JP2005136976A - Image processing apparatus, image processing method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology of obtaining a pleasant print output with smooth gradation and high edge reproducibility while keeping an inter-dot interval. <P>SOLUTION: The image processing apparatus receives image data the each pixel of which has a prescribed gradation value, starts searching an initial pixel, and selects pixels configuring a cell until the total sum of the gradation values of pixels including the searched pixels reaches a prescribed threshold or over. When an average gradation value of the pixels in the cell is a prescribed value (e.g., "30") or over in this case, the threshold is changed from "255" used so far into "510" for example in order to keep the inter-dot interval. Further, when a difference between a maximum gradation value and a minimum gradation value in the cell is greater than a prescribed value (e.g., "50"), it is discriminated to be an edge area and the revised threshold value "510" is decreased into "255" to increase doth density. Further, the pixels are selected from pixels closest to the gravity center position. When the pixels are selected as the average gradation value reaches the threshold or over, a dot is generated to the gravity center position. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to halftone processing of gradation image data in an image processing apparatus such as a laser printer. Specifically, a pixel constituting a cell is selected until the threshold value is reached, and processing is performed so that dots are generated at the reference point of the cell, but the threshold value is changed according to the input gradation value of the pixel. About.

  2. Description of the Related Art Conventionally, an image output device such as a printer converts gradation data having multi-value gradation values for each pixel into binary values representing the presence or absence of dots, and prints on printing paper. It is made like that. In general, a process of converting a multi-value gradation value into a binary value is referred to as a halftone process.

  As a halftone processing method, a threshold matrix in which a predetermined threshold value is stored is used, and a dither method for binarizing depending on whether an input gradation value is larger or smaller than the threshold value, or a threshold value is set and the threshold value is set. There is also an error diffusion method in which binarization is performed depending on whether it is larger or smaller than that and the error is distributed to surrounding pixels to proceed with processing.

Furthermore, a halftone process is also known in which a pixel constituting a cell is selected until the sum of gradation values of each pixel of the input image reaches a predetermined threshold value, and a dot is generated at the center position of the cell ( For example, Patent Document 1).
JP 11-27528 A

  However, in the above-mentioned Japanese Patent Application Laid-Open No. 11-27528, selection of pixels constituting a cell is performed using a predetermined table, but the pixels existing on the left side in the table are often already processed, and as a result The generated cells are distorted. In such a case, if a dot is generated at the center of a cell, an adjacent dot is overlapped and formed at a certain portion. Since the print output in such halftone processing is overlapped at a portion where dots are present, the dots are present in a granular manner when viewed as a whole, resulting in a visually unpleasant print output.

  In particular, in the case of a laser printer, for example, when dots are generated every other pixel, the dots may overlap depending on the performance, and it may not be possible to generate dots that express a smooth gradation.

  Further, in Japanese Patent Laid-Open No. 11-27528 described above, the shape of the cell tends to be distorted at the edge portion where the change in gradation value is large, and even if the dot is hit at the center of the cell, the dot is formed in the edge region. The reproducibility was not good because it was too far away or overlapped, and the print output was unpleasant.

  Accordingly, an object of the present invention is to obtain a comfortable print output with a smooth gradation and high reproducibility of edges while maintaining a distance between dots.

  In order to achieve the above object, the present invention converts gradation data having N or more level values for each pixel into data having M (M is an integer) level values for each pixel. In the image processing apparatus, a pixel group generation unit that generates a pixel group by selecting a pixel based on a first reference point for pixel selection until a total of N kinds of level values for each pixel becomes equal to or greater than a threshold value A pixel group reference point determining means for determining the second reference point of the pixel group generated by the pixel group generating means, and M types of pixels located at the second reference point determined by the pixel group reference point determining means. And a pixel group generation unit that updates the first reference point, selects a pixel based on the first reference point, and further generates a level value in the pixel group It is characterized in that the threshold value is updated according to the above. Thereby, for example, a comfortable print output expressing smooth gradation can be obtained while maintaining the distance between dots.

  Furthermore, the present invention is characterized in that, in the image processing apparatus, the pixel group generation means updates the threshold value according to an average value of level values of pixels in the generated pixel group. Thereby, for example, if the area where a pixel is selected as a cell has a high average gradation value, it is determined that the area is a high density area, and the inter-dot distance can be maintained by changing the threshold value.

  Further, in the image processing apparatus according to the aspect of the invention, the pixel group generation unit may increase the threshold value from the previous second value to the second value when the level value in the generated pixel group is larger than the first value. It is characterized by updating to the third value. Thereby, for example, if the change in the level value is large, the threshold value is increased accordingly, so that the cell size can be maintained and the inter-dot distance can be maintained.

  Furthermore, in the image processing apparatus according to the aspect of the invention, the pixel group generation unit may set the threshold value when the difference between the maximum value and the minimum value of the level value in the generated pixel group is larger than the fourth value. It is characterized by updating from a value of 5 to a sixth value lower than the fifth value. Thereby, for example, in the edge region, the dot density can be increased to improve the reproducibility thereof.

  Furthermore, the present invention is characterized in that, in the image processing apparatus, the second reference point is a barycentric position calculated from each pixel position of a pixel group and N kinds of level values of each pixel. Thereby, for example, since dots are hit at the center of gravity of the pixel group grown in a circle, a comfortable print output in which the intervals between the dots are maintained at a constant distance can be obtained.

  Furthermore, the present invention is characterized in that, in the image processing apparatus, the pixel group generation means randomly selects one pixel when there are a plurality of pixels to be selected based on the first reference point. Thereby, for example, the pixel group easily grows in a circular shape, and if a dot is hit at the second reference point, the distance between the dots is kept constant.

  In order to achieve the above object, the present invention provides gradation data having N (N is an integer) or more level values for each pixel, and data having M (M is an integer) types of level values for each pixel. A pixel group generating unit that selects pixels having N level values and selects unprocessed pixels until the sum of the level values for each selected pixel reaches a threshold value in the image processing apparatus for converting to And an adding means for giving M kinds of level values to predetermined pixels of the pixel group generated by the pixel group generating means, the pixel group generating means depending on the generated level value in the pixel group The threshold value is updated. Thereby, for example, a comfortable print output expressing smooth gradation can be obtained while maintaining the distance between dots.

  Furthermore, in order to achieve the above object, the present invention provides gradation data having N (N is an integer) or more level values for each pixel, and data having M (M is an integer) types of level values for each pixel. In the image processing method for converting to a pixel group, a pixel group is generated by selecting a pixel based on a first reference point for pixel selection until a sum of N kinds of level values for each pixel is equal to or greater than a threshold value. A pixel group reference point determination step for determining a second reference point of the pixel group generated in the generation step, a pixel group generation step, and a pixel located at the second reference point determined in the pixel group reference point determination step. A pixel group generation step that updates the first reference point, selects a pixel based on the first reference point, and further generates a pixel in the generated pixel group. The threshold value is updated according to the level value of There. Thereby, for example, a comfortable print output expressing smooth gradation can be obtained while maintaining the distance between dots.

  Furthermore, in order to achieve the above object, the present invention provides gradation data having N (N is an integer) or more level values for each pixel, and data having M (M is an integer) types of level values for each pixel. In the image processing method for converting to a pixel group, a pixel group generation step of selecting a pixel having N types of level values and generating a pixel group by selecting unprocessed pixels until the sum of the level values for each selected pixel reaches a threshold value And an assigning step of assigning M kinds of level values to predetermined pixels of the pixel group generated by the pixel group generating means, and the pixel group generating step includes a threshold value according to the generated level value in the pixel group. It is characterized by updating. Thereby, for example, a comfortable print output expressing smooth gradation can be obtained while maintaining the distance between dots.

  Furthermore, in order to achieve the above object, the present invention provides M (M is an integer) type level values for each pixel with respect to gradation data having N (N is an integer) type level values or more for each pixel. In a program for causing a computer to execute a process of outputting data, pixels are selected based on the first reference point for pixel selection until the sum of N level values for each pixel is equal to or greater than a threshold value. A pixel group generation process for generating a pixel group, a pixel group reference point determination process for determining a second reference point of the pixel group generated by the pixel group generation process, and a second determined by the pixel group reference point determination process A pixel group generation process that updates the first reference point and selects a pixel based on the first reference point. Furthermore, it corresponds to the level value in the generated pixel group. It is characterized by updating the threshold Te. Thereby, for example, a comfortable print output expressing smooth gradation can be obtained while maintaining the distance between dots.

  Furthermore, in order to achieve the above object, the present invention provides M (M is an integer) type level values for each pixel with respect to gradation data having N (N is an integer) type level values or more for each pixel. In a program for causing a computer to execute a process of outputting data having N values, pixels having N types of level values are selected, and unprocessed pixels are selected until the sum of the level values for each selected pixel becomes a threshold value. A pixel group generation process to be generated, and an adding process to give M kinds of level values to predetermined pixels of the pixel group generated by the pixel group generation unit. The pixel group generation process is a level within the generated pixel group. The threshold value is updated according to the value. Thereby, for example, a comfortable print output expressing smooth gradation can be obtained while maintaining the distance between dots.

  The best mode for carrying out the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing the configuration of the entire system to which the present invention is applied. As a whole, the host computer 10 and the image output device 20 are configured.

  The host computer 10 includes an application unit 11 and a rasterizing unit 12. The application unit 11 generates data to be printed such as character data, graphic data, and bitmap data. For example, character data, graphic data, and the like are generated by operating the keyboard or the like using an application program such as a word processor or a graphic tool in the host computer 10. These generated data are output to the rasterizing unit 12.

