JP5921134B2 - Halftone processing apparatus, halftone processing method, and program - Google Patents

Description

  The present invention relates to an apparatus for performing error diffusion processing.

  2. Description of the Related Art Conventionally, in image processing apparatuses such as printers, copiers, and facsimiles, an error diffusion method and an average density storage method are generally known as pseudo halftone processing (halftone processing) methods with high gradation reproducibility. The error diffusion method binarizes the multi-valued image data of the pixel of interest, gives a predetermined weight to the error between the binarization result and the multi-valued image data before binarization, and uses the weighted value for the pixel of interest. This is added to the pixel data. In the average density storage method, weighted average density is obtained using already binarized data in the vicinity of the target pixel and compared with the image data of the target pixel, and the target pixel is binarized according to the comparison result. The error generated at the time of conversion is diffused to neighboring pixels that have not yet been binarized. FIG. 8A shows an example of an error diffusion coefficient in a typical Floyd-Steinberg method among error diffusion methods. In this case, 7/16 of the error is passed to the pixel immediately after the target pixel (*), and 3/16, 5/16, and 1/16 of the error are distributed to the three pixels below one line, respectively. Become.

  In recent years, in order to meet the demand for higher image quality, digital printing technology has been improved in resolution. However, it is in a situation where isolated dots with the minimum pixels cannot be formed stably. In other words, the pulse width of the recording control signal becomes narrower as the resolution becomes higher, and the potential distribution of the electrostatic latent image on the photosensitive member becomes a lesser shape as the resolution becomes narrower, making it difficult to reproduce dots. .

  As a method for solving this, a technique of switching an error diffusion coefficient in accordance with the output resolution is known (Patent Document 1). In the method disclosed in Patent Document 1, the range in which an error generated in a pixel of interest is diffused is widened (see FIG. 8B), and for example, an image quality equivalent to 600 DPI is obtained from error diffusion processed by 1200 DPI. be able to.

JP 2002-135583 A

  However, in the case of the method disclosed in Patent Document 1, it is necessary to prepare more buffers (memory) in order to save the error. For example, when an error diffusion table for 600 DPI is used, only a buffer for storing an error one line before the line where the target pixel exists is sufficient, but when an error diffusion table for 1200 DPI is used, an error two lines before is stored. A buffer is also required to do this. As a result of requiring such a large number of buffers, the cost increases.

The halftone processing apparatus according to the present invention includes a quantization processing unit that performs quantization processing on each of a plurality of pixels included in a target pixel group including at least the target pixel and a pixel adjacent to the target pixel in the sub-scanning direction. Diffusion means for diffusing a quantization error obtained from the result of the quantization process to a plurality of pixels in a pixel group located in the vicinity of the target pixel group, and a plurality of pixels included in the target pixel group result of the quantization process, if the pixel of interest is isolated dots, comprising changing means for changing the results of the previous SL amount coca processing such that the magnitude of the isolated dot becomes smaller, and the changing means, By taking the logical product of the matrix corresponding to the pixel group of interest and the result of the quantization processing of the pixel group of interest, a dot in which some of the plurality of pixels included in the pixel group of interest are missing is obtained. Change, and controls such that the magnitude of the isolated dot is reduced.

  The range of error diffusion can be expanded without increasing the number of buffers for storing errors.

1 is a block diagram illustrating a configuration of an image forming apparatus. It is a figure which shows the internal structure of an image process part. 2 is a block diagram illustrating an internal configuration of a halftone processing unit according to Embodiment 1. FIG. 3 is a flowchart showing a flow of halftone processing in Embodiment 1. It is a figure which shows the relationship between an attention pixel and an attention pixel group. It is a figure which shows the condition where an accumulation error is added with respect to the attention pixel in a attention pixel group. It is a figure which shows an example of the internal structure of a cumulative error memory. (A) is a figure which shows the error diffusion coefficient in Floyd-Steinberg method, (b) is a figure which shows an example at the time of extending the range which diffuses an error. 6 is a block diagram illustrating an internal configuration of a halftone processing unit according to Embodiment 2. FIG. 12 is a flowchart illustrating a flow of halftone processing according to the second embodiment. It is a figure which shows the relationship between the predetermined coefficient used for calculation of average density | concentration, and an attention pixel group. It is a figure which shows the relationship between the binarization result used for calculation of average density | concentration, and an attention pixel group. FIG. 9 is a block diagram illustrating an internal configuration of a halftone processing unit according to a third embodiment. 12 is a flowchart illustrating a flow of halftone processing according to the third embodiment. (A), (b), (c), (d) is a figure which shows an example of control MATRIX, respectively. It is a figure explaining the position of an attention pixel. It is a figure which shows an example of the predetermined pattern used for pattern matching. FIG. 9 is a block diagram illustrating an internal configuration of a halftone processing unit according to a fourth embodiment.

