Image processing device, image processing method, and program

Abstract

To suppress subsidiary reduction in image quality while improving sharpness of a printed image such as a character or line.SOLUTION: Recording data is generated, the recording data being used by a recording device that records an image in a recording medium by attaching a recording agent of one or more colors to the recording medium according to the recording data of one or more colors. An edge in an N-value image (N is a natural number of 2 or more) is detected, the N-value image indicating a result of threshold processing on a gray image obtained from an input image. The recording data is generated on the basis of the input image and a detection result of the edge.SELECTED DRAWING: Figure 17

Description

[Technical field]

[0001] The present invention relates to an image processing apparatus, an image processing method, and a program, and more particularly to an inkjet printing apparatus having a print head that performs printing by ejecting ink. [Background technology]

[0002] In inkjet printers, a technique for suppressing bleeding caused by contact between multiple ink droplets is known. For example, Patent Document 1 discloses thinning out black pixels and color pixels located at the boundary between a black image portion and a color image portion to every other dot. According to the method of Patent Document 1, contact between black ink droplets and color ink droplets is suppressed, so that ink bleeding is reduced. Furthermore, Patent Document 2 discloses performing multi-pass printing so that recording of high density areas at the edge portion is biased to one pass, and recording of low density areas at the edge portion is biased to one pass. According to the method of Patent Document 2, the arrival timing of ink droplets is separated between adjacent areas at the edge portion, so that bleeding caused by contact between ink droplets is reduced. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 6-152902 [Patent Document 2] JP 2019-72890 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the method of Patent Document 1 has a problem in that the outlines of characters and lines printed in black ink become faint, reducing visibility, and the method of Patent Document 2 has a problem in that white areas tend to appear at edges due to misalignment of printing positions between passes.

[0005] An object of the present invention is to improve the sharpness of images such as printed characters or lines while suppressing secondary degradation of image quality. [Means for solving the problem]

[0006] An information processing device according to an embodiment of the present invention has the following configuration. 1. An image processing device for generating recording data for use in a recording device that records an image on a recording medium by depositing one or more colors of recording material on the recording medium in accordance with one or more colors of recording data, A detection means for detecting edges in an N-value image (N is a natural number equal to or greater than 2) that indicates the result of threshold processing on a grayscale image obtained from an input image; a generating means for generating the recording data based on the input image and the edge detection result; Equipped with. Effect of the Invention

[0007] It is possible to improve the sharpness of images such as printed characters or lines while suppressing secondary degradation of image quality. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a perspective view of a recording apparatus according to an embodiment. [Diagram 2] FIG. 1 is a diagram showing an example of the hardware configuration of an image processing apparatus according to an embodiment. [Diagram 3] 1 is a flowchart of an image processing method according to an embodiment. [Figure 4] FIG. 4 is a diagram showing an example of a dot arrangement at an edge portion. [Diagram 5] FIG. 4 is a diagram showing an example of input image data. [Figure 6] FIG. 4 is a graph showing the relationship between ink amount and lightness. [Figure 7] 5A to 5C are diagrams for explaining edge pattern detection. [Figure 8] 6A and 6B are diagrams for explaining rotation and phase shift of pattern matching data. [Figure 9]5A to 5C are diagrams illustrating a tone correction process. [Figure 10] 4 is a flowchart of a color separation process and a quantization process. [Figure 11] FIG. 13 is a diagram for explaining an index expansion process. [Figure 12] 4A and 4B are diagrams for explaining conversion of an input image into a luminance image. [Figure 13] FIG. 4 is a diagram for explaining conversion of a luminance image into a binary image. [Figure 14] FIG. 11 is a diagram showing the results of edge detection processing. [Figure 15] FIG. 13 is a diagram showing an example of quantized data. [Figure 16] 5A to 5C are diagrams showing examples of dot arrangement patterns. [Figure 17] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 18] 1 is a flowchart of an image processing method according to an embodiment. [Figure 19] FIG. 13 is a diagram for explaining an index expansion process. [Figure 20] FIG. 11 is a diagram showing the results of edge detection processing. [Figure 21] FIG. 13 is a diagram showing an example of quantized data. [Figure 22] 5A to 5C are diagrams showing examples of dot arrangement patterns. [Diagram 23] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 24] FIG. 13 is a diagram showing an example of quantized data. [Diagram 25] 5A to 5C are diagrams showing examples of dot arrangement patterns. [Figure 26] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 27] FIG. 2 is a diagram showing a schematic configuration of a recording head. [Figure 28] FIG. 13 is a diagram showing an example in which the third end portion is continuous. [Figure 29] 1 is a flowchart of an image processing method according to an embodiment. [Diagram 30] 4A to 4C are diagrams for explaining ink color separation. [Diagram 31] 4 is a flowchart of a color separation process and a quantization process. [Diagram 32] 5A to 5C are diagrams illustrating a tone correction process. [Diagram 33] FIG. 13 is a diagram for explaining an index expansion process. [Diagram 34] 6A and 6B are diagrams showing examples of binary images corresponding to ink amount information. [Diagram 35] FIG. 11 is a diagram showing the results of edge detection processing. [Diagram 36] FIG. 11 is a graph showing density values ​​after tone correction. [Figure 37] FIG. 13 is a diagram showing an example of quantized data. [Figure 38] 5A to 5C are diagrams showing examples of dot arrangement patterns. [Figure 39] FIG. 13 is a diagram showing an example of a dot arrangement. [Diagram 40] FIG. 4 is a diagram for explaining multi-pass printing. [Diagram 41] FIG. 13 is a diagram for explaining an index expansion process. [Diagram 42] A diagram showing an example of index data [Diagram 43] FIG. 11 is a diagram showing an example of a mask pattern. [Diagram 44] FIG. 13 is a diagram showing an example of a mask decode table. [Diagram 45] 5A to 5C are diagrams showing examples of dot arrangement patterns. [Diagram 46] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 47] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 48] FIG. 11 is a diagram showing an example of threshold setting. [Figure 49-1] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 49-2] FIG. 13 is a diagram showing an example of a dot arrangement. [Figure 49-3] FIG. 13 is a diagram showing an example of a dot arrangement. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, the embodiments will be described in detail with reference to the attached drawings. Note that the following embodiments do not limit the invention according to the claims. Although the embodiments describe a number of features, not all of these features are essential to the invention, and the features may be combined in any manner. Furthermore, in the attached drawings, the same reference numbers are used for the same or similar configurations, and duplicated descriptions are omitted.

[0010] <Structure of the recording device> The structure of the recording apparatus in each embodiment will be described below with reference to Fig. 1. Fig. 1 is a perspective view showing an overview of a recording unit in a printer 2 which is a recording apparatus according to one embodiment.

[0011] The recording medium P (hereinafter also referred to simply as the recording medium) fed to the recording section is transported in the -Y direction (sub-scanning direction) by the nip portion between a transport roller 101 arranged on the transport path and a pinch roller 102 driven by the transport roller 101 as the transport roller 101 rotates.

[0012] The platen 103 is provided at a recording position opposite the surface (nozzle surface) on which nozzles are formed of the recording head H of the inkjet recording method, and supports the back surface of the recording medium P from below, thereby maintaining a constant distance between the surface of the recording medium P and the nozzle surface of the recording head H.

[0013] The recording medium P in the area where recording has been performed on the platen 103 is nipped between the discharge roller 105 and a spur 106 driven by the discharge roller 105, transported in the -Y direction as the discharge roller 105 rotates, and discharged onto the discharge tray 107.

[0014] The print head H is detachably mounted on a carriage 108 with its nozzle surface facing the platen 103 or a print medium. The carriage 108 is reciprocated in the X direction, which is the main scanning direction, along two guide rails 109 and 110 by the driving force of a carriage motor (not shown), and during this movement the print head H performs an ejection operation in response to an ejection signal.

[0015] The ±X directions in which the carriage 108 moves are directions that intersect with the -Y direction in which the recording medium is transported, and are called the main scanning direction. In contrast, the -Y direction in which the recording medium is transported is called the sub-scanning direction. An image is formed in stages on the recording medium P by alternating between main scanning (movement accompanied by ejection) of the carriage 108 and the recording head H and transport of the recording medium (sub-scanning). The above is the content regarding the structure of the recording device in this embodiment.

[0016] A recording device according to an embodiment records an image on a recording medium by applying one or more colors of recording material to the recording medium according to recording data for one or more colors. The recording device described below records an image on the recording medium by applying multiple colors of recording material to the recording medium. Specifically, the recording device records an image by applying cyan ink, magenta ink, yellow ink, and black ink to the recording medium according to nozzle data for cyan (C), magenta (M), yellow (Y), and black (K). As described below, the nozzle data corresponds to recording data.

[0017] <Recording head structure> The structure of the print head in this embodiment will be described below with reference to FIG. 27. FIGS. 27(a) to 27(c) are schematic diagrams of the nozzle surface of the print head H used in this embodiment as viewed from the +Z direction. The print head H has print chips 2705 and 2706, each of which receives a print signal from the printing apparatus body via a contact pad (not shown), and is supplied with power required to drive the print head. As shown in FIG. 27(a), the print chip 2705 is provided with a nozzle row 2701 (hereinafter also referred to as a black nozzle row) in which a plurality of nozzles for ejecting black ink are arranged in the Y direction. Similarly, the print chip 2706 is provided with a nozzle row 2702 for ejecting cyan ink, a nozzle row 2703 for ejecting magenta ink, and a nozzle row 2704 for ejecting yellow ink.

[0018] FIG. 27(b) is an enlarged view of the black nozzle row 2701. FIG. 27(c) is an enlarged view of one of the three nozzle rows, i.e., the nozzle rows 2702, 2703, and 2704, i.e., the cyan, magenta, and yellow nozzle rows. This enlarged view is common to all color inks. On both sides of the ink liquid chambers 2707 and 2710, nozzles 2708 and 2711 that eject ink are arranged. Directly below (on the +Z direction side) each nozzle are ejection heaters 2709 and 2712. When a voltage is applied to both of the ejection heaters 2709 and 2712, they generate heat to generate bubbles, causing ink to be ejected from each nozzle. The nozzle 2708 has 832 nozzles, and the nozzle 2711 has 768 nozzles. The nozzle 2708 ejects black ink, and is arranged in an Ev row and an Od row arranged in the Y direction at a pitch of 600 dpi. The Ev row is shifted by half a pitch in the -Y direction from the Od row. By performing a print scan using the black nozzle row 2701 configured in this way, it is possible to print on the print medium at a print density of 1200 dpi. The cyan nozzle row 2702, magenta nozzle row 2703, and yellow nozzle row 2704 have the same configuration as the black nozzle row 2701.

[0019] The print head H of this embodiment has a configuration including a print chip having a black nozzle row, and a print chip having a cyan nozzle row, a magenta nozzle row, and a yellow nozzle row, but it is not necessary to be limited to this configuration. Specifically, the black nozzle row, the cyan nozzle row, the magenta nozzle row, and the yellow nozzle row may all be mounted on one chip. Alternatively, a print head mounting a print chip having a black nozzle row and a print head mounting a print chip having a cyan nozzle row, a magenta nozzle row, and a yellow nozzle row may be separate. Alternatively, the black nozzle row, the cyan nozzle row, the magenta nozzle row, and the yellow nozzle row may each be mounted on a separate print head. In addition, the print head H of this embodiment is a so-called bubble jet type that applies a voltage to a heater to generate heat to eject ink, but it is not necessary to be limited to this configuration. Specifically, it may be a configuration in which ink is ejected using an electrostatic actuator or a piezoelectric element.

[0020] The above is the content regarding the structure of the recording head in this embodiment.

[0021] FIG. 2(a) is a diagram showing an example of the configuration of a printing system including an image forming apparatus 10 equipped with a printer 2. FIG. 2(a) shows a cloud print system in which a terminal apparatus 11, a cloud print server 12, and an image forming apparatus 10 are connected via a network 13. The cloud print server 12 is a server apparatus that provides a cloud print service. That is, in the configuration of FIG. 2(a), the image forming apparatus 10 is a cloud print-compatible printer. The network 13 is a wired network, a wireless network, or a network including both. The network 13 may be, for example, the Internet, a WAN, or a VPN environment. However, the printing system is not limited to the cloud print system. For example, the network 13 may be configured as an in-house LAN, or the terminal apparatus 11 and the image forming apparatus 10 may be configured to be directly connected without going through the network 13. FIG. 2(a) shows one each of the terminal apparatus 11 and the image forming apparatus 10, but there may be a plurality of the terminal apparatus 11 and the image forming apparatus 10. The cloud print server 12 may also be a server system configured of a plurality of information processing apparatuses. Furthermore, the printing system may be a cloud print system in which a plurality of cloud print services cooperate with each other.

[0022] The terminal device 11 is an information processing device such as a PC, tablet, or smartphone, and has a cloud printer driver for the cloud print service installed. A user can run any application software on the terminal device 11. For example, a print job and print data are generated via the cloud printer driver based on image data generated on a print application. The print job and print data are transmitted to an image forming device 10 registered in the cloud print service via a cloud print server 12. The image forming device 10 is a device that prints on a recording medium such as paper, and prints an image on the recording medium based on the received print data.

[0023] <Control system configuration> The configuration of the control system in this embodiment will be described below with reference to FIG. 2(b). FIG. 2(b) is a schematic diagram of the image processing device 100. In this embodiment, the image processing device 100 will be described as being included in the image forming device 10. However, the image processing device 100 may be configured as a device connected to the image forming device 10 including the printer 2 and the scanner 202. For example, the image processing device 100 may be configured inside the host computer 201. In that case, the image processing device 100 may not include the printhead control unit 213 and the scanner IF control unit 205.

[0024] The host computer 201 is an information processing device that creates print jobs consisting of input image data and recording condition information required for recording, and corresponds to, for example, the terminal device 11 in Fig. 2(a). Note that the recording condition information is information such as the type and size of recording paper, and recording quality.

[0025] The scanner 202 is a scanner device connected to the image processing apparatus, which reads information on a document placed on the scanner table, generates analog data, and converts the data into digital data via an AD converter. The execution of reading by the scanner 202 is controlled by sending a scan job from the host computer 201 to the image processing apparatus 100, but is not limited to this. It may also be replaced by a dedicated UI device connected to the image processing apparatus 100 or the scanner 202.

[0026] The ROM 206 is a readable memory that stores a program for controlling the image processing device 100 .

[0027] The CPU 203 controls the image processing device 100 by executing a program stored in the ROM 206 .

[0028] A host IF control unit 204 communicates with the host computer 201 , transmits print jobs, and stores them in a RAM 207 .

[0029] The RAM 207 is a readable and writable memory used as an execution area for programs and a storage area for data.

[0030] The image processing unit 208 generates nozzle data broken down into printable nozzle units from the input image data stored in the RAM 207 in accordance with the printing conditions included in the print job. The generated nozzle data is stored in the RAM 207. The image processing unit 208 is internally composed of a decoder unit 209, a scan image correction unit 216, an image analysis unit 210, a color separation quantization unit 211, and a nozzle separation processing unit 212.

[0031] The printhead control unit 213 controls the printhead H in the printer 2 based on control data obtained based on the nozzle data stored in the RAM 207 .

[0032] The shared bus 215 is connected to each of the processing units, the CPU 203, the host IF control unit 204, the scanner IF control unit 205, the ROM 206, the RAM 207, and the image processing unit 208. The connected units can communicate with each other through the shared bus 215. The above is the content of the configuration of the control system in this embodiment.

[0033] The image processing device according to an embodiment can be realized by a computer including a processor and a memory. In this case, a processor such as a CPU 203 executes a program stored in a memory such as a RAM 207 or a ROM 206 to realize each function of an image processing unit 208. However, some or all of the functions of the image processing device may be realized by dedicated hardware. Moreover, the image processing device according to an embodiment may be configured by a plurality of information processing devices connected via a network, for example.

[0034] [Embodiment 1] In this embodiment, nozzle data indicating a dot arrangement is generated so that dots in edge regions are thinned out. In particular, in this embodiment, information indicating an edge detection result is assigned to a pixel, and dots in edge regions are thinned out based on this information. On the other hand, in this embodiment, edge regions are detected based on a luminance image. In this case, if a pixel in an input image has a pixel value that belongs to a low-brightness region, information indicating an edge is not assigned to this pixel. For this reason, the generation of nozzle data is controlled so that dots in edge regions are not thinned out in low-brightness regions.

[0035] <Overall flow> The flow of edge processing according to this embodiment will be described below. Fig. 3(a) is a flow diagram showing the processing performed by the image processing unit 208 in this embodiment. In this embodiment, the input image data can be converted into nozzle data by the processing in Fig. 3(a).

[0036] In S301 , the image processing unit 208 acquires input image data from the RAM 207 .

[0037] In S302, the decoder unit 209 performs a decoding process on the acquired input image data. There are various storage formats for the input image data, and a compression format such as JPEG is generally used to reduce the amount of communication between the host computer 201 and the image processing device 100. If the storage format is JPEG, the decoder unit 209 decodes the JPEG and converts it into a bitmap format (an information format in which an image is recorded as continuous pixel values). If the host computer 201 communicates with the image processing device 100 via a dedicated driver or the like, a dedicated storage format may be used. If a dedicated storage format that is convenient for both the driver and the image processing device 100 is available, the decoder unit 209 can convert to the dedicated storage format. For example, it is possible to apply a storage format with different compression rates between an area where information needs to be stored with fine accuracy and an area where it is not, according to the characteristics of the inkjet recording device. If emphasis is placed on image quality rather than on reducing the amount of communication, the input image data may be in a bitmap format, and in that case, the decoder unit 209 may simply output the bitmap format as the conversion result.

