JP4261564B2 - Image forming method, image processing apparatus, and storage medium - Google Patents

Description

  The present invention relates to various apparatuses that handle images such as copiers, printers, scanners, and fax machines, and more particularly to an image forming method and an image processing apparatus that use an error diffusion method.

  For example, in a laser printer, a digital copying machine, and other various image processing apparatuses, a dither method or an error diffusion method is generally used to reproduce the gradation of an image in a pseudo manner.

  In general, the dither method has an advantage that it is excellent in graininess and can express a halftone image smoothly, but also has a disadvantage. For example, in the dither method (area gradation method typified by), the resolution is deteriorated in order to obtain gradation. In addition, in the dither method for generating a periodic image, moire tends to occur on a printed image such as a halftone dot.

  On the other hand, the error diffusion method can obtain a resolution that is faithful to the original image and is suitable for reproducing a character image. However, in a halftone image such as a photograph, isolated dots are dispersed or irregularly connected to each other, so that granularity is poor and a peculiar texture may occur. In particular, in an electrophotographic printer, an image is formed with isolated dots, so that the image is unstable, and graininess deterioration and banding are likely to occur due to density unevenness.

Regarding the error diffusion method, in order to improve the texture due to irregular connection of dots, a dither threshold is used as the quantization threshold, and the following improvements including the method of improving the texture by disturbing dot connection Technology has been proposed.
(1) A dither threshold is used for the purpose of eliminating the occurrence of pseudo contours and unique stripe patterns, and the larger the edge amount, the larger the error diffusion amount (Patent Document 1).
(2) For the purpose of preventing white spots in a non-edge low density portion and preventing occurrence of a character notch, a fixed threshold is used in the edge portion of the image, a variation threshold is used in the non-edge portion, and the level of the variation threshold The lower the concentration, the lower the value (Patent Document 2).
(3) When using a multi-value printer of three or more values, a dither signal having a magnitude corresponding to the edge amount is added to the image data at the edge portion of the image for the purpose of preventing the occurrence of moire and pseudo contour. In the edge portion, a fixed value is added to the image data, and the image data after the addition is subjected to multilevel quantization using a fixed threshold value (Patent Document 3).
(4) In order to smoothly reproduce an image portion in which the density changes smoothly and reproduce a fine line or a character with emphasis on the sharpness, the non-edge portion has a large number of quantization levels (for example, 16 values). The quantization result by the means is selected, and the quantization result by the quantization means having a small number of quantization levels (for example, 6 values) is selected at the edge portion (Patent Document 4).

JP-A-3-34772 Japanese Patent No. 2755307 Japanese Patent No. 2801195 Japanese Patent No. 2851662

  The present invention relates to an improvement in an image forming method and an image processing apparatus using an error diffusion method. Tone characteristics at low and medium density levels, image formation with good stability, and images with few white spots at high density levels A novel image forming method capable of forming a high-resolution image of a character portion, a line drawing portion of an image, a halftone dot image portion having a relatively low number of lines, a smooth reproduction of a joint portion of regions having different image characteristics, and the like An object of the present invention is to provide an image processing apparatus.

The image forming method of the present invention is to prevent the occurrence of white spots by suppressing the generation of small dots in the low and medium density portions to improve the stability of the image and increasing the number of dots in the high density portions. In an image forming method in which a multi-tone image is quantized by an n-value (n ≧ 3) by an error diffusion method, and the pixels of the quantized image are expressed by larger dots as the quantization level is higher, the multi-tone image The m value (m <n) quantization is performed at the low and medium concentration portions.
According to a first aspect of the present invention, in such an image forming method, for quantization of a multi-tone image, (n−1) quantization thresholds periodically changing in an image space are used, This is because at least two quantization threshold values corresponding to the low and medium density levels of the multi-tone image have the same value at specific positions within the fluctuation period.
According to a second aspect of the present invention, in the image forming method according to the first aspect, dots are concentrated in the low / medium density portion to improve gradation and stability, and in the high density portion, around the concentrated dots. In order to prevent dots from being generated by dispersing the dots, the specific position is set as a central portion within the fluctuation cycle of the quantization threshold.
Features of the invention of claim 3, wherein, in the image forming method according to claim 1, wherein, by switching the processing according to the image characteristic to the dither tones from the error diffusion keynote, to multi-tone images having various image features In order to enable optimal image formation, the variation range of the quantization threshold is changed according to the image characteristics of the multi-tone image.
According to a fourth aspect of the present invention , there is provided an image forming method according to the third aspect of the present invention, in order to improve resolution by processing an error diffusion tone in a character portion or a line drawing portion of a multi-tone image. That is, the variation range of the quantization threshold is decreased as the edge degree of the image is larger.
According to a fifth aspect of the present invention, in the image forming method according to the third aspect of the present invention, in order to smoothly reproduce the boundary portion of the image regions having different image characteristics, the average value of the quantization threshold value is kept substantially constant. It is that.
According to a sixth aspect of the present invention, in the image forming method of the first aspect, the m-valued region is used in order to suppress the generation of small dots in the low / medium density portion and increase the number of dots in the high density portion. And (n-1) quantization thresholds that periodically vary are generated using a plurality of dither threshold matrices each consisting of an n-valued area and the same in the m-valued areas of at least two dither threshold matrices. The position threshold is equal.
Features of the invention of claim 7, wherein, in the image forming method of the invention described in claim 1, in order to prevent white spots reduces the continuous white pixels in the high density portion, and the m-valued area and n value conversion region (N-1) dither threshold matrixes which are arranged and enlarged by shifting a plurality of threshold matrices made of a half-phase in the main scanning direction or the sub-scanning direction. , And the thresholds at the same position in the m-valued region of at least two dither threshold matrices are equal.
The invention according to claim 8 is characterized in that, in the image forming method according to claim 6 or 7 , white spots are prevented in the high density portion and small dots are not generated in the low / medium density portion. In order to improve the stability of the image, the threshold values at the same position in the m-valued region of all the dither threshold value matrices are made equal.
The invention according to claim 9 is characterized in that, in the image forming method according to claim 6, 7 or 8 , the dots are concentrated in the low / medium density portion to improve gradation and stability, and in the high density portion. In order to prevent white spots, the n-valued area of the dither threshold matrix is arranged around the m-valued area.
According to a tenth aspect of the present invention, there is provided an image forming method according to the ninth aspect, wherein the m value of the dither threshold matrix is used in order to increase the stability of the image by suppressing the generation of small dots in the low and medium density portions. In other words, the maximum threshold value in the digitized region is made smaller than the minimum threshold value in the n-valued region.
According to the eleventh aspect of the present invention, in the image forming method according to the tenth aspect of the present invention, gradation is enhanced by increasing the concentration of dots in the low / medium density portion and suppressing the occurrence of small dots. In order to increase the stability, the difference between the minimum threshold value in the n-valued region and the maximum threshold value in the m-valued region of the dither threshold matrix is made larger than the step width of the threshold value in the m-valued region. .
According to a twelfth aspect of the present invention, in the image forming method according to any one of the sixth to eleventh aspects, the probability of occurrence of small dots in the high density portion is increased, white spots are prevented, and gradation is prevented. In order to improve the performance, all threshold values in the n-valued area of the dither threshold matrix for generating the highest quantization threshold value are set to a constant value larger than the threshold values in the n-valued areas of other dither threshold matrixes. It is.
According to a thirteenth aspect of the present invention, in the image forming method according to any one of the sixth to twelfth aspects, each quantization threshold is optimized in order to optimize the quantization threshold according to the image characteristics of the multi-tone image. The dither threshold value matrix used for generating the image is switched according to the image characteristics of the multi-tone image.
According to a fourteenth aspect of the present invention, there is provided an image forming method according to the thirteenth aspect of the present invention, in order to obtain a high resolution by performing error diffusion tone processing in a character portion or a line drawing portion of a multi-tone image. This is because the fluctuation range of the quantization threshold is reduced in the region where the edge degree of the image is large.
According to a fifteenth aspect of the present invention, in the image forming method according to the fourteenth aspect of the present invention, only large dots are generated from the low / medium density portion to the high density portion, and a character portion or line drawing portion of a multi-tone image is generated. In order to obtain a stable image with no white spots, the same dither threshold matrix, which consists of m-valued areas as a whole, is used to generate all quantization thresholds in areas where the edge degree of multi-tone images is large. It is.
According to a sixteenth aspect of the present invention, there is provided the image forming method according to the fourteenth aspect, wherein a binary error diffusion process with a fixed threshold value is performed from a low / medium density portion to a high density portion, so that a character portion of a multi-tone image is obtained. In order to obtain an image with excellent resolution and stability in the line drawing section, etc., all quantization thresholds are fixed to the same constant value in the region where the edge degree of the multi-tone image is maximum. is there.

