JP2011223082A - Image processing apparatus and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing apparatus and a program that smoothly perform a gradation-value transformation process adapted for an pixel edge status.SOLUTION: Pixels of input image data are binarized by storing, in storage means, a blue noise mask BL having an arrangement of random threshold values having no regularity and a Bayer mask BY having an arrangement of threshold values having such regularity that differences at predetermined intervals serve as predetermined gradation values, and using threshold values of the blue noise mask BL for non-edge pixels among the pixels of the input image data, while using threshold values of the Bayer mask BY for edge pixels among the pixels of the input image data. This can prevent a binarization-caused pattern from appearing in portions other than the edge portions, and can prevent the degree of discreteness of a binarization result from being deviated in the edge portions. One set of threshold values in either mask is switched to another: there is no difference in technique for binarization between both masks. Smooth binarization can be performed since masks are stored in the storage means.

Description

The present invention relates to an image processing apparatus and a program that input multi-tone image data and perform binarization processing.

Conventionally, as this type of image processing apparatus, an apparatus that inputs multi-gradation image data and converts it into a predetermined gradation value has been proposed (for example, see Patent Document 1). In this apparatus, it is determined whether each pixel constituting the input image data belongs to an edge region or a non-edge region, and an error diffusion method is applied to the edge region and blue is applied to the non-edge region. By converting the tone value by applying the noise mask method, conversion results suitable for each area can be realized by preventing the conversion result that dots are scattered unnaturally in the edge area around the character. It is said.

JP 2008-227759 A

Here, in the error diffusion method, since it is necessary to reflect the error generated when converting the gradation value by dispersing it in surrounding pixels, the threshold value is simply compared with the gradation value of the pixel to be processed to obtain the gradation value. Compared with the blue noise mask method to be converted, processing time is required. For this reason, depending on the degree of the edge area included in the image, the area using the error diffusion method increases, and the time required for the gradation value conversion process becomes longer. In addition, by switching between two processes having different processing methods such as the error diffusion method and the blue noise mask method, the configuration of the apparatus may be complicated.

The main object of the image processing apparatus and program according to the present invention is to more smoothly perform gradation value conversion processing suitable for the edge state of a pixel.

The image processing apparatus and program of the present invention employ the following means in order to achieve the main object described above.

The image processing apparatus of the present invention
An image processing apparatus that inputs multi-gradation image data and performs binarization processing,
A first mask in which random thresholds having no regularity are arranged in a matrix and a second mask in which thresholds having regularity are arranged in a matrix so that a difference at a predetermined interval becomes a predetermined gradation value. Storage means for storing a mask;
Edge pixel determination means for determining whether each pixel of the input image data is an edge pixel corresponding to an edge portion;
Binarization is performed using the threshold value of the first mask for pixels that are not determined as edge pixels, and binarization is performed using the threshold value of the second mask for pixels that are determined as edge pixels. And a means.

In the image processing apparatus according to the present invention, a first mask in which random thresholds having no regularity are arranged in a matrix and a threshold having regularity so that a difference at a predetermined interval becomes a predetermined gradation value. A second mask arranged in a matrix is stored, and pixels that are not determined as edge pixels among the pixels of the input image data are binarized using the threshold value of the first mask, The determined pixel is binarized using the threshold value of the second mask. As a result, regularity does not appear in the binarization result in the part other than the edge part, and it is possible to prevent the pattern resulting from the binarization from appearing and in the edge part, the binarization result has a predetermined interval. Since the regularity of each appears, it is possible to prevent a deviation from occurring in the degree of discreteness. Further, both the first mask and the second mask are stored in the storage means,
Since it is a binarization method that is simply compared with the threshold value, compared with the case of using it together with the error diffusion method,
It can be binarized smoothly. As a result, the gradation value conversion process suitable for the edge state of the pixel can be performed more smoothly.

In such an image processing apparatus of the present invention, the first mask is a mask having a larger matrix size than the second mask, and the binarizing means is configured such that the input image data is the first or the second mask. When the size of the matrix is larger than that of the second mask, the matrix may be a means for binarization by repeatedly using the first or second mask. By doing so, it is possible to suppress the appearance of a block-like pattern based on the matrix size in a portion other than the edge portion. In the image processing apparatus according to the aspect of the present invention, the second mask is a mask in which the gradation values of the multi-gradation are arranged once, and the first mask is the second mask. It is also possible to use a mask in which each gradation value of the multi-gradation is arranged in a matrix having a size that is an integral multiple of the mask. Here, the first mask may be a blue noise mask, and the second mask may be a Bayer-type dither mask.

