JP2017135569A - Image processing device, image forming apparatus and threshold matrix shaping method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a threshold matrix in which connections of dots are reduced and an image quality after screen processing is improved.SOLUTION: An image processing device comprises: a probability conversion part for converting an initial pattern into a probability distribution space; a repetitive processing part for weighting a density distribution space with the probability distribution space, repeating processing for determining a pixel indicating a maximum value in the weighted density distribution space as a pixel for dotting and determining an order of determining the dotting as a dotting order of the dots, and updating the density distribution space in accordance with a position of the pixel each time the pixel is determined; and a threshold conversion part for converting the dotting order of the dots into a threshold and obtaining a threshold matrix. The probability conversion part uses a threshold matrix that is created beforehand so as to form dots, as the initial pattern and obtains the probability conversion space by performing smoothing processing after gradation conversion processing is performed on the initial pattern.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an image processing apparatus, an image forming apparatus, and a threshold matrix shaping method.

  Conventionally, a simulated annealing method, a void & cluster method, a BIPPCCA method, and the like are known as methods for creating a threshold matrix for an FM screen (see, for example, Patent Document 1 and Non-Patent Document 1). In either case, the threshold value of each pixel is determined based on an initial pattern in which dots are randomly arranged by random numbers.

If the halftone dots formed by the threshold value matrix have large irregularities or irregular shapes, the screen-processed image tends to have a grainy and rough texture.
Therefore, a method has been proposed in which a threshold value matrix that forms halftone dots is used as an initial pattern, and each threshold value of the threshold value matrix is determined in accordance with the density distribution of dots in the initial pattern (see, for example, Patent Document 2). ). According to this method, the shape of the halftone dots formed by the threshold matrix can be made close to a circle, and the graininess of the image after screen processing can be improved.

US Pat. No. 6,798,537 Japanese Patent No. 5601333

Modern Digital HARFTONING, Daniel El Low et al., 2nd edition, CRC Press publication (USA)

  In the image screened by the threshold matrix for FM screen, although the arrangement of halftone dots is not regular, the halftone dots are connected in a line in a high density region, and locally irregular shading tends to occur. This shading is observed as noise. When there is a connection direction such as halftone dots being more easily connected in the main scanning direction than in the sub-scanning direction, there is a difference in image reproducibility between the main scanning direction and the sub-scanning direction, leading to a reduction in image quality.

  An object of the present invention is to provide a threshold value matrix having few dot connections and high image quality after screen processing.

According to the invention of claim 1,
A probability conversion unit that converts the initial pattern into a probability distribution space that represents the probability that a dot is hit, and
According to the probability distribution space, the density distribution space representing the density of the pixel for which the dot dot is determined is weighted, and the pixel having the maximum value in the weighted density distribution space is determined as a dot dot dot pixel. The process of determining the order in which the dots are determined as the dot ranking of the dots is repeated until all the pixels of the initial pattern are determined, and the density is determined in accordance with the position of the pixels every time the pixels to be doted are determined. An iterative processing unit for updating the distribution space;
A threshold conversion unit that converts dot dot ranks of all the pixels of the initial pattern into threshold values and obtains a threshold matrix; and
The probability conversion unit obtains the probability distribution space by using a threshold matrix prepared in advance so as to form a halftone dot as the initial pattern, and performing a smoothing process on the initial pattern after a gradation conversion process. An image processing apparatus is provided.

According to invention of Claim 2,
The repetitive processing unit updates the dot dot space indicating each pixel for which the dot dot has been determined every time the pixel for dot dot determination is determined, and performs the update by smoothing the dot dot space. The image processing apparatus according to claim 1, wherein a density distribution space is obtained.

According to invention of Claim 3,
The image processing apparatus according to claim 1, wherein the repetitive processing unit updates only the density distribution space.

According to invention of Claim 4,
The said probability conversion part makes the length of the range of the said smoothing process shorter than the distance between the halftone dots formed by the said threshold value matrix, The Claim 1 characterized by the above-mentioned. An image processing apparatus is provided.

According to the invention of claim 5,
The image processing apparatus according to claim 1, wherein the probability conversion unit varies the length of the smoothing processing range between a main scanning direction and a sub-scanning direction. Provided.