  The best mode for carrying out the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram illustrating a configuration of an image forming apparatus (printer) according to the present embodiment.

  An image reading unit 101 reads an image of a document set on a document reading table (not shown) and outputs image data.

  An image processing unit 102 performs various types of image processing such as density adjustment processing, color conversion processing, gamma correction processing, and halftone processing. Here, the image processing unit 102 has been described as a component of the printer, but it can also function as an independent device (image processing device). Details of the image processing unit 102 will be described later.

  A storage unit 103 includes a ROM, a RAM, a hard disk (HD), and the like. Various control programs and image processing programs are stored in the ROM. The RAM is used as a reference area or work area where the CPU 104 stores data and various types of information. Further, image data to be subjected to print processing is stored in the RAM or HD.

  Reference numeral 104 denotes a CPU, which executes various control programs and image processing programs stored in the ROM to control each unit in an integrated manner.

  An image output unit 105 forms an image on a recording medium such as recording paper and outputs the image.

  Reference numeral 106 denotes an internal bus of the image forming apparatus that connects the above-described units.

  The image forming apparatus can be connected to a server that manages image data, a personal computer (PC) that instructs execution of printing, and the like via a network.

  Next, various image processing executed by the image processing unit 102 will be described.

  FIG. 2 is a diagram illustrating an internal configuration of the image processing unit 102.

  The image processing unit 102 includes a compression processing unit 201, a development processing unit 202, a color conversion processing unit 203, a density adjustment processing unit 204, a gamma correction processing unit 205, a halftone processing unit 206, and an internal bus 207.

  The compression processing unit 201 performs a compression process of compressing the image data acquired by the image reading unit 101 into a format such as JPEG in order to store the image data in the HD in the storage unit 103. Note that image data to be subjected to compression processing is provided to the image processing unit 102 by the CPU 104 in predetermined block units (for example, 8 × 8 pixel block units). The 8 × 8 pixel block unit image data compressed by the compression processing unit 201 is stored in the HD via the RAM.

  The expansion processing unit 202 expands the compressed 8 × 8 pixel block image data and returns the raster image to a raster image. At this time, the 8 × 8 pixel block is developed in the line memory.

  A color conversion processing unit 203 performs color conversion processing for matching the developed image data with the color space of the printer.

  The density adjustment processing unit 204 performs a process of adjusting the print density at the time of output.

  The gamma correction processing unit 205 performs gamma correction on the image data based on printer gamma characteristic data at the time of image formation determined for each printer.

  The halftone processing unit 206 performs halftone processing on the color-converted image data using an error diffusion method or the like. Here, halftone processing refers to the gradation value (for example, 256 gradations) of image data obtained by reading a document into N-value (for example, binary) image data that can be printed by the image output unit 105. It means converting.

  Next, details of processing executed by the halftone processing unit 206 will be described.

  FIG. 3 is a block diagram showing the internal configuration of the halftone processing unit 206 according to the present embodiment, which performs halftone processing using the error diffusion method, and FIG. 4 is a flowchart showing the flow of the halftone processing.

The halftone processing unit 206 according to the present embodiment includes an error adding unit 301, a quantizing unit 302, an error diffusing unit 303, and an accumulated error memory 304, and the input 256 gradation image data is converted into binary image data. Shall be converted to
In step 401, the halftone processing unit 206 sequentially receives data of each pixel constituting image data from the gamma correction processing unit 205.

In step 402, the error adding unit 301 adds the value of the accumulated error E (x) corresponding to the horizontal pixel position x to the input pixel data. If the received pixel data is I, and the pixel data after adding the accumulated error is I ′, the relationship is as shown in Equation (1).
I ′ = I + E (x) (1)

  In the cumulative error addition process in this step, the same cumulative error E (x) is added to each pixel constituting the target pixel group.