The image processing apparatus of the present invention enables image processing by the image forming method of the present invention as set forth in claims 1 to 16 , and the main feature thereof is that multi-tone image data as set forth in claim 17. A first means for adding an error to the second data, a second means for quantizing the image data after the error is added by the first means (n ≧ 3), the quantized data by the second means, and the first data A third means for obtaining an error to be added to the multi-tone image data from the image data after the error addition by the means and giving to the first means, and (n-1) pieces which periodically change in the image space. A fourth means for generating and giving the quantization threshold to the second means, wherein at least two quantization thresholds take the same value at a specific position within the fluctuation period.
The feature of the invention according to claim 18 is the image processing device according to claim 17 , wherein at least two quantization threshold values corresponding to the low / medium density levels of the multi-tone image data are in the middle of the fluctuation period. The same value is taken in the part.
According to a nineteenth aspect of the present invention, in the image processing apparatus according to the seventeenth aspect of the present invention, the fourth means changes the fluctuation range of the quantization threshold according to the image characteristics of the multi-tone image data. That is.
Features of the invention of claim 20, wherein, in the image processing apparatus of the invention of claim 19 wherein, as a fourth means, to reduce the fluctuation range of the quantization threshold as a region edge degree of the multi-tone image data is large It is that.
According to a twenty- first aspect of the present invention, in the image processing apparatus according to the nineteenth aspect, the fourth means keeps the average value of the quantization thresholds at a substantially constant value.
According to a twenty-second aspect of the present invention, in the image processing apparatus according to the seventeenth aspect, the fourth means includes (n−1) dither threshold matrixes each including an m-valued area and an n-valued area. Use (n-1) periodically varying quantization thresholds to equalize thresholds at the same position in the m-valued region of the dither threshold matrix to generate at least two quantization thresholds That is.
According to a twenty- third aspect of the present invention, in the image processing apparatus according to the seventeenth aspect, the fourth means includes a plurality of threshold matrixes composed of m-valued areas and n-valued areas, in the main scanning direction or in the sub-scanning direction. Using (n−1) dither threshold matrixes arranged and enlarged by shifting by half a period in the scanning direction, (n−1) periodically varying quantization thresholds are generated, and at least two dithers are generated. The threshold value at the same position in the threshold value matrix m-valued region is the same.
According to a twenty-fourth aspect of the present invention, in the image processing apparatus according to the twenty-second or twenty- third aspect, the thresholds at the same position in the m-valued region of all the dither threshold matrixes are made equal.
The feature of the invention of claim 25 is the image processing apparatus of claim 22 or 23, wherein the n-valued area of the dither threshold matrix is arranged around the m-valued area, and the m-valued area is dot-concentrated. The mold area.
According to a twenty-sixth aspect of the present invention, in the image processing apparatus according to the twenty-fifth aspect, the maximum threshold value in the m-valued area of the dither threshold matrix is made smaller than the minimum threshold value in the n-valued area. It is that.
According to a twenty-seventh aspect of the present invention, in the image processing apparatus according to the twenty-sixth aspect, the difference between the minimum threshold value in the n-valued region and the maximum threshold value in the m-valued region of the dither threshold value matrix is calculated. That is, it is larger than the threshold step width in the m-valued area.
The invention according to claim 28 is the image processing device according to claim 25, 26 or 27 , wherein all threshold values in the n-valued region of the dither threshold value matrix for generating the highest quantization threshold value are set as other values. This is that a constant value larger than the threshold value in the n-valued region of the dither threshold value matrix for generating the quantization threshold value is set.
According to a twenty- ninth aspect of the present invention, in the image processing apparatus according to any one of the twenty-second to twenty-eighth aspects, the image processing apparatus further comprises fifth means for extracting and outputting image characteristics of the multi-tone image data. The fourth means switches the dither threshold matrix used for generating each quantization threshold according to the image feature output from the fifth means.
According to a thirty- third aspect of the present invention, in the image processing apparatus according to the twenty- ninth aspect, the fifth means outputs the edge degree of the multi-tone image data as the image feature, and the fourth means is more than the fifth means. That is, the variation range of the quantization threshold is reduced as the output edge degree is larger.
The feature of the invention described in claim 31 is that, in the image processing apparatus according to claim 30 , when the fourth means has a large degree of edge output from the fifth means, the whole comprises the m-valued region. The dither threshold matrix is used to generate all quantization thresholds.
According to a thirty-second aspect of the present invention, in the image processing apparatus of the thirty- third aspect, when the fourth means has the maximum edge degree output from the fifth means, all the quantization thresholds are set to the same constant value. It is to be fixed to.
The feature of the invention of claim 33 is that, in the image processing device of claim 30, 31 or 32 , the fifth means outputs the edge degree subjected to the area expansion process in order to reproduce the halftone image satisfactorily. This is what I did.
According to a thirty-fourth aspect of the invention, in the image processing apparatus according to the thirty-third aspect, in order to satisfactorily reproduce the halftone dot image having the number of lines used for general printed matter, the extension width of the area extension processing is set to the image. This is to select within 0.5 mm in space.

A feature of the computer-readable storage medium according to the thirty-fifth aspect of the present invention is that any one of the seventeenth to thirty-fourth aspects enables the invention of the first to sixteenth aspects to be easily implemented using a general purpose or dedicated computer. A program for causing a computer to realize the functions of the respective means of the image processing apparatus according to the first aspect is recorded.

The invention of claim 36, 37 or 38, wherein the printer to which the image forming method of the invention of claims 1 to 16, wherein, the display, scanner, fax, for realizing a copying machine, any of claims 17 to 34 An image forming means for forming an image using dots for outputting quantized data to the image processing apparatus according to claim 1, and inputting multi-tone image data by optically scanning a document One or both of the image reading means are further provided.

According to the image forming method and the image processing apparatus of the present invention, the gradation property at the low / medium density portion, the image formation with good stability, the image formation with less white spots at the high density portion, the character portion of the image, High-resolution image formation of a line drawing portion, a relatively low number of halftone dot image portions, and smooth image formation of a joint portion between regions having different image characteristics are possible.

According to the storage medium of the present invention, an image processing apparatus of the present invention, by using a general purpose or special purpose computer can be easily realized, thus, it can be easily carried out the image forming method of the present invention.

According to the image processing apparatus of the present invention, it is possible to printer to which the image forming method of the present invention, a display, a scanner, fax, etc. copier realized.

  Hereinafter, an image processing apparatus according to an embodiment of the present invention will be described with reference to the accompanying drawings. The image forming method of the present invention can be implemented by supplying quantized data from the image processing apparatus of the present invention described below to an appropriate image forming means. In addition, in order to avoid duplication of description, the same reference number is used for the same part or a corresponding part in several drawings in the attached drawings.