Further, in the image processing apparatus of the present invention in which the image data is input via an image reading apparatus that reads an image formed on a medium, the edge pixel determination unit is a determination target of whether or not it is the edge pixel. It may be a means for performing the determination based on a gradation value change between a predetermined number of pixels adjacent to the pixel. Here, in the image data input through such an image reading device, the change in the gradation value of the pixel at the edge portion is more gradual than the original steep change depending on the reading characteristics of the device. There is. If the threshold value of the first mask is used for the edge pixel of such a slowly changing portion, the binarization result becomes unnaturally dispersed or aggregated. Since it can prevent such a tendency from appearing, the significance of applying the present invention becomes high. In this aspect of the image processing apparatus of the present invention, the edge pixel determination means is a means for performing the determination when a gradation value of a pixel to be determined as to whether or not it is the edge pixel indicates black or gray. It can also be. Edge portions such as black and gray are clearer than other colors and are easily noticeable, so that the significance of applying the present invention is higher.

The program of the present invention is for causing a computer to function as any of the above-described image processing apparatuses of the present invention. This program may be recorded on a computer-readable recording medium (for example, hard disk, ROM, FD, CD, DVD, etc.) or from a computer via a transmission medium (communication network such as the Internet or LAN). It may be distributed to another computer, or may be exchanged in any other form. If this program is executed by a computer, the same effects as those of the image processing apparatus described above can be obtained.

1 is a configuration diagram showing an outline of the configuration of a multifunction printer 10. FIG. 7 is a flowchart illustrating an example of a copy mode processing routine. The flowchart which shows an example of an edge determination process. Explanatory drawing which shows an example of the change of the gradation value D of a x direction. The flowchart which shows an example of a binarization process. Explanatory drawing which shows an example of the blue noise mask BL. Explanatory drawing which shows an example of Bayer type mask BY. Explanatory drawing which shows a mode that image data is binarized. Explanatory drawing which shows the mode of the difference of the binarization result by the difference of a mask.

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram showing an outline of a configuration of a multi-function printer 10 which is an embodiment of an image processing apparatus of the present invention. As shown in the figure, the multifunction printer 10 of the present embodiment includes a printer unit 20 that executes printing on a sheet S based on image data, a scanner unit 30 that reads a document placed on a glass table 36, and a memory card. A memory card controller 44 for inputting / outputting a file storing data to / from the memory card 42 inserted in the slot 40, a button group 54 for displaying various types of information on the display unit 52 and user instructions.
And an operation panel 50 for inputting via the above operation, and a main controller 60 for controlling the entire apparatus. The multifunction printer 10 is configured such that the printer unit 20, the scanner unit 30, the memory card controller 44, and the main controller 60 can exchange various control signals and data with each other via the bus 12.

The printer unit 20 includes a printer ASIC 22 and a printer mechanism 24. The printer ASIC 22 is an integrated circuit that controls the printer mechanism 24. Upon receiving a print command from the main controller 60, the printer ASIC 22 controls the printer mechanism 24 to print an image on the paper S based on the image data that is the target of the print command. To do. The printer mechanism 24 is a well-known ink jet color that performs printing by ejecting C (cyan), magenta (M), yellow (Y), and black (B) ink from the print head to the paper S. It is configured as a printer mechanism.

The scanner unit 30 includes a scanner ASIC 32 and a scanner mechanism 34. The scanner ASIC 32 is an integrated circuit that controls the scanner mechanism 34. When the scanner ASIC 32 receives a reading command from the main controller 60, the scanner ASIC 32 controls the scanner mechanism 34 to read a document placed on the glass table 36 as image data. The scanner mechanism 34 is configured as a well-known image scanner, and the reflected light after emitting light toward the document is red (
A well-known color image sensor is provided which is separated into each color of R), green (G), and blue (B) and used as image data. The image data (RGB data) has pixels arranged in a vertical and horizontal matrix, and each value of the arranged pixels is represented by 256 gradations (8 bits) of values 0 to 255 according to the shade. To do.