According to the invention of claim 6,
The image processing apparatus according to claim 1, wherein the gradation conversion process is a binarization process.

According to the invention of claim 7,
The gradation conversion process outputs a minimum value when the input value is smaller than the first threshold value, and outputs an output value corresponding to the input value in a linear expression with a slope of 1 or more when the input value is greater than or equal to the first threshold value and less than or equal to the second threshold value. The image processing apparatus according to claim 1, wherein the image processing apparatus outputs the maximum value when the value is larger than the second threshold value.

According to the invention described in claim 8,
The image processing apparatus according to claim 1, wherein the threshold matrix used as the initial pattern is a threshold matrix having a blue noise characteristic or a green noise characteristic.

According to the invention of claim 9,
The image processing apparatus according to claim 1, wherein the threshold value matrix used as the initial pattern is a threshold value matrix obtained by the threshold value conversion unit.

According to the invention of claim 10,
The image processing apparatus according to claim 1, further comprising a screen processing unit that performs screen processing on an image using the threshold value matrix obtained by the threshold value conversion unit. .

According to the invention of claim 11,
An image forming apparatus comprising the image processing apparatus according to any one of claims 1 to 10 is provided.

According to the invention of claim 10,
(A) converting the initial pattern into a probability distribution space representing the probability that a dot is hit;
(B) According to the probability distribution space, the density distribution space representing the density of the pixel for which the dot dot is determined is weighted, and the pixel indicating the maximum value in the weighted density distribution space is determined as the dot dot pixel. The process of determining the order in which the dot dots are determined as the dot dot rank is repeated until all the pixels in the initial pattern are determined, and each time the dot dot is determined, the position of the dot is determined. Updating the density distribution space;
(C) converting dot dot ranks for all the pixels of the initial pattern into threshold values to obtain a threshold value matrix,
In the step (a), a threshold value matrix prepared in advance so as to form a halftone dot is used as the initial pattern, and the initial pattern is subjected to a gradation conversion process and then a smoothing process. A method for shaping a threshold matrix is provided.

  According to the present invention, it is possible to provide a threshold value matrix with few dot connections and high image quality after screen processing.

FIG. 2 is a block diagram illustrating a main configuration of the image forming apparatus according to the present embodiment for each function. It is a block diagram which shows the structure of the shaping process part of FIG. 1 for every function. It is a figure which shows an example of the gradation conversion characteristic of the gradation conversion process applicable to a threshold value matrix. It is a figure which shows an example of the smoothing filter which can be used for a smoothing process. It is a figure which shows an example of probability distribution space, density distribution space, dot dot space, and dot rank space. It is a flowchart which shows the procedure of the shaping process which a shaping process part performs. It is a figure which shows the process result of the screen process by the presence or absence of the shaping process with respect to a threshold value matrix.

  Embodiments of an image processing apparatus, an image forming apparatus, and a threshold matrix shaping method according to the present invention will be described below with reference to the drawings.

FIG. 1 shows the main configuration of the image forming apparatus G according to the embodiment of the present invention for each function.
As shown in FIG. 1, the image forming apparatus G includes a control unit 11, a storage unit 12, an operation unit 13, a display unit 14, a communication unit 15, an image generation unit 16, an image reading unit 17, an image memory 18, and an image processing device. 19 and an image forming unit 20.

The control unit 11 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and the like, and controls each unit by reading and executing various programs from the storage unit 12.
For example, the control unit 11 causes the image processing device 19 to perform image processing on the original image generated by the image generation unit 16 or the image reading unit 17 and held in the image memory 18, and based on the original image after image processing. Then, the image forming unit 20 forms an image on the sheet.

  The storage unit 12 stores a program that can be read by the control unit 11, a file that is used when the program is executed, and the like. As the storage unit 12, a large-capacity memory such as a hard disk can be used.

  The operation unit 13 generates an operation signal corresponding to a user operation and outputs the operation signal to the control unit 11. As the operation unit 13, a keypad, a touch panel configured integrally with the display unit 14, or the like can be used.

  The display unit 14 displays an operation screen and the like according to instructions from the control unit 11. As the display unit 14, an LCD (Liquid Crystal Display), an OELD (Organic Electro Luminescence Display), or the like can be used.

