KR19990025680A - Block superposition based image binarization apparatus and method - Google Patents

Abstract

The present invention provides a block-based image binarization method in which binarization is performed by dividing a binarization target image in units of blocks, and determining the area of the block of interest while dividing the binarization target image so as to overlap the horizontal and vertical directions in units of blocks. By performing the method of performing binarization on all pixels located within the block of interest without performing binarization on the pixels located at the boundary of the block of interest by performing the binarization on all blocks. A block overlap-based image binarization apparatus and method capable of binarizing while sufficiently reflecting distribution characteristics of pixels are provided.

According to the present invention, the uneven phenomenon or false boundary caused by excessively performing binarization may be minimized by applying the same region determination result to the pixels located at the boundary of the block or the pixels located inside the block. The picture quality improvement can be achieved.

Description

Block superposition based image binarization apparatus and method

The present invention relates to an Apparatus and Method for Block Overlap Based Image Binarization, and more particularly, to a block-based image binarization method for performing binarization by dividing a binarization target image into block units. The pixels to be separated are overlapped in units of blocks so that the pixels positioned at the boundary of the block of interest are not binarized according to the determination result of the block of interest, and are binarized only for pixels located inside the block of interest. The present invention relates to a block overlap-based image binarization apparatus and method that can be binarized while sufficiently reflecting distribution characteristics of adjacent pixels by repeatedly performing a process of performing the process.

One of the typical imaging systems, the scanner (ie, reading means) is the most common means for reading printed texts, photographs, human-written letters or pictures, and includes a multifunction device, a document translator, and CAD (Computer Aided Design). Essential components such as computer, facsimile, document recognizer and digital copier.

In recent years, as office electronics, which are at the crossroads of development, is rapidly increasing in demand for office automation devices such as printers, scanners, and facsimile machines, each office automation device is developed with high performance to expand its own functions. In addition, it is a trend to produce and provide a product that performs a complex document output function at the same time to reduce the economic burden and installation space for the user by developing each office automation equipment that was used independently as an integrated.

In particular, in devices in which a scanner capable of reading a document is built in, such as a facsimile machine or a multi-function peripheral (MFP), an image having multiple levels of gradation is read through a scanner and then read into the read data. The characteristics of the binarization method have become the main evaluation items in evaluating the performance of the scanner.

In general, "image binarization" refers to a gray level image expressed as a multi-step gray level in an image system, in which a quantization level is 1 bit, that is, '1'. An image processing technique used to represent a binary image (or binary image) of '0', which is capable of preserving the characteristic information of a target image while reducing the amount of processing data. It is widely used in the fields of image compression, image display and printing, image recognition, and the like.

Due to such a wide range of applications, various image binarization methods are known through image processing related books, papers and patent publications.

The most common image binarization method is to obtain a binarized image by determining an arbitrary threshold, assigning 1 to a pixel value greater than this threshold and 0 to a pixel value less than that threshold. Thus, image binarization processing is a kind of threshold processing for classifying gray levels up and down based on a threshold value. The most important thing in the threshold processing is the determination of a threshold value, and the simplest method is to take the median value of image gray levels as a threshold value. There are also ways to determine more precise thresholds through intensity histograms.

On the other hand, in order to properly binarize the edge regions of the visually important image, there is a method of performing binarization after emphasizing the outline of the original image through a predetermined edge enhancement filter. It is known that a good binarization image can be obtained as compared with the method that does not include the emphasis process.

An example thereof will be described with reference to FIG. 1.

If the gradation value of the original image is given in Table 1,

78 77 76 93 152 176 227 222

When spatial filtering is performed using a 1 × 3 edge enhancement mask as shown in Table 2,

-One 3 -One

You can get the result with the outline outlined in Table 3.

77 58 51 187 149 255

Fig. 1 is an exemplary diagram showing the gradation value of the original image and the gradation value of which the outline is highlighted, wherein the solid line represents the gradation value of the original image, and the dotted line represents the gradation value of the image as a result of emphasizing the outline.

In this case, when the threshold is set to 150, 78, 58, and 51 in Table 3 are set to 0, 187 is set to 1, 149 is set to 0, 255 is set to 1, and the pixel set to 1 is represented by white, and 0 When the pixel set to is black, two lines of outlines appear in each of the 187, 149, and 255 areas where the outlines are distributed, based on 149, respectively, resulting in deformation or distortion of the original image. do.

As such, even when the binarization method including the contour enhancement process, which is known to have relatively superior binarization characteristics, is used, a false pixel is generated in the contour region as described with reference to FIG. 1. ), A local adaptive image binarization method that can reflect the distribution characteristics of adjacent pixels is expected.

Applicants of the present invention has applied for a number of local image binarization apparatus and methods in order to meet such technical trends, and various local image binarization methods are known by patent applications of a plurality of applicants in addition to the applicant of the present invention. It is a state.

The present invention is directly related to local image binarization in which binarization is performed by reflecting characteristics of adjacent pixels located around the pixel of interest in performing image binarization. Hereinafter, representative localization among localized image binarization methods according to the prior art will be described. Embodiments of the image binarization method will be briefly described as follows.

2 is an exemplary view for explaining an embodiment of a local image binarization method according to the prior art.

In one embodiment of the localized image binarization method according to the related art, a plurality of adjoining pixels positioned around the same line as a pixel of interest, which is a pixel to be binarized at this point in time, are selected, and the pixel of interest and the adjacent pixel are selected. An adjacent pixel group composed of pixels is generated. For example, as shown in FIG. 2A, when n adjacent pixels are selected to the left and right of the pixel of interest, the total number of pixels including the pixel of interest and the adjacent pixels, that is, the size of the adjacent pixel group is 2n +. It becomes one. Alternatively, the size of the adjacent pixel group may be set to 16 pixels, which are macro block sizes.

Subsequently, the pixels having the maximum gray value are compared and calculated in the adjacent pixels group including the selected adjacent pixels and the center pixel, and the proportional constant preset to the maximum gray value is calculated. After determining the product multiplied by) as the local threshold, pixels of interest smaller than the local threshold are binarized to zero, otherwise the pixels of interest are binarized to one.

For example, this will be described with the example shown in FIG. 2B. In general, the example of FIG. 2B is extreme as even the system between pixels is doubled around the outline. It can be appreciated that this is not an example.

In (b) of FIG. 2B, when pixels (ie, pixels having a gradation value of 81) located in 2 rows and 6 columns are selected as the pixel of interest and the size n of the adjacent pixels is set to 3 for convenience of description, the adjacent pixel group is 78, 77,76,81,152,157,158. Here, the pixel having the maximum gray value is 158, and as in the usual case, the proportional constant ( Is 0.5, the local pixel of interest is 76, so that the pixel of interest 81 is binarized to one.