<< Example 1 and Modification >>
The image processing apparatus according to the first embodiment has a block configuration as shown in the block diagram of FIG. 1, and an error diffusion process that quantizes multi-tone input image data 101 by an error diffusion method and outputs quantized data 103. And a quantization threshold value generation unit 200 for supplying a quantization threshold value periodically varying in the image space to the error diffusion processing unit 100. In this embodiment and each embodiment described later, the input image data 101 is image data of 256 gradations representing a density level of 0 to 255 at 8 bits / pixel.

  The error diffusion processing unit 100 is supplied from the quantization threshold generation unit 200 with an error addition unit 104 for adding an error to the input image data 101 and the image data 102 with the error added by the error addition unit 104. Quantization unit 105 that quantizes using quantization threshold and outputs quantized data 103, error detection unit 106 that detects an error (quantization error) between quantized data 103 and image data 102, and detected error Error storage unit 107 that temporarily stores the error, and an error to be added to the pixel (target pixel) to be processed next using the error data stored in error storage unit 107 and given to error addition unit 104 The calculation unit 108 is configured.

  In the present embodiment, the quantization threshold value given from the quantization threshold value generation unit 200 to the quantization unit 105 is one, and the quantization unit 105 converts the error-added image data 102 to quantization level 0 or quantization level 1. Binary quantization. Therefore, in the present embodiment, the error detection unit 106 determines that the output value corresponding to the quantization level 0 is 0 (decimal) when the error between the quantization data 103 and the error-added image data 102 is obtained. The output value corresponding to the conversion level 1 is handled as 255 (decimal).

  Further, in this embodiment, as shown in the block of the error calculation unit 108, with the mark * as the position of the target pixel, the pixels at the positions a, b, and c on the previous line and the target pixel are on the same line. A value obtained by dividing a sum of values obtained by multiplying the error detected by the pixel at the immediately previous position d by the coefficients 1, 5, 3, and 7 by 16 is output as an error to be added to the target pixel (*). Therefore, for example, a two-line line memory is used as the error storage unit 107. However, it is possible to change the arrangement and number of pixels around the target pixel to which the error data is referenced in order to calculate the error to be added to the target pixel, and also to change the coefficient (weight) that multiplies the error in each pixel. It is.

  Note that a calculation unit corresponding to the error calculation unit 108 is provided between the error detection unit 106 and the error storage unit 107, and the error detected by the error detection unit 106 and the error data stored in the error storage unit 107 are used. The error to be added to the surrounding unprocessed pixels is sequentially recalculated by the calculation means, and the error data in the error storage unit 107 is updated based on the result, and the error to be added to the next pixel to be processed is updated from the error storage unit 107. A configuration in which the data is directly read and given to the error adding unit 104 can also be adopted.

  In the present embodiment, the quantization threshold value generation unit 200 generates a quantization threshold value that varies in a cycle of 4 pixels in each of the main and sub scanning directions using a 4 × 4 dither threshold value matrix shown in FIG. Such a quantization threshold generation unit 200 generates a read address of the ROM storing the dither threshold matrix shown in FIG. 2 by using a counter that counts the main and sub-scan timing signals of the input image data 101, for example. The threshold value corresponding to the target pixel position of the threshold value matrix can be read from the ROM and output.

  As shown in FIG. 2, the size and arrangement of the threshold values in the dither threshold matrix (within the quantization threshold 4 × 4 pixel fluctuation period) are shown in FIG. The threshold value that sequentially increases up to 104 is a dot-concentrated region in a spiral arrangement, and the surrounding region is a dot-distributed region in which threshold values that increase from 160 to 208 with a step width of 8 are distributed. It is said that. The threshold value in the dot concentration type region is selected to be lower than the threshold value of the dot distribution type region. However, in order to increase the concentration of dots in the low and medium density portions of the image, the minimum threshold value ( 160) and the maximum threshold value (104) of the dot-concentrated area are set to be larger than the step width 8 of the threshold value.

  Since the quantization threshold is generated using such a dither threshold matrix, the influence of error propagation is ignored, and the low density level region of the input image data corresponds to the dot concentration type region within the 4 × 4 pixel period. The 2 × 2 pixels in the center are quantized to the quantization level 1, and the 3 × 3 pixels corresponding to the dot concentration type region are quantized to the quantization level 1 in the medium density level region. In the high density level region, 3 × 3 pixels corresponding to the dot concentration type region are quantized to the quantization level 1, and the flying pixels corresponding to the surrounding dot dispersion type region are also quantized to the quantization level 1. Will come to be.

  Therefore, if the quantized data 103 of the image processing apparatus of the present embodiment is given to a binary image forming apparatus in which only dots of quantization level 1 are hit, dots are formed in the low density portion of the image as shown in FIG. In the middle density portion of the image, dots are generated as shown in FIG. 4, and in the high density portion of the image, dots are generated as shown in FIG. As seen in FIGS. 3 and 4, since dots are concentrated in the low / medium density area, it is less affected by dot gain, has good gradation, and has little banding and uneven density. An image having good properties and stability is formed. Further, as shown in FIG. 5, in the high density portion, dots are generated dispersedly around 3 × 3 dots generated in a concentrated manner, and white pixels are difficult to be continuous, thereby preventing white spots from occurring. it can. In practice, since the dots spread due to the influence of dot gain, white spots in the high density portion become even less noticeable particularly when an electrophotographic printer having a large dot enlargement rate is used. 3 to 5 show the tendency of dot generation. Strictly speaking, dots are not always formed as shown in FIGS. 3 to 5 due to the influence of error propagation.

  Incidentally, when the quantization threshold is generated using the dot concentration type dither threshold matrix as shown in FIG. 28 as a whole, an image with good gradation can be obtained. However, in the high density portion, as shown in FIG. White pixels are continuously generated and are easily recognized as white spots.

  According to the modification of the present embodiment, the quantization threshold value generator 200 uses an 8 × 8 dither threshold value matrix as shown in FIG. 6 to set a quantization threshold value that fluctuates in a period of 8 pixels in the main and sub scanning directions. appear. The dither threshold matrix shown in FIG. 6 has a configuration in which four 4 × 4 dither threshold matrices shown in FIG. 7 are arranged while being shifted by two pixels (half phase) in the sub-scanning direction. In the 4 × 4 dither threshold matrix of FIG. 7, the shaded 2 × 3 region is a dot concentration type in which thresholds increasing in a step width of 8 from 64 to 104 are arranged in a spiral shape, and its peripheral region is from 112 to 184. This is a dot dispersion type in which threshold values increasing with a step width of 8 are dispersedly arranged. A similar dither threshold matrix shifted by half a phase in the main scanning direction may be used.

  However, the quantization threshold value generation unit 200 does not necessarily have such an 8 × 8 dither threshold value matrix. For example, the control has the 4 × 4 dither threshold matrix shown in FIG. 7 and shifts the readout position of the dither threshold matrix in the sub-scanning direction by two pixels (half phase) in a 4-pixel cycle in the main scanning direction. Accordingly, a quantization threshold value substantially similar to that when the dither threshold value matrix of FIG. 6 is used may be generated.

  When an 8 × 8 dither threshold matrix in which four dither threshold matrices shown in FIG. 7 are arranged without shifting in the sub-scanning direction is used for generating a quantization threshold, 2 × as shown in FIG. White pixels in two areas are generated, and white spots are generated. On the other hand, in this modification, the quantization threshold is generated using the dither threshold matrix of FIG. 6, and therefore, as shown in FIG. Is unlikely to occur.

  In this modification as well, since the dots corresponding to the dot concentration type region of the dither threshold matrix are concentrated in the middle / low density area, the gradation and stability in the low / medium density area are good. is there.

<< Example 2 and its modification >>
The image processing apparatus according to the second embodiment has a block configuration as shown in the block diagram of FIG. The difference in the block configuration from the first embodiment is that an image feature extraction unit 300 that extracts image features in the region near the target pixel from the input image data 101 is added, and the output signal 301 is given to the quantization threshold value generation unit 200. In order to adjust the timing between the image feature extraction unit 300 and the error diffusion processing unit 100, a signal delay unit 150 such as a line memory is passed as necessary, and the input image data 101 is converted into the error diffusion processing unit 100. To be input.

