JP2015095713A - Image processing device, image processing method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an excellent halftone processing result by properly correcting screen processing (systematic dither method) by a simple method.SOLUTION: There is provided an image processing device including: an edge detection part 101 which detects an edge from input image data and outputs determination data indicative of whether a pixel of interest is a pixel of an edge part; a gamma conversion processing part 102; a screen processing part 103; a selection part 104 which selectively outputs one of gamma conversion image data and screen image data according to the determination data; a first in-cell total calculation part 105 which sums up selection data by cells so as to calculate a first in-cell total value; a second in-cell total calculation part 106 which sums up the input image data by cells so as to calculate a second in-cell total value; a subtraction part 107 which outputs difference data; and a normalization part 108 which calculates a correction amount for each pixel from the difference data. A correction amount (absolute value) of the edge part is smaller than a correction amount (absolute value) of a non-edge part.

Description

  The present invention relates to an image processing apparatus, an image processing method, and a program. Specifically, the present invention relates to an image processing apparatus, an image processing method, and a program that appropriately correct a result of screen processing and output a good halftone image.

  In recent years, printing of image data processed by an image processing apparatus such as a computer has been widely performed. However, the number of gradations that can be expressed by an output device such as a printing device or a display device is generally smaller than the number of gradations represented by image data on a computer. For this reason, halftone processing is often performed in which the number of gradations of image data is converted to the number of gradations that can be expressed by an output device.

  As one of the halftone processes, a screen process (organized dither method) that determines an output value by comparing image data with a periodically varying threshold value is known. By performing screen processing, a screen image in which the density of the image is expressed by area gradation can be obtained. In particular, in a flat portion where there is little change in gradation, halftone dots having the same shape are formed at equal intervals, so that a good image can be obtained. However, visually repetitive patterns (moire) may occur in the screen-processed image data, and the image quality may deteriorate. In addition, when an image includes a fine line, the reproducibility of the fine line may deteriorate depending on the density and angle of the fine line.

  Techniques have been proposed for suppressing image deterioration as described above and obtaining good screen processing results. For example, there is a method of detecting a position where moiré is likely to occur in an image to be subjected to halftone processing, and in a region where moiré is likely to occur, by removing high frequency components that cause moiré in advance, thereby suppressing moiré ( Patent Document 1).

  A method of combining screen processing (organized dither method) and error diffusion method has also been proposed. By combining the systematic dither method and the error diffusion method, it is possible to reduce the error bias before and after processing due to the dither processing (Patent Document 2).

JP-A-9-238259 JP 2005-252583 A

  However, although the method described in Patent Document 1 can reduce the occurrence of moire, high-frequency components are lost in some areas. When the high frequency component is removed from the edge portion of medium density, the contrast of the edge portion is lowered and unclear, and a preferable image may not be obtained.

  In the method described in Patent Document 2, it is selected whether to apply dither processing or error diffusion processing to each pixel. However, in order to perform an appropriate selection, complicated processing is required. An object of the present invention is to appropriately correct the screen processing (organized dither method) by a simple method and obtain a good halftone processing result.

  In order to solve the above problems, an image processing apparatus according to the present invention comprises the following arrangement. That is, the invention described in claim 1 of the present invention comprises edge detection means for detecting pixels constituting an edge in input image data and outputting determination data indicating whether or not the pixel of interest is an edge portion pixel. Gamma conversion processing means for gamma conversion processing of the input image data; screen processing means for screen processing of the input image data; either gamma conversion image data or screen image data for each pixel according to the determination data A selection means for outputting data for which the pixel values constituting these are determined as selection data, and a first for calculating the sum total of the selection data for each cell used in the screen processing means as a first in-cell total value. And a second in-cell total for calculating a total sum of the input image data for each cell as a second in-cell total value. A calculating means; a subtracting means for calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data; and a correction amount for each pixel based on the difference data. Normalization means for calculating, wherein the normalization means makes the correction amount (absolute value) in the edge portion smaller than the correction amount (absolute value) in the non-edge portion according to the determination data. To do.

  The dots on the edge are made continuous in a line to stabilize the dots. Thereby, the image quality of a character and a line drawing improves.

1 is a block diagram of an image processing apparatus according to a first embodiment. FIG. 6 is a block diagram of an image processing apparatus according to a second embodiment. It is explanatory drawing of the example of the cell used with a screen process part. 6 is an explanatory diagram of image processing data according to Embodiment 1. FIG. FIG. 10 is a block diagram of an image processing apparatus according to a third embodiment. FIG. 10 is a block diagram of an image processing apparatus according to a fourth embodiment. 6 is an explanatory diagram of image processing data according to Embodiment 2. FIG.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. As image data used in the following embodiments, monochrome scale image data having only a density component is assumed.