Subsequently, when the pixels positioned in two rows and seventh columns (that is, pixels having a gray scale value of 151) are selected as the pixel of interest in the raster scanning order, the adjacent pixel group is 77,76,81,152,157,158,160. Therefore, the maximum gray value is 160, and the proportional constant ( Is 0.5, the local pixel of interest is binarized to 1 as the local threshold becomes 80.

Intuitively, in FIG. 2 (b), the outline is bounded by the pixel 81 and the pixel 152, so that the pixel 81 is preferably zero, and the pixel 152 is binarized to 1, but the opposite results are provided. You can check it.

As another embodiment of the local image binarization method according to the prior art, another example using the global threshold and the local threshold will be introduced.

First, the global threshold value is determined by half of the maximum gray scale value of the neighboring pixel group according to the conventional art as described above, or the sum of all gray scale values of the binarization target image is obtained to average the total gray scale value. The value is determined by multiplying the proportional constant or is determined by averaging the pixel values of adjacent pixel groups.

Subsequently, when the size of the adjacent pixel group is 16 pixels, the maximum gray value is obtained from 16 pixels, and the average value of the maximum gray value and the global threshold is set as a local threshold to perform binarization. However, if the maximum gray value is smaller than the global threshold, binarization is performed by applying a global threshold.

As described above, although the local image binarization method according to the related art performs local image binarization in consideration of information on adjacent pixels, it is difficult to sufficiently consider complicated and irregular contour components by simply using only the maximum gray level value of the adjacent pixel group. have. Also, when binarizing an image having a clear contrast, an excellent result may be generally obtained. Otherwise, a large amount of outline information may be lost.

On the other hand, when considered from another aspect, according to one embodiment and another embodiment of the localized image binarization method according to the prior art, in a document image containing a character, an adjacent width in which a width inside the character is selected and constituted by an adjacent pixel. When larger than the size of the pixel group, a problem as shown in FIG. 3 occurs. This is due to the fact that the local threshold determined by reflecting the gradation distribution of adjacent pixels is not continuously maintained inside the character.

3 is an exemplary diagram for explaining a problem in binarizing character images in one embodiment and another embodiment of a local image binarization method according to the prior art.

In other words, assuming that the size of the adjacent pixel group is 16 pixels, as shown in FIG. 3, when the internal width b1 of the character exceeds 16 pixels, 8 pixels (a1, a2) near both contours of the character. In this case, the binarization is properly performed, whereas in the region b2 inside the character excluding the two pixels, the local threshold value and the pixel value of interest do not differ so much that binarization is not performed properly.

On the other hand, as another embodiment of the local image binarization method according to the prior art, Shinji Tetsudani and Hiroshi Ochi, which are considered to belong to the similar category of the present invention, are published in the Journal of the Institute of Electronics and Telecommunications in Japan. An image binarization method by block-based area separation, such as a binarization processing method of a manuscript, in which a binarized image and a shaded image contributing to J67-B No. 7 are mixed, will be described with reference to the accompanying drawings.

Another embodiment of the local image binarization method according to the related art includes pixels located in a block of interest by dividing a binarization target image having a mixture of characters and photos in a unit of a block of a predetermined size so as not to overlap with respect to the horizontal and vertical directions. The maximum gray value and the minimum gray value are compared and calculated in the block pixel group consisting of the? Then, the maximum block that is even the system of maximum and minimum gray values is obtained, and even the maximum system is compared with a preset even threshold, and the block of interest depends on whether or not even the maximum system is larger than the threshold. Determine whether it is an area or a photograph area. If the maximum system is even larger than the threshold, the character is binarized to a fixed threshold; otherwise, it is determined to be a photographic domain and then binarized by the dither method, a well-known halftoning process method. do.

The above-described process is repeated for all blocks included in the binarization target image.

However, another embodiment of the local image binarization method according to the prior art, which is one of the relatively advanced local binarization methods in the prior art, determines whether each block is a character area or a photo area after dividing the binarization target image in units of blocks. As shown in FIG. 4, when switching from the background to the text area or the text area to the background at the boundary of a block, both blocks adjacent to it are determined to be the picture area, which causes deterioration in image quality. To become a big factor.

FIG. 4 is an exemplary view for explaining a problem of another embodiment of the localized image binarization method according to the prior art. FIG. 4A illustrates an example of an image signal that can be seen at the boundary of a character. 4 (b) shows a dither threshold matrix used when binarization is performed by a dither method, which is a kind of halftoning process technique when it is determined as a photographic region.

For simplicity of explanation, here we specify the block size At 4, the number of levels of the pixel gradation value is set to 16 gradations (that is, 4 bits), and even the threshold is set to 7.

In FIG. 4A, the left 4 Since the maximum gray value is 8 and the minimum gray value is 1 in 4 blocks, the block is determined as the picture area P because even the maximum system is 7 and even the maximum system is equal to the threshold. Also, the right 4 Since the maximum gray value is 16 and the minimum gray value is 14 in four blocks, this block is also determined as the photographic area P because even the maximum system is 2, even the maximum system is smaller than the threshold.

Figure 4 (c) shows the result of such a decision, Figure 4 (d) performs the binarization through the dither threshold matrix shown in Figure 4 (b) as two blocks are determined to be a picture area. One result is shown.

As can be seen from the binarization result of FIG. 4 (d), FIG. 4 is a block diagram of a local image binarization method according to another embodiment of the prior art. Since the binarized result determined in the photographic region appears in the form of irregularities, it is a factor that directly inhibits the image quality.

A defect of another embodiment according to the prior art is that the pixels or blocks located at the boundary of the block are preferably not binarized because information on the grayscale values of adjacent pixels is insufficient in the case of the pixels located at the boundary of the block. This is caused by excessively performing binarization by applying the same region determination result to the pixels located inside of.

Accordingly, the present invention has been devised in view of such a problem, and in the block-based image binarization method of dividing the binarization target image into block units and performing binarization, the binarization target images are superimposed on the horizontal and vertical directions in units of blocks. The region determination of the block of interest is performed while dividing so as to overlap, and the binarization is performed only on pixels located inside the block of interest without performing binarization on the pixel located at the boundary of the block of interest. It is an object of the present invention to provide a block overlap-based image binarization apparatus and method that can be binarized while sufficiently reflecting a distribution characteristic of adjacent pixels by repeatedly performing a process of performing the process.

