JP2007208638A - Image processing method, image processing device, image recording device, program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically optimize the dither design problem, taking into consideration granularity, gradation, texture and stability. <P>SOLUTION: An adaptation degree calculating means 2 obtains the adaptation degree, by evaluating a plurality of characteristics, such as the granularity, gradation, stability and texture of dither, while output level corresponding to the gradation of a dither matrix is used. In obtaining the adaptation degree, a point is reduced, when one of the plurality of evaluation characteristics is very much better than the other evaluation items to prevent being captured by local resolution. An adaptation degree selecting means 3 selects a dither matrix, having the highest adaptation degree, and outputs it to an output means 4. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image processing method, an image processing apparatus, an image recording apparatus (image forming apparatus), a program, and a recording medium for printing multi-value image data with high definition and high gradation.

  There is an image input / output system that outputs multi-valued image data read by an input device such as a scanner or a digital camera to an output device such as a printer or a display. At that time, multi-valued image data read by the input device (for example, 256 gradations for 8-bit accuracy) is converted into image data of the number of gradations that can be output by the output device, and pseudo continuous gradation is expressed. As a method for doing this, there is a pseudo halftone process.

  In particular, when the output device can express only binary values of ON / OFF of dots, a halftone image is binarized to create a pseudo halftone image. As such pseudo halftone creating means, various dither methods for binarizing an image according to a threshold matrix (dither matrix) table are widely used. There is also a multi-value dither process applied to the number of gradations of three or more values. In the dither processing, only a comparison operation is performed with a threshold corresponding to the input value and the input pixel position of the dither matrix, and high-speed processing is possible.

  The dither design is performed as follows. A basic matrix (hereinafter referred to as a cell) is determined in accordance with the desired number of dither lines and angles. In many cases, the number of gradations that can be expressed by a cell is small, and it is preferable that the shape is rectangular for processing by a computer. Therefore, a supercell is designed so that a plurality of cells are combined to be periodic and rectangular.

  In addition, the dither characteristics vary greatly depending on the order in which the dots in the cell are grown. In general, there are Bayer type, center-of-gravity preservation type, spiral type, line type, etc. for dither growth. As a rule of thumb, Bayer-type ink jets, etc., where dots are stable, and when the dots to be developed change due to fluctuations in the development potential, such as electrophotography, the center-of-gravity / swirl type or line-type are better. It was. The dot growth order in the cell has been designed based on dither designers' empirical rules or geometric rules.

  By the way, an image output by a printer such as an electrophotographic or inkjet printer is captured by a scanner, or an image created by simulation is converted into a Fourier space and various filters are used to quantify and quantify the image quality of the output image. Evaluation has become possible.

  Patent Document 1 selects dot reflectance distribution data for a pixel of interest from a plurality of types of dot reflectance distribution data according to the state of neighboring pixel data of the pixel of interest of input image data, and selects each pixel. This is a technique that can obtain a simulation image by calculating the reflectance distribution of an image that will be output when input image data is given to the image recording apparatus using the dot reflectance distribution data that has been obtained. .

  Patent Document 2 is a method in which the evaluation method such as RMS granularity is developed from the evaluation of granularity in a photograph, and is not effective for a pseudo halftone image in which periodicity of processing is perceived regardless of the pattern. On the other hand, the technique relates to a technique for evaluating the granularity in consideration of the periodicity of dither. Further, Patent Document 3 frequency-converts a dither pattern of a certain gradation and a smoothed version thereof, obtains a difference value of frequency components from both, and uses the multi-channel model of the human visual system for the difference value. This technique relates to a technique for evaluating a dither texture by weighting with a plurality of spatial filter groups based on.

  Depending on the output device, the dot reproduction instability in the highlight portion and the relationship between the dot gain input value and the average density of the output patch may not be linear from the highlight portion to the dark portion. In such a case, the gradation is evaluated by calculating a deviation from the ideal average density of each gradation using a linear expression of the gradation and the average density obtained from the average density of the highlight portion and the dark portion. be able to. In addition, the gradation can be evaluated by obtaining the average number of dots in each gradation from a linear expression of gradation and average density in advance.

  Although there are problems such as banding due to fluctuations in development potential and driving vibration as in electrophotography, it has been empirically known that such problems are less affected by concentration of dots. Stability can be evaluated by obtaining the perimeter of a dot from a simulation image as in Patent Document 1 or a binary image obtained from dither.