  FIG. 5 is a diagram illustrating the relationship between the input pixel (target pixel) and the target pixel group. In FIG. 5, the main scanning direction is from left to right, and the sub-scanning direction is from top to bottom. Each rectangular area delimited by a solid line indicates an individual target pixel group, and each area delimited by a broken line in each target pixel group indicates each pixel to be input. Each input pixel and each target pixel group are specified by coordinates. Here, the target pixel group is composed of a total of four pixels of 2 pixels × 2 pixels. For example, in the case of the target pixel group A indicated by the reference numeral 501, the position is specified by the coordinates A (1, 1). The target pixel group A specified by the coordinates A (1, 1) has four pixels specified by the coordinates (2, 2), (3, 2), (2, 3), and (3, 3), respectively. Consists of pixels. The error adding unit 301 derives a target pixel group to which the input target pixel belongs from the coordinate information of the input pixel data, and adds the accumulated error E (x).

FIG. 6 shows a situation where the accumulated error E (x) is added to the target pixel (a, b, c, d) in the target pixel group A. For each pixel of interest (a, b, c, d), an average value of errors generated in neighboring pixel groups B, C, D, E is added. For example, assuming that the error diffusion coefficient in the above-mentioned Floyd-Steinberg method is adopted, and the error occurring in each pixel in each pixel group B, C, D, E is e (n ○○), the target pixel group The accumulated error E (x) added to each pixel of A is obtained by the following equation (2).
E (x) = {(e (n46) + e (n45) + e (n36) + e (n35)) * (1/16) + (e (n44) + e (n43) + e (n34) + e (n33)) * (5/16)
+ (e (n42) + e (n41) + e (n32) + e (n31)) * (3/16) + (e (n26) + e (n25) + e (n16) + e (n15)) * (7/16)} / 4
... Formula (2)

  The accumulated error memory 304 stores the value of the accumulated error E (x) calculated by the above equation (2). FIG. 7 is a diagram illustrating an example of the internal structure of the cumulative error memory 304. In the case of the present embodiment, since the target pixel group is composed of 4 pixels of 2 pixels × 2 pixels, the cumulative error E (x (x)) is stored in the storage area of the number (W / 2) of the number of horizontal pixels (W) of the image. ) Will be stored respectively. For example, if the number of horizontal pixels (main scanning direction) of the image data is 1000 (pixels), the cumulative pixels stored in the cumulative error memory 304 are E (1) to E (500). Conventionally, if the error diffusion range is expanded to the same range, all the errors generated at the pixel position n46, the pixel position n45, the pixel position n36, and the pixel position n35 (that is, the error two lines before) ) It was necessary to hold. On the other hand, in the method of this embodiment, it is only necessary to hold the average value of errors in four pixels (that is, one line is sufficient), so there is no need to increase the memory capacity.

  Thus, in this step, the same cumulative error E (x) is added to each pixel constituting the target pixel group.

In step 403, the quantization unit 302 performs quantization processing. Specifically, the pixel data I ′ to which the accumulated error is added is compared with a predetermined threshold value T, and the pixel value to be output is determined. The threshold value T for binarization is set to the median value of the maximum input pixel value and the minimum input pixel value (the threshold value T is 127 to 128 if the input pixel value is an integer value in the range of 0 to 255). The In this case, the output pixel value V is determined by the following equation (3).
V = 0 (when I ′ ≦ T)
V = 255 (when I ′> T) Expression (3)

  In step 404, the error diffusion unit 303 calculates a difference (quantization error E) between the pixel data I ′ to which the accumulated error is added and the output pixel value V, and performs error diffusion processing based on a predetermined error diffusion coefficient. I do. The accumulated error E (x) is calculated based on the error newly generated by the diffusion process and stored in the accumulated error memory 304.

  In step 405, it is determined whether or not the processing has been completed for all the pixels in the image data. If there are unprocessed pixels, the processing returns to step 401 and the processing in steps 401 to 404 is repeated.

  As described above, according to the halftone processing according to the present embodiment, the range of the error diffusion coefficient can be expanded without increasing the memory capacity. Thereby, even if the output resolution is increased, a high-quality image can be obtained without causing an increase in cost due to an increase in memory capacity.