The communication unit 15 communicates with external devices on the network, for example, user terminals, servers, other image forming systems, and the like.
The communication unit 15 receives vector data in which an instruction content for forming an image is described in a page description language (PDL) via a network from a user terminal.

  The image generation unit 16 rasterizes the vector data received by the communication unit 15 to generate a bitmap-format original image. In the original image, each pixel has pixel values of four colors of C (cyan), M (magenta), Y (yellow), and K (black). The pixel value is a data value representing the density of an image. For example, an 8-bit data value represents a density of 0 to 255 gradations.

  The image reading unit 17 includes an automatic document feeder, a scanner, and the like, and reads a document surface set on a document table to generate a bitmap-format original image. In the original image generated by the image reading unit 17, each pixel has pixel values of three colors of R (red), G (green), and B (blue). This original image is color-converted into an original image having pixel values of four colors C, M, Y, and K by a color conversion unit (not shown).

  The image memory 18 is a buffer memory that temporarily holds an original image generated by the image generation unit 16 or the image reading unit 17. As the image memory 18, a DRAM (Dynamic RAM) or the like can be used.

The image processing device 19 reads an original image from the image memory 18 and performs various image processing such as image rotation, enlargement, reduction, page number addition, layout processing such as page aggregation, color conversion processing, and density correction processing. .
Further, the image processing device 19 includes a screen processing unit 191 as shown in FIG. 1, and the screen processing unit 191 screen-processes the original image using a threshold value matrix. The screen processing is image processing for expressing a pseudo halftone using a systematic dither method.

  The image processing device 19 includes a shaping processing unit 30. The shaping processing unit 30 shapes the given threshold value matrix and outputs it to the screen processing unit 191. The screen processing unit 191 uses the threshold value matrix obtained by the shaping processing unit 30 for screen processing.

The image forming unit 20 forms an image of four colors C, M, Y, and K on a sheet according to the pixel values of the four colors of each pixel of the original image subjected to image processing by the image processing device 19.
Specifically, the image forming unit 20 exposes the photosensitive drum on the photosensitive drum according to the pixel value of the color formed by each pixel, forms an electrostatic latent image, and then develops a color such as toner by the developing unit. The material is supplied and developed to form an image on the photosensitive drum. The image forming unit 20 transfers the images of the respective colors on the image carrier such as an intermediate transfer belt from the photosensitive drum, and further transfers the obtained four-color image onto the paper conveyed from the paper feed tray. Then, a fixing process is performed in which the sheet is heated and pressed by the fixing device.

FIG. 2 shows the configuration of the shaping processing unit 30 for each function.
As illustrated in FIG. 2, the shaping processing unit 30 includes a probability conversion unit 31, an iterative processing unit A, and a threshold conversion unit 35.

[Probability converter]
As shown in FIG. 2, the probability conversion unit 31 includes an inversion unit 311, a gradation conversion unit 312, and a smoothing processing unit 313, and converts the initial pattern into a probability distribution space U. The probability distribution space U represents the probability that a dot is hit.

  The probability conversion unit 31 uses a threshold matrix TH0 created in advance so as to form halftone dots by a known threshold matrix creation method as an initial pattern. In the case of an FM screen, a threshold matrix TH0 created by a simulated annealing method, a void and cluster method, a BIPPCCA method, or the like can be used as an initial pattern. Note that the shaping processing unit 30 can perform shaping processing not only on the FM screen but also on a threshold matrix created so as to form halftone dots of the AM screen.

  The threshold matrix TH0 is an m × n pixel pattern in which a threshold is set for each pixel. During screen processing, the threshold matrix is compared with the original image in units of m × n pixels, and by comparing each threshold of the threshold matrix with the pixel value of each pixel corresponding to the position, the dot dot or non-dot in each pixel is determined. decide. The threshold value of the threshold value matrix TH0 created by the known creation method is set so that dots are gathered in a circle to form halftone dots, and the formed halftone dots are distributed at regular intervals.

The threshold matrix TH0 is preferably a threshold matrix having blue noise characteristics or green noise characteristics.
Thus, a threshold matrix capable of forming an image having blue noise characteristics or green noise characteristics can be obtained by screen processing.