  FIG. 1 is a block diagram illustrating the configuration of the image processing apparatus according to the present embodiment. The image processing apparatus shown in FIG. 1 converts input image data into halftone image data having a smaller number of gradations. In this specification, image data input to the image processing apparatus is abbreviated as “input image data”. The image processing apparatus includes an edge detection unit 101, a white pixel detection unit 111, a gamma conversion processing unit 102, a screen processing unit 103, a selection unit 104, an in-cell total calculation unit 105/106, a subtraction unit 107, a normalization unit 108, and an addition. Part 109 and stabilization part 110. Hereinafter, an outline of each component will be described.

  The edge detection unit 101 detects an edge from the input image data, and generates determination data indicating whether the pixel is an edge portion or not for each pixel. The edge detection unit 101 outputs the determination data to the selection unit 104, the normalization unit 108, and the stabilization unit 110.

  The white pixel detection unit 111 detects white pixels in the input image data. A white pixel is a pixel having a pixel value of 0 in the input image data. White pixel determination data indicating a white pixel is generated for each pixel, and the white pixel determination data is output to the in-cell total calculation units 105 and 106 and the normalization unit 108.

  The gamma conversion processing unit 102 performs gamma conversion processing on the input image data. Image data obtained as a result of the γ conversion processing is output to the selection unit 104 as gamma converted image data.

  The screen processing unit 103 performs screen processing (systematic dither processing) on the input image data, and converts the input image data into image data having a smaller number of gradations (referred to as “screen image data” in this specification). . The screen processing unit 103 performs quantization by comparing the input pixel value with the corresponding threshold value for each pixel. In this specification, the value of each pixel obtained as a result of quantization by the screen processing unit 103 is referred to as a “screen pixel value”. The screen image data is composed of screen pixel values for each pixel. That is, the screen processing unit 103 converts each input pixel value constituting the input image data into a screen pixel value constituting the screen image data by screen processing, and outputs the screen image data to the selection unit 104.

  The selection unit 104 selects either a screen pixel value constituting screen image data or a pixel value constituting gamma conversion processed image data for each pixel based on the detection result by the edge detection unit 101. Each of the screen image data and the gamma conversion processed image data corresponds to a pixel position. The selection unit 104 selects, for each corresponding pixel position, data that selects either the screen pixel value or the pixel value constituting the gamma conversion processed image data as selection data in the in-cell total calculation unit 105 and the addition unit 109. Output.

  The in-cell total calculation unit 105 calculates the sum of pixel values representing pixels included in the cell for each cell in the selection data output from the selection unit 104. Similarly, the in-cell total calculation unit 106 calculates the sum of pixel values representing pixels included in the cell for each cell with respect to the input image data. That is, the in-cell total calculation unit 105 and the in-cell total value calculation unit 106 have the same operation except for the data to be processed. The in-cell total calculation unit 105 and the in-cell total value calculation unit 106 output the sum of pixel values for each cell to the subtraction unit 107.

  The subtraction unit 107 calculates a difference value between the input image data and the selection data obtained from the selection unit 104 for each pixel. Each pixel position corresponds to the input image data and the selection data. The subtraction unit 107 calculates a difference between a pixel value constituting the input image data and a pixel value constituting the selection data for a certain pixel position, and determines the difference as a difference value of the pixel position. Then, the subtraction unit 107 outputs the difference data formed by the difference value of each pixel to the normalization unit 108.

  The normalization unit 108 calculates a correction amount for each pixel based on the difference data obtained from the subtraction unit 107, and generates correction data. The normalizing unit 108 outputs the correction data to the adding unit 109. At this time, the normalization unit 108 also refers to the determination data from the edge detection unit 101, and changes the correction amount calculation method for each pixel according to the value.

  The addition unit 109 adds the correction amount of the corresponding pixel in the correction data obtained by the normalization unit 108 to each pixel of the selection data obtained by the selection unit 104, and outputs intermediate data for stabilization. Output to the unit 110.

  The stabilizing unit 110 detects data in which dots are not stable in the intermediate data obtained from the adding unit 109, and replaces the data with data that can be output stably by the subsequent image forming apparatus.

  The above is the outline of the components of the image processing apparatus. Hereinafter, the processing of each unit will be described in detail.

  The edge detection unit 101 detects an edge present in the input image data, and outputs the result as determination data for each pixel. Many methods for detecting edges have already been proposed, and are not particularly defined here. For example, an edge can be detected by the following method. Four differences are calculated by subtracting the pixel values of the upper, lower, left, and right pixels of the target pixel from the pixel value of the target pixel constituting the input image data, and the maximum value of the obtained four differences is equal to or greater than a predetermined threshold value. The edge detection unit 101 determines that the pixel of interest is an edge portion. That is, in this method, it is determined that the high density side of the edge is the edge portion. The edge detection unit 101 outputs “1” as determination data when it is determined that the target pixel in the input image data is an edge portion, and outputs “0” as determination data when it is determined that the target pixel is not an edge portion.