The memory card controller 44 inputs / outputs a file storing data with the memory card 42 inserted into the memory card slot 40. The memory card controller 44 has a memory card 42 in the memory card slot 40.
Is connected, the file stored in the memory card 42 is read and transmitted to the main controller 60, or a command is input from the main controller 60 and the file is stored in the memory card 42 based on the command. .

The main controller 60 is configured as a microprocessor centered on the CPU 62, and is a ROM 6 storing various processing programs, various data, various tables, and the like.
4, a RAM 66 that temporarily stores scan data, print data, and the like, a flash memory 68 that can retain data even when the power is turned off, and an internal communication interface (I / F) that enables communication with the operation panel 50. 69). Note that the ROM 64 stores a blue noise mask BL and a Bayer mask BY used in the binarization processing by the dither method, and details thereof will be described later. The main controller 60 inputs various operation signals and various detection signals from the printer unit 20, the scanner unit 30, and the memory card controller 44, and inputs operation signals generated according to the operation of the button group 54 of the operation panel 50. To do. Also, a read command for reading a file from the memory card 42 and outputting it to the main controller 60 is sent to the memory card controller 4.
4, a print command is output to the printer unit 20 to execute printing of image data, and a document placed on the glass table 36 is converted into image data based on a scan instruction of the button group 54 of the operation panel 50. As a result, a reading command is output to the scanner unit 30 or a control command for the display unit 52 is output to the operation panel 50. The operation panel 5
Various modes can be selected by operating the button group 54 of 0. As selectable modes, a copy mode for scanning and copying a document placed on the glass table 36 and a memory card 42 are stored. There is a memory card mode in which an image is printed using image data, a document is scanned, converted into data, and stored in the memory card 42.

Next, the operation of the multifunction printer 10 of the present embodiment configured as described above,
In particular, the operation when the copy mode is selected will be described. FIG. 2 is a flowchart illustrating an example of a copy mode processing routine executed by the main controller 60. This routine is executed when the copy mode is selected by the user operating the button group 54 and a copy start instruction is given.

When the processing routine in the copy mode is executed, the CPU 62 of the main controller 60
First, by outputting a reading command to the scanner unit 30, the original placed on the glass table 36 is read (scanned) to generate RGB color system image data (step S100). Next, edge determination processing is performed for determining whether each pixel of the generated image data is an edge pixel corresponding to the edge portion of the image or a non-edge pixel not corresponding to the edge portion (step S110). A color conversion process is performed on the 8-bit RGB color system image data into 8-bit CMYK color system image data (step S120). continue,
The blue noise mask BL and Bayer mask BY stored in the ROM 64 are stored in the RAM 6.
6 and using the threshold values of these masks, binarization processing is performed by a dither method in which CMYK 8-bit image data after color conversion processing is converted into 2-bit binary data, respectively (step S130). ). When the binarization process is performed, the image data after the binarization process is developed in the order in which the print head of the printer mechanism 24 forms dots to generate image data that is the target of the print command (step S140), and the printer By outputting a print command to the unit 20, a print process based on the image data after the expansion process is performed (step S150), and this routine is terminated. Hereinafter, details of the edge determination processing in step S110 and the binarization processing in step S130 will be described in order.

First, the edge determination process will be described. FIG. 3 is a flowchart illustrating an example of the edge determination process. In this edge determination process, first, the index value y is set to a value 1 (step S200), the index value x is set to a value 1 (step S210), and pixels corresponding to the set index value xy are processed pixels. (Step S220
). Here, the index value xy is used as a value indicating the position coordinates of the pixels arranged in a matrix in the image data, and the index value x, y of the upper left pixel in the image data is set to 1 respectively. The right direction is the x direction, the lower direction is the y direction, the rightmost pixel is defined as the index value xmax, and the lowermost pixel is defined as the index value ymax. Next, it is determined whether or not R = G = B except for the gradation values indicating white as the RGB gradation values of the set processing target pixel (step S230). This process is a process for determining whether or not the color of the pixel to be processed indicates black or gray. Although the details of the reason for performing such determination will be described later, different processing is performed during binarization processing on edge pixels and non-edge pixels determined in the edge determination processing. This is because the effect of switching the processing for the edge pixel becomes significant.

