JP3757644B2 - Image processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to processing of digital image data, and more particularly to edge processing of character images.
[0002]
[Prior art]
The image processing apparatus processes digital image data obtained by reading a document and outputs digital print data for printing. The image is reproduced based on the digital print data.
The image processing apparatus performs various processes on the digital image data obtained by reading the image of the document in order to better reproduce the image of the document. For character manuscripts, it is desirable to emphasize the edge of the characters in order to reproduce the character image. For this reason, various edge determination methods and data enhancement methods based on edge determination results have been proposed.
[0003]
[Problems to be solved by the invention]
There is a method of dividing a pixel into a plurality of sub-pixels and controlling the density in units of sub-pixels (see, for example, JP-A-9-240053). In this case, if the input gradation data is binary data, the position of the density centroid is always constant. Therefore, even if the density is controlled in units of subpixels, the effect of smoothing the edge cannot be obtained.
[0004]
An object of the present invention is to provide an image processing apparatus that smoothes edges even when input gradation data is binary data.
[0005]
[Means for Solving the Problems]
An image processing apparatus according to the present invention outputs, based on multi-valued image data, digital image data for dividing a pixel into a plurality of sub areas in the main scanning direction and forming an image in units of sub areas. It is. The edge determination means determines the edge direction of the target pixel from the difference in density level between the target pixel and its surrounding pixels based on the multivalued image data. The density level generation means performs a smoothing process on the target pixel using an asymmetric filter centered on the target pixel only for the pixel determined to be an edge by the edge determination unit, and newly sets the density of the target pixel. Generate a level. In this way, the pixel of interest becomes a multi-value gradation even in the case of a binary image, by recalculation from the gradation of the peripheral pixels. The density control means changes the density distribution of the target pixel by setting density levels in a plurality of sub-regions in the target pixel according to the edge direction of the target pixel determined by the edge determination means.
Further, in this image processing apparatus, the density control unit sets density level setting parameters for each sub-region in the target pixel in accordance with the edge direction of the target pixel determined by the edge determination unit. And density level setting means for setting the density level in each of a plurality of sub-regions in the target pixel based on the density level of the target pixel using the density level setting parameter set by the density level control means. Become.
In this image processing apparatus, the filter used by the density level correction unit is asymmetric in the sub-scanning direction.
Further, in this image processing apparatus, the density level correction means selects and uses one from a plurality of filters. Preferably, in the image processing apparatus, the density level correction unit selects a filter that can obtain a minimum density level from the results of smoothing using a plurality of filters.
The image processing method according to the present invention outputs digital image data for forming an image in units of sub-regions by dividing a pixel into a plurality of sub-regions in the main scanning direction based on multi-value image data. This is an image processing method. Based on the multi-value image data, the edge direction of the target pixel is determined from the difference in density level between the target pixel and its surrounding pixels. On the other hand, a smoothing process using an asymmetric filter centered on the target pixel is performed to newly generate a density level of the target pixel, and in accordance with the determined edge direction of the target pixel, a plurality of sub-regions in the target pixel are generated. A feature is that the density distribution of the pixel of interest is changed by setting the density level.
[0006]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an image processing apparatus according to an embodiment of the present invention will be described with reference to the accompanying drawings. In the drawings, the same reference symbols denote the same or equivalent ones.
FIG. 1 shows an embodiment of an image processing apparatus of the present invention. The image processing apparatus generates digital image data for forming an image on the photosensitive member by exposing the photosensitive member based on digital image data input from a document reading device, a computer, and the like, and outputs the digital image data to a printer. . Here, the gradation difference between the target pixel and the surrounding pixels is obtained, and the edge direction in the main scanning direction is identified by the combination of the gradation differences. Then, with respect to the pixel of interest at the edge portion, a density level is recalculated in consideration of surrounding pixels using a filter that is asymmetric in the sub-scanning direction. Then, the density centroid is changed according to the identified edge direction.