  The gamma conversion processing unit 102 performs gamma correction optimized for edges on the input image data in order to stabilize the dots in the edge portion continuously. Various methods have been proposed as a method of performing one-to-one conversion such as gamma correction, and a known method may be used. Usually, a configuration of referring to a table is often used.

  The screen processing unit 103 performs screen processing on the input image data and outputs screen processing data. FIG. 3 shows threshold groups used in the screen processing unit 103. In FIG. 3, a region surrounded by a thick line indicates a cell, and a number in the cell indicates a threshold value. Here, there are 32 threshold values in the cell, and each threshold value corresponds to each pixel constituting the input image data. The screen processing unit 103 binarizes the corresponding input pixel value based on each threshold value in the threshold value group. In the screen image data obtained by the screen processing, there is a case where all pixel values in a cell are 0 to a case where all screen pixel values in the cell are 1. Accordingly, the sum of the screen pixel values in the cell is 0 or more and 32 or less, so that 33 gradations can be expressed for each cell. As described above, the screen processing unit 103 binarizes the input pixel value, but actually does not output the binary value of “0” and “1”, but a value corresponding to the quantized representative value. Is output. That is, the screen processing unit 103 outputs “0” when the input pixel value is smaller than the threshold, and “255” when larger than the threshold.

The selection unit 104 selects and outputs either screen image data or gamma converted image data for each pixel according to the determination data obtained by the edge detection unit 101. Specifically, when the value of the target pixel in the input determination data is “1”, that is, when the target pixel is an edge portion pixel, the selection unit 104 performs the gamma conversion image at the position of the target pixel. Select and output the pixel value of the data. On the other hand, if the value of the target pixel in the determination data is “0”, that is, indicates that the target pixel is a non-edge pixel, the selection unit 104 selects the pixel value of the screen image data at the target pixel position. And output. If divided into cases, it can be described as follows.
(I) When the judgment data is “1”, the selection data = gamma conversion image data (ii) When the judgment data is “0”, the selection data = the screen image data. In many cases, line breaks and jaggies cause image quality degradation. Therefore, for the edge portion, the selection unit 104 outputs gamma converted image data to prevent line breaks due to screen processing, thereby preventing image quality deterioration. The selection data obtained in this way has a stable structure obtained by screen processing excluding the edge portion. However, since the edge portion is not data obtained by screen processing, it cannot be said that the desired image forming apparatus is stable data for forming dots. That is, the pixel for which gamma conversion image data is selected is represented by data that is unstable for forming dots.

  The in-cell total calculation unit 105 calculates the sum of pixel values for each cell for the selection data output from the selection unit 104. The in-cell total calculation unit 105 outputs the sum of pixel values for each cell as first total value data. The in-cell total calculation unit 105 also refers to white pixel determination data from the white pixel detection unit 111 when calculating the sum of pixel values. For pixels whose white pixel determination data indicates a white pixel, even if the selection data has a pixel value greater than 0, it is not added to the sum of the pixel values in the cell. That is, the in-cell total calculation unit 105 excludes the pixels of the selection data corresponding to the pixel position having a pixel value of 0 in the input image data when taking the sum. If the pixel value of the white pixel is not converted from 0 to a larger value by gamma conversion or screen processing, 0 is added as the white pixel, so the in-cell total calculation unit 105 determines the white pixel. There is no need to reference the data. The cell used here is a cell having the same shape as the cell used in the screen processing unit 103 (see FIG. 3).

  The in-cell total calculation unit 106 is the same in operation as the in-cell total calculation unit 105 except for the input data as described above. The in-cell total calculation unit 106 receives the input image data, calculates a total value for each cell, and outputs it as second total value data.

  The subtraction unit 107 calculates the density fluctuation amount of the selection data using the input image data as a reference. That is, for each corresponding cell, the first total value data is subtracted from the second total value data and output as difference data. The difference data represents the density fluctuation amount in cell units. The difference data represents a density variation caused by gamma conversion or screen processing, and is used for correction amount calculation in the normalization unit 108 at the next stage.