  In this embodiment, as shown in FIG. 11, the image feature extraction unit 300 includes an edge detection unit 302 for detecting the degree of edge, detection of periodicity of density level change (in a halftone dot region in a specific line number range). This is a block configuration comprising an area expansion processing unit 303 for identification. The edge detection unit 302 uses, for example, four types of differential filters as shown in FIG. 12 to detect a total of four edge amounts in the main scanning direction, the sub-scanning direction, and the direction inclined ± 45 ° with respect to the main scanning direction. Then, 2-bit edge degree data indicating the edge degree from the level 0 (non-edge) to the level 3 (edge degree maximum), for example, is output as the maximum value (absolute value). The region expansion processing unit 303 performs region expansion on the image space for the edge degree data given from the edge detection unit 302. Specifically, for example, reference is made to edge degree data in a 7 × 7 pixel area around the pixel of interest (a range of 3 pixels before and after the main scanning direction and 3 pixels before and after the sub scanning direction), and the maximum value is noted. The pixel edge degree data is selected and output as an output signal 301.

  In this embodiment, the fluctuation range of the quantization threshold is controlled according to the edge degree level indicated by the output signal 301 of the image feature extraction unit 300. In order to control the fluctuation range of the quantization threshold, the quantization threshold value generator 200 has a block configuration including a dither threshold value generator 201, a multiplier 202, and an adder 203 as shown in FIG.

  The dither threshold value generation unit 201 outputs, for example, a threshold value corresponding to the target pixel position using a 4 × 4 dither threshold value matrix shown in FIG. In this dither threshold matrix, the shaded 3 × 3 region is a dot-concentrated type in which thresholds that sequentially increase from −7 to +1 with a step width of 1 are arranged in a spiral shape, and the peripheral region is from 2 A dot dispersion type in which threshold values increasing in step width 1 up to 8 are dispersedly arranged. The use of the dither threshold matrix in which the dot dispersion type area is arranged around the dot concentration type area as described above in connection with the first embodiment concentrates the dots in the low and medium density portions of the image. This is to improve the gradation and stability by generating the dots and to generate dots by dispersing them around the concentrated dots in the high density portion, thereby making it difficult for white spots to occur. Such a dither threshold value generation unit 201 can be easily performed by, for example, a ROM storing the dither threshold value matrix shown in FIG. 14 and a counter that counts main and sub-scan timing signals of image data to generate a read address of the ROM. realizable.

  The multiplication unit 202 outputs a value obtained by multiplying the threshold value output from the dither threshold value generation unit 201 by a coefficient corresponding to the edge degree level indicated by the output signal 301 of the image feature extraction unit 300. In this embodiment, the coefficient at level 0 (non-edge) is 8, the coefficient at level 1 is 4, the coefficient at level 2 is 2, and the coefficient at level 3 (maximum edge degree) is 0. is there. Therefore, the output value of the multiplication unit 202 fluctuates in the maximum range from −56 to +64 when the edge degree is level 0 (non-edge), the fluctuation range decreases at the edge degree level 1, and the edge degree level 2 The fluctuation range further decreases. When the edge degree is 3, the fluctuation range becomes zero.

  The adder 203 adds +128 corresponding to the center level of the image data to the output value of the multiplier 202. The output value of the adder 203 is given to the quantizer 105 as a quantization threshold. Accordingly, when the edge degree level is 0, the quantization threshold varies with a maximum fluctuation range from +72 to +192 around +128, and when the edge degree level is 3, the quantization threshold is fixed to +128. As described above, the fluctuation range of the quantization threshold is controlled to decrease as the edge degree level increases.

  Therefore, in an image region having a high edge degree level such as a character part or a line drawing part of an image, the quantization threshold is fixed to +128 or can be changed with a small fluctuation range. The threshold binary error diffusion process or the error diffusion keynote process is performed, and an image with good resolution can be formed. On the other hand, in an image region with a low edge degree level such as a flat portion of the image, the quantization threshold is changed with a large fluctuation range centering on +128 according to the dither threshold matrix of FIG. In the same manner as in the first embodiment, dots are concentrated in the low / medium density portion, and an image having good gradation and stability can be formed, and concentrated in the high density portion. Since dots are generated by being dispersed around the generated dots, it is possible to form an image in which white spots are unlikely to occur. In addition, the quantization threshold varies periodically, and the fluctuation range also changes according to the edge degree level. However, since the time average of the quantization threshold is maintained at a substantially constant value (+128), the quantization threshold is switched. In other words, in other words, it is possible to prevent a delay in dot generation at the boundary between the image areas having different image characteristics, and thus the boundary between the image areas can be expressed smoothly.

  When the input image data 101 is read from a document at a resolution of 600 dpi, the expanded width of 7 pixels in the area expansion process is about 0.3 mm on the document, which corresponds to a halftone dot period of about 86 Lpi. Therefore, the halftone dot image portion having a line number of 86 Lpi or more is evaluated as an edge, and the error diffusion processing unit 100 uses the fixed quantization threshold value or the quantization threshold value with a small fluctuation width to obtain an error diffusion process with good resolution or Since error diffusion tone processing is performed, halftone dots can be faithfully reproduced with high resolution, and generation of moire can be prevented. In the halftone dot image portion having a lower number of lines than that, the inside of each halftone dot is evaluated as a non-edge, and, similarly to the flat portion of the image, a dither tone process is performed using a quantization threshold of a large vibration width, Stability is good and white spots are less likely to occur in high density areas. It should be noted that such a low-line-number halftone dot image does not have image resolution information inside each halftone dot, so there is no inconvenience in the dither tone processing. As described above, according to this embodiment, it is possible to reproduce a high-quality image including characters, line drawings, photographs, halftone dots, and the like.

  It is the image change point that affects the sharpness of the image. Generally, it is sufficient in terms of image quality if a halftone dot having a relatively low number of lines up to about 50 Lpi can be faithfully reproduced. Therefore, strictly speaking, it is necessary to consider the influence of the MTF characteristics of the scanner used for image reading, the characteristics of the edge detection filter, the period difference caused by the density change of the halftone image, and the like. If it is selected within 0.5 mm in space (within 12 pixels at 600 dpi), a halftone image can generally be reproduced with sufficient image quality. In addition, image data scanned by scanning a document with a scanner is generally passed through a smoothing filter in order to express halftones smoothly. Usually, the image data is smoothed from about 150 Lpi, and from 175 Lpi. The periodic amplitude of halftone dots having a higher number of lines than about 200 Lpi does not remain. Accordingly, such a high-line-number halftone dot image portion is determined to be a non-edge portion by the image feature extraction unit 300, and there is no concern that moire will occur even if the dither tone processing is performed in the same manner as the flat portion of the image. Absent.

  In this embodiment, the halftone dot image area in the specific line number range is identified by the edge degree data area expansion process, but the halftone dot image area in the specific line number range is directly detected from the input image data 101. A means is provided separately, and the quantization threshold generation section 200 generates a fixed or small fluctuation threshold for the halftone dot image area detected by the means regardless of the edge degree detected by the edge detection section 302. Thus, the error diffusion keynote process may be performed.

  According to the modification a of the present embodiment, the dither threshold value generation unit 201 uses a dither threshold value matrix as shown in FIG. This dither threshold matrix is a dot distribution in which thresholds increasing by 1 from 3 to 9 are distributed around a shaded dot-concentrated area in which thresholds increasing by 1 from -8 to 0 are spirally arranged. It consists of a mold area. The difference between the minimum threshold value (3) in the dot dispersion type region and the maximum threshold value (0) in the dot concentration type region is selected to be 3 which is larger than the threshold step width (1) in the dot concentration type region. ing. When such a dither threshold matrix is used, the quantization threshold varies in the range of 64 to 200 in the flat image portion, but the maximum quantization threshold (128) of the dot concentration type region and the minimum of the dot dispersion type region. Since the difference from the quantization threshold (152) is as large as 24, the probability that dots are generated in the dot dispersion type region at low / medium density is reduced, the concentration of dots is increased, and the gradation of the low / medium density portion is increased. And stability are improved.

