JPH07254982A - Coexisting image processor - Google Patents

Abstract

PURPOSE:To reproduce a high-quality image by preventing the image from being distorted at a border between the different kinds of areas by calculating an intermediate area, which does not belong to any areas of a character area, dot area and photograph area, from a coexisting original and further precisely classifying the intermediate area corresponding to the classes of areas at adjacent blocks. CONSTITUTION:This device is provided with an image input part 1 for fetching a source image, in which characters, dots and photographs coexist, for the unit of a picture element, area classifying means 31-36 for classifying the input images into the character area, dot area, photograph area and intermediate area, which does not belong to any one of these areas, for each block in a prescribed size, and intermediate area redeciding circuit 37 for precisely classifying the intermediate area corresponding to the classes of areas at the blocks adjacent to any one of row and column directions. At the same time, the device is provided with a half tone processing part 4 for reproducing the coexisting image by using a dither matrix corresponding to the classes of areas in the classified (precisely classified) images.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus of a digital image forming apparatus such as a digital copying machine or a facsimile machine, and is particularly compatible with a boundary of an original image in which characters and pictures (dots and photographs) are mixed. The present invention relates to a mixed image processing device for reproducing a certain digital image.

[0002]

2. Description of the Related Art It is known that high resolution and gradation are required as conditions for reproducing an original image. In the case of a mixed image including pictures (dots and photographs) in addition to characters, a dither matrix suitable for each image type is applied and image reproduction processing is performed in order to maintain both of the above qualities.

Conventionally, various techniques for identifying the image type of such characters and pictures have been proposed. JP-A-2-10368
In Japanese Patent Laid-Open No. 5 (1994), the larger number of regions is measured based on information in a block having a predetermined number of pixels, that is, the number of regions measured based on white pixels and the number of regions measured based on black pixels. The number of regions of this block is taken as the number of regions of this block, or the average density of this block is added to the number of regions measured based on white pixels and the number of regions measured based on black pixels. , A method in which halftone dot areas are separated is described. Japanese Patent Laid-Open No. 2-16
Japanese Patent No. 2477 discloses that the pixel area of interest is determined to be a character area or a photograph area depending on whether or not the pixel density of the pixel data of interest is rapidly changing in a specific direction. There is. JP-A-3-2081
In Japanese Patent Laid-Open No. 84, the dispersion processing is performed on each pixel of the input image data on a 3 × 3 matrix, and the mesh is divided into 2 × 2 meshes, and each mesh is subjected to edge processing such as a photograph area or a character area. Those that are classified are listed.

Further, when reproducing a document in which characters / dots / photos are mixed, halftone dot area detection utilizing peak information of halftone dot dots and edge area detection utilizing continuity of black and white images. An algorithm that separates a character area from a picture area by combining the two is known ("Image area separation method for mixed text / picture (halftone dot, photograph) image", IEICE Transactions D −II Vol.J75-D-II No.1, P39
~ 47, January 1992).

[0005]

The above-mentioned prior arts are all methods of allocating each block to a character area, a halftone dot area, or a photographic area in a mixed image, and therefore, to each of the allocated areas. Although suitable image reproduction processing is performed and a reproduced image of high quality can be provided, since all areas are uniformly distributed to any one of the above-mentioned three types, the distribution condition is relaxed. It cannot be denied that, because of this, distortion occurs in the reproduced image at the boundary between different types of areas, noise caused during image capture, or erroneous determination during sorting processing. Therefore, there is a risk that different types of regions may be accidentally isolated in a region surrounded by the same type, and it is difficult to always provide high image quality.

The present invention has been made in view of the above,
The original image in which characters, halftone dots, and photographs are mixed is assigned to each predetermined block as a character region, a halftone dot region, a photograph region, and an intermediate region that does not belong to any of the above regions, and this intermediate region is a region of neighboring blocks. Provide a mixed image processing device capable of preventing image distortion at the boundary of different types of images or correcting isolated errors in images of the same type to reproduce high-quality images by finely classifying according to types The purpose is to do.