  The rasterizing unit 12 converts the print target data output from the application unit 11 into 8-bit gradation data for each pixel (or for each dot). Therefore, each pixel has a gradation value (level value) from 0 to 255. The generation of gradation data in the rasterizing unit 12 is actually processed by a driver installed in the host computer 10. The gradation data output from the rasterizing unit 12 is output to the image output device 20. In the present embodiment, the gradation data will be described below as monochrome data.

  The image output apparatus 20 includes an image processing unit 21 and a print engine 22 as a whole. In the image processing unit 21, halftone processing or the like is performed on the gradation data output from the host computer 10. The processed data is output to the print engine 22, and the print engine 22 actually prints on a recording medium such as print paper.

The image processing unit 21 includes a halftone processing unit 211 and a pulse width conversion unit 212. The halftone processing unit 211 receives gradation data output from the host computer 10, converts it into a multi-value (level value) of two or more values, and outputs quantized data. As the halftone processing, in this embodiment, processing using the center of gravity (Circular Cell method, hereinafter referred to as CC method) is performed in which a cell composed of a plurality of pixels is formed and a dot is generated at the center of gravity. Specific contents of this processing will be described later.

  The pulse width modulation unit 212 receives the quantized data output from the halftone processing unit 211, and generates drive data such as whether or not there is a laser drive pulse for each dot with respect to the quantized data. The generated drive data is output to the print engine 22.

  The print engine 22 includes a laser driver 221 and a laser diode (LD) 222. The laser driver 221 receives the drive data from the pulse width modulation unit 212, generates control data indicating the presence or absence of a drive pulse based on the drive data, and outputs the control data to the laser diode 222. The laser diode 222 is driven based on the control data output from the laser driver 221. Further, a photosensitive drum and a transfer belt (not shown) are driven, and data from the host computer 10 is actually printed on a recording medium such as printing paper. Will be.

  Next, a specific configuration of the image output apparatus 20 will be described with reference to FIG. Here, in the image output apparatus 20 of FIG. 1, the halftone processing unit 211 and the pulse width modulation unit 212 correspond to the CPU 24, the ROM 25, and the RAM 26 in FIG.

  The image output device 20 as a whole is composed of an input interface (I / F) 23, a CPU 24, a ROM 25, a RAM 26, and a print engine 22, and are connected to each other via a bus. The input I / F 23 serves as an interface between the host computer 10 and the image output apparatus 20. The input I / F 23 receives gradation data from the host computer 20 transmitted by a predetermined transmission method, and is converted into data that can be processed by the image output device 20. The input gradation data is temporarily stored in the RAM 26.

  The CPU 24 is connected to the input I / F 23, the ROM 25, the RAM 26, and the print engine 22 via a bus, reads out a program stored in the ROM 25, and performs various processes such as a halftone process. Details thereof will be described later. The RAM 26 serves as a working memory for each process executed by the CPU 24 and stores various data after the process. In addition, drive data for driving the laser driver 221 of the print engine 22 is also stored. The specific configuration of the RAM 26 will also be described later.

  The print engine 22 has the same configuration as the print engine of FIG. 1, and the drive data stored in the RAM 26 is input under the control of the CPU 24, and the above-described print processing is performed.

  Next, processing in the halftone processing unit 212 will be specifically described below, but before that, an outline of the CC method will be briefly described. In the halftone processing by the CC method, an initial pixel is first selected, and unprocessed pixels around the initial pixel are selected until a predetermined threshold value is reached, thereby forming a cell composed of a plurality of pixels. When selecting an unprocessed pixel constituting a cell, a pixel closest to the center of gravity of the cell constituted so far is selected. When the threshold value is reached, a value indicating dot generation (for example, “255”) is assigned to the pixel located at the center of gravity of the cell, and a value indicating that dot generation is not performed for other pixels (for example, “0”). ). The CC method obtains quantized data by performing the above processing on all data of the input image.

  In the above-mentioned Japanese Patent Laid-Open No. 11-27528, a pixel at a predetermined position is selected using a table for selecting an unprocessed pixel to be taken into a cell. On the other hand, since the CC method uses the position of the center of gravity to select the unprocessed pixels to be taken into the cell, the cell grows in a circular shape and the inter-dot distance is constant as compared with JP-A-11-27528. Therefore, the dispersibility of dots can be improved, and a visually comfortable print output can be obtained.

  On the other hand, in the case of a laser printer, a latent image is formed by irradiating a photosensitive drum with laser light in an exposure unit. When a dot for one pixel is formed, the size of one pixel is accurately determined by the characteristics of the light. Dots cannot be formed. The amount of light gradually increases from adjacent pixels, and becomes the highest amount of light in the pixel to be irradiated, and gradually decreases. That is, light is irradiated across adjacent pixels. For this reason, when dots are formed every other pixel, dots may be continuously generated depending on the laser printer to be used, and this also causes the generation of granular dots.

  Here, when processing by the CC method is applied to a laser printer, if the threshold value for configuring the cell is made constant, the sum of the gradation values cannot easily reach the threshold value in a region where the gradation value of each pixel is small. Therefore, if the size of the cell is relatively large and dots are generated at the center of gravity, the distance between the dots is kept long, so the probability of suppressing the generation of granular dots increases. However, since the threshold value is reached immediately in a region where the gradation value is large, the cell size is relatively small, and even if dots are hit at the center of gravity, adjacent dots overlap and the inter-dot distance cannot be maintained. The probability of generating granular dots increases. Therefore, in the present invention, by changing the threshold value for configuring the cell according to the gradation value of the pixel in the cell, it is intended to suppress the generation of granular dots and realize a smooth gradation expression. Specifically, when the average value of the gradation values of the pixels constituting the cell is equal to or greater than a certain value, the threshold value is increased and the cell size is kept large, thereby maintaining the inter-dot distance and granularity. This suppresses the generation of new dots.

  Further, as described above, it is difficult for a laser printer to stably output small dots such as the size of one pixel. In such a printer, when the threshold value by the CC method is fixed, in an area where the input gradation value is low, for example, dots often occur only in an area for one pixel. In this case, as described above, a dot is generated up to the adjacent pixel, and a granular dot is generated. Therefore, the threshold value is updated in accordance with the input gradation value so that dots having a stable size such as dots for two pixels and dots for three pixels are generated as much as possible. In this way, in the present invention, it is possible to reduce the probability of generating a small dot even in a region having a low gradation value, and to generate a dot having a stable size that is not granular. Therefore, even in a laser printer that generates granular dots when the dot size is small, changing the threshold value according to the input tone value increases the inter-dot distance, but makes each dot stable. As a result, a non-granular dot is generated and a comfortable print output can be obtained.

  FIG. 3 shows an output example from the laser printer. FIG. 3A shows an example when the threshold value is made constant, and FIG. 3B shows an example when the threshold value is increased by the average gradation value of the pixels in the cell. Both of the input tone values gradually increase from the left to the right in the drawing. If the threshold value is made constant (in the case of FIG. 3A), the dots are densely gathered in the region where the gradation value is large (the right side of the drawing), and the dots are stuck in places due to the characteristics of the laser printer described above. Dots are generated. On the other hand, in the case of FIG. 3B, the inter-dot distance is maintained even in the region where the gradation value is large, and no granular dots are generated. It can be said that such dots express a smooth gradation as a whole.

  Also, in the case of FIG. 3A, the gradation expression itself is expressed by arranging dots so that the dots gradually gather toward the right side. On the other hand, in the case of FIG. 3B, gradation is expressed by increasing each dot gradually toward the right while maintaining the distance between dots. This is because when the threshold value is set high in the case of processing by the CC method, the total size of the pixels captured in the cell increases as the size of the cell increases, and the dot is not for one pixel but for two pixels. This is because it grows as large as 3 pixels.

  However, even if the threshold value is changed in this way, there are the following problems. In other words, if the distance between dots suddenly increases in a region where the tone value of the pixel changes (hereinafter referred to as an edge), the reproducibility of the edge deteriorates. As the distance increases, the expression of the edge deteriorates. In such an edge portion, the resolution should be improved by reducing the cell size and increasing the dot density. FIG. 4 shows an output example when processing is performed such that no threshold is detected and the threshold value is increased when the average gradation value of the pixels constituting the cell increases. As shown in the drawing, the reproducibility of the edge is not good because the inter-dot distance is excessively wide or the dots are overlapped particularly in the curve portion of the character “S”. In this figure, since the input gradation value is set to “0” in the region of the character “S”, the cell does not grow in a circular shape at the edge portion, and is distorted, and a dot is hit at the center of gravity. This is because the dots of adjacent cells may overlap.

  Therefore, if an area where the change in the input tone value is large is detected, the threshold value is reduced, so that the cell size is reduced and the dot density is increased accordingly, so that the edge reproducibility is improved. An output example when the processing according to FIG. 5 is performed is shown. In the curve portion of the letter “S”, the size of the dot is smaller than that of the surrounding dots, but more dots are hit, and the reproducibility is better than in the case of FIG. In this case, the above-described characteristics of the laser printer, that is, the probability that dots are stuck to each other increases, but the resolution should be increased to improve the edge reproducibility. Even in the case of FIG. 5, there is a portion where an adjacent dot is stuck at the edge portion, but on the contrary, this emphasizes the letter “S” and it can be said that the reproducibility is high. Of course, in order to keep the distance between the dots as much as possible and to improve the reproducibility, it is only necessary to set a threshold value that does not cause the dots to be caught.

  In this way, by changing the threshold value for configuring the cell according to the input gradation value, a smooth gradation is expressed while maintaining the distance between dots as shown in FIG. 3, and when the change of the input gradation value is large, By reducing the threshold value, a print output with excellent edge reproducibility can be obtained as shown in FIG.

  The details of the halftone processing by the CC method according to the present invention will be described below with reference to the drawings. 6 to 11 are flowcharts showing processing operations, and FIGS. 12 to 29 show examples of gradation data stored in the RAM 26. FIG. Hereinafter, description will be made along the flowchart.