  In the first embodiment, the case of halftone processing using the error inequality method has been described. Next, an embodiment of halftone processing using the average density preservation method will be described as a second embodiment. Note that description of parts common to the first embodiment is simplified or omitted, and here, differences will be mainly described.

  FIG. 9 is a block diagram showing an internal configuration of the halftone processing unit 206 according to the present embodiment that performs halftone processing using the average density preservation method, and FIG. 10 is a flowchart showing the flow of the halftone processing. is there. The halftone processing unit 206 according to this embodiment includes an error addition unit 901, a quantization unit 902, an error diffusion unit 903, an accumulated error memory 904, a quantization result delay memory, and an average density calculation unit 906. As in the first embodiment, an example of converting input 256-gradation image data into binary image data will be described.

  In step 1001, the halftone processing unit 206 sequentially receives data of each pixel constituting the image data from the gamma correction processing unit 205.

In step 1002, the error addition unit 901 adds the value of the accumulated error E (x) corresponding to the horizontal pixel position x to the input pixel data. If the received pixel data is I, and the pixel data after adding the accumulated error is I ′, the relationship is as shown in Equation (4).
I ′ = I + E (x) (4)

  Also in the cumulative error addition processing in this step, the same cumulative error E (x) is added to each pixel constituting the pixel group of interest as in the first embodiment.

In step 1003, the average density calculation unit 906 acquires the quantization result (binarization result) stored in the quantization result delay memory 905, and sets a predetermined coefficient and a delayed binarization result. A product-sum operation is performed to generate a threshold value M (average density) used in the quantization unit 902. FIG. 11 is a diagram illustrating a relationship between a predetermined coefficient used for calculating the average density and the target pixel group. FIG. 12 is a diagram illustrating the relationship between the binarization result used for calculating the average density and the pixel group of interest. The average density that becomes the threshold value M can be obtained by Expression (5).
M = (K35 * (N6A + N69 + N5A + N59) + K34 * (N68 + N67 + N58 + N57) + K33 * (N66 + N65 + N56 + N55) + K32 * (N64 + N63 + N54 + N53) + K31 * (N62 + N61 + N52 + N51) + K25 * (N4A + N3A + N49 + N39) + K24 * (N48 + N47 + N38 + N37) + K23 * (N46 + N45 + N36 + N35) + K22 * (N44 + N43 + N34 + N33) + K21 * (N42 + N41 + N32 + N31) + K15 * (N2A + N29 + N1A + N19) + K14 * (N28 + N27 + N18 + N17)) / 4
... Formula (5)

In the conventional method, for example, when one line is 4992 pixels, it is necessary to hold the binarization result (4line * 4992pix * 1bit = 19968bit) for four lines. According to the present embodiment, it is only necessary to hold the binarization result in units of 4 pixels (pixel group G) in total of 2 pixels × 2 pixels. That is, it is sufficient to hold only the binarization result (2line * 2496pix * 3bit = 14976bit) of 2 lines and 3 bits per pixel group (to hold 0, 1, 2, 3, 4).
In step 1004, the quantization unit 902 compares the pixel data I ′ after adding the accumulated error E (x) with the threshold value M generated by the average density calculation unit 906 to determine a pixel value to be output. That is, the output pixel value V is determined by the following equation (6).
V = 0 (when I ′ ≦ M)
V = 255 (when I ′> M) Expression (6)

  In step 1005, the error diffusion unit 903 calculates a difference (quantization error E) between the pixel data I ′ to which the accumulated error is added and the output pixel value V, and performs error diffusion processing based on a predetermined error diffusion coefficient. I do. Then, the error newly generated by the diffusion process is stored in the accumulated error memory 904.

  In step 1006, it is determined whether or not the processing has been completed for all the pixels in the image data. If there are unprocessed pixels, the processing returns to step 1001 and the processing in steps 1001 to 1005 is repeated.

  As described above, the range of the error diffusion coefficient can be expanded without increasing the memory capacity even by the halftone process using the average density method according to the present embodiment.