[0038] In S303, the image analysis unit 210 detects edges from the input image. The input image is an image indicated by the input image data acquired by the image processing unit 208, and includes a bitmap image output by the decoder unit 209. In order to detect edges, the image analysis unit 210 can analyze the image using the bitmap image that is the decoded result. In this embodiment, the image analysis unit 210 detects edges in an N-value image indicating the result of threshold processing on a grayscale image obtained from the input image. Here, N is a natural number of 2 or more. N can be, for example, 2, 3, or 4. In this embodiment, the image analysis unit 210 detects pixels inside the edge corresponding to both sides (for example, top and bottom or left and right) of the edge (hereinafter referred to as edge pixels or first edge pixels) and pixels outside the edge (hereinafter referred to as edge adjacent pixels or second edge pixels). In the following description, the edge region includes these edge pixels and edge adjacent pixels.

[0039] In this embodiment, the input image is analyzed to estimate, based on the features of the input image, whether the target pixel is at an edge of a white portion of the paper or an edge of a portion formed with an ink different from that of the target pixel. In one embodiment, it is estimated whether the target pixel is at the top, bottom, left, or right edge of a shape such as a character.

[0040] FIG. 3(b) shows an internal processing flow of the image analysis process performed in S303. In S401, the image analysis unit 210 converts the input image into a grayscale image. In the grayscale image, one pixel value is assigned to each pixel. The image analysis unit 210 can calculate a pixel value for one pixel of the grayscale image using multiple pixel values ​​for the corresponding pixels of the input image. The grayscale image can be, for example, a luminance image or a brightness image. That is, the grayscale image can indicate the luminance or brightness on the recording medium for each pixel when the image is recorded on the recording medium according to each pixel value of the input image. In this embodiment, the image analysis unit 210 converts the input image into a luminance image. For example, when the bitmap image data includes information of three channels of RGB, the image analysis unit 210 can convert this information into information of one channel of luminance Y. That is, the image analysis unit 210 can calculate the luminance value Y for each pixel based on the R pixel value, the B pixel value, and the G pixel value. Note that, when the luminance value is transmitted from the application, S401 can be omitted.

[0041] Conversion from three-channel RGB information to one-channel luminance Y information can be performed according to the following formula (1). Y=R×0.299+G×0.587+B×0.114 ... conversion formula (1)

[0042] The image analysis unit 210 may convert the input image into a grayscale image according to the type of recording medium. The image analysis unit 210 may also generate such a grayscale image by converting pixel values ​​of the input image according to a conversion table. In this case, the image analysis unit 210 can use the conversion table according to the type of recording medium.

[0043] FIG. 5 shows an example of input image data. In FIG. 5(a), FIG. 5(b), and FIG. 5(c), the character "E" is shown represented by pixels 52. The background color of the character "E" is different between FIG. 5(a), FIG. 5(b), and FIG. 5(c). Here, the color of pixel 52 is RGB=(0, 0, 0). The color of pixel 51 is RGB=(255, 255, 255). The color of pixel 53 is RGB=(255, 255, 0). The color of pixel 54 is RGB=(0, 0, 255). FIG. 5(a) shows an example in which black characters are arranged on a white background. FIG. 5(b) shows an example in which black characters are arranged on a yellow background. FIG. 5(c) shows an example in which black characters are arranged on a blue background. According to conversion formula (1), the luminance Y_51 of pixel 51 is 255, the luminance Y_52 of pixel 52 is 0, the luminance Y_53 of pixel 53 is 226, and the luminance Y_54 of pixel 54 is 29.

[0044] In this embodiment, image analysis is performed using an index called luminance. Luminance Y is obtained by weighting R, G, and B pixel values ​​by coefficients as shown in conversion formula (1). By using luminance Y, it is possible to express differences in the brightness of colors on a recording medium. This point will be described with reference to FIG. 6. FIG. 6(a) shows a graph plotting the luminance on paper surface against the density value of each ink color when recording in a certain recording mode. FIG. 6(b) shows a graph plotting the luminance on paper surface against the density value of a secondary color obtained with the same ink. The luminance shown in FIG. 6 is the L* value of the CIE L*a*b* value obtained by measuring the color of a recorded object obtained by recording according to each density value with a colorimeter or the like.

[0045] As can be seen from FIG. 6(a), even if the amount of yellow ink applied to the recording medium is increased, the change in brightness is gradual compared to paper white. On the other hand, when black ink is applied, the change in brightness is large even if the amount applied is small. Furthermore, FIG. 6(b) shows that the change in brightness is particularly large for blue, which is a secondary color of cyan ink and magenta ink. The luminance Y_53 of pixel 53 in FIG. 5(b), which represents yellow, is 226, indicating that this pixel is bright. Moreover, the luminance Y_54 of pixel 54 in FIG. 5(c), which represents blue, is 29, indicating that this pixel is dark. The luminance relationship obtained by such conversion matches the luminance relationship of the ink colors on the recording medium shown in FIG. 6. This is the basis for using luminance values ​​for edge detection.

[0046] In S402, the image analysis unit 210 performs threshold processing on the grayscale image obtained in S401, and generates an N-value image indicating the result of the threshold processing. In this embodiment, the image analysis unit 210 generates a binary image by performing binarization processing. Specifically, the image analysis unit 210 converts the luminance Y data into binary data for edge detection. In this way, the image analysis unit 210 converts the luminance image into a binary image.

[0047] As an example, the image analysis unit 210 converts the luminance Y into binary data (Bin) according to the following formula (2), using a threshold value Th that is prepared in advance in accordance with the recording mode of the printer. The threshold value Th will be described later. The binary data generation formula is merely an example, and the method of binary conversion is not particularly limited. For example, the design of the inequality condition and the form of the formula may be different. IF Y > Th : Bin = 0 else : Bin = 1 ……(2)

[0048] In this embodiment, edges are detected from the input image, and the number of dots of the ink color that make up the edge pixels and the number of dots of the ink color that make up the edge adjacent pixels can be controlled. Figure 4 shows an example of the dot arrangement when such control is performed.

[0049] Fig. 4(a) shows a dot arrangement corresponding to the input image data shown in Fig. 5(a). In Fig. 4(a), the number of dots is controlled so as to reduce the amount of black ink applied to the detected edge pixels. As shown in Fig. 4(a), by reducing the amount of black ink applied to the edge pixels, bleeding of the black ink on the recording medium is suppressed. This results in a high-quality recorded image with high sharpness.

[0050] Also, by limiting the amount of color ink deposited at the detected edge adjacent pixels, it is possible to suppress bleeding between colors. Figure 4(b) shows a dot arrangement corresponding to the input image data in Figure 5(c). In Figure 4(b), the number of dots is controlled so as to suppress the amount of black ink deposited at the edge pixels, and further limit the amount of color ink deposited at the edge adjacent pixels.

[0051] However, when the color pixels adjacent to the edge are dark as shown in FIG. 4(b), reducing the amount of black or color ink applied to the edge pixels or adjacent pixels reduces the contrast of the boundary (e.g., between the character and the background). This can reduce the visibility of the printed matter. In particular, when small characters or thin lines are printed, the outline or line of the character becomes faint, and the tendency for visibility to decrease is noticeable.

[0052] As described below, a binary image is generated according to the threshold value Th, and edges are detected based on this binary image. As described above, the luminance Y mimics the luminance of the ink on the recording medium. Therefore, setting the threshold value Th of the luminance Y used in binarization is synonymous with specifying the luminance area of ​​the background color that is the target for edge detection. Therefore, based on the relationship between the luminance of the edge-adjacent pixels and the visibility, the threshold value Th can be determined so that the visibility is less likely to decrease.

[0053] For example, as shown in FIG. 6(a), even if the amount of yellow ink applied to the recording medium increases, the change in brightness of the recorded portion is gradual, and the brightness difference between the recorded portion and the white paper is small. Therefore, even if both or either of the black ink and the yellow ink are thinned out, the contrast between the pixel 52 to which the black ink is applied and the pixel 53 to which the yellow ink is applied is unlikely to decrease. In this case, even if a white portion occurs around the pixel 52, this is unlikely to be noticeable. On the other hand, as shown in FIG. 5(b), at the boundary between the pixel 53 to which a large amount of yellow ink is applied and the adjacent pixel 52 to which the black ink is applied, the color bleeding is particularly noticeable due to the brightness difference between the two inks on the medium. Here, by thinning out both or either of the black ink and the yellow ink, it is possible to improve the sharpness of the black image. Therefore, the threshold value Th can be set so that the pixel 52 to which the black ink is applied and which is adjacent to the pixel with the brightness Y_53 is detected as an edge pixel. For example, it is also possible to set a binary image near the brightness Y_53. In this case, for pixels 52 and 53, the number of dots of each color of ink can be controlled.

[0054] In S403, the image analysis unit 210 detects edges in the N-value image obtained in S402. In this embodiment, the image analysis unit 210 detects edge patterns in a binary image.

[0055] 7(a) and (b) show examples of pattern information for edge pattern detection. There are two types of pattern information: "for generating pattern matching data" and "for generating edge pattern detection results." For the pattern matching data generation information, a bit AND process is performed on each pixel of the rectangular area of ​​binary data obtained in S402. The pattern matching data obtained as a result of the bit AND process is in a form in which only the information necessary to detect an edge pattern in the rectangular area is extracted. The edge pattern detection result generation information is information for performing pattern matching processing on the pattern matching data. If a perfect match is found as a result of the pattern matching processing, the rectangular area is determined to be a specified edge pattern. The determination result is linked to the central pixel in the rectangular area.

[0056] FIG. 7(a) shows pattern information for determining that the target pixel is "the left and right ends of a 1-dot vertical line." The pattern matching data generation information is set to a value so that edge pattern detection is performed on 3×3 pixels including the target pixel. A pixel assigned with "0" in the pattern matching data generation information is considered to be a pixel that is not considered in pattern matching, regardless of how the binary data is formed. Next, the edge pattern detection result generation information corresponds to the above-mentioned predetermined edge pattern, and in this example, it is a pattern in which only the central row and vertical three pixels in the 3×3 pixels are set to 1. This information is equivalent to determining whether the central row and vertical three pixels are low-luminance and the other six pixels are high-luminance. If the pattern matching data completely matches this pattern, it can be seen that at least there are high-luminance pixels on the left and right, and the target pixel and low-luminance pixels above and below it.

[0057] Figure 7(b) shows that the target pixel is not only "the left or right end of a 1dot vertical line," but also "part of a 1dot1space." A "1dot1space" refers to a pattern in which multiple 1dot vertical lines are arranged at 1dot intervals. By expanding the range of the information for generating pattern matching data to 7x3 pixels, it is possible to include information about the surroundings of the 1dot line to which the target pixel belongs in the judgment.

[0058] Figure 7(c) shows the results of sequentially performing pattern matching on binary data using Figures 7(a) and (b). When the pattern matching data generation information and edge pattern detection result generation information in Figure 7(a) are applied to the target binary data, the judgment result is "match." When the pattern matching data generation information and edge pattern detection result generation information in Figure 7(b) are applied, the judgment result is "mismatch." Based on the detection results of the two patterns, it can be seen that the target binary data "is the left or right end of a 1-dot vertical line" and "is not part of a 1-dot 1-space."

[0059] Based on the above method, it is possible to detect various edge patterns. In this embodiment, the target of pattern matching is 7×7 pixels, but this is only one example. For example, if it is sufficient to be able to detect the patterns in FIG. 7(a)(b), the target of pattern matching is 7×3 pixels. On the other hand, if it is desired to individually detect line shapes of 4 dots or more, 7×7 pixels is not sufficient, and a wider area may be set as the target. By expanding the target, more work memory for holding the binary data to be compared and work memory for holding the pattern information are required. The work memory corresponds to the RAM 207. When the image analysis unit 210 is a dedicated circuit, if it is desired to process multiple pixels by parallel clk for pattern matching, the number of processing registers and processing circuits will also increase. In addition, since the pattern information needs to be stored in the ROM 206 of the image processing device 100 in advance, the capacity of the ROM 206 is also required. When checking edge patterns in detail and in a variety of ways, a lot of pattern information must be stored, so the design is made in consideration of such memory capacity and the increase in analysis time due to the increase in the number of comparisons.

[0060] Implementing a determination means in which "0" in the pattern matching data generation information = "not considered in pattern matching" contributes to reducing memory capacity and the number of comparisons. As another means of reducing memory capacity, it is also possible to perform another variation of pattern matching by processing such as rotation and phase shift, as shown in Figure 8. In Figure 8(a), the pattern information shown in Figure 7(a) is rotated 90 degrees, and using the processed pattern information it is possible to determine that "these are the upper and lower ends of a 1-dot horizontal line."

[0061] Furthermore, by devising the pattern information as in Figure 8(c), it is also possible to detect "one pixel to the right of the leftmost pixel" as an edge pixel. Also, by rotating the pattern in Figure 8(c), it is also possible to detect "one pixel to the left of the rightmost pixel," "one pixel below the topmost pixel," and "one pixel above the bottommost pixel." This method makes it possible to detect not only the edge pixel at the very end, but also the second pixel from the very end. In other words, it becomes possible to detect two pixels from the edge as edge pixels.

[0062] In FIG. 8, the variation is increased by processing the pattern information, but it is also possible to increase the variation by processing the binary data side.

[0063] As shown in Fig. 7(c), it is also effective to narrow down the judgment result by sequentially applying multiple pattern information and to obtain information that could not be obtained from individual pattern information. For example, in Fig. 7(c), when the pattern in Fig. 7(a) is found, it is not necessary to carry out the judgment regarding the 2-dot line or more that was prepared in advance. Since it is only necessary to apply the pattern information that judges the more detailed information of the 1-dot line as shown in Fig. 7(b), it is effective in reducing the number of comparisons. In addition, by applying Fig. 7(a)(b), it is possible to know that the target binary data is "the left and right ends of a 1-dot vertical line" and "not part of a 1-dot 1-space". By deriving this information from the results of both Fig. 7(a)(b) rather than preparing individual pattern information that can be obtained, it is effective in reducing memory capacity.

[0064] Furthermore, the image analysis unit 210 may further detect pixels adjacent to the edge detected as described above. For example, the image analysis unit 210 may detect such pixels using pattern information. In FIG. 8(b), the pattern information shown in FIG. 7(a) is shifted horizontally by one pixel, and the processed pattern information can be used to determine that the pixel is "adjacent to a 1-dot vertical line." Furthermore, the image analysis unit 210 may further detect edge-adjacent pixels based on the detection result of the edge pixels.

[0065] As described above, this embodiment can determine whether a pixel requires special processing, such as thinning out dots or changing the dot arrangement. This processing means for detecting edges from binary data is one example, and other detection means may be used.

[0066] The result of the image analysis process in S303 is output in an information format suitable for processing in the subsequent steps. For example, the result of the determination can be expressed as a 3-bit multi-value such as non-detection (not applicable to any detection pattern) = 0, upper end detection = 1, lower end detection = 2, left end detection = 3, right end detection = 4, adjacent to some end = 5. Alternatively, each bit within 5 bits can be expressed as non-detection = 00000, upper end detection = 00001, lower end detection = 00010, left end detection = 00100, right end detection = 01000, adjacent to some end = 10000. The former has a small amount of data and can be transmitted to the next process. The latter has the advantage of being able to use bit processing in the next process, thereby reducing the processing load. Here, it is described that five pieces of information are transmitted to the subsequent steps, but as explained in the explanation in S303 that "pattern information can be expressed in various ways," it is also possible to detect and transmit more information than the control information required in the subsequent processing steps.

[0067] In the following, a case where the image analysis unit 210 detects the pixel at the edge end in S303 will be described. In one embodiment, the image analysis unit 210 can detect edge pixels by the above-mentioned method. In another embodiment, the image analysis unit 210 can separately detect the pixel at the top end, the pixel at the left end, the pixel at the bottom end, and the pixel at the right end. Then, the image analysis unit 210 can output the edge determination result for each pixel. Here, the pixels at the top end, the left end, the bottom end, and the right end are pixels in a region having a pixel value of "1" in a binary image, and refer to the pixels at the top end, the left end, the bottom end, and the right end of this region.

[0068] For example, the image analysis unit 210 can output "1" as the determination result for a pixel at the top or left end. The image analysis unit 210 can also output "2" as the determination result for a pixel at the bottom or right end. The image analysis unit 210 can also output "0" as the determination result for a pixel that does not fall into any of the above categories. However, in this embodiment, it is not essential to distinguish between pixels at the top or left end and pixels at the bottom or right end.

[0069] Furthermore, the image analysis unit 210 can detect edge adjacent pixels by the above method. In the following description, the edge adjacent pixels refer to pixels in a region having a pixel value of "0" in a binary image and adjacent to an edge pixel. In one embodiment, the edge adjacent pixels are adjacent to the top, left, bottom, or right of the edge pixel. In this example, the pixel value of the edge pixel in the grayscale image is smaller than the pixel value of the edge adjacent pixel. However, in this embodiment, it is not essential to detect the edge adjacent pixels.

[0070] Such information indicating the relationship between each pixel and an edge may be referred to as edge information hereinafter. This edge information may indicate whether or not each pixel constitutes an edge. Also, this edge information may indicate the type of edge each pixel constitutes (e.g., top edge or bottom edge). Furthermore, this edge information may indicate whether or not each pixel is adjacent to a pixel that constitutes an edge. This edge information may indicate the classification of each pixel (e.g., top, left, bottom, or right edge pixel, or edge adjacent pixel).

[0071] In S304 to S306, the color separation quantization unit 211 and the nozzle separation processing unit 212 generate print data corresponding to at least one color for a pixel of interest from the value of the pixel of interest of the input image. In this embodiment, print data corresponding to C, M, Y, and K is generated for each pixel of the input image. Also, in this embodiment, print data for at least one color is generated according to the edge detection result. In the following example, print data corresponding to K is generated according to the edge detection result. Particularly, in this embodiment, print data for a pixel of interest corresponding to at least one color is generated from the value of the pixel of interest of the input image in a manner according to whether the pixel of interest of the input image is located at an edge or not. In the following example, print data for a pixel of interest corresponding to K is generated from the value of the pixel of interest of the input image in a manner according to whether the pixel of interest of the input image is located at a first end or a second end or not.

[0072] FIG. 10 shows a flowchart of the color separation process in S304 and the quantization process in S305 performed by the color separation and quantization unit 211.