The normalizing unit 108 obtains a density correction amount per pixel based on the difference data. The normalization unit 108 stores or sets the cell shape in advance, and counts the number of pixels at the edge portion in the cell from the input determination data. At the same time, the number of white pixels in the cell is counted with reference to the white pixel determination data. Next, referring to the determination data for each pixel, if the determination data is “1” and the target pixel is an edge portion pixel, the difference data is calculated from the cell area (the number of pixels in the cell) to white. The correction amount is calculated by dividing the number of pixels by the value obtained by subtraction. For pixels whose determination data is “0” and which are adjacent to the pixel of determination data “1” in any of the four neighborhoods of the top, bottom, left and right, the difference data is calculated from the cell area to the number of pixels in the edge portion. Then, the correction amount is calculated by dividing by the value obtained by subtracting the number of white pixels. For pixels whose determination data is “0” and whose determination data is not “1”, the difference data is corrected by dividing by the value obtained by subtracting the number of white pixels from the cell area. Calculate the amount. These can be described as follows:
(I) When the determination data is “1” (when the target pixel is an edge pixel)
Correction amount = difference data / (cell area−number of white pixels)
(Ii) When the determination data is “0” and the determination data is adjacent to the pixel “1” (when the target pixel is a non-edge portion pixel and adjacent to the edge portion pixel)
Correction amount = difference data / (cell area−number of pixels of edge portion−number of pixels of white pixel)
(Iii) When the determination data is “0” and the determination data is not adjacent to the pixel “1” (when the target pixel is a non-edge portion pixel and not adjacent to the edge portion pixel)
Correction amount = difference data / (cell area−number of white pixels)
As can be seen from the above equation, the correction amount obtained under the condition (i) is smaller than the correction amount obtained under the condition (ii). As described above, for the edge portion, not the screen image data but the gamma conversion data is selected by the selection unit 104, so that the edge portion is continuous in a linear shape without interruption. In addition, since the gamma conversion data has an effect in the direction of increasing the density, the correction amount is often a negative value in order to reduce the effect. For this reason, if the same correction is performed on the edge portion as in the vicinity of the edge, the edge portion may be interrupted and the continuity may be lost. Therefore, in the first embodiment, the correction amount at the edge portion (in the case of (i) above) is made smaller than the correction amount at the edge vicinity portion. In addition, since the screen image data is selected and the portion that is not adjacent to the pixel of the edge portion (that is, in the case of (iii) above), the value is already suitable for the image forming apparatus. In fact, no correction is required. In this embodiment, a relatively small correction amount is set as in (i).

  The adding unit 109 adds the correction amount of the corresponding pixel position in the correction data to the selection data for each pixel to generate intermediate data.

  The stabilizing unit 110 converts the input intermediate data into a value that can be output stably by the image forming apparatus. Although the stabilization processing method is not particularly limited, for example, stabilization is possible by the following method.

  Here, an example will be described in which an operation for guaranteeing the minimum pulse width in a subsequent pulse width modulation circuit (hereinafter referred to as a PWM circuit) (not shown) is performed. When the intermediate pixel value in the target pixel is less than the preset minimum value and there is a pixel having an intermediate pixel value other than 0 in the neighboring pixels, the stabilizing unit 110 determines that the intermediate pixel value in the target pixel is the preset minimum value. Try to be greater than or equal to the value Specifically, the intermediate pixel value having the smaller value is added to the intermediate pixel value having the larger value among the intermediate pixel values of the target pixel or the intermediate pixel values of the neighboring pixels. Then, the intermediate pixel value having the smaller value is set to zero. Here, when the summed intermediate pixel value is still less than the preset minimum value, the summed intermediate pixel value is replaced with the preset minimum value. Further, when there are pixels having intermediate pixel values other than 0 on the left and right of the target pixel, the phase of the PWM circuit may be controlled so that the pulses are continuous. On the other hand, when the intermediate pixel value of the target pixel is less than the preset minimum value and there is no pixel having an intermediate pixel value other than 0 in the neighboring pixels, the intermediate pixel value of the target pixel is replaced with the preset minimum value. You may make it do.

  The process so far will be described with reference to FIG. FIG. 4A shows input image data. First, edge detection is performed on the input image data by the edge detection unit 101, and determination data shown in FIG. 4B is obtained. In the determination data shown in FIG. 4B, “1” is stored for pixels detected as edge portions, and “0” is stored for pixels that are not detected as edge portions. As described above, FIG. 3 shows threshold groups used for screen processing. When screen processing is performed on the input data using the threshold group shown in FIG. 3, screen processing data shown in FIG. 4C is obtained. FIG. 4D shows selection data, and either screen image data or gamma conversion data is selected for each pixel according to the determination data. A pixel value 187 in FIG. 4D is gamma conversion data, and represents the result of gamma conversion of 96 of the input data. The in-cell total value of FIG. 4D calculated by the in-cell total calculation unit 105 is 187 × 8 + 255 × 5 = 22771, and the in-cell total value of FIG. 4A calculated by the in-cell total calculation unit 106 is 96 × 21 = 2016. Next, the subtraction unit 107 calculates the difference between these two in-cell total values. Next, the normalization unit 108 calculates a correction amount for each pixel. In the example of FIG. 4, the cell area (the number of pixels in the cell) is 32, the number of white pixels in the cell is 11, and the number of pixels in the edge portion in the cell is 8. Accordingly, the correction amount for the edge portion can be calculated as -35 by calculating (2016-2771) / (32-11). Further, the correction amount for the edge adjacent portion can be calculated as −58 by calculating (2016-2771) / (32-11-8). From this result, it can be seen that the correction amount for the edge portion is smaller in magnitude (absolute value) than the correction amount for the edge adjacent portion. The correction amount obtained in this way is added to the selection data shown in FIG. 4D by the adding unit 109 to obtain intermediate data shown in FIG. Thereafter, the stabilization unit 110 performs stabilization processing on the obtained intermediate data.