  Although not shown, according to another modification b of the present embodiment, the quantization threshold value generator 200 sets each threshold value of the dither threshold value matrix shown in FIG. 14 or 15 by 8 times, 4 times, 2 times, There are four types of 4 × 4 dither threshold matrixes for edge degree levels 0, 1, 2, and 4 obtained by multiplying by 0 and adding 128, and the edge degree indicated by the output signal 301 of the image feature extraction unit 300 among them. The threshold value of the dither threshold value matrix selected according to the level is output as the quantization threshold value. According to such a configuration, the multiplication unit 202 and the addition unit 203 in the configuration shown in FIG. 13 are not required, so that the hardware can be easily realized, and the multiplication that requires a processing time is not required even in the case of software. There is an advantage to become. It is obvious that the same processing as in the present embodiment or the modification a can be performed by the modification b. Further, the average value of the quantization threshold is kept at a substantially constant value (+128).

  Although not shown, according to another modification c of the present embodiment, in the present embodiment or each of the embodiments a and b thereof, the area expansion processing unit 303 is omitted from the image feature extraction unit 300, and the edge detection unit The edge degree data output from 302 is output as the output signal 301 as it is. According to this modification, it is possible to form a high-resolution image in a region where the density changes drastically, such as image characters and line drawings, and the gradation and stability of the flat portion of the image are good, and there is no white spot. An image can be formed.

  Also in this embodiment and its modifications a, b, and c, an enlargement in which a plurality of dither threshold matrices as shown in FIG. 14 or FIG. 15 are arranged while being shifted by a half phase in the sub-scanning method or main scanning direction. A dither threshold matrix may be used for generating quantization thresholds.

<< Example 3 and its modification >>
The present embodiment 3 and the following embodiment 4 are related to the present invention. The image processing apparatus according to the third embodiment has a block configuration as shown in the block diagram of FIG. 1 as in the first embodiment, but differs from the first embodiment in the following points.

  One difference is that the quantization threshold value generator 200 generates two quantization threshold values Thr1 and Thr2 (0 <Thr1 ≦ Thr2 <255). Another difference is that the quantization unit 105 quantizes the post-error-added image data 102 when Thr1 <Thr2, but if Thr1 = Thr2, the post-error-added image data 102 is binary quantized. It is to become. That is, if the level of the error-added image data 102 is L, if Thr <Thr2, the quantization level is 0 when L <Thr1, the quantization level is 1 when Thr1 ≦ L <Thr2, and L ≧ Thr2. Quantization level 2 is output respectively. On the other hand, when Thr1 = Thr2, the quantization level 0 is output when L <Thr1 (= Thr2), and the quantization level 2 is output when L ≧ Thr2 (= Thr1). Another difference is that when the error detection unit 106 obtains an error between the quantized data 103 and the error added image data 102, the output value corresponding to the quantization level 0 is set to 0 (decimal), and the quantization level is set. For example, the output value corresponding to 1 is treated as 128 (decimal), and the output value corresponding to the quantization level 2 is treated as 255 (decimal).

  The quantization threshold value generator 200 generates the quantization threshold values Thr1 and Thr2 using different dither threshold matrices. An example of two such dither threshold matrices is shown in FIG. The dither threshold matrixes shown in (a) and (c) are for occurrence of Thr1, and the dither threshold matrixes shown in (b) and (c) are for occurrence of Thr2. The dither threshold matrix pair of (a) and (b) may be used, or the dither threshold matrix pair of (c) and (d) may be used.

  Both dither threshold matrices are composed of a shaded 3 × 3 binarized area and its surrounding ternary areas. In the binarized areas of both dither threshold matrices, the threshold values of the corresponding positions are the same value. That is, the pixels corresponding to the binarized area are binarized and quantized. Also, the binarized area is a dot-concentrated area in which thresholds that increase by 8 from 42 to 106 are arranged in a spiral shape, but this is to concentrate dots in the low and medium density portions of the image. It is.

  In the ternary area around the binarized area, the threshold values for Thr1 and Thr2 are different, and the threshold value for Thr2 is selected to be larger than the threshold value for Thr1. Therefore, in the pixel corresponding to the ternary region, Thr1 <Thr2, and ternary quantization is performed. Further, the three-valued region of the Thr1 dither threshold matrix is a dot concentration type in which thresholds increasing in increments of 8 from 156 to 204 are sequentially arranged in the dither threshold matrix of (a). The threshold matrix is a dot dispersion type in which similar thresholds are dispersedly arranged. The three-valued region of the Thr2 dither threshold matrix is also a dot-concentrated type in which thresholds increasing in increments of 8 from 192 to 240 are sequentially arranged in the dither threshold matrix in (b), but the dither threshold matrix in (d). In the dot dispersion type, similar threshold values are dispersedly arranged. Further, in order to increase the concentration of dots, the difference between the minimum threshold value in the ternary area and the maximum threshold value in the binarization area is selected to be larger than the step width of the threshold value in the binarization area. .

  According to the image processing apparatus of this embodiment, since the binary quantization is performed in the low / medium density portion of the image, no small dots are generated, and since the binarized area is a dot concentrated area, the large dots Therefore, an image with good gradation and stability can be formed. Further, since the three-level quantization is performed in the high density portion of the image, the probability of occurrence of small dots is high and white spots are unlikely to occur.

  In order to clarify such advantages, the quantized data 103 of the image processing apparatus according to the present embodiment is supplied to, for example, an electrophotographic printer that forms an image using small dots and large dots to form an image. FIG. 17 shows how the dots are generated in the case of occurrence. As described above, the threshold value in the binarized region of the dither threshold matrix for generating Thr1 and Thr2 is lower than the threshold value in the ternary region, and the binarized region is a dot concentration type region. Therefore, large dots corresponding to the quantization level 2 are intensively generated in a spiral shape in the low density part and the middle density part of the image, as shown in FIGS. 17A and 17B, respectively. A stable and stable image is formed. Since the ternary region of the dither threshold matrix is arranged around the binarized region, the quantization level 1 is set around the concentrated large dots in the high density portion as shown in FIG. Corresponding small dots are generated, and white spots are hardly generated.

  According to the modification a of the present embodiment, the quantization threshold value generator 200 uses the dither threshold value matrix pairs shown in FIGS. 18 to 20 instead of the dither threshold value matrix pairs shown in FIG. These dither threshold matrices differ from the dither threshold matrix of FIG. 16 only in the positional relationship between the binarized area (dot concentrated area) and the ternary area.

  According to another modification b of the present embodiment, the dither threshold value matrix pairs shown in FIGS. 21 to 24 are used for generating the quantization threshold values Thr1 and Thr2. In these dither threshold matrix pairs, all threshold values in the ternary area of the dither threshold matrix for Thr2 are selected to be a relatively low constant value (216). When such a dither threshold matrix is used, in the ternary region, since all small dots are generated and then the tendency to generate large dots becomes strong, the probability of generating white pixels is further reduced, and more reliable white spots are generated. Prevention is possible.

The number of pixels in the binarized area and the ternary area of the dither threshold matrix can be uniquely determined as follows based on the lowest density level at which white pixels are not desired to be generated. For example, it is assumed that white pixels are not desired to be generated at a level 200 or higher of multi-tone image data after gamma conversion. If the number of pixels in the binarized area in the 4 × 4 dither threshold matrix is X, the number of pixels in the binarized area is (16−X). As described above, if the output value of a small dot (quantization level 1) is 128 and the output value of a large dot (quantization level 2) is 255,
(255 * X / 16) + (128 * (16-X) / 16) = 200
In this case, X = 9.1. Therefore, if the binarized area is 9 pixels and the ternary area is 7 pixels, white pixels are not generated at the image data level 200 or higher.