1 is an exemplary diagram showing a gray value of an original image and a gray value with an outline highlighted;

2 is an exemplary diagram for explaining an embodiment of a localized image binarization method according to the prior art;

3 is an exemplary diagram for explaining a problem in binarizing character images in one embodiment and another embodiment of a local image binarization method according to the prior art;

4 is an exemplary diagram for explaining a problem of another embodiment of the local image binarization method according to the prior art;

5 is a flowchart of a preferred embodiment of a block overlap based image binarization method according to the present invention;

6 is a block diagram showing a shuttle scanner module that is a key component of a shuttle scanning based image input system;

7 is an exemplary view showing a shuttle block for scanning an A4 size original with a shuttle scanner module;

8 is an exemplary diagram for describing a method in which one shuttle block is scanned;

9 is a four pixel according to the prior art. An illustration showing an example of block division of four pixels,

Figure 10 is a four pixel according to a preferred embodiment of the present invention An illustration showing an example of block division of four pixels,

11 is an exemplary view for explaining an application example of a preferred embodiment of the block overlap-based image binarization method according to the present invention;

12 is a flowchart of another embodiment of a block overlap based image binarization method according to the present invention;

13 is a block diagram showing a preferred embodiment of a block overlap based image binarization apparatus according to the present invention;

14 is a block diagram illustrating another embodiment of a block overlap-based image binarization apparatus according to the present invention.

<Description of the code used in the main part of the drawing>

101: block overlap division 102: block selection unit

110: gradation information calculation unit 111: gradation comparison unit

112: maximum gradation value storage unit 113: minimum gradation value storage unit

120, 220: threshold calculation section 121: subtraction section

122: adder 123: divider

124: multiplication unit 130: even the comparison unit

140: first block-based binarization unit 150: dithering binarization unit

160, 260: output multiplexer 151: dither threshold matrix storage

250: second block-based binarization unit

In order to achieve the object of the present invention, the block superposition-based image binarization method of the present invention, in the image binarization method for performing binarization by reflecting the pixel distribution characteristics of the binarization target image:

A block overlap division step of dividing the binarization target image into blocks having a predetermined size so as to overlap at least one of a horizontal direction and a vertical direction to generate a block group composed of a plurality of blocks;

A block selection step of sequentially selecting a block of interest to be binarized at this time in the block group according to a predetermined scanning order;

A block classification step of classifying a pixel distribution characteristic in a block pixel group including pixels located in the block of interest and classifying an area to which the block of interest belongs as one of two or more areas; And

And a block-based binarization step of binarizing pixels included in the target block through different thresholds according to a region to which the target block belongs.

In the feature according to the method of the present invention, the step of classifying the block comprises: comparing and calculating a maximum gray value and a minimum gray value in a block pixel group consisting of pixels located in the block of interest; And

And an area determining step of calculating an even maximum gradation even between the maximum gradation value and the minimum gradation value and judging the area to which the block of interest belongs according to whether even the maximum gradation is greater than a preset threshold. desirable.

Further, in the block-based binarization step, if the maximum system is even greater than the threshold as a result of the determination of the area, the pixels included in the block of interest are averaged between the maximum gray value and the minimum gray value. It is preferred to include a first block based binarization step of binarizing through a first threshold proportional to the value.

On the other hand, in the block-based binarization step, if the maximum system is even smaller than the threshold as a result of the determination of the region, binarizing the pixels included in the block of interest through a dither method using a dither threshold matrix It is even more desirable to further include a second block based binarization step.

Alternatively, in the block-based binarization step, if the maximum system is smaller than the threshold even as a result of the determination of the region, the pixels included in the block of interest are relatively smaller than the first threshold. The method may further include a second block-based binarization step of binarizing through the threshold.

On the other hand, the block superposition-based image binarization apparatus of the present invention, in the image binarization apparatus for performing binarization by reflecting the distribution characteristics of the block pixel group consisting of pixels located in a block of a predetermined size,

A block overlapping division unit for dividing the binarization target image so as to overlap at least one of a horizontal direction and a vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selector for sequentially selecting a block of interest to be binarized at this time in the block group;

A gray scale information calculating unit for comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

A threshold calculation unit for calculating a maximum threshold value that is even the gradation value between the maximum gradation value and the minimum gradation value, and a first threshold obtained by averaging the maximum gradation value and the minimum gradation value, respectively;

A comparison unit which receives the maximum system even from the threshold calculation unit and compares a threshold even the maximum system with a predetermined system and outputs a comparison result;

A first block-based binarization unit which receives the first threshold value from the threshold calculator and binarizes pixels included in the block of interest;

A dither binarizing unit for binarizing pixels included in the block of interest through a dither method using a dither threshold matrix; And

The outputting unit outputs the output of the first block-based binarization unit if the comparator even outputs the comparison result that the maximum system is even greater than the threshold, and otherwise outputs the output of the dithering binarization unit. It is characterized by being configured.

In still another aspect of the present invention, in the image binarization apparatus performing binarization by reflecting distribution characteristics of a block pixel group composed of pixels located in a block having a predetermined size,

A block overlapping division unit for dividing the binarization target image so as to overlap at least one of a horizontal direction and a vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selector for sequentially selecting a block of interest to be binarized at this time in the block group;

A gray scale information calculating unit for comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

Calculating a first threshold value obtained by averaging the maximum gray value and the minimum gray value, and a second threshold value obtained by multiplying the maximum gray value by a proportional constant, even the maximum gray level that is even between the maximum gray value and the minimum gray value; A threshold calculator;

A comparison unit which receives the maximum system even from the threshold calculation unit and compares a threshold even the maximum system with a predetermined system and outputs a comparison result;

A first block-based binarization unit which receives the first threshold value from the threshold calculator and binarizes pixels included in the block of interest;

A second block-based binarization unit which receives the second threshold value from the threshold calculator and binarizes pixels included in the block of interest; And

The comparator outputs the output of the first block-based binarization unit if the comparator outputs a comparison result that even the maximum system is even greater than the threshold, otherwise outputs the output of the second block-based binarization unit. A block overlap based image binarization apparatus including an output multiplexer.

Hereinafter, a preferred embodiment of a block overlap based image binarization method according to the present invention will be described with reference to FIG. 5.

5 shows a flowchart of a preferred embodiment of a block overlap based image binarization method according to the present invention.

A preferred embodiment of the block overlap-based image binarization method according to the present invention is an image binarization method in which binarization is performed by reflecting distribution characteristics of a block pixel group composed of pixels located in a block having a predetermined size. In

A block overlap division step (S10) of dividing the binarization target image so as to overlap each other in the horizontal direction and the vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selection step (S20) of sequentially selecting a block of interest to be binarized at this time in the block group;

A gray scale information calculating step (S30) of comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

An area determining step (S40) of calculating an even maximum system which is even a system between the maximum gray value and the minimum gray value to determine an area to which the block of interest belongs according to whether even the maximum system is greater than a preset threshold; And

As a result of the determination in the area determining step S40, when even the maximum system is even greater than the threshold value, the pixels included in the block of interest may have a first threshold value obtained by averaging the maximum gray value and the minimum gray value. Binarizing first block-based binarization step S50; And

As a result of the determination in the region determining step S40, if even the maximum system is smaller than the threshold, the second block performs binarization of pixels included in the block of interest through a dither method using a dither threshold matrix. It is configured to perform the base binarization step (S60).