[0007]

SUMMARY OF THE INVENTION The present invention is a mixed image processing apparatus for performing mixed image processing using a dither matrix corresponding to the type of classified images, and is an original in which characters, halftone dots, and photographs are mixed. An image input unit that captures images in pixel units,
Classification of the input image for each block of i rows x j columns (i and j are integers of 2 or more) pixels into a character area, a halftone dot area, a photograph area, and an intermediate area that is not classified into any of the above areas. And a sub-classifying means for sub-classifying the intermediate area in either the row direction or the column direction according to the area type of a neighboring block (claim 1).

Further, the sub-classification means performs sub-classification according to the area types of the blocks that are adjacent to each other (claim 2).

Further, the sub-classifying means classifies into the same type as the type of the block when the area types of the blocks adjacent to each other are the same type (claim 3).

Further, according to the present invention, for the intermediate area where the blocks on both sides are the halftone dot area and the photograph area, i rows ×
A dither matrix processing means for applying a j-column dither matrix is provided (claim 4).

Further, according to the present invention, for an intermediate area in which only one of the blocks on both sides is a character area, a dither matrix consisting of an integer k smaller than the above i (or j) is connected in the direction connecting both sides. A small dither matrix processing means to be applied is provided (Claim 5), and the integer k.
May be 1 (claim 6).

[0012]

According to the first aspect of the invention, the original image in which characters, halftone dots, and photographs are mixed is captured in pixel units via the optical reading means or the like. Number of pixels N rows x M columns (N >> i, M>
Areas are allocated to the input image of> j) in units of blocks of i rows × j columns (i and j are integers of 2 or more) pixels. That is, the character area, the halftone dot area, the photograph area, and the intermediate area that does not belong to any of the above areas are sorted, and further, for the intermediate area, either in the row direction or the column direction. According to the area type of the neighboring blocks, the images are subdivided so as to have a natural continuity (no image distortion) with the neighboring images. Then, a dither matrix corresponding to the type of the classified (including subclassified) images is adopted, and image processing with affinity is performed as a whole.

According to the second aspect of the present invention, the sub-classification is performed according to the area type of the blocks which are adjacent to each other in either the row direction or the column direction. That is, when the area types of the blocks on both sides are character and character, halftone dot and halftone dot, photo and photo, and character and halftone dot, character and photo, halftone dot and photo, there are a total of six types. Subclassification is performed. It should be noted that if at least the blocks on both sides are characters and halftone dots, characters and photos, and halftone dots and photos, an intermediate area classification is performed to reproduce an image with natural continuity for different types of images. To be done.

According to the third aspect of the invention, when the area types of the blocks on both sides are the same type, the blocks are sub-classified into the same type as the type of the block, and the isolated error is corrected.

According to the fourth aspect of the invention, the i × j dither matrix is applied to the intermediate area in which the blocks on both sides are the halftone area and the photographic area, and the halftone dot and the photographic area are applied. We are aiming for an affinity with the boundary with.

According to the fifth aspect of the invention, for the intermediate area in which only one of the blocks on both sides is a character area, an integer k smaller than the above i (or j) is formed in the direction connecting both adjacent blocks. By applying a dither matrix, the boundary between the character and the halftone dot or the boundary between the character and the photograph is made compatible. For example, in the case of i, j = 4, if k = 2, a dither matrix of 2 rows × 4 columns (when both sides are in the row direction) or 4 rows × 2 columns (when both sides are in the column direction) is Two will be used for each.

According to the invention described in claim 6, the integer k is set.
Is 1, so i is 1 row × j column, or i row × 1
J column dither matrices will be used.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT FIG. 1 is an overall block diagram of a mixed image processing apparatus according to the present invention. Reference numeral 1 denotes an image input unit for reading and storing a so-called mixed document in which characters, patterns (dots, photographs), etc. are mixed in the document, and the image reading unit 11 and the pixel memory 1 are provided.
2 and an input line buffer 13.

The image reading means 11 is, for example, an image scanner or the like in which optical reading elements are arranged in at least a line or a matrix, and is capable of relative scanning with respect to a document. This input section 1
Has a resolution of, for example, about 400 DPI (dots / inch), and is capable of reading each pixel at multiple levels, for example, at a power of 2 (a = 2 or more integer, for example, a = 8) level.