[0073] In S1001, the color separation and quantization unit 211 performs color correction processing on the bitmap image obtained in S302. In this example, the bitmap image is data of three channels of RGB, and each channel of RGB has an 8-bit pixel value of 256 gradations. In S1001, the color separation and quantization unit 211 converts the RGB data of the input image data into device R'G'B' data, which is a color space specific to the recording device. The color separation and quantization unit 211 can perform this conversion by, for example, referring to a lookup table (LUT) stored in advance in memory.

[0074] Next, in S1002, the color separation quantization unit 211 separates the color data of the image into ink color data. The color separation quantization unit 211 can separate the converted R′G′B′ data into 8-bit density data of four colors, C (cyan), M (magenta), Y (yellow), and K (black), which are the ink colors of the recording device. At this stage, a single-channel gray image of four planes is generated, and the gray images of the four planes correspond to each of the four colors. In addition, each gray image indicates the density value of each pixel. For example, the color separation quantization unit 211 can perform this separation process by referring to a lookup table (LUT) stored in advance in a memory. The following describes the process for the density value K of the K plane. Unless otherwise specified, the same process is performed for the density values ​​C, M, and Y of the C, M, and Y planes.

[0075] In S1003, the color separation and quantization unit 211 performs a gradation correction process on the density value K. For example, the color separation and quantization unit 211 can obtain a density value K' by performing a gradation correction process on the density value K using a gradation correction table. The gradation correction process is performed so that the input density value and the optical density expressed on the recording medium have a linear relationship. FIG. 9 shows an example of the setting of the gradation correction process. In FIG. 9, In ​​corresponds to the density value K which is an input value to the gradation correction process, and Out corresponds to the density value K' which is an output value of the gradation correction process. For ease of explanation, FIG. 9 shows an example where In and Out have a linear relationship.

[0076] In S1004, the color separation and quantization unit 211 performs a quantization process on the density value K'. For example, the color separation and quantization unit 211 performs a predetermined quantization process on the density value K' to generate 4-bit 3-value quantized data of "0000", "0001", or "0010". Similarly, the color separation and quantization unit 211 performs a quantization process on the density values ​​C', M', and Y' to generate 4-bit 3-value quantized data C", M", and Y" of "0000", "0001", or "0010".

[0077] Furthermore, in S1005 to S1009, the color separation and quantization unit 211 sets a value indicating the determination result in S303 to the quantized data obtained in S1004. The color separation and quantization unit 211 sets a value to the most significant 2 bits of the quantized data based on edge information of the pixel to be processed. In this way, the color separation and quantization unit 211 outputs 4-bit quantized data K". This quantized data can be called color separation data corresponding to the printing material used by the printing device for printing. This data also indicates the printing amount for each pixel (lower 2 bits) and the edge detection result for each pixel (upper 2 bits).

[0078] Specifically, in S1005, the color separation and quantization unit 211 determines whether the pixel is at the top end or the left end. Hereinafter, the top end or the left end will be referred to as the first end. If it is detected that the pixel is at the first end, the process proceeds to S1009. In S1009, the color separation and quantization unit 211 sets the value "01" to the most significant 2 bits.

[0079] If it is not detected that the pixel is at the first end, in S1006, the color separation and quantization unit 211 determines whether the pixel is at the bottom end or the right end. Hereinafter, the top end or the left end will be referred to as the second end. If it is detected that the pixel is at the second end, the process proceeds to S1008. In S1008, the color separation and quantization unit 211 sets the value "10" to the most significant 2 bits. If it is not detected that the pixel is at the second end, the process proceeds to S1007. In S1007, the color separation and quantization unit 211 sets the value "00" to the most significant 2 bits.

[0080] On the other hand, in this embodiment, the color separation and quantization unit 211 does not add end information indicating the edge determination result to the most significant 2 bits of the quantized data C'', M'', and Y''. Therefore, the quantized data C'', M'', and Y'' are "0000", "0001", or "0010". Hereinafter, the quantized data C'', M'', and Y'' will be collectively referred to as quantized data CL''.

[0081] In S306, the nozzle resolution processing unit 212 generates nozzle data to be used as print data. The nozzle resolution processing unit 212 generates nozzle data by index expansion processing on the quantized data K" obtained in S305. In the index expansion processing in this embodiment, the quantized data K" of 600 x 600 dpi is converted into composite nozzle data Kp of 600 x 1200 dpi using a prepared index pattern. In this example, data for one pixel is converted into data for two pixels linked in the vertical direction. In this manner, in this embodiment, composite nozzle data with a higher resolution than the quantized data is generated based on the multi-value (more than two values, three values ​​in this embodiment) quantized value of the quantized data. In addition, the composite nozzle data has data for a plurality of pixels (two pixels in this embodiment) corresponding to one pixel of the input image.

[0082] In this embodiment, the nozzle resolution processing unit 212 generates print data corresponding to K for a pixel of interest from quantized data K" obtained based on the value of the pixel of interest in the input image, using a method depending on whether the pixel of interest in the input image is at the first end or the second end. The nozzle resolution processing unit 212 generates nozzle data K based on the detection results of edge pixels (i.e. the first end and second end). In the following example, such control is performed based on a dot arrangement pattern and a reference index pattern. Figures 11(a) to (c) show an example of a dot arrangement pattern and a reference index pattern used in the index expansion process. Figure 11(a) shows the dot arrangement pattern indicated by the composite nozzle data Kp obtained by the index expansion process.

[0083] When the quantized data K'' indicates "0000," "0100," or "1000," no dot is placed in either of the two pixels.

[0084] When the quantized data K" indicates "0001," a dot is placed in the upper of the two pixels in pattern A, and a dot is placed in the lower of the two pixels in pattern B. When the quantized data K" indicates "0010," a dot is placed in both of the two pixels.

[0085] When the quantized data K" indicates "0101" or "0110", a dot is placed in the upper of the two pixels, and no dot is placed in the lower of the two pixels. When the quantized data K" indicates "1001" or "1010", a dot is placed in the lower of the two pixels, and no dot is placed in the upper of the two pixels.

[0086] Here, attention is paid to the case where the quantized data K" indicates "0010", "0110" or "1010". As described above, the upper 2 bits of the quantized data K" are given a value according to the detection result in S403 as end information. This end information indicates the detection result of the edge pixels (i.e., the first end and the second end). In this example, although the lower 2 bits have a common value of "10", the arrangement and number of dots change depending on the value of the upper 2 bits. That is, when the quantized data K" is "0010", dots are arranged above and below. On the other hand, when the quantized data K" is "0110" or "1010", dots are arranged only on the upper side or the lower side. In this way, even if the quantized value obtained in S1004 is the same, the method of generating the print data is controlled depending on the value of the upper 2 bits of the quantized data K". Specifically, in this embodiment, the number of dots is controlled depending on the value of the upper 2 bits of the quantized data K". In this example, the nozzle resolution processing unit 212 generates print data so that when it is determined that a pixel is on an edge (first end or second end), the amount of printing for the pixel is reduced compared to when it is determined that the pixel is not on an edge. In this way, it is not essential to distinguish between the first end and the second end in order to control the number of dots. On the other hand, in this embodiment, the arrangement of dots is also controlled according to the value of the most significant 2 bits of the quantized data K". In this way, in this embodiment, the number or arrangement of dots can be controlled based on the end information.

[0087] In this embodiment, with respect to the quantized data K″, if the pixel is not on an edge (if the upper two bits of the quantized data are “00”), a maximum of two dots are placed on this pixel. In this case, the maximum recording rate is 100%. If the pixel is on an edge (first end or second end) (if the upper two bits of the quantized data are “01” or “10”), a maximum of only one dot is placed on this pixel. For this reason, the maximum recording rate is 50%. The maximum recording rate can be the ratio of the number of dots that can be placed under specific conditions to the number of dots that can be placed on one pixel when generating nozzle data. In this way, the nozzle resolution processing unit 212 can generate recording data such that the maximum recording amount (i.e., the number of dots) for pixels on an edge is less than the maximum recording amount for pixels that are not on an edge. Also, the nozzle resolution processing unit 212 can generate recording data such that the maximum recording rate for pixels on an edge is lower than the maximum recording rate for pixels that are not on an edge.

[0088] 11(b) shows the reference index pattern. Each square corresponds to one pixel of 600 x 600 dpi. This reference index pattern determines whether dots are arranged according to pattern A or pattern B at each pixel.

[0089] FIG. 11C shows binary composite nozzle data (600 dpi in the X direction and 1200 dpi in the Y direction) obtained by index expansion processing when all quantized data for each pixel is "0001". The nozzle resolution processing unit 212 assigns data of an upper pixel (nozzle data K1p) of the composite nozzle data Kp corresponding to the quantized data K" of one pixel to the Ev nozzle of the black nozzle column 2701 corresponding to the pixel. Furthermore, the nozzle resolution processing unit 212 assigns data of a lower pixel (nozzle data K2p) of the composite nozzle data Kp corresponding to the quantized data K" of one pixel to the Od nozzle of the black nozzle column 2701 corresponding to the pixel. Then, for one pixel of the input image, both recording at an upper position using the Ev nozzle and recording at a lower position using the Od nozzle are performed. In this way, the nozzle resolution processing unit 212 generates recording data for multiple positions corresponding to one pixel of the input image (i.e., data of two pixels connected in the vertical direction) based on the quantized data for this pixel.

[0090] The nozzle resolution processing unit 212 also performs index expansion processing on the quantized data C'', M'', and Y'' obtained in S305 in a similar manner, thereby generating composite nozzle data Cp, Mp, and Yp for the color inks.

[0091] Through the above process, composite nozzle data of 600×1200 dpi is obtained based on each pixel of the 600×600 dpi input image data. This composite nozzle data specifies whether or not each nozzle of the black nozzle row 2701 is to print. In order to print on the print medium, the print head H can eject ink according to such nozzle data. That is, the composite nozzle data contains data of a plurality of pixels (two pixels in this embodiment) corresponding to one pixel of the input image, indicating the number of ink dots used to print this pixel. Furthermore, when the maximum number of ink dots corresponding to one pixel of the input image indicated by the composite nozzle data is M dots (two dots in this embodiment), the nozzle resolution processing unit 212 can limit the number of ink dots for pixels at edges to less than M.

[0092] In the following, the process of converting the image data shown in FIG. 5 into nozzle data will be further described with reference to FIGS. 12 to 17 according to the flowchart shown in FIG. 3(a). FIG. 12 shows an image obtained by performing luminance conversion on the image data of FIGS. 5(a) to (c) in S401. The luminance image obtained from the input image shown in FIG. 5(a) is composed of pixels with a luminance value Y=255 and pixels with a luminance value Y=0 as shown in FIG. 12(a). The luminance image obtained from the input image shown in FIG. 5(b) is composed of pixels with a luminance value Y=226 and pixels with a luminance value Y=0 as shown in FIG. 12(b). The luminance image obtained from the input image shown in FIG. 5(c) is composed of pixels with a luminance value Y=29 and pixels with a luminance value Y=0 as shown in FIG. 12(c).

[0093] Fig. 13 shows a binary image obtained by performing binarization processing on the luminance image shown in Fig. 12 using a threshold value Th=225 in S402. The binary images shown in Fig. 13(a) to (c) correspond to the luminance images shown in Fig. 12(a) to (c), respectively. The luminance image shown in Fig. 12(c) is composed of pixels with a luminance Y=29 and pixels with a luminance Y=0, so that all pixel values ​​are converted to "1" by the binarization processing.

[0094] FIG. 14 shows the result of edge detection processing performed in S403 on each pixel of the binary image shown in FIG. 13. FIG. 14 shows edge information for each pixel. As shown in FIGS. 14(a) and (b), edges are detected from the binary images shown in FIGS. 13(a) and (b). On the other hand, as shown in FIG. 14(c), edges are not detected from the binary image shown in FIG. 13(c). In this way, the edge of the black character "E" adjacent to a high density color having a luminance Y=26 is not detected. For this reason, edge information indicating that this pixel is an edge pixel is not given to the edge pixels of the black character "E".

[0095] FIG. 15 shows quantized data obtained by performing the processes in S304 and S305 using the end information shown in FIG. 14. FIGS. 15(a) to (c) show quantized data K" corresponding to the image data in FIGS. 5(a) to (c). FIGS. 15(d) to (f) show quantized data CL" corresponding to the image data in FIGS. 5(a) to (c).

[0096] In the quantized data K″, the pixels detected as the first end and the second end are assigned the values ​​“0110” and “1010”, respectively. In addition, the pixels not detected as the end are assigned the value “0010”. In the quantized data K″ obtained from the image data of FIG. 5(c) in which no end is detected as shown in FIG. 14(c), the values ​​of all the pixels corresponding to the black character “E” are “0010”. As described above, the upper two bits of the quantized data CL″ are not set to a value based on the end information. Therefore, in the quantized data CL″ shown in FIGS. 15(e) and (f), the values ​​of all the pixels corresponding to the color data are “0010”. Note that pixels in which the lower two bits of the quantized data are “00” are not shown.

[0097] FIGS. 16(a) to (f) show dot arrangement patterns corresponding to the quantized data K" of FIGS. 15(a) to (c) and the quantized data CL" of FIGS. 15(d) to (f), obtained by the index expansion process in S306. It can be seen that in the dot arrangement pattern K shown in FIGS. 16(a) and (b), the number and arrangement of dots are controlled in accordance with edge information. Compared with FIGS. 16(a) and (b), it can be seen that in the dot arrangement pattern K shown in FIG. 16(c), the number and arrangement of dots are also controlled in accordance with edge information indicating that each pixel is not an edge pixel. Note that in this embodiment, no control of the number of dots is performed in the dot arrangement pattern CL.

[0098] FIG. 17 shows the final dot arrangement obtained by performing the processes of S301 to S306 on the input image shown in FIG. 5. FIG. 17(a) corresponds to FIG. 5(a) and shows the dot arrangement when the black character "E" composed of black dots indicated by K_1 has white adjacent colors. It can be seen that the number and arrangement of dots in the edge region are controlled based on the values ​​indicating the edge information given to each pixel of the quantized data. This process makes it possible to suppress bleeding of the black ink K on the recording medium at the edge pixels and suppress a decrease in the sharpness of the characters.

[0099] FIG. 17(b) corresponds to FIG. 5(b) and shows a dot arrangement when the density of color dots of ink CL_1 adjacent to the black character "E" composed of black dots indicated by K_1 is low. In FIG. 17(b), CL_1 is yellow ink. As in FIG. 17(a), it can be seen that the number and arrangement of dots corresponding to the edge pixels of the black character "E" are controlled. This processing makes it possible to suppress bleeding of the black ink K. It is also possible to suppress bleeding between the black ink K and the adjacent yellow ink CL_1.

[0100] FIG. 17(c) corresponds to FIG. 5(c) and shows a dot arrangement in which the color dots of ink CL_2 adjacent to the black character "E" composed of black dots indicated by K_1 are dense. In FIG. 17(c), CL_2 is composed of magenta ink and cyan ink. Unlike FIG. 17(a) and (b), edge pixels are not detected in this example. Therefore, the number and arrangement of dots corresponding to all pixels that compose the black character "E" are controlled in the same way as non-edge pixels. In such a case, if the number of black ink dots that compose the edge pixels is controlled, the outline of the character becomes thin, especially when the character size is small, and the contrast between the character and the background decreases, and visibility may decrease. According to this embodiment, when the brightness of a pixel adjacent to an edge pixel is lower than a predetermined threshold, edge information indicating that this pixel is an edge pixel is not added. Therefore, it is possible to suppress a decrease in visibility or the appearance of a noticeable white border around the black character.

[0101] According to the dot arrangement pattern shown in FIG. 11(a), the dot arrangement is controlled depending on whether the pixel is at the first end (upper 2 bits="01") or the pixel is at the second end (upper 2 bits="10") Specifically, when the pixel is at the first end (upper end or left end), recording is performed at the upper position among the multiple recording positions, and recording is not performed at the lower positions. Also, when the pixel is at the second end (lower end or right end), recording is performed at the lower positions, and recording is not performed at the upper positions. By controlling the dot arrangement in this way, it becomes easier to align the positions of each edge compared to the case where dots are randomly thinned out, as shown in FIG. 17(a).

[0102] Here, when a pixel is at the first end (top end or left end), it is not essential to always perform printing at the upper position. For example, the same effect can be obtained by biasing the dot arrangement to one of the multiple printing positions according to the end information. In one embodiment, among the multiple printing positions corresponding to the first edge pixel, printing is performed more at the position adjacent to the second edge pixel than at the position not adjacent to the second edge pixel. The nozzle resolution processing unit 212 can generate printing data for the colors of the first group in this way. Also, in this embodiment, the number of dots is controlled based on the end information in the composite nozzle data having a higher resolution than the input image and the quantized data. For this reason, when the quantized values ​​obtained in S1004 for the edge pixel are 0, 1, and 2, the number of dots for this edge pixel can be controlled to be 0, 1, and 1. However, it is not essential to control the number of dots in such high-resolution data. For example, a binary (0 or 1) quantized value may be obtained in S1004. Then, among the edge pixels with a quantization value of 1, the number of dots may be controlled (thinned out) so that half of the edge pixels have a number of dots of 0 and the remaining edge pixels have a number of dots of 1.

[0103] According to this embodiment, the number of dots in the edge region is controlled based on edge information detected using an N-value image that indicates the result of threshold processing on a grayscale image (e.g., a luminance image). This makes it possible to improve the sharpness of images such as printed characters or lines while suppressing secondary image quality degradation such as reduced visibility.

[0104] (Variation 1) In the above-described embodiment, the number of dots to be placed on edge pixels and their arrangement are controlled. Hereinafter, a configuration for controlling the number of dots to be placed on edge adjacent pixels and their arrangement will be described.

[0105] 18 shows a flowchart of the color separation process in S304 and the quantization process in S305 performed by the color separation and quantization unit 211 in this modified example. Steps S1801 to S1804 are performed in the same manner as steps S1001 to S1004.