  As described above, the first embodiment has described the configuration in which when the normalization unit 108 determines the correction amount, the correction amount at the edge portion is smaller than the correction amount at the portion near the edge. With this configuration, it is possible to prevent the edges and lines expressed by the gamma conversion data selected by the selection unit 104 with respect to the edge portions from being interrupted by the correction, and to make the edge portions continuous in a linear shape. Therefore, the image quality of characters / line drawings can be improved.

  In the second embodiment, a configuration without the gamma conversion processing unit 102 as compared with the first embodiment will be described. FIG. 2 is a block diagram illustrating the configuration of the image processing apparatus according to the second embodiment. The difference between the second embodiment and the first embodiment is that the gamma conversion processing unit 102 is not provided and the operation of the normalization unit is different (see FIGS. 1 and 2). In FIG. 2, the same reference numerals as those in FIG. Therefore, hereinafter, only differences between the second embodiment and the first embodiment will be described.

  First, since there is no gamma conversion unit, input image data is input to the selection unit 104 instead of gamma conversion data. The selection unit 104 is the same as that of the first embodiment except that the data handled is different. Specifically, when the target pixel is “1” in the determination data from the edge detection unit 101, the selection unit 104 selects and outputs the pixel value of the target pixel position in the input image data. On the other hand, when the target pixel is “0” in the determination data from the edge detection unit 101, the selection unit 104 selects and outputs the pixel value of the target pixel position in the screen processing data.

The normalization unit 208 refers to the determination data and the white pixel determination data. When the determination data is “1” and the target pixel is the edge portion pixel, the normalization unit 208 calculates the difference data from the cell area to the white pixel. A correction amount is calculated by dividing a value obtained by subtracting the number of pixels by a value obtained by multiplying by a constant (M times). Here, M may be determined in advance as a constant or may be set by a register. In order to obtain a good processing result, it is better to use a value of about 1.5 or more and 2 or less as the value of M. On the other hand, when the input determination data is “0”, the correction amount is calculated by dividing the difference data by the value obtained by subtracting the number of white pixels from the cell area. These can be described as follows:
(I) When the determination data is “1” (when the target pixel is an edge pixel)
Correction amount = difference data / ((cell area−number of white pixels) × M)
(Ii) When the determination data is “0” (when the target pixel is a non-edge pixel)
Correction amount = difference data / (cell area−number of white pixels)
As described above, in the second embodiment, similarly to the first embodiment, the correction amount obtained under the condition (i) is smaller in magnitude (absolute value) than the correction amount obtained under the condition (ii). Therefore, the edge can be prevented from being interrupted by the correction as in the first embodiment.

  A specific example will be described with reference to FIG. As in the first embodiment, the input image data is shown in FIG. 4A, and the threshold value group used in the screen processing unit 103 is shown in FIG. At this time, the determination data output by the white pixel detection unit 111 is as shown in FIG. 4B, and the screen processing data is as shown in FIG. 4C. These are the same as those in the first embodiment. However, in the second embodiment, input image data is selected as selection data instead of gamma processing data at the edge portion, and the selection data is as shown in FIG. Therefore, the value calculated by the in-cell total calculation unit 105 is 96 × 8 + 255 × 5 = 2043. Since the number of pixels in the cell excluding data whose input pixel data is “0” is 21 as in the first embodiment, the correction amount for the edge portion is ((2016-2043) / (21 X1.5) is calculated to be 0. However, M is set to 1.5, while the correction amount for the non-edge portion is calculated to be ((2016-2043) / 21 to -1. Therefore, also in Example 2, it can be seen that the correction amount for the edge portion is smaller in magnitude (absolute value) than the correction amount for the non-edge portion.

  As described above, in the second embodiment, the selection unit 104 selects and outputs either input image data or screen processing data according to the edge determination data. Thereby, the input image data is selected at the edge portion. Then, the normalization unit 208 generates appropriate correction data according to the processing in the previous stage. With this configuration, it is possible to prevent discontinuity of continuous lines at the edge portion. Therefore, the image quality of characters / line drawings can be improved.

  In the third embodiment, unlike the configuration for correcting the edge portion described in the first embodiment, a configuration for correcting a white pixel boundary will be described. FIG. 5 is a block diagram illustrating the configuration of the image processing apparatus according to the third embodiment. The third embodiment is different from the first embodiment in that the white pixel boundary detection unit 501 is used instead of the edge detection unit 101 and the operation of the normalization unit is different (see FIGS. 1 and 5). reference). Accordingly, only the differences between the third embodiment and the first embodiment, that is, the white pixel boundary detection unit 501 and the normalization unit 508 will be described below.