  FIG. 6 shows a flowchart of the entire halftone process according to the CC method of the present invention. First, as shown in FIG. 6, the CPU 24 reads the program for executing this processing from the ROM 25, and the processing is started (step S10). It is assumed that input gradation data from the host computer 10 is stored in the RAM 26. An example is shown in FIG. As shown in this figure, the RAM 26 has an input data storage area 261a, and input gradation data from the host computer 10 is stored in the input data area 261a under the control of the CPU 24. In this area 261a, each coordinate position corresponds to a pixel position of the entire image. For example, the coordinate position (0, 0) in the input data storage area 261a corresponds to the pixel located at (0, 0) in the entire image. The value stored at each coordinate position indicates the input gradation value of the pixel. Note that the example shown in FIG. 12A has a structure of 7 rows and 12 columns, but this is for ease of explanation, for example, a structure that can store an image for one page (one frame). Alternatively, the number of rows and columns may be smaller than in this figure. Further, the RAM 26 has an output data area 261b for storing quantized data. Similarly to the input data area 261a, the coordinate position of this area 261b corresponds to the pixel position of the entire image, and the configuration is 7 rows and 12 columns, but other configurations may be used. Further, the RAM 26 has a working memory area 270 as shown in FIG. 12B, and stores the coordinates of the barycentric position as the reference point and the total value of the gradation values of the pixels selected so far as the cells. 270a, an area 270b for storing the maximum and minimum gradation values of the pixels in the cell, an area 270c for storing the average value of the gradation values in the cell, and a pixel constituting the cell And an area 270d for storing the threshold value. The areas 270a to 270d are calculated by the CPU 24 and the values are stored. It is assumed that “255” is stored in the area 270d as a default value of the threshold value.

  Returning to FIG. 6, the CPU 24 then determines an initial pixel and sets the reference point to the coordinates of the initial pixel (step S11). When determining the initial pixel, the CPU 24 searches the input data area 261a in the raster direction. Here, the raster direction is an image generated by the host computer 10, that is, the uppermost leftmost direction in the input data storage area 261a. By searching in the raster direction in this way, a cell is arranged immediately below between the two cells grown in a circular shape by the CC method. If dots are hit at the center of gravity of the cell, the dot in the lower row is arranged around the middle between the dots in the upper row when viewed as the whole printing paper. That is, the dots are formed in an oblique direction as a whole. The dots arranged in this way are visually inconspicuous and can provide a comfortable print output. If the dots in the upper row and the dots in the lower row are arranged at the same position in the vertical direction, the dots are arranged too accurately and the dots are visually noticeable. In order to prevent such dot arrangement, the search is performed in the raster direction. In the example of FIG. 12B, when searching in the raster direction, the pixel located at (0, 0) in the input data area 261a becomes an unprocessed pixel (see the input data area 261a in FIG. 12C). Therefore, the CPU 24 selects the pixel located at (0, 0) as the initial pixel. The CPU 24 stores “−1” at the pixel position in the input data area 261a for the pixel that has been processed by the CC method. This is to prevent the subsequent process from being selected as an unprocessed pixel constituting the cell. Therefore, whether or not the pixel is an unprocessed pixel searched in the raster direction is determined based on whether or not “−1” is stored in the selected pixel position.

Further, the CPU 24 calculates the position of the center of gravity in this step S11. This is for the purpose of pixel selection and finally the process of placing dots in the cell. The calculation formula shown below is used for the calculation of the center of gravity position.
(Formula 1)
x _centroid = {{(x coordinate of centroid of cell) × (sum of cell gradation values)}
+ {(X coordinate of selected unprocessed pixel) × (tone value of selected unprocessed pixel)}}
/ (Total tone value of cells + tone value of selected unprocessed pixel)
y _centroid = {{(y coordinate of the centroid of the cell) × (total sum of cell gradation values)}
+ {(Y coordinate of selected unprocessed pixel) × (tone value of selected unprocessed pixel)}}
/ (Total tone value of cells + tone value of selected unprocessed pixel)
(X _{center of gravity} , y _center of gravity are coordinates of the _center of gravity)
This calculation formula is stored in the ROM 25, and the CPU 24 reads out from the ROM 25 and performs the calculation when calculating the position of the center of gravity. Incidentally, in the case of the initial pixel, the sum of the x and y coordinates of the center of gravity of the cell and the gradation value of the cell is calculated as 0 in (Equation 1). Note that a cell is a group of pixels composed of a plurality of pixels as described above, and the “cell centroid” in the above equation is the centroid of the cell constructed so far. When the centroid for the initial pixel is calculated, x _centroid = 0 and y _centroid = 0. The center-of-gravity position calculated in step S11 is stored in the area 270a of the working memory 270 by the CPU 24 (see the area 270a in FIG. 13A). Further, since the CPU 24 selects only the total value of the gradation values, in this case, the pixel of (0, 0), the gradation value “20” of the pixel is also stored in the area 270a.

  Further, in this step S11, the CPU 24 stores the maximum gradation value (max) and minimum value (min) of the gradation values of the pixels constituting the cell and the average gradation value (mean) of the selected pixel in each of the areas 270b and 270c. To do. In the example of FIG. 13A, since only one pixel (0, 0) is selected, the maximum value “20”, the minimum value “20”, and the average gradation value “20” are the respective regions 270b and 270c. Will be stored.

  Further, when the CPU 24 selects a pixel constituting the cell, as described above, the CPU 24 stores “−1” in the pixel position so that it is not selected as an unprocessed pixel thereafter. As shown in the input data area 261a of FIG. 13A, since the pixel located at (0, 0) is selected as the initial pixel, the CPU 24 stores “−1” at this pixel position. The above is the processing performed by the CPU 24 in step S11.

  Returning to FIG. 6, the CPU 24 then determines whether or not the determined initial pixel has a low density (step S12). The case of low density is a case where the selected pixel has a small black component and a large white component. The case of high density is a case where the selected pixel has a large amount of black component and a large amount of white component. It is not always preferable to perform the same processing on the low density and high density pixels. That is, if processing is performed to configure a cell until a threshold value is reached for a high concentration component as in the case of a low concentration, the threshold value is reached immediately, and the cell size cannot be secured, so that Dots occur adjacent to each other in matched cells. Such dots are conspicuous as a whole print output, and a comfortable print output cannot be obtained. Therefore, the process proceeds separately for pixels with a large black component and pixels with a large white component. In this determination, the CPU 24 reads a predetermined value (“127” in this embodiment) stored in the ROM 25, and determines whether the gradation value of the selected pixel is larger or smaller than that value. In the example of FIG. 13A, since the gradation value “20” of the pixel (0, 0) is smaller than the predetermined value “127”, it is determined to be low density (“YES” in step S12), and step S13. Migrate to In the case of high density (“NO” in step S12), the process proceeds to white density processing (step S24), the details of which will be described later.

Returning to FIG. 6, the CPU 24 then selects an unprocessed pixel (step S13). The unprocessed pixel is selected by selecting a pixel at a position closest to the coordinates of the barycentric position stored in the working memory 270. Since the unprocessed pixels that make up the cell are selected using the center of gravity, as described above, the cell tends to be circular, and by forming dots at the center of gravity, a comfortable print output with a constant distance between dots is possible. You can get it. In the example of FIG. 13A, since the barycentric position x _barycenter = 0 and y _barycenter = 0 are stored in the area 270a of the working memory area 270, the pixel position closest to this position is selected. In this example, there are two pixels located at (1, 0) and (0, 1) as the pixels closest to the center of gravity. In this case, the CPU 24 selects any one pixel at random. When selected in a fixed order (for example, the unprocessed pixel to be selected next is the pixel on the right, the pixel to be selected next is the pixel below the pixel, etc.), the cells that are finally configured are regularly arranged. If it is too large, a periodic pattern in which dots are periodically generated in a certain area of print output is generated. Such a pattern is easily visible and a comfortable print output cannot be obtained. For this reason, if there are a plurality of pixels to be selected, they are selected at random. In the example of FIG. 13A, the CPU 24 randomly selects a pixel located at (1, 0) as an unprocessed pixel (see the input data area 261a in FIG. 13B).

  Returning to FIG. 6, the CPU 24 then performs a threshold control process (step S14). This is because by changing the threshold value for configuring the cell, smooth gradation is expressed and edge reproducibility is enhanced. FIG. 8 shows a flowchart of the threshold control process.

  First, when proceeding to this process (step S14), the CPU 24 calculates the maximum value, the minimum value, and the average gradation value including the newly selected pixel (step S141). In the example of FIG. 13A, since a new pixel (1,0) is selected as an unprocessed pixel (step S13), the maximum value in the cell including this pixel is calculated. In this case, since the gradation value of the selected pixel (1, 0) is “20”, the maximum value is “20”, the minimum value is “20”, and the average gradation value is “20” (= (20 + 20) / 2. The CPU 24 stores these values in the areas 270b and 270c (see the areas 270b and 270c in FIG. 13B).

