JP6256747B2 - Image conversion processing method, program for executing the same, and image conversion processing apparatus - Google Patents

Description

  The present invention relates to an image conversion processing method, a program for executing the image conversion method, and an image conversion processing device, and more particularly, to a method, program, and device for converting a grayscale image that is a multi-valued image into a hybrid screening image.

  In the field of lithographic printing, a grayscale image, which is an input multi-valued image, is expressed using a screening technique that represents a binary image with gradation values “0” and “1”. Has been done. Conventional screening methods are roughly classified into a halftone dot AM (Amplitude Modulation) method and an FM (Frequency Modulation) screening method.

  The halftone dot method regularly arranges halftone dots based on the multilevel tone values of the grayscale image and modulates the dot size of each halftone dot according to the multilevel tone values of the grayscale image. This is a technique for expressing gradation in a pseudo manner.

  On the other hand, the FM screening method is a method of performing gradation expression by modulating the dot arrangement density. Conventionally, when creating an image by the FM screening method, for example, a dither method or an error diffusion process is used. In the dither method and the error diffusion process, the gray scale image is reproduced using two chromatic pixels of “black” and “white” using the illusion of human eyes.

  However, an image created using the halftone method (halftone image) is arranged on a grid in which halftone dots are regularly arranged, so that an image having no graininess can be expressed. While having the advantage of being hardly affected by the printing environment, it has the disadvantage of generating interference fringes such as moire and rosette patterns.

  In addition, it is possible to suppress the generation of interference fringes by shifting the screen angles of the 4th version of CMWK, but it is not possible to completely suppress the generated interference fringes, and moire is also applied to the striped pattern portion. Is observed.

  In addition, in an image (FM screening image) created using the FM screening technique, the arrangement positions of the dots are random and not regular, so that interference fringes such as moire and rosetta patterns are less likely to occur. Compared with the halftone dot method, the degree of freedom of dot placement is high, so it has the advantage that images with higher resolution and high image quality can be obtained. It is difficult to transfer the image from the plate to the paper, and there is a disadvantage that a visually rough feeling is generated on the printed matter on the highlight side.

  Therefore, recently, with the penetration of CTP (Computer To Plate) and digital workflow, it has high accuracy and high resolution even in the newspaper printing field where commercial printing and mass printing that require high image quality and high added value are required. A hybrid screening technique that combines the advantages of the halftone technique and the conventional FM screening technique has been proposed and is widely used.

  For example, in Patent Documents 1 and 2, the center of the screen of halftone dots is used as a generating point, the position of the generating point is changed by a random number, and Voronoi division is performed using the generating point as a core. The input image is divided into a plurality of unit regions. A method for generating a screening image in which the periodicity of halftone dots is eliminated is described.

  Patent Document 3 describes a method of creating a hybrid screening image from a multivalued image. In this method, first, based on the DBS (Direct Binary Search) method, a binary image is created by performing a simple binarization process on a multi-valued image, and a binary image is created on the created binary image. A multi-valued image is restored by dimensional filtering. Then, an error between the restored multi-value image and the original multi-value image is calculated, and by determining the pixels of the binary image so that the calculated error becomes a luminance value pattern satisfying a predetermined condition, the multi-value image A hybrid screening image is created from

JP 2005-341089 A Japanese Patent No. 4599511 JP 2004-304543 A

  However, in the methods described in Patent Documents 1 and 2, the non-periodic cluster shape representing dots becomes larger as the density value increases, and a pattern like a turtle shell is generated, giving an unnatural impression to the entire image. There was a problem that. In addition, the dot shape of the negative image and the dot shape of the inverted positive image are different, and there is a possibility that the continuity of gradation like a halftone dot is lost.

  Furthermore, although the method described in Patent Document 3 has attracted attention as a new hybrid screening process, it is necessary to create a binary image from a multi-valued image first, and two-dimensional convolution operation with filter coefficients for all pixels Therefore, there is a problem that the processing takes a long time.

  Therefore, the present invention has been made in view of the above-described problems in the prior art, and an input multi-valued image can be produced with high image quality and high addition without generating interference fringes such as moire and rosetta patterns. It is an object of the present invention to provide an image conversion processing method that has value and can be converted into a hybrid screening image suitable for a printing process, a program for executing the method, and an image conversion processing apparatus.

In order to achieve the above object, the present invention provides an image conversion processing method for converting a grayscale image into a hybrid screening image, wherein the grayscale image has a halftone cell size based on resolution information of the grayscale image. A corresponding conversion block is set for the grayscale image, the grayscale image is converted into a first binary image composed of black pixels and white pixels for each conversion block, and the number of black pixels in the conversion block In addition, a random weighting factor that can be adjusted to improve the image quality is added to the detected number of black pixels, thereby detecting the first number of black pixels in the conversion block for each conversion block. A binarization step for detecting a conversion error that occurs when the first binary image is created, and the first black pixel count and A conversion error distribution step of converting the first binary image into a second binary image based on the conversion error and detecting a second number of black pixels included in the second binary image; An FM screening image creation step for converting the second binary image into an FM screening image based on the detected number of second black pixels, and a plurality of adjacent pixels formed from the created FM screening image A pixel fusion step of detecting a predetermined pixel pattern to be performed, performing a predetermined pixel fusion process on the detected predetermined pixel pattern, and creating a hybrid screening image.

The present invention is also a program for causing a computer apparatus to execute an image conversion processing method for converting a grayscale image into a hybrid screening image, and based on the resolution information of the grayscale image, the halftone of the grayscale image A conversion block corresponding to the size of the cell is set for the grayscale image, and the grayscale image is converted into a first binary image including black pixels and white pixels for each conversion block, It detects the number of black pixels, by adding an adjustable randomized weighting factors in order to improve the image quality for the detected black-pixel count, converting a first number of black pixels in the converted block A binarization step is detected for each block and detects a conversion error occurring when the first binary image is created. And converting the first binary image into a second binary image based on the first black pixel count and the conversion error, and the second black image included in the second binary image. A conversion error distribution step for detecting the number of pixels, an FM screening image creation step for converting the second binary image into an FM screening image based on the detected number of the second black pixels, and the created FM A pixel fusion step of detecting a predetermined pixel pattern formed by a plurality of adjacent pixels from the screening image, performing a predetermined pixel fusion process on the detected predetermined pixel pattern, and creating a hybrid screening image; A computer apparatus is caused to execute an image conversion processing method comprising:

Furthermore, the present invention is an image conversion processing device for converting a grayscale image into a hybrid screening image, and based on the resolution information of the grayscale image, a conversion block corresponding to the size of a halftone cell of the grayscale image is provided. Set for the grayscale image, convert the grayscale image to a first binary image composed of black pixels and white pixels for each conversion block, detect the number of black pixels in the conversion block, by adding an adjustable randomized weighting factors in order to improve the image quality for the detected black-pixel count, detecting the number of first black pixel in the converted block for each transform block, the first A binarization processing unit that detects a conversion error that occurs when the binary image is generated, and based on the first black pixel number and the conversion error. A conversion error dispersion processing unit for converting the first binary image into a second binary image and detecting a second number of black pixels included in the second binary image; Based on the second number of black pixels, an FM screening image creation processing unit that converts the second binary image into an FM screening image, and a plurality of adjacent pixels formed from the created FM screening image A pixel fusion processing unit that detects a predetermined pixel pattern, performs a predetermined pixel fusion process on the detected predetermined pixel pattern, and creates a hybrid screening image;

  As described above, according to the present invention, an input multi-valued image can be converted into a hybrid screening image having high image quality and high added value and suitable for a printing process.

It is a block diagram which shows one Embodiment of the image conversion processing apparatus which concerns on this invention. It is the schematic for demonstrating the binarization process which converts a gray scale image into a binary image. It is the schematic for demonstrating conversion error dispersion | distribution processing. It is the schematic for demonstrating the effect of a conversion error dispersion | distribution process. It is the schematic for demonstrating the production process of FM screening image. It is the schematic which shows an example of the FM screening image produced by the FM screening image production process. It is the schematic for demonstrating the FM screening image produced by the FM screening image creation process. It is the schematic for demonstrating the FM screening image (gradation image) produced by the FM screening image creation process. It is the schematic for demonstrating the FM screening image (gradation image) produced by the FM screening image creation process. It is the schematic for demonstrating a 1st pixel fusion process. It is the schematic for demonstrating the 2nd pixel fusion process. It is the schematic which shows an example of the screening image produced by the two-dimensional pixel fusion process. It is the schematic for demonstrating the 3rd pixel fusion process. It is the schematic which shows an example of the hybrid screening image produced by the one-dimensional pixel fusion process. It is the schematic for demonstrating a roughness correction process. It is the schematic which shows an example of the hybrid screening image produced by the roughness correction process. It is the schematic which compared the FM screening image and the hybrid screening image. It is a flowchart for demonstrating the flow of each process in an image conversion processing apparatus. It is a flowchart for demonstrating the flow of a binarization process and a conversion error dispersion | distribution process. It is a flowchart for demonstrating the flow of FM screening conversion process. It is a flowchart for demonstrating the flow of an arrangement | positioning process. It is a flowchart for demonstrating the flow of a two-dimensional pixel fusion process. It is a flowchart for demonstrating the flow of a one-dimensional pixel fusion process. It is a flowchart for demonstrating the flow of a roughness correction process. It is the schematic for demonstrating the converted hybrid screening image. It is the schematic which shows an example at the time of converting a gradation image into a hybrid screening image.