  According to another modification c of the present embodiment, a plurality of 4 × 4 dither threshold matrixes as shown in FIG. 16, FIG. 18 to FIG. 20, or FIG. 22 to FIG. An enlarged dither threshold matrix arranged in a shifted direction is used for generating the quantization thresholds Thr1 and Thr2. Examples of such a dither threshold matrix are shown in FIGS. The dither threshold value matrix for Thr1 shown in FIG. 25 corresponds to the four 4 × 4 dither threshold value matrixes shown in FIG. 16A that are shifted by a half phase (2 pixels) in the sub-scanning direction. The dither threshold value matrix for Thr2 shown in FIG. 26 corresponds to the four 4 × 4 dither threshold value matrixes shown in FIG. 16B, which are shifted by half a phase in the sub-scanning direction. Such a dither threshold matrix is effective in preventing white spots because continuation of the ternary regions can be suppressed. Note that the quantization threshold value generation unit 200 does not necessarily have such an 8 × 8 dither threshold matrix, and has, for example, the 4 × 4 dither threshold matrix shown in FIGS. By performing a control to shift the readout position of the dither threshold matrix in the sub-scanning direction by 2 pixels at a four-pixel period in the direction, the quantization is substantially the same as when the dither threshold matrix of FIGS. 25 and 26 is used. A threshold value may be generated.

  In the present embodiment and its modifications a, b, and c, the high density portion is quantified in three values and the low and medium density portions are binary quantized. However, the number of quantizations is not limited thereto. For example, three region threshold values Thr1... Are obtained by using three dither threshold value matrices in which a four-valued region having a larger threshold value is arranged around the binarized region. By generating Thr2 and Thr3 and setting the threshold value at the same position in the binarized region of the dither threshold value matrix for the quantization threshold values Thr1, Thr2 and Thr3 to binary values, the binary value is obtained at the low / medium density portion of the image. It is also possible to quantize and perform quaternary quantization in the high density part. Further, for example, four quantization threshold values Thr1, Thr2, Thr3, Thr4 are generated using four dither threshold matrixes in which a quinary region having a larger threshold value is arranged around the ternary region, and Thr1, The threshold value at the same position in the three-valued region of the dither threshold value matrix for Thr2 is set to an equal value, and the threshold value at the same position in the three-valued region of the dither threshold value matrix for Thr3 and Thr4 is set to an equal value. Thus, it is possible to perform ternary quantization in the low / medium concentration portion and quinary quantization in the high concentration portion.

  More generally, (n−1) quantization thresholds are generated using (n−1) dither threshold matrixes consisting of m-valued regions and n-valued regions (n> m), and In the high density area, n-value quantization increases the probability of small dots and prevents white spots. On the other hand, in low and medium density areas, m-value quantization reduces the occurrence of small dots and stabilizes the image. Can increase the sex. The same applies to the fourth embodiment described below and its modifications.

<< Example 4 and its modification >>
The image processing apparatus according to the fourth embodiment has the block configuration shown in FIG. The image feature extraction unit 300 has the same configuration as that shown in FIG. The quantization threshold value generation unit 200 generates two quantization threshold values Thr1 and Thr2 as in the third embodiment. The dither threshold value matrix used for generating the quantization threshold values is output from the image feature extraction unit 300. The point of switching according to the edge degree level indicated by the signal 301 is different from the third embodiment.

  As in the third embodiment, the quantization unit 105 of the error diffusion processing unit 100 quantizes the post-error addition image data 102 to levels 0, 1, and 2 when the quantization threshold values Thr1 and Thr2 are different, If the quantization thresholds Thr1 and thr2 are equal, the post-error addition image data 102 is binary quantized to levels 0 and 2.

  FIG. 27 shows an example of a dither threshold value matrix used at each edge degree level in the quantization threshold value generation unit 200. When the edge level is 0, a pair of dither threshold matrices having a large threshold fluctuation range shown in the uppermost part of FIG. 27 is used. The dither threshold matrix pair is the same as the dither threshold matrix pair of FIG. 21 used in the modification of the third embodiment.

  In order to generate Thr1 and Thr2 at the edge level 1, the dither threshold matrix pair shown in the second row of FIG. 27 is used. These dither threshold matrices also have a structure in which ternary regions are arranged around a 3 × 3 binarized region, but the threshold values of each region are different from the dither threshold value matrix for edge degree level 0. That is, the minimum threshold value in the binarized area is selected as 86, which is larger than that at the edge degree level 0, and the step width is selected as 4, which is smaller than that at the edge degree level 0. In the ternary region of the Thr1 dither threshold matrix, the minimum threshold value is selected to be 138 smaller than the edge degree level 0, and the step width is selected to be 4 smaller than the edge degree level 0. . Further, the threshold value of the ternary region of the Thr2 dither threshold value matrix is selected to be a constant value (172) smaller than the edge degree level 0.

  In order to generate Thr1 and Thr2 at the edge level 2, the dither threshold matrix pair shown in the third row of FIG. 27 is used. These dither threshold matrices are entirely binarized regions, and the minimum value of the threshold is selected to be a larger value (114) than that at the edge degree level 1, and the step width is also smaller than that at the edge degree level 1 ( 2).

  In order to generate Thr1 and Thr2 at the edge degree level 3, a pair of dither threshold matrixes shown at the bottom of FIG. 27 is used. These dither threshold matrixes are also made as binarized areas as a whole, but the threshold values are all selected to be the same value (128).

  Since such a dither threshold matrix pair is used by switching according to the edge degree level, it is possible to perform an appropriate process according to the feature of the image. That is, in a character part or a line drawing part having a high edge degree level, a normal binary error diffusion process or a binary error diffusion tone process is performed using a fixed threshold value or a quantization threshold value having a small vibration width, so that sharpness is excellent. Reproducible images. A halftone dot region that is evaluated as an edge portion by the image feature extraction unit 300 and that has a relatively low number of lines greater than a certain number of lines is processed in the same manner, so that the halftone dots can be faithfully reproduced. On the other hand, a halftone dot region having a lower number of lines than the number of lines and a halftone dot region having a high number of lines whose periodicity is removed by a smoothing filter are subjected to dither tone processing in the same manner as a flat portion of an image. Therefore, it is possible to form a stable image in which white spots do not occur easily.

  Note that the positional relationship between the binarized area and the ternary area of the dither threshold matrix shown in FIG. 27 may be changed. Further, as described in the above embodiments, a dither threshold matrix in which a plurality of dither threshold matrices as shown in FIG. 27 are arranged in the main scanning direction or the sub-scanning direction may be used.

  Alternatively, the region extension processing unit 303 may be omitted from the image feature extraction unit 300 and the edge degree data of the edge detection unit 302 may be output as the output signal 301 as it is. Even in this case, as in the case of the present embodiment, it is possible to clearly reproduce an image area having a large density change such as a character portion or a line drawing portion of an image and to reproduce a flat portion such as a photograph or an image with high quality. Can do.

<< Implementation of the present invention by software >>
The image processing apparatus according to each of the embodiments and the modifications thereof may be realized by software using a general purpose or dedicated computer. In this case, a program for realizing the function of each unit of the image processing apparatus on a computer is read from various storage media such as a floppy disk, an optical disk, a magneto-optical disk, and a semiconductor storage element on which the program is recorded, or The image processing apparatus of the present invention can be realized on a computer by receiving it from an external computer or the like via a network, loading it into the main memory of the computer, and causing the CPU to execute it. For example, a main memory is used as a storage area such as a line memory necessary for storing various data. Various computer-readable storage media on which such programs are recorded are also included in the present invention.