[0106] In S1805 to S1813, the color separation and quantization unit 211 sets a value indicating the determination result in S303 to the quantized data obtained in S1804. That is, the color separation and quantization unit 211 outputs 4-bit quantized data C", M", Y", and K" by setting a value to the most significant 2 bits of the quantized data based on the edge information. The processes in S1805 to S1813 are performed for each of the quantized data C", M", Y", and K". Here, the color separation and quantization unit 211 determines the value of the most significant 2 bits by processing according to each color. Here, the multiple colors used for recording can be classified into two or more groups including a first group and a second group. Then, the color separation and quantization unit 211 can perform different processing for each color group. Hereinafter, K is defined as a first group color, Y as a second group color, and C and M as a third group color.

[0107] In S1805, the color separation and quantization unit 211 determines whether the quantized data to be processed corresponds to the first group of colors. If the quantized data to be processed corresponds to the first group of colors, the process proceeds to S1806, and if not, the process proceeds to S1811. For the quantized data K″ of the first group of colors, a value of the most significant 2 bits is set by the processes of S1806 to S1810. This process is similar to that of S1005 to S1009.

[0108] In S1811, the color separation and quantization unit 211 determines whether the quantized data to be processed corresponds to a color in the second group. If the quantized data to be processed corresponds to a color in the second group, the process proceeds to S1812, and if not, the process proceeds to S1808. For the quantized data Y″ of the color in the second group, a value of the most significant 2 bits is set by the processes in S1812 to S1813 and S1808.

[0109] In S1812, the color separation and quantization unit 211 determines whether the pixel is an edge adjacent pixel. As described above, the image analysis unit 210 can detect edge adjacent pixels. Hereinafter, a portion adjacent to the first end or the second end is referred to as the third end. For example, the first end and the third end, or the second end and the third end, correspond to both sides of an edge, respectively. The third end can be present in an area having a pixel value of "0" in a binary image. In this example, the edge adjacent pixel is a pixel of the third end. Also, the edge pixel and the edge adjacent pixel corresponding to both sides of an edge are adjacent to each other, and the value of the edge pixel and the value of the edge adjacent pixel in the binary image are different. If the pixel is in the third end, the process proceeds to S1813. In S1813, the color separation and quantization unit 211 sets the value "11" to the upper 2 bits. If the pixel is not in the third end, the process proceeds to S1808, where the color separation and quantization unit 211 sets the value "00" to the upper 2 bits.

[0110] If the quantized data to be processed corresponds to a color in the third group, the process proceeds to S1808. That is, the color separation and quantization unit 211 sets the value "00" to the most significant 2 bits of the quantized data C" and M" of the color in the third group.

[0111] In S306, the nozzle resolution processing unit 212 performs index expansion processing on the quantized data C", M", Y", and K" output in S305 in the same manner as in the above embodiment. As described above, the quantized data for the first group of colors is provided with end information indicating the detection results of edge pixels (i.e., the first end and second end). Therefore, the nozzle resolution processing unit 212 can generate printing data for the first group of colors based on the detection results of the edge pixels.

[0112] Figures 19(a) to (c) show an example of a dot arrangement pattern and a reference index pattern used in the index expansion process for the quantized data Y" corresponding to the second group of colors. Note that the dot arrangement pattern and the reference index pattern used in the index expansion process shown in Figures 11(a) to (c) are used in the index expansion process for the quantized data K" corresponding to the first group of colors.

[0113] Figure 19(a) shows the dot arrangement pattern indicated by the composite nozzle data Yp obtained by index expansion processing corresponding to the quantized data Y". When the quantized data Y" indicates "0000" or "1100", no dots are placed in either of the two pixels.

[0114] When the quantized data Y" indicates "0001," in pattern A a dot is placed in the upper of the two pixels, and in pattern B a dot is placed in the lower of the two pixels. When the quantized data Y" indicates "0010," a dot is placed in both of the two pixels. These are the same dot arrangement patterns as those shown in FIG. 11(a).

[0115] When the quantized data Y'' indicates "1101" or "1110", as with "0001", in pattern A a dot is placed in the upper of the two pixels, and in pattern B a dot is placed in the lower of the two pixels.

[0116] In this way, even though "1110" and "0010" have the same lower 2 bits "10", the arrangement and number of dots change depending on the value of the upper 2 bits. That is, when the quantized data Y" is "0010", dots are arranged above and below. On the other hand, when the quantized data Y" is "1110", dots are arranged only on the upper or lower side. In this way, even if the quantized value obtained in S1804 is the same, the arrangement or number of dots can be controlled depending on the value of the upper 2 bits of the quantized data Y".

[0117] As described above, the quantized data for the second group color is provided with end information indicating the detection result of the edge adjacent pixel (i.e., the third end). Therefore, the nozzle resolution processing unit 212 can generate print data for the second group color based on the detection result of the edge adjacent pixel. For example, when it is determined that the pixel of interest is on an edge (the third end), the nozzle resolution processing unit 212 can generate print data such that the print amount for the pixel of interest is reduced compared to when it is determined that the pixel of interest is not on an edge. In this way, the dot arrangement for the first group color and the dot arrangement for the second group color can be controlled based on different types of edge detection results (for example, the detection result of the edge pixel or the edge adjacent pixel).

[0118] In this embodiment, with regard to the quantized data Y", if the pixel is not on an edge (if the most significant 2 bits of the quantized data are "00"), a maximum of 2 dots are placed on this pixel. In this case, the maximum recording rate is 100%. If the pixel is on an edge (third end portion) (if the most significant 2 bits of the quantized data are "11"), a maximum of only 1 dot is placed on this pixel. For this reason, the maximum recording rate is 50%. In this way, the nozzle resolution processing unit 212 can generate recording data such that the maximum recording amount (i.e., the number of dots) for pixels on an edge is less than the maximum recording amount for pixels not on an edge. Also, the nozzle resolution processing unit 212 can generate recording data such that the maximum recording rate for pixels on an edge is lower than the maximum recording rate for pixels not on an edge.

[0119] FIG. 19(b) shows the reference index pattern referenced when the quantized data Y" is "1101" or "1110". When the quantized data Y" is other than "1101" or "1110", the reference index pattern shown in FIG. 11(b) is referenced. In this way, by switching the reference index pattern to be referenced depending on the most significant 2 bits of the quantized data, it is possible to control the dot arrangement in edge areas.

[0120] Figure 19(c) shows binary composite nozzle data (600 dpi in the X direction and 1200 dpi in the Y direction) obtained by index expansion processing according to the reference index pattern shown in Figure 19(b) when the quantized data for each pixel is all "1101." By using the reference index pattern shown in Figure 19(b), in the area of ​​pixels with "11" assigned to the most significant 2 bits, pixels with dots placed above and pixels with dots placed below are alternately placed.

[0121] The nozzle resolution processing unit 212 assigns the data of the upper pixel (nozzle data Y1p) of the composite nozzle data Yp corresponding to the quantized data Y″ of one pixel to the Ev nozzle of the yellow nozzle row corresponding to the pixel. Furthermore, the nozzle resolution processing unit 212 assigns the data of the lower pixel (nozzle data Y2p) of the composite nozzle data Yp corresponding to the quantized data Y″ of one pixel to the Od nozzle of the yellow nozzle row corresponding to the pixel.

[0122] The nozzle decomposition processing unit 212 also performs index expansion processing on the quantized data C'' and M'' obtained in S305 in a similar manner to generate composite nozzle data Cp and Mp.

[0123] The process of converting the image data shown in FIG. 5 into nozzle data will be further described below with reference to FIGS. 20 to 23, in accordance with the flowchart shown in FIG. 3(a).

[0124] Figs. 14(a)-(c) and 20(a)-(c) show the results of edge detection processing performed on each pixel of the binary images shown in Figs. 13(a)-(c). Figs. 14(a)-(c) show the "first end" and "second end" detected in S403. Figs. 20(a)-(c) show the detection results of a third end adjacent to either end. In Figs. 20(a) and (b), a pixel adjacent to the first end or the second end of the black character "E" is detected as the third end. On the other hand, neither the first end nor the second end is detected in Fig. 14(c), and therefore the third end is not detected in Fig. 20(c).

[0125] FIG. 21 shows quantized data obtained by performing the processes in S304 and S305 using the end information shown in FIG. 14 and FIG. 20. FIG. 21(a)-(c) shows quantized data K" corresponding to the image data of FIG. 5(a)-(c), which is the same as FIG. 15(a)-(c). FIG. 21(d)-(f) shows quantized data Y" corresponding to the image data of FIG. 5(a)-(c). In the quantized data Y" shown in FIG. 21(e), the value "1110" is assigned to pixels detected as the third end. The value "0010" is assigned to pixels not detected as the third end. Note that FIG. 21 does not show pixels whose lower 2 bits of quantized data are "00".

[0126] 22(a)-(f) show dot arrangement patterns corresponding to the quantized data K" of FIGS. 21(a)-(c) and the quantized data CL" of FIGS. 21(d)-(f), obtained by the index expansion process in S306. The dot arrangement pattern K shown in FIGS. 22(a)-(c) is the same as that of FIGS. 16(a)-(c). In the dot arrangement pattern CL shown in FIG. 22(e), it can be seen that the number and arrangement of dots in edge-adjacent pixels are controlled in accordance with end information indicating the third end. Furthermore, the dot arrangement pattern is controlled so that pixels detected as the third end are printed by either the upper dots (Ev nozzles) or the lower dots (Odd nozzles). On the other hand, the dot arrangement pattern is controlled so that pixels not detected as the third end are printed by both the upper nozzles and the lower nozzles. 22(f), for example, the background pixels are not detected as the third edge, so the dot arrangement pattern is controlled so that the pixels adjacent to the black character "E" are also printed by both the upper and lower nozzles. In this way, it can be seen that the number and arrangement of dots are controlled in accordance with the edge information in dot arrangement pattern Y as well.

[0127] FIG. 23(a)-(c) show the final dot arrangement obtained by performing the processes of S301-S306 on the input images shown in FIG. 5(a)-(c). FIG. 23(a) and (c) are the same as FIG. 17(a) and (c). FIG. 23(b) corresponds to FIG. 5(b) and shows the dot arrangement when the color dots of ink CL_1 adjacent to the black character "E" composed of black dots indicated by K_1 have low density. In FIG. 23(b), CL_1 is yellow ink. As in FIG. 23(a), it can be seen that the number and arrangement of dots corresponding to the edge pixels of the black character "E" are controlled. This process can suppress bleeding of the black ink K on the recording medium at the edge pixels and suppress the deterioration of the sharpness of the characters. In this example, the number of yellow dots indicated by CL_1 adjacent to the edge pixels of the black character "E" is also limited. Therefore, bleeding between colors can be suppressed more than in the example of FIG. 17(b). In particular, in the example of Fig. 23(b), the arrangement of the yellow ink CL_1 dots adjacent to K_1 has been devised. In other words, by using the index pattern shown in Fig. 19(b), the yellow ink CL_1 dots are evenly arranged in the adjacent portions of the edge pixels of the black character "E". With this configuration, bleeding between colors can be suppressed even more effectively.

[0128] (Variation 2) In the above modified example, K was classified into the first group color, Y into the second group color, and C and M into the third group color. Then, for the first group color, nozzle data was generated to control the number and arrangement of ink dots in edge pixels. Also, for the second group color, nozzle data was generated to control the number and arrangement of ink dots in edge adjacent pixels. However, the classification method is not particularly limited. For example, K may be classified into the first group color, and C, M, and Y into the second group color.

[0129] Also, a color other than black may be classified as the first group color. In one embodiment, at least one color is classified as the first group color, and the other colors are classified as the second group color or the third group color. In this case, the color classified as the first group color may be a color with a relatively low brightness (for example, K, C, or M). In another embodiment, at least one color is classified as the second group color, and the other colors are classified as the first group color or the third group color. In this case, the color classified as the second group color may be a color with a relatively high brightness (for example, Y, C, or M). This configuration also provides the effect of suppressing at least one of the decrease in sharpness of characters or the bleeding between colors, while suppressing secondary deterioration in image quality. In one embodiment, the optical density of all the recording materials of the first group color is higher than the optical density of any of the recording materials of the second group color. Optical density (OD) indicates the attenuation rate of light expressed in logarithm.

[0130] For example, at least one of K, C, and M may be classified into the first group color, and Y may be classified into the second group color. In the following, a case where K, C, and M are classified into the first group color, and Y is classified into the second group color will be described. The ink constituting the edge pixel is not limited to black ink. For example, the pixel constituting the character "E" shown in Figs. 5(a) and (b) may be blue (RGB=(0, 0, 255)). Even in such a case, in S303, the first end and the second end can be detected as shown in Figs. 14(a) and (b), and the third end can be detected as shown in Figs. 20(a) and (b). In the following, an example will be described in which the number and arrangement of dots of each ink are controlled based on end information when pixel 52 of the input image shown in Figs. 5(a) and (b) is blue with RGB=(0, 0, 255).

[0131] The color separation process in S304 and the quantization process in S305 in this modified example can be performed according to FIG. 18. As described above, the quantized data C", M", and K" correspond to the first group of colors, and the quantized data Y" corresponds to the second group of colors. The most significant two bits of the quantized data C", M", and K" of a pixel determined to be the first end or the second end are set to a value of "01" or "10". The most significant two bits of the quantized data Y" of a pixel determined to be the third end are set to a value of "11". The most significant two bits of the other quantized data are set to a value of "00".

[0132] The index expansion process in S306 is performed in the same manner as in the above-mentioned modified example 1. That is, the dot arrangement pattern and reference index pattern shown in Fig. 11 are used for the index expansion process corresponding to the quantized data C'', M'', and K''. Also, the dot arrangement pattern and reference index pattern shown in Fig. 19 are used for the index expansion process corresponding to the quantized data Y''.

[0133] The process of converting the image data into nozzle data as described above will be further described below with reference to the flow chart shown in FIG. 3(a) and with reference to FIGS.

[0134] 24(a) and (c) show quantized data obtained by performing the processes in S304 and S305 using the input images shown in FIG. 5(a) and FIG. 5(b). Also, FIG. 24(b) and (d) show quantized data obtained by performing the processes in S304 and S305 using the input images shown in FIG. 5(a) and FIG. 5(b). FIG. 24(a) and (b) show quantized data C″ and M″, and FIG. 24(c) and (d) show quantized data Y″. The values ​​of the quantized data C″ and M″ corresponding to the first group color for pixels detected as the first end portion or the second end portion are “0110” or “1010”. The values ​​of the quantized data C″ and M″ for pixels not detected as the first end portion or the second end portion are “0010”. The value of the quantized data Y″ corresponding to the second group color for pixels detected as the third end portion is “1110”. Furthermore, for pixels that are not detected as the third end portion, the value of the quantized data Y'' is "0010." Note that pixels whose lowest 2 bits of the quantized data are "00" are not shown.

[0135] Fig. 25 shows dot arrangement patterns obtained by the index expansion process in S306. The dot arrangement patterns C and M shown in Figs. 25(a) and (b) correspond to the quantized data C" and M" shown in Figs. 24(a) and (b), which are similar to the dot arrangement pattern K shown in Figs. 22(a)-(b). Moreover, the dot arrangement pattern Y shown in Figs. 25(c) and (d) corresponds to the quantized data Y" shown in Figs. 24(c) and (d), which is similar to the dot arrangement pattern Y shown in Figs. 22(d) and (e).

[0136] FIG. 26 shows a final dot arrangement obtained by performing the processes of S301 to S306 on the input image as described above. FIG. 26(a) shows a dot arrangement when the adjacent color of the character "E" composed of color dots indicated by CL_2 is white. The color indicated by CL_2 is a darker color than CL_1. In this example, the color dots indicated by CL_2 are a combination of dots of cyan ink and dots of magenta ink. Also, in this example, the character "E" is a blue character. In this case, too, it can be seen that the number and arrangement of dots in the edge region are controlled based on the value indicating the edge information given to each pixel of the quantized data. By such a process, it is possible to suppress bleeding of cyan and magenta inks on the recording medium at the edge pixels and suppress a decrease in the sharpness of the characters.

[0137] FIG. 26(b) shows a dot arrangement in the case where the density of color dots of ink CL_1 adjacent to the character "E" composed of color dots indicated by CL_2 is low. In FIG. 26(b), CL_1 is yellow ink. As in FIG. 26(a), it can be seen that the number and arrangement of dots corresponding to edge pixels of the blue character "E" are controlled. This process can suppress bleeding of cyan ink and magenta ink on the recording medium at edge pixels, and suppress deterioration of character sharpness. In this example, the number of yellow dots indicated by CL_1 adjacent to edge pixels of the blue character "E" is also limited. This makes it possible to suppress bleeding between the yellow ink and the cyan and magenta inks.

[0138] Inter-color bleeding does not only occur between black ink and color ink, but also between color inks. Inter-color bleeding also occurs between cyan ink and magenta ink. On the other hand, when the difference in brightness between the single color inks is large, inter-color bleeding is likely to be noticeable. For example, the density of yellow ink as a single color ink is low, and the density of cyan ink and magenta ink as single color inks is high. For this reason, inter-color bleeding between yellow ink and cyan ink and magenta ink is relatively likely to be noticeable. In this embodiment, cyan ink and magenta ink are classified as the first group color in order to suppress bleeding between yellow, which has a high brightness, and blue, which is a secondary color of cyan ink and magenta ink.

[0139] This method can improve the sharpness of images such as characters or lines printed in color ink while suppressing secondary degradation of image quality such as reduced visibility. In addition, this method can suppress bleeding between color inks.

[0140] In addition, when K is classified into the first group color, and C, M, and Y are classified into the second group color, the upper two bits of the quantized data C″ and M″ for the pixel detected as the “third end” can be given “11” in the same manner as the quantized data Y″. Here, in the index expansion process in S306 for the quantized data C″ and M″, the dot arrangement pattern (FIG. 19(a)) used for the quantized data Y″ can be used. On the other hand, in the index expansion process for the quantized data C″ and M″, another dot arrangement pattern may be used. For example, when the quantized data is “1101”, a dot is arranged in the upper pixel of the two pixels in pattern A, and a dot is arranged in the lower pixel of the two pixels in pattern B. On the other hand, when the quantized data is “1110”, dots may be arranged in both of the two pixels. In this way, when the upper two bits are “11”, the same dot arrangement pattern as when the upper two bits are “00” may be used. In this way, by setting a dot arrangement pattern to be used in the index expansion process for each ink color, it is possible to control the dot arrangement according to the added end information for each ink color.