Next, the CPU 24 changes a threshold value for selecting a pixel by a function using the average gradation value of the cell as an argument (step S142). When the average gradation value is high, the threshold value is reached as soon as the pixel is selected until the threshold value is reached, and as described above, the dots are densely formed and granular dots are generated. The threshold value is changed in order to express a smooth gradation. In this embodiment, when the average gradation value (mean) is smaller than “30”, the threshold is changed to “255”, when the average gradation value (mean) is greater than “30” and smaller than “80”, the threshold is changed to “510”. If it is above, change it to “765”. Of course, this is an example, and other average values may be used, or other threshold values may be used. Further, the threshold value is set in three stages, but it may be set in two stages or four or more stages. The value may be set for each image output device 20 so as not to generate granular dots. An example of the function in this case is shown in FIG. This function is a so-called program read function (threshold ()). As shown in this figure, the output threshold value (threshold) is changed according to the value of the average gradation value (mean). Of course, the threshold value may be changed by an arithmetic expression using a mathematical expression as a function. For example, the following arithmetic expression.
(Formula 2)
threshold = (2 + 75 × (total value of gradation values) / (number of selected pixels) / 255) × 255 (threshold is a threshold value)
In this equation, “2” indicates a characteristic value that matches the characteristic of the image output apparatus 20, and “75” is a coefficient for linearly changing the threshold value. The average gradation value is expressed by (total gradation value) / (number of selected pixels). This function is also set so that the threshold value increases as the average gradation value increases. It should be noted that both the equation and the calling function shown in FIG. 23 are stored in the ROM 25 and the like, and each value is obtained by the CPU 24 reading and executing it appropriately during this processing (step S14). That is, as shown in FIG. 13B, the average gradation value “20” stored in the region 270c in step S141 is read and substituted into (Equation 2) or the calling function (threshold ()), etc. “255” is obtained and stored in the area 270d by the CPU 24 again.

  Returning to FIG. 8, next, the CPU 24 determines whether or not the difference between the maximum value and the minimum value of the gradation values in the cell which is the pixel group is larger than a predetermined value (step S143). This is for detecting whether or not the region is an edge region from the difference in input gradation values as described above. Since both the maximum gradation value and the minimum gradation value are stored in the area 270b in step S141, the CPU 24 performs processing by reading out these values from the area 270b. In this embodiment, it is assumed that “50” is set as the predetermined value and is stored in the ROM 25. In the example of FIG. 13B, the difference between the maximum value “20” and the minimum value “20” is “0”, which is smaller than the predetermined value “50”. Therefore, “NO” is selected in this step S143 and the threshold value control process is performed. Is completed and the process proceeds to step S15 in FIG. Note that when the value is larger than the predetermined value in this step S143 (“YES”), a process of lowering the threshold value (step S144) will be described in detail later.

  Returning to FIG. 6, when the threshold control process is completed, the CPU 24 determines whether or not the sum of the gradation values exceeds the threshold (step S15). This is because the pixels constituting the cell are selected until the threshold value is reached based on the threshold value changed in the threshold value control process (step S14). In the example of FIG. 13B, the sum of the gradation values of the pixels constituting the cell is “40”, which is the sum of the gradation values of the (0, 0) and (1, 0) pixels. Does not exceed “255”. Therefore, “NO” is selected in step S15, and the process proceeds to step S16.

In step S16, the CPU 24 calculates a reference point including unprocessed pixels, that is, the position of the center of gravity. This is because the barycentric position of the cell including the pixel selected as an unprocessed pixel (step S13) is updated. The calculation of the center of gravity position uses the above-described (Equation 1). When the center-of-gravity position is calculated in the example of FIG. 13B, x- _centroid = 0.5 and y- _centroid = 0 are obtained and stored in the area 270a of the memory area 270 by the CPU 24 (area 270a in FIG. 13B). reference).

  Next, the CPU 24 calculates the total of the gradation values and stores it in the area 270a of the working memory 270 (step S17). As shown in FIG. 13B, since the sum of the gradation values of the selected pixels ((0, 0) and (1, 0)) is “40”, this value is stored in the area 270a. Further, the CPU 24 stores “−1” in the position of the input data area 261a of the pixel selected as the unprocessed pixel in step S13 (see the input data area 261a in FIG. 5). This is to prevent selection as an unprocessed pixel in subsequent processing. Further, the CPU 24 stores a value (in this case, “0”) indicating that a dot is not placed at the pixel position (1, 0) of the corresponding output data area 261b (see the output data area 261b in the figure).

  Next, the CPU 24 determines whether or not the sum of the gradation values stored in the area 270a of the working memory 270 is equal to the threshold value (step S18). Since the sum of the gradation values and the threshold value are stored in the areas 270a and 270d, respectively, the CPU 24 reads and compares these values as appropriate for processing. In the example of FIG. 13B, the sum of the gradation values is “40”, and the threshold value is “255”, so that “NO” is selected in step S18, and the process proceeds to step S13 again.

  Then, the CPU 24 selects unprocessed pixels until the sum of the gradation values of the selected pixels is equal to the threshold (“YES” in step S18) or exceeds the threshold (step S15). In the example of FIG. 13C, four pixels (pixels (0, 0), (1, 0), (0, 1), and (1, 1)) are selected as the pixels constituting the cell. The state when the process proceeds to step S18 is shown. Up to this point, the total of the gradation values is “80”, and the gradation values of the selected pixels are all the same value, so that the maximum value, the minimum value, and the average gradation value are both “20”, and the average gradation value is also “ “20” remains “255” because it is not “30” or more. Further, the position of the center of gravity is (0.5, 0.5) as a result of calculation according to (Equation 1) (see regions 270a to 270d in the same figure). In step S18, the CPU 24 selects “NO” in this step because the total tone value “80” is not equal to the threshold value “255”, and selects an unprocessed pixel again (step S13).

  As shown in FIG. 13C, after selecting four pixels, the pixels closest to the center of gravity (0.5, 0.5) are (2, 0), (2, 1), (1, 2 ), There are four pixels located at (0, 2), but the pixel of (2, 0) is selected at random (see step S13, input data area 261a in FIG. 14A). Then, the process proceeds to threshold value control processing (step S14). Since the gradation value of the selected pixel (2, 0) is “25” in step S141, the maximum gradation value of the pixels constituting the cell is “25”. Furthermore, the average gradation value is also updated to “21” (= (80 + 25) / 5) (see the areas 270b and 270c in FIG. 14B).

  Next, the CPU 24 performs a threshold value changing process (step S142). However, since the average gradation value is “21” and “30” or less, the threshold value is not changed and “255” is stored in the area 270d. The difference between the maximum value “25” and the minimum value “20” of the gradation value is “5”, which is equal to or less than the predetermined value “50”. Therefore, “NO” is selected in step S143, and the threshold value control process ends without lowering the threshold value.

  After that, since the total of the gradation values is “105”, “NO” is selected in step S15, and the CPU 24 calculates the position of the center of gravity (step S16), calculates the gradation value (step S17), and FIG. ) Is obtained in the area 270a. Further, the CPU 24 stores “−1” in the input data area 261a and stores “0” in the output data area 261b at the pixel position (2, 0) (input and output in FIG. 14B). (Refer to each data area 261a, 261b). Even in this case, since the total tone value “105” is equal to or smaller than the threshold value “255”, “NO” is selected in step S18, and an unprocessed pixel is selected again (step S13).

  Hereinafter, in order from FIG. 14C, the CPU 24 next selects (2, 1) as an unprocessed pixel (step S13). The gradation value of the pixel is “25” and the average gradation value is updated to “21.7” (≈ (105 + 25) / 6) regardless of the maximum value (see the memory area 270 in FIG. 14C). ). Therefore, since it is “30” or less, the threshold value is not changed and the threshold value is not lowered. At this time, since the sum of the gradation values of the pixels constituting the cell is “130” and does not reach the threshold value “255”, “NO” is selected in step S18 and the unprocessed pixel is selected again (step S13).

  Next, as shown in FIG. 15A, the CPU 24 selects (1, 2) as the pixel closest to the barycentric position as the unprocessed pixel. The gradation value of the pixel is “20”, and the average gradation value is “21.4” (≈ (130 + 20) / 7). Since the average gradation value is “30” or less, the threshold value is not changed, and since the maximum and minimum difference is also “50” or less, the threshold value is not reduced.

  By repeating such processing, the state shown in FIG. This is the state immediately before the sum of the gradation values of the pixels selected so far is “250” and reaches the threshold value “255”. The barycentric position, maximum value, minimum value, average gradation value, and threshold value in this state are values shown in the respective areas 270a to 270d of the memory 270 in FIG. Even in this state, since the sum of the gradation values is not equal to the threshold value, “NO” is selected in step S18, and the unprocessed pixel is selected again (step S13).

  In step S13, the CPU 24 selects a pixel closest to this position from the barycentric positions stored in the area 270a. If there are multiple, select randomly. As shown in FIG. 15C, the CPU 24 selects the pixel located at (2, 3). Since the gradation value of the pixel is “25”, the maximum value remains “25” and the average gradation value becomes “22.7” (step S141). Since it is “30” or less, the threshold value is not changed (step S142), and since the difference between the maximum value and the minimum value is “50” or less, the threshold value is not reduced (“NO” is selected in step S143). The process proceeds to S15.

  In step S15, the CPU 24 adds “275” (= 250 + 25) to the total of the gradation values and exceeds the threshold value “255”. Therefore, “YES” is selected in this step. The process proceeds to the calculation process (step S22) when the threshold value is exceeded. A flowchart of this process is shown in FIG.

  When shifting to this process, the CPU 24 first performs a subtraction process of the sum of the density from the threshold, that is, the sum of the gradation values of the cells configured so far (step S221). In the example of FIG. 15C, “5” is obtained by subtracting the total “250” of the gradation values so far from the threshold “255”. As described above, since the total of the gradation values of the pixels selected so far is stored in the area 270a, the CPU 24 performs processing by appropriately reading the values from the area 270a.