The pixel memory 12 stores each pixel data of the original image read by the image reading means 11, and N
It has an address of (row) × M (column). The writing of the image data to the pixel memory 12 is performed in synchronization with the scanning of the document by the image reading means 11 by the write address from the control circuit 2. Therefore, only the image is digitized with the document as it is. It is designed to be remembered. The input line buffer 13 stores the pixel data for each row output from the pixel memory 12 for three rows, and in this state, the pixel data for three rows is synchronized in the column direction, and the block memory 31 and the area described later are stored. It is output to the separate processing circuit 43. The image input unit 1 is adapted to output image data that has already been read and temporarily stored and stored in a recording medium such as a floppy disk to the pixel memory 12 of the present apparatus or to the input line buffer 13. Things are also included.

Reference numeral 3 denotes an area classification unit for allocating areas based on pixel data from the image input unit 1, which includes a block memory 31, a density value calculation circuit 32, an edge amount calculation circuit 33,
The change amount calculation circuit 34, the area determination circuit 35, the determination result storage circuit 36, and the intermediate area redetermination circuit 37 are included. As will be described later, the area classification unit 3 sorts into a character area, a halftone dot area, a photographic area, and an indeterminate intermediate area that does not belong to any of these areas. In this way, by providing the intermediate area, it becomes possible to make each judgment criterion of the character area, the halftone dot area, and the photograph area strict, and it is possible to reduce the classification error.

The block memory 31 stores the pixel data for three rows from the input line buffer 13 and has an address of 3 rows × 3 columns.

The control circuit 2 generates a write address and a read address of pixel data to the pixel memory 12, and outputs a required operation timing signal to the area classification section 3 and the halftone processing section 4. It is a thing.

Concentration value calculation circuit 32, edge amount calculation circuit 3
3 and the change amount calculation circuit 34 perform a predetermined density distribution statistical processing calculation for area classification, and lead the calculation result to the area determination circuit 35. The area determination circuit 35 determines whether each block is divided into a character area, a halftone dot area, a photograph area, and an intermediate area based on the input calculation results from the respective circuits 32 to 34. The determination result is the determination result. It is temporarily stored in the storage circuit 36. Intermediate area re-determination circuit 3
Reference numeral 7 is for further subclassifying the area determined as the intermediate area by the area determination circuit 35.

The area classification process will be described below with reference to FIG. As shown in FIG. 2, S (1,1) to S (N, M) are stored in the pixel memory 12 as pixel data. Then, the classification is performed in block units of a predetermined size. In this embodiment, for the sake of convenience of description, squares of 3 rows × 3 columns are adopted. In addition,
The size of this block can be i rows × i columns (i = 2 or more integer, i << N, M), and may be i = 6 or i = 9, for example. As the value of i, it is preferable to adopt a value suitable for the system used.

The control circuit 2 generates a read address for reading the pixel data S in the pixel memory 12 into the block memory 31 through the input line buffer 13 by 3 rows × 3 columns, and leads the read address to the pixel memory 11. For example, regarding the first block, the block memory 31 stores S (1,
1), S (1, 2), S (1, 3), S (2, 1), S
(2,3), S (2,3), S (3,1), S (3,3
2), pixel data of S (3, 3) are captured, and then S (1, 4), S (1, 5), S (1, 6), S
(2,4), S (2,5), S (2,6), S (3
4), S (3,5) and S (3,6) are taken in, and finally, S (N-2, M-2) and S (N-
2, M-1), S (N-2, M), S (N-1, M-)
2), S (N-1, M-1), S (N-1, M), S
Pixel data of (N, M-2), S (N, M-1), S (N, M) are captured. However, N and M are multiples of three. Every time 9 × 3 × 9 pieces of pixel data are input, the block memory 31 sequentially reads them out to the density value calculation circuit 32, the edge amount calculation circuit 33, and the change amount calculation circuit 34 in the same order as at the time of input. Then, the respective circuits 32 to 3
In 4, the calculation for the concentration distribution statistical processing based on the conditional expression described later is executed.

Now, for a pixel S (n, m) in a certain block, the maximum density value is represented by Max, the minimum density value is represented by Min, the average density value is represented by ZZ, and a density threshold value of TH1 to TH8 is 8 It is assumed that four values KH, KV, HL, LL, and KD are prepared as the determination threshold values.