The threshold value matrix TH0 may be a threshold value matrix obtained by the threshold value conversion unit 35 through the shaping process.
As a result, the threshold value matrix TH0 can be subjected to shaping processing a plurality of times to obtain a threshold value matrix with fewer dots connected.

  The inversion unit 311 inverts the threshold value set for each pixel of the threshold value matrix TH0. The threshold value after inversion represents the probability that a dot will be hit when screen processing is performed using the threshold value matrix TH0. The threshold value after inversion can be obtained by obtaining a difference obtained by subtracting the threshold value from the maximum value in the threshold value range. For example, when the threshold value set in the range of 0 to 255 is 50, the threshold value after inversion is 205 (= 255-50).

The gradation conversion unit 312 performs gradation conversion processing on the inverted threshold matrix TH0 so that the probability that dots are dotted between halftone dots in the probability distribution space U is reduced.
As a result, the threshold matrix TH0 can be shaped so that the connection between the halftone dots is reduced.

The gradation conversion process performed by the gradation conversion unit 312 is preferably a binarization process.
Specifically, the gradation conversion unit 312 compares each threshold value of the inverted threshold value matrix TH0 with the threshold value α of the binarization process, and if it is larger than the threshold value α, the maximum value is 1; Convert to the value 0. The threshold value α can be determined as appropriate. As the threshold value α is larger, the connection between halftone dots formed by screen processing tends to decrease. Therefore, the threshold value α when the desired image quality is obtained may be obtained by varying the threshold value α.

The gradation conversion process performed by the gradation conversion unit 312 may be a gradation conversion process having gradation conversion characteristics as shown in FIG.
As shown in FIG. 3, in this gradation conversion process, when the input value is smaller than the first threshold value β, the minimum value 0 is output, and when the input value is not less than the first threshold value β and not more than the second threshold value α, the gradient is 1 or more. In the linear expression, the output value corresponding to the input value is output, and when the value is larger than the second threshold value α, the maximum value 1 is output. In FIG. 3, the input value is represented by a value converted to 0 to 1. For example, when the value range is an input value of 0 to 255, the input value is divided by 256 and converted to a value of 0 to 1. A linear expression between the first threshold value β and the second threshold value α in FIG. 3 is y = ax + b (y represents an output value of 0 to 1, x represents an input value of 0 to 1, a is a slope, b Is an intercept.), And the slope a is a linear function of 1 or more.

The smoothing processing unit 313 obtains the probability distribution space U by smoothing the threshold value matrix TH0 after the gradation conversion processing.
By the smoothing process, it is possible to adjust the probability that dots are hit so that the density difference changes smoothly from the center of the halftone dot to the periphery. As a result, the threshold matrix can be shaped so that natural halftone dots can be formed.

In order to prevent the halftone dot shape from being excessively blurred by the smoothing process, the smoothing processing unit 313 makes the length of the smoothing process range shorter than the distance between the halftone dots formed by the threshold matrix TH0. It is preferable to do. The distance between halftone dots is the distance between the centers of adjacent halftone dots.
The range of the smoothing process refers to a pixel range in which the pixel value is weighted. In the smoothing filter, the range of pixels where the weighting parameter is not 0 is the range of the smoothing process.
For example, when a smoothing filter in which the set weighting parameters are not all 0 is used for the smoothing process, if the distance between halftone dots is 5 pixels, the size of the smoothing filter is set to one side of 5 × 5 pixels. A short size such as 4 × 4 pixels, 3 × 3 pixels, or the like may be used.

Further, when there is directionality in the ease of connecting halftone dots due to the characteristics of the image forming apparatus G, the smoothing processing unit 313 sets the length of the smoothing processing range in the main scanning direction and the sub-scanning direction. It is preferable to make them different.
Thereby, the difference in connection between the main scanning direction and the sub-scanning direction can be alleviated, and irregular shading caused by the direction can be prevented from being observed as noise.
For example, when halftone dots are easily connected in the main scanning direction, the weighting range in the smoothing process can be an elliptical range that is longer in the main scanning direction than in the sub scanning direction.