Next, the CPU 24 calculates the gravity center position using the value subtracted in step S221 as the gradation value of the pixel selected last (step S222). That is, if the gradation value of the last selected pixel is used as it is, the threshold value is exceeded, so that the centroid calculation is performed with the value just equal to the threshold value as the gradation value of the pixel. The following formula is used as the arithmetic expression.
(Formula 3)
x _centroid = {{(x coordinate of centroid of cell) × (sum of cell gradation values)}
+ {(X coordinate of selected unprocessed pixel) × (gradation value necessary until threshold)}}
/ (Total tone value of cells + required tone value up to threshold)
y _centroid = {{(y coordinate of the centroid of the cell) × (total sum of cell gradation values)}
+ {(Y coordinate of selected unprocessed pixel) × (gradation value necessary until threshold)}}
/ (Total tone value of cells + required tone value up to threshold)
(X _{center of gravity} , y _center of gravity are coordinates of the _center of gravity)
The difference from (Equation 1) is that the gradation value of the selected unprocessed pixel is not used as it is, but the value in which the sum of the gradation values of the cells is the same as the threshold value (“the gradation value necessary until the threshold value”). ) Is different. This arithmetic expression is stored in advance in the ROM 25 as in the case of (Expression 1), and is processed by being read from the ROM 25 when the CPU 24 executes this step. When calculating in the example of FIG. 15C, the center-of-gravity positions are x- _{centroid≈1.75} and y- _{centroid≈0.96} . The calculated barycentric position is stored in the area 270a of the working memory 270 under the control of the CPU 24 (see the area 270a in FIG. 15C). Then, the CPU 24 ends the gradation value calculation process when the threshold value is exceeded, and proceeds to step S23 in FIG.

  In step S <b> 23, the CPU 24 performs processing for returning the remaining gradation values to the unprocessed pixels. In the example of FIG. 15C, the gradation value of the pixel (2, 3) selected last is “25”, and is calculated as a gradation value of “5” in the centroid calculation. (= 25-5) is returned to the pixel position (2, 3) in the input data area 261a (refer to the input data area 261a in the figure). Although it is conceivable to return to a pixel other than the pixels constituting the cell, it has a value that is not the gradation value of that pixel, so that it is not possible to generate a dot that is faithful to the input gradation value. Therefore, the value after subtraction is returned to the last selected pixel so that it can be selected again as an unprocessed pixel. Further, the CPU 24 stores “0” in the output data area 261b corresponding to the selected unprocessed pixel. As shown in the output data area 261b of FIG. 15C, “0” is stored in (2, 3) in this example.

  And CPU24 performs the process which produces | generates a black dot in a gravity center position (step S23). That is, since the centroid position of the cell configured so far is stored in the area 270a of the working memory 270, the CPU 24 reads this value and stores “255” at the pixel position where the centroid position exists in the output data area 261b. It will be. In the example of FIG. 15C, since the center of gravity position (1.75, 0.96) calculated so far is stored in the memory 270, the pixel (2, 1) of the output data area 261b at this position is stored. “255” indicating dot generation is stored (see the output data area 261b in FIG. 16A). Then, the process proceeds to step S20.

  On the other hand, also in the case where it is equal to the threshold value in the step S18 (in the case of “YES”), since the sum of the gradation values already constituting the cell has reached the threshold value, similarly, a process of generating black dots at the center of gravity position ( Step S19) will be performed. Similarly, the process proceeds to step S20.

  In step S20, the CPU 24 determines whether there is an unprocessed pixel. In the example of FIG. 16 (a), processing has been performed so far so that one cell is formed and a black dot is generated at the center of gravity of the cell. It goes. Whether or not the pixel is an unprocessed pixel is determined based on whether or not “−1” is stored at all pixel positions since “−1” is stored in the pixel processed in the input data area 261a. Judgment can be made. In the example of FIG. 16A, since “−1” is not stored in all pixel positions, “YES” is selected in this step S20, and the process proceeds to step S11 again.

  In step S11, the CPU 24 searches for an unprocessed pixel in the raster direction as described above, and selects the pixel (4, 0) as the initial pixel as shown in FIG. Thereafter, when the above-described processing is performed, four pixels are selected (pixels (4, 0), (5, 0), (4, 1), and (5, 1)), and again as unprocessed pixels. A case where (6, 0) is selected (step S13) will be described below. The average gradation value when these four pixels are selected is “30” (= (30 × 4) / 4), but the gradation value of (6, 0) is “35”. When the included average gradation value is calculated (step S141), "31" (= (120 + 35) / 5) is obtained. Since this value exceeds “30” from the calling function shown in FIG. 23, the threshold value is changed from “255” to “510” (step S142). The CPU 24 stores the average gradation value “31” and the changed threshold value “510” in the respective areas 270c and 270d of the working memory 270 (see the respective areas 270c and 270d in FIG. 16C). Since it is determined that the input gradation value has a relatively large value, a larger cell size is secured. As described above, this is because the cell size is secured and the inter-dot distance is maintained by setting the threshold value high in this way. In addition, as described above, although the cell becomes large in the region where the input gradation value is large as described above, one dot is increased by the changed threshold value, so that gradation expression faithful to the input gradation value is obtained as a result. It has been realized. That is, a smooth gradation can be expressed. Of course, in the region where the threshold value is not changed, the higher the input gradation value, the smaller the cell size and the higher the dot density. As a result, even in this case, a smooth gradation can be expressed.

  This will be specifically described with reference to FIG. When the unprocessed pixel (6, 0) is selected in FIG. 16C (step S13) and the threshold value control process (step S14) is completed, the total gradation value “155” (= 120 + 35) is changed. It does not exceed “510”. Accordingly, “NO” is selected in step S15, the center of gravity position is calculated (step S16), the total value of the gradation values is calculated (step S17), and the value is stored in each area 270a and the like (FIG. 17 (a)). Since the sum of the gradation values is not equal to the threshold value, “NO” is selected in step S18, and the unprocessed pixels constituting the cell are selected again (step S13). By repeating this process, the values shown in FIG. 17B are obtained. This figure shows a state immediately before the sum of the gradation values reaches the threshold value “510”. In this figure, eight pixels have been selected so far. The values such as the position of the center of gravity are values shown in the respective areas 270a to 270d in FIG. Then, the CPU 24 selects the pixel (6, 4) shown in FIG. 17C as the unprocessed pixel closest to the gravity center position (5.67, 1.4). The gradation value of the pixel is “35”, the total gradation value of the cells configured so far is “490”, and therefore the total gradation value is “525”, which exceeds the threshold value “510”. In step S15, “YES” is selected, and the process proceeds to step S22.

  In the gradation value calculation process in step S22, first, the CPU 24 subtracts the total value “490” of the gradation values of the cells configured so far from the threshold value “510” (step S221). The center of gravity is calculated by using (Equation 3) as the gradation value of the pixel (6, 4) that is the last selected subtraction value “20” (step S222). The position of the center of gravity is stored in the area 270a of the memory 270. Then, returning to FIG. 6, the remaining density “15” (= 35−20) is returned to the selected pixel (6, 4) (see the input data area 261a in FIG. 17C), and a black dot is generated at the center of gravity. (Step S23).

  At the time of black dot generation, since “510” is set as a threshold value and pixels are selected until the total of gradation values reaches “510”, a value “255” indicating dot generation is assigned to two pixels. That is, as shown in the output data area 261a of FIG. 18A, the CPU 24 first assigns “255” to the pixel (6, 2) existing at the center of gravity. At this time, the total value of the gradation values is “255” (= 510-255) (see region 270a in FIG. 5). Furthermore, “255” is assigned to the pixel (5, 2) closest to the center of gravity (see the output data area 261a in FIG. 18B). By doing in this way, the value for the gradation value selected as a cell is allocated.

  Therefore, in a region where the input tone value is large, the threshold value is set high to ensure the cell size, maintain the inter-dot distance, and express a smooth tone corresponding to the input tone value. Moreover, the number of pixels to be taken into the cell may increase by setting the threshold value higher than a preset value. At this time, when a dot is hit at the center of gravity, the number of pixels to which a value indicating dot generation is assigned increases as the threshold is increased as described above. That is, although the size of the cell increases, the dot easily grows accordingly, so that the gradation according to the input gradation value can be expressed while maintaining the distance between the dots. In FIG. 3B described above, the dots on the right side are larger in size than the dots on the left side, and as a whole, the dot density on the right side seems to be higher. It can be said that it expresses a smooth gradation in which dots do not stick to each other according to the change in the input gradation value.

  When returning to the processing again, the CPU 24 repeats the above-described processing until there is no unprocessed pixel (“NO” in step S20 in FIG. 7). The result is shown in FIG. The value stored in the output data area 261a is output as quantized data from the halftone processing unit 211, output to the print engine 22 via the pulse width modulation unit 212, and print data generated by the host computer 10 is printed. It is.

  In the above-described example, the input gradation value is gradually increased from the left side in the drawing, so that the edge detection is not performed as a result, and “NO” is always selected in step S143 in the threshold control process of FIG. Processing to lower the threshold was not performed. As a next example, processing in the edge region will be described. Hereinafter, description will be made with reference to the drawings shown in FIG. The entire process itself is exactly the same as described above, and the process is performed according to the flowchart shown in FIG.

  It is assumed that the gradation value shown in FIG. 19A is input from the host computer 10 to the input data area 261a. The gradation value changes abruptly at the pixel position (5, 0).

  First, the CPU 24 searches the input data area 261a in the raster direction in order to select the initial pixel in the same manner as the above-described processing. Then, the pixel located at (0, 0) is selected as the initial pixel (step S11). Then, the position of the center of gravity is calculated, and the pixels are continuously selected until the pixels constituting the cell are equal to the threshold value (“YES” in step S18) or exceeded (“YES” in step S15). As a result, the output data shown in the output data area 261b of FIG. 19B is obtained. In the middle of this selection, since the average gradation value does not exceed “30”, the threshold value is not changed (step S142), and the difference between the maximum and minimum of the gradation value does not exceed “50”. The lowering process (step S144) is also not performed.

  The initial pixel is searched again in the raster direction, the pixel located at (4, 0) is selected as the initial pixel (see step S11, input data area 261a in FIG. 19B), and the pixel is selected until the threshold value is exceeded. select. In the middle of this selection, four pixels (pixels of (4,0), (4,1), (5,0), (5,1)) are configured as cells, and (6,0) as unprocessed pixels. FIG. 19C shows a state when “” is selected.