  The white pixel boundary detection unit 501 determines whether the target pixel is a pixel on the white pixel boundary, and outputs the result as determination data. A white pixel boundary means a pixel adjacent to a pixel indicating a white pixel. As an example, a white pixel boundary can be detected by the method described below. The white pixel boundary detection unit 501 has a threshold value. Judgment is made when the pixel of interest has a value that exceeds the threshold value, and the values of four adjacent pixels positioned above, below, left, and right of the pixel of interest are compared with the threshold value, and the value of any adjacent pixel is equal to or less than the threshold “1” is output as data. If this condition is not satisfied, “0” is output as determination data. Here, in this embodiment, it is indicated that the pixel value 0 in the input image data is a white pixel. In this case, by setting the threshold value to 0, “1” can be output as determination data only when the adjacent pixel has a density value of 0. The threshold value may be stored in advance, or may be determined by setting a register or the like.

The normalization unit 508 is the same as the normalization unit 108 in that the difference data is input to obtain the density correction amount per pixel. However, unlike the case where the normalization unit 108 considers whether or not the pixel of interest is an edge pixel when calculating the correction amount, the normalization unit 508 in this embodiment calculates the correction amount. Sometimes, it is considered whether or not the pixel of interest is a pixel at the white pixel boundary. In addition, the determination data of “1” indicates that the target pixel is a pixel at the edge portion in the first embodiment, but the target pixel is a pixel at a white pixel boundary in the third embodiment. ing. These can be described as follows:
(I) When the determination data is “1” (when the target pixel is a pixel on the white pixel boundary)
Correction amount = difference data / (cell area−number of white pixels)
(Ii) When the determination data is “0” and the determination data is adjacent to the pixel “1” (when the target pixel is a pixel on the non-white pixel boundary and adjacent to the pixel on the white pixel boundary) )
Correction amount = difference data / (cell area−number of pixels of white pixel−number of pixels of white pixel boundary)
(Iii) When the determination data is “0” and the determination data is not adjacent to a pixel of “1” (when the target pixel is a pixel at a non-white pixel boundary and is not adjacent to a pixel at a white pixel boundary) )
Correction amount = difference data / (cell area−number of white pixels)
The effect by Example 3 is demonstrated. In the configuration for determining the edge portion described in the first and second embodiments, the contrast of the latent image in the electrophotographic process cannot be secured at the determined edge portion, and the latent image is attracted to surrounding dots to form the edge portion. There is a risk that it will be difficult. Therefore, in the third embodiment, dots are prevented from being interrupted by determining and correcting a white pixel boundary where the contrast of the latent image is clear instead of the edge portion. As a result, the image quality of characters / line drawings can be improved.

  In the fourth embodiment, unlike the configuration for correcting the edge portion described in the second embodiment, a configuration for correcting a white pixel boundary will be described. FIG. 6 is a block diagram illustrating the configuration of the image processing apparatus according to the fourth embodiment. The fourth embodiment differs from the second embodiment in that a white pixel boundary detection unit 501 is used instead of the edge detection unit 101, and a normalization unit 608 is used instead of the normalization unit 208. (See FIGS. 2 and 6). In addition, the white pixel boundary detection unit 501 according to the fourth embodiment performs the same operation as that described in the third embodiment. Accordingly, only the normalization unit 608 will be described below.

The normalization unit 608 is the same as the normalization unit 108 and the normalization unit 208 in that the density correction amount per pixel is obtained by inputting the difference data. However, in the fourth embodiment, like the normalization unit 508 shown in the third embodiment, the meaning of the determination data referred to when the normalization unit 608 generates correction data is different. That is, the determination data “1” indicates that the pixel of interest is an edge portion pixel in the second embodiment, but the pixel of interest is a white pixel boundary pixel in the fourth embodiment. ing. These can be described as follows: The formula itself is the same as that shown in the second embodiment.
(I) When the determination data is “1” (when the target pixel is a pixel on the white pixel boundary)
Correction amount = difference data / ((cell area−number of white pixels) × M)
(Ii) When the determination data is “0” (when the target pixel is a pixel at a non-white pixel boundary)
Correction amount = difference data / (cell area−number of white pixels)
The effect of the fourth embodiment is the same as that described in the third embodiment. That is, since the contrast of the latent image is clear at the white pixel boundary, it is possible to prevent the dots from being interrupted at the white pixel boundary, and to improve the image quality of characters and line drawings.

[Other Examples]
In the above-described embodiment, the selection unit 104 is configured to select either screen processing image data or gamma conversion processing data for each pixel. For example, the selection unit 104 can perform similar processing by using screen image data as a base and overwriting some pixels with corresponding pixel values of gamma conversion processing data according to the determination data.

  An object of the present invention is to supply a storage medium storing software program codes for realizing the functions of the above-described embodiments to a system or apparatus, and the system or apparatus (specifically, CPU or MPU) stores the program code. This can also be achieved by executing read. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the program code constitutes the present invention.

  As a storage medium for supplying the program code, for example, a flexible disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, magnetic tape, nonvolatile memory card, ROM, DVD, or the like is used. I can do it.