First, the density value calculation circuit 32 and the change amount calculation circuit 34 execute the calculation of z + = S [n] [m] (1) to cumulatively add the densities in the block. Next, the average value of the densities in the block is calculated by ZZ = z / n × m (2) On the other hand, the density change amount is calculated by MM = Max-Min (3), and the area determination circuit 35 executes the conditional expression (4) if (MM> TH1) (4). If the condition is satisfied, the calculation result by the edge amount calculation circuit 33 is used as described later, while if the condition is not satisfied, the density value calculation circuit 3 is used.
The calculation result of 2 is used.

The edge amount calculation circuit 33 first determines if (abs (S [n] [m] -S [n] [m + 1])> MM / TH2) (5) for the pixels in the area. ) Is executed and the value Kh is set to 1 only when the condition of Expression (5) is satisfied.
Is cumulatively added, and if (abs (S [n] [m] -S [n + 1] [m])> MM / TH2) ... (6) is executed, the condition of expression (6) is satisfied. Value Kv is 1 only when
Is cumulatively added.

Further, if (S [n] [m]> TH3) (7) is executed, and only when the formula (7) is satisfied, 1 is cumulatively added to the value Hl to determine the number of black pixels. Along with the calculation, if (S [n] [m] <TH4) (8) is executed, and only when the formula (8) is satisfied, 1 is cumulatively added to the value Ll to obtain the number of white pixels, Further, Kd = abs (Kh-Kv) (9) is calculated. Then, the area determination circuit 35 executes if (Kh> KH and Kv> KV and Kd> KD) (10) using the results of the above equations (5) to (9) to satisfy the equation (10). Then, it is classified as a halftone dot area. On the other hand, if expression (10) is not satisfied, if (Hl> HL or Ll> LL) (11) is executed, and if expression (11) is satisfied, it is classified as a character area as a line sentence. On the other hand, if the formula (11) is not satisfied, it is classified as the intermediate region.

Meanwhile, if (Max> TH6 and ZZ> TH7) (12) is executed using the density maximum value Max and the average value ZZ obtained by the density value calculation circuit 32, the expression (12) is satisfied. Then, the text area is classified as a black text. On the other hand, if expression (12) is not satisfied, if (ZZ <TH8) (13) is executed, and if expression (13) is satisfied, it is classified as a white area as a character area. Then, if the expression (13) is not satisfied, it is classified as a photograph area.

In this way, the area determination circuit 35 sequentially classifies each block into the area type, and the result is stored in the determination result storage circuit 36. In addition, since the character areas are divided into line sentences, black sentences, and white sentences, it is possible to apply a reproduction dither matrix having a suitable threshold value for each type, even in the same character region. Therefore, it is possible to reproduce an image that matches each characteristic.

For the blocks classified as the intermediate area, the intermediate area determination circuit 37 executes a conditional expression for further classification. That is, the intermediate area determination circuit 37 determines whether the area type determined by the area determination circuit 35
The block classified as the intermediate region is subjected to isolated point removal processing and boundary area processing. In the isolated point removal processing, when the classification type of the k-th block is represented by B (k), if (B (k) = intermediate region) and if (B (k-1) = B (k + 1)) then (B (k) = B (k-1))… (14). However, k-1 and k + 1 that are on both sides of kth
The number is the column direction in this embodiment. Therefore, the intermediate region re-determination circuit 37 is simultaneously guided from the determination result storage circuit 36 at least the k-th and k + 1-th adjacent k-1 and k + 1-th determination results. In addition, when the type determination of the intermediate area is other than the k-1th and k + 1th, for example, when further improving the accuracy by considering the determination results of the blocks k-2th and k + 2nd on both sides thereof, Are also guided to the intermediate area re-determination circuit 37 at the same time.

When the boundary region processing does not satisfy the equation (14), it is reclassified as follows according to the types of B (k-1) and B (k + 1). That is, if B (k-1): B (k + 1) is character: halftone dot (hereinafter referred to as sentence / halftone), halftone dot: photograph (hereinafter referred to as halftone / copy), photograph: character This is the case (hereinafter referred to as a copy / sentence). When the determination results of the k-2nd block and the k + 2nd block are taken into consideration, the accuracy can be improved as in the case of the correction of the isolated error.