When it is determined in step S230 that R = G = B, the difference Δx between the gradation values D of two pixels adjacent in the x direction of the processing target pixel and the processing target based on the following expressions (1) and (2) The difference Δy between the gradation values D of two pixels adjacent to each other in the y direction of the pixel is calculated (step S2).
40). Since the case of R = G = B is considered now, the gradation value D may be any gradation value of R, G, B. Further, when the pixel to be processed corresponds to the pixel at the end and there is no adjacent pixel, the gradation value D is calculated as a value 0. Next, an edge gradient a is calculated based on the following equation (3) using the calculated differences Δx and Δy (step S250), and it is determined whether or not the calculated edge gradient a exceeds a predetermined threshold value aref. (Step S260)
. When it is determined that the edge gradient a exceeds the predetermined threshold value aref, it is determined that the processing target pixel is an edge pixel, and the edge determination result is associated with the index value xy and registered in the RAM 66 (step S270). On the other hand, in step S260, the edge gradient a is a predetermined threshold value are.
When it is determined that f is not exceeded, or when it is determined that R = G = B is not satisfied in step S230 described above, it is determined that the processing target pixel is a non-edge pixel, and the edge determination result is associated with the index value xy. Registration is made in the RAM 66 (step S280). For example, the predetermined threshold value aref is set to a value such that the processing target pixel is determined to be an edge pixel when any of the differences Δx and Δy exceeds 200 for all 256 gradation values. It was supposed to be. Here, an example of the change of the gradation value D in the x direction is shown in FIG. FIG. 4 (a)
Indicates a case where the change in the gradation value D is not so large and the difference Δx does not exceed the value 200, and the processing target pixel is determined as a non-edge pixel. On the other hand, FIGS. 4B and 4C show a case where the gradation value D changes greatly from the value 0 to the value 255, and the difference Δx changes beyond the value 200. It is determined as an edge pixel. In this way, it is determined whether or not the processing target pixel is an edge pixel by using the difference between the gradation values D of two pixels adjacent to the processing target pixel. In the RGB data generated by the scan processing as in the present embodiment, the pixels at the edge portion in the image to be scanned are originally set to the gradation value 0 and the gradation value due to the reading characteristics of the scanner unit 30 and the like. Even when 255 is adjacent, a gentle edge may occur. Specifically, the edge does not suddenly rise between the pixels as shown in FIG. 4B (in this case, from the pixel adjacent to the left of the pixel to be processed to the pixel to be processed). As shown in (c), it rises relatively gently between a plurality of two or more pixels (here, two pixels from the pixel adjacent to the left of the processing target pixel to the pixel adjacent to the right of the processing target image). Become an edge.

Δx = D (x-1, y) -D (x + 1, y) (1)
Δy = D (x, y-1) -D (x, Y + 1) (2)
a = √ (Δx ^ 2 + Δy ^ 2) (3)

In this way, determining whether it is an edge pixel or a non-edge pixel and registering the result,
The index value x is incremented by 1 (step S290), and it is determined whether or not the incremented index value x exceeds the maximum value xmax (step S300).
. When it is determined that the maximum value xmax is not exceeded, the process returns to step S220 and is repeated. On the other hand, when it is determined in step S310 that the maximum value xmax is exceeded, the index value y is incremented by 1 (step S310), and it is determined whether or not the incremented index value y exceeds the maximum value ymax (step S320). ). Maximum value y
When it is determined that the value does not exceed max, the process returns to step S210 and the process is repeated. When it is determined that the value exceeds the maximum value ymax, it is determined that the edge determination for all the pixels is completed,
This process ends.