Another embodiment of the block superposition-based image binarization method according to the present invention relates to an image binarization method in which binarization is performed by reflecting distribution characteristics of a block pixel group composed of pixels located in blocks having a predetermined size. In

A block overlap division step (S10) of dividing the binarization target image so as to overlap each other in the horizontal direction and the vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selection step (S20) of sequentially selecting a block of interest to be binarized at this time in the block group;

A gray scale information calculating step (S30) of comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

An area determining step (S40) of calculating an even maximum system which is even a system between the maximum gray value and the minimum gray value to determine an area to which the block of interest belongs according to whether even the maximum system is greater than a preset threshold; And

As a result of the determination in the area determining step S40, when even the maximum system is even greater than the threshold value, the pixels included in the block of interest may have a first threshold value obtained by averaging the maximum gray value and the minimum gray value. Binarizing first block-based binarization step S50; And

As a result of the determination in the area determining step S40, when even the maximum system is even less than the threshold, binarization is performed through a second threshold value of pixels included in the block of interest by multiplying the maximum gray value by a proportional constant. It is configured to perform a second block-based binarization step (S60) to perform.

A detailed description will now be made of a process of performing a preferred embodiment of the block overlap-based image binarization method according to the present invention configured as described above with reference to FIG. 5.

First of all, prior to explaining the process of carrying out the preferred embodiment according to the method of the present invention, the shuttle scanning based image input system, which is one of the image systems to which the present invention can be applied to facilitate understanding of the present invention, will be briefly described. Let's look at it.

6 is a block diagram illustrating a shuttle scanner module that is a key component of a shuttle scanning-based image input system. Referring to the configuration and operation of the shuttle scanner module in more detail, the shuttle scanner module 21 includes a CCD board, a lens, and a light source. It is installed on the same horizontal moving shaft 23 as the print head 22, and the two carriage return motor (hereinafter referred to as 'CR motor') 24 and two through the moving belt 26 The module can be driven.

FIG. 7 is an exemplary view showing a shuttle block for scanning an A4 sized document with a shuttle scanner module, and FIG. 8 is an exemplary view for explaining how one shuttle block is scanned.

The CCD (Charge Coupled Device) used in the shuttle scanner module is a typical example of using a 128-dot CCD.

In the case of a shuttle scanning scanner employing a 128-dot CCD, when scanning at 300 dpi when the A4 (210 mm x 297 mm) format document is used, the number of dots is 2551 dots x 3507 dots. Techniques for scanning by dividing into shuttle blocks are known.

In detail, when an A4 size original is scanned with a shuttle scanner module 21 using a 128-dot CCD (2551 × 3507; based on 300 DPI), the original is divided into 27 shuttle blocks, and the shuttle scanner module ( 21) is transported horizontally by the CR motor 24 while reading one shuttle block vertically one line, and thus reads all from the vertical one line to the 2551 line of the shuttle block. When the scanning of one shuttle block is completed, the shuttle scanner module 21 may vary depending on the method, but is normally returned to the first line position by the CR motor 24 and the line feed motor 25 The originals are transferred by), and the next shuttle block is located in the shuttle scanner module 21. When the above steps are repeated to scan all shuttle blocks from [BLOCK1] to [BLOCK 27], all the image data for one page of an A4 size document is read.

By doing so, it is possible to complete the entire data reading on the original to be scanned while overcoming the problem of low resolution CCD by using limited resources (ie, 128 dot CCD).

A preferred embodiment of the block overlap based image binarization method according to the present invention can be easily implemented in software by a shuttle based image input system or a central processing unit (not shown) of a conventional image system as described above, In addition, it may be implemented and implemented by providing separate hardware.

For convenience of description, a preferred embodiment of the block overlap based image binarization method according to the present invention will be described based on the former case. Therefore, in the following description, it is well known that the hardware subject who implements the method of the present invention becomes the central processing unit of the image system.

On the other hand, the latter case will be described in detail later through the block overlap-based image binarization apparatus according to the present invention shown in the accompanying drawings.

First, in the block overlap division step S10, a binarization target image to be binarized is received from an external memory or other reading means, and as shown in FIG. 10, the binarization target image is horizontally and vertically arranged in blocks of a predetermined size. A block group consisting of a plurality of blocks is generated by dividing so as to overlap in a direction, which will be described in more detail.

The gradation of the binarization target image used in the present invention is 256 gradations representing a single pixel as 8 bits as in a normal case, and the bits allocated according to the respective application fields can be added or subtracted. While it is possible to express pixels in detail, it is well known that a relatively large amount of resources should be allocated, and the amount of computation must be exponentially increased during signal processing.

Meanwhile, the block size is 4 pixels It is preferable that it is four pixels, and eight pixels 8 pixels are also available. 16 pixels, such as macro blocks, depending on application and processing speed 16 pixels may be used, but generally 4 pixels It is desirable to set the size close to four pixels.

In terms of the accuracy of the area setting, the larger the size of the block is, the more sufficient the state of the distribution of adjacent pixels can be considered, so that the accurate area determination can be made. As binarization, binarization performance may be degraded. On the other hand, if the size of the block is set relatively small, it is impossible to sufficiently consider the distribution state of adjacent pixels, which may cause an inaccurate region determination, and an inaccurate region determination may cause an object to generate an inappropriate threshold or dither threshold matrix. As it can be binarized through, it is obvious that the block size should be appropriately set in consideration of application fields, processing speeds, characteristics of binarized images, and the like.

In particular, when the size of the block is set too large, care must be taken as the local distribution characteristic of the pixel may be removed and rather only the global characteristic may be reflected.

Meanwhile, in the preferred embodiment of the present invention, a block having a square shape may be preferable as the shape of the block. However, a rectangular block having a different ratio in the vertical direction and the horizontal direction may be used, and thus, a two-dimensional block shape. In addition, it can be used in the form of a one-dimensional block.

9 is a four pixel according to the prior art. It is an exemplary figure which shows the example of block division of 4 pixels.

Figure 10 (a) is a four pixel according to a preferred embodiment of the present invention It is an exemplary figure which shows the example of block division of four pixels, and FIG. 10 (b) is an enlarged illustration which expands and shows a part of FIG.

The block division according to the prior art divides the blocks so that they do not overlap as shown in FIG. 9, while the block division according to the preferred embodiment of the present invention shows that the blocks have a horizontal direction and a vertical direction as shown in FIG. 10A. Partitioning to overlap is a key feature.

10A is two pixels It can be observed that it is divided into two pixel blocks, but this can be observed by dividing the overlapping half of the block size in the horizontal and vertical directions as shown in FIG. 10B. To generate a block group composed of a plurality of overlapping blocks.

Subsequently, in the block selection step S20, the block of interest to be binarized at this point in the block group is sequentially selected according to a raster scanning order.

For example, in the case of Fig. 10B, the selection order of the blocks is selected in the order of block A1, block B1, block C1, and then in the next block line in the same manner as block A2, block B2, block C2.