  Quasi-optimization has become possible by using these simulation images and evaluation methods. Patent Document 4 discloses a dither matrix optimization method for improving graininess by using a quasi-optimization method such as a genetic algorithm (hereinafter referred to as GA) or stochastic annealing. Patent Document 5 discloses a dither matrix optimization technique using an electrophotographic simulation image and GA.

JP 2001-127997 A JP 2002-245465 A JP 2004-272565 A JP-A-7-177351 JP 2000-152004 A

  However, when quasi-optimization is performed using a plurality of evaluations in the dither matrix optimization method, one of the plurality of evaluation items becomes good, and the other items do not become good, so that a so-called local solution is used. There is a case. In addition, among the plurality of evaluation items, there are items having contrary characteristics. Specifically, when the dither matrix is optimized as a desired gradation characteristic, the gradation characteristic having a pair of input value and average density as shown in FIG. It becomes the number of patterns. On the other hand, in the evaluation such as graininess and stability, it becomes a binarization threshold value at which all dots are output at a certain gradation, and the number of possible patterns is optimized to be two patterns of white and black. End up. In the optimization, a function for maintaining the number of patterns having a desired gradation characteristic, that is, the number of gradations having a desired gradation characteristic has been required. That is, in the automatic optimization of the dither matrix, it has become necessary to have a function that is not constrained by local solutions and that maintains the number of gradations of the desired gradation characteristics.

The present invention has been made in view of such problems,
An object of the present invention is to provide an image processing method, an image processing apparatus, an image recording apparatus, a program, and a recording medium capable of automatically optimizing a dither design problem in consideration of graininess, gradation, texture, stability, and the like. is there.

  In particular, it is an object of the present invention to perform N-value dither matrix optimization that is not constrained by local solutions and that is optimal for a printing press that is assumed by processing using simulation images.

  Further, the inventions of claims 2 and 4 are less likely to be caught by a local solution, and an object of the invention is to perform N-value dither matrix optimization in a single time process by a process using an N-value pattern image.

  Further, the invention according to claim 5 maintains that the number of gradations of a desired gradation characteristic is maintained, and that N-value dither matrix optimization optimal for a printing press assumed by processing using a simulation image is performed. Objective.

  According to the sixth aspect of the invention, it is possible to maintain the number of gradations of a desired gradation characteristic, and to perform N-value dither matrix optimization in a single time process by a process using an N-value pattern image. Objective.

  Another object of the present invention is to optimize a good dither matrix by using a genetic algorithm that is considered not to be caught by a local solution as an optimization step.

  In order to solve the above problems, the image processing method, the image processing apparatus, the image recording apparatus, the program, and the recording medium according to the present invention are not constrained by the local solution and have the desired gradation characteristics in the automatic optimization of the dither matrix. By adding a function for maintaining the number of gradations, a dither matrix with good image quality is obtained.

  In order to achieve this object, according to the first aspect of the present invention, an N-value dither output level corresponding to a certain gradation of an N-value (M> N ≧ 2) dither matrix is calculated from a given M value. A value dither output level simulation step, an evaluation step for evaluating a plurality of characteristics such as graininess, gradation, stability, texture, etc. of the N value dither using the N value dither output level at a plurality of gradations, and the plurality In the image processing method (N-value dither matrix optimization method) comprising the step of obtaining the fitness of the N-value dither matrix information based on the evaluation and the optimization step of changing the N-value dither matrix based on the fitness. The step of determining the fitness is characterized in that a point is deducted when one of a plurality of evaluation characteristics is extremely better than the other evaluation items.

  The invention according to claim 2 is an N-value dither pattern calculation step for calculating an N-value dither pattern according to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value, An evaluation process for evaluating a plurality of characteristics such as graininess, gradation, stability, and texture of N-value dither using the N-value dither pattern in gradation, and the fitness of N-value dither matrix information based on the plurality of evaluations And an optimization process for changing the N-value dither matrix based on the fitness, the evaluation characteristics include a plurality of steps for obtaining the fitness. One of the features is that points are deducted when they are extremely better than the other evaluation items.