[0141] Also, there may be no colors classified into the first group of colors. In this case, the number or arrangement of dots in the edge region for the second group of colors is controlled by the above-mentioned method. For example, when the density of the color dots is low, the amount of color ink used to print in the edge region can be reduced to suppress bleeding between the black ink and the color ink. On the other hand, when the density of the color dots is high, the amount of color ink used to print in the edge region is not reduced, so that noticeable fringing around black characters can be suppressed.

[0142] (Variation 3) In the above-mentioned modified examples 1 and 2, the number of dots of yellow ink in the pixel determined to be the third end adjacent to the edge pixel is limited. On the other hand, as shown in FIG. 28, there is a possibility that the pixels determined to be the third end are consecutive. FIG. 28(a) is a luminance image obtained by luminance conversion in S401 of a gradation image from yellow to a slightly darker similar color. FIG. 28(b) is a binary image obtained by binarizing this luminance image in S402. Here, when the luminance image shown in FIG. 28(a) has a gradation of luminance Y near the threshold value Th, there are cases where pixels with pixel values ​​"0" and pixels with pixel values ​​"1" alternate in the binary image. The result of the edge detection process in S403 for the binary image shown in FIG. 28(b) is shown in FIG. 28(c). For simplification, only the "third end" is shown in FIG. 28(c). In FIG. 28(c), there is a place where two pixels detected as the "third end" are consecutive in a direction perpendicular to the edge. This is because, in the above example, the pixels adjacent to the "first end" or "second end" are detected as the "third end." In this way, if the "third end" is continuous in a direction perpendicular to the edge, even if the yellow ink has a high brightness, this part may appear lighter than the surrounding area due to the continuous pixels with a limited number of dots.

[0143] Therefore, when two or more lines of edge-adjacent pixels adjacent to each other are detected, the image analysis unit 210 can exclude at least one line of edge-adjacent pixels from the detection result. For example, in S403, the image analysis unit 210 can detect pixels between the "1dot2Space vertical line" or the "1dot2Space horizontal line". The "1dot2Space vertical line" and the "1dot2Space horizontal line" respectively refer to two vertical lines and horizontal lines each having a width of one dot and spaced two dots apart. The image analysis unit 210 can detect such pixels using appropriate pattern information. In this case, the color separation and quantization unit 211 does not add end information indicating the "third end" to the quantization data for the pixels between the "1dot2Space vertical line" or the "1dot2Space horizontal line". FIG. 28(d) shows the detection result of the third end according to such conditions. In this way, by detecting the edge area using appropriate pattern information, it is possible to avoid limiting the number of dots of yellow ink in two consecutive pixels in a direction perpendicular to the edge. It should be noted that the pattern information used to avoid consecutive pixels detected as the "third end" is not limited to the above.

[0144] The types of edges detected by the image analysis unit 210 are not limited to those described above. As described above, the image analysis unit 210 can detect various other types of edges using pattern matching. The color separation and quantization unit 211 and the nozzle separation processing unit 212 can then appropriately control the arrangement or number of dots, taking into account various types of edges.

[0145] (Variation 4) In the above embodiment, in S402, the image analysis unit 210 generates a binary image by performing binarization processing on a grayscale image (e.g., a luminance image). However, the image analysis unit 210 may generate an N-value image other than a binary image. For example, the image analysis unit 210 can convert the luminance Y into ternary data (Bin) according to the following formula (3). IF Y > Th_1 :Bin = 0 else IF Th_1 ≧ Y > Th_2:Bin = 1 else :Bin = 2 Equation (3)

[0146] When performing binarization, all brightness values ​​Y lower than the threshold value Th are converted to "1". Therefore, edge information is not added to the area of ​​pixels having such brightness values ​​Y, and therefore dot placement control according to the edge information is not performed. On the other hand, by converting the brightness Y to ternary data, it is possible to increase the types of edges detected in S403. For example, it is possible to set the threshold value Th_1 to the same value as the threshold value Th, and further set one threshold value Th_2. With this configuration, it is possible to specify the brightness area in which the edge is detected in more detail. In this way, it is possible to control the types of edges and pixels to be detected.

[0147] For example, using pattern matching, a "first end", "second end", or "third end" pixel at an edge between a region with a pixel value of "0" and a region with a pixel value of "2" can be detected as described above. Furthermore, a "first end", "second end", or "third end" pixel at an edge between a region with a pixel value of "0" and a region with a pixel value of "1" can be detected as described above.

[0148] Also, by using pattern matching, it is possible to detect pixels at the "third end" at the edge between the area of ​​pixel value "1" and the area of ​​pixel value "2". In this case, it is not necessary to detect the pixels at the "first end" and the "second end". The area of ​​pixel value "1" adjacent to the area of ​​pixel value "2" is a mid-tone brightness area that is darker than the area of ​​pixel value "0". With this configuration, the number of dots of color ink, for example yellow ink, is controlled, so that the decrease in visibility due to the blurring of the outline of black characters can be suppressed. Also, with this configuration, by controlling the color ink, it is possible to suppress bleeding between colors in the mid-tone brightness area.

[0149] (Variation 5) In the first embodiment, an edge is detected from the binary image in S303. Then, the edge information for controlling the dot arrangement pattern for the first group color and the edge information for controlling the dot arrangement pattern for the second group color are set based on the same edge detection result. However, the binary image (or N-value image) used to detect the edge may be different for each color group or each ink color. For example, the threshold value used when converting the grayscale image into a binary image may be different for each color group or each ink color. With this configuration, it is possible to adjust the determination of whether to add edge information for each color group or each ink color.

[0150] (Other variations) In the first embodiment, an example in which the edge pixel is the top, bottom, left, and rightmost end, that is, one pixel inside the edge, has been mainly described. However, as already shown, the second pixel from the end may also be treated as an edge pixel. In this way, the first edge pixel (edge ​​pixel) and the second edge pixel (edge ​​adjacent pixel) may each be a predetermined number of pixels on both sides of the edge. In this case, a predetermined number of pixels from the end can be treated as edge pixels. With this configuration, it becomes easier to improve the sharpness of images such as characters or lines. This predetermined number, that is, the width of the edge area included in the edge, may be predetermined corresponding to the recording mode of the printer, or may be set by the user. For example, the operation mode of the printer may include a first recording mode in which this predetermined number is a first value, and a second recording mode in which this predetermined number is a second value different from the first value.

[0151] Furthermore, the threshold value Th may be set according to the above-mentioned predetermined number. An example of the setting will be described with reference to FIG. 48 and FIG. 49-1 to FIG. 49-3. FIG. 48(a) shows an image in which a background with a luminance value of 255 and a character 481 with a luminance value of 0 are arranged in an area 482, and a background with a luminance value of 254 and a character 481 with a luminance value of 0 are arranged in an area 483 adjacent to the area 482. FIG. 48(b) shows an image in which a background with a luminance value of 226 and a character 481 with a luminance value of 0 are arranged in an area 484, and a background with a luminance value of 225 and a character 481 with a luminance value of 0 are arranged in an area 485 adjacent to the area 484. FIG. 48(c) shows an image in which a background with a luminance value of 29 and a character 481 with a luminance value of 0 are arranged in an area 486, and a background with a luminance value of 28 and a character 481 with a luminance value of 0 are arranged in an area 487 adjacent to the area 486.

[0152] The dot arrangements obtained when the images shown in each of Figs. 48(a) to (c) are converted into nozzle data using a threshold value Th=225 according to the flow chart shown in Fig. 3A and the second pixel from the edge is processed in the same way as the first pixel is processed, are shown in Figs. 49-1(a) to (c). Here, the pixel group 491 at the center of the horizontal line consisting of three pixels wide in the figure is processed as the second end in Fig. 14. Also, dots with a luminance value of 254 corresponding to the background are omitted because their number is small. Also, dots with a luminance value of 225 corresponding to the background are expressed in the same way as dots with a luminance value of 226 corresponding to the background because the difference between them is small and dots with a luminance value of 226 corresponding to the background. Similarly, dots with a luminance value of 28 corresponding to the background are expressed in the same way as dots with a luminance value of 29 corresponding to the background because the difference between them is small and dots with a luminance value of 29 corresponding to the background. As shown in Figure 49-1(b), when the background luminance value is 226, two pixels' worth of dots are thinned out from the end of the edge of the character, whereas when the background luminance value is 225, the dots on the edge of the character are not thinned out. In this way, in adjacent areas where the difference in background luminance value is small, character images in which edge dots are thinned out and character images in which dots are not thinned out are mixed. When there are a large number of edge pixels subject to dot thinning and there is significant ink bleeding on the recording medium, the quality difference between the character images tends to become greater.

[0153] According to the inventor's study, when ink bleeding on a recording medium is relatively large, such quality differences are less noticeable when the brightness of the background is relatively low. In this case, even if the number of pixels to be thinned out of dots at the edge is large, the quality differences are likely to be hidden by the density of the background. Figures 49-2 (d) to (f) show dot arrangements obtained when a threshold value Th=28 is used. By reducing the threshold value Th when the number of pixels to be thinned out of dots is large, as shown in Figures 49-2 (d) and (e), dots at the edge of characters are thinned out in an image with a high-brightness background in which the quality difference due to the presence or absence of thinning is easily noticeable. On the other hand, as shown in Figure 49-2 (f), the presence or absence of thinning is controlled in an image with a low-brightness background in which the quality difference due to the presence or absence of thinning is relatively less noticeable.

[0154] On the other hand, when the ink bleeding on the recording medium is relatively small, such a difference in quality becomes less noticeable when the background brightness is relatively high. Even if the number of pixels to be thinned out at the edge is large, the faintness of the color of the edge pixel group 492 to be thinned out is relatively close to the faintness of the background, so the difference in quality is relatively difficult to see. Figures 49-3(g) to (i) show dot arrangements obtained when a threshold value Th=254 is used. By increasing the threshold value Th when the number of pixels to be thinned out is large, as shown in Figures 49-3(h) and (i), dots at the edge of characters are not thinned out in an image having a low-luminance background where the pixels to be thinned out tend to be relatively faint compared to the background. On the other hand, as shown in Figure 49-3(g), dots at the edge of characters are thinned out only in an image having a high-luminance background where the pixels to be thinned out tend to be faint.

[0155] In this manner, in this modified example, different threshold values ​​Th are set according to the number of pixels to be subjected to dot thinning at the edge, i.e., the above-mentioned predetermined number. However, the same threshold value Th may be set regardless of the above-mentioned predetermined number. For example, the threshold value Th may be set according to the recording mode of the printer. That is, the operation mode of the printer may include a first recording mode in which the threshold value Th is a first value, and a second recording mode in which the threshold value Th is a second value different from the first value.

[0156] As an example, a first threshold value (e.g., Th=225) may be set in recording mode 1. In this case, the same threshold value (e.g., Th=225) may be set in recording mode 2, which is different from recording mode 1. Furthermore, another threshold value (e.g., Th=28) may be set in recording mode 3, which is different from recording mode 1 (and 2). Here, the threshold value set in recording mode 1 (and 2) may correspond to a lighter color in the grayscale image than the threshold value set in recording mode 3. Also, the width of the edge region in recording mode 1 may be different from the width of the edge region in recording mode 3 (and 2). For example, in recording mode 1, the edge pixel may be one pixel inside the extreme edge on the top, bottom, left, and right. Also, in recording modes 2 and 3, the second pixel from the extreme edge may be treated as an edge pixel. In this configuration, the user can select a recording mode from recording mode 1 and recording mode 3, which has a larger number of pixels to be thinned out at the edge than recording mode 1 and thins out dots at the edge even when the background brightness is low. Furthermore, the user can also select recording mode 2, which has a larger number of pixels to be thinned at the edge than recording mode 1, and does not change the condition (threshold Th) related to the brightness of the background when thinning out the dots. Recording mode 3 can be set as a mode for recording a high-quality (or clear) image with higher sharpness of characters and thin lines than recording mode 1 when ink bleeding on the recording medium is relatively large. Recording mode 3 can also be set as a mode for recording a code image (e.g., a barcode image, a QR code (registered trademark) image, or a machine-readable code image) with higher quality than recording mode 1. Recording mode 2 can be set as a mode for recording a code image with higher quality (or clearer) than recording mode 1, regardless of the ink bleeding on the recording medium. Of course, it is not necessary to make all of recording modes 1 to 3 selectable. For example, only recording modes 2 and 3 may be selectable.

[0157] [Embodiment 2] In the first embodiment, the number of ink dots in edge pixels and edge adjacent pixels is limited based on edge information. Alternatively, edge pixels may be emphasized based on edge information. In this embodiment, a process is performed to enhance the visibility of characters or thin lines by emphasizing edge pixels.

[0158] As described above, the image processing unit 208 generates nozzle data broken down into nozzle units from the input image data according to the recording conditions included in the recording job. The recording device can operate in a recording mode such as a draft mode or an eco mode in which the amount of ink used for recording is suppressed more than in the standard mode. In such a case, the recording conditions can include information indicating the recording quality, for example, information indicating the draft mode or the eco mode. In such a mode, the amount of ink may be reduced by half compared to the standard mode. If the amount of ink is half that of the standard setting, the resulting recorded matter will be light overall. In such a mode, the amount of ink constituting black characters and thin lines is also reduced by half, so that bleeding of black ink on the recording medium is less likely to occur. However, for example, when pixels adjacent to characters or thin lines are high-density color pixels, contrast may decrease and visibility may decrease. Therefore, visibility can be improved by detecting edge pixels adjacent to high-density color pixels and emphasizing the detected edge pixels. The process of detecting edge pixels adjacent to high-density color pixels will be described below. The case where a mode in which the amount of ink is reduced by half is specified as the recording conditions will be described below. This mode will be referred to as a draft mode below.

[0159] The configurations of the printing system and image processing device according to this embodiment are the same as those in Figs. 1-2 and 27. The process according to this embodiment can also be performed according to the flowchart shown in Fig. 3(a). The following describes the differences from the first embodiment.

[0160] FIG. 29 shows an internal processing flow of the image analysis process performed in S303 in this embodiment. In S2901, the image analysis unit 210 converts the input image into a grayscale image. The grayscale image used in this embodiment indicates, for each pixel, the amount of at least one recording material used to record an image on a recording medium according to each pixel value of the input image. For example, the grayscale image can indicate ink amount information A for each pixel. The ink amount information A indicates the total amount of ink of at least one color that is applied to each pixel. In this embodiment, the image analysis unit 210 converts the pixel values ​​of the input image into ink amount information A. For example, the image analysis unit 210 can convert the information of three channels of RGB of the input image into ink amount information. In this way, the image analysis unit 210 can generate a grayscale image indicating the ink amount information for each pixel.

[0161] As a specific conversion method, the image analysis unit 210 can generate such a grayscale image by converting pixel values ​​of an input image according to a conversion table. Here, the image analysis unit 210 can use a conversion table according to the recording mode of the recording device. For example, conversion tables corresponding to the draft mode and other modes can be stored in memory. In this embodiment, the image analysis unit 210 can obtain the total ink amounts of cyan ink and magenta ink corresponding to the RGB values ​​in the set recording mode by referring to the LUT stored in advance in the memory.

[0162] The conversion to ink amount information is performed in order to grasp the ink amount for each pixel. In this embodiment, the inks to be determined for the ink amount are cyan ink and magenta ink. In one embodiment, the yellow ink amount is not taken into consideration. This is because, as described above, the lightness of yellow ink is high, so even if the ink amount of yellow ink increases, the lightness change on the recording medium is gradual. As described above, when a pixel adjacent to a character or a thin line is a high-density color pixel, the contrast decreases, so in such a case, visibility can be improved by emphasizing the edge. Therefore, by performing edge emphasis on a pixel adjacent to a pixel with a large total ink amount of cyan ink and magenta ink, visibility is effectively improved. On the other hand, in order to suppress the occurrence of bleeding between colors due to edge emphasis, the yellow ink amount is not taken into consideration in this embodiment.

[0163] FIG. 30 shows the relationship between the RGB values ​​of the input image and the ink color data. FIG. 30(a) shows the density values ​​of the corresponding ink colors when the RGB values ​​change from white to blue and then to black. When the RGB values ​​indicate white, the density values ​​of all the ink colors are 0. It can also be seen that blue is expressed as a secondary color of cyan ink and magenta ink. In this example, almost equal amounts of cyan ink and magenta ink are applied. Although it depends on the color of the cyan ink and magenta ink, blue is often expressed with equal amounts of these two colors of ink unless a special color ink is used. Similarly, green is expressed as a secondary color of cyan ink and yellow ink. Red is also expressed as a secondary color of magenta ink and yellow ink. In this case, the amount of these inks applied is equal, and the amount of ink applied often increases as the green and red become darker.

[0164] Figure 30(b) shows the density values ​​of each corresponding ink color when the RGB values ​​change from white to yellow and then to black. It can be seen that yellow is expressed as the primary color of yellow ink. Similarly, RGB=(0, 255, 255) corresponding to cyan is often expressed as the primary color of cyan ink, and RGB=(255, 255, 0) corresponding to magenta is often expressed as the primary color of magenta ink.

[0165] The density value data of each color corresponding to the input RGB values ​​can be prepared in advance as an ink color separation table. Based on such a table, an LUT can be generated that converts the pixel values ​​of the input image referred to in S2901 into ink amount information. For example, for each printing condition, the ink color separation table can be referenced to obtain the ink amount of each ink color corresponding to the RGB values, and the above total ink amount can be calculated based on the ink amount of each ink color. The LUT thus obtained, which indicates the total ink amount corresponding to the RGB values, can be stored in memory in advance.