FIG. 4 shows an example of a smoothing filter whose weighting range is elliptical.
As shown in FIG. 4, the smoothing filter has a size of 5 × 5 pixels, but the weighting parameter that is not 0 is set and the weighting effective range is the length (major axis) in the main scanning direction x. It is an ellipse having 5 pixels and a length (minor axis) in the sub-scanning direction y of 3 pixels. Each parameter of the weight is set so that the weight becomes larger as it is closer to the central pixel of 5 × 5 pixels.
According to such a smoothing filter, it is possible to suppress the connection of adjacent halftone dots in the main scanning direction, and it is possible to reduce image quality degradation due to a difference in image reproducibility between the main scanning direction and the sub-scanning direction.

When the length of the smoothing process range is different between the main scanning direction and the sub-scanning direction, the shorter one of the length in the main scanning direction and the length in the sub-scanning direction is set between the halftone dots. It is preferable to make it shorter than the distance.
Thereby, not only the image quality deterioration due to the directionality but also the collapse of the halftone dot shape due to the smoothing process can be prevented.

[Repetition processing section]
As shown in FIG. 2, the iterative processing unit A includes a weighting calculation unit 32, a dot dot unit 33, and an update unit 34.

The weighting operation unit 32 performs an operation of weighting the density distribution space CM that represents the density of the pixels for which dot dots are determined, using the probability distribution space U output from the probability conversion unit 31.
The density distribution space CM is also referred to as a concentration matrix, and is an m × n pixel pattern representing the density of pixels for which dot dots are determined. The value of each pixel in the density distribution space CM has a value range of 0 to 1, and the initial value is 1.

The weighting calculator 32 multiplies the values of corresponding pixels in the probability distribution space U and the density distribution space CM to obtain a weighted density distribution space CMnew.
The weighting calculation of the density distribution space CM can be expressed as follows.
CMnew [i] = CMold [i] × U [i]
Here, CMnew [i] represents the value (density) of the i-th pixel in the density distribution space CM after the weighting calculation. CMold [i] represents the value (density) of the i-th pixel in the density distribution space CM before the weighting calculation. U [i] represents the value (probability) of the i-th pixel in the probability distribution space U.

  Since the initial value of the density distribution space CM is 1, when determining the first pixel to dot, the probability value of the probability distribution space U input from the probability conversion unit 31 is directly obtained as the density distribution space CMnew. It is done. Since the density distribution space CM updated by the updating unit 34 is given when determining pixels for dot printing thereafter, the weighting calculation unit 32 uses the updated density distribution space CM as the density distribution space CMold. The probability distribution space U is multiplied.

The dot dot unit 33 determines the pixel having the maximum value in the density distribution space CMnew weighted by the weighting calculation unit 32 as a pixel to dot, and the dot dot dot in the order in which the dot dot of the pixel is determined Decide as a ranking.
After determining the first pixel to dot, since there is a dot that has already been doted, even if the dot dot portion 33 is a pixel having the maximum value in the density distribution space CMnew, the dot dot portion Ig is already doted in the dot dot space Ig. Excluded pixels (pixels having a value of 1) are excluded, and a pixel to be doted is determined.

The dot dot unit 33 updates the dot dot space Ig and the dot dot rank space Num each time the dot dot dot and the dot dot rank are determined.
The dot dot space Ig is an m × n pixel pattern representing a pixel that determines the dot dot. Each pixel in the dot dot space Ig is set to an initial value 0 representing a non-dot dot, and the dot dot portion 33 represents the dot dot from the initial value 0 to the pixel value for which the dot dot has been determined. Update to 1.

The dot ranking space Num is an m × n pixel pattern representing dot dot ranking, and is expressed as follows.
Num [i] = k
Here, Num [i] represents the dot ranking of the i-th pixel in the dot ranking space Num. k is an integer of 1 to m × n representing dot dot ranking.

  The weighting calculation unit 32 and the dot dot unit 33 weight the density distribution space CM, determine the pixel indicating the maximum value in the weighted density distribution space CMnew as the dot dot pixel, and determine the dot dot of the dot The process of determining the order as the dot ranking of the dots is repeated until all the pixels of the threshold value matrix TH0 that is the initial pattern are determined.

The update unit 34 updates the density distribution space CM weighted by the weighting calculation unit 32 every time the dot dot dot unit 33 determines a dot dot dot according to the pixel position where the dot dot dot is determined.
The updating unit 34 includes an inversion unit 341 and a smoothing processing unit 342 as illustrated in FIG.

