this case, the quantization patterns 30 to 31 are two dots that are adjacent to the recording dots of the image portion and the character portion, but the quantization patterns 32 to 33 are one dot that is adjacent. is there. In order to avoid bleeding, a quantization pattern with a small number of adjacent recording dots is selected.

  As described above, in this embodiment, when quantizing the image image data, the quantized character image data is referred to, so that the quantum of the image image data overlaps with the pixels of the recording dots On of the character image data. To do. As a result, it is possible to reduce white spots at the boundary while reducing bleeding that occurs at the boundary between the image portion and the character portion. It is also possible to reduce the amount of transfer data and the amount of processing on the printer side.

(Other embodiments)
In the above-described embodiment, image image data and character image data are separated under a predetermined condition, quantization is performed for each separated data, and edge detection is performed only for character image data. The present invention is not limited to this method. For example, an object edge may be detected based on multivalued image data such as RGB data captured by copying or the like, or binary data after quantization. As a method for detecting an edge, for example, a pixel having a predetermined value such as R = G = B = 0 may be determined as an edge, or another known method may be used. The recorded image is not limited to the above as long as the image data serving as the background of the detected object is moved to adjacent pixels so as to overlap the edge of the object.

Claims (12)

An image processing method for generating data for recording an image on a recording medium using a recording head for discharging ink droplets,
Character data for recording a character object based on image data to be recorded and indicating ink droplet ejection or non-ejection, and image data for recording an image object and ink droplet ejection Or a generation process for generating image data indicating non-ejection,
A detection step of detecting edge pixels corresponding to edges of the character object in the character data;
In the image data, data indicating non-ejection of ink droplets is set for pixels corresponding to pixels adjacent to the edge pixels of the character data indicating ink droplet ejection, and adjacent to the set pixels. And a setting step of setting data indicating ejection of ink droplets for pixels corresponding to the edge pixels of the character data.   The image processing method according to claim 1, wherein the character object includes a character and a line drawing.   In the generating step, a pixel satisfying a predetermined condition is determined as a pixel corresponding to a character object, a pixel not satisfying the predetermined condition is determined as a pixel corresponding to an image object, and the character data is based on the determination result The image processing method according to claim 1, wherein the image data is generated.   The image processing method according to claim 3, wherein a pixel value satisfying the predetermined condition has a pixel value of R = G = B = 0.   The image processing method according to claim 3, wherein the pixel satisfying the predetermined condition is not a bitmap object.   The image processing method according to claim 3, wherein the pixel satisfying the predetermined condition is a vector drawing object.   4. The image processing method according to claim 3, wherein one of a pixel value satisfying the predetermined condition has a pixel value of 0 among R, G, and B, and the remaining two are 255. 5.   8. The recording head according to claim 1, wherein the recording head is capable of ejecting ink droplets of a plurality of inks, and the image data generated in the generating step includes a plurality of data corresponding to the plurality of inks. The image processing method according to claim 1.   9. The image processing method according to claim 8, wherein the plurality of inks include black ink, and the character data is data for recording black characters using the black ink. An image processing apparatus that generates data for recording an image on a recording medium using a recording head that discharges ink droplets,
Character data for recording a character object based on image data to be recorded and indicating ink droplet ejection or non-ejection, and image data for recording an image object and ink droplet ejection Or acquisition means for acquiring image data indicating non-ejection,
Detecting means for detecting edge pixels corresponding to edges of the character object in the character data;
In the image data, data indicating non-ejection of ink droplets is set for pixels corresponding to pixels adjacent to the edge pixels of the character data indicating ink droplet ejection, and adjacent to the set pixels. And setting means for setting data indicating ejection of ink droplets to the pixels corresponding to the edge pixels of the character data. A recording apparatus for recording an image by applying ink onto a recording medium using a recording head that discharges ink droplets,
When an image image is recorded on the recording medium at a position adjacent to the character image, the ink droplet for forming the image image should be applied to a position adjacent to the edge of the character image. An ink droplet is applied so as to overlap the edge of the character image. A recording method for recording an image by applying ink onto a recording medium using a recording head that discharges ink droplets,
When an image image is recorded on the recording medium at a position adjacent to the character image, the ink droplet for forming the image image should be applied to a position adjacent to the edge of the character image. An ink droplet is applied so as to overlap the edge of the character image.

Images