<< Other Embodiments >>
The image processing apparatus according to each of the embodiments and modifications thereof is related to an image forming apparatus such as a printer or a display, an apparatus related to image reading such as a scanner or a fax machine, and both image reading and image forming. Can be incorporated into a device such as a digital copier. As an example of such an embodiment, an example of a digital copying machine to which the present invention is applied will be described below.

  FIG. 30 is a schematic cross-sectional view for explaining the structure of an example of a digital copying machine according to the present invention. This digital copying machine includes an image reading unit 400 that optically scans and reads a document, a laser printer 411 as an electrophotographic image forming unit, and a circuit unit (not shown) (see FIGS. 31 and 32). Composed.

  The image reading unit 400 illuminates a document placed on a flat document table 403 with an illumination lamp 502, and forms a reflected light image on an image sensor 507 such as a CCD via mirrors 503 to 505 and a lens 506. At the same time, the document is sub-scanned by the movement of the illumination lamp 502 and the mirrors 503 to 505, whereby the image information of the document is read and converted into an electrical image signal. An analog image signal output from the image sensor 507 is input to a circuit unit (not shown) and processed. The image data output from the circuit unit 550 is input to the laser printer 411 to form an image.

  In the laser printer 411, the writing optical unit 508 converts image data input from a circuit unit (not shown) into an optical signal, and exposes an image carrier made of a photosensitive member, for example, a photosensitive drum 509, whereby an original document is obtained. An electrostatic latent image corresponding to the image is formed. For example, the writing optical unit 508 emits a laser beam whose intensity is modulated by driving a semiconductor laser by the light emission drive control unit with the image data, and deflecting and scanning the laser beam by the rotary polygon mirror 510, and an f / θ lens. Then, the photosensitive drum 509 is irradiated through the reflection mirror 511. The photosensitive drum 509 is driven to rotate by a drive unit and rotates clockwise as indicated by an arrow, and is uniformly charged by a charger 512 and then exposed by a writing optical unit 508 to form an electrostatic latent image. Is done. The electrostatic latent image on the photosensitive drum 509 is developed by the developing device 513 to become a toner image. Further, the sheet is fed to the registration roller 520 from any of the plurality of sheet feeding units 514 to 518 and the manual sheet feeding unit 519. The registration roller 520 sends a sheet to the toner image on the photosensitive drum 509 in time. The transfer belt 521 is applied with a transfer bias from a transfer power source, transfers the toner image on the photosensitive drum 509 to a sheet, and conveys the sheet. The sheet onto which the toner image has been transferred is conveyed to the fixing unit 522 by the transfer belt 521 and the toner image is fixed, and then is discharged to the paper discharge tray 523. Further, the photosensitive drum 509 is cleaned by the cleaning device 524 after the toner image is transferred, and is further discharged by the charge eliminator 525 to prepare for the next image forming operation.

  FIG. 31 is a block diagram schematically showing an example of a circuit unit inside the digital copying machine. The input of this circuit unit is an analog image signal read at, for example, 600 dpi by the image sensor 507 of the image reading unit 400. This analog image signal is adjusted in level by the AGC circuit 551 and then converted into 8-bit digital image data per pixel by the A / D conversion circuit 552. Further, the shading correction circuit 553 further converts the analog image signal for each pixel of the image sensor 507. Variations in sensitivity and illuminance are corrected. Next, the image data is sent to a filter processing circuit 554, for example, subjected to MTF correction, and then subjected to smoothing filter processing for smoothly expressing a halftone image. The image data that has been subjected to such processing is sent to the gamma correction circuit 555, where it is subjected to γ correction for conversion to writing density. The image data after γ correction is input to the halftone processing unit 556. As the halftone processing unit 556, the image processing apparatus according to each of the above-described embodiments or modifications thereof is used. Output image data of the image processing apparatus is sent to the light emission drive control unit of the semiconductor laser in the writing optical unit 508. Since the halftone processing unit 556 performs the above-described quantization processing, an image read from the document can be reproduced with high image quality.

  When the image processing apparatus according to the second or fourth embodiment or the modified example is used as the halftone processing unit 556, the image feature extraction unit 300 can be provided in the halftone processing unit 556, but from the halftone processing unit 556. It can also be provided in the previous stage. For example, the image feature extraction unit 300 may be provided at a position as shown in FIG.

  It should be noted that in a digital copying machine, in practice, scaling processing, background removal processing, flare removal processing, and other image editing processing on image data are often possible, but the description thereof is omitted. Further, the present invention can be similarly applied to a digital copying machine having an image reading unit that moves an original table and a digital copying machine having image forming means other than a laser printer.

  As described above, the present invention can be applied to various devices such as a scanner and a fax machine. In this case, although not shown, a reading means such as the image reading unit 400 of the digital copying machine for inputting image data to the image processing apparatus according to each of the embodiments or modifications thereof, an image of the quantized data. One or both of image forming means such as the laser printer 411 of the digital copying machine are added for forming.

It is a block diagram which shows an example of the block configuration of the image processing apparatus by this invention. It is a figure which shows the dither threshold value matrix used in Example 1 of this invention. It is a figure which shows the mode of the dot generation | occurrence | production in the low concentration part in Example 1 of this invention. It is a figure which shows the mode of the dot generation | occurrence | production in the middle density part in Example 1 of this invention. It is a figure which shows the mode of the dot generation | occurrence | production in the high concentration part in Example 1 of this invention. It is a figure which shows the shift type expansion dither threshold value matrix used in the modification of Example 1 of this invention. FIG. 7 is a diagram showing a dither threshold matrix that is a basis of the shift-type enlarged dither threshold matrix shown in FIG. 6. FIG. 7 is a diagram showing how dots are generated in a high density portion when the shift type enlarged dither threshold matrix shown in FIG. 6 is used. It is a figure which shows the mode of the dot generation | occurrence | production in a high density part at the time of using a non-shift type expansion dither threshold value matrix. It is a block diagram which shows another example of the block configuration of the image processing apparatus by this invention. It is a block diagram which shows an example of an image feature extraction part. It is a figure which shows the example of the differential filter for edge detection. It is a block diagram which shows an example of a quantization threshold value generation part. It is a figure which shows the dither threshold value matrix used in Example 2 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 2 of this invention. It is a figure which shows the dither threshold value matrix used in Example 3 of this invention. It is a figure which shows the mode of the dot generation | occurrence | production in the low density part in Example 3 of this invention, a medium density part, and a high density part. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the dither threshold value matrix used in the modification of Example 3 of this invention. It is a figure which shows the shift-type expansion dither threshold value matrix used for Thr1 generation | occurrence | production in the modification of Example 3 of this invention. It is a figure which shows the shift type expansion dither threshold value matrix used for Thr2 generation | occurrence | production in the modification of Example 3 of this invention. It is a figure which shows the pair of the dither threshold value matrix for generation | occurrence | production of Thr1 and Thr2 of the fluctuation range corresponding to an edge degree level in Example 4 of this invention. It is a figure which shows the dither threshold value matrix of the dot concentration type as a whole. It is a figure which shows the mode of the dot generation | occurrence | production at the time of using the dither threshold value matrix of FIG. 1 is a schematic cross-sectional view showing a schematic configuration of an example of a digital copying machine according to the present invention. FIG. 32 is a block diagram illustrating an example of a block configuration of a circuit unit of the digital copying machine of FIG. 30. FIG. 32 is a block diagram illustrating another example of a block configuration of a circuit unit of the digital copying machine of FIG. 30.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Error diffusion process part 101 Input image data 102 Image data after error addition 103 Quantization data 104 Error addition part 105 Quantization part 106 Error detection part 107 Error storage part 108 Error calculation part 150 Signal delay part 200 Quantization threshold value generation part 201 Dither threshold value generation unit 202 Multiplication unit 203 Addition unit 300 Image feature extraction unit 302 Edge detection unit 303 Area expansion processing unit 400 Image reading unit 403 Document table 411 Laser printer 502 Illumination lamps 503, 504, 505 Mirror 506 Lens 507 Image sensor 508 Writing Optical unit 509 Photosensitive drum 510 Rotating polygon mirror 511 f / θ lens and reflection mirror 512 Charger 513 Developing device 514 to 518 Paper feeding unit 519 Manual paper feeding unit 520 Registrho et al. 521 Transfer bell 522 fixing unit 523 discharge tray 524 cleaning device 525 discharger 551 AGC circuit 552 A / D converter circuit 553 the shading correction circuit 554 filter circuit 555 gamma correction circuit 556 halftone processing section