  Next, an embodiment for carrying out the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows an embodiment of an image conversion processing apparatus according to the present invention. This image conversion processing apparatus 1 is roughly divided into a memory 2, a binarization processing unit 3, a conversion error dispersion processing unit 4, and an FM screening. An image creation processing unit 5, a pixel fusion processing unit 6, and a roughness correction processing unit 7 are provided.

  Input to the image conversion processing apparatus 1 is input grayscale image data (input multivalued image data) 10 that is output from an external input device such as a scanner and converted into a grayscale image having multivalued gradation values. After the input grayscale image data 10 is subjected to various processes, a black pixel whose tone value is “1” and a white pixel whose tone value is “0” are two. Output binary image data 20 expressed by values is output.

  The memory 2 stores input grayscale image data 10 input from the outside, and also includes a binarization processing unit 3, a conversion error dispersion processing unit 4, an FM screening image creation processing unit 5, a pixel fusion processing unit 6, and roughness correction. This is a storage medium for storing image data that has been subjected to various types of processing in the processing unit 7, information necessary for each processing, and the like.

  The binarization processing unit 3 reads the input grayscale image data 10 stored in the memory 2 and determines the multilevel gradation value of the input grayscale image data 10 from only the binary gradation values of black pixels and white pixels. And the number of black pixels included in the first binary image is detected. In addition, when detecting the number of black pixels included in the first binary image, an error occurs as described later, and this error is detected as conversion error information. The converted first binary image, the first black pixel number information indicating the number of detected black pixels, and the conversion error information are stored in the memory 2.

  The conversion error variance processing unit 4 reads out the first binary image, the first black pixel number information, and the conversion error information stored in the memory 2, and based on the first black pixel number information and the conversion error information, The binary image converted by the binarization processing unit 3 is subjected to processing for suppressing a tone jump phenomenon that occurs during image conversion by the binarization processing unit 3 to create a second binary image. The number of black pixels included in the second binary image is detected. The generated second binary image and the second black pixel number information indicating the detected number of black pixels are stored in the memory 2.

  The FM screening image creation processing unit 5 reads the second binary image and the second black pixel number information stored in the memory 2 and creates an FM screening image using the second black pixel number information. The created FM screening image is stored in the memory 2. Further, the FM screening image creation processing unit 5 has a required arrangement pixel number counter and an arranged pixel number counter (not shown) used when creating an FM screening image. These counters will be described later.

  The pixel fusion processing unit 6 reads the FM screening image stored in the memory 2 and detects a predetermined pixel pattern from the FM screening image. Then, the pixel fusion processing unit 6 performs pixel fusion processing on the detected predetermined pixel pattern to create a hybrid screening image. The created hybrid screening image is stored in the memory 2.

  The roughness correction processing unit 7 reads out the hybrid screening image stored in the memory 2, performs a correction process for suppressing the feeling of roughness of the hybrid screening image, and creates output binary image data 20. The created output binary image data 20 is stored in the memory 2.

  The binarization processing unit 3, the conversion error dispersion processing unit 4, the FM screening image creation processing unit 5, the pixel fusion processing unit 6, and the roughness correction processing unit 7 described above are not necessarily configured by hardware. Part or all may be configured by software (program).

  Next, processing of each unit in the image conversion processing apparatus 1 having the above configuration will be described.

  First, the binarization processing by the binarization processing unit 3 will be described with reference to FIG. In the binarization processing, first, a halftone cell (halftone cell) block is included in the input grayscale image data 10 based on the resolution information of the input grayscale image data 10 as in the case of creating halftone dots. It is set as a conversion block which is a detection unit for detecting black pixels to be detected. Then, the input grayscale image data 10 is converted into a binary image composed of black pixels and white pixels for each conversion block.

  Next, an average gradation value of the input grayscale image data 10 in the conversion block is calculated based on the gradation value of each pixel included in the conversion block, and the corresponding first gradation value is calculated using the calculated average gradation value. A density value indicating the ratio between the number of black pixels and the number of all pixels in one binary image conversion block is acquired. Based on the density value of the conversion block, the number of black pixels corresponding to the average gradation value in the conversion block in the gray scale image is detected.

  That is, when the conversion block size is set to “Lx” pixels and “Ly” pixels in the horizontal and vertical directions, and the gradation value of the grayscale image is P, the number of black pixels s in the binary image obtained by the conversion Can be obtained based on Equation (1).

  Next, a randomization weight coefficient γ generated using a uniform random number is added to the number of black pixels s of the binary image acquired based on Equation (1), and the black pixels of the first binary image are added. The number S is determined. The randomized weighting coefficient γ is a coefficient that allows the strength to be adjusted in order to improve image quality such as sharpness and sharpness. This black pixel number S can be obtained based on Equation (2).

  Here, in the binarization process, the grayscale values of the pixels included in the conversion block of the grayscale image are averaged to obtain the number of black pixels in the first binary image, and therefore expressed by the size of the halftone cell block. The maximum number of gradations that can be made can be determined. Further, the number of gradations that can be expressed changes depending on the number of pixels included in the halftone cell block, and the change in gradation is expressed by the number of pixels.

  Therefore, when the size of the halftone cell block is large, the converted first binary image can increase the number of gradations that can be expressed, but it becomes a grainy image and the image quality deteriorates. On the other hand, when the size of the halftone cell block is small, the converted first binary image has a small number of gradations that can be expressed, but becomes a fine-grained image, and the image quality can be improved.

  Since the binary image converted using the binarization processing in the present embodiment has the same conversion image quality as that obtained when blur processing using an average filter is performed on a grayscale image, the size of the conversion block can be made as small as possible. It is necessary to make it smaller. The inventor has verified through an evaluation experiment that a good conversion image quality can be obtained by setting the size of the conversion block to about “4 × 4” pixels to “6 × 6” pixels. The example shown in FIG. 2 shows the correspondence between the gradation value of the multi-value image and the number of black pixels of the binary image when the size of the conversion block is “6 × 6”.

  Next, conversion error dispersion processing by the conversion error dispersion processing unit 4 will be described with reference to FIG.

  As described above, when binarization processing is performed by setting a halftone cell block as a conversion block, the number of gradations that can be expressed changes according to the size of the halftone cell block. For example, when the size of the halftone cell block is “16 × 16” pixels, 257 (16 × 16 + 1) gradation representations are possible.

  However, when a grayscale image in which the size of the halftone cell block is set in this way is converted into a binary image of 1200 dpi (dot per inch), the number of lines becomes 75 (= 1200/16) lines, The size of the dots becomes large enough to be visible to the human eye, and the image quality of the entire image is deteriorated.

  Here, the human eye can generally identify about 200 gradations. For example, the halftone cell block size is reduced, and the number of gradations that can be expressed is 200 or less. When the block size is set (for example, about “14 × 14” pixels in the case of a halftone cell), a “tone jump phenomenon” in which a boundary occurs between the halftones occurs.

  In the binarization processing described above, the number of black pixels s acquired based on the equation (1) is the number of black pixels in the binary image corresponding to the multi-value gradation value pixels of the conversion block, and can be expressed as black pixels. It includes an integer part and a part below the decimal point that cannot be expressed as a black pixel. At this time, the portion below the decimal point that cannot be expressed as a black pixel becomes a conversion error when the grayscale image is converted into a binary image, which causes a tone jump phenomenon.

  Therefore, in the present embodiment, the conversion error dispersion process is performed on the first binary image in order to suppress the influence of the image quality blur due to the binarization process and to obtain a finer image quality. As a result, the tone generated by setting the transform block size to be small as described above, adding a randomization weight to the number of black pixels in the transform block, and reducing the tone representation by reducing the transform block size. Jump phenomenon can be prevented.

  In the conversion error distribution process, as shown in FIG. 3, an error (tone value conversion error) ε generated in the binarization process is distributed at a predetermined ratio with respect to adjacent surrounding conversion blocks. For example, for a conversion block having a high correlation with the conversion block of interest, the gradation value conversion error ε is distributed at a higher rate than a conversion block with a low correlation. Then, in the conversion block to be converted, the gradation value conversion error ε distributed from the surrounding conversion blocks is used to obtain the black pixel number S ″ considering the conversion error based on the equation (3), and A second binary image is generated by performing conversion error dispersion processing on one binary image.

  FIG. 4 shows the effect when the conversion error dispersion processing is performed, and FIG. 4A shows the first binary image (gradation image) when the halftone block size is set small, and FIG. b) shows a second binary image when the conversion error dispersion processing is performed on the first binary image.

  In this way, by performing the conversion error dispersion process, a binary image can be created that has a smooth gradation expression and does not generate moire or the like without causing a tone jump phenomenon such as a pseudo contour. .