  The invention according to claim 3 is an N-value dither output level simulation step of calculating an N-value dither output level according to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value, A gradation evaluation step for evaluating whether a dither having a desired gradation characteristic is performed using the N-value dither output level, and a dither having a desired granular characteristic is evaluated using the N-value dither output level. A granularity evaluation step, a texture evaluation step for evaluating whether the dithering has a desired texture characteristic using the N-value dither output level, and a dither having a desired stability using the N-value dither output level. A step of evaluating the stability, a step of obtaining a gradation degree from each of the gradation evaluations in a plurality of gradations, and a granularity evaluation of each in a plurality of gradations A step of obtaining a degree of texture, a step of obtaining a texture from each of the texture evaluations at a plurality of gradations, a step of obtaining a stability from each of the stability evaluations at a plurality of gradations, and the gradation and granularity An image processing method (N-value dither matrix optimization) comprising a step of obtaining the adaptability of N-value dither matrix information from texture and stability and an optimization step of changing the N-value dither matrix based on the fitness In the method), the step of obtaining the fitness is characterized in that a point is deducted when one of a plurality of evaluation items is extremely better than the other evaluation items.

  According to a fourth aspect of the present invention, there is provided an N-value dither pattern calculating step for calculating an N-value dither pattern corresponding to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value, Gradation evaluation step for evaluating whether the dither has a desired gradation characteristic using a value dither pattern, and granularity evaluation step for evaluating whether the dither has a desired granular characteristic using the N-value dither pattern A texture evaluation step for evaluating whether the dither pattern has a desired texture characteristic using the N-value dither pattern, and a stability evaluation for evaluating whether the dither has a desired stability using the N-value dither pattern. A step of obtaining a degree of gradation from each of the gradation evaluations in a plurality of gradations, a step of obtaining a granularity from each of the granularity evaluations in a plurality of gradations, A step of obtaining a texture degree from each of the texture evaluations at a plurality of gradations, a step of obtaining a stability from the respective stability evaluations at a plurality of gradations, and N from the gradation degree, granularity, texture degree, and stability. In an image processing method (N-value dither matrix optimization method) comprising a step of obtaining a fitness of value dither matrix information and an optimization step of changing an N-value dither matrix based on the fitness. The step to be obtained is characterized in that points are deducted when one of a plurality of evaluation items is extremely better than the other evaluation items.

  The invention according to claim 5 is an N-value dither output level simulation step of calculating an N-value dither output level corresponding to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value, A gradation evaluation step for evaluating whether a dither having a desired gradation characteristic is performed using the N-value dither output level, and a dither having a desired granular characteristic is evaluated using the N-value dither output level. A granularity evaluation step, a texture evaluation step for evaluating whether the dithering has a desired texture characteristic using the N-value dither output level, and a dither having a desired stability using the N-value dither output level. A step of evaluating the stability, a step of obtaining a gradation degree from each of the gradation evaluations in a plurality of gradations, and a granularity evaluation of each in a plurality of gradations A step of obtaining a degree of texture, a step of obtaining a texture from each of the texture evaluations at a plurality of gradations, a step of obtaining a stability from each of the stability evaluations at a plurality of gradations, and the gradation and granularity An image processing method (N-value dither matrix optimization) comprising a step of obtaining the adaptability of N-value dither matrix information from texture and stability and an optimization step of changing the N-value dither matrix based on the fitness In the method), the step of obtaining the fitness is characterized by deducting points when the number of gradations is smaller than a desired gradation characteristic.

  According to the sixth aspect of the present invention, an N-value dither pattern calculation step for calculating an N-value dither pattern corresponding to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value; Gradation evaluation step for evaluating whether the dither has a desired gradation characteristic using a value dither pattern, and granularity evaluation step for evaluating whether the dither has a desired granular characteristic using the N-value dither pattern A texture evaluation step for evaluating whether the dither pattern has a desired texture characteristic using the N-value dither pattern, and a stability evaluation for evaluating whether the dither has a desired stability using the N-value dither pattern. A step of obtaining a degree of gradation from each of the gradation evaluations in a plurality of gradations, a step of obtaining a granularity from each of the granularity evaluations in a plurality of gradations, A step of obtaining a texture degree from each of the texture evaluations at a plurality of gradations, a step of obtaining a stability from the respective stability evaluations at a plurality of gradations, and N from the gradation degree, granularity, texture degree, and stability. In an image processing method (N-value dither matrix optimization method) comprising a step of obtaining a fitness of value dither matrix information and an optimization step of changing an N-value dither matrix based on the fitness. The step of obtaining is characterized in that points are deducted when the number of gradations is smaller than a desired gradation characteristic.