Claims (38)

In an image forming method in which a multi-tone image is quantized by an n-value (n ≧ 3) using an error diffusion method, and the pixels of the quantized image are represented by larger dots as the quantization level is higher. An image forming method for quantizing m values (m <n) in a density portion ,
For quantization of a multi-tone image, (n−1) quantization thresholds periodically changing in the image space are used, and at least two quantization thresholds have the same value at specific positions within the change period. And an image forming method. The image forming method according to claim 1, wherein the specific position is a central portion within a fluctuation cycle of a quantization threshold. 2. The image forming method according to claim 1, wherein the fluctuation range of the quantization threshold is changed in accordance with the image characteristics of the multi-tone image. 4. The image forming method according to claim 3, wherein the fluctuation range of the quantization threshold is reduced in a region where the edge degree of the multi-tone image is large. 4. The image forming method according to claim 3, wherein an average value of the quantization threshold value is maintained at a substantially constant value. Using (n−1) dither threshold matrixes composed of m-valued areas and n-valued areas, periodically (n−1) quantized threshold values are generated, and at least two dithers are generated. 2. The image forming method according to claim 1, wherein the threshold values at the same position in the m-valued area of the threshold value matrix are equal. A cycle using a plurality of (n−1) dither threshold matrixes arranged and enlarged by shifting a half phase in the main scanning direction or the sub-scanning direction by a plurality of threshold value matrices composed of m-valued regions and n-valued regions. 2. The image forming method according to claim 1, wherein (n-1) quantization threshold values that vary in number are generated, and threshold values at the same position in an m-valued region of at least two dither threshold value matrices are equal. Method. 8. The image forming method according to claim 6, wherein the threshold values at the same position in the m-valued area of all dither threshold value matrices are equal. 9. The image forming method according to claim 6, wherein the n-valued area of the dither threshold matrix is arranged around the m-valued area, and the m-valued area is a dot concentrated area. 10. The image forming method according to claim 9, wherein the maximum threshold value in the m-valued area of the dither threshold value matrix is made smaller than the minimum threshold value in the n-valued area. The difference between the minimum threshold value in the n-valued region and the maximum threshold value in the m-valued region of the dither threshold matrix is larger than the step width of the threshold value in the m-valued region. Image forming method. All the threshold values in the n-valued area of the dither threshold matrix for generating the highest quantization threshold value are set to a constant value larger than the threshold values in the n-valued areas of other dither threshold matrixes. Item 12. The image forming method according to any one of Items 6 to 11. 13. The image forming method according to claim 6, wherein a dither threshold matrix used for generating each quantization threshold is switched according to image characteristics of a multi-tone image. 14. The image forming method according to claim 13, wherein the fluctuation range of the quantization threshold is reduced in a region where the edge degree of the multi-tone image is large. 15. The image forming method according to claim 14, wherein the same dither threshold value matrix consisting of m-valued regions as a whole is used for generating all quantization threshold values in a region where the edge degree of a multi-tone image is large. 15. The image forming method according to claim 14, wherein all quantization thresholds are fixed to the same constant value in a region where the edge degree of the multi-tone image is maximum. A first means for adding an error to the multi-tone image data; a second means for quantizing the image data after the error is added by the first means; and an quantization by the second means. A third means for obtaining an error to be added to the multi-tone image data from the data and the image data after the error addition by the first means and giving the error to the first means, and periodically fluctuates in the image space ( and n-1) a fourth means for generating and giving to the second means, at least two quantization thresholds take the same value at a specific position within the fluctuation period, An image processing apparatus. The image processing apparatus according to claim 17, wherein the at least two quantization threshold values have the same value in a central portion within the fluctuation period. 18. The image processing apparatus according to claim 17, wherein the fourth means changes a fluctuation range of the quantization threshold according to the image characteristics of the multi-tone image data. The image processing apparatus according to claim 19, wherein the fourth means reduces the fluctuation range of the quantization threshold in a region where the edge degree of the multi-tone image data is larger. 20. The image processing apparatus according to claim 19, wherein the fourth means maintains an average value of the quantization threshold value at a substantially constant value. The fourth means generates (n−1) periodically varying quantization thresholds using (n−1) dither threshold matrixes composed of m-valued regions and n-valued regions, 18. The image processing apparatus according to claim 17, wherein the thresholds at the same position in the m-valued region of at least two dither threshold matrices are equal. The fourth means includes (n−1) dithers in which a plurality of threshold matrixes each including an m-valued area and an n-valued area are arranged and enlarged by being shifted by a half cycle in the main scanning direction or the sub-scanning direction. The threshold matrix is used to generate (n-1) periodically varying quantization thresholds, and the thresholds at the same position in the m-valued region of at least two dither threshold matrices are equal. Item 17. The image processing apparatus according to Item 17. 24. The image processing apparatus according to 22 or 23, wherein the threshold values at the same position in the m-valued region of all dither threshold value matrices are equal. The image processing apparatus according to claim 22 or 23, wherein the n-valued area of the dither threshold matrix is arranged around the m-valued area, and the m-valued area is a dot concentrated area. 26. The image processing apparatus according to claim 25, wherein the maximum threshold value in the m-valued area of the dither threshold value matrix is made smaller than the minimum threshold value in the n-valued area. 27. The difference between the minimum threshold value in the n-valued region and the maximum threshold value in the m-valued region of the dither threshold matrix is greater than the step width of the threshold value in the m-valued region. Image processing device. All thresholds in the n-valued region of the dither threshold matrix for generating the highest quantization threshold are constant values greater than the threshold values in the n-valued region of the dither threshold matrix for generating other quantization thresholds. 28. The image processing apparatus according to claim 25, 26 or 27, wherein The apparatus further comprises fifth means for extracting and outputting image features of the multi-tone image data, and the fourth means is used to generate each quantization threshold according to the image features output from the fifth means. The image processing apparatus according to any one of claims 22 to 28, wherein the dither threshold value matrix is switched. The fifth means outputs the edge degree of the multi-tone image data as an image feature, and the fourth means reduces the fluctuation range of the quantization threshold as the edge degree outputted from the fifth means is larger. 30. The image processing apparatus according to claim 29, wherein: The fourth means uses the same dither threshold matrix composed entirely of m-valued regions for generating all quantization thresholds when the degree of edge output from the fifth means is large. 30. The image processing apparatus according to 30. 31. The image processing apparatus according to claim 30, wherein the fourth means fixes all quantization thresholds to the same constant value when the edge degree output from the fifth means is maximum. 33. The image processing apparatus according to claim 30, 31 or 32, wherein the fifth means outputs the degree of edge subjected to the area expansion process. 34. The image processing apparatus according to claim 33, wherein an expansion width of the region expansion processing is selected within 0.5 mm on the image space. 35. A computer-readable storage medium on which a program for causing a computer to realize the functions of the respective units of the image processing apparatus according to claim 17 is recorded. 35. The image processing apparatus according to claim 17, further comprising image forming means for forming an image using dots for outputting an image of quantized data. 35. The image processing apparatus according to claim 17, further comprising image reading means for inputting multi-gradation image data by optically scanning a document. An image reading unit that inputs multi-tone image data by optically scanning a document, and an image forming unit that forms an image using dots for outputting an image of quantized data. 35. The image processing apparatus according to any one of claims 17 to 34.

Images