[0166] In S2902, the image analysis unit 210 performs threshold processing on the grayscale image obtained in S2901, and generates an N-value image that indicates the result of the threshold processing. In this embodiment, the image analysis unit 210 converts the ink amount information A into binary data for edge detection. In this embodiment, as an example, the image analysis unit 210 converts the ink amount information A into binary data (Bin) according to the following formula (4), using a threshold Th that is preset in accordance with the recording mode of the recording apparatus. The threshold Th will be described later. The binary data generation formula is merely an example, and the method of binary conversion is not particularly limited. For example, the design of the inequality condition and the form of the formula may be different. IF Ink amount information A > Th : Bin = 0 else : Bin = 1 Equation (4)

[0167] In this embodiment, pixels in the input image where the total amount of cyan ink and magenta ink is equal to or greater than the threshold value Th are detected. The number of black ink dots for pixels adjacent to these pixels is then controlled. Therefore, based on the relationship between the density of color pixels adjacent to pixels formed with black ink and the effect of improving visibility by emphasizing edges and increasing the contrast of edge portions, the threshold value Th can be determined so as to obtain such an effect.

[0168] In S2903, the image analysis unit 210 detects an edge pattern using the N-value image obtained in S2902. In this embodiment, the image analysis unit 210 detects an edge pattern in a binary image. An example of pattern information used to detect an edge pattern has already been described in the first embodiment. In this embodiment, the image analysis unit 210 detects a portion adjacent to any end (third end). The image analysis unit 210 can output "3" as a determination result for a pixel at the third end. Also, it can output "0" as a determination result for other pixels. Here, the pixel at the third end is a pixel adjacent to a pixel whose total ink amount of cyan ink and magenta ink is equal to or greater than the threshold value Th. Note that, as shown in FIG. 30, a portion constituting a black character or a thin line is highly likely to exist in the area of ​​Bin=1. Therefore, a pixel at the end of a black character or a thin line adjacent to a pixel with a large total ink amount is highly likely to be at the third end.

[0169] 31 shows a flowchart of the color separation process in S304 and the quantization process in S305 performed by the color separation and quantization unit 211. S3101 to S3102 are the same as S1001 to S1002.

[0170] In S3103 to S3105, color separation and quantization unit 211 performs gradation correction processing. In S3103, color separation and quantization unit 211 determines whether the pixel to be processed has been detected as the "third end" and the pixel to be processed is density value K. If the pixel to be processed is the "third end", color separation and quantization unit 211 performs a second gradation correction processing on density value K in S3104. If the pixel to be processed is not the "third end", color separation and quantization unit 211 performs a first gradation correction processing on density value K in S3105.

[0171] FIG. 32 shows an example of the settings of the first tone correction process and the second tone correction process. In FIG. 32, In corresponds to the density value K which is the input value to the tone correction process, and Out corresponds to the density value K' which is the output value of the tone correction process. In this embodiment, when the draft mode is set as the recording condition, recording according to the draft mode is realized by the tone correction process. In this example, the output value is half of the input value in the first tone correction. For example, In=255 corresponds to OUT=128. In this way, the first tone correction reduces the amount of ink used by half compared to the standard setting. On the other hand, in the second tone correction, a larger output value is obtained compared to the first tone correction for the same input value. In this example, the amount of ink used is reduced by half in the first tone correction, while the second tone correction is performed so that the amount of ink used is twice as much as in the first tone correction. For example, In=255 corresponds to Out=255. In this way, the first tone correction and the second tone correction are switched based on the information of the "third end". In this way, it is possible to increase the amount of black ink for pixels adjacent to high density color pixels.

[0172] The second tone correction may be performed so that the amount of black ink at the end of the high-density black region is increased and the amount of black ink at the end of the gradation region is maintained. For example, in the second tone correction, the density value may not be corrected if the input value is smaller than a threshold value, and the density value may be corrected if the input value is equal to or greater than the threshold value. This threshold value may be, for example, 255.

[0173] In this embodiment, information about the "third end" is not used in the gradation correction process for the density values ​​C, M, and Y. In S3105, the color separation and quantization unit 211 performs a first gradation correction for the density values ​​C, M, and Y. That is, the first gradation correction for the density values ​​C, M, and Y reduces the values ​​of the density values ​​C', M', and Y' by half.

[0174] In S3106, the color separation and quantization unit 211 performs a quantization process on the density values ​​C', M', Y', and K'. The color separation and quantization unit 211 performs a predetermined quantization process to convert the density values ​​C'', M'', Y'', and K'' into 2-bit 3-value quantized data of "00", "01", or "10".

[0175] In S306, the nozzle decomposition processing unit 212 performs index expansion processing on the quantized data C", M", Y", and K" obtained in S305. In the index expansion processing in this embodiment, a prepared index pattern is used to convert the 600 x 600 dpi quantized data C", M", Y", and K" into 600 x 1200 dpi composite nozzle data Cp, Mp, Yp, and Kp. In this example, data for one pixel is converted into data for two pixels linked vertically.

[0176] FIG. 33 shows an example of a dot arrangement pattern used in the index expansion process. When the quantized data indicates "00", no dot is arranged in either of the two pixels. When the quantized data indicates "01", a dot is arranged in the upper pixel of the two pixels in pattern A, and a dot is arranged in the lower pixel of the two pixels in pattern B. When the quantized data indicates "10", a dot is arranged in both of the two pixels. In this embodiment, the reference index pattern shown in FIG. 11(b) is used. When the quantized data for each pixel is all "01", the composite nozzle data obtained by the index expansion process is the same as that shown in FIG. 11(c).

[0177] Through the above processing, composite nozzle data of 600 x 1200 dpi is obtained based on each pixel of input image data of 600 x 600 dpi. This composite nozzle data specifies whether or not each nozzle is to print.

[0178] The process of converting the image data shown in Fig. 5 into nozzle data will be further described below with reference to Figs. 34 to 36 in accordance with the flowchart shown in Fig. 3(a). Figs. 34(a) to (c) show binary images obtained by carrying out the processes of S2901 and S2902 on the image data shown in Figs. 5(a) to (c). Figs. 35(a) to (c) show the results of edge detection processing in S2903 on the binary images shown in Figs. 34(a) to (c). Here, information indicating whether each pixel is located at the third edge is obtained as edge information.

[0179] The values ​​of all pixels in Fig. 34(a) and (b) are "0". In the input image shown in Fig. 5(a), the black letter "E" is placed on a white background. As shown in Fig. 30(a), the ink amounts of cyan ink and magenta ink, which correspond to white and black, are 0. In the input image shown in Fig. 5(b), the black letter "E" is placed on a yellow background. As shown in Fig. 30(b), the ink amounts of cyan ink and magenta ink used to express yellow are 0. Therefore, the ink color information A for each pixel in these input images is 0, and the value after binarization is also 0. In these cases, as shown in Figs. 35(a) and (b), no edges are detected in S2903.

[0180] In the input image shown in FIG. 5(c), the black character "E" is placed on a blue background. As shown in FIG. 30(a), blue is expressed as a secondary color of cyan ink and magenta ink. Therefore, when the ink amount information A is binarized using a predetermined threshold value Th, the ink amount information A for the blue pixels is converted to "1". On the other hand, as described above, the ink amount information A for the black pixels is converted to "0". By performing pattern matching on the binary image shown in FIG. 34(c) obtained in this way, a "third end" that is a pixel adjacent to one of the ends is detected. FIG. 35(c) shows the detection result of the third end. In this way, pixels adjacent to high-density color pixels can be detected. In this example, the pixel that constitutes the black character "E" adjacent to the blue pixel (i.e., the third end) is detected.

[0181] Figures 36(a)-(c) show density values ​​K' obtained by performing the processes S3101-S3105 shown in Figure 31 on the input images shown in Figures 5(a)-(c) based on the edge information shown in Figures 35(a)-(c). Since edge information is not detected from the input images shown in Figures 5(a) and (b), a first gradation correction process is performed on each pixel in S3105. As a result, a pixel with a density value K=255 is converted to a density value K'=128.

[0182] For the input image shown in FIG. 5(c), the second gradation correction process is performed on the pixel detected as the "third end" as shown in FIG. 35(c), and the first gradation correction process is performed on the other pixels. Therefore, the density value K=255 of the pixel detected as the "third end" is converted to a density value K'=255. Furthermore, the density value K=255 of the other pixels is converted to a density value K'=128. In this way, a process for obtaining the density value K' of the black ink for the pixel adjacent to the high density color pixel and a process for obtaining the density value K' of the black ink for the other pixels are switched. Note that the first gradation correction process is performed on the density values ​​C, M, and Y as described above. In this way, correction is performed so that the density values ​​C', M', and Y', which are the output values, are halved.

[0183] FIG. 37 shows quantized data obtained by the quantization process in S3106. FIGS. 37(a) to (c) show quantized data K" obtained by quantizing the density value K' shown in FIGS. 36(a) to (c). FIG. 37(d) shows quantized data C", M", and Y" obtained by performing the processes S3101 to S3106 on the input image shown in FIG. 5(a). FIG. 37(e) shows quantized data Y" obtained by performing the processes S3101 to S3106 on the input image shown in FIG. 5(b). FIG. 37(f) shows quantized data C" and M" obtained by performing the processes S3101 to S3106 on the input image shown in FIG. 5(c).

[0184] FIG. 38 shows dot arrangement patterns obtained by index expansion processing based on the dot arrangement patterns and reference index patterns shown in FIG. 33 and FIG. 11(b) in S305. FIG. 38(a)-(c) shows dot arrangement pattern K corresponding to FIG. 5(a)-(c). FIG. 38(d) shows dot arrangement patterns C, M, and Y corresponding to FIG. 5(a). FIG. 38(e) shows dot arrangement pattern Y corresponding to FIG. 5(b). FIG. 38(f) shows dot arrangement patterns C and M corresponding to FIG. 5(c). In dot arrangement pattern K shown in FIG. 38(c), two dots are arranged at the edge portion of black character E, and more dots are arranged than in other pixels of black character E. In addition, in the other dot patterns shown in FIG. 38(a)-(b) and (d)-(e), it can be seen that printing is performed with an ink amount (i.e., the number of dots) according to the draft mode set as a printing condition.

[0185] In the case of this embodiment, with regard to the quantized data K", if a pixel is on an edge (third end), then as a result of the second gradation correction process, a maximum of two dots will be placed on this pixel. In this case, the maximum recording rate is 100%. Also, if the pixel is not on an edge (third end), then as a result of the first gradation correction process, a maximum of only one dot will be placed on this pixel. For this reason, the maximum recording rate is 50%. In this way, the nozzle resolution processing unit 212 can generate recording data such that the maximum recording amount (i.e. the number of dots) for pixels on an edge is greater than the maximum recording amount for pixels that are not on an edge. Also, in this way, the nozzle resolution processing unit 212 can generate recording data such that the maximum recording rate for pixels on an edge is greater than the maximum recording rate for pixels that are not on an edge.

[0186] FIG. 39 shows the final dot arrangement obtained by performing the processes of S301 to S306 on the input image shown in FIG. 5 in this embodiment. FIG. 39(a) corresponds to FIG. 5(a) and shows the dot arrangement when the adjacent color of the black character "E" composed of black dots indicated by K_1 is white. In this embodiment, the draft mode is set as the printing condition. Therefore, the number of black dots is half (1 dot per pixel) compared to the first embodiment. As described above, since the ink amount is small, the influence of ink bleeding caused only by black ink is small. Therefore, in such a mode, it is not necessary to further thin out the black dots based on the edge information as in the first embodiment. With such a configuration, it is possible to improve the contrast and increase the visibility.

[0187] FIG. 39(b) corresponds to FIG. 5(b) and shows a dot arrangement in which the density of color dots of ink CL_1 adjacent to the black character "E" composed of black dots indicated by K_1 is low. In FIG. 39(b), CL_1 is yellow ink. In this embodiment, the draft mode is set as the printing condition. Since the number of black dots is reduced, bleeding between colors is also suppressed compared to the standard setting. Therefore, in such a mode, it is not necessary to further thin out the black dots and yellow dots based on the edge information as in the first embodiment. With such a configuration, it is possible to improve the contrast and increase visibility.

[0188] FIG. 39(c) corresponds to FIG. 5(c) and shows a dot arrangement in which the color dots of ink CL_2 adjacent to the black character "E" composed of black dots designated K_1 have a high density. In FIG. 39(c), CL_2 is composed of magenta ink and cyan ink. Unlike FIGS. 39(a) and (b), it can be seen that the number of black dots designated K_1 is controlled so that the number of black dots at the edge portions of the black character "E" is greater than the non-edge portions. In this way, when the adjacent color pixels are high density, the contrast of the black character "E" can be increased by increasing the number of black dots at the edge portions of the black character "E".

[0189] According to this embodiment, when the pixels adjacent to the edge of the portion where black ink is printed are high-density color pixels, the visibility can be improved. Specifically, the visibility of the characters or thin lines can be improved by emphasizing the pixels that constitute the characters or thin lines and that are adjacent to the high-density color pixels.

[0190] [Embodiment 3] In this embodiment, printing in each pass in multi-pass printing is controlled according to the detection results of edge pixels. In this embodiment as well, edge information is selectively added to pixels in an input image that belong to a specific signal value region. Nozzle data can be generated according to such edge information so that printing of black ink at edge pixels is biased to certain passes. Furthermore, nozzle data can be generated according to such edge information so that printing of color ink at edge adjacent pixels is biased to certain passes.

[0191] The configurations of the printing system and image processing device according to this embodiment are the same as those in Figs. 1-2 and 27. The process according to this embodiment can also be performed according to the flowchart shown in Fig. 3(a). The following describes the differences from the first embodiment.

[0192] In this embodiment, the recording apparatus performs multi-pass printing to form an image on a recording medium by multiple recording scans. First, multi-pass printing will be described. In this embodiment, bidirectional two-pass recording is performed. FIG. 40 is a schematic diagram for explaining bidirectional two-pass multi-pass recording executed under the control of the CPU 203 in the printer 2. Here, for ease of explanation, a recording operation using the black nozzle row 2701 out of the multiple nozzle rows of the recording head H shown in FIG. 27 will be described.

[0193] In the two-pass printing, printing is performed on the printing area 1 by the first printing scan and the second printing scan. In the first printing scan, the printing head H moves in the +X direction, which is the forward direction. Also, while the printing head is moving, the black nozzle row 2701 performs a discharge operation on the printing area 1. In the second printing scan, the printing head H moves in the -X direction, which is the backward direction opposite to the first printing scan. Also, while the printing head is moving, the black nozzle row 2701 performs a discharge operation on the printing area 1. The printing medium is not conveyed between the first printing scan and the second printing scan. After the second printing scan, the printing medium is conveyed in the -Y direction. The conveyance amount corresponds to the nozzle array length in the sub-scanning direction. FIG. 40 shows the relative positional relationship between each printing area and the printing medium. As shown in FIG. 40, the black nozzle row 2701 moves relatively in the +Y direction with respect to the printing medium.

[0194] Next, printing is performed on the printing area 2 by the third printing scan and the fourth printing scan. In the third printing scan, the printing head H moves in the +X direction, which is the forward direction, as in the first printing scan. Also, the black nozzle row 2701 performs a discharge operation on the printing area 2. In the fourth printing scan, the printing head H moves in the opposite -X direction, as in the second printing scan. Also, the black nozzle row 2701 performs a discharge operation on the printing area 2. Furthermore, after the fourth printing scan, the printing medium is transported in the -Y direction. In this way, two printing scans are performed on the same printing area, and the printing medium is repeatedly transported in the -Y direction, thereby performing two-pass multi-pass printing in both directions.

[0195] In this embodiment, 4-bit quantized data C″M″Y″K″ is generated for each ink color according to S301 to S305 shown in FIG. 3(a). The generation of quantized data can be performed in the same manner as in Modification 1. Here, the first group color may be black, and the second group color may be at least one of cyan, magenta, and yellow. In the following example, K is classified into the first group color, and Y, C, and M are classified into the second group color. That is, the quantized data K″ is 4-bit data in which end information of “first end” or “second end” is added to edge pixels. In addition, the quantized data C″, M″, and Y″ are 4-bit data in which end information of “third end” is added to edge adjacent pixels. Hereinafter, the quantized data C″, M″, and Y″ are collectively referred to as quantized data CL″.

[0196] In the following, a case will be described in which processing is performed using the input images shown in Figures 5(a) to (c). In this case, the quantized data is as shown in Figures 15(a) to (f).

[0197] In S306, the nozzle resolution processing unit 212 performs index expansion processing on the quantized data C", M", Y", and K" obtained in S305. The nozzle resolution processing unit 212 performs index expansion processing based on the dot arrangement patterns shown in Figs. 41(a)-(b) and the reference index pattern shown in Fig. 11(b). The nozzle resolution processing unit 212 converts the quantized data of 600 x 600 dpi into index data of 600 x 1200 dpi. Here, as in the first embodiment, data of one pixel is converted into data of two pixels connected in the vertical direction. In the first embodiment, the composite nozzle data has 1-bit data corresponding to each nozzle and indicating printing or non-printing. However, in this embodiment, the index data has 2-bit data corresponding to each nozzle. This index data is set according to edge information contained in the quantized data. In this embodiment, by combining this index data with a multi-value (2-bit) mask pattern, printing in each pass of multi-pass printing is controlled for pixels to which edge information is added.

[0198] FIG. 41(a) shows a dot arrangement pattern used in the index expansion process for pixels to which edge information is not added. If the upper two bits of the quantized data for a pixel are "00", that is, if the quantized data is "0000", "0001", or "0010", edge information is not added to this pixel. For such pixels, "00" or "01" is assigned as index data. Specifically, if the lower two bits of the quantized data are "00", "00" is set as index data for both two pixels. If the lower two bits of the quantized data are "01", in pattern A, index data "01" is set for the lower pixel of the two pixels, and index data "00" is set for the upper pixel. In addition, in pattern B, index data "01" is set for the upper pixel of the two pixels, and index data "00" is set for the lower pixel. If the lower two bits of the quantized data indicate "10", "01" is set as index data for both two pixels.