Next, the binarization process will be described. FIG. 5 is a flowchart illustrating an example of the binarization process. In this binarization process, first, the index value y is set to a value 1 (step S400), the index value x is set to a value 1 (step S410), and a pixel corresponding to the set index value xy is processed. A pixel is set (step S420). This index value xy is used as a value indicating the position coordinates of the pixel in the image data of the CMYK color system after the color conversion process, and is set similarly to the index value xy in the edge determination process of FIG. Next, it is determined from the edge determination result registered in the RAM 66 in the edge determination process of FIG. 3 whether the processing target pixel is an edge pixel (step S430). When it is determined that the processing target pixel is not a non-edge pixel, that is, an edge pixel, a threshold value at a position corresponding to the index value xy of the processing target pixel is extracted from the blue noise mask BL (step S440), and the extracted threshold value and the processing target pixel are extracted. The gray level value is compared with a binary value and binarized (step S460). An example of the blue noise mask BL is shown in FIG. As shown,
A threshold value for 128 pixels is arranged in the x direction, and a threshold value for 64 pixels is arranged in the y direction. In this blue noise mask BL, each gradation value of all 256 gradations from 0 to 255 appears at random 32 times, and is configured as a mask having so-called blue noise characteristics. As the threshold value of the blue noise mask BL extracted in step S440, the index value x of the processing target pixel in the x direction is divided by a value 128 that is the number of pixels in the x direction (threshold number) of the blue noise mask BL. The threshold value corresponding to the position of the remainder numerical value is extracted in the y direction by dividing the index value y by the value 64 which is the number of pixels (threshold number) in the y direction of the blue noise mask BL. Will be.

On the other hand, when it is determined in step S430 that the processing target pixel is an edge pixel, a threshold value at a position corresponding to the index value xy of the processing target pixel is extracted from the Bayer mask BY (step S450), and the threshold value extracted in step S460 is extracted. And the gradation value of the pixel to be processed are compared and binarized. An example of this Bayer type mask BY is shown in FIG. As shown in the figure, thresholds for 16 pixels are arranged in the x and y directions, respectively. Therefore, all 256
Each gradation value of the gradation appears once each. In the Bayer mask, in the x direction, for example, the uppermost threshold value has values 0, 128, 32, 160, 8, 13
6 and so on, and the difference between the value 0 and the value 128 is a value 12 that is the median value of all 256 gradations.
The threshold is arranged so that the difference between the value 32 and the value 160 and the difference between the value 8 and the value 136 are also the value 128. Similarly, the threshold values other than the uppermost level are also set to a difference of 12 at every predetermined interval.
The threshold value is arranged to be 8. On the other hand, in the y direction, for example, the threshold values of the leftmost column are values 0, 192, 48, 240,..., And the threshold values are arranged so that the difference becomes the value 192 at every predetermined interval. The threshold values in the second column from the left end are values 128, 64, and 176.
, 112,..., And the thresholds are arranged so that the difference becomes 64 at every predetermined interval. Such a column in which the difference is a value 192 and a column in which the difference is a value 64 are alternately arranged. The average value of the value 192 and the value 64 is a value 128 that is the median value of all 256 gradation values. As described above, in the Bayer mask BY, the thresholds are arranged with regularity so that the difference at each predetermined interval becomes a predetermined value (value 128, value 196, value 64). Step S
As the threshold value of the Bayer mask BY extracted at 440, the index value x and y of the pixel to be processed in the x and y directions is the value 16 that is the number of pixels (threshold number) in the x and y directions of the Bayer mask BY, respectively. The threshold value corresponding to the position of the remaining numerical value divided by is extracted.

When binarization is performed in this way, the index value x is incremented by 1 (step S470), and it is determined whether or not the incremented index value x exceeds the maximum value xmax (step S480). If it is determined that the maximum value xmax is not exceeded, the process returns to step S420 and the process is repeated. On the other hand, if it is determined in step S480 that the maximum value xmax is exceeded, the index value y is incremented by 1 (step S490).
), It is determined whether or not the incremented index value y exceeds the maximum value ymax (
Step S500). If it is determined that the maximum value ymax is not exceeded, step S410 is performed.
The process is repeated, and when it is determined that the maximum value ymax is exceeded, it is determined that the edge determination of all the pixels has been completed, and this process ends.