  Next, FM screening image creation processing by the FM screening image creation processing unit 5 will be described with reference to FIG.

  In the FM screening image creation process, the FM pixels are randomly arranged on the basis of the number of black pixels S ″ detected by the conversion error dispersion process by the conversion error dispersion processing unit 4 to thereby calculate FM. Create screening images.

  When black pixels are randomly arranged, a virtual layout diagram for arranging black pixels, which is referred to as a chess board type pixel layout diagram, is used.

  Specifically, first, as shown in FIG. 5A, arrangement pixel coordinates (m, n) are defined for each pixel of the transform block. In the example of FIG. 5A, the arrangement pixel coordinates from the upper left pixel to the lower right pixel are (0, 0), (1, 0),..., (0, Ly),. Lx, Ly). At this time, in order to ensure that the image information about the created FM screening image is evenly distributed in the four directions, up, down, left, and right, the horizontal coordinate value “Lx” and the vertical coordinate value “Ly” Are defined to be the same.

  When the arrangement pixel coordinates are defined in this way, the size of the transform block is “(Lx + 1) × (Ly + 1)”, and therefore the total number of pixels in the transform block is an odd number of “Lx” and “Ly”. In this case, the number is even, and when the values of “Lx” and “Ly” are even, the number is odd.

  Next, a chess board type pixel arrangement diagram is set for the conversion block in which the arrangement pixel coordinates are defined in this way. As shown in FIG. 5B, the chess board type pixel arrangement diagram is the same size as the conversion block, and the first arrangement layer and the second arrangement layer for arranging the arrangement pixels are alternately set in a checkered pattern. Is done. Also, coordinates (0, 0) to (Lx, Ly) similar to the arrangement pixel coordinates defined for the conversion block are set for each pixel in the chessboard pixel arrangement diagram.

  Based on the number of black pixels S ″ obtained by the conversion error distribution process, the black pixel as the arrangement pixel is one of the first arrangement layer and the second arrangement layer set in advance in the chessboard pixel arrangement diagram. The number of black pixels detected is randomly arranged on the layer.

  Here, in the FM screening image creation processing according to the present embodiment, a pixel arrangement determination condition for determining whether to arrange a black pixel in one of the first arrangement layer and the second arrangement layer is set in advance. To do. In this example, the pixel arrangement determination condition is set when the sum of the coordinates in the x direction and the y direction is an even number.

  Next, the total number of arranged pixels in the first arrangement layer and the second arrangement layer according to the set pixel arrangement determination condition is calculated. For example, when Lx in the chessboard pixel arrangement diagram is “odd”, the total number of pixels in the arrangement diagram is “even”, so that the pixel arrangement determination condition is “even” or “odd”. The total number of arranged pixels in each arrangement layer is the same number of pixels, and is “((Lx + 1) × (Ly + 1)) / 2”.

  On the other hand, when Lx is “even”, the total number of pixels in the layout drawing is “odd”. For this reason, the total number of arranged pixels when the pixel arrangement determination condition is “even” is larger by “1” than the total arrangement pixel number when the pixel arrangement determination condition is “odd”.

  That is, when Lx is “even” and the pixel arrangement determination condition is “even”, the total number of arranged pixels is “((Lx + 1) × (Ly + 1) +1) / 2”. Further, when Lx is “even” and the pixel arrangement determination condition is “odd”, the total number of arranged pixels is “((Lx + 1) × (Ly + 1) −1) / 2”.

  Specifically, for example, in the case of a “6 pixel × 6 pixel” conversion block, the total number of pixels is 36 (even), and the pixel arrangement determination condition is “even” or “odd”. In this arrangement layer, the total number of arranged pixels is 18.

  On the other hand, in the case of a conversion block of “5 pixels × 5 pixels”, the total number of pixels is 25 (odd number), so the total number of arranged pixels in the arrangement layer when the pixel arrangement determination condition is “even” is 13 Thus, when the pixel arrangement determination condition is “odd number”, the total number of arranged pixels in the arrangement layer is 12.

Next, the black pixels as the arrangement pixels are randomly arranged with respect to the arrangement layer corresponding to the preset pixel arrangement determination condition (in this example, an even number). Regarding the coordinates (m, n) of the black pixel (arranged pixel) P to be arranged, the number of pixels in the horizontal direction is “Lx + 1” and the number of pixels in the vertical direction is “Ly + 1” in the chessboard pixel arrangement diagram. In this case, the following formula (4) is used.
m = x (determined by a random number that can take values from “0” to “Lx”)
n = y (determined by a random number that can take values from “0” to “Ly”) (4)
Pixel arrangement determination condition TS = m + n (even number or odd number)

  The horizontal coordinate “m” in the arrangement pixel P is determined by generating a uniform random number that can take values from “0” to “Lx”. Similarly, the vertical coordinate “n” in the arrangement pixel P (m, n) is determined by generating a uniform random number that can take values from “0” to “Ly”.

  Then, based on “m” and “n” determined in this manner, a pixel arrangement determination condition TS is calculated, and the calculation result (“even” or “odd”) is set to a predetermined pixel arrangement determination condition. Compare As a result of the comparison, when the two pixel arrangement determination conditions are different, specifically, when the calculated pixel arrangement determination condition TS is an odd number, the arrangement is performed by generating a uniform random number again without arranging the arrangement pixels. Determine the new pixel coordinates for the pixel.

  On the other hand, if the two pixel arrangement determination conditions match, it is determined whether or not the arrangement pixel has already been arranged at the position indicated by the coordinates of the arrangement pixel P (m, n). As a result of the determination, if the arrangement pixel has already been arranged at the position indicated by the coordinate, the coordinates of the arrangement coordinate P (m, n) are discarded and a uniform random number is generated again, whereby a new arrangement pixel is obtained. Determine the coordinates.

  On the other hand, when the arrangement pixel is not arranged at the position indicated by the coordinates, the arrangement pixel is arranged. Then, the value of the required arrangement pixel number counter provided in the FM screening image creation processing unit 5 is decremented by “1”, and the value of the arranged pixel number counter is incremented by “1”.

  Here, the required number of pixels to be arranged counter is for counting the number of pixels to be arranged. The number of black pixels in the conversion block is set as an initial value, and the value is set every time the arranged pixels are arranged. Decrement by “1”.

  The arranged pixel number counter is for counting the number of arranged pixels that have been arranged. The value “0” is set as an initial value, and the value is set to “1” every time a arranged pixel is arranged. Increment by one.

  When the value of the arranged pixel number counter exceeds the total number of arranged pixels of the arrangement layer, it indicates that the arranged pixels are arranged at all coordinate positions where the arrangement layer can be arranged.

  Here, there may be a case where the number of required arrangement pixels is larger than the total number of arrangement pixels of the arrangement layer. In such a case, the preset pixel arrangement determination condition is changed, and the arrangement pixel is set to be arranged in the next arrangement layer.

  Specifically, for example, the pixel arrangement determination condition is set to “even”, and the arrangement pixel arrangement process is performed under this condition. Then, after arrangement pixels are arranged at all arrangement coordinates that can be arranged, if arrangement pixels to be arranged remain, the pixel arrangement determination condition is changed to “odd number”, and arrangement processing for the remaining arrangement pixels is performed. I do.

  In this way, the arrangement process described above is repeated until the value of the required arrangement pixel number counter becomes “0”. When the value of the arrangement pixel number counter becomes “0”, the arrangement pixel to be arranged is determined. It is determined that the arrangement has been completed, and the arrangement processing of the arrangement pixels is completed. Thereby, the halftone image corresponding to a predetermined conversion block can be converted into an FM screening image.

  In this example, the pixel arrangement determination condition is set to “even”. However, the present invention is not limited to this, and the pixel arrangement determination condition may be set to “odd”. In this case, the same processing as described above is performed. Thus, the halftone image can be converted into an FM screening image.

  FIG. 6 shows an example of an FM screening image created by the above-described FM screening image creation process. FIGS. 6A to 6F show FM screenings created when the density of black pixels in the conversion block is 10%, 25%, 40%, 50%, 75%, and 90%, respectively. Images are shown. This FM screening image is characterized by having pixel distribution characteristics in four directions, top, bottom, left, and right, while balancing the global uniform distribution and the local random distribution.

  7 shows an FM screening image (FIG. 7A) created by the FM screening image creation processing according to the present embodiment and an FM screening image created by the conventional error diffusion processing (FIG. 7B). ). As shown in FIG. 7 (a), the FM screening image created by the FM screening image creation processing according to the present embodiment has no unnatural turtle shell pattern generated in the conventional FM screening image, and the image quality is improved. I understand that.

  FIG. 8 shows an FM screening image of a gradation image created by the FM screening image creation processing according to the present embodiment. FIG. 8A shows an FM screening image created by the FM screening image creation processing according to the present embodiment. A screening image is shown, FIG.8 (b) shows the conventional FM screening image. FIG. 9 shows an FM screening image of a gradation image (300 dpi) created by the FM screening image creation processing according to the present embodiment. As described above, in the FM screening image created by the FM screening image creation processing according to the present embodiment, it is possible to suppress unevenness such as moire that causes image quality degradation.