  In addition, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also the OS running on the computer performs one of the actual processing based on the instruction of the program code. Part or all may be performed. The case where the functions of the above-described embodiments are realized by the processing of the OS is also included in the scope of the present invention.

  Furthermore, the program code read from the storage medium is written in the memory provided in the function expansion board inserted into the computer or the function expansion unit connected to the computer, and then the process is executed based on the instruction of the program code. May be. In addition, the CPU provided in the function expansion board or function expansion unit may execute part or all of the actual processing, and the functions of the above-described embodiments are realized by the execution processing of the CPU provided in the function expansion board or function expansion unit. Such cases are also included in the scope of the present invention.

  Further, the program code for realizing the functions of the above-described embodiments may be executed by one computer (CPU, MPU) or may be executed by cooperation of a plurality of computers. Further, the program code may be executed by a computer, or hardware such as a circuit for realizing the function of the program code may be provided. Alternatively, a part of the program code may be realized by hardware and the remaining part may be executed by a computer.

Claims (13)

Edge detection means for detecting pixels constituting the edge in the input image data and outputting determination data indicating whether or not the target pixel is a pixel of the edge portion;
Gamma conversion processing means for performing gamma conversion processing on the input image data;
Screen processing means for screen-processing the input image data;
Selection means for outputting, as selection data, data in which a pixel value constituting either gamma conversion image data or screen image data is determined for each pixel according to the determination data;
First in-cell total calculation means for calculating the sum of the selection data for each cell used in the screen processing means as a first in-cell total value;
Second in-cell total calculation means for calculating the sum of the input image data for each cell as a second in-cell total value;
Subtraction means for calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
Normalization means for calculating a correction amount for each pixel based on the difference data,
The image processing apparatus according to claim 1, wherein the normalization unit makes a correction amount (absolute value) in an edge portion smaller than a correction amount (absolute value) in a non-edge portion according to the determination data. The normalizing means includes
If the pixel of interest is an edge pixel, the difference data is divided by a value obtained by subtracting the number of white pixels in the cell from the cell area (number of pixels in the cell), and a correction amount is calculated.
When the pixel of interest is a non-edge part pixel and is adjacent to the edge part pixel, the difference data is a value obtained by subtracting the number of pixels of the edge part and the number of white pixels from the cell area. Calculate the correction amount by dividing,
If the target pixel is a non-edge pixel and not adjacent to the edge pixel, the difference data is divided by the value obtained by subtracting the number of white pixels in the cell from the cell area. The image processing apparatus according to claim 1, wherein: Edge detection means for detecting an edge from input image data and outputting determination data indicating whether the pixel of interest is a pixel of an edge portion;
Screen processing means for screen-processing the input image data;
According to the determination data, selecting means for selecting either the input image data or the screen image data and outputting as selection data;
First in-cell total calculating means for calculating the first in-cell total value by taking the sum of the selection data for each cell used in the screen processing means;
Second in-cell total calculation means for calculating a second in-cell total value by taking the sum of the input image data for each cell;
Subtraction means for calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
Normalization means for calculating a correction amount for each pixel from the difference data,
The image processing apparatus according to claim 1, wherein the normalization unit makes a correction amount (absolute value) in an edge portion smaller than a correction amount (absolute value) in a non-edge portion according to the determination data. The normalizing means includes
When the target pixel is an edge portion pixel, the difference data is divided by a constant multiple of a value obtained by subtracting the number of white pixels in the cell from the cell area, thereby calculating a correction amount.
The correction amount is calculated by dividing the difference data by a value obtained by subtracting the number of white pixels in the cell from the cell area when the target pixel is a pixel in a non-edge portion. The image processing apparatus according to 3. Boundary detection means for detecting a white pixel boundary from input image data and outputting determination data indicating whether or not the target pixel is a pixel at the white pixel boundary;
Gamma conversion processing means for performing gamma conversion processing on the input image data;
Screen processing means for screen-processing the input image data;
According to the determination data, selecting means for selecting either gamma-converted image data or screen image data, and outputting as selection data;
First in-cell total calculating means for calculating the first in-cell total value by taking the sum of the selection data for each cell used in the screen processing means;
Second in-cell total calculation means for calculating a second in-cell total value by taking the sum of the input image data for each cell;
Subtraction means for calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
Normalization means for calculating a correction amount for each pixel from the difference data,
The image processing apparatus according to claim 1, wherein the normalization unit makes a correction amount (absolute value) at a white pixel boundary smaller than a correction amount (absolute value) at a non-white pixel boundary according to the determination data. When the pixel of interest is a pixel at a white pixel boundary, the normalizing means divides the difference data by a value obtained by subtracting the number of white pixels in the cell from the cell area (number of pixels in the cell). Calculate the correction amount,
If the pixel of interest is a non-white pixel boundary pixel and is adjacent to a white pixel boundary pixel, the difference data is subtracted from the cell area by the number of white pixels and the number of white pixel boundaries in the cell. The correction amount is calculated by dividing by the calculated value,