In the case of performing the intermediate area determination for the blocks in the column direction as in the present embodiment, the area determination circuit 35 performs the determination processing only for B (1), ..., B (M / 3). The re-determination process by the intermediate area re-determination circuit 37 is started at the time of execution. Alternatively, the processing speed can be increased by sequentially performing the intermediate area determination at the time when the area determination of the adjacent block in the column direction, which is necessary for re-determining the intermediate area, is completed.

FIG. 3 is a flow chart for explaining the area type determination operation. First, the pixel data of the block to be determined is loaded into the block memory 31, and the arithmetic circuit 3
Statistical processing of concentration distribution according to 2-34 (formula (1) -formula (13))
Is performed (steps S2 and S4). Region determination is performed based on this processing (step S6), and the regions are classified into a character region, a halftone dot region, a photograph region, and other intermediate regions, and stored in the determination result storage circuit 36 (step S).
8 to S14). Next, for the blocks classified into the intermediate area, the intermediate area re-determination circuit 37 further determines the sentence / mesh area, halftone / copy area, copy / sentence area, or isolated area according to the area type of the peripheral block in the column direction. A subclassification of point reprocessing is performed (steps S16 and S18). Then, it is determined as a sentence / halftone area, halftone / copy area or copy / text area (step S
20 to S24), and if it is an isolated point reprocessing, the above equation (1
4) is executed (step S26).

Further, when the intermediate area type is determined based on the relationship with the blocks adjacent in the row direction, it may be performed after the determination processing of the area determination circuit 35 is completed for all the blocks of the document. Good. In this case, since the image reproduction is performed in the column direction as described later, a memory for temporarily storing the determination result including the intermediate area type is required. As a memory for this purpose, a memory corresponding to the judgment result storage circuit 36 may be used, and the judgment result of each block may be directly output from the judgment result storage circuit 36 to the halftone processing unit 4 in the column direction in sequence.

The operation for determining the area type in this case will be described with reference to the flowchart of FIG.

First, the pixel data of the block to be judged is fetched into the block memory 31, and the arithmetic circuits 32 to 34.
Thus, the concentration distribution statistical processing (Equations (1) to (13)) is performed (steps S32 and S34). Region determination is performed based on this processing (step S36), and the regions are classified into a character region, a halftone dot region, a photograph region, and other intermediate regions, and stored in the determination result storage circuit 36 (step S36).
38-S44). Next, for the blocks classified into the intermediate area, the intermediate area re-determination circuit 37 further determines, depending on the area type of the peripheral block in the row direction, the sentence / halftone area, halftone / copy area, copy / sentence area, or isolated area. A subclassification of point reprocessing is performed (steps S46 and S48). Then, it is determined to be a sentence / halftone area, halftone / copy area or copy / text area (steps S50 to S54), and if it is an isolated point reprocessing, the above formula
It is executed by replacing (14) in the row direction (step S5).
6).

The halftone processing section 4 comprises an area-specific processing pattern generating circuit 41, a processing pattern selector 42 and an area-specific processing circuit 43.

The area-specific processing pattern generation circuit 41 uses predetermined dither patterns for each pattern for characters, halftone dots, and photographic areas, and for sentence / mesh, halftone / copy, and copy / sentence for intermediate areas. It has a pattern ROM pre-written in the form of a matrix and a threshold ROM for storing a threshold value for determining the density given to each dot forming the dither matrix of each pattern. The pattern of the dither matrix selected by the device 42 and the threshold data corresponding thereto are output.

Each of the above patterns is 1 for the character area.
It is a dither matrix of × 1 and is composed of a 3 × 3 dither matrix for halftone dots and photo areas, and a 3 × 1 dither matrix for each area of sentence / mesh and copy / sentence. Thus, the halftone / image area is composed of a 3 × 3 dither matrix.

The sentence / mesh and copy / sentence areas are defined as a 3 × 1 dither matrix pattern. One of the two adjacent areas is a character area, and this character area is made compatible with the halftone dot area and the photograph area. In order to do so, it is considered that the boundary between regions can be made continuous without unnaturalness by performing gradation processing with a 3 × 1 dither matrix and outputting this for three columns.