Subsequently, when the maximum gray value and the minimum gray value are compared and calculated in the block pixel group composed of the pixels located in the block of interest in the gray scale information calculating step S30, the maximum gray value and the minimum in the area determining step S40. Even the maximum system, which is a system even between gray levels, is calculated to determine the area to which the block of interest belongs, depending on whether even the maximum system is larger than the threshold.

Here, even the maximum system is a cost function for determining whether the area to which the block of interest belongs is a character area or a picture area. If even the maximum system is even larger than the threshold, it means that the block of interest belongs to the character area, otherwise it means that the block of interest belongs to the picture area.

Finally, when the determination result of the area determining step S40 indicates that even the maximum system is larger than the threshold value, in the first block-based binarization step S50, the pixels included in the block of interest are compared with the maximum gray value. The minimum gray value is binarized through an average of a first threshold value.

On the other hand, when the determination result of the area determining step S40 indicates that even the maximum system is smaller than the threshold value, in the second block-based binarization step S60, the pixels included in the block of interest are divided into a dither threshold matrix. By performing binarization through the dither method used, the process of performing the preferred embodiment of the block overlap-based image binarization method according to the present invention is completed.

For example, the pixels included in each of the blocks of interest for blocks A1, B1, and C1 of FIG. 10B each include pixels belonging to the inner block A, the inner block B, and the inner block C, respectively.

In order to understand the preferred embodiment according to the method of the present invention, the first block-based binarization step S50 and the second block-based binarization step S60 will be described step by step. First, the first block-based binarization step S50 will be described.

As a result of the determination in the area determining step S40, when even the maximum system is even larger than the preset value, in step S51 of the first block-based binarizing step S50, as shown in Equation 1, the maximum gray value and the minimum The first threshold value obtained by averaging the gray scale values is obtained.

Accordingly, in step S52, it is determined whether the pixel of interest is larger than the first threshold, and if the pixel of interest is larger than the first threshold, in step S53, the pixel of interest is binarized to white (that is, binary 1); The pixel of interest is binarized to black (ie, binary 0) through S54.

Subsequently, in step S55, it is determined whether binarization of the last pixel is completed among all pixels included in the block of interest, and if binarization of the last pixel is not completed, the process returns to step S52 to return to step S52. The binarization is repeated for all pixels included in the.

On the other hand, if it is determined in step S55 that the binarization of the last pixel is completed, it is determined in step S70 whether binarization of all blocks included in the block group is completed, and if binarization of all blocks is completed, all the processes are performed. If not, the process returns to the block selection step S20 to continuously repeat the binarization of the next block in the raster scanning order.

Next, the second block-based binarization step S60 will be described step by step.

As a result of the determination in the area determining step S40, if even the maximum system is smaller than the preset system, in step S61 of the second block-based binarization step S60, a dither threshold matrix for the block of interest is selected, Accordingly, in step S62, it is determined whether the pixels belonging to the block of interest are larger than the elements of the corresponding dither threshold matrix, and if the pixel of interest is larger than the elements of the corresponding dither threshold matrix, the pixel of the block of interest becomes white in step S63. (I.e., binary 1), otherwise, the corresponding pixel belonging to the block of interest is binarized to black (i.e. binary 0) through step S64.

Thereafter, in step S65, it is determined whether binarization of the last pixel is completed among all pixels included in the block of interest, and if binarization of the last pixel is not completed, the process returns to step S62 to return to step S62. The binarization is repeated for all pixels included in the.

On the other hand, if it is determined in step S65 that the binarization of the last pixel is completed, it is determined in step S70 whether binarization of all blocks included in the block group is completed, and if binarization of all blocks is completed, all the processes are performed. If not, the process returns to the block selection step S20 to continuously repeat the binarization of the next block in the raster scanning order.

On the other hand, to help understand the present invention, a dither method, which is a kind of halftone processing technique used in the present invention, will be described.

Halftone processing is an image processing technique that makes a device appear to be close to the original image while reducing the quantized grayscale per pixel in a device having limited grayscale value reproduction characteristics. This technique can be used for various display devices such as printing presses, ink-jet printers, laser printers and liquid crystal displays (LCDs), and cathode ray tubes (CRTs). Widely used in image output apparatuses and the like.

It's been a long time since we've been trying to develop an appropriate conversion technique that can reduce the conversion time, where halftone images look as visually indistinguishable from the original continuous tone image as possible, while at the same time significantly affecting print speed. This is an issue to be solved across.

The most common halftone processing techniques are the well-known ordered dither method and the error diffusion method.

This is an alignment dither, written by Robert Ulichney and filed by Digital Halftoning and Chan, published by the Massachusetts Institute Technology press, issued US Patent No. 5,140,432. And Method and System for Reproducing Monochrome and Color Images Using Ordered Dither and Error Diffusion.

First, in contrast to the random dither method generating random thresholds, the aligned dither method has a simple thresholding process, that is, a dither threshold matrix, which is a preset threshold array. , A simple threshold processing), wherein the dither threshold matrix includes a plurality of constituent elements to occupy a physical space in the pixel space.

The dither threshold matrix is mapped onto a continuous-tone image, and each pixel of the continuous grayscale image is compared pixel by pixel with each element of the corresponding dither threshold matrix. If the pixel value of the continuous grayscale image is larger than the element value of the corresponding dither threshold matrix, the dot at the corresponding position of the halftone image is binarized to white, otherwise the dot is made black. By binarizing, it is desired to generate a natural halftone image as a whole.

At this time, if the size of the space occupied by the dither threshold matrix is smaller than the size of the continuous grayscale image, the dither threshold matrix is replicated to fit the size of the entire continuous grayscale image, thereby performing halftone processing without being limited to the image size. can do.

Meanwhile, the error diffusion method, which is another method, is continuously processed through a threshold process for binarizing grayscale values larger than the threshold value to white based on a fixed threshold value, and binarizing the grayscale values to black. The gradation image is converted into a halftone image and processed.

In this case, a lack of accuracy may occur due to the use of a fixed threshold value. Accordingly, in the error diffusion method, an error between a pixel value of a continuous grayscale image and a thresholded binary pixel value is obtained. This problem is overcome by compensating for error by diffusing adjacent pixels.

In other words, the above-mentioned problem is solved by propagating an error between the pixel value of the continuous grayscale image being processed at this point and the thresholded binary pixel value to adjacent pixels participating in the weight determination. The same approach is even better when it comes to expressing continuous tones over a small area surrounded by a few few dots.

On the other hand, in 1991, Sullivan, L. Ray and R. Miller, at least in the journal System, Man, and Cybernetics of the Institute of Electrical and Electronics Engineers (IEEE), Based on the contribution of the Design of Minimum Visual Modulation Halftone Patterns, research and interest in binary database-based halftone processing using binary databases have been increasing.