  Here, in the FM screening image shown in FIG. 8, the pixels are isolated at intervals of one pixel from the pixels adjacent in the four directions, top, bottom, left, and right, for each pixel due to the pixel distribution characteristics of the chessboard type pixel arrangement diagram. In this state, since the isolated pixel is one pixel, the dot which is the minimum information unit of the image is small, the ink is not deposited on the paper by the rotary press, and the dot is smaller than the halftone image. Since the gain is very large, it is not suitable for the printing process.

  Therefore, in this embodiment, in order to realize screening suitable for a printing process in which the ink deposition property on the paper is improved by a rotary press while maintaining the high quality of the FM screening image, the following pixels are provided. By the fusing process, isolated pixels are converted into dots having a size suitable for printing.

  Next, pixel fusion processing by the pixel fusion processing unit 6 will be described with reference to FIGS.

  In the pixel fusion processing, first, raster scanning is performed on the FM screening image created by the FM screening image creation processing unit 5. Then, a predetermined pixel pattern in which black pixels and white pixels are distributed at one-pixel intervals is detected from a pixel pattern formed by a plurality of adjacent pixels, and two-dimensional pixel fusion is performed on the detected predetermined pixel pattern. Process.

  Specifically, first, a first pixel pattern including four pixels of “2 pixels × 2 pixels” shown in FIG. 10A is detected from the FM screening image. Specifically, the first pixel pattern is a pattern in which the upper left and lower right pixels are black pixels, and the upper right and lower left pixels are white pixels.

  When the first pixel pattern is detected, a first pixel fusion process is performed on the first pixel pattern. In the first pixel fusion process, as shown in FIG. 10B, an exchange process is performed in which the lower right black pixel in the first pixel pattern is replaced with the adjacent upper right or lower left white pixel. At this time, the replacement probability of the lower right black pixel and the upper right or lower left white pixel is both 50%.

  That is, in the first pixel fusion process, an exchange process for replacing the lower right black pixel and the upper right white pixel shown in FIG. 10C with a probability of 50% (hereinafter referred to as “first pixel fusion process 100a”). ), And an exchange process (hereinafter referred to as “first pixel fusion process 100b”) for exchanging the lower right black pixel and the lower left white pixel shown in FIG. I do.

  Next, in the two-dimensional pixel fusion process, the second pixel pattern shown in FIG. Specifically, the second pixel pattern is a pattern in which the upper left and lower right pixels are white pixels, and the upper right and lower left pixels are black pixels.

  When the second pixel pattern is detected, a second pixel fusion process is performed on the second pixel pattern. In the second pixel fusion process, as shown in FIG. 11B, an exchange process is performed in which the lower right white pixel in the second pixel pattern is replaced with the adjacent upper right or lower left black pixel. At this time, the exchange probability between the lower right white pixel and the upper right or lower left black pixel is set to 50%.

  That is, in the second pixel fusion process, an exchange process (hereinafter referred to as “second pixel fusion process 200a”) that replaces the lower right white pixel and the upper right black pixel shown in FIG. 11C with a probability of 50%. ), And an exchange process (hereinafter referred to as “second pixel fusion process 200b”) for exchanging the lower right white pixel and the lower left black pixel shown in FIG. I do.

  Finally, in the two-dimensional pixel fusion process, the first and second pixel fusion processes are alternately performed until the first and second pixel patterns cannot be detected from the FM screening image.

  As described above, in the two-dimensional pixel fusion processing, the black pixels and the white pixels are exchanged so that the black pixels in the pixel pattern in which the black pixels and the white pixels are distributed at intervals of one pixel are continuously distributed. A predetermined screening image is created.

  FIG. 12 is a screening image created by performing a two-dimensional pixel fusion process on the FM screening image when the density of the black pixel shown in FIG. 6D is 50%. As shown in FIG. 12, by performing the two-dimensional pixel fusion processing, it can be seen that pixels distributed in isolation in FIG. 6D are connected and formed in a linear shape.

  Next, in pixel pixel fusion processing, raster scanning is performed on the screening image converted by the above-described two-dimensional pixel fusion processing. A predetermined pixel pattern in which black pixels and white pixels are distributed at one-pixel intervals is detected from a pixel pattern formed by a plurality of pixels adjacent in the angular direction, and the detected predetermined pixel pattern will be described later. Perform one-dimensional pixel fusion processing.

  In the one-dimensional pixel fusion processing, first, a third pixel pattern including four pixels that are adjacent to each other in the angular direction from the screening image and in which each pixel is arranged as “white / black / white / black” is detected.

  When a third pixel pattern is detected, a third pixel fusion process is performed on the third pixel pattern. In the third pixel fusion process, the third pixel pattern arranged as “white / black / white / black” is converted into a pixel pattern arranged as “white / white / black / black”.

  Here, the third pixel pattern is detected at 45 ° intervals in a 360 ° space of the image, as shown in FIG. Specifically, in the third pixel fusion process, when the third pixel pattern is detected in the 0 ° direction, as shown in FIG. 13B, “white” is displayed in the 0 ° direction from the left to the right. A third pixel pattern aligned with “black / white / black” (hereinafter referred to as “third pixel pattern (0 °)”) is converted into a pixel pattern aligned with “white / white / black / black” ( Third pixel fusion process (0 °).

  When the third pixel pattern is detected in the 45 ° direction, as shown in FIG. 13C, the third pixel pattern aligned with “white / black / white / black” from the lower left to the upper right in the 45 ° direction. A pixel pattern (hereinafter referred to as “third pixel pattern (45 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (45 °)).

  When the third pixel pattern is detected in the 90 ° direction, as shown in FIG. 13D, the third pixel pattern aligned with “white / black / white / black” from the bottom to the top in the 90 ° direction. A pixel pattern (hereinafter referred to as “third pixel pattern (90 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (90 °)).

  When the third pixel pattern is detected in the 135 ° direction, as shown in FIG. 13E, the third pixel pattern in which “white / black / white / black” is aligned in the 135 ° direction from the lower right to the upper left is arranged. The pixel pattern (hereinafter referred to as “third pixel pattern (135 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (135 °)). .

  When the third pixel pattern is detected in the 180 ° direction, as shown in FIG. 13F, the third pixel pattern aligned with “white / black / white / black” from the right to the left in the 180 ° direction. A pixel pattern (hereinafter referred to as “third pixel pattern (180 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (180 °)).

  When the third pixel pattern is detected in the 225 ° direction, as shown in FIG. 13G, the third pixel pattern aligned with “white / black / white / black” from the upper right to the lower left in the 225 ° direction. A pixel pattern (hereinafter referred to as “third pixel pattern (225 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (225 °)).

  When the third pixel pattern is detected in the 270 ° direction, as shown in FIG. 13H, the third pixel pattern aligned with “white / black / white / black” from the top to the bottom in the 270 ° direction. A pixel pattern (hereinafter referred to as “third pixel pattern (270 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (270 °)).

  When the third pixel pattern is detected in the 315 ° direction, as shown in FIG. 13 (i), the third pattern in which “white / black / white / black” is arranged in the 315 ° direction from the upper left to the lower right is arranged. The pixel pattern (hereinafter referred to as “third pixel pattern (315 °)”) is converted into a pixel pattern aligned with “white / white / black / black” (third pixel fusion processing (315 °)). .

  Since the 360 ° direction is the same as the 0 ° direction, the third pixel pattern is not detected.

  Thus, in the third pixel fusion processing, the third pixel pattern is sequentially detected at 45 ° intervals with respect to the 360 ° space of the image, and the above-described conversion is performed until the third pixel pattern cannot be detected from the screening image. The process is repeated.

  And the image in which the 1st-3rd pixel fusion process was performed by the pixel fusion process in this way turns into the hybrid screening image which converted the FM screening image.

  FIG. 14 is a hybrid screening image created by performing one-dimensional pixel fusion processing on the screening image shown in FIG. By performing the pixel fusion process in this way, the number of isolated pixels can be reduced and converted into large dots formed by a plurality of pixels.

  Next, the roughness correction processing by the roughness correction processing unit 7 will be described with reference to FIGS. 15 and 16.

  As described above, the hybrid screening image created by performing each process on the input grayscale image data 10 has a strong randomness of the FM screening image, and thus may have a rough feeling as image quality. It is considered that this rough feeling is mainly caused by end point pixels and isolated point pixels present in the created hybrid screening image.

  Here, the end point pixel and the isolated point pixel are pixels defined by the correlation between the target pixel and pixels adjacent to the target pixel when a predetermined pixel is the target pixel.

  For example, the end point pixel indicates a target pixel when a plurality of continuous pixels (for example, five pixels or more) among the eight pixels adjacent to the periphery of the target pixel are pixels having a different color from the target pixel. Specifically, as illustrated in FIG. 15A, when the target pixel is a black pixel, and when five consecutive pixels among the eight pixels around the target pixel are white pixels, the target pixel is an end pixel. It becomes. In addition, when the target pixel is a white pixel, and when five consecutive pixels among the eight pixels around the target pixel are black pixels, the target pixel is an end point pixel.