5 and 6 show intermediate processing patterns. FIG. 5 shows a case where the columns are aligned with each other.
(A) is a 3 × 1 sentence / halftone pattern when both sides are a text area and a halftone area, and (b) is a 3 × 3 halftone / printing pattern when both sides are a halftone area and a photographic area. (C) shows a 3 × 1 copy / sentence pattern when both sides are a photograph area and a character area. 6A and 6B show a case where the lines are aligned with each other, where FIG. 6A shows a 1 × 3 sentence / halftone pattern when both sides are a text area and a halftone area, and FIG. In the case of 3 * 3 halftone / printing pattern, (c) shows a 1 * 3 copy / sentence pattern when both sides are a photograph area and a character area.

If the threshold value of each dot forming the dither matrix is 1, and the value 128 is set for 256 density gradations, the input density of the pixel data from the input line buffer 13 is assumed. Is 128 or more, it is output as black, and 127 or less is output as white. On the other hand, a plurality of thresholds may be set, and in this case, the corresponding density value is captured in the corresponding dot by the magnitude logic of the input density and each threshold, and the number corresponding to the number of thresholds is taken. The density gradation signal may be output to each dot by using an output device such as a laser.

The region-specific processing circuit 43 selects the pixel data output from the input line buffer 13 in units of 3 rows × 3 columns from the region-based processing pattern generation circuit 41 according to the determined region type and outputs the selected pixel data. The replacement is performed with the pattern. For example, if the pixel data of 3 rows × 3 columns is for the character area, the 1 × 1 dither matrix for the character area is output together with the threshold value, and the density values of the 9 pixel data are compared and compared. Replaced by multi-valued data. If the pixel data of 3 rows × 3 columns is a halftone dot or a photograph area, a 3 × 3 dither matrix for the halftone dot or photograph area is output together with the threshold value of each dot, and 9 pixel data And the density value for each dot is compared and replaced with binary or multi-valued data. Further, if the pixel data of 3 rows × 3 columns is for a certain intermediate area, the 3 × 1 dither matrix (in the case of sentence / mesh, copy / sentence area) for the corresponding intermediate area is three times, Alternatively, a 3 × 3 dither matrix (in case of halftone / image area) is output once together with the threshold value of each dot, and the density value is compared for each dot in units of dither matrix with each of the 9 pixel data, and then 2 Value or multivalued (when multiple thresholds are set) data is replaced. When determining the type of the intermediate area in the row direction, as shown in FIG.
The 1 dither matrix is a 1 × 3 dither matrix.

The image output section 5 is for making a hard copy of the image replaced with the image for reproduction, and comprises an output line buffer 51 and an image output means 52. The output line buffer 51 temporarily stores the pixel data replaced by 3 × 3 for each column direction, and the pixel data for each row is sequentially output serially in the column direction on the output side. Has been done. The image output unit 52 is composed of a laser light emitting element and a polygon mirror for surface scanning, and the pixel data for one row output from the line buffer 51 is time-sequentially applied to the photoconductor charged to a predetermined potential. The image is output and an electrostatic latent image as a reproduced image is formed on the surface of the photoconductor. The electrostatic latent image is transferred to a transfer sheet fed in synchronism with the electrostatic latent image forming operation by a sheet feeding means (not shown) so that a hard copy can be obtained.

In this embodiment, the block is i rows × i.
Although the column has a square shape, the present invention is not limited to this, and may have a rectangular shape of i rows × j columns. Then, the pattern of the dither matrix in the intermediate area may be set according to the block shape.

[0049]

According to the present invention described in claim 1, a character area, a halftone dot area, a photograph area and an intermediate area which does not correspond to any of the above areas for each block of i rows × j columns with respect to an input image. And a sub-classification means for sub-classifying the intermediate area according to the area type of a neighboring block in one of the row direction and the column direction. It is possible to reproduce an image having natural continuity with the image and having no image distortion, and it is possible to reproduce a high-quality image having affinity as a whole.

According to the second aspect of the invention, if the sub-classification is performed according to the area type of the blocks on both sides,
The configuration of the fine classification means can be simplified.