[0199] Figure 41(b) shows the dot arrangement pattern used in the index expansion process for pixels that have been assigned edge information. If the upper 2 bits of the quantized data for a pixel are not "00", then the pixel has been assigned edge information. For such pixels, index data of "00" or "10" is assigned. The method of assigning index data is the same as in Figure 41(a), except that index data "10" is assigned instead of "01".

[0200] In the first embodiment, the composite nozzle data i is set with information indicating printing or non-printing of each nozzle. On the other hand, in this embodiment, in order to control the printing pass, the information indicating printing of each nozzle is classified into "10" and "01" depending on the presence or absence of edge information. In this example, both of the index data "10" and "01" are information indicating that ink is ejected to form dots. Incidentally, in the first embodiment, for pixels to which edge information is not given, when the quantization value obtained in S1004 increases to 0, 1, and 2, the number of dots increases to 0, 1, and 2. On the other hand, for pixels to which edge information is given, when the quantization value increases to 0, 1, and 2, the number of dots is adjusted to 0, 1, and 1. In the above example, in order to make it easier to understand the effect of this embodiment, the index data is set so that the number of dots does not change depending on the presence or absence of edge information. Specifically, regardless of the presence or absence of edge information, when the quantization value increases to 0, 1, and 2, the number of dots increases to 0, 1, and 2.

[0201] Figures 42(a) to (c) respectively show index data K obtained in this embodiment by performing index expansion processing on the quantized data K" shown in Figures 15(a) to (c) in S306. Also, Figures 42(d) to (f) respectively show index data CL obtained in S306 by performing index expansion processing on the quantized data CL" shown in Figures 15(d) to (f). The index data CL is data corresponding to any one of the C (cyan), M (magenta), and Y (yellow) color nozzle arrays.

[0202] In the index data K shown in Figures 42(a) and (b), the edge pixels (first and second ends) of the black character "E" are set to a value of "10," and the non-edge pixels inside are set to a value of "01." In Figure 42(c), since no edge pixels are detected, the pixels located at the ends of the black character "E" are not set to a value of "10," and are set to a value of "01," just like the internal pixels.

[0203] In the index data CL shown in Fig. 42(e), the edge adjacent pixels outside the edge pixels of the character "E" are set to a value of "10" in accordance with the edge information indicating the third edge. The other pixels whose quantization value is not 0 are set to a value of "01". In Fig. 42(f), since no edge pixels are detected, the edge adjacent pixels outside the edge pixels of the black character "E" are not set to a value of "10" and are set to a value of "01" like the other pixels.

[0204] In S306, the nozzle resolution processing unit 212 further generates composite nozzle data by applying a mask pattern to the index data generated as described above. The composite nozzle data in this embodiment is data that indicates printing or non-printing by each nozzle in each printing scan.

[0205] Fig. 43 shows an example of a mask pattern used in this embodiment. Fig. 43(a) and (b) respectively show a first mask pattern and a second mask pattern applied to index data K. The first mask pattern is a mask pattern used to generate nozzle data for a first pass print scan of two print scans. The second mask pattern is a mask pattern used to generate nozzle data for a second pass print scan. In this way, nozzle data for controlling printing by the black nozzle row 2701 is generated.

[0206] Similarly, Figures 43(c) and (d) respectively show the third mask pattern and the fourth mask pattern to be applied to the index data CL. The third mask pattern is a mask pattern used to generate nozzle data for the first pass print scan. The fourth mask pattern is a mask pattern used to generate nozzle data for the second pass print scan. In this way, nozzle data is generated to control printing by the C (cyan), M (magenta), and Y (yellow) color nozzle arrays.

[0207] Each mask pattern is 600 dpi x 1200 dpi data. Each pixel in each mask pattern has a 2-bit value of "00", "01", "10", or "11". Below, this value will be referred to as mask data. The meaning of the value will be explained later.

[0208] Fig. 44 is a table showing nozzle data corresponding to combinations of 2-bit index data and 2-bit mask data for each pixel. This table defines printing or non-printing for combinations of index data "00", "01", and "10" and mask data "00", "01", "10", and "11". In the table of Fig. 46, a "●" represents "printing" and a blank represents "non-printing". In other words, when the table represents "printing" for a combination of index data and mask data for a pixel, printing (i.e., ink ejection) is performed for that pixel.

[0209] The first mask pattern corresponding to the black nozzle row 2701 shown in Fig. 43(a) has only mask data "10" and "11". Also, as shown in Fig. 44, "printing" is defined for the combination of index data "10" and mask data "10" or "11". Therefore, printing by the black nozzle row 2701 in the first pass is performed on pixels to which index data "10" is assigned, that is, pixels to which edge information is added and where dots are formed.

[0210] On the other hand, the mask data of the second mask pattern corresponding to the black nozzle row 2701 shown in Fig. 43(b) is only "00" and "01". Also, as shown in Fig. 44, "non-printing" is defined for the combination of index data "10" and mask data "00" or "01". Therefore, printing by the black nozzle row 2701 in the second pass is not performed on pixels to which index data "10" is assigned, that is, pixels to which edge information is added and where dots are formed.

[0211] Furthermore, when the corresponding mask data is "01" or "11", the pixel to which the index data "01" is assigned is printed by the black nozzle row 2701. Here, the pixel to which the index data "01" is assigned is a pixel to which no end information is added and to which a dot is formed. In this way, the pixel to which the index data "01" is assigned is printed in the first pass when the mask data "11" is set in the first mask pattern, and is printed in the second pass when the mask data "01" is set in the second mask pattern. Incidentally, when the first mask pattern and the second mask pattern are superimposed, the mask data "11" in the first mask pattern and the mask data "01" in the second mask pattern complement each other. As a result, printing of the pixel to which the index data "01" is assigned is divided into two passes.

[0212] Unlike the mask patterns shown in Figures 43(a) and (b) corresponding to the black nozzle row 2701, the first mask pattern shown in Figure 43(c) corresponding to the color nozzle row has mask data only of "00" and "01". Also, the second mask pattern shown in Figure 43(d) corresponding to the color nozzle row has mask data only of "10" and "11". That is, printing by the color nozzle row for pixels assigned with index data "10" is performed in the second pass. Also, printing by the color nozzle row for pixels assigned with index data "01" is divided into the first and second passes.

[0213] As described above, the nozzle resolution processing unit 212 can generate print data such that the printing timing of the printing material of the first group color for edge pixels is biased to some of the printing scans among the multiple printing scans. Also, the nozzle resolution processing unit 212 can generate print data such that the printing timing of the printing material of the second group color for edge adjacent pixels is biased to other some of the multiple printing scans. In the above example, the printing by the black nozzle row 2701 for pixels to which edge information is assigned (index data = "10") is biased to some passes. On the other hand, the printing by the color nozzle row for pixels to which edge information is assigned (index data = "10") is biased to another pass. With this configuration, the printing timing of the black ink and the color ink can be shifted in the edge region. For example, the printing timing of the black ink for edge pixels constituting characters or lines and the printing timing of the color ink for edge adjacent pixels that are the background of these can be shifted in time. This can reduce bleeding due to contact of the ink before it permeates the print medium.

[0214] Figures 45(a)-(c) show pixels printed in a first pass when a mask pattern is applied to the index data K shown in Figures 42(a)-(c) and the index data CL shown in Figures 42(d)-(f). Figures 45(d)-(f) show pixels printed in a second pass in a similar case. Figures 45(a)-(f) show whether each pixel is printed by the black nozzles, the color nozzles, or by neither nozzle.

[0215] As shown in FIG. 45(b), all of the recording of the inner edge pixels of the letter "E" by the black nozzle array 2701 is performed in the first pass. Furthermore, recording of the outer edge adjacent pixels of the letter "E" by the color nozzle array is not performed in the first pass. On the other hand, recording of the inner edge pixels of the letter "E" by the black nozzle array 2701 is not performed in the second pass. Furthermore, recording of the outer edge adjacent pixels of the letter "E" by the color nozzle array is all performed in the second pass. In this way, recording of pixels to which edge information has been added by the black nozzle array 2701 can be biased toward the first pass, and recording of pixels to which edge information has been added by the color nozzle array can be biased toward the second pass.

[0216] On the other hand, as shown in FIG. 45(b), printing for pixels to which no edge information has been added is distributed to the first and second passes so as to avoid bias.

[0217] In the above example, the printing by the black nozzle array 2701 for pixels to which edge information has been added is biased to the first pass, and the printing by the color nozzle array for pixels to which edge information has been added is biased to the second pass. On the other hand, the mask pattern corresponding to the black nozzle array 2701 and the mask pattern corresponding to the color nozzle array may be interchanged. In this case, printing by the color nozzle array for pixels adjacent to an edge is performed in the first pass, and printing by the black nozzle array 2701 for edge pixels is performed in the second pass.

[0218] FIG. 46(a) shows a schematic diagram of a dot arrangement printed in the first pass when printing is performed as shown in FIG. 45(b). FIG. 46(b) shows a schematic diagram of a dot arrangement printed in the second pass when printing is performed as shown in FIG. 45(e). It can be seen that black dots are printed on the edge pixels of the character "E" in the first pass, and color dots are printed on all the adjacent edge pixels of the character "E" in the second pass. In this way, in multi-pass printing, the printing of black dots on the edge pixels of the character "E" and the printing of color dots on the edge adjacent pixels can be biased to different printing passes. This configuration can suppress bleeding caused by the ink coming into contact with the recording medium before the ink penetrates the paper.

[0219] FIG. 46(c) shows a state where the dot arrangement shown in FIG. 46(a) and the dot arrangement shown in FIG. 46(b) are superimposed without any misalignment. On the other hand, FIG. 46(d) shows a state where the dot arrangement obtained in the second pass shown in FIG. 46(b) is superimposed with a shift of −21 μm in the X direction with respect to the dot arrangement obtained in the first pass shown in FIG. 46(a). In this example, all edge pixels of the character “E” are printed in the first pass, so that the sharpness of the edge of the character is maintained even if the printing position shift occurs between passes, as shown in FIG. 46(b). In the case of FIG. 46(c), a continuous white background is generated along the boundary between the character “E” and the background due to the printing position shift between passes. However, as already described, in this embodiment, edge information is detected when the brightness of the background adjacent to the character is high (for example, in the case of FIG. 5(b)). Therefore, printing pass control in the edge area as shown in FIG. 46(a) and (b) is performed when the brightness of the background is high. For example, in FIG. 46(c), most of the color dots are yellow dots. In this way, in FIG. 46(c), the brightness of the background is sufficiently high, so the white background along the boundary is not easily noticeable. In this way, when the brightness of the background is sufficiently high, Based on the edge information, the recording of pixels in edge regions can be biased towards some passes.

[0220] On the other hand, in this embodiment, edge information is not detected when the brightness of the background adjacent to the character is low (for example, in the case of FIG. 5(c)). In such a case, control is performed so that the recording of pixels in the edge area is not biased toward some passes.

[0221] Fig. 47(a) shows a schematic diagram of the dot arrangement printed in the first pass when printing is performed as shown in Fig. 45(c). Fig. 47(b) shows a schematic diagram of the dot arrangement printed in the second pass when printing is performed as shown in Fig. 45(f). In this example, no edge information is added to any pixel. Therefore, for the entire image, the printing of black dots and the printing of color dots are divided into two passes.

[0222] FIG. 47(c) shows a state where the dot arrangement shown in FIG. 47(a) and the dot arrangement shown in FIG. 47(b) are superimposed without any misalignment. On the other hand, FIG. 47(d) shows a state where the dot arrangement obtained in the second pass shown in FIG. 47(b) is superimposed with a shift of −21 μm in the X direction on the dot arrangement obtained in the first pass shown in FIG. 47(a). Unlike FIG. 46(d), the printing in the edge area is also divided into two passes, so that the printing position shift between the passes causes the edge of the character to be disturbed. However, in this case, the contrast between the character part and the background part is relatively small, so the edge disturbance is difficult to perceive. On the other hand, a continuous white background does not occur along the boundary between the character part and the background part as seen in FIG. 46(d). When the brightness of the background is relatively low, for example, when most of the color dots in the background are composed of cyan ink or magenta ink, such a white background is easily noticeable. In this embodiment, control is performed so that the printing passes in the edge area are not biased in this case, so degradation of image quality can be suppressed.

[0223] As described above, in this embodiment, the printing of edge areas in each pass is controlled based on edge information detected from an N-value image. That is, edge information is detected according to the brightness of the background side, and control is performed to bias printing of edge pixels to different printing passes according to the color of the ink. With this configuration, when the brightness of the background is high, the sharpness of the characters is maintained even if there is a printing position shift between passes. On the other hand, when the brightness of the background is low, the occurrence of white background at the boundary between the characters and the background can be suppressed even if there is a printing position shift between passes.

[0224] In this embodiment, the method of classifying colors into groups is not particularly limited. For example, instead of classifying all of C, M, and Y into the second group color, any of C, M, and Y may be classified into the third group color. In this embodiment, dots of edge pixels and edge adjacent pixels are not thinned out. On the other hand, as in the first embodiment, dots of edge pixels or edge adjacent pixels may be thinned out. In this case, index data can be set so that when the quantization values ​​are 0, 1, and 2, the number of dots is 0, 1, and 1, respectively.

[0225] In the present embodiment, the recording apparatus performs two-pass recording. However, multi-pass printing with three or more passes may be performed. For example, when performing three-pass recording, all recording using black ink for edge pixels can be performed in the first pass, and all recording using color ink for edge adjacent pixels can be performed in the third pass.

[0226] In this embodiment, the printing of edge pixels, edge adjacent pixels, and non-edge pixels in each pass is controlled using a combination of 2-bit index data and a 2-bit mask pattern, but different mask patterns may be used for edge pixels and non-edge pixels.

[0227] Also, in this embodiment, all printing of edge pixels was performed in a specific pass. However, this is not limited to such a configuration. The reason for biasing the printing of edge pixels between printing passes is to suppress disturbance of dot alignment due to misalignment of printing positions between multiple printing passes. In other words, by making the printing rate for edge pixels in a specific pass higher than the maximum printing rate of each pass for non-edge pixels, disturbance of dot alignment in edge pixels compared to non-edge pixels is suppressed. Therefore, even in such a case, the effect of improving the sharpness of characters or lines can be obtained.

[0228] For example, for the printing material of the first group of colors, the printing rate of this pixel in the print scan in which the printing rate of the non-edge pixel is maximum can be made smaller than the printing rate of this pixel in the print scan in which the printing rate of the edge pixel is maximum. Specifically, in a four-pass print, if the printing rate of the non-edge pixel in each pass is 25%, the printing rates of the edge pixel in each pass for the first group of colors can be 0%, 50%, 0%, and 50%. In this example, the print is split into two passes. In this case, the printing rate of this pixel in the print scan in which the printing rate of the non-edge pixel is maximum is 25%, and the printing rate of this pixel in the print scan in which the printing rate of the edge pixel is maximum is 50%.

[0229] Furthermore, the print data can be generated so that the print scan in which the print rate of edge pixels by the print material of the first group color is maximized is different from the print scan in which the print rate of edge adjacent pixels by the print material of the second group color is maximized. Here, for the print material of the second group color, the print rate of this pixel in the print scan in which the print rate of a pixel not on an edge is maximized can be made smaller than the print rate of this pixel in the print scan in which the print rate of an edge adjacent pixel is maximized. Specifically, in the above four-pass printing, when the print rate of non-edge pixels in each pass for the second group color is 25%, the print rates of edge pixels in each pass can be 50%, 0%, 50%, and 0%. In this case, the print rate of this pixel in the print scan in which the print rate of non-edge pixels is maximized is 25%, and the print rate of this pixel in the print scan in which the print rate of edge pixels is maximized is 50%. In addition, the printing scans (second and fourth passes) in which the printing rate of edge pixels using the printing material of the first group of colors is maximized are different from the printing scans (first and third passes) in which the printing rate of edge-adjacent pixels using the printing material of the second group of colors is maximized.

[0230] If print head constraints limit the maximum printing rate in a single scan, then printing can be split into multiple passes so that the printing rate in a particular pass for edge pixels is greater than the maximum printing rate of each pass for non-edge pixels, as described above.

[0231] (Other embodiments) In the above embodiment, a case where printing is performed using a serial type recording device has been described. However, the configuration of the recording device is not limited to this. For example, the recording device may have a line head. Also, serial type ink heads may be arranged. Furthermore, in the above embodiment, a case where the recording device is an inkjet printer has been described. However, the configuration of the recording device is not limited to this. For example, the recording device may be a laser printer that performs printing using toner, or a copier.

[0232] In the above-described embodiment, the recording device records an image on the recording medium by applying recording materials of multiple colors to the recording medium. However, the recording device may record an image using only one color of recording material. In this case, for example, in the first embodiment, the sharpness of images such as printed characters or lines can be improved, while secondary degradation of image quality such as reduced visibility when the background density is high can be suppressed.

[0233] In the above embodiment, the luminance image showing the luminance value Y and the image showing the total ink amount A of cyan ink and magenta ink are given as the gradation image used for edge detection. However, the type of gradation image is not limited to these. For example, each pixel of the gradation image may be mounted on a printing device and may show the total ink amount of each ink used for recording. Also, a contribution rate may be set for each ink color. In this case, each pixel of the gradation image may show a weighted total ink amount of each ink used for recording based on the contribution rate. In this case, for example, the contribution rate of yellow ink can be lowered and the contribution rates of cyan, magenta, and black can be higher. As described above, an LUT showing these total ink amounts corresponding to RGB values ​​can be created in advance. When generating a gradation image, such an LUT can be referenced. Also, such an LUT may be prepared for each recording condition. The LUT for each recording condition can be created by obtaining ink amount information corresponding to RGB values ​​based on an ink color separation table prepared for each recording condition.