Here, how the image data is binarized will be described. FIG. 8 shows that the image data is 2
It is explanatory drawing which shows a mode that it is digitized. In FIG. 8, only a part of the image data, the blue noise mask BL, and the Bayer mask BY is shown in an enlarged manner. As shown in the figure, for the non-edge pixels other than the pixels determined to be edge pixels (pixels within the bold line frame in the figure), a threshold value corresponding to the pixel position (index value xy) is used from the blue noise mask BL. (Indicated by dotted arrows in the figure). On the other hand, for pixels determined to be edge pixels, a threshold value corresponding to the pixel position from the Bayer mask BY is used (indicated by a solid arrow in the figure).
Here, as described above, in the blue noise mask BL, random thresholds having no regularity are arranged in a large matrix as compared with the Bayer type mask BY, while the Bayer type mask BY is the blue noise mask BL. A threshold having regularity is arranged in a matrix having a smaller size compared to the above so that a difference at a predetermined interval becomes a predetermined value. For this reason, for example, when the Bayer mask BY is used for the entire image, the block pattern due to the small matrix size becomes conspicuous and a pseudo contour is generated in the binarized image. . On the other hand, when a blue noise mask is used for the entire image, it is possible to suppress the occurrence of a block pattern due to the large matrix size. However, even in a portion where edge pixels having substantially the same gradation value, such as a peripheral portion of a character or a line drawing, are continuous, the binarization result is unnaturally dispersed due to a random threshold having no regularity. The part which gathers or gathers arises. Therefore, for the edge pixel, the threshold value of the Bayer mask BY is used to obtain a binarized result having regularity. Here, FIG. 9 is an explanatory diagram showing the difference in the binarization result due to the difference in mask. FIG. 9A shows the difference with respect to the pixel whose gradation value D is the value 190, and FIG. 9B shows the difference with respect to the pixel whose gradation value D is the value 120. Note that, as a binarization result, pixels whose threshold value is smaller than the gradation value D are illustrated in a dark color, and pixels whose threshold value is larger than the gradation value D are illustrated in white. As shown in the figure, it can be seen that in either case, the application of the Bayer mask is more regular than the application of the blue noise mask BL. Note that the binarization processing in the present embodiment switches which mask threshold value is used between the blue noise mask BL and the Bayer mask BY, and the binarization method itself is not different. There is no complicated process. These blue noise mask BL and Bayer mask BY are both RO.
Since it is stored in M64 and developed in the RAM 66 during binarization processing, a threshold value is extracted.
Compared to the case of binarizing while diffusing the binarization error to surrounding pixels, such as the error diffusion method, it can be processed smoothly. In the edge determination process of FIG. 3 described above,
The reason for determining whether or not the pixel is an edge pixel for black or gray pixels is that the edge portion of black or gray is clearer than other colors, making it easier to see. This is to prevent the formation of unnatural dots in such a portion.

Here, the correspondence between the components of the present embodiment and the components of the present invention will be clarified. ROM 64 and RA for storing the blue noise mask BL and the Bayer mask BY of the present embodiment
M66 corresponds to the “storage means” of the present invention, and the main controller 60 that executes the processing of step S110 (edge determination processing of FIG. 3) in the copy mode processing routine of FIG.
Corresponds to “edge pixel determination means”, and the main controller 60 that executes the processing of step S130 (binarization processing of FIG. 5) in the copy mode time processing routine of FIG. 2 corresponds to “binarization means”.

According to the multi-function printer 10 of the present embodiment described in detail above, the blue noise mask BL in which random thresholds having no regularity are arranged and regularity so that the difference at every predetermined interval becomes a predetermined gradation value. A Bayer mask BY with a predetermined threshold value stored in advance, and it is determined whether the pixel of the input image data is an edge pixel corresponding to the edge portion or a non-edge pixel not corresponding to the edge portion. Since each pixel is binarized using the threshold value of the blue noise mask BL for the non-edge pixel and the threshold value of the Bayer mask BY for the edge pixel, the binarization result is obtained in the portion other than the edge portion. It is possible to prevent a pattern resulting from binarization from appearing without regularity, and at the edge portion, binarization results are displayed at predetermined intervals. It is possible to prevent the deviation in the discrete degree occurs regularity appears. In addition, both the blue noise mask BL and the Bayer mask BY are stored in advance in the ROM 64, and compared with a case where the error diffusion method is used together,
It can be binarized smoothly. As a result, the gradation value conversion process suitable for the edge state of the pixel can be performed more smoothly.

In addition, this invention is not limited to the embodiment mentioned above at all, and it cannot be overemphasized that it can implement with a various aspect, as long as it belongs to the technical scope of this invention.