According to the third aspect of the present invention, when the area types of the blocks on both sides are the same type, the blocks are classified into the same type as the block type. Therefore, the isolated error can be corrected and the isolated points can be removed. obtain.

According to the fourth aspect of the invention, the dither matrix processing means for applying the i × j dither matrix to the intermediate area in which the blocks on both sides are the halftone dot area and the photographic area is provided. Therefore, the affinity of the boundary between the halftone dot and the photograph can be suitably achieved.

According to the fifth aspect of the invention, for the intermediate area in which only one of the blocks on both sides is a character area, an integer k smaller than the above i (or j) is formed in the direction connecting both sides. Since the small dither matrix processing means for applying the dither matrix is provided, the boundary between the character and the halftone dot,
Alternatively, the affinity of the boundary between the character and the photograph can be suitably achieved.

According to the sixth aspect of the invention, the dither matrix has one row (or one column) for the intermediate area in which only one of the blocks on both sides is the character area. The sex can be further enhanced.

[Brief description of drawings]

FIG. 1 is an overall block diagram of a mixed image processing apparatus according to the present invention.

FIG. 2 is pixel data S (1,1) to S in a pixel memory.
It is a figure which shows the relationship between (N, M) and an address.

FIG. 3 is a flowchart illustrating an operation of determining a region type in a column direction.

FIG. 4 is a flowchart illustrating an operation of determining a region type in a row direction.

5A and 5B are pattern diagrams of an intermediate area in the case of making them compatible with each other in the column direction. FIG. 5A is a 3 × 1 sentence / halftone pattern when both sides are a character area and a halftone area, and FIG. 3 × 3 halftone / printing pattern when adjacent to halftone dot area and photo area, (c)
Shows a 3 × 1 copy / sentence pattern when both sides are a photograph area and a character area.

FIG. 6 is a pattern diagram of an intermediate area in the case of matching in the row direction, (a) shows 1 when both sides are a character area and a halftone dot area.
3 × 3 sentence / mesh pattern, (b) 3 × 3 mesh / copy pattern when both sides are a dot area and a photo area, (c) 1 × when both sides are a photo area and a character area 3 shows a copy / sentence pattern of No. 3.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 image input section 11 image capturing means 12 pixel memory 13 input line buffer 2 control circuit 3 area classification section 31 block memory 32 density value calculation circuit 33 edge amount calculation circuit 34 change amount calculation circuit 35 area determination circuit 36 determination result storage circuit 37 Intermediate Area Re-Determination Circuit 4 Halftone Processing Section 41 Area-Specific Processing Pattern Generation Circuit 42 Processing Pattern Selector 43 Area-Specific Processing Circuit 5 Image Output Section 51 Output Line Buffer 52 Image Output Means

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location 7459-5L G06F 15/70 330 E

Claims (6)

[Claims] 1. A mixed image processing apparatus for performing mixed image processing using a dither matrix according to the type of classified images, which is an image input for fetching an original image in which characters, halftone dots, and photographs are mixed in pixel units. Parts and i rows xj for the input image
Classification means for allocating each block of pixels of column (i and j are integers of 2 or more) to a character area, a halftone dot area, a photograph area and an intermediate area not classified into any of the above areas, and the intermediate area in the row direction. And a sub-classification unit for sub-classifying according to the area type of a neighboring block in one of the column direction and the column direction. 2. The mixed image processing apparatus according to claim 1, wherein the sub-classification unit performs sub-classification in accordance with the area types of blocks that are adjacent to each other. 3. The subclassifying means, when the area types of the blocks that are adjacent to each other are of the same type, classifies the blocks into the same type as the type of the block. Mixed image processing device. 4. The mixed image processing apparatus according to claim 1, wherein the blocks adjacent to each other are an i-row × j-column dither for an intermediate area in which a halftone dot area and a photographic area are present. A mixed image processing apparatus comprising a dither matrix processing means for applying a matrix. 5. The mixed image processing apparatus according to any one of claims 1 to 4, wherein, for an intermediate area in which only one of both adjacent blocks is a character area, the i in the direction connecting both adjacent (Or j) A mixed image processing device comprising a small dither matrix processing means for applying a dither matrix having an integer k smaller than (or j). 6. The mixed image processing apparatus according to claim 5, wherein the integer k is 1.