  The inversion unit 341 inverts the value of each pixel in the dot dot space Ig. Since the value of the dot dot space Ig is 0 or 1, it is sufficient to invert 0 to 1 and 1 to 0.

The smoothing processing unit 342 obtains an updated density distribution space CM by smoothing the dot dot space Ig after inversion.
By the smoothing process, it is possible to increase the density of the pixel in which the dot spot is determined in the density distribution space CM and the pixels located around the pixel. A dot distributed within a certain distance from the pixel where the dot dots are determined, and the density can be adjusted so that the dots are more likely to be spotted when they are more than a certain distance away. A point matrix can be shaped into a threshold matrix.

  The updated density distribution space CM is used to determine a pixel for hitting the next dot and its dot rank. In other words, the weighting calculation unit 32 weights the updated density distribution space CM with the probability distribution space U, and determines the dot dot ranking and dot ranking in the weighted density distribution space CMnew.

FIG. 5 shows a density distribution space CM that is weighted by the probability distribution space U and updated by the dot dot space Ig. In FIG. 5, the value of each pixel is expressed as a higher density as it is closer to 1 and as a lower density as it is closer to 0.
As shown in FIG. 5, a probability distribution space U with a low dot dot probability between halftone dots is obtained by binarizing and smoothing the threshold matrix TH0 input and inverted as an initial pattern. . The probability distribution space U indicates a probability distribution in which the center of the halftone dot of the threshold matrix TH0 is the highest and decreases concentrically from the center.

  The density distribution space CMold is weighted by the probability distribution space U, and the dot dot spot space Ig and the dot spot rank space Num are obtained by determining the dot dot dot and the dot dot rank in the obtained density distribution space CMnew. In the dot dot space Ig, the pixel for which the dot dot is determined has a value of 1, and the pixel for which the dot dot has not yet been determined has a value of 0. The updated density distribution space CMold is obtained by inverting the value of the dot spot space Ig and performing the smoothing process.

  In the updated density distribution space CM, the density of pixels around which dots are dotted and the surrounding area are reduced. By weighting the updated density distribution space CM with the probability distribution space U, it is possible to separate the position of the pixel that has already determined the dot spot from the pixel that newly determines the dot spot by a certain distance or more. .

  In addition, as described in Japanese Patent No. 5601333, the update unit 34 multiplies a pair correlation function so that the shape of the halftone dots becomes more concentric, or the halftone dots are spaced at regular intervals. The probability distribution space U can also be updated so that However, the probability distribution space U obtained by the gradation conversion process and the smoothing process is sufficiently adjusted in probability so that the halftone dots are formed concentrically, and the density distribution space CM updated by the smoothing process is the density distribution space CM. Since the density is adjusted so that the points are separated by a certain distance or more, only the density distribution space CM needs to be updated.

[Threshold conversion unit]
When the dot dot ranking is determined for all the pixels of the initial pattern, the threshold conversion unit 35 converts the dot ranking into threshold values to obtain a shaped threshold matrix TH1.
Specifically, the threshold conversion unit 35, when dots are doted on all the pixels in the dot dot space Ig and the dot rank space Num representing the dot rank of all the pixels is obtained, the dot rank space Num is converted into the threshold matrix TH1. Convert. This transformation can be expressed as:
TH1 [i] = {Num [i] / (m × n)} × g
Here, TH1 [i] represents the threshold value of the i-th pixel among m × n pixels. Num [i] represents the dot ranking of the dot of the i-th pixel among m × n pixels. g represents the maximum value of the pixel value range.

  For example, when the size m × n of the threshold matrix TH0 is 32 × 32 pixels, dot rankings of 1 to 1024 are set for each pixel in the dot ranking space Num. If the threshold value of the threshold matrix TH0 is 8-bit data, the maximum value g = 255, and therefore, the dot rank Num [i] of each pixel is input to TH1 [i] = Num [i] / 1024 × 255. Thus, the threshold value TH1 [i] for each pixel can be obtained.

  The threshold value matrix TH1 obtained by the threshold value conversion unit 35 is output to the screen processing unit 191. The screen processing unit 191 uses the threshold matrix TH1 during screen processing.