[0007]
More specifically, the image data output device 10 outputs digital image data input from a document reading device, a computer or the like not shown here. Here, the image data is output as 8-bit gradation data per pixel. The edge determination unit 12 uses the gradation data output from the image data output device 10 to obtain the gradation difference between the target pixel and the surrounding pixels, and in the main scanning direction by combining the gradation differences. Identifies the edge direction. Then, based on the determination result of the edge determination unit 12, the density level control unit 14 generates a parameter signal for controlling the density centroid in the pixel in units of sub-pixels obtained by dividing the target pixel in the main scanning direction. . On the other hand, the gradation generation unit 16 newly generates gradation data from the gradation data input from the image data output device 10 based on the determination result of the edge determination unit 12. The gamma correction unit 18 performs non-linear conversion of the tone data output from the density generation unit 16 and corrects the non-linear tone distortion of the print unit 24. The density level setting unit 20 controls the density level of the data corrected by the gamma correction unit 18 using the density control parameter signal generated by the density level control unit 14 to change the density centroid in the pixel. The D / A converter 22 converts the digital gradation data obtained by the density level setting unit 20 into an analog signal and outputs the analog signal to the laser drive circuit of the printing unit 24. The print unit 24 modulates the intensity of the laser beam in units of subpixels based on the input data, and forms a halftone image on the recording medium by raster scanning.
[0008]
The edge determination unit 12 determines the edge direction in the main scanning direction of the target pixel by dividing into the following four cases. Based on this determination result, it is determined in which direction the edge is moved. Here, the main scanning direction is the left-right direction. “Right edge” refers to an edge on the right side of a character, that is, an edge when a character portion is on the left side of a pixel of interest. “Left edge” refers to an edge on the left side of a character, that is, an edge when a character portion is on the right side of a pixel of interest. The “thin line edge” refers to an edge when a character portion is in the center of the target pixel, that is, when a right edge and a left edge are present in one target pixel. If none of the above applies, it is a “non-edge portion”.
[0009]
FIG. 2 is a block diagram of the edge determination unit 12. The edge is determined using, for example, a 3 × 3 pixel matrix. First, the gradation difference between the target pixel and the surrounding eight pixels is calculated and divided into a pixel having a higher density and a lower density than the target pixel. As shown in FIG. 3, in the 3 × 3 pixel matrix, V33 represents the gradation data of the target pixel, and V22, V23, V24, V32, V34, V42, V43, and V44 are eight adjacent to the target pixel. Represents pixel gradation data. In the edge determination unit 12 shown in FIG. 2, the eight gradation difference signal generation circuits 120 include gradation data V33 of the target pixel and gradation data V22, V23, V24, V32, V34, V42 of the surrounding eight pixels. , V43, V44 are inputted, and the difference (gradation difference signal) of the gradation data between the peripheral pixel and the target pixel is obtained. The combination determination circuit 122 receives the gradation difference between the target pixel and the surrounding pixels, and determines the edge direction based on the combination of the gradation differences. That is, the gradation difference between the target pixel and the surrounding eight pixels is calculated and divided into a pixel having a higher density and a lower density than the target pixel. Then, the edge direction is identified from the relationship between the density values of the target pixel and the surrounding pixels. Specifically, the combination determination circuit 122 determines the edge direction (right edge, left edge, etc.) in the main scanning direction by combining these eight gradation difference signals, and outputs the right-justified signal MARKR and the left-justified signal MARKL. Generate. The right justified signal MARKR indicates that a right edge exists, and the left justified signal MARKL indicates that a left edge exists. A logic circuit 124 including a NAND gate, two AND gates, and three selectors (selecting A when S = L) determines the edge direction from the right-justified signal MARKR and the left-justified signal MARKL, and determines the edge direction. The signal EDG is output. That is, if both MARKR and MARKL are output, EDG = “01” (thin line edge) is output, and if MARKR or MARKL is output, EDG = “03” (right edge) or EDG = “02”. If “(left edge)” is output and neither MARKR nor MARKL is output, EDG = “00” (non-edge portion) is output.
[0010]
FIG. 4 is a block diagram of the density level control unit 14. The edge direction signal EDG, which is the output of the edge determination unit 12, is input as an address signal, and eight density control parameter signals A1, A2, A3, A4, B1, B2, B3 from the table stored in the eight parameter RAMs 140 are input. , B4. The obtained density control parameter is sent to the density level setting unit 20.