  When the CPU 24 selects the pixel (6, 0) as an unprocessed pixel (step S13), the CPU 24 proceeds to threshold control (step S14). First, the CPU 24 calculates the maximum value, minimum value, and average gradation value of the gradation values of the pixels in the cell including the newly selected pixel (6, 0) and stores them in the corresponding areas 270b and 270c. (Step S141). The minimum value is unchanged, “30”, the maximum value is the gradation value “100” of the pixel of (6, 0), and the average gradation value is “44” (= (120 + 100) / 5) (FIG. 20 ( a) 270b, 270c).

  Next, the CPU 24 reads out the function and changes the threshold value using the average gradation value as an argument (step S142). In this case, since the average gradation value is in a range larger than “30” and not larger than “80”, the threshold value is changed to “510” from the function shown in FIG. 23 (the region in FIG. 20A). 270d). It is determined that the input gradation value is a high region.

  Next, the CPU 24 determines whether or not the difference between the maximum gradation value and the minimum gradation value of the pixels in the cell exceeds a predetermined value “50” (step S143), but the values stored in the areas 270b and 270c. Is read out, the difference is “70” (= 100−30), which exceeds the predetermined value “50”. Therefore, “YES” is selected in this step, and the process of lowering the threshold value is performed. Here, it is determined that this area is an edge area.

The process of lowering the threshold value may be determined by calling a function in the same manner as in step S142, or may be determined using an arithmetic expression such as (Expression 2). In this embodiment, as shown in FIG. 23, the function threshold () is called to determine its value. That is, as shown in the latter half of FIG. 23, when the difference between the maximum value (max_value) and the minimum value (min_value) exceeds 50, the threshold value is uniformly set to “255”. As shown in a region 270d in FIG. 20B, the threshold value is changed from “510” to “255”. As another process for lowering the threshold value, for example, the threshold value may be lowered stepwise in accordance with the average gradation value as in the case of changing the threshold value in step S142. For example, the threshold when “30” or less remains “255”, and can be set to “255” when it is greater than “30” and “80” or less, and “510” when it is greater than “80”. . What is necessary is just to set so that the distance between dots may be kept smaller than the threshold value so far.

  In the case of FIG. 20B, when the threshold control process (step S14) is completed, the CPU 24 calculates the center of gravity (step S16) and the gradation value (step S17), and each calculated value is stored in the area 270a. Further, “−1” is stored in the unprocessed pixel (6, 0) of the input data area 261a, and “0” is stored in the output data area 261b. Since the total of the gradation values is “220” and does not exceed the threshold value “255”, “NO” is selected in step S15.

  Next, the CPU 24 selects “NO” in step S18, and again selects the unprocessed pixels constituting the cell from the pixels closest to the centroid position (step S13). In this example, as shown in FIG. 21A, the pixel (6, 1) is selected. When the process proceeds to the threshold control process (step S14), the CPU 24 calculates an average gradation value or the like (step S141). In this case, the average value is “53.3” (≈ (220 + 100) / 6), and the function is The threshold value is again changed to “510” (see step S142, area 270d in FIG. 21A). However, the difference between the maximum gradation value and the minimum gradation value of the cells configured so far is also “70”, which is equal to or greater than the predetermined value, so “YES” is selected in step S143 to lower the threshold value ( Step S144). This is also because it is determined that the region is an edge region. Similarly, the threshold value is changed to “255” and stored in the area 270d by the CPU 24 (see the area 270d in FIG. 21B). Then, the threshold control process (step S14) ends, and the process proceeds to step S15 in FIG.

  Similarly, the CPU 24 determines whether or not the sum of the gradation values of the cells including the last selected pixel exceeds the threshold value (step S15). In this case, the threshold value is “255” and the last selected value is selected. Since the sum of the gradation values of the cells including the pixel (6, 1) is “320”, “YES” is selected in this step. Then, as described above, the gradation value of the pixel of (6, 1) is set to “35” (= 255−220) by the gradation value calculation process (step S22) when the threshold value is exceeded (Expression 3). ) Is used to calculate the center of gravity (step S222). The remaining gradation value “65” (= 100−35) is returned to the pixel position, and “255” indicating black dot generation is assigned to the calculated gravity center position (see FIG. 21C).

  By repeating this process until there is no unprocessed pixel (“NO” in step S20 in FIG. 7), the process ends (step S21). The result is shown in FIG.

  By detecting the edge in this way and lowering the set threshold value, the cell size is reduced, so that the dot density is increased and the print output with improved edge reproducibility can be obtained. As shown in the output data area 261b of FIG. 22A, the edge area from (5, 0) to (5, 6) is reproduced by a larger number of smaller dots than the surrounding pixels, so that the resolution A comfortable print output with improved edge reproducibility can be obtained without falling.

  Next, an example in which the output dot size is output in units smaller than “255” will be described. In the above-described example, since the output of one dot is all “255”, the example has been described in which dots are applied to all areas for one pixel. In this example, dots are applied to a predetermined area width for one pixel. Is the case.

  The example of the input data of Fig.24 (a) is shown. The configuration of the input data area 261a in this case is a 7 × 8 configuration. For convenience of explanation, it is of course the same as the above example that other configurations may be used. Similarly, each coordinate position corresponds to each pixel position of the input image. Also in the example of this figure, the gradation value gradually increases from the left side to the right side on the input data area 261a. The working memory 270 and the output data area 261b have the same configuration as in the above example.

  The processing operations performed by the CPU 24 also follow the flowcharts shown in FIGS. That is, first, the CPU 24 reads a program for executing this process and starts the process (step S10 in FIG. 6). Next, the initial pixel is determined, and the reference point is set at the coordinate position of the initial pixel (step S11). When the input data area 261a is searched in the raster direction as shown in FIG. 24B, since the pixel (0, 0) is an unprocessed pixel (since “−1” is not stored), this pixel is changed to the initial pixel. And Then, the maximum gradation value, the minimum gradation value, and the average gradation value among the pixels constituting the cell are stored in the respective areas 270b and 270c of the working memory 270 (see the respective areas 270b and 270c in the same figure). Further, since the average gradation value is “70”, the CPU 24 sets the threshold value for selecting the pixels constituting the cell to “510” and stores this value in the area 270d of the memory 270 (area 270d in the figure). reference). Further, the CPU 24 calculates the position of the center of gravity using the above (Equation 1), stores the calculated value and the total value of the gradation values of the pixels constituting the cell in the area 270a, and selects the input data area 261a. “−1” is stored in the pixel position of the initial pixel, and “0” is stored in the corresponding output data area 261b (see FIG. 24C).

  Returning to FIG. 6, since the initial density (gradation value) “70” of the initial pixel is smaller than “127”, it is determined that the density is low in step S11 (“YES”), and step S13 and subsequent steps are executed. That is, the unprocessed pixel closest to the barycentric position stored in the area 270a of the working memory 270 is selected (step S13), and the sum of the gradation values of the pixels constituting the cell is equal to the threshold (“YES” in step S18). ") Until the threshold value is exceeded (" YES "in step S15), the selection of this unprocessed pixel is continued. At this time, if the average value of the gradation values of the pixels constituting the cell in step S14 is 30 or less, for example, the threshold value for pixel selection is left at “255”, and if it is greater than 30 and 80 or less, the threshold value is set. If “510” or greater than 80, “765” is set (step S142). This is because the distance between dots is maintained by increasing the cell size to some extent. When the difference between the maximum gradation value and the minimum gradation value of the pixels constituting the cell is larger than “50”, for example (“YES” in step S143), the threshold value is set to “255” as an edge region. "(Step S144). This is because the reproducibility of the edge is improved by increasing the dot density.

FIG. 25A shows an example in which six pixels are selected as unprocessed pixels, and (2,1) is selected as the seventh unprocessed pixel. At this time, the sum of the gradation values of the seven pixels constituting the cell is “500”, and the threshold value for pixel selection is “510” because the average gradation value is “71” (see FIG. Area 270a, 270d). This is just an example immediately before the sum of the gradation values becomes the threshold value.

  Then, the pixel of (2, 0) is selected as the unprocessed pixel closest to the barycentric position stored in the area 270a (see step S13, input data area 261a in FIG. 25B). At this time, since the total value of the gradation values is “580” (= 510 + 70) and exceeds the threshold value “510” stored in the area 270d, “YES” is selected in step S15, and the threshold value is exceeded. A gradation value calculation process is performed (step S22).

  That is, the total value “500” of the gradation values of the pixels selected as the cells so far is subtracted from the threshold value “510” (step S221), and the subtraction value “10” is the value of the last selected pixel (2, 0). The center of gravity is calculated as the gradation value (step S222). The calculation result is stored in the area 270a by the CPU 24 (see the area 270a in FIG. 25C), the remaining gradation value is returned to the pixel of (2, 0), and black dots are generated (step S23).

With this black dot generation, in the above example, first, a value “255” indicating dot generation is assigned to the pixel existing at the centroid position, and when the threshold is “510”, the pixel closest to the centroid position is selected. When the threshold is “765”, the second closest pixel from the center of gravity position is selected, and the value “255” indicating dot generation is assigned in order from the pixel near the center of gravity position. However, in this example, when “255” is first assigned to the pixel present at the center of gravity, a processed pixel adjacent to the selected pixel is selected when the threshold is higher than that. If there are a plurality of adjacent pixels, they are allocated equally. In this way, finer gradation expression can be achieved than when all the dots are printed in the size of one pixel.