FIG. 6 shows the procedure of threshold matrix shaping processing executed by the shaping processor 30.
As shown in FIG. 6, in the shaping processing unit 30, the probability converting unit 31 inputs the shaping target threshold value matrix TH0 as an initial pattern (step S1).
The probability conversion unit 31 inverts the value of each pixel of the threshold matrix TH0, performs gradation conversion processing, and then performs smoothing processing, thereby converting the threshold matrix TH0 into the probability distribution space U (step S2).

  The weighting calculation unit 32 weights the density distribution space CM by the probability distribution space U obtained by the probability conversion unit 31 (step S3). In the first density distribution space CM, the value of each pixel of m × n pixels is the initial value 1.

  The dot dot portion 33 determines a pixel to be dot-dotted in the density distribution space CMnew weighted by the weight calculator 32, and sets the value of the pixel in the dot dot space Ig to 1 indicating that the dot has been dot-dotted. Thus, the dot dot space Ig is updated. Further, the dot dot unit 33 determines the order of pixels in which dot dots are determined as the dot rank, sets the value of the pixel in the dot rank space Num to a value representing the dot rank, and sets the dot rank space Num. Is updated (step S4).

When the dot dots of all the pixels are not determined in the dot dot space Ig (step S5: N), the updating unit 34 inverts the value of the dot dot space Ig and performs smoothing processing, thereby performing the density distribution space CM. Is updated (step S6).
Thereafter, the processing returns to step S3, and the above-described processing is repeated until dot dots are determined for all the pixels in the dot dot space Ig. That is, the weighting calculation unit 32 weights the updated density distribution space CM with the probability distribution space U, and the dot spotting unit 33 determines the dot spotting position and dot ranking in the weighted density distribution space CMnew.

  When the dot dots for all the pixels are determined (step S5: Y), the threshold conversion unit 35 converts the dot rank in the dot rank space Num into a threshold value to obtain a shaped threshold matrix TH1 (step S7). ).

FIG. 7 shows the processing result of the screen processing depending on the presence / absence of the shaping processing on the threshold matrix. 7 indicates that the higher the density of the threshold matrix, the smaller the threshold (the dot dot probability is higher).
In FIG. 7, the comparative example shows a processing result obtained by performing screen processing on an image in a low density area, an image in a halftone area, and an image in a high density area using the threshold matrix TH0 input as an initial pattern.
The embodiment shows a processing result obtained by subjecting the same image as the comparative example to screen processing using the threshold value matrix TH1 obtained by the shaping processing.

  As shown in FIG. 7, the halftone dots in the image of the comparative example are connected in a line as the density increases, and the connected portion is likely to be observed as noise due to local and irregular density changes. On the other hand, the halftone dots in the image of the embodiment have almost no connected portion regardless of the density region of the image and are concentric, so that a very high quality image is obtained.

  As described above, the image processing apparatus 19 according to the present embodiment uses the probability conversion unit 31 that converts the initial pattern into the probability distribution space U that represents the probability that dots are hit, and the probability distribution space U allows dot dots to be set. The density distribution space CM representing the density of the determined pixel is weighted, and the pixel showing the maximum value in the weighted density distribution space CMnew is determined as a dot dot pixel, and the order in which the dot dot is determined is determined by the dot The repetition processing unit A that repeats the process of determining the dot ranking for all pixels in the initial pattern until it determines all the pixels in the initial pattern, and updates the density distribution space CM according to the position of the pixel every time a pixel for dot doting is determined. The threshold conversion unit 3 converts dot dot ranks for all pixels of the initial pattern into threshold values to obtain a threshold matrix TH1. It has a, and. The probability conversion unit 31 uses a threshold matrix TH0 created in advance so as to form a halftone dot as an initial pattern, and performs a smoothing process on the initial pattern after performing a gradation conversion process. obtain.

  As a result, it is possible to provide a threshold value matrix with few dot connections and high image quality after screen processing.