[0234] The gradation image may also be a brightness image corresponding to the input image. This brightness image may represent the brightness of the color on the recording medium obtained by performing recording on the recording medium according to the color information shown in the input image. Here, the brightness is a value indicating the brightness of the color, and the type is not particularly limited. For example, the LUT referred to when generating the gradation image may indicate the relationship between specific color information and the L* value of the CIE L*a*b* value obtained when an image according to this color information is recorded on the recording medium and the image is measured by a colorimeter or the like. The LUT referred to when generating the gradation image may also indicate the relationship between specific color information and the optical density of the color recorded according to this color information.

[0235] In the above-mentioned embodiment 1, the generation of nozzle data is controlled so that dots in edge regions are not thinned in regions where the background is low brightness. Also, in the above-mentioned embodiment 3, the generation of nozzle data is controlled so that paths are not biased in regions where the background is low brightness. For these purposes, in the above-mentioned embodiments, edges are detected in an N-value image (N is a natural number of 2 or more) that indicates the result of threshold processing on a grayscale image so that edges are not detected in regions where the background is low brightness. However, in order to achieve such purposes, it is not essential to perform edge detection on an N-value image. For example, the image analysis unit 210 can detect edges in a grayscale image (for example, a luminance image or a brightness image) by any method. At this time, the image analysis unit 210 may detect edges using an edge detection filter. In this case, the color separation quantization unit 211 and the nozzle separation processing unit 212 may generate print data based on the input image, the edge detection result, and the pixel value at the edge of the grayscale image. For example, the color separation quantization unit 211 and the nozzle separation processing unit 212 may generate recording data based on the input image, the edge detection results, and the luminance or brightness of the side with higher luminance or brightness (i.e., the background side) at the edge of the luminance image or brightness image.

[0236] As a specific example, the color separation quantization unit 211 may add the above-mentioned end information (upper 2 bits) to the quantization data for the edge pixel or edge adjacent pixel detected by the image analysis unit 210 when the luminance or lightness of the background side is equal to or greater than a threshold value. The nozzle separation processing unit 212 may generate the recording data as described above according to the quantization data thus generated. In another embodiment, the nozzle separation processing unit 212 may control the amount of dot thinning in the edge pixel or edge adjacent pixel according to the magnitude of the luminance or lightness of the background side. For example, when the luminance or lightness of the background side is higher, the amount of dot thinning in the edge pixel or edge adjacent pixel can be increased. In addition, the nozzle separation processing unit 212 can bias the recording of the pixel to some passes according to the magnitude of the luminance or lightness of the background side. For example, when the luminance or lightness of the background side is higher, the bias of the recording of the pixel to some passes can be increased.

[0237] Also, in the above-mentioned second embodiment, the generation of nozzle data is controlled so that the recording amount of adjacent black dots increases in an area where the total ink amount of the background is large. For this purpose, an edge is detected in an N-value image (N is a natural number of 2 or more) showing the result of threshold processing on a grayscale image showing the total ink amount. However, it is not essential to perform edge detection on an N-value image to achieve such a purpose. As described above, the color separation quantization unit 211 and the nozzle separation processing unit 212 may generate recording data based on an input image, an edge detection result, and a pixel value at an edge of the grayscale image. For example, the color separation quantization unit 211 and the nozzle separation processing unit 212 may generate recording data based on an input image, an edge detection result, and a total ink amount on the side where the total ink amount is large (i.e., the background side) at an edge of a grayscale image showing the total ink amount. As a specific example, when the total ink amount for an edge pixel detected by the image analysis unit 210 is equal to or greater than a threshold, the color separation quantization unit 211 may apply a second gradation correction process to an edge adjacent pixel adjacent to this edge pixel.

[0238] The present invention can also be realized by a process in which a program for implementing one or more of the functions of the above-described embodiments is supplied to a system or device via a network or a storage medium, and one or more processors in a computer of the system or device read and execute the program. The present invention can also be realized by a circuit (e.g., ASIC) that implements one or more of the functions.

[0239] The disclosure of this specification includes the following image processing device, image processing method, and program. (Item 1) 1. An image processing device for generating recording data for use in a recording device that records an image on a recording medium by depositing one or more colors of recording material on the recording medium in accordance with one or more colors of recording data, A detection means for detecting edges in an N-value image (N is a natural number equal to or greater than 2) that indicates the result of threshold processing on a grayscale image obtained from an input image; a generating means for generating the recording data based on the input image and the edge detection result; An image processing device comprising: (Item 2) 2. The image processing device according to item 1, wherein the grayscale image is a luminance image or a brightness image corresponding to the input image. (Item 3) the grayscale image indicates, for each pixel, a luminance or brightness on the recording medium when an image is recorded on the recording medium in accordance with each pixel value of the input image; 3. The image processing device according to any one of items 1 to 2, characterized in that the grayscale image is obtained by converting pixel values ​​of the input image according to a conversion table corresponding to the type of the recording medium. (Item 4) 4. The image processing device according to any one of items 1 to 3, characterized in that the grayscale image indicates for each pixel an amount of at least one recording agent used to record an image on the recording medium according to each pixel value of the input image. (Item 5) 5. The image processing device according to item 4, wherein the grayscale image is obtained by converting pixel values ​​of the input image according to a conversion table corresponding to a recording mode of the recording device. (Item 6) The image processing device according to any one of items 1 to 5, characterized in that the generating means generates the recording data corresponding to the pixel of interest of the input image for at least one color from the value of the pixel of interest of the input image in a manner depending on whether the pixel of interest of the input image is on the edge or not. (Item 7) 7. The image processing device according to any one of items 1 to 6, wherein the generating means generates the recording data so that a maximum recording amount for pixels on the edge is less than a maximum recording amount for pixels not on the edge. (Item 8) 7. The image processing device according to any one of items 1 to 6, wherein the generating means generates the recording data so that a maximum recording amount for pixels on the edge is greater than a maximum recording amount for pixels not on the edge. (Item 9) the recording device records an image on the recording medium by depositing a plurality of color recording materials onto the recording medium; The plurality of colors are classified into two or more groups, including a first group and a second group; the detection means detects a first edge pixel and a second edge pixel corresponding to both sides of the edge, The generating means generates the recording data for the colors of the first group based on a detection result of the first edge pixels, and generates the recording data for the colors of the second group based on a detection result of the second edge pixels. 9. The image processing device according to any one of items 1 to 8, (Item 10) Item 10. The image processing device described in item 9, characterized in that the generation means generates the recording data for the colors of the first group so that a maximum recording amount for the first edge pixels is less than a maximum recording amount for pixels not on the edge, and generates the recording data for the colors of the second group so that a maximum recording amount for the second edge pixels is less than a maximum recording amount for pixels not on the edge. (Item 11) the first group of colors being at least one of black, cyan, and magenta; The second group of colors is yellow. 11. The image processing device according to any one of items 9 to 10, (Item 12) 12. The image processing device according to any one of items 9 to 11, characterized in that when the detection means detects two or more lines of second edge pixels adjacent to each other, the detection means excludes at least one line of second edge pixels from the detection result. (Item 13) the generating means generates recording data for a plurality of positions corresponding to one pixel of the input image; The generating means generates the recording data for the colors of the first group such that, among the plurality of positions corresponding to the first edge pixels, recording is performed more frequently at positions adjacent to the second edge pixels than at positions not adjacent to the second edge pixels. 13. The image processing device according to any one of items 9 to 12, (Item 14) the recording device performs multi-pass printing to form an image on the recording medium by performing a plurality of recording scans; 14. The image processing device according to any one of items 9 to 13, characterized in that the generation means generates the recording data such that recording timing of the first edge pixels by the recording material of the first group of colors is biased to a portion of the multiple recording scans, and recording timing of the second edge pixels by the recording material of the second group of colors is biased to another portion of the multiple recording scans. (Item 15) a print scan in which the print rate of the first edge pixel by the print material of the first group of colors is maximized is different from a print scan in which the print rate of the second edge pixel by the print material of the second group of colors is maximized, a recording rate of the non-edge pixels by the recording material of the first group of colors in a recording scan in which the recording rate of the non-edge pixels by the recording material of the first group of colors is maximized is smaller than the recording rate of the first edge pixels by the recording material of the first group of colors in a recording scan in which the recording rate of the first edge pixels by the recording material of the first group of colors is maximized, The recording rate of the non-edge pixels by the recording material of the second group of colors in a recording scan in which the recording rate of the non-edge pixels by the recording material of the second group of colors is maximized is smaller than the recording rate of the second edge pixels by the recording material of the second group of colors in a recording scan in which the recording rate of the second edge pixels by the recording material of the second group of colors is maximized. Item 15. An image processing device according to item 14. (Item 16) the first group of colors is black; The second group of colors is at least one of cyan, magenta, and yellow. 16. The image processing device according to any one of items 14 to 15, (Item 17) In the grayscale image, a pixel value of the first edge pixel is smaller than a pixel value of the second edge pixel, 17. The image processing device according to any one of items 9 to 16, wherein the optical density of all recording agents for the colors of the first group is higher than the optical density of any recording agent for the colors of the second group. (Item 18) An image processing device that generates recording data for use in a recording device that records an image on a recording medium by applying at least one color of recording material to the recording medium in accordance with recording data corresponding to each color, the image processing device comprising: a detection means for detecting edges in a grayscale image corresponding to an input image; a generating means for generating the recording data based on the input image, the edge detection result, and pixel values ​​at the edges of the grayscale image; An image processing device comprising: (Item 19) A detection means for detecting edges in an N-value image (N is a natural number equal to or greater than 2) that indicates the result of threshold processing on a grayscale image obtained from an input image; a generating means for generating color separation data indicating a recording amount for each pixel and the edge detection result for each pixel based on the input image and the edge detection result, and corresponding to a recording material used by a recording device for recording; An image processing device comprising: (Item 20) a first recording mode in which a threshold in the threshold processing is a first value, and a second recording mode in which the threshold is a second value different from the first value; 20. The image processing device according to any one of items 1 to 19, comprising: (Item 21) In the first recording mode, a width of an edge region included in the edge detected by the detection means is a first width, and in the second recording mode, a width of the edge region is a second width different from the first width. 21. The image processing device according to item 20, (Item 22) 22. The image processing device according to item 20 or 21, further comprising a third recording mode in which the width of the edge region is the second width and the threshold value is the first value. (Item 23) The second recording mode is a recording mode for recording a code image clearly, and the first value is on the lighter side of the grayscale image than the second value. 23. The image processing device according to any one of items 20 to 22, (Item 24) The third recording mode is a recording mode for recording a code image clearly, and the first value is on the lighter side of the grayscale image than the second value. 23. An image processing device according to item 22. (Item 25) 1. An image processing method for use in a recording device that records an image on a recording medium by depositing one or more colors of recording materials on the recording medium in accordance with one or more colors of recording data, the method being performed by an image processing device that generates the recording data, the method comprising: detecting edges in an N-valued image (N is a natural number equal to or greater than 2) that indicates the result of threshold processing on a grayscale image obtained from an input image; generating the recording data based on the input image and the edge detection result; 13. An image processing method comprising: (Item 26) 25. A program for causing a computer to function as the image processing device according to any one of items 1 to 24.

[0240] The invention is not limited to the above-described embodiments, and various modifications and variations are possible without departing from the spirit and scope of the invention. Accordingly, the following claims are appended to apprise the public of the scope of the invention. [Explanation of symbols]

[0241] 100: image processing device, 208: image processing section, 209: decoder section, 210: image analysis section, 211: color separation and quantization section, 212: nozzle separation processing section

Claims

1. 1. An image processing device for generating recording data for use in a recording device that records an image on a recording medium by depositing one or more colors of recording material on the recording medium in accordance with one or more colors of recording data, comprising: a detection means for detecting edges in an N-value image (N is a natural number equal to or greater than 2) that indicates the result of threshold processing on a grayscale image obtained from an input image; a generating means for generating the recording data based on the input image and the edge detection result; An image processing device comprising:

2. The image processing apparatus according to claim 1 , wherein the grayscale image is a luminance image or a brightness image corresponding to the input image.

3. the grayscale image indicates, for each pixel, a luminance or brightness on the recording medium when an image is recorded on the recording medium in accordance with each pixel value of the input image; 2. The image processing apparatus according to claim 1, wherein the grayscale image is obtained by converting pixel values ​​of the input image in accordance with a conversion table corresponding to the type of the recording medium.

4. 2. An image processing apparatus according to claim 1, wherein the greyscale image indicates for each pixel an amount of at least one recording material to be used in recording an image on the recording medium in accordance with each pixel value of the input image.

5. 5. The image processing apparatus according to claim 4, wherein the grayscale image is obtained by converting pixel values ​​of the input image in accordance with a conversion table corresponding to a printing mode of the printing apparatus.

6. 2. The image processing device according to claim 1, wherein the generating means generates the recording data corresponding to the pixel of interest of the input image for at least one color from the value of the pixel of interest of the input image in a manner depending on whether the pixel of interest of the input image is on the edge or not.

7. 2. The image processing apparatus according to claim 1, wherein said generating means generates said recording data so that a maximum recording amount for pixels on said edge is less than a maximum recording amount for pixels not on said edge.

8. 2. The image processing apparatus according to claim 1, wherein said generating means generates said recording data so that a maximum recording amount for pixels on said edge is greater than a maximum recording amount for pixels not on said edge.

9. the recording device records an image on the recording medium by depositing a plurality of color recording materials onto the recording medium; the plurality of colors are classified into two or more groups including a first group and a second group; the detection means detects a first edge pixel and a second edge pixel corresponding to both sides of the edge, The generating means generates the recording data for the colors of the first group based on a detection result of the first edge pixels, and generates the recording data for the colors of the second group based on a detection result of the second edge pixels.

2. The image processing device according to claim 1 .

10. 10. The image processing device according to claim 9, wherein the generation means generates the recording data for the colors of the first group so that a maximum recording amount for the first edge pixels is less than a maximum recording amount for pixels not on the edge, and generates the recording data for the colors of the second group so that a maximum recording amount for the second edge pixels is less than a maximum recording amount for pixels not on the edge.

11. the first group of colors being at least one of black, cyan, and magenta; The second group of colors is yellow. The image processing device according to claim 10 .

12. 10. The image processing apparatus according to claim 9, wherein the detection means, when detecting two or more lines of second edge pixels adjacent to each other, excludes at least one line of second edge pixels from the detection result.

13. the generating means generates recording data for a plurality of positions corresponding to one pixel of the input image; The generating means generates the recording data for the colors of the first group such that, among the plurality of positions corresponding to the first edge pixels, recording is performed more frequently at positions adjacent to the second edge pixels than at positions not adjacent to the second edge pixels.

10. The image processing device according to claim 9.

14. the recording device performs multi-pass printing to form an image on the recording medium by performing a plurality of recording scans; 10. The image processing device according to claim 9, wherein the generation means generates the printing data such that printing timing of the first edge pixels by the printing material of the first group of colors is biased to a portion of the multiple printing scans, and printing timing of the second edge pixels by the printing material of the second group of colors is biased to another portion of the multiple printing scans.

15. a print scan in which the print rate of the first edge pixel by the print material of the first group of colors is maximized is different from a print scan in which the print rate of the second edge pixel by the print material of the second group of colors is maximized, a recording rate of the non-edge pixels by the recording material of the first group of colors in a recording scan in which the recording rate of the non-edge pixels by the recording material of the first group of colors is maximized is smaller than the recording rate of the first edge pixels by the recording material of the first group of colors in a recording scan in which the recording rate of the first edge pixels by the recording material of the first group of colors is maximized, The recording rate of the non-edge pixels by the recording material of the second group of colors in a recording scan in which the recording rate of the non-edge pixels by the recording material of the second group of colors is maximized is smaller than the recording rate of the second edge pixels by the recording material of the second group of colors in a recording scan in which the recording rate of the second edge pixels by the recording material of the second group of colors is maximized. The image processing device according to claim 14 .

16. the first group of colors is black; The second group of colors is at least one of cyan, magenta, and yellow. The image processing device according to claim 14 .

17. In the grayscale image, a pixel value of the first edge pixel is smaller than a pixel value of the second edge pixel, 10. An image processing apparatus according to claim 9, wherein the optical density of all recording materials for the colors of said first group is higher than the optical density of any recording material for the colors of said second group.

18. An image processing device for generating recording data for use in a recording device that records an image on a recording medium by depositing at least one color of recording material on the recording medium in accordance with recording data corresponding to each color, the image processing device comprising: a detection means for detecting edges in a grayscale image corresponding to an input image; a generating means for generating the recording data based on the input image, the edge detection result, and pixel values ​​at the edges of the grayscale image; An image processing device comprising:

19. a detection means for detecting edges in an N-value image (N is a natural number equal to or greater than 2) that indicates the result of threshold processing on a grayscale image obtained from an input image; a generating means for generating color separation data indicating a recording amount for each pixel and the edge detection result for each pixel based on the input image and the edge detection result, and corresponding to a recording material used by a recording device for recording; An image processing device comprising:

20. a first print mode in which a threshold in the threshold processing is a first value, and a second print mode in which the threshold is a second value different from the first value; The image processing device according to claim 1 , further comprising:

21. In the first recording mode, a width of an edge area included in the edge detected by the detection means is a first width, and in the second recording mode, a width of the edge area is a second width different from the first width.

21. The image processing device according to claim 20.

22. 22. The image processing apparatus according to claim 21, further comprising a third recording mode in which the width of the edge region is the second width and the threshold value is the first value.

23. The second recording mode is a recording mode for recording a code image clearly, and the first value is on the lighter side of the grayscale image than the second value.

21. The image processing device according to claim 20.

24. The third recording mode is a recording mode for recording a code image clearly, and the first value is on the lighter side of the grayscale image than the second value.

23. The image processing device according to claim 22.

25. 1. An image processing method for a recording device that records an image on a recording medium by depositing one or more colors of recording materials on the recording medium in accordance with one or more colors of recording data, the image processing method being performed by an image processing device that generates the recording data, the image processing device comprising: detecting edges in an N-valued image (N being a natural number of 2 or more) that indicates the result of threshold processing on a grayscale image obtained from an input image; generating the recording data based on the input image and the edge detection result; 13. An image processing method comprising:

26. A program for causing a computer to function as the image processing device according to any one of claims 1 to 24.

Images