  The invention according to claim 7 is characterized in that the optimization step is a genetic algorithm.

  According to the present invention, in the automatic optimization of the dither matrix, the number of gradations having a desired gradation characteristic can be maintained without being restricted by the local solution.

  In particular, the inventions according to claims 1 and 3 are not constrained by local solutions, and N-value dither matrix optimization that is optimal for a printing press assumed by processing using a simulation image is performed.

  In the inventions according to claims 2 and 4, the N-value dither matrix optimization is performed in a single time process by the process using the N-value pattern image, which is not easily caught by the local solution.

  In the invention described in claim 5, the N-value dither matrix optimization that is optimal for the printing press assumed by the processing using the simulation image is performed while maintaining the number of gradations of the desired gradation characteristics.

  In the invention according to claim 6, the N-value dither matrix optimization is performed in a single time process by the process using the N-value pattern image while maintaining the number of gradations of a desired gradation characteristic.

  In addition, the invention according to claim 7 can be optimized to a good dither matrix by using a genetic algorithm which is considered to be less susceptible to local solutions as an optimization step.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 shows a flowchart of N-value dither matrix optimization processing (image processing) according to the present invention (particularly, a flowchart of a processing unit that performs design processing characteristic of the present invention). FIG. 1 is a flowchart when a genetic algorithm (hereinafter referred to as GA) is used as an optimization means. This embodiment will be described by taking the case of optimizing binary dither as an example. As shown in Patent Document 5, a dither matrix is optimized using GA by defining a gene.

  Starting from step S101, in step S102, a plurality of given initial values are duplicated, and each duplicated individual is mutated, that is, a change is applied to the dither matrix. In step S103, the fitness of each of a plurality of individuals is obtained. The method for obtaining the fitness in step S103 will be described later (FIG. 2). In step S104, a ranking is assigned according to the fitness of each of a plurality of individuals. In step S105, end determination is performed. The termination condition is either the desired generation evolution, whether the individual with the best fitness has reached the desired value, or whether a predetermined time has passed. If the end determination in step S105 is Yes, the process proceeds to step S107, and if No, the process proceeds to step S106. In step S106, crossover, duplication / mutation, and the like are performed based on the fitness ranking to form a new generation of individuals, and the process returns to step S103. In step S107, the individual with the highest fitness at the end of the calculation is selected as a semi-optimal solution, and the process ends in step S108. FIG. 5 shows the configuration of the N-value dither matrix optimization apparatus of the present invention. 1 is an input means for inputting a dither matrix, 2 is an fitness calculation means for calculating fitness, 3 is an fitness selection means for selecting an individual with high fitness, and 4 is an output means for outputting an optimal dither matrix. .

  FIG. 2 shows a flowchart for sequentially obtaining the fitness of each of a plurality of individuals in step S103 of FIG. Starting from step S201, a simulation image of a certain gradation T is created in step S202. As shown in Patent Document 1, the simulation image selects dot reflectance distribution data for the target pixel from a plurality of types of dot reflectance distribution data according to the state of the neighboring pixel data of the target pixel of the input image data, Using the dot reflectance distribution data selected for each pixel, a simulation image is obtained by calculating the reflectance distribution of an image that will be output when input image data is given to the image recording apparatus. .

  In step S203, graininess is obtained using a simulation image. As shown in Patent Document 2, the granularity is obtained by converting the frequency of a simulation image, weighting the frequency component of the frequency spectrum with the sensitivity function of the human visual system, obtaining an integrated value, and setting the integrated value as the granularity. .

  In step S204, the gradation is obtained using the simulation image. As shown in FIG. 3, the ideal average density and simulation of each gradation are obtained by using the linear expression of the gradation and the average density obtained from the average density of the highlight part and the dark part, and the average density obtained from the simulation image. The absolute value of the difference value of the average density of the image is assumed to be gradation.

  In step S205, the texture level is obtained using the simulation image. As shown in Patent Document 3, the texture level is obtained by frequency-converting a simulation image and a smoothed version of the simulation image, obtaining a difference value of frequency components from both, and using the multi-channel model of the human visual system for the difference value. An integral value of values weighted by a plurality of spatial filter groups based on the above is obtained, and this integral value is set as a texture level.