  On the other hand, for example, an isolated point pixel indicates a target pixel when all the eight pixels adjacent to the periphery of the target pixel are pixels having a different color from the target pixel. Specifically, when the target pixel is a black pixel and all eight pixels around the target pixel are white pixels, the target pixel is an end point pixel. In addition, when the target pixel is a white pixel and all eight pixels around the target pixel are black pixels, the target pixel is an end point pixel.

  In the present embodiment, correction processing is performed to suppress the feeling of roughness and improve the image quality by reducing the end point pixels and the isolated point pixels.

  In the roughness correction process, first, a raster scan is performed on the hybrid screening image created by the pixel fusion processing unit 6. Next, based on the definition of the end point pixel and the isolated point pixel set in advance, it is determined whether or not the predetermined target pixel is the end point pixel or the isolated point pixel.

  When the target pixel is an end point pixel or an isolated point pixel, it is determined whether or not an end point pixel or an isolated point pixel having a color different from that of the target pixel exists in a predetermined area with the target pixel as a center. Specifically, for example, it is sequentially determined whether or not surrounding pixels that exist clockwise around the target pixel are end point pixels or isolated point pixels.

  When an end point pixel or isolated point pixel having a color different from that of the target pixel is detected, as shown in FIG. 15B, the target pixel that is the end point pixel or the isolated point pixel, and the detected end point pixel or Replace isolated point pixels.

  In this way, by replacing the end point pixels or isolated point pixels of different colors with each other, the end point pixels and the isolated point pixels that cause the rough feeling can be reduced. In addition, since the pixels of different colors are replaced in the replacement process, the number of black pixels and white pixels in the entire image can be maintained as the number before the replacement, so that the gray level information of the image before and after the process is maintained. However, the image quality can be improved by suppressing the feeling of roughness.

  FIG. 16 shows a hybrid screening image corrected by performing a roughness correction process on the hybrid screening image shown in FIG. FIG. 17 shows a comparison between an FM screening image created by this embodiment and a hybrid screening image. In FIGS. 17A to 17D, the left side shows the FM screening image, and the right side shows the hybrid screening image. FIGS. 17A to 17D show images when the density of black pixels is 10%, 25%, 50%, and 75%, respectively.

  As shown in FIGS. 16 and 17, by reducing the end point pixels and isolated point pixels present in the hybrid screening image, it is possible to suppress the feeling of roughness present in the image and improve the image quality.

  Next, the flow of each process in the image conversion processing apparatus 1 will be described with reference to the flowcharts shown in FIGS.

  First, processing of the entire image conversion processing apparatus 1 that converts input grayscale image data 10 input from the outside into output binary image data 20 and outputs it will be described with reference to FIG.

  When input grayscale image data 10 is input from the outside and stored in the memory 2, the binarization processing unit 3 reads the input grayscale image data 10 stored in the memory 2 and relates to the input grayscale image data 10. Based on the resolution information, a conversion block is set for the input grayscale image data 10. And the binarization process which detects the 1st black pixel number while converting the image for every conversion block into a 1st binary image is performed (step S1). The binarization processing unit 3 temporarily stores in the memory 2 the first binary image, the first black pixel number information, and the conversion error information obtained by the binarization process.

  Next, the conversion error variance processing unit 4 reads out the first binary image, the first black pixel number information, and the conversion error information stored in the memory 2, and based on these images and information, the first binary image is read out. A value image is converted into a second binary image, and a conversion error distribution process for detecting the second number of black pixels is performed (step S2). The conversion error variance processing unit 4 temporarily stores the second binary image and the second black pixel number information obtained by the conversion error variance processing in the memory 2.

  Next, the FM screening image creation processing unit 5 reads the second black pixel number information stored in the memory 2, and based on the number of black pixels for each conversion block indicated by the second black pixel number information, the conversion block An FM screening conversion process is performed to convert the second binary image, which is the input grayscale image data 10, into an FM screening image using the same size chessboard pixel arrangement diagram (step S3). The FM screening image creation processing unit 5 temporarily stores the FM screening image created by the FM screening conversion process in the memory 2.

  Next, the pixel fusion processing unit 6 reads out the FM screening image stored in the memory 2, and performs a two-dimensional pixel fusion process (first and second pixel fusion processing) on the FM screening image (Step 1). S4). The pixel fusion processing unit 6 temporarily stores the screening image created by the two-dimensional pixel fusion processing in the memory 2.

  Next, the pixel fusion processing unit 6 reads out the screening image created by the two-dimensional pixel fusion processing from the memory 2, and performs one-dimensional pixel fusion processing (third pixel fusion processing) on the screening image ( Step S5). The pixel fusion processing unit 6 temporarily stores the hybrid screening image created by the one-dimensional pixel fusion processing in the memory 2.

  Then, the roughness correction processing unit 7 reads out the hybrid screening image created by the one-dimensional pixel fusion processing from the memory 2, and performs the roughness correction processing on the hybrid screening image (step S6). As a result, the input grayscale image data 10 is converted into a hybrid screening image with reduced roughness. The roughness correction processing unit 7 stores the converted hybrid screening image in the memory 2. The hybrid screening image that has been subjected to the roughness correction process and is stored in the memory 2 is output as output binary image data 20.

  In the example shown in FIG. 18, it has been described that the processes of step S1, step S2, and step S3 are sequentially performed. However, the present invention is not limited to this. May be performed.

  Next, the flow of binarization processing and conversion error dispersion processing in steps S1 and S2 of FIG. 18 will be described with reference to FIG.

  First, in step S11, the binarization processing unit 3 calculates the number of black pixels s of the binary image using the above-described equation (1) based on the gradation value of each pixel included in the set conversion block. . Next, the binarization processing unit 3 adds a randomization weight coefficient to the number of black pixels s in the transform block using Equation (2), and calculates the number of black pixels S of the first binary image. (Step S12).

  Next, in step S <b> 13, the binarization processing unit 3 performs an error dispersion process by adding dispersion errors from transform blocks located around the transform block to be processed. Finally, the binarization processing unit 3 calculates the number of black pixels S ″ to be arranged in the conversion block by using Equation (3), and calculates the dispersion error ε for the surrounding conversion blocks (step S14). .

  Next, the flow of the FM screening conversion process in step S3 of FIG. 18 will be described with reference to FIG.

  First, in step S <b> 21, the FM screening image creation processing unit 5 reads the input grayscale image data 10 from the memory 2, and converts the input grayscale image data 10 into a conversion block based on the resolution information regarding the input grayscale image data 10. Set.

  Next, the FM screening image creation processing unit 5 determines the maximum coordinate values in the x direction and the y direction according to the set size of the conversion block (step S22). Further, the FM screening image creation processing unit 5 sets a chessboard type pixel arrangement diagram having the same size as the set conversion block.

  Then, the FM screening image creation processing unit 5 performs a raster scan for each conversion block on the input grayscale image data 10 (step S23), and detects the number of black pixels in each conversion block for each conversion block. (Step S24).

  Next, in step S <b> 25, the FM screening image creation processing unit 5 performs pixel arrangement processing based on a preset pixel arrangement determination condition A (in this example, the pixel arrangement determination condition is an even number).

  In step S <b> 26, the FM screening image creation processing unit 5 determines whether or not the pixel arrangement process based on the pixel arrangement determination condition A has been completed. When it is determined that the pixel arrangement processing based on the pixel arrangement determination condition A has been completed, all the black pixels are arranged at positions corresponding to the pixel arrangement determination condition A, and there are remaining black pixels to be arranged (step S26; Yes) ) Proceeds to step S27.

  On the other hand, if the FM screening image creation processing unit 5 determines that the pixel placement processing based on the pixel placement judgment condition A has not ended (step S26; No), the process returns to step S25, and the FM screening image creation processing is performed. The unit 5 repeats the process of step S25 until black pixels are arranged at all positions corresponding to the pixel arrangement determination condition A.

  In step S27, the FM screening image creation processing unit 5 changes the pixel arrangement determination condition to a pixel arrangement determination condition B (in this example, the pixel arrangement determination condition is an odd number), which is a condition different from the pixel arrangement determination condition A. Then, pixel arrangement processing based on the pixel arrangement determination condition B is performed.

  Finally, the FM screening image creation processing unit 5 determines whether all the black pixels to be arranged have been arranged and the arrangement of the black pixels has been completed (step S28). When the FM screening image creation processing unit 5 determines that the arrangement of the black pixels has been completed (step S28; Yes), a series of processing ends.

  On the other hand, when the FM screening image creation processing unit 5 determines that the arrangement of the black pixels is not completed (step S28; No), the process returns to step S27, and the FM screening image creation processing unit 5 The process of step S27 is repeated until black pixels are arranged.

  In this way, by creating FM screening images from grayscale images in units of transform blocks and creating FM screening images for all transform blocks, grayscale images can be converted into FM screening images.

  Next, the flow of arrangement processing in step S25 and step S27 in FIG. 20 will be described with reference to FIG.