In the above-described embodiment, the threshold value of the blue noise mask BL is used for non-edge pixels and the threshold value of the Bayer mask BY is used for edge pixels. However, the present invention is limited to the blue noise mask BL and the Bayer mask BY. Not a non-edge pixel, a threshold value of a mask in which a random threshold value having no regularity is arranged is used, and a threshold value having a regularity is arranged for an edge pixel so that a difference at a predetermined interval becomes a predetermined value. Any mask may be used as long as the mask threshold is used. The blue noise mask BL has a larger matrix size than the Bayer mask BY, but the present invention is not limited to this, and a mask having the same size may be used. However, in order to prevent the generation of block patterns, it is preferable that the blue noise mask BL is a relatively large matrix as in the present embodiment. Furthermore, the matrix size of the blue noise mask BL and the Bayer mask BY is an example, and any size may be used as long as it is a size based on a multiple of all gradation values. In particular, as the blue noise mask BL, when all gradation values are 255, the size is 128 pixels × 128 pixels or 256 pixels ×
The size may be 256 pixels.

In the above-described embodiment, it is determined whether or not a pixel in which R = G = B, that is, a pixel indicating black or gray is an edge pixel except for a gradation value indicating a white pixel. However, the present invention is not limited to this, and it may be determined whether all pixels are edge pixels regardless of the color of the pixels.

In the above-described embodiment, it is determined whether or not the pixel is an edge pixel based on the edge gradient a calculated from the difference between the gradation values D of the pixels adjacent to the processing target pixel. It is good also as what determines based on a difference. Or, it is not limited to the one based on the difference of the gradation value D between adjacent pixels, but other first-order differential filter such as Sobel filter or Laplacian
Based on a calculation result obtained by multiplying the gradation value D of pixels in the vicinity of the upper, lower, left, and right 4 centering around the pixel to be processed by using a second-order differential filter such as a filter, or the 8 neighboring pixels including the diagonal, by a filter coefficient. It is good also as what performs a determination.

10 Multifunction printer, 12 buses, 20 printer units, 2
2 Printer ASIC, 24 Printer mechanism, 30 Scanner unit, 32 Scanner ASIC, 34 Scanner mechanism, 36 Glass stand, 40 Memory card slot, 42 Memory card, 44 Memory card controller, 50 Operation panel, 52 Display unit, 54 Button group, 60 main controller, 62 CPU, 64
ROM, 66 RAM, 68 Flash memory, 69 Internal communication interface (
I / F), BL blue noise mask, BY Bayer type mask, S paper.

Claims (7)

An image processing apparatus that inputs multi-gradation image data and performs binarization processing,
A first mask in which random thresholds having no regularity are arranged in a matrix and a second mask in which thresholds having regularity are arranged in a matrix so that a difference at a predetermined interval becomes a predetermined gradation value. Storage means for storing a mask;
Edge pixel determination means for determining whether each pixel of the input image data is an edge pixel corresponding to an edge portion;
Binarization is performed using the threshold value of the first mask for pixels that are not determined as edge pixels, and binarization is performed using the threshold value of the second mask for pixels that are determined as edge pixels. An image processing apparatus comprising: means. The image processing apparatus according to claim 1,
The first mask is a mask having a larger matrix size than the second mask;
The binarization means binarizes the input image data when the size of the matrix is larger than the first or second mask by repeatedly using the first or second mask. An image processing apparatus. The image processing apparatus according to claim 2,
The second mask is a mask in which each gradation value of the multi-gradation is arranged once.
The image processing apparatus according to claim 1, wherein the first mask is a mask in which each gradation value of the multi-gradation is arranged in a matrix having a size that is an integral multiple of the second mask. The image processing apparatus according to claim 3,
The first mask is a blue noise mask;
The image processing apparatus, wherein the second mask is a Bayer-type dither mask. The image processing apparatus according to any one of claims 1 to 4, wherein the image data is input via an image reading apparatus that reads an image formed on a medium.
The image processing apparatus according to claim 1, wherein the edge pixel determination unit is a unit that performs the determination based on a gradation value change between a predetermined number of pixels adjacent to a pixel that is a determination target of whether or not the pixel is the edge pixel. The image processing apparatus according to claim 5, wherein the edge pixel determination unit is a unit that performs the determination when a gradation value of a pixel that is a determination target of whether or not the pixel is the edge pixel indicates black or gray. A program for causing a computer to function as the image processing apparatus according to any one of claims 1 to 6.