The above embodiment is a preferred example of the present invention, and the present invention is not limited to this. Modifications can be made as appropriate without departing from the spirit of the present invention.
For example, although the example in which the shaping process is performed by the image processing apparatus 19 mounted on the image forming apparatus G has been shown, the control unit 11 causes the control unit 11 to execute the above processing procedure of the shaping processing unit 30 by reading the program. You can also. In addition to the image forming apparatus G, the program can be read by a computer such as a general-purpose PC, and the above processing procedure can be executed.
Further, as a computer-readable medium for the program, a non-volatile memory such as a ROM and a flash memory, and a portable recording medium such as a CD-ROM can be applied. A carrier wave (carrier wave) can also be applied as a medium for providing program data via a communication line.

G Image forming device 11 Control unit 12 Storage unit 19 Image processing device 191 Screen processing unit 20 Image forming unit 30 Shaping processing unit 31 Probability conversion unit 311 Inversion unit 312 Tone conversion unit 313 Smoothing processing unit 32 Weighting calculation unit 33 Dot dot Unit 34 update unit 341 inversion unit 342 smoothing processing unit 35 threshold conversion unit

Claims (12)

A probability conversion unit that converts the initial pattern into a probability distribution space that represents the probability that a dot is hit, and
According to the probability distribution space, the density distribution space representing the density of the pixel for which the dot dot is determined is weighted, and the pixel having the maximum value in the weighted density distribution space is determined as a dot dot dot pixel. The process of determining the order in which the dots are determined as the dot ranking of the dots is repeated until all the pixels of the initial pattern are determined, and the density is determined in accordance with the position of the pixels every time the pixels to be doted are determined. An iterative processing unit for updating the distribution space;
A threshold conversion unit that converts dot dot ranks of all the pixels of the initial pattern into threshold values and obtains a threshold matrix; and
The probability conversion unit obtains the probability distribution space by using a threshold matrix prepared in advance so as to form a halftone dot as the initial pattern, and performing a smoothing process on the initial pattern after a gradation conversion process. An image processing apparatus.   The repetitive processing unit updates the dot dot space indicating each pixel for which the dot dot has been determined each time the pixel for dot dot determination is determined, and smoothes the dot dot space, thereby updating the dot dot space. The image processing apparatus according to claim 1, wherein a density distribution space is obtained.   The image processing apparatus according to claim 1, wherein the iterative processing unit updates only the density distribution space.   The said probability conversion part makes the length of the range of the said smoothing process shorter than the distance between the halftone dots formed by the said threshold value matrix, The Claim 1 characterized by the above-mentioned. Image processing device.   The image processing apparatus according to claim 1, wherein the probability conversion unit varies a length of the smoothing process range between a main scanning direction and a sub-scanning direction.   The image processing apparatus according to claim 1, wherein the gradation conversion process is a binarization process.   The gradation conversion process outputs a minimum value when the input value is smaller than the first threshold value, and outputs an output value corresponding to the input value in a linear expression with a slope of 1 or more when the input value is greater than or equal to the first threshold value and less than or equal to the second threshold value. The image processing apparatus according to claim 1, wherein the image processing apparatus outputs the maximum value when the value is larger than the second threshold value.   The image processing apparatus according to claim 1, wherein the threshold matrix used as the initial pattern is a threshold matrix having a blue noise characteristic or a green noise characteristic.   The image processing apparatus according to claim 1, wherein the threshold value matrix used as the initial pattern is a threshold value matrix obtained by the threshold value conversion unit.   The image processing apparatus according to claim 1, further comprising a screen processing unit that performs screen processing on an image using the threshold value matrix obtained by the threshold value conversion unit.   An image forming apparatus comprising the image processing apparatus according to claim 1. (A) converting the initial pattern into a probability distribution space representing the probability that a dot is hit;
(B) According to the probability distribution space, the density distribution space representing the density of the pixel for which the dot dot is determined is weighted, and the pixel indicating the maximum value in the weighted density distribution space is determined as the dot dot pixel. The process of determining the order in which the dot dots are determined as the dot dot rank is repeated until all the pixels in the initial pattern are determined, and each time the dot dot is determined, the position of the dot is determined. Updating the density distribution space;
(C) converting dot dot ranks for all the pixels of the initial pattern into threshold values to obtain a threshold value matrix,
In the step (a), a threshold value matrix prepared in advance so as to form a halftone dot is used as the initial pattern, and the initial pattern is subjected to a gradation conversion process and then a smoothing process. A method of shaping a threshold matrix, characterized by: