JP2009171014A - Image processor, image recorder and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor which solves the problems of search time and search section distortion occurring in DBS method, and to provide image recorder and program. <P>SOLUTION: An entire sheet image to be output is stored in a memory. The image thus stored is divided, a first section of divided image is set in a memory, a DBS method is performed locally for the section thus set as a search section and a binary image is created. A determination is made an whether or not the section size is changed in the binary image thus created. When the section size is changed, the image is divided into images each having a size larger than that of the above divided image, and a plurality of larger divided images are held. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus, an image recording apparatus, and a program 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. There is a method called pseudo halftone processing.

  As a method for pseudo halftone processing, a technique using a dither created in Patent Document 1 and an error diffusion method such as Patent Document 2 are widely used.

  The dither method is a method in which the original multi-valued image is divided into blocks of the same size, and the luminance value of the corresponding binary image is determined using a dither matrix for each block.

  The error diffusion method determines the luminance value of the corresponding binary image pixel while scanning the pixels of the input multilevel image in the raster scan order, and the luminance value of the input multilevel image and the binary image generated at the time of determination are determined. This is a method of diffusing a luminance value error to surrounding luminance value undetermined pixels.

  However, in the dither method, since processing is performed for each block, when pixels having the same luminance value in the original multi-valued image are in a continuous wide area, each block outputs a binary image having the same pattern. An unsightly pattern with a long period is likely to occur.

  If the block is enlarged, the gradation of the original multi-valued image can be reproduced with high accuracy, but the outline is blurred and a clear image cannot be obtained.

  Further, in the error diffusion method, an unsightly pattern or moire is relatively less likely to occur than in the dither method, but there is a problem that the outline is blurred because the error is diffused around.

  As a halftoning processing method in which these problems are unlikely to occur, there is a Direct Binary Search (Direct Binary Search, hereinafter referred to as DBS) method. The DBS method is disclosed in Non-Patent Document 1, for example.

  The DBS method will be described below.

  First, one pixel is specified while scanning a binary image having an arbitrary luminance value for each pixel in units of one pixel. For the specified pixel (hereinafter referred to as the pixel of interest), the luminance value of the pixel of interest is inverted from 0 to 1 or 1 to 0, or replaced with the luminance value of the adjacent pixel. A binary image obtained by replacing values is subjected to two-dimensional filter processing to generate a restored multi-value image, and an error between the restored multi-value image and the original multi-value image (hereinafter referred to as post-substitution error) is calculated. .

  An error from the original multi-valued image (hereinafter referred to as pre-replacement error) is also calculated for the restored multi-valued image obtained by performing two-dimensional filtering on the binary image before the pixel of interest is replaced.

  When the error after replacement is smaller than the error before replacement, the luminance value of the target pixel of the binary image is updated with the luminance value after replacement. If the error after replacement is larger than the error before replacement, the same process is performed for the next pixel of interest without updating the luminance value. Thereafter, the same processing is repeated for the target pixel sequentially identified in the scanning process.

  In such a DBS method, the initial binary image used at the beginning of the halftoning process is changed to a random noise image, so that the periodic and annoying patterns and moire that are problems in the dither method and the error diffusion method can be obtained. Occurrence can be avoided. Further, since the operation of diffusing the luminance value in the vicinity is not performed unlike the error diffusion method, there is an advantage that a binary image having a clear outline can be obtained.

  Moreover, there exists patent document 3 as a technique using DBS method. Patent Document 3 discloses a technique for restoring a binary image to a multi-valued image for each DBS search section and calculating an error by comparing the restored image with the original image. If Patent Document 3 is used, a binary image with less unevenness and roughness in gradation can be obtained, and dot missing, image blurring, and whiteout are less likely to occur.

  In addition, there is Non-Patent Document 2 that uses a quasi-optimization technique for iterative exploration in the DBS method. Non-Patent Document 2 uses a genetic algorithm to divide an input image into a plurality of search intervals so that the difference in image quality evaluation value between the input image and the output image with the reduced number of gradations is minimized in each search interval. This is a method for searching for an optimal solution.

  However, although the DBS method is excellent in gradation and sharpness, there are two problems of search time and search section distortion.

  The search time can be solved by narrowing the search space, which is generally one of the means for speeding up by the semi-optimization method. In Non-Patent Document 2, the search section is divided into a plurality of sections, and the search time can be reduced by reducing the search section. However, if the search interval is reduced, distortion occurs in the second problem / search interval.

  Here, the distortion in the search section means that edge portions such as continuous lines extending over the search section are not connected.

In response to this problem, Patent Document 4 eliminates search section distortion by processing while shifting the search section little by little.
JP 2005-192195 A Japanese Patent No. 3732470 JP 2004-304543 A JP 2005-144873 A "Model based halftoning using direct binary search", author "M. Analoui, JP Allebach", publication "Proceedings SPIE Human Vision, Visual Proceedings, and Digital Display III" (Volume 1666, pages 96-108, August 1992) Kobayashi, Saito, "Pseudo grayscale display method using genetic algorithm", IEICE Transactions DII, Vol.J78-D-II, No.10, pp.1450-1459,1995

  However, in order to eliminate the search section distortion, there remains a problem that the ratio of the boundary portion of the search section in the entire image must be reduced by increasing the search section.

  If the search interval is increased, the search time becomes longer, and it becomes difficult to obtain a binary image having the desired gradation and sharpness with a practical search time. However, the search section distortion occurs.

  Therefore, the present invention has been made in view of such problems, and it is possible to provide an image processing apparatus, an image recording apparatus, and a program that can solve the search time and search section distortion problems that occur in the DBS method.

  In order to achieve the above object, the invention described in claim 1 is a structure in which a binary image is obtained by a direct binary exploration method using a sub-optimization method that performs iterative processing, and depends on the number of evaluations or the evaluation value. And an image processing apparatus having means for expanding the search range.

  According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the size ratio for expanding the search range is not an integer.

  According to a third aspect of the present invention, in the image processing apparatus according to the first or second aspect, the size ratio for expanding the search range is a real number of 1 or more and less than 2.

  According to a fourth aspect of the present invention, in the image processing apparatus according to the first aspect, a genetic algorithm is used as the semi-optimization method.

  According to a fifth aspect of the present invention, in the image processing apparatus according to the first aspect, the evaluation value is a value composed of at least one evaluation result of graininess, sharpness, and gradation. To do.

  The invention according to claim 6 is a structure in which a binary image is obtained by a direct binary search method using a sub-optimization technique that performs iterative processing, and means for expanding a search range according to the number of evaluations or an evaluation value. It is an image recording apparatus having the above.

  A seventh aspect of the present invention is the image recording apparatus according to the sixth aspect, wherein the size ratio for expanding the search range is not an integer.

  According to an eighth aspect of the present invention, in the image recording apparatus according to the sixth or seventh aspect, the size ratio for expanding the search range is a real number of 1 or more and less than 2.

  The invention according to claim 9 is the image recording apparatus according to claim 6, wherein a genetic algorithm is used as the semi-optimization method.

  According to a tenth aspect of the present invention, in the image recording apparatus according to the sixth aspect, the evaluation value is a value consisting of at least one evaluation result of graininess, sharpness, and gradation. To do.

  The invention according to claim 11 is a program for causing a computer to execute a process of acquiring a binary image by a binary search method directly using a sub-optimization technique for performing an iterative process, and depending on the number of evaluations or an evaluation value. It has a code for a process of expanding the search range.

  According to the present invention, in an image processing apparatus that obtains a binary image by a binary search method directly using a semi-optimization method that performs iterative processing, by expanding the search range according to the number of evaluations or the evaluation value, It is possible to provide an image processing apparatus, an image forming apparatus, and a program capable of obtaining a binary image having a desired image quality without causing a search section distortion in a practical search time.

DBS (Direct Binary Search)
DBS is a method for searching for an optimal binary image by comparing a binarized image with an original image and evaluating gradation, sharpness, graininess, and the like. It is said that a better quantum image can be obtained than a dither method in which a quantum image is determined by one pixel, an error diffusion method that calculates a quantization error of a neighboring image, and an average error minimum method. However, since the execution time is very long because the quantization / evaluation is performed a plurality of times, I do not hear the story put into practical use in EP / IJ.

  In order to obtain high-speed and high-quality images by DBS (Direct Binary Search Method), the entire image area (A4: 210 mm × 297 mm) is stored and optimized. Partial optimization is performed by dividing into m × m size at the time of optimization. Further, the division size is increased in accordance with the number of optimizations and the evaluation value (score / cost function).

  The present invention achieves the following using the DBS.

  In an image processing apparatus that obtains a binary image by a binary search method directly using a sub-optimization method that performs iterative processing, it is practical by expanding the search range according to the number of evaluations or the evaluation value. In the search time, a binary image having a desired image quality is obtained without causing distortion in the search section.

  In addition, assuming that the size ratio for expanding the search range is not an integer, a binary image having a desired image quality can be obtained in a practical search time without causing distortion in the search section.

  Further, since the size ratio for expanding the search range is a real number of 1 or more and less than 2, a binary image having a desired image quality can be obtained in a practical search time without causing search section distortion.

  Further, by using a genetic algorithm as the quasi-optimization method, a binary image having a desired image quality can be obtained in a practical search time without being limited by a local solution and without causing search section distortion.

  Furthermore, the evaluation value is a value composed of at least one evaluation result among graininess, sharpness, and gradation, so that a binary image having a desired image quality is obtained.

(First embodiment)
DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described in detail with reference to the accompanying drawings.
Components are distinguished by adding symbols.
FIG. 1 is a diagram showing the configuration of the image recording apparatus according to this embodiment.
FIG. 2 is a diagram showing a configuration of the image processing apparatus according to the present embodiment (particularly a diagram showing a configuration of an image processing unit that performs image processing characteristic of the present invention).
FIG. 3 shows a configuration of an image input / output system configured using the image processing apparatus of the present embodiment.

  In FIG. 1, a sheet on which an image is to be formed is set on the main body tray 101 or the manual feed tray 102, and sheet conveyance is started from the tray 101 or 102 by the sheet feed roller 103. Prior to the conveyance of the sheet by the sheet feeding roller 103, the photosensitive member (photosensitive drum) 104 rotates, the surface of the photosensitive member 104 is cleaned by the cleaning blade 105, and then uniformly charged by the charging roller 106. The

  Here, the laser beam modulated in accordance with the image signal is exposed from the laser optical system unit 107, developed by the developing roller 108, and the toner adheres. Made. A high voltage bias is applied to the developing roller 108 by a bias circuit 114, and the overall density of the image can be controlled by controlling the bias in the bias circuit 114.

  The sheet fed from the sheet feeding roller 103 is conveyed while being sandwiched between the photosensitive drum 104 and the transfer roller 109, and at the same time, a toner image is transferred to the sheet. The transferred toner remaining on the photosensitive member 104 is again scraped off by the cleaning blade 105.

  A toner density sensor 110 is provided in front of the cleaning blade 105, and the density of the toner image formed on the photoreceptor 104 can be measured by the toner density sensor 110. The paper on which the toner image is placed is transported to the fixing unit 111 along the transport path, and the toner image is fixed on the paper in the fixing unit 111.

  The printed sheets finally pass through the paper discharge roller 112 and are discharged in page order with the recording surface down. By the way, a video control unit 171 and an LD driving circuit 172 are connected to the laser optical system unit 107, and the video control unit 171 controls or holds an image signal from a personal computer or a workstation. The generated evaluation chart (test pattern) signal or the like is generated. The light beam emitted from the laser optical system unit 107 is reflected and deflected by the reflection mirror 121 and the reflection mirror 122, passes through the fθ optical system, and is guided to the photosensitive drum 104.

FIG. 2 is a perspective view showing an example of a positional relationship with a photosensitive drum as a latent image carrier on which a light beam emitted by the laser optical system unit 107 of FIG. 1 is written.
The symbols used in FIG. 2 are partially different from those in FIG.

  In FIG. 2, an image processing unit 1 includes a laser diode (semiconductor laser) 11, a laser diode 12, a collimator lens 13, a collimator lens 14, an optical member for optical path synthesis 15, a quarter wavelength plate 16, and a quarter wavelength plate. 16. Each optical element of the beam shaping optical system 17 and the beam shaping optical system 18 is included.

  Each of these optical elements constitutes a laser light source section (beam light source) Sou. The two light beams P1 emitted from the laser light source unit Sou are converted into parallel light beams by the collimator lenses P1 and P2, and are guided to the polygon mirror 19 constituting a part of the scanning optical system. Reflected and deflected in the main scanning direction Q1 by the surfaces 20a to 20f.

  The reflected and deflected light beam is guided to reflecting mirrors 21 and 22 constituting a part of the fθ optical system, and the reflected and deflected light beam passes through the fθ optical system 23 and is obliquely reflected. 24 and is guided to the surface 26 of the photosensitive drum 25 as a latent image carrier by the oblique reflection mirror 24. The surface 26 of the photosensitive drum 25 is linearly scanned in the main scanning direction Q1 by the light beam P1. This surface 26 is a surface to be scanned by the light beam P1, and writing is performed on this surface to be scanned.

  The laser optical system unit 107 is provided with synchronization sensors 27 and 28 on both sides in the longitudinal direction of the reflection mirror 24 (main scanning direction Q1 of the light beam). The synchronization sensor 27 is used for determining the write start timing, and the sync sensor 28 is used for determining the write end timing.

  An image input device 301 indicates an input device such as a scanner or a digital camera. If the input image has an 8-bit accuracy, the image input device 301 is fetched as 256-gradation image data. This multi-value image data is input to the image processing apparatus 302 of this embodiment.

  The image processing device (image processing unit) 302 performs processing for converting the 256-gradation image data input from the image input device 301 into the number of gradations that can be output by the subsequent image output device 303. In this gradation number conversion processing, multilevel error diffusion or a multilevel average error minimum method may be used. The image data quantized by the image processing apparatus 302 is sent to an image recording apparatus (image forming apparatus, image output apparatus) 303 as shown in the configuration of FIG.

  The processing method according to the present invention can be applied to the image recording apparatus 303 even when an image is recorded (image formation) using an inkjet method or gravure printing.

  In the system configuration diagram of FIG. 3, each device is illustrated as independent depending on the processing. However, the present invention is not limited to this, and the form in which the function of the image processing device 302 exists in the image input device 301 or the image Some forms exist in the output device 303.

  FIG. 4 is a flowchart showing processing in the image processing apparatus 302 of the present embodiment shown in FIG.

The entire image output by the image recording apparatus 303 is set in the memory (step S1001).
If an A4 size image is output at 600 dpi × 600 dpi, an image of 4960 dots × 7015 dots is stored.

All the images stored in step S1001 are divided as shown in FIG. 5 (step S1002).
Now, assume that the division size in FIG. 5 is 64 dots × 64 dots.

  The first section of the image divided in step S1002 is set in the memory (step S1003).

The DBS method is executed locally using the set section as a search section to create a binary image (step S1004).
The processing for executing the DBS method locally will be described later.

  In the process of step S1005, it is determined whether binarization has been completed for all of the sections divided in the process of step S1002 (step S1005).

  If not completed (step S1005 / NO), the process proceeds to step S1006, and if completed (step S1005 / YES), the process proceeds to S1007. In the process of step S1006, the next section that has not been binarized is set in the memory, and the process proceeds to step S1004 (step S1006).

  It is determined whether or not to change the partition size (step S1007). When the partition size is changed (step S1007 / YES), the process proceeds to step S1008. When the partition size is not changed (step S1007 / NO), all processes are ended.

Next, in the process of step S1008, as shown in FIG. 6, it divides | segments into a larger size than FIG. 5 (step S1008). Thereafter, the process proceeds to S1003.
The division size in FIG. 6 is larger than the division size shown in FIG. 5 but is not an integral multiple, and is assumed to be 80 dots × 80 dots.

  In the processing of S1008, the partition size to be changed is not limited to FIGS. 5 and 6, but a plurality of partition sizes further increased as shown in FIGS.

  According to the present embodiment, in an image processing apparatus that obtains a binary image by a direct binary search method using a sub-optimization method that performs iterative processing, the search range is expanded according to the number of evaluations or the evaluation value. In a practical search time, it is possible to obtain a binary image having a desired image quality without causing distortion in the search section.

(Second Embodiment)
In the present embodiment, local DBS performed in the process of S1004 in FIG. 4 will be described with reference to FIG.
In FIG. 9, a binary image is semi-optimized using a genetic algorithm.

First, as shown in FIG. 10, a plurality of sections set in the memory by the process of step S1003 or the process of step S1006 are duplicated and each is mutated (step S2001).
The mutation here is to binarize each replicated section by a random dither method.

  The fitness is determined for each copied individual (step S2002). The fitness will be described later.

  Ranking is performed according to the fitness of each individual obtained in the process of S2002 (step S2003).

  Next, end determination is performed (step S2004).

  If the process is to end (step S2004 / YES), the process proceeds to step S2006, and if the process is to be continued (step S2004 / NO), the process proceeds to step S2005.

  The termination condition here is whether it has been performed a predetermined number of times or there was an individual who has obtained the highest fitness.

  In the process of step S2006, the individual with the highest fitness is output, and the process ends (step S2006).

  In the process of step S2005, a new generation is generated according to the rank according to the fitness performed in the process of step S2003 (step S2005).

  The new generation is generated by crossing, duplicating, and mutating based on the upper individual according to the roulette rule, leaving the upper individual, deleting the lower individual, and performing the new generation according to the general genetic algorithm.

  As an example of the way of crossing here, as shown in FIG. 11, the individual A and the individual B with higher fitness are divided into left and right at the center and exchanged to create two new individuals. As shown, the individual A and the individual B are divided into upper and lower parts at the center and exchanged to create two new individuals.

  For duplication / mutation, individuals with higher fitness are selected at random, the contents of the selected individuals are duplicated, and one point in the image is randomly selected as shown in FIG. Any one of white pixels and black pixels may be set at random.

  According to the present embodiment, in addition to the effects of the first embodiment, by using a genetic algorithm as a semi-optimization method, the search section distortion does not occur in a practical search time without being caught by a local solution. A binary image having a desired image quality can be obtained.

(Third embodiment)
In the present embodiment, the operation for obtaining the fitness performed in the process of S2002 in FIG. 9 will be described with reference to FIG.

  First, the first individual is read and set in the memory (step S3001).

  The granularity is obtained using the binary pattern image of the individual set in the process of S3001 (step S3002).

  The granularity is obtained by converting the frequency of an image as described in JP-A-2002-245465 and weighting the frequency component of the frequency spectrum with a sensitivity function of a human visual system to obtain an integral value. This integral value is set as the granularity G of the individual being evaluated.

  Next, the gradation is evaluated (step S3003). Gradation is evaluated as in equation (1).

ここで、Ｔq（ｉ）は、評価個体中のｉ番目の画素位置における濃度である。
評価個体は、２値画像だから０または２５５の値を持つ。
また、Ｔo（ｉ）、は元画像中のｉ番目の画素位置における濃度である。
階調性の評価値Ｔは、２値画像と元画像の階調総和の差分値である。

Here, Tq (i) is the density at the i-th pixel position in the evaluation individual.
Since the evaluation individual is a binary image, it has a value of 0 or 255.
To (i) is the density at the i-th pixel position in the original image.
The gradation evaluation value T is a difference value of the total gradation of the binary image and the original image.

  Next, the sharpness is evaluated (step S3004). Sharpness is evaluated as in equation (2).

ここで、ｓ（ｉ）は、画像中のｉ番目の画素位置における鮮鋭性評価値である。鮮鋭性評価値ｓ（ｉ）は、式（３）により定義される。

Here, s (i) is a sharpness evaluation value at the i-th pixel position in the image. The sharpness evaluation value s (i) is defined by equation (3).

ここでＴｑ（ｉ）は、評価個体中のｉ番目の画素位置における濃度であり、Ｌ（ｉ）は元画像中のｉ番目の画素位置におけるラプラシアン操作後のデータを２値化したものであり、０または２５５の値をもつ。

Here, Tq (i) is the density at the i-th pixel position in the evaluation individual, and L (i) is the binarized data after the Laplacian operation at the i-th pixel position in the original image. , 0 or 255.

  An example of filter coefficients used for Laplacian operation is shown in FIG. The threshold value for binarizing the data after the Laplacian operation is set according to the desired image quality.

  Now, the binarization threshold is set to 127. It is determined whether L (i) is 255, that is, whether a dot is output at the edge of the image.

  Next, a comprehensive evaluation value is obtained (step S3005).

  The comprehensive evaluation value E is expressed by the following equation (4) based on the granularity G obtained in the process of step S3002, the gradation evaluation value T obtained in the process of step S3003, and the sharpness evaluation value S obtained in the process of step S3004. Ask for.

ここでα・β・γは粒状度、階調性、鮮鋭性の重み付け係数である。今、α＝β＝γ＝１としておく。

Here, α, β, and γ are weighting coefficients for granularity, gradation, and sharpness. Now, α = β = γ = 1.

  Next, end determination is performed (step S3006).

  When the process is to be ended (step S3006 / YES), all the processes are ended. When continuing (step S3006 / NO), it progresses to the process of step S3007.

  Here, the end condition is whether all individuals have been evaluated. In the process of step S3007, the individual to be evaluated next is set in the memory, and the process proceeds to step S3002 (step S3007).

  As described above, the multi-value error diffusion processing in the image processing unit is performed by the configuration of FIG.

  Next, why such a process is effective will be described.

  Here, the partition size is changed in the process of step S1007 in FIG. 4 in the first embodiment. However, in the process of step S1008, as shown in FIG.

  Now, the division size in FIG. 5 is 64 dots × 64 dots, and the division size in FIG. 6 is 80 dots × 80 dots.

  In this way, if the division size is not an integral multiple, the pixel position that becomes the boundary in FIG. 5 is less likely to become the boundary in FIG.

  The pixel position 320 dots × 320 dots, which is the least common multiple of the division sizes in FIGS. 5 and 6, becomes a boundary even if any of the divisions in FIGS. 5 and 6 is used. However, there are only a few points in the entire image.

  Further, by using FIG. 7 and FIG. 8 as well as FIG. 5 and FIG. 6 as the division size, the search section distortion that could not be eliminated by the division size of FIG. 5 and FIG. This can be solved by a combination of division sizes or a combination of division sizes in FIGS.

  Here, if the division size ratio between FIG. 5 and FIG. 6 is an integer ratio such as 2 or 3 or a simple one such as 1.5, the least common multiple of the division sizes in FIGS. 5 and 6 is small. Since the value is a value, it is difficult to improve the search section distortion. Therefore, the ratio for expanding the search range is preferably a real number of 1 or more.

  Further, if the search size is increased, the search space becomes larger and it becomes difficult to reach the sub-optimal solution. Therefore, the ratio for expanding the search range is preferably less than 2. Therefore, the ratio for expanding the search range is preferably a real number of 1 or more and less than 2.

  According to the present embodiment, in addition to the effects according to the first embodiment and the second embodiment, the size ratio for expanding the search range is not an integer, so that the search section distortion can be achieved with a practical search time. It is possible to obtain a binary image having a desired image quality without occurrence of the above.

  Further, since the size ratio for enlarging the search range is a real number of 1 or more and less than 2, it is possible to obtain a binary image having a desired image quality without causing a search section distortion in a practical search time. Become.

  Furthermore, since the evaluation value is a value composed of at least one evaluation result among graininess, sharpness, and gradation, a binary image having a desired image quality can be obtained.

  Note that the present invention can be applied to a system including a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), or an apparatus including a single device (for example, a copier, a facsimile machine, etc.). You may apply.

  Further, the present invention supplies a storage medium storing software program codes for realizing the functions of the above-described embodiments to a system or apparatus, and the computer (CPU or MPU) of the system or apparatus is stored in the storage medium. This can also be achieved by reading and executing the program code. 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 functions of the above-described embodiments are realized, but also an OS (operating system) operating on the computer based on the instruction of the program code. A case where part or all of the actual processing is performed and the functions of the above-described embodiments are realized by the processing is also included.

  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.

  The embodiment of the present invention has been described above. The first embodiment, the second embodiment, and the third embodiment described above show examples of preferred embodiments of the present invention, and the present invention is not limited thereto. Various modifications can be made without departing from the scope of the invention.

1 is a diagram illustrating a configuration of an image recording apparatus according to a first embodiment. 1 is a block diagram of an image processing apparatus according to a first embodiment. 1 illustrates a configuration of an image input / output system configured using an image processing apparatus according to a first embodiment. It is a flowchart figure which shows the operation | movement which calculates | requires the fitness concerning 1st Embodiment. It is a 1st figure which shows an example of division | segmentation size. It is a 2nd figure which shows an example of division | segmentation size. It is a 1st figure which shows an example of the search area distortion which was not able to be eliminated by division size. It is a 2nd figure which shows an example of the search area distortion which was not able to be eliminated by division size. It is a flowchart figure which shows the operation | movement which calculates | requires the fitness concerning 2nd Embodiment. It is a figure which shows the example which gives a variation | mutation to each of the division which replicated two or more. It is a figure which shows the example which divides right and left individual | organism | solid A and individual | organism | solid B in the center at right and left, and produces each two by exchanging each. It is a figure which shows the example which divides | segments the individual | organism | solid A and the individual | organism | solid B up and down in the center, and makes two each by exchanging each. It is a figure which shows the example which selects one point in an image at random, and sets any white pixel and black pixel of a binary image at random. It is a flowchart figure which shows the operation | movement which calculates | requires the fitness concerning 3rd Embodiment. It is a figure which shows an example of the filter coefficient used for Laplacian operation.

Explanation of symbols

11, 12 Laser diode (semiconductor laser)
DESCRIPTION OF SYMBOLS 13, 14 Collimating lens 15 Optical path composition optical member 16 1/4 wavelength plate 17, 18 Beam shaping optical system 19 Polygon mirror 21, 22, 121, 122 Reflection mirror 23 fθ optical system 24 Oblique reflection mirror 25, 104 Photoconductor Drum 26 Surface of Photosensitive Drum 101 Main Body Tray 102 Manual Tray 103 Paper Feed Roller 105 Cleaning Blade 106 Charging Roller 107 Laser Optical System Unit 108 Developing Roller 109 Transfer Roller 110 Toner Concentration Sensor 111 Fixing Unit 112 Paper Discharge Roller 114 Bias Circuit 171 Video Control unit 172 LD drive circuit

Claims (11)

It uses a sub-optimization method that performs iterative processing, and is a structure that obtains a binary image by direct binary exploration.
An image processing apparatus comprising means for expanding a search range according to the number of evaluations or an evaluation value.   The image processing apparatus according to claim 1, wherein a size ratio for expanding the search range is not an integer.   The image processing apparatus according to claim 1, wherein a size ratio for expanding the search range is a real number greater than or equal to 1 and less than 2. 4.   The image processing apparatus according to claim 1, wherein a genetic algorithm is used as the semi-optimization method.   The image processing apparatus according to claim 1, wherein the evaluation value is a value including at least one evaluation result among graininess, sharpness, and gradation. It uses a sub-optimization method that performs iterative processing, and is a structure that obtains a binary image by direct binary exploration.
An image recording apparatus comprising means for expanding a search range according to the number of evaluations or an evaluation value.   The image recording apparatus according to claim 6, wherein a size ratio for enlarging the search range is not an integer.   The image recording apparatus according to claim 6 or 7, wherein a size ratio for expanding the search range is a real number of 1 or more and less than 2.   The image recording apparatus according to claim 6, wherein a genetic algorithm is used as the semi-optimization method.   7. The image recording apparatus according to claim 6, wherein the evaluation value is a value composed of at least one evaluation result among graininess, sharpness, and gradation. A program that causes a computer to execute a process of obtaining a binary image by a direct binary search method using a sub-optimization technique that performs iterative processing.
A program having a code for a process of expanding a search range according to the number of evaluations or an evaluation value.

Images