In order to meet the technical trend, the applicant of the present invention has applied for a binary database based halftone processing method using a binary database through Korean Patent Application No. 97-26633, a halftone processing method of an image system. have.

Korean Patent Application No. 97-26633, The halftone processing method of an image system classifies a grayscale group up and down based on one of the grayscale values of the entire grayscale range, and then refers to a binary reference pattern for the grayscale values. A database generation step of generating a binary database composed of respective dot profiles generated by inverting the dots so as to include the number of white dots proportional to each gray value; And middle color of the original image by indexing binary data for cyan, magenta, and yellow from the binary database to each color component gradation value and each pixel position by a cyclic indexing method using a predetermined pixel interval. And a color database indexing step of generating a crude image.

The indexing of the database may include: a dot profile indexing step of indexing the dot profile with the grayscale value of the original image from the binary database;

A binary data indexing step of indexing binary data located at pixel positions of an original image from the dot profile;

It is preferable that a single color halftone image generation step of generating a single color halftone image by positioning the binary data at the same pixel position as the original image.

In the second block-based binarization step S60 of the present invention, it is well known that Korean Patent Application No. 97-26633, a halftone processing method of an image system, can be used as an embodiment.

FIG. 11 is an exemplary view for explaining an application example of the preferred embodiment of the block superposition-based image binarization method according to the present invention, to which the preferred embodiment of the apparatus of the present invention is applied to the example shown in FIG. 4.

As in Figure 4, the size of the block is 4 At 4, when the number of levels of the pixel grayscale value is set to 16 grayscales (i.e., 4 bits) and even the threshold is set to 7, as shown in FIG. 11 (a), block A is a photo area P and block B is As the character area C, the block C is classified into the picture area P. FIG.

As can be seen in (b) of FIG. 11, block A has a maximum gray value of 8, a minimum gray value of 1, and even a maximum gray level of 7, so that even the maximum gray level (7) is not larger than the threshold (7). The pixels are classified into the photo area according to the pixel region, and the pixels located inside the block A (that is, the inner block A) are binarized through a dither threshold matrix. For example, the pixels 1, 1, 1, 1 belonging to the inner block A are binarized to 0, 0, 0, 0 by the dither threshold matrix shown in FIG. 4B.

In addition, block B has a maximum gray value of 16, a minimum gray value of 8, a first threshold of 12, and even a maximum gray of 8, so that even the maximum gray level 12 is larger than the threshold 7, the character area. And the pixels located inside the block B (ie, the inner block B) are binarized through the first threshold. For example, the pixels 8, 14, 8, 14 belonging to the inner block B are binarized to 0, 1, 0, 1 by the first threshold 12, respectively.

Block C is classified into a photographic area as the maximum gray value is 16, the minimum gray value is 16, and even the maximum gray level is 0, even the maximum gray level (0) is smaller than the threshold (7). Pixels located inside C (ie, inner block C) are binarized by a dither threshold matrix. For example, the pixels 16, 16, 16, and 16 belonging to the inner block B are binarized to 1, 1, 1, 1 by the dither threshold matrix shown in FIG. 4B.

Figure 11 (c) shows the results binarized through such a process, through which the preferred embodiment according to the method of the present invention can be obtained that can be obtained a desirable binarization result is excluded erroneous boundary or irregularities phenomenon Can be.

Hereinafter, another embodiment of a block overlap based image binarization method according to the present invention will be described with reference to FIG. 12.

12 is a flowchart of another embodiment of a block overlap based image binarization method according to the present invention, and for convenience of description, the same as the preferred embodiment of the block overlap based image binarization method according to the present invention as described with reference to FIG. 5. In the performing step, the same identification number is assigned.

Another embodiment of the block overlap-based image binarization method according to the present invention relates to an image binarization method in which binarization is performed by reflecting distribution characteristics of a block pixel group composed of pixels located in blocks having a predetermined size. In

A block overlap division step (S10) of dividing the binarization target image so as to overlap each other in the horizontal direction and the vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selection step (S20) of sequentially selecting a block of interest to be binarized at this time in the block group;

A gray scale information calculating step (S30) of comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

An area determining step (S40) of calculating an even maximum system which is even a system between the maximum gray value and the minimum gray value to determine an area to which the block of interest belongs according to whether even the maximum system is greater than a preset threshold; And

As a result of the determination in the area determining step S40, when even the maximum system is even greater than the threshold value, the pixels included in the block of interest may have a first threshold value obtained by averaging the maximum gray value and the minimum gray value. A first block based binarization step S50 for binarizing; and

As a result of the determination in the area determining step S40, when even the maximum system is even less than the threshold, binarization is performed through a second threshold value of pixels included in the block of interest by multiplying the maximum gray value by a proportional constant. It is configured to perform a second block-based binarization step (S70) to perform.

Referring to FIG. 12, a process of performing another embodiment of the block overlap-based image binarization method according to the present invention configured as described above will be described in detail.

The second block-based binarization step (S60) of the preferred embodiment of the block overlap-based image binarization method according to the present invention is replaced by the second block-based binarization step (S70) of another embodiment of the block overlap-based image binarization method according to the present invention. Except for the rest.

In other words, another embodiment of the block overlap based image binarization method according to the present invention includes a block overlap division step S10, a block selection step S20, a gray scale information calculation step S30, an area determination step S40, and a first step. The process of performing the block-based binarization step S50 is the same as the preferred embodiment of the block superposition-based image binarization method according to the present invention. Accordingly, the present application excludes repeated descriptions of overlapping portions within the allowable range of those skilled in the art.

Another embodiment of the block overlap based image binarization method according to the present invention sequentially performs a block overlap division step S10, a block selection step S20, a gray scale information calculation step S30, and an area determination step S40.

Subsequently, when the determination result of the area determining step S40 indicates that even the maximum system is even larger than the threshold value, in the first block-based binarizing step S50, the pixels included in the block of interest include the maximum gray value and the maximum gray value. Binary the minimum gray value through a first threshold value.

On the other hand, when the determination result of the area determining step S40 indicates that even the maximum system is even smaller than the threshold value, in the second block-based binarizing step S70, the pixels included in the block of interest are the maximum gray value. The binarization-based image binarization method according to the present invention is completed by performing binarization through a second threshold multiplied by a proportional constant.

In order to understand another embodiment according to the method of the present invention, the first block-based binarization step S50 and the second block-based binarization step S70 will be described in detail step by step. First, the first block-based binarization step S50 will be described.

As a result of the determination in the area determining step S40, when even the maximum system is even larger than the preset value, in step S51 of the first block-based binarizing step S50, as shown in Equation 1, the maximum gray value and the minimum The first threshold value obtained by averaging the gray scale values is obtained.

Accordingly, in step S52, it is determined whether the pixel of interest is larger than the first threshold, and if the pixel of interest is larger than the first threshold, in step S53, the pixel of interest is binarized to white (that is, binary 1); The pixel of interest is binarized to black (ie, binary 0) through S54.