If the pixel of interest is a non-white pixel boundary pixel and not adjacent to a white pixel boundary pixel, the difference data is divided by the value obtained by subtracting the number of white pixels in the cell from the cell area. The image processing apparatus according to claim 5, wherein a correction amount is calculated. Boundary detection means for detecting a white pixel boundary from input image data and outputting determination data indicating whether or not the target pixel is a pixel at the white pixel boundary;
Screen processing means for screen-processing the input image data;
According to the determination data, selecting means for selecting either the input image data or the screen image data and outputting as selection data;
First in-cell total calculating means for calculating the first in-cell total value by taking the sum of the selection data for each cell used in the screen processing means;
Second in-cell total calculation means for calculating a second in-cell total value by taking the sum of the input image data for each cell;
Subtraction means for calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
Normalization means for calculating a correction amount for each pixel from the difference data,
The image processing apparatus according to claim 1, wherein the normalization unit makes a correction amount (absolute value) at a white pixel boundary smaller than a correction amount (absolute value) at a non-white pixel boundary according to the determination data. The normalizing means includes
When the target pixel is a pixel at the white pixel boundary, the correction amount is obtained by dividing the difference data by a constant multiple of a value obtained by subtracting the number of white pixels in the cell from the cell area (number of pixels in the cell). To calculate
When the target pixel is a pixel at a non-white pixel boundary, the correction amount is calculated by dividing the difference data by a value obtained by subtracting the number of white pixels in the cell from the cell area. Item 8. The image processing device according to Item 7. An edge detection step of detecting pixels constituting an edge in the input image data and outputting determination data indicating whether or not the target pixel is a pixel of an edge portion;
A gamma conversion processing step for gamma conversion processing the input image data;
A screen processing step of screen-processing the input image data;
A selection step of outputting, as selection data, data in which pixel values constituting either gamma conversion image data or screen image data are determined for each pixel according to the determination data;
A first in-cell total calculation step for calculating a sum of the selection data for each cell used in the screen processing step as a first in-cell total value;
A second in-cell total calculation step for calculating a sum of the input image data for each cell as a second in-cell total value;
A subtraction step of calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
A normalization step of calculating a correction amount for each pixel based on the difference data,
In the normalization step, the correction amount (absolute value) in the edge portion is made smaller than the correction amount (absolute value) in the non-edge portion according to the determination data. An edge detection step of detecting an edge from input image data and outputting determination data indicating whether or not the target pixel is a pixel of an edge portion;
A screen processing step of screen-processing the input image data;
A selection step of selecting either the input image data or the screen image data according to the determination data, and outputting as selection data;
A first in-cell total calculation step of calculating a first in-cell total value by taking the sum of the selection data for each cell used in the screen processing step;
A second in-cell total calculation step of calculating a second in-cell total value by taking the sum of the input image data for each cell;
A subtraction step of calculating a difference between the first in-cell total and the second in-cell total and outputting difference data;
A normalization step of calculating a correction amount for each pixel from the difference data,
In the normalization step, the correction amount (absolute value) in the edge portion is made smaller than the correction amount (absolute value) in the non-edge portion according to the determination data. A boundary detection step of detecting a white pixel boundary from the input image data and outputting determination data indicating whether or not the target pixel is a pixel of the white pixel boundary;
A gamma conversion processing step for gamma conversion processing the input image data;
A screen processing step of screen-processing the input image data;
A selection step of selecting either gamma-converted image data or screen image data according to the determination data and outputting as selection data;
A first in-cell total calculation step of calculating a first in-cell total value by taking the sum of the selection data for each cell used in the screen processing step;
A second in-cell total calculation step of calculating a second in-cell total value by taking the sum of the input image data for each cell;
A subtraction step of calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
A normalization step of calculating a correction amount for each pixel from the difference data,
In the normalization step, the correction amount (absolute value) at a white pixel boundary is made smaller than the correction amount (absolute value) at a non-white pixel boundary in accordance with the determination data. A boundary detection step of detecting a white pixel boundary from the input image data and outputting determination data indicating whether or not the target pixel is a pixel of the white pixel boundary;
A screen processing step of screen-processing the input image data;
A selection step of selecting either the input image data or the screen image data according to the determination data, and outputting as selection data;
A first in-cell total calculation step of calculating a first in-cell total value by taking the sum of the selection data for each cell used in the screen processing step;
A second in-cell total calculation step of calculating a second in-cell total value by taking the sum of the input image data for each cell;
A subtraction step of calculating a difference between the first in-cell total value and the second in-cell total value and outputting difference data;
A normalization step of calculating a correction amount for each pixel from the difference data,
In the normalization step, the correction amount (absolute value) at a white pixel boundary is made smaller than the correction amount (absolute value) at a non-white pixel boundary in accordance with the determination data.   A program for causing a computer to function as the image processing apparatus according to any one of claims 1 to 8.

Images