[0011]
FIG. 5 shows the tone generation unit 16. The line (V3) including the target pixel and the image data (V1, V2, V4, V5) of the two lines before and after the line are input to the two types of smoothing circuits 160 and 162. The smoothing circuit 160 and the smoothing circuit 162 send the processing results to the selector 164 and the comparator 166, respectively. The selector 164 selects one of the outputs in accordance with the output of the comparator 166. Here, the selector 164 selects the one with the smaller smoothed gradation value and outputs it as the signal VH1. The selector 166 selects either the output of the selector 164 or the signal V3 (VH2) of the line including the target pixel. Here, the selector 166 selects either one according to the NOR gate 168 to which MARKR and MARKL are input, but here, only in the case of an edge, the output signal VH1 of the smoothing circuit 160 is output as the gradation level VF. In other cases, the signal V3 (VH2) of the target pixel is output as the gradation level VF.
[0012]
FIG. 6 is a block diagram of the density level setting unit 22. In the four density level calculation units 180, four types of combinations A1 and B1, A2 and B2, A3 and B3, A4 and B4 of the density control parameter signal are applied to the gradation data VG nonlinearly converted by the gamma correction unit 12. Are used to perform a primary operation (VH = A * (VG−B)) as shown in each block. As a result, four gradation signals VH1, VH2, VH3, and VH4 are obtained from VG. Next, the selector 182 uses the gray level signal VH1, VH2, VH3 obtained by the primary calculation in the density level calculation unit 180 using the pixel clock CLK and the double speed clock XCLK having a frequency twice that of the pixel clock. , VH4 are switched for each sub-pixel within one pixel to generate a density level signal VD. As a result, one pixel is divided into four sub-pixels, and a density level VD is output for each sub-pixel.
[0013]
FIGS. 7 to 10 show how the density in one pixel changes for each of the edge types determined by the edge determination unit 12. It shows how the digital gradation data given to the four sub-pixels in one pixel change as the gradation of gradation data nonlinearly converted by the gamma correction unit 16 increases. In the figure, the height of the black portion in each sub-pixel represents the density level.
FIG. 7 shows the change in the case of the right edge. Here, the density control parameter signal is as follows. A1 = A2 = A3 = A4 = 4. B1 = 0. B2 = 64. B3 = 128. B4 = 192. In the figure, the gradation levels are 0, 32, 64, 96, 128, 160, 192, 224, and 255. As is apparent in the figure, since it is the right edge, the density is increased in order from the left sub-pixel. In this way, the center of gravity of the pixel density shifts sequentially from the left to the center.
[0014]
FIG. 8 shows the change in the case of the left edge. This is symmetrical with the right edge in FIG. Here, the density control parameter signal is as follows. A1 = A2 = A3 = A4 = 4. B1 = 192. B2 = 128. B3 = 64. B4 = 0. The figure shows the case where the gradation levels are 0, 32, 64, 96, 128, 160, 192, 224, and 255. As is clear in the figure, since it is the left edge, the density is increased in order from the right sub-pixel. In this way, the center of gravity of the pixel density shifts sequentially from the right to the center.
[0015]
FIG. 9 shows the change in the non-edge case. Here, the density control parameter signal is as follows. A1 = A2 = A3 = A4 = 1. B1 = B2 = B3 = B4 = 0. The figure shows the case where the gradation levels are 0, 32, 64, 96, 128, 160, 192, 224, and 255. As can be seen in the figure, since there is no edge, all four sub-pixels have the same density, and therefore the center of gravity of the pixel density is always in the center. The density increases corresponding to the gradation level.
[0016]
FIG. 10 shows the change in the case of a thin line edge. Here, the density control parameter signal is as follows. A1 = A2 = A3 = A4 = 2. B1 = 128, B2 = B3 = 0. B4 = 128. The figure shows the case where the gradation levels are 0, 32, 64, 96, 128, 160, 192, 224, and 255. Since the right edge and the left edge are thin line edges that exist at the same time, as clearly shown in the figure, first, the density of the central two sub-pixels increases corresponding to the gradation level. The center of gravity of the pixel density is always in the center. Next, the density of the two sub-pixels on both sides increases corresponding to the gradation level. The center of gravity of the pixel density is always in the center, but the density distribution gradually spreads to the left and right when the gradation level 128 is exceeded. Although FIGS. 7 to 10 mainly describe the change in the center of gravity of the density, as shown in FIG. 11, the density is not always increased from the end in the target pixel. Also, it is not only the center of gravity that changes, but also the density distribution.
[0017]
FIG. 11 shows the state of edge portion reproduction by the smoothing process. The original image shown on the left side is a binary image. When smoothing processing is performed on this original image, an image shown in the center is obtained. Here, the shaded portion indicates the density level between white and black. The image shown on the right side is an image when density centroid control is further performed using an asymmetric filter in units of subpixels, and it can be seen that the edges are expressed smoothly.