  In the first and second embodiments, the range of the error diffusion coefficient and the average density coefficient is expanded while suppressing the memory capacity. For example, when processing is performed in units of 1200 DPI, it is possible to obtain an image quality equivalent to 600 DPI. In this case, the minimum isolated point is 2 pixels × 2 pixels, but may be too large depending on the performance of the printer. In order to obtain a higher image quality, it is necessary to finely control the size of the minimum isolated point in accordance with the performance of the printer. Therefore, a mode for freely controlling the size of the minimum isolated point will be described as a third embodiment. In addition, description is simplified or abbreviate | omitted about the part which is common in Example 1 and 2, and suppose that it demonstrates centering around a difference here.

  FIG. 13 is a block diagram showing the internal configuration of the halftone processing unit 206 according to this embodiment, and FIG. 14 is a flowchart showing the flow of the halftone processing. The halftone processing unit 206 according to the present embodiment has the same configuration as that of the second embodiment (an error addition unit 901, a quantization unit 902, an error diffusion unit 903, a cumulative error memory 904, a quantization result delay memory, an average density calculation unit) 906), the following parts are added. That is, a random number generation unit 1301, a control matrix selection unit 1302, a logical product processing unit 1303, a pattern matching unit 1304, and a selector 1305 are further provided. In the present embodiment, as in the first and second embodiments, a case where input 256-gradation image data is converted into binary image data will be described as an example.

  Steps 1401 to 1405 are the same as steps 1001 to 1005 in the second embodiment. That is, the error diffusion process is performed through addition of the accumulated error E (x) to the input pixel data (S1401 and S1402), generation of a threshold M (average density) (S1403), and determination of an output pixel value (S1404). It is executed (S1405).

  In step 1406, the random number generation unit 1301 generates pseudo random numbers (for example, M series). The generated pseudo random number is sent to the control MATRIX selection unit 1302.

  In step 1407, the control MATRIX selection unit 1302 randomly selects one control MATRIX using the input pseudo-random number. FIGS. 15A to 15D show examples of control MATRIX, and one control MATRIX is randomly selected from a plurality of control MATRIX prepared in advance using pseudorandom numbers. Is done. The selected control MATRIX is sent to the logical product processing unit 1303.

  In step 1408, the logical product processing unit 1303 performs a logical product of the output result of the quantization unit 902 and the control MATRIX selected by the control MATRIX selection unit 1302. That is, the result of the quantization process is changed. For example, assume that the selected control MATRIX is (a) in FIG. 15 and the target pixel is at the position j in FIG. In this case, since a1 in FIG. 15A corresponds to j in FIG. 16, a logical product with this a1 is calculated. That is, in this case, a result is output such that the position of j in FIG. The result of the logical product calculated in this way is sent to the selector 1305.

  In step 1409, the pattern matching unit 1304 compares the quantization result around the pixel of interest stored in the quantization result delay memory 905 with a predetermined pattern prepared in advance, so that the pixel of interest is an isolated point (isolated). It is determined whether it is a dot). In this case, when the quantization result matches a predetermined pattern, the target pixel is determined to be an isolated point. FIG. 17 is a diagram illustrating an example of a predetermined pattern prepared and held in advance. In FIG. 17, n, o, p, and q each indicate a target pixel. In this case, the output value of the pixel of interest n, o, p, q is 255, and the quantization result of its periphery (s15, s16, s25, s26, s31-s36, s41-s46 in FIG. 17) is 0. Then, the target pixel is determined to be an isolated point. If it is determined that the target pixel is an isolated point, the process proceeds to step 1410. If it is determined that the target pixel is not an isolated point, the process proceeds to step 1411. Note that the result of this determination is sent to the selector 1305 as an isolated point signal.

  In step 1410 and step 1411, the selector 1305 switches the output according to the isolated point signal output from the pattern matching unit 1304. Specifically, when it is determined that the target pixel is an isolated point (step 1410), a logical product result is output, and when it is determined that the target pixel is not an isolated point (step 1411), quantization is performed. The result of is output as it is.

In step 1412, it is determined whether or not the processing has been completed for all the pixels in the image data. If there are unprocessed pixels, the processing returns to step 1401 and the processing in steps 1401 to 1411 is repeated.
As described above, according to this embodiment, it is possible to finely control the size of isolated points. Thereby, it becomes possible to obtain a higher quality image.