  In the example of FIG. 25C, first, a value “255” indicating dot generation is assigned to the pixel (1, 1) existing at the center of gravity (0.76, 0.98) (see the output data area 261b in FIG. 25). ). At this time, the threshold value is “510” and it is necessary to assign dots for the gradation value of “255”. Select the processed pixels (pixels located at (1, 0), (0, 1), (1, 1), (1, 2)) adjacent to the pixel existing at this barycentric position and select four pixels 64, 64, 64, and 63 are equally allocated to “255” as a whole (see the output data area 261b in FIG. 26A). The output data is output as quantized data to the print engine 22 via the pulse width modulation unit 212. An example of print output in this area is shown in FIG.

  As shown in the example of the print output, the value “64” indicating the generation of the same dot is assigned to the pixel (0, 1) and the pixel (2, 1), but the former pixel is a dot from the right. The latter pixel has dots growing from the left. That is, the dot on the left side of the pixel located at the center of gravity is grown from the right, and conversely, the dot located on the right side is grown from the left. The CPU 24 outputs, to the pulse width modulation unit 212, the data indicating which dot should grow from the left or right side depending on whether the pixel is located on the right side or the left side of the pixel where the center of gravity exists. is there.

  If the processing is further continued in the example of FIG. 26A, when a black dot is generated at the center of gravity position, since there are still unprocessed pixels, the unprocessed pixels are selected again ("YES" is selected in step S20, The process proceeds to step S11). When searching in the raster direction, the pixel position (2, 0) is an unprocessed pixel, so this pixel is selected. At this time, although the gradation value of this pixel is “70”, for example, when it is “20”, if (3, 0) is selected as the unprocessed pixel next, the maximum of the pixels constituting the cell There is a possibility that the difference between the gradation value and the minimum / minimum value exceeds 50 and is detected as an edge. Therefore, the CPU 24 does not use the original gradation value but returns the returned data in step S23 if “0”, which is the process area, has already been stored in the pixel of the output data area 261b corresponding to the pixel selected as the unprocessed pixel. Recognizing that there is, the calculation of the maximum gradation value, the minimum gradation value, and the average gradation value is skipped. This prevents erroneous detection as an edge.

  Thereafter, it is assumed that an unprocessed pixel constituting the cell is selected and (4, 0) is selected as shown in FIG. At this time, the average gradation value is “83”, and the threshold value is changed to “765” (step S142) in the threshold value control process (step S14 in FIG. 6). Similarly, the selection of unprocessed pixels is continued to obtain the values shown in FIG. This is a state immediately before the sum of the gradation values of the pixels constituting the cell exceeds the threshold value “765”.

  As shown in FIG. 27A, when the pixel of (5, 2) is selected, the sum of the gradation values is “850” (= 760 + 90) and exceeds the threshold value “765” (step in FIG. 6). If “YES” in S15), the process proceeds to a calculation process (step S22) when the threshold value is exceeded. The total “760” of the gradation values so far is subtracted from the threshold value “765”, and “5” is calculated as the gradation value of the pixel (5, 2) (steps S221 and S222, FIG. 27). (See (c)).

  Then, the remaining gradation value “85” (= 90−5) is returned to the last selected pixel (5, 2) (see the input data area 261a in FIG. 27B), and black dots are generated ( Step S23). In generating the black dots, first, “255” is assigned to the pixel (4, 1) existing at the center of gravity (3.74, 0.79) (see the output data area 261b in FIG. 27C). Since the total number of gradation values is still “510”, four processed pixels ((4, 0), (3, 1), (5, 1) adjacent to the (4, 1) pixel are present. ), (4, 2) pixels) are equally assigned 128, 128, 127, 127 (a total of “510”) (see the output data area 261b in FIG. 28A). In this case, output dots shown in FIG. 29B are obtained. Since values indicating dot generation are equally assigned from surrounding pixels, it is possible to generate dots that are more faithful to the input gradation value and that are more excellent in gradation expression. Also in this case, since the pixel position (3, 1) is on the left side of the barycentric position, the dot grows from the right, and since the pixel of (5, 1) is on the right side of the barycentric position, the dot is hit so that the dot grows from the left side. Be drunk.

  Thereafter, the above-described processing is repeated to obtain output data shown in FIG. As described above, even in the image output apparatus 20 that can generate dots in a region having a predetermined width out of the dot width of one pixel, a smooth gradation can be expressed and the reproducibility of the edge as in the above-described example. Can output improved dots.

  Next, white density processing (FIGS. 10 to 11) will be described. In the above-described example, after the initial pixel is selected (step S11), the gradation value has a value lower than “127”. Therefore, it is determined that the density is low in step S12, and the following processing is performed. Here, a case where the gradation value of the initial pixel is not low density will be described. For example, if the data shown in the input data area 261a in FIG. 30A is input, it is assumed that the unprocessed pixel is searched in the raster direction and the (0, 0) pixel is selected as described above. At this time, since the gradation value of the pixel located at (0, 0) is “185”, which is higher than “127”, “NO” is determined in step S12, and the process proceeds to white density processing (step S24). Become.

  Flow charts showing the operation of the white density processing are shown in FIGS. This is almost the same as the processing at the low density (step S13 to step S20 in FIG. 6). Until the pixel constituting the cell is equal to the threshold value ("YES" in step S248) or exceeded ("YES" in step S245), the pixel closest to the center of gravity position is continuously selected. If there are multiple, select randomly. The threshold control process (step S244, see FIG. 8) is the same. When the average gradation value is equal to or greater than a predetermined value, the threshold value so far is set to a larger value (step S142), and the maximum gradation value and the minimum gradation value are set. When the difference between the gradation values is equal to or greater than a predetermined value, the threshold value thus set is set to a low value as being an edge portion (step S144).

  The difference between the white density process and the low density process is that the gradation value of the selected unprocessed pixel is not used as it is, but from the maximum value of the input gradation value (here, “255”). And the value is used as the gradation value of the pixel to calculate the center of gravity. For example, in the example of FIG. 30A, the pixel (0, 0) has a gradation value of “185”, but when subtracted from the maximum value, “70” (= 255-185) is obtained. By performing this for each unprocessed pixel selection, the value shown in the input data area 261a in FIG. 30B is obtained. However, in processing, since each unprocessed pixel is selected and then subtracted (step S242), the processing is not performed after the values shown in FIG. In this case, the process is eventually performed assuming that it has the same value as that shown in FIG.

  Also, the difference from the low density processing is that a value “0” indicating that no dot is generated at the pixel position of the corresponding output data area 261b when the unprocessed pixel is selected is stored. In the case of white density processing, a value (here, “255”) indicating that a dot is to be generated is stored at the corresponding pixel position (step S243 in FIG. 10). As shown in FIG. 30 (c), when the pixel (0, 0) is selected as an unprocessed pixel, “255” is stored in the corresponding (0, 0) of the output data area 261b by the control of the CPU 24. become. If the sum of the gradation values of the pixels constituting the cell is equal to the threshold value (“YES” in step S248) or exceeds (“YES” in step S245), a dot is generated at the center of gravity position. Unlike the case of density, a value indicating that dots are not generated (also “0” here) is assigned to the pixel position existing at the center of gravity (see FIG. 30D). However, when it is necessary to assign “0” to a plurality of pixels, the pixel selection is performed by selecting the pixel closest to the center of gravity as in the low-density processing, or even for the pixels adjacent to the first selected pixel. The method of assigning to may be used. In the case of even allocation, the value is subtracted from the value “255” of surrounding pixels.

  Furthermore, when the total of the gradation values exceeds the threshold value, when the density is low, the gradation value of the last selected pixel is set to a value that is exactly equal to the threshold value, and the rest is returned to the pixel. Similarly, the process of returning the remaining gradation value to the pixel is performed (step S253). However, since the gradation value is subtracted in step S242, the original gradation value is restored. For example, as shown in FIG. 30D, (2, 0) is finally selected and the gradation value of the pixel is “10” (the gradation value after subtraction in step S242 is “80”). Since the total value of the cell gradation values was “500”, it was calculated as “10” in order to make it equal to the threshold value “510”, and the remaining gradation value was “70”. This is not the original gradation value but the value obtained by subtraction in step S242. Therefore, the original gradation value “185” is restored and the pixel is restored (see the input data area 261a in the figure).

  As described above, the processing is different between the case of low density with many white components in each pixel and the case of high density with many black components (when performing white density processing). If the same processing as that for low density is performed, the threshold value is reached immediately at two or three pixels, so the size of the cell cannot be secured. This is because the distance cannot be maintained and dots are overlapped, resulting in an unpleasant print output.

  Also in the white density processing as described above, since the content of the threshold control processing itself is the same as that of the low density processing, it is possible to obtain a print output in which smooth gradation is expressed and edge reproducibility is improved.

  In the above example, both low density and high density have been processed as an integer multiple of “255” (510, 765,...) As a threshold value for pixel selection. It may be. As shown in FIG. 31A, when “450” is set as the threshold value, “255” is assigned to the center of gravity position, and the remaining value “195” is processed four pixels adjacent to this pixel position. Are allocated evenly. Further, as shown in FIG. 31B, “255” may be assigned to the center of gravity position, and the remaining value “195” may be assigned to the pixel closest to the center of gravity position. Even in such a case, when the average gradation value is equal to or higher than a predetermined value, the threshold for pixel selection is changed from “450” to, for example, “480”, thereby ensuring the cell size and maintaining the inter-dot distance. be able to. In short, a threshold value corresponding to the average gradation value may be set using a function having the average gradation value as an argument.

  Further, the same effect is obtained in the following modified example. That is, in the above-described example, the halftone processing unit 211 to which the present invention is applied is processed in the image output apparatus 20, but the halftone processing unit 211 is also connected to the host computer as shown in FIG. Even when the present invention is applied to the halftone processing unit 211, the same effect as described above can be obtained. In this case, the quantized data output from the halftone processing 211 is input from the host computer 10 to the image output device 20, and control data indicating dot generation is output from the pulse width modulation 212 to the print engine 22 for printing. It will be.