  In step S206, stability is obtained using a simulation image. For stability, the perimeter of dots included in the simulation image is obtained, and the perimeter is regarded as stability.

  In step S207, it is determined whether or not the gradation T has reached the designated gradation. If the dither matrix to be optimized is a dither matrix used for quantization from 256 gradations to binary values, the granularity, gradation, texture level, and stability may be examined for all 256 gradations. Then, it is possible to examine only a simulation image of a specified number of gradations out of 256 gradations. If the end determination in Step S207 is Yes, the process proceeds to Step S209, and if No, the process proceeds to Step S208. In step S208, T is changed to the next gradation to be examined, and the process returns to step S202.

  In step S209, the granularity is calculated. The granularity is the total value of the granularity in a plurality of gradations. In step S210, the gradation is calculated. The gradation is a total value of gradation in a plurality of gradations. In step S211, the texture level is calculated. The texture level is a total value of texture levels in a plurality of gradations. In step S212, the stability is calculated. The stability is a total value of stability in a plurality of gradations. In step S213, the fitness is calculated. The fitness is the total value of granularity, gradation, texture, and stability.

  In step S214, it is confirmed whether or not the user has the number of gradations having a desired gradation characteristic. If the number of gradations of the desired gradation characteristic is 256, the number of gradations is checked to see if the individual being evaluated possesses 256 gradations. If the determination result is Yes, the process proceeds to step S216, and if the determination result is No, the process proceeds to step S215. In step S215, the fitness obtained in step S213 is deducted. Here, the deduction means that the fitness is manipulated so that the ranking in step S104 in FIG. When the GA is executed as a better individual as the fitness level is lower, the deduction processing means adding a certain value to the fitness level. The value to be deducted is such a large value that this individual will not remain in the next generation.

  In step S216, it is confirmed whether the individual is likely to be a local solution. If each of the granularity, texture degree, and stability is within ± 20% of the average value obtained from the granularity, texture degree, and stability, it is determined that the local solution is not obtained. If Yes, that is, if it is a local solution, the process proceeds to step S217, and if No, the process proceeds to step S218. In step S217, the fitness obtained in step S213 is deducted.

  Next, in the automatic optimization of the dither matrix, the point that the above-described processing of the present invention can maintain the number of gradations of a desired gradation characteristic and can be made difficult to be caught by a local solution will be described below.

  The problem of maintaining the number of gradations with desired gradation characteristics will be described. In the calculation of the fitness in step S213, the fitness is the total value of the granularity, the gradation, the texture, and the stability. If the weighting of gradation is doubled and tripled and importance is attached to the gradation, it becomes easier to own the number of gradations of the desired gradation characteristics, but the number of gradations of the desired gradation characteristics is not necessarily required. That is, it is not guaranteed that the number of gradations of a desired gradation characteristic is maintained. In addition, the optimization result is similar to the desired gradation characteristic, and the fitness is improved, and the granularity and the texture may be deteriorated.

  Therefore, in step S214 in FIG. 2, the above problem can be solved by confirming whether or not the user possesses the number of gradations having a desired gradation characteristic. In step S214, the number of gradations of the desired gradation characteristic is the same, but it may not be the same as the desired gradation characteristic, and the degree of freedom is given to the design, so the optimum dither matrix is automatically optimized. It becomes possible.

  Further, in the calculation of the fitness in step S213, the fitness is composed of a plurality of evaluation items such as granularity, gradation, texture, and stability. Although it depends on the search space, it is empirically known that a local solution often has one good item among a plurality of evaluation items. Therefore, in step S216, if each of the granularity, the texture degree, and the stability is within ± 20% compared to the average value obtained from the granularity, the texture degree, and the stability, it is determined that it is not a local solution. . When items other than granularity, gradation, texture, and stability are added as items used for optimization, the comparison range ± 20 may be changed. Since the gradation is confirmed in step S214, the gradation may be excluded from the determination as to whether or not it is a local solution in step S216.

  1 is automatic optimization of a binary dither matrix, it is also possible to perform automatic optimization of a multi-value dither matrix by using a simulation image corresponding to ternary and quaternary values. . In the dither matrix automatic optimization according to the present invention, the fitness is composed of four evaluation items of granularity, gradation, texture, and stability, but this may be changed. Depending on the execution time, evaluation items such as stability and texture may be reduced, or evaluation items such as moire and sharpness may be increased. In addition, halftone images used in various types of printing machines such as electrophotographic printers and ink jet printers can be created using the dither matrix created by the dither matrix automatic optimization processing of the present invention.