[0018]
FIG. 12 schematically shows the effect of the smoothing process when an asymmetric filter is used and when a symmetric filter is used. When a symmetrical filter is used for the original image shown on the left side, the image shown on the upper right side is obtained. Here, the density centroid is shifted to the right by 1/8 for the first pixel in the edge portion, and the centroid position is not changed for the third pixel by shifting to the right by 3/8 for the second pixel. On the other hand, when the smaller gradation value generated using two asymmetric filters is selected for the original image shown on the left, the image shown on the lower right is obtained. Here, the density centroid is shifted to the right by about 1/3 for the first pixel in the edge portion, and is shifted to the right by about 2/3 for the second pixel, and the barycentric position is not changed for the third pixel. When an asymmetric filter is used, the edge portion is smoothly reproduced in the reproduced image.
[0019]
【The invention's effect】
In the reproduced image, the edge portion is reproduced smoothly.
[Brief description of the drawings]
FIG. 1 is a block diagram of an embodiment of an image processing apparatus.
FIG. 2 is a block diagram of an edge determination unit.
FIG. 3 is a diagram showing a signal distribution of nine pixels for edge determination.
FIG. 4 is a block diagram of a density level control unit.
FIG. 5 is a block diagram of a density generation unit.
FIG. 6 is a block diagram of a density level setting unit.
FIG. 7 is a diagram showing a change in density with respect to a gradation level in the case of a right edge.
FIG. 8 is a diagram showing a change in density with respect to a gradation level in the case of a left edge.
FIG. 9 is a diagram illustrating a change in density with respect to a gradation level in the case of a non-edge portion.
FIG. 10 is a diagram showing a change in density with respect to a gradation level in the case of a thin line edge.
FIG. 11 is a diagram illustrating an example of a result of smoothing processing.
FIG. 12 is a diagram illustrating an example of a result of smoothing processing using an asymmetric filter.
[Explanation of symbols]
10 image data output device, 12 edge determination unit,
14 density level control unit, 16 gradation generation unit, 18 gamma correction unit,
20 Density level setting section.

Claims (6)

An image processing apparatus that outputs digital image data for forming an image in units of sub-regions by dividing a pixel into a plurality of sub-regions in the main scanning direction based on multi-valued image data.
An edge determination means on the basis of the multivalued image data, to determine the edge direction of the pixel of interest pixel of interest from the difference in density level of the surrounding pixels,
Only for pixels that are determined to be edges by the edge determination means , a density level that performs a smoothing process using an asymmetric filter centered on the target pixel with respect to the target pixel and newly generates a density level of the target pixel Generating means;
A density control unit configured to change a density distribution of the target pixel by setting density levels in a plurality of sub-regions in the target pixel according to the edge direction of the target pixel determined by the edge determination unit. Image processing device. The concentration control means is
Density level control means for setting a density level setting parameter for each sub-region in the target pixel according to the edge direction of the target pixel determined by the edge determination means;
Using density level setting parameters set by the density level control means, based on the density level of the pixel of interest, and density level setting means for setting the density level in each of a plurality of sub-regions within the pixel of interest. The image processing apparatus according to claim 1, wherein the image processing apparatus is characterized. The image processing apparatus according to claim 1, wherein the filter used in the density level correction unit is asymmetric in the sub-scanning direction. The image processing apparatus according to claim 1, wherein the density level correcting unit selects and uses one of a plurality of filters. The image processing apparatus according to claim 4, wherein the density level correction unit selects a filter that obtains a minimum density level from a result of smoothing using a plurality of filters. An image processing method for outputting digital image data for dividing a pixel into a plurality of sub-regions in the main scanning direction based on multi-value image data and forming an image in units of sub-regions,
Based on the multi-valued image data, determine the edge direction of the target pixel from the difference in density level between the target pixel and its surrounding pixels,
Only a pixel determined to be an edge is subjected to a smoothing process using an asymmetric filter centered on the target pixel with respect to the target pixel, and a new density level of the target pixel is generated.
A method of image processing characterized in that, according to the determined edge direction of a target pixel, density levels are set in a plurality of sub-regions within the target pixel to change the density distribution of the target pixel.

Images