In addition to the host computer 10, for example, in a mobile phone, a PDA (Personal Digital Assistance), or other information portable terminal, data to be printed may be generated and output to the image output device 20. Further, the terminal may include a halftone processing unit 211 to which the present invention is applied. Furthermore, the halftone processing 211 of the present invention is provided in so-called electronic paper in which paper and a display are combined, and a document or an image is displayed with a binarized (or multivalued) value, or an image output A print output may be obtained in combination with the apparatus 20. Even in these cases, the above-described effects can be obtained.

  In addition, when selecting a pixel constituting a cell by processing by the CC method, the pixel is selected using the calculated center of gravity position, but other than that, the pixel closest to the center position of only the coordinate position without considering the gradation value May be selected. That is, the pixels constituting the cell are selected from the coordinate position (x, y) calculated by (sum of coordinate positions) / (number of selected pixels). For example, when two pixels (0, 0) and (1, 0) are selected as the pixels constituting the cell, the center position is (0.5, 0), which is not closest to the center position. A processing pixel is selected. Further, in the case of three pixels to which (0, 1) is added, the center position is (0.5, 0.5), and the pixel position closest to the pixel position is selected as an unprocessed pixel. In short, when selecting unprocessed pixels constituting a cell, it is not always fixed and selected using a table or the like, but the position of the reference point for pixel selection (the centroid position described in the above example, The center position calculated only by the coordinate position without considering the gradation value may be moved. By moving in this way, there is a higher possibility that the cell shape will grow more circularly than in the case where the cell shape is fixed, and a comfortable image output in which the inter-dot distance is maintained can be obtained.

  In the above example, the average gradation value is calculated every time the pixel constituting the cell is selected. For example, after the initial pixel selection, the average gradation value is obtained until a fixed number of pixels such as one pixel, two pixels, etc. are selected. The same effect can be obtained even when the calculation is performed and thereafter the average gradation value is not calculated. By selecting several pixels near the center of gravity and calculating the average gradation value and setting a threshold value, the calculation amount can be reduced rather than calculating each time the pixel is selected. You can also speed up.

  Furthermore, in the above example, the threshold value is updated by the average gradation value, but the threshold value may be changed by the number of pixels. For example, when the minimum number of pixels is set and more pixels are selected as the pixels constituting the cell, the threshold value can be set from “255” to “510” so far.

In the above example, the edge is detected by calculating from the difference between the maximum gradation value and the minimum gradation value in the cell. However, for example, it may be calculated from the dispersion value or standard deviation of the gradation value. Exhibits exactly the same effect. That is, if the average gradation value is μ and the variance is σ2, the variance and standard deviation are
σ 2 = {1 / (number of pixels)} × Σ (gradation value of each pixel−μ) 2
Standard deviation = √σ2
It becomes. When this variance value or standard deviation value is greater than or equal to a certain value, the threshold is updated as an edge.

  Furthermore, in the above example, the threshold value is updated by the average gradation value, but the threshold value may be changed by the number of pixels. For example, when the minimum number of pixels is set and more pixels are selected as the pixels constituting the cell, the threshold value can be set from “255” to “510” so far.

  Furthermore, in the above-described example, the data output from the host computer 10 has been described as monochrome data. However, processing may be performed on color data of CMYK (cyan, magenta, yellow, black). That is, the rasterizing unit 12 generates RGB (red, green, blue) gradation data for each pixel, and a color conversion unit is provided in the image output device 20 to use the RGB data by using a table or a function. Conversion into CMYK color data. The above-described CC method is performed on each converted CMYK data (each data of cyan, magenta, yellow, and black). Even in this case, in the area where the average gradation value is high for each color component, the threshold value is made larger than the previous value to maintain the inter-dot distance, and in the area where the gradation value difference is large, the increased threshold value is lowered as an edge portion. This increases the dot density and improves edge reproducibility. Of course, the same effect can be obtained even when there are six colors including light cyan and light magenta in addition to the four colors C, M, Y, and K as a color component that expresses a color and a plurality of colors including other color components. Can be obtained.

It is a figure which shows the structure of the whole system to which this invention is applied. It is a figure which shows the specific structure of an image output device. This is an example of print output when the input gradation value gradually changes. It is an example of print output when no edge is detected. It is an example of print output when the threshold value is changed by detecting an edge. It is a flowchart which shows the operation | movement of the process by CC method. It is a flowchart which shows the operation | movement of the process by CC method. It is a flowchart which shows operation | movement of a threshold value control process. It is a flowchart which shows the operation | movement of the gradation value calculation process at the time of exceeding a threshold value. It is a flowchart which shows the operation | movement of a white density calculation process. It is a flowchart which shows the operation | movement of a white density calculation process. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. It is a flowchart which shows the operation | movement of the whole process with respect to C and M. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. It is an example of the function which changes a threshold value. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. It is an example of dot printout. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. 2 is a diagram illustrating an example of an internal configuration of a RAM 26. FIG. It is a figure which shows the structure of the other whole system with which this invention is applied.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Host computer 20 Image output device 21 Image processing part 211 Halftone processing part 212 Pulse width modulation part 22 Print engine 24 CPU 25 ROM 261a Input data storage area 261b Output data storage area 270 Working memory area 270a The area where the total value is stored 270b The area where the maximum gradation value and the minimum gradation value of the cell are stored 270c The area where the average gradation value is stored 270d The area where the threshold value is stored

Claims (11)

In an image processing apparatus that converts gradation data having N (N is an integer) or more level values for each pixel into data having M (M is an integer) level values for each pixel.
Pixel group generation means for selecting a pixel and generating a pixel group based on a first reference point for pixel selection until the sum of the N types of level values for each pixel is equal to or greater than a threshold;
Pixel group reference point determining means for determining a second reference point of the pixel group generated by the pixel group generating means;
Providing means for assigning the M types of level values to the pixels located at the second reference point determined by the pixel group reference point determining means;
The pixel group generation unit updates the first reference point, selects the pixel based on the first reference point, and further updates the threshold according to the generated level value in the pixel group. An image processing apparatus. The image processing apparatus according to claim 1.
The image processing apparatus, wherein the pixel group generation unit updates the threshold value according to an average value of the level values of the generated pixels in the pixel group. The image processing apparatus according to claim 1.
When the level value in the generated pixel group is larger than the first value, the pixel group generation unit updates the threshold value from the second value so far to a third value larger than the second value. An image processing apparatus characterized by that. The image processing apparatus according to claim 1.
When the difference between the maximum value and the minimum value of the level value in the generated pixel group is larger than the fourth value, the pixel group generation means changes the threshold value from the fifth value so far to the fifth value. An image processing apparatus that updates to a lower sixth value. The image processing apparatus according to claim 1.
The image processing apparatus according to claim 2, wherein the second reference point is a barycentric position calculated from each pixel position of the pixel group and the N kinds of level values of each pixel. The image processing apparatus according to claim 1.
The image processing apparatus, wherein the pixel group generation unit randomly selects one pixel when there are a plurality of pixels to be selected based on the first reference point. In an image processing apparatus that converts gradation data having N (N is an integer) or more level values for each pixel into data having M (M is an integer) level values for each pixel.
A pixel group generation unit that selects pixels having the N types of level values, selects unprocessed pixels until a sum of level values for each selected pixel reaches a threshold value, and generates a pixel group;
Providing means for assigning the M kinds of level values to predetermined pixels of the pixel group generated by the pixel group generating means;
The image processing apparatus, wherein the pixel group generation unit updates the threshold value according to the generated level value in the pixel group. In an image processing method for outputting gradation data having N (N is an integer) or more level values for each pixel and data having M (M is an integer) level values for each pixel,
A pixel group generation step of selecting a pixel and generating a pixel group based on a first reference point for pixel selection until the sum of the N types of level values for each pixel is equal to or greater than a threshold;
A pixel group reference point determining step for determining a second reference point of the pixel group generated in the pixel group generating step;
An applying step of applying the M kinds of level values to the pixels located at the second reference point determined in the pixel group reference point determining step,
The pixel group generation step updates the first reference point, selects the pixel based on the first reference point, and further updates the threshold value according to the generated level value in the pixel group. An image processing method. In an image processing method for outputting gradation data having N (N is an integer) or more level values for each pixel and data having M (M is an integer) level values for each pixel,
A pixel group generation step of selecting pixels having the N types of level values, and selecting unprocessed pixels until a sum of level values for the selected pixels reaches a threshold value to generate a pixel group;
An application step of applying the M kinds of level values to predetermined pixels of the pixel group generated by the pixel group generation means,
The pixel group generation step updates the threshold value according to the generated level value in the pixel group. For causing a computer to execute a process of outputting data having M (M is an integer) type of level value for each pixel with respect to gradation data having N or more (N is an integer) type of level value for each pixel. In the program
Based on a first reference point for pixel selection, a pixel group generation process for generating a pixel group by selecting pixels until the sum of the N kinds of level values for each pixel is equal to or greater than a threshold value;
A pixel group reference point determination process for determining a second reference point of the pixel group generated by the pixel group generation process;
A provisioning process for imparting the M kinds of level values to the pixels located at the second reference point determined by the pixel group reference point determination process,
The pixel group generation process updates the first reference point, selects the pixel based on the first reference point, and further updates the threshold according to the generated level value in the pixel group. A program characterized by that. In a program for causing a computer to execute a process of outputting data having M (M is an integer) type of level value for each pixel with respect to gradation data having N (N is an integer) type of level value or more for each pixel. ,
A pixel group generation process of selecting pixels having the N types of level values, selecting unprocessed pixels until a sum of level values for each selected pixel reaches a threshold value, and generating a pixel group;
A provisioning process for imparting the M kinds of level values to predetermined pixels of the pixel group generated by the pixel group generation unit,
The pixel group generation processing updates the threshold value according to the generated level value in the pixel group.

Images