Subsequently, in step S55, it is determined whether binarization of the last pixel is completed among all pixels included in the block of interest, and if binarization of the last pixel is not completed, the process returns to step S52 to return to step S52. The binarization is repeated for all pixels included in the.

On the other hand, if it is determined in step S55 that the binarization of the last pixel is completed, it is determined through step S80 whether binarization of all blocks included in the block group is completed, and if binarization of all blocks is completed, all the processes are performed. If not, the process returns to the block selection step S20 to continuously repeat the binarization of the next block in the raster scanning order.

Next, the second block-based binarization step S70 will be described step by step.

As a result of the determination in the area determining step S40, when even the maximum system is even smaller than the preset system, in step S71 of the second block-based binarizing step S70, the maximum proportional value is proportional to the maximum gray value as shown in Equation 2. a constant( Find the second threshold multiplied by). Where the proportional constant ( ) Is 0.5, and can be easily changed according to the application field and application environment.

Thereafter, in step S72, it is determined whether the pixel of interest is greater than the second threshold, and if the pixel of interest is greater than the second threshold, the pixel of interest is binarized to white (ie, binary 1) (S73), otherwise, the pixel of interest is The pixel is binarized to black (that is, binary 0) (S74).

Thereafter, in step S75, it is determined whether binarization of the last pixel is completed among all pixels included in the block of interest, and if binarization of the last pixel is not completed, the process returns to step S72 to return to step S72. The binarization is repeated for all pixels included in the.

On the other hand, if it is determined in step S75 that the binarization of the last pixel is completed, it is determined through step S80 whether binarization of all blocks included in the block group is completed, and if binarization of all blocks is completed, all the processes are performed. If not, the process returns to the block selection step S20 to continuously repeat the binarization of the next block in the raster scanning order.

In a preferred embodiment according to the method of the present invention, the second threshold value is calculated through Equation 2, but the second threshold value used in the present invention generally has a global threshold characteristic and thus is half of the maximum gray value. In addition, the total threshold value may be obtained by summing all gray values of the binarization target image, and then multiplying the total average values by a proportional constant to determine a global threshold value. It can also be set by value.

In this case, the overall average value is for reflecting the statistical characteristics of the image to the threshold value, and the proportional constant is a parameter that can add or subtract a ratio of black or white in the photo area. In addition to using only the overall mean to reflect the statistical properties of the image in the determination of the global threshold, the standard deviation and variance of the data may be compared to the overall mean to reflect more accurate statistical properties. It may be considered together.

Hereinafter, a preferred embodiment of the block overlap-based image binarization apparatus according to the present invention will be described with reference to FIG.

13 is a block diagram showing a preferred embodiment of a block overlap based image binarization apparatus according to the present invention.

As shown in FIG. 13, a block overlap-based image binarization apparatus according to an exemplary embodiment of the present invention provides an image binarization method in which binarization is performed by reflecting distribution characteristics of a block pixel group including pixels located in blocks having a predetermined size. In

A block overlapping division unit 101 for dividing the binarization target image so as to overlap each other in at least one of a horizontal direction and a vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selector (102) for sequentially selecting a block of interest to be binarized at this point in the block group;

A gray scale information calculating unit (110) for comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

A threshold calculation unit (120) for calculating a maximum threshold that is even between the maximum gray value and the minimum gray value, and a first threshold obtained by averaging the maximum gray value and the minimum gray value;

A comparison unit 130 which receives the maximum system even from the threshold calculation unit 120 and compares a threshold even the maximum system with a predetermined system and outputs a comparison result;

A first block-based binarizer (140) for receiving the first threshold value from the threshold calculator (120) and binarizing pixels included in the block of interest;

A dither binarizer 150 for binarizing pixels included in the block of interest through a dither method using a dither threshold matrix; And

When the comparison unit 130 outputs a comparison result that even the maximum system is even larger than the threshold, the output of the first block-based binarization unit 140 is output. Otherwise, the dithering binarization unit ( And an output multiplexer 160 for outputting the output of 150.

The operation of the preferred embodiment of the block overlap-based image binarization apparatus according to the present invention configured as described above will be described in detail with reference to FIG.

In the following description, there may be overlapping contents described while explaining embodiments according to the method of the present invention, repeated to the extent that it is acceptable to those having ordinary children in the art to avoid this The description will be weak.

First, the block overlap division unit 101 divides the binarization target image in units of blocks having a predetermined size so as to overlap each other in at least one of the horizontal direction and the vertical direction to generate a block group composed of a plurality of blocks.

The gradation of the binarization target image used in the present invention is 256 gradations representing a single pixel as 8 bits as in the normal case, as in the embodiments according to the method of the present invention, and bits allocated according to respective application fields. It is possible to add and subtract, and the more bits that are allocated, the more precise and detailed the pixels can be represented, while the more resources must be allocated and the amount of computation must be exponentially increased during signal processing. It is true.

Meanwhile, the block size is 4 pixels It is preferable that it is four pixels, and eight pixels 8 pixels are also available. 16 pixels, such as macro blocks, depending on application and processing speed. 16 pixels may be used, but generally 4 pixels It is desirable to set the size close to four pixels.

In terms of the accuracy of the area setting, the larger the size of the block is, the more sufficient the state of the distribution of adjacent pixels can be considered, so that the accurate area determination can be made. As it is binarized, binarization performance may be degraded. On the other hand, if the size of the block is set relatively small, it is impossible to sufficiently consider the distribution state of adjacent pixels, which may cause an inaccurate region determination, and an inaccurate region determination may cause an object to generate an inappropriate threshold or dither threshold matrix. As it can be binarized through, it is obvious that the block size should be appropriately set in consideration of application fields, processing speeds, characteristics of binarized images, and the like.

In particular, when the size of the block is set too large, care must be taken as the local distribution characteristic of the pixel may be removed and rather only the global characteristic may be reflected.

Meanwhile, in the present invention, a block having a square shape is preferable in the present invention, but rectangular blocks having different ratios in the vertical direction and the horizontal direction may be used, and thus, in addition to the two-dimensional block shape, It may also be used in the form of blocks.

As described above, when a block group is generated by the block overlap divider 101, the block selector 102 selects a block of interest to be binarized at this point in the block group of the block overlap divider 101 in the raster scanning order. Select sequentially.

In this case, the raster scanning order is not necessarily required, but according to the present invention, since the blocks are overlapped and divided, it is impossible to select a block of interest based on a scanning method such as zigzag scanning or fractal scanning. Do.

Thereafter, the gray scale information calculating unit 110 compares and calculates a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest, which will be described in more detail as follows.

When the gray comparison unit 111 receives the block pixel group from the block selection unit 102 and outputs the maximum gray value and the minimum gray value through gray level comparison, the maximum gray value storage unit 112 in the gray level comparison unit 111. The maximum gray level value is received and stored, and the minimum gray value storage unit 113 receives and stores the minimum gray value from the gray level comparison unit 111.