  Other examples are shown below. In the flowchart shown in FIG. 2, a simulation image is obtained in step S202. In step S202, if a dither pattern of a certain gradation T is optimized, a binary pattern is created. Based on this, graininess, gradation, texture level and stability are obtained.

  For the gradation in step S204 in FIG. 2, an LUT having an ideal average number of dots for each gradation and each gradation shown in FIG. 4 is created using the linear expression of the input value and the average density shown in FIG. The absolute value of the difference value between the average number of dots included in the LUT and the binary pattern is defined as gradation.

  When automatic optimization of a multi-value dither matrix such as ternary or quaternary is performed using a pattern, the calculation for obtaining the stability in step S206 in FIG. If execution times are compared, it is overwhelmingly faster to obtain a binary pattern than to obtain a simulation image. On the other hand, in the case of simulation, an image close to the assumed output environment can be optimized with a dither matrix that is more faithful to the image including the average density and machine drive noise.

  The dither obtained according to the present invention is applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.) or a device composed of a single device (for example, a copier, a facsimile device, etc.). You may apply.

  Another object of the present invention is to supply a storage medium storing software program codes for realizing the function of quantizing with the dither obtained in the above-described embodiment to a system or apparatus, and a computer ( This can also be achieved by reading and executing the program code stored in the storage medium by the CPU or MPU. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiment. As a storage medium for supplying the program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used. Further, by executing the program code read by the computer, not only the above-described dither function is realized, but an OS (operating system) running on the computer is actually executed based on the instruction of the program code. This includes a case where part or all of the processing is performed and the functions of the above-described embodiments are realized by the processing. Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. This includes a case where the CPU or the like provided in the board or the function expansion unit performs part or all of the actual processing, and the functions of the above-described embodiments are realized by the processing.

6 shows a processing flowchart of N-value dither matrix optimization according to the present invention. The flowchart of fitness calculation is shown. The relationship between input value and average density is shown. The ideal average number of dots for each gradation is shown. The structure of the N value dither matrix optimization apparatus of this invention is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Input means 2 Fitness calculation means 3 Fitness selection means 4 Output means

Claims (11)