  First, in step S31, the FM screening image creation processing unit 5 sets a pixel arrangement determination condition (for example, pixel arrangement determination condition A). Next, the number of black pixels detected in step S24 of FIG. 20 is set in the required arrangement pixel number counter, and the value of the arranged pixel number counter is reset to “0” (steps S32 and S33).

  In step S34, the FM screening image creation processing unit 5 generates a uniform random number for determining the coordinates of the arrangement pixel P based on the above-described equation (4), and determines the arrangement pixel coordinates (m, n). .

  Next, in step S35, the FM screening image creation processing unit 5 determines whether or not the pixel arrangement determination condition TS matches the pixel arrangement determination condition set in step S31 based on Expression (4). If the FM screening image creation processing unit 5 determines that the pixel arrangement determination condition TS matches (step S35; Yes), the process proceeds to step S36.

  On the other hand, when the FM screening image creation processing unit 5 determines that the pixel arrangement determination condition TS does not match (step S35; No), the FM screening image creation processing unit 5 sets the arrangement pixel coordinates (m, n) determined in step S34. ) And the process returns to step S34.

  In step S36, the FM screening image creation processing unit 5 determines whether or not the arrangement pixel has already been arranged at the arrangement pixel coordinate (m, n) determined in step S34. When it is determined that the arrangement pixel has not been arranged (step S36; No), the FM screening image creation processing unit 5 arranges the arrangement pixel at the arrangement pixel coordinate (m, n) and sets the value of the necessary arrangement pixel number counter. Only "1" is decremented (steps S37 and S38). On the other hand, when the FM screening image creation processing unit 5 determines that the arrangement pixel has been arranged (step S36; Yes), the processing returns to step S34.

  Next, in step S <b> 39, the FM screening image creation processing unit 5 determines whether or not the value of the required arrangement pixel number counter is larger than “0”. If it is determined that the value of the required arrangement pixel number counter is larger than “0” (step S39; Yes), the FM screening image creation processing unit 5 increments the value of the arranged pixel number counter by “1” (step S39). S40).

  On the other hand, when the FM screening image creation processing unit 5 determines that the value of the required arrangement pixel number counter is “0” or less (step S39; No), the FM screening image creation processing unit 5 arranges all the arrangement pixels. Is completed, a series of processing ends.

  In step S41, the FM screening image creation processing unit 5 compares the value of the arranged pixel number counter with the total number of arranged pixels corresponding to the maximum coordinate value obtained in step S22 of FIG. It is determined whether or not the counter value is equal to or greater than the total number of arranged pixels.

  When the arranged pixel number counter value is equal to or greater than the total number of arranged pixels (step S41; Yes), the FM screening image creation processing unit 5 determines that all the arranged pixels are arranged, and determines the pixel arrangement set in step S31. The condition is changed to a different pixel arrangement determination condition (for example, pixel arrangement determination condition B) (step S42). Then, the process returns to step S34.

  On the other hand, when the FM screening image creation processing unit 5 determines in step S41 that the arranged pixel number counter value is less than the total arranged pixel number (step S41; No), the process returns to step S34.

  In this way, by arranging the black pixels as the arrangement pixels on the chess board type pixel arrangement diagram, an FM screening image can be created in units of conversion blocks.

  Next, the flow of the two-dimensional pixel fusion processing (first and second pixel fusion processing) in step S4 of FIG. 18 will be described with reference to FIG.

  First, in step S51, the pixel fusion processing unit 6 reads an FM screening image that is a processing target image from the memory 2 and performs raster scanning. Next, the pixel fusion processing unit 6 determines whether or not the first pixel pattern is detected from the “2 pixels × 2 pixels” pixel pattern obtained in step S51 (step S52).

  If it is determined that the first pixel pattern is detected (step S52; Yes), in step S53, the pixel fusion processing unit 6 generates a random number. In this example, a uniform random number in which all real numbers appear with the same probability in the interval from “0” to “1” is used as the random number.

  Next, in step S54, the pixel fusion processing unit 6 determines whether or not the value of the random number generated in step S53 is less than “0.5”. When the random value is 0.5 or more (step S54; No), the pixel fusion processing unit 6 performs the first operation on the first pixel pattern detected in step S52 as shown in FIG. 10C. The pixel fusion processing 100a is performed (step S55).

  If the random number value is less than 0.5 (step S54; Yes), the pixel fusion processing unit 6 applies the first pixel pattern detected in step S52 as shown in FIG. A first pixel fusion process 100b is performed (step S56).

  On the other hand, when the pixel fusion processing unit 6 determines in step S52 that the first pixel pattern is not detected (step S52; No), the process proceeds to step S57.

  In step S57, the pixel fusion processing unit 6 determines whether or not the second pixel pattern is detected from the “2 pixels × 2 pixels” pixel pattern obtained in step S51. If it is determined that the second pixel pattern has been detected (step S57; Yes), the pixel fusion processing unit 6 generates a uniform random number having a value from “0” to “1” (step S58).

  Next, the pixel fusion processing unit 6 determines whether or not the value of the random number generated in step S58 is less than “0.5” (step S59). When the random number value is 0.5 or more (step S59; No), the pixel fusion processing unit 6 performs the second operation on the second pixel pattern detected in step S57 as shown in FIG. The pixel fusion process 200a is performed (step S60).

  On the other hand, when the random number value is less than 0.5 (step S59; Yes), the pixel fusion processing unit 6 applies the second pixel pattern detected in step S57 as shown in FIG. A second pixel fusion process 200b is performed (step S61).

  On the other hand, if the pixel fusion processing unit 6 determines in step S57 that the second pixel pattern has not been detected (step S57; No), the process proceeds to step S62.

  Next, in step S62, the pixel fusion processing unit 6 determines whether scanning for the FM screening image is completed. If the pixel fusion processing unit 6 determines that the scan has ended (step S62; Yes), the process proceeds to step S63. On the other hand, when the pixel fusion processing unit 6 determines that the scan has not ended (step S62; No), the process returns to step S51, and the scan for the FM screening image is continued.

  In step S63, the pixel fusion processing unit 6 determines whether or not the first and second pixel patterns cannot be detected from the processing target image. When the pixel fusion processing unit 6 determines that the first and second pixel patterns cannot be detected (step S63; Yes), the series of processing ends. On the other hand, when the pixel fusion processing unit 6 determines that the first and second pixel patterns can be detected (step S63; No), the process returns to step S51.

  As described above, the pixel fusion processing unit 6 performs the one-dimensional pixel fusion processing by performing the two-dimensional pixel fusion processing on the FM screening image created by the FM screening image creation processing unit 5. To be screened and stored in the memory 2.

  Next, the flow of the one-dimensional pixel fusion process (third pixel fusion process) in step S5 of FIG. 18 will be described with reference to FIG.

  First, in step S71, the pixel fusion processing unit 6 reads out a screening image created by the two-dimensional pixel fusion processing, which is a processing target image, from the memory 2, and performs a raster scan.

  Next, the pixel fusion processing unit 6 determines whether or not a third pixel pattern (0 °) in the 0 ° direction is detected from the pixel pattern obtained in step S71 (step S72). When it is determined that the third pixel pattern (0 °) has been detected (step S72; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (0 °) as shown in FIG. 13B. Then, a third pixel fusion process (0 °) is performed (step S73).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (0 °) is not detected (step S72; No), the process proceeds to step S74.

  Next, in step S74, the pixel fusion processing unit 6 determines whether or not a third pixel pattern (45 °) in the 45 ° direction has been detected. When it is determined that the third pixel pattern (45 °) has been detected (step S74; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (45 °) as shown in FIG. Then, a third pixel fusion process (45 °) is performed (step S75).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (45 °) is not detected (step S74; No), the process proceeds to step S76.

  Next, in step S76, the pixel fusion processing unit 6 determines whether a third pixel pattern (90 °) in the 90 ° direction is detected. When it is determined that the third pixel pattern (90 °) has been detected (step S76; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (90 °) as shown in FIG. Is subjected to a third pixel fusion process (90 °) (step S77).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (90 °) is not detected (step S76; No), the process proceeds to step S78.

  Next, in step S78, the pixel fusion processing unit 6 determines whether a third pixel pattern (135 °) in the 135 ° direction has been detected. When it is determined that the third pixel pattern (135 °) has been detected (step S78; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (135 °) as shown in FIG. Then, a third pixel fusion process (135 °) is performed (step S79).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (135 °) is not detected (step S78; No), the process proceeds to step S80.

  Next, in step S80, the pixel fusion processing unit 6 determines whether or not a third pixel pattern (180 °) in the 180 ° direction has been detected. If it is determined that the third pixel pattern (180 °) has been detected (step S80; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (180 °) as shown in FIG. Then, a third pixel fusion process (180 °) is performed (step S81).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (180 °) is not detected (step S80; No), the process proceeds to step S82.

  Next, in step S82, the pixel fusion processing unit 6 determines whether or not a third pixel pattern (225 °) in the 225 ° direction has been detected. When it is determined that the third pixel pattern (225 °) has been detected (step S82; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (225 °) as shown in FIG. Is subjected to the third pixel fusion processing (225 °) (step S83).