  In the third embodiment, the aspect in which the isolated point is controlled by pattern matching has been described. In order to cope with various patterns by the method of the third embodiment, it is necessary to hold many patterns, and there is a concern that the circuit scale increases.

  Therefore, a mode for controlling isolated points from input data without using pattern matching will be described as a fourth embodiment. In addition, since most part is common with Example 3, suppose that it demonstrates centering on difference with Example 3.

  FIG. 18 is a block diagram showing the internal configuration of the halftone processing unit 206 according to this embodiment. A difference from FIG. 13 according to the third embodiment is that a density determination unit 1801 exists instead of the pattern matching unit 1304. The density determination unit 1801 determines that the pixel of interest is an isolated point when the pixel of interest is in a low density region.

  The flow of halftone processing in this embodiment is the same as that in the third embodiment. The difference is the contents of step 1409 for determining whether the pixel of interest is an isolated point. Hereinafter, the processing in step 1409 in the present embodiment will be described.

After step 1401 to step 1408, in step 1409, the density determination unit 1801 compares the preset coefficient L with the input pixel data I to determine whether the pixel of interest is an isolated point. Specifically, when the value of the input pixel data I is smaller than the coefficient L, it is determined as an isolated point (isolated dot), and the process proceeds to Step 1410. On the other hand, when the value of the input pixel data I is equal to or larger than the coefficient L, it is determined that the input pixel data I is not an isolated point, and the process proceeds to step 1411. This processing is done.
The subsequent processing in Steps 1410 to 1412 is the same as that in the third embodiment, and will be omitted.

  As described above, according to the present embodiment, it is possible to deal with various patterns without increasing the circuit scale.

(Other examples)
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, etc.) of the system or apparatus reads the program. It is a process to be executed.

Claims (7)

Quantization processing means for performing a quantization process on each of a plurality of pixels included in a target pixel group including at least the target pixel and a pixel adjacent to the target pixel in the sub-scanning direction;
Diffusion means for diffusing a quantization error obtained from the result of the quantization processing to a plurality of pixels in a pixel group located in the vicinity of the target pixel group;
The target result of the quantization process for a plurality of pixels included in the pixel group, the pixel of interest may be isolated dot, to change the result of the previous SL amount coca processing such that the magnitude is decreased of the isolated dots Change means,
Equipped with a,
The changing unit calculates a logical product of a matrix corresponding to the pixel group of interest and a result of quantization processing of the pixel group of interest, thereby missing a part of a plurality of pixels included in the pixel group of interest. The halftone processing apparatus is controlled so as to reduce the size of the isolated dot . 2. The change unit according to claim 1 , wherein the change unit performs the change by randomly determining one matrix to be applied from among a plurality of matrices capable of obtaining dots lacking different portions. The halftone processing apparatus described. The halftone processing apparatus according to claim 2, wherein the changing unit determines the one matrix to be applied using a pseudo random number. Whether the pixel of interest is the isolated dot, claims 1 to 3, further comprising a determining means by pattern matching for comparing the quantization result of the periphery of the target pixel with a predetermined pattern The halftone processing device according to any one of the above. Wherein the pixel of interest is whether the isolated dot, the halftone processing apparatus according to any one of claims 1 to 3, further comprising a determining means by the concentration of the target pixel. A quantization processing step for performing a quantization process on each of the plurality of pixels included in the target pixel group including at least the target pixel and the pixel adjacent to the target pixel in the sub-scanning direction;
A diffusion step of diffusing a quantization error obtained from the result of the quantization process to a plurality of pixels in a pixel group located in the vicinity of the target pixel group;
A change step of changing the result of the quantization processing so that the size of the isolated dot is reduced when the pixel of interest is an isolated dot as a result of the quantization processing for a plurality of pixels included in the pixel-of-interest group. When,
Only including,
In the changing step, a result of the quantization process is obtained by performing a logical product of a matrix corresponding to the pixel group of interest and a result of the quantization process of the pixel group of interest, thereby a plurality of pixels included in the pixel group of interest. A halftone processing method , wherein a part of a pixel is changed to be a missing dot, and the size of the isolated dot is controlled to be small . A program for causing a computer to function as the halftone processing apparatus according to any one of claims 1 to 5 .

Images