  An N-value dither output level simulation step for calculating an N-value dither output level corresponding to a certain gradation in an N-value (M> N ≧ 2) dither matrix from a given M value, and the N value in a plurality of gradations An evaluation process for evaluating a plurality of characteristics such as graininess, gradation, stability, and texture of N-value dither using the dither output level, and the fitness of N-value dither matrix information is obtained based on the plurality of evaluations. In the image processing method including a process and an optimization process for changing an N-value dither matrix based on the fitness, the process for obtaining the fitness is characterized in that one of a plurality of evaluation characteristics is much higher than the other evaluation items. An image processing method characterized by deducting points when favorable.   An N-value dither pattern calculation step for calculating an N-value dither pattern corresponding to a certain gradation in an N-value (M> N ≧ 2) dither matrix from a given M value, and the N-value dither pattern in a plurality of gradations An evaluation step of evaluating a plurality of characteristics such as graininess, gradation, stability, texture, etc. of N-value dither using the step, and a step of determining the fitness of N-value dither matrix information based on the plurality of evaluations; In the image processing method comprising the optimization step of changing the N-value dither matrix based on the fitness level, the step of obtaining the fitness level is when one of a plurality of evaluation characteristics is extremely better than the other evaluation items. An image processing method characterized by deducting points.   An N-value dither output level simulation step for calculating an N-value dither output level corresponding to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value, and the N-value dither output level is used. A gradation evaluation step for evaluating whether the dither has a desired gradation characteristic, a granularity evaluation step for evaluating whether the dither has a desired granular characteristic using the N-value dither output level, and the N A texture evaluation step for evaluating whether the dither has a desired texture characteristic using a value dither output level, and a stability evaluation step for evaluating whether the dither has a desired stability using the N-value dither output level. A step of obtaining a gradation degree from each of the gradation evaluations in a plurality of gradations; a step of obtaining a granularity from each of the granularity evaluations in a plurality of gradations; A step of obtaining a texture from each of the texture evaluations in a plurality of gradations, a step of obtaining a stability from each of the stability evaluations in a plurality of gradations, and the gradation, granularity, texture, and stability In the image processing method comprising the step of obtaining the fitness of the N-value dither matrix information and the optimization step of changing the N-value dither matrix based on the fitness, the step of obtaining the fitness includes a plurality of evaluation items. An image processing method characterized in that points are deducted when one of the items is extremely better than the other evaluation items.   An N-value dither pattern calculation step for calculating an N-value dither pattern corresponding to a certain gradation in an N-value (M> N ≧ 2) dither matrix from a given M value, and a desired value using the N-value dither pattern A gradation evaluation step for evaluating whether the dither has gradation characteristics, a granularity evaluation step for evaluating whether the dither has desired granular characteristics using the N-value dither pattern, and the N-value dither pattern. A texture evaluation step for evaluating whether the dither has a desired texture characteristic, a stability evaluation step for evaluating whether the dither has a desired stability using the N-value dither pattern, and a plurality of gradations. A step of obtaining a gradation degree from each of the gradation evaluations; a step of obtaining a granularity from each of the granularity evaluations in a plurality of gradations; The step of obtaining the texture degree from the texture evaluation, the step of obtaining the stability from the stability evaluation of each of a plurality of gradations, and the adaptation of the N-value dither matrix information from the gradation degree, granularity, texture degree, and stability In the image processing method comprising the step of obtaining the degree of optimization and the optimization step of changing the N-value dither matrix based on the degree of fitness, the step of obtaining the degree of fitness includes: An image processing method characterized in that a point is deducted when it is extremely good.   An N-value dither output level simulation step for calculating an N-value dither output level corresponding to a certain gradation of an N-value (M> N ≧ 2) dither matrix from a given M value, and the N-value dither output level is used. A gradation evaluation step for evaluating whether the dither has a desired gradation characteristic, a granularity evaluation step for evaluating whether the dither has a desired granular characteristic using the N-value dither output level, and the N A texture evaluation step for evaluating whether the dither has a desired texture characteristic using a value dither output level, and a stability evaluation step for evaluating whether the dither has a desired stability using the N-value dither output level. A step of obtaining a gradation degree from each of the gradation evaluations in a plurality of gradations; a step of obtaining a granularity from each of the granularity evaluations in a plurality of gradations; A step of obtaining a texture from each of the texture evaluations in a plurality of gradations, a step of obtaining a stability from each of the stability evaluations in a plurality of gradations, and the gradation, granularity, texture, and stability In the image processing method including the step of determining the fitness of the N-value dither matrix information and the optimization step of changing the N-value dither matrix based on the fitness, the step of determining the fitness includes a desired gradation An image processing method characterized by deducting points when the number of gradations is smaller than the characteristic number.   An N-value dither pattern calculation step for calculating an N-value dither pattern corresponding to a certain gradation in an N-value (M> N ≧ 2) dither matrix from a given M value, and a desired value using the N-value dither pattern A gradation evaluation step for evaluating whether the dither has gradation characteristics, a granularity evaluation step for evaluating whether the dither has desired granular characteristics using the N-value dither pattern, and the N-value dither pattern. A texture evaluation step for evaluating whether the dither has a desired texture characteristic, a stability evaluation step for evaluating whether the dither has a desired stability using the N-value dither pattern, and a plurality of gradations. A step of obtaining a gradation degree from each of the gradation evaluations; a step of obtaining a granularity from each of the granularity evaluations in a plurality of gradations; The step of obtaining the texture degree from the texture evaluation, the step of obtaining the stability from the stability evaluation of each of a plurality of gradations, and the adaptation of the N-value dither matrix information from the gradation degree, granularity, texture degree, and stability In the image processing method comprising the step of obtaining the degree of optimization and the optimization step of changing the N-value dither matrix based on the degree of fitness, the step of obtaining the degree of fitness is less than the number of gradations of the desired gradation characteristic An image processing method characterized by deducting points in some cases.   The image processing method according to claim 1, wherein the optimization step is a genetic algorithm.   An image processing apparatus using the N-value dither matrix created by the image processing method according to claim 1.   An image recording apparatus using the N-value dither matrix created by the image processing method according to claim 1.   A program for causing a computer to implement the image processing method according to any one of claims 1 to 7.   A computer-readable recording medium recording a program for causing a computer to implement the image processing method according to claim 1.

Images