  On the other hand, if the pixel fusion processing unit 6 determines that the third pixel pattern (225 °) has not been detected (step S82; No), the process proceeds to step S84.

  Next, in step S84, the pixel fusion processing unit 6 determines whether or not a third pixel pattern (270 °) in the 270 ° direction has been detected. When it is determined that the third pixel pattern (270 °) has been detected (step S84; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (270 °) as shown in FIG. Then, a third pixel fusion process (270 °) is performed (step S85).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (270 °) is not detected (step S84; No), the process proceeds to step S86.

  Next, in step S86, the pixel fusion processing unit 6 determines whether or not a third pixel pattern (315 °) in the 315 ° direction has been detected. If it is determined that the third pixel pattern (315 °) has been detected (step S86; Yes), the pixel fusion processing unit 6 performs the third pixel pattern (315 °) as shown in FIG. 13 (i). Is subjected to the third pixel fusion processing (315 °) (step S87).

  On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern (315 °) is not detected (step S86; No), the process proceeds to step S88.

  Next, in step S88, the pixel fusion processing unit 6 determines whether scanning for the screening image is completed. If the pixel fusion processing unit 6 determines that the scan has ended (step S88; Yes), the process proceeds to step S89. On the other hand, when the pixel fusion processing unit 6 determines that the scan has not ended (step S88; No), the process returns to step S71, and the pixel fusion processing unit 6 continues scanning the screening image.

  In step S89, the pixel fusion processing unit 6 determines whether or not the third pixel pattern cannot be detected from the processing target image. When the pixel fusion processing unit 6 determines that the third pixel pattern cannot be detected (step S89; Yes), a series of processing ends. On the other hand, when the pixel fusion processing unit 6 determines that the third pixel pattern can be detected (step S89; No), the process returns to step S71.

  As described above, the pixel fusion processing unit 6 creates a hybrid screening image by performing the one-dimensional pixel fusion processing on the screening image created by performing the two-dimensional pixel fusion processing, and stores it in the memory 2. To do.

  Next, the flow of the roughness correction process in step S6 of FIG. 18 will be described with reference to FIG.

  First, in step S91, the roughness correction processing unit 7 reads out the hybrid screening image created by the one-dimensional pixel fusion processing that is the processing target image from the memory 2, and performs raster scanning.

  Next, the roughness correction processing unit 7 determines whether or not scanning for all the pixels in the hybrid screening image has been completed (step S92). When the roughness correction processing unit 7 determines that scanning for all the pixels has not been completed (step S92; No), the process proceeds to step S93. On the other hand, when the roughness correction processing unit 7 determines that the scanning for all the pixels has been completed (step S92; Yes), a series of processing ends.

  In step S93, the roughness correction processing unit 7 determines whether an end point pixel or an isolated point pixel is detected. When it is determined that an end point pixel or an isolated point pixel has been detected (step S93; Yes), the roughness correction processing unit 7 sets a detection range centering on the target pixel (step S94), and based on a predetermined detection method, An end point pixel or isolated point pixel having a color different from that of the pixel of interest is detected (step S95). On the other hand, when the roughness correction processing unit 7 determines that no end point pixel or isolated point pixel has been detected (step S93; No), the process returns to step S91.

  Next, in step S <b> 96, the roughness correction processing unit 7 determines whether an end point pixel or an isolated point pixel having a color different from that of the target pixel is detected. When it is determined that an end point pixel or an isolated point pixel having a color different from that of the target pixel has been detected (step S96; Yes), the roughness correction processing unit 7 replaces the target pixel with the detected end point pixel or isolated point pixel. Processing is performed (step S97). Then, the process returns to step S91.

  On the other hand, when the roughness correction processing unit 7 determines that no end point pixel or isolated point pixel having a color different from that of the pixel of interest has been detected (step S96; No), the process returns to step S95.

  FIG. 25A shows a halftone image obtained by converting the input grayscale image data 10 input to the image conversion processing apparatus 1 according to the present embodiment by a conventional method, and FIG. The output binary image data 20 (hybrid screening image) obtained by performing the processing shown in FIGS. 18 to 24 described above on the grayscale image data 10 is shown. FIG. 26 shows a gradation image created by the hybrid screening image creation method by the image conversion processing device 1 according to the present embodiment.

  As described above, the image conversion processing apparatus 1 according to the present embodiment can create a hybrid screening image suitable for a print image with dots smaller than halftone dots that are randomly distributed without periodicity. That is, in the image conversion processing device 1 according to the present embodiment, a hybrid screening image having the advantages of the conventional halftone dot and FM screening can be created.

  As described above, in the image conversion processing device according to the present embodiment, the first grayscale image that is composed of black pixels and white pixels using the randomized weighting coefficient that can be adjusted to improve the image quality of the input grayscale image. Convert to binary image. Next, the first black pixel number included in the converted first binary image is detected for each conversion block, and an error (conversion error) that occurs when the first black pixel number is detected is taken into consideration. A second binary image is created. Then, the number of black pixels included in the created second binary image is detected for each change block, an FM screening image is created based on the detected number of black pixels, and the created FM screening image First to third pixel fusion processing is performed to create a hybrid screening image. In addition, the roughness of the hybrid screening image created by the pixel fusion process is corrected to create a hybrid screening image that is an output image.

  This makes it possible to improve the image quality, create a hybrid screening image that combines the advantages of being suitable for the printing process of the halftone image and easy to handle, and the high quality of the FM screening image. can do.

  In addition, although a part or all of said embodiment can be described also as the following additional remarks, it is not restricted to the following.

(Appendix 1)
An image conversion processing method for converting a multi-value image having multi-value gradation values into a hybrid screening image,
The multi-valued image is converted into a first binary image composed of black pixels and white pixels using a randomized weighting factor that can be adjusted to improve the image quality, and the first binary image included in the first binary image A binarization step for detecting the number of black pixels and detecting a conversion error that occurs when the first binary image is created;
Based on the first number of black pixels and the conversion error, the first binary image is converted into a second binary image, and the second number of black pixels included in the second binary image is determined. A conversion error distribution step to detect;
An FM screening image creation step of converting the second binary image into an FM screening image based on the detected number of the second black pixels;
A predetermined pixel pattern formed by a plurality of adjacent pixels is detected from the created FM screening image, and a predetermined pixel fusion process is performed on the detected predetermined pixel pattern to create a hybrid screening image. And a pixel fusion step.

(Appendix 2)
The binarization step includes
Based on the resolution information of the grayscale image, a conversion block corresponding to the size of the halftone cell of the grayscale image, which is a creation unit of the FM screening image, is set for the grayscale image,
The number of black pixels in the conversion block is detected, and the random weighting coefficient is added to the detected number of black pixels, whereby the first black pixel number in the conversion block is determined for each conversion block. The image conversion processing method according to attachment 1, wherein the image conversion processing method is detected.

(Appendix 3)
The image conversion processing method according to appendix 2, wherein the conversion error distribution step distributes the conversion error in the predetermined conversion block to surrounding conversion blocks adjacent to the conversion block.

(Appendix 4)
The FM screening image creation step includes:
Set horizontal and vertical coordinates for all pixels in the transform block;
The image conversion processing method according to appendix 2 or 3, wherein the FM screening image is created by arranging the black pixel at a coordinate position determined based on a random number generated corresponding to the coordinate.

(Appendix 5)
The FM screening image creation step includes:
The image conversion processing method according to appendix 4, wherein the black pixels are arranged in the order of the total value of the horizontal and vertical coordinates based on the random number from even to odd, or from odd to even.

(Appendix 6)
The pixel fusion step includes:
From the created FM screening image, the predetermined pixel pattern that is formed by a plurality of adjacent pixels and in which the black pixels and the white pixels are distributed at intervals of one pixel is detected.
The image conversion according to any one of appendices 1 to 5, wherein the pixel fusion processing is performed to exchange the black pixels and the white pixels so that the detected black pixels in the predetermined pixel pattern are continuously distributed. Processing method.

(Appendix 7)
The pixel fusion step includes:
A pixel pattern in which black pixels and white pixels are distributed at intervals of one pixel is detected from a pixel pattern formed by four adjacent pixels of “2 pixels × 2 pixels”, and the black pattern in the detected pixel pattern is detected. A two-dimensional pixel fusion process for exchanging the black pixels and the white pixels so that the pixels are continuously distributed;
A pixel pattern in which black pixels and white pixels are distributed at intervals of one pixel is detected from a pixel pattern formed by four pixels adjacent to each other in the angular direction for the image subjected to the two-dimensional pixel fusion processing. The image conversion processing method according to appendix 6, wherein a one-dimensional pixel fusion process is performed in which the black pixels and the white pixels are exchanged so that the black pixels in the pixel pattern are continuously distributed.

(Appendix 8)
The two-dimensional pixel fusion process
When the first pixel pattern in which the upper left and lower right pixels are black pixels and the upper right and lower left pixels are white pixels is detected, the lower right black pixels are replaced with the upper right or lower left white pixels. First pixel fusion processing to
When the second pixel pattern in which the upper left and lower right pixels are white pixels and the upper right and lower left pixels are black pixels is detected, the lower right white pixels and the upper right or lower left black pixels are exchanged. And performing a second pixel fusion process
The image conversion processing method according to appendix 7, wherein the first and second pixel fusion processes are alternately performed until the first and second pixel patterns are no longer detected.

(Appendix 9)
The one-dimensional pixel fusion process is:
When a third pixel pattern composed of four pixels adjacent to each other in the angular direction and arranged in the form of white pixels, black pixels, white pixels, and black pixels is detected, the third pixel pattern is converted into a white pixel and a white pixel. -Perform a third pixel fusion process to convert to a pixel pattern lined up like black pixels / black pixels,
The image conversion processing method according to appendix 7 or 8, wherein the third pixel fusion processing is sequentially performed at 45 ° intervals until the third pixel pattern is not detected.

(Appendix 10)
The image conversion processing method according to any one of supplementary notes 1 to 9, further comprising a roughness correction step for performing a roughness correction process on the hybrid screening image created in the pixel fusion step.

(Appendix 11)
The roughness correction step includes
Detecting end point pixels or isolated point pixels in the hybrid screening image;
The image according to appendix 10, wherein, when an end point pixel or an isolated point pixel having a color different from that of the detected end point pixel or isolated point pixel is detected from pixels around the detected end point pixel or isolated point pixel, the pixels are replaced with each other. Conversion processing method.

(Appendix 12)
The end point pixel is a color in which a plurality of continuous pixels among eight adjacent pixels are different from each other,
The image conversion processing method according to attachment 11, wherein the isolated point pixel is a color in which all eight adjacent pixels have different colors.

(Appendix 13)
A program for causing a computer device to execute an image conversion processing method for converting a grayscale image into a hybrid screening image,
The multi-valued image is converted into a first binary image composed of black pixels and white pixels using a randomized weighting factor that can be adjusted to improve the image quality, and the first binary image included in the first binary image A binarization step for detecting the number of black pixels and detecting a conversion error that occurs when the first binary image is created;
Based on the first number of black pixels and the conversion error, the first binary image is converted into a second binary image, and the second number of black pixels included in the second binary image is determined. A conversion error distribution step to detect;
An FM screening image creation step of converting the second binary image into an FM screening image based on the detected number of the second black pixels;
A predetermined pixel pattern formed by a plurality of adjacent pixels is detected from the created FM screening image, and a predetermined pixel fusion process is performed on the detected predetermined pixel pattern to create a hybrid screening image. A program that causes a computer apparatus to execute an image conversion processing method including a pixel fusion step.

(Appendix 14)
An image conversion processing device for converting a multi-value image having multi-value gradation values into a hybrid screening image,
The multi-valued image is converted into a first binary image composed of black pixels and white pixels using a randomized weighting factor that can be adjusted to improve the image quality, and the first binary image included in the first binary image A binarization processing unit that detects the number of black pixels and detects a conversion error that occurs when creating the first binary image;
Based on the first number of black pixels and the conversion error, the first binary image is converted into a second binary image, and the second number of black pixels included in the second binary image is determined. A conversion error variance processing unit to detect;
An FM screening image creation processing unit that converts the second binary image into an FM screening image based on the detected number of the second black pixels;
A predetermined pixel pattern formed by a plurality of adjacent pixels is detected from the created FM screening image, and a predetermined pixel fusion process is performed on the detected predetermined pixel pattern to create a hybrid screening image. An image conversion processing device comprising a pixel fusion processing unit.

DESCRIPTION OF SYMBOLS 1 Image conversion processing apparatus 2 Memory 3 Binarization process part 4 Conversion error dispersion | distribution process part 5 FM screening image creation process part 6 Pixel fusion process part 7 Roughness correction process part 10 Input grayscale image data 20 Output binary image data

Claims (9)

An image conversion processing method for converting a grayscale image into a hybrid screening image,
Based on the resolution information of the grayscale image, a conversion block corresponding to the size of the halftone cell of the grayscale image is set for the grayscale image, and the grayscale image is converted into a black pixel and a white for each conversion block. Convert to a first binary image consisting of pixels, detect the number of black pixels in the conversion block, and add an adjustable randomization weight coefficient to improve the image quality for the detected number of black pixels by a first of the number of black pixels detected for each transform block, binarization step of detecting a conversion error occurring during the creation of the first binary image of the transform block,
Based on the first number of black pixels and the conversion error, the first binary image is converted into a second binary image, and the second number of black pixels included in the second binary image is determined. A conversion error distribution step to detect;
An FM screening image creation step of converting the second binary image into an FM screening image based on the detected number of the second black pixels;
A predetermined pixel pattern formed by a plurality of adjacent pixels is detected from the created FM screening image, and a predetermined pixel fusion process is performed on the detected predetermined pixel pattern to create a hybrid screening image. An image conversion processing method comprising: a pixel fusion step. 2. The image conversion processing method according to claim 1 , wherein the conversion error distribution step distributes a conversion error in the predetermined conversion block to surrounding conversion blocks adjacent to the conversion block. The FM screening image creation step includes:
Set horizontal and vertical coordinates for all pixels in the transform block;
The determined coordinate position based on the random number is generated in correspondence with the coordinates, by arranging the black pixels, the image conversion process according to claim 1 or 2, characterized in that to create the FM screening image Method. The pixel fusion step includes:
From the created FM screening image, the predetermined pixel pattern that is formed by a plurality of adjacent pixels and in which the black pixels and the white pixels are distributed at intervals of one pixel is detected.
Claim 1, 2 or 3, wherein the black pixels of the detected within the predetermined pixel pattern and performing the pixel fusion process the black pixels and the white pixels to continuously distributed to interchangeable The image conversion processing method described in 1. The pixel fusion step includes:
A pixel pattern in which black pixels and white pixels are distributed at intervals of one pixel is detected from a pixel pattern formed by four adjacent pixels of “2 pixels × 2 pixels”, and the black pattern in the detected pixel pattern is detected. A two-dimensional pixel fusion process for exchanging the black pixels and the white pixels so that the pixels are continuously distributed;
A pixel pattern in which black pixels and white pixels are distributed at intervals of one pixel is detected from a pixel pattern formed by four pixels adjacent to each other in the angular direction for the image subjected to the two-dimensional pixel fusion processing. 5. The image conversion process according to claim 4 , wherein one-dimensional pixel fusion processing is performed in which the black pixels and the white pixels are exchanged so that the black pixels in the pixel pattern are continuously distributed. Method. Image conversion processing method according to any one of claims 1 to 5, further comprising a correction step roughness performs correction of roughness for said hybrid screen image created by the pixel fusion step. The roughness correction step includes
Detecting end point pixels or isolated point pixels in the hybrid screening image;
The pixel is replaced with each other when an end point pixel or an isolated point pixel having a color different from that of the detected end point pixel or isolated point pixel is detected from pixels around the detected end point pixel or isolated point pixel. Item 7. The image conversion processing method according to Item 6 . A program for causing a computer device to execute an image conversion processing method for converting a grayscale image into a hybrid screening image,
Based on the resolution information of the grayscale image, a conversion block corresponding to the size of the halftone cell of the grayscale image is set for the grayscale image, and the grayscale image is converted into a black pixel and a white for each conversion block. Convert to a first binary image consisting of pixels, detect the number of black pixels in the conversion block, and add an adjustable randomization weight coefficient to improve the image quality for the detected number of black pixels by a first of the number of black pixels detected for each transform block, binarization step of detecting a conversion error occurring during the creation of the first binary image of the transform block,
Based on the first number of black pixels and the conversion error, the first binary image is converted into a second binary image, and the second number of black pixels included in the second binary image is determined. A conversion error distribution step to detect;
An FM screening image creation step of converting the second binary image into an FM screening image based on the detected number of the second black pixels;
A predetermined pixel pattern formed by a plurality of adjacent pixels is detected from the created FM screening image, and a predetermined pixel fusion process is performed on the detected predetermined pixel pattern to create a hybrid screening image. A computer program causing a computer apparatus to execute an image conversion processing method including a pixel fusion step. An image conversion processing device for converting a grayscale image into a hybrid screening image,
Based on the resolution information of the grayscale image, a conversion block corresponding to the size of the halftone cell of the grayscale image is set for the grayscale image, and the grayscale image is converted into a black pixel and a white for each conversion block. Convert to a first binary image consisting of pixels, detect the number of black pixels in the conversion block, and add an adjustable randomization weight coefficient to improve the image quality for the detected number of black pixels by the said converting first detects the number of black pixels for each transform block in the block, the first detecting a conversion error occurring when creating a binary image binarizing processor,
Based on the first number of black pixels and the conversion error, the first binary image is converted into a second binary image, and the second number of black pixels included in the second binary image is determined. A conversion error variance processing unit to detect;
An FM screening image creation processing unit that converts the second binary image into an FM screening image based on the detected number of the second black pixels;
A predetermined pixel pattern formed by a plurality of adjacent pixels is detected from the created FM screening image, and a predetermined pixel fusion process is performed on the detected predetermined pixel pattern to create a hybrid screening image. An image conversion processing device comprising: a pixel fusion processing unit.