Here, the maximum gray value storage unit 112 and the minimum gray value storage unit 113 continuously store the previous maximum gray value and the minimum gray value until the new maximum gray value and the minimum gray value are input, respectively. When the new maximum and minimum gray values are input, respectively, the old maximum and minimum gray values are replaced with the new maximum and minimum gray values.

The threshold calculator 120 receives the maximum gray value and the minimum gray value from the maximum gray value storage 112 and the minimum gray value storage 113, respectively, and subtracts the maximum gray value through the subtraction unit 121. Compute even the maximum gradation even between the minimum gradation value, and calculates the first threshold value by averaging the maximum gradation value and the minimum gradation value through the adder 122 and the divider 123 as shown in equation (1). do.

On the other hand, even the comparison unit 130 receives even the maximum system from the subtraction unit 121 of the threshold calculation unit 120 and compares even the maximum system and the predetermined system threshold value and outputs a comparison result. For example, as an example of the output format of the comparator 130, even if the maximum system is larger than the threshold even, the system outputs a logic high state. Otherwise, the region of interest pixel is generated by generating a logic low state output. You can identify whether each is a text area or a picture area.

The first block-based binarization unit 140 receives a block of interest from the block selector 102, receives the first threshold value from the divider 123 of the threshold calculator 120, and then internalizes the block of interest. Binarize the pixels included in the. That is, it is determined whether each pixel included in the block of interest is greater than the first threshold, and if the gray value of each pixel is greater than the first threshold, the pixel is binarized to white (ie, binary 1). Otherwise, the pixel is binarized to black (that is, binary 0) and output.

The dither digitizer 150 receives a block of interest from the block selector 102, receives a dither threshold matrix from the dither threshold matrix storage unit 151 for storing a dither threshold matrix, and is included in the block of interest. The pixels are binarized and output through a dither method using a dither threshold matrix.

Dithering threshold 150 of the present invention is the same as in the second block-based binarization step (S60) of the preferred embodiment according to the method of the present invention, Korean Patent Application No. 97-26633 filed by the applicant of the present invention The halftone processing method of the call and image system can be used, and other well-known general halftone processing techniques or dither methods can be easily adopted and applied.

Finally, the output multiplexer 160 outputs a comparison result even when the comparator 130 outputs the comparison result that even the maximum system is larger than the threshold even when the maximum system is equal to the system (that is, even when the comparator 130 is in a logic high state). Outputting the output), outputting the output of the first block-based binarization unit 140 otherwise (that is, even if the comparator 130 outputs a logic low state output), the dithering binarization unit 150 By outputting the output of), block overlap based image binarization according to the present invention is completed.

Hereinafter, another embodiment of a block overlap-based image binarization apparatus according to the present invention will be described with reference to FIG. 14.

14 is a block diagram illustrating another embodiment of a block overlap-based image binarization apparatus according to the present invention.

According to another embodiment of the block superposition-based image binarization apparatus according to the present invention, as shown in FIG. 14, the binarization-based image binarization method reflects a distribution characteristic of a block pixel group composed of pixels located in a block having a predetermined size. In

A block overlapping division unit 101 for dividing the binarization target image so as to overlap each other in at least one of a horizontal direction and a vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks;

A block selector (102) for sequentially selecting a block of interest to be binarized at this point in the block group;

A gray scale information calculating unit (110) for comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest;

Calculating a first threshold value obtained by averaging the maximum gray value and the minimum gray value, and a second threshold value obtained by multiplying the maximum gray value by a proportional constant, even the maximum gray level that is even between the maximum gray value and the minimum gray value; A threshold calculator 220;

A comparison unit 130 which receives the maximum system even from the threshold calculator 220 and outputs a comparison result by comparing a threshold even the maximum system with a predetermined system;

A first block-based binarizer 140 for inputting the first threshold value from the threshold calculator 220 and binarizing pixels included in the block of interest;

A second block-based binarizer (250) for receiving the second threshold value from the threshold calculator (220) and binarizing pixels included in the block of interest; And

Even if the comparison unit 130 outputs a comparison result that the maximum system is even greater than the threshold value, and outputs the output of the first block-based binarization unit 140, otherwise, the second block-based And an output multiplexer 260 for outputting the output of the binarization unit 250.

The operation of another embodiment of the block overlap-based image binarization apparatus according to the present invention configured as described above will be described in detail with reference to FIG.

In the following description, there may be overlapping contents as described while explaining preferred embodiments according to the apparatus of the present invention is repeated in a range that can be tolerated by those of ordinary skill in the art to avoid this Repeated explanations will be weak.

First, the block overlap division unit 101 divides the binarization target image in units of blocks having a predetermined size so as to overlap each other in at least one of the horizontal direction and the vertical direction to generate a block group composed of a plurality of blocks.

The gradation of the binarization target image used in the present invention is 256 gradations representing a single pixel as 8 bits as in the normal case, as in the embodiments according to the method of the present invention, and bits allocated according to respective application fields. It is possible to add and subtract, and the more bits that are allocated, the more precise and detailed the pixels can be represented, while the more resources must be allocated and the amount of computation must be exponentially increased during signal processing. It is true.

Meanwhile, the block size is 4 pixels It is preferable that it is four pixels, and eight pixels 8 pixels are also available. 16 pixels, such as macro blocks, depending on application and processing speed 16 pixels may be used, but generally 4 pixels It is desirable to set the size close to four pixels.

Meanwhile, in the present invention, a block having a square shape is preferable in the present invention, but rectangular blocks having different ratios in the vertical direction and the horizontal direction may be used, and thus, in addition to the two-dimensional block shape, It may also be used in the form of blocks.

As described above, when a block group is generated by the block overlap divider 101, the block selector 102 selects a block of interest to be binarized at this point in the block group of the block overlap divider 101 in the raster scanning order. Select sequentially.

In this case, the raster scanning order is not necessarily required, but according to the present invention, since the blocks are overlapped and divided, it is impossible to select a block of interest based on a scanning method such as zigzag scanning or fractal scanning. Do.

Thereafter, the gray scale information calculating unit 110 compares and calculates a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest, which will be described in more detail as follows.

When the gray comparison unit 111 receives the block pixel group from the block selection unit 102 and outputs the maximum gray value and the minimum gray value through gray level comparison, the maximum gray value storage unit 112 in the gray level comparison unit 111. The maximum gray level value is received and stored, and the minimum gray value storage unit 113 receives and stores the minimum gray value from the gray level comparison unit 111.

Here, the maximum gray value storage unit 112 and the minimum gray value storage unit 113 continuously store the previous maximum gray value and the minimum gray value until the new maximum gray value and the minimum gray value are input, respectively. When the new maximum and minimum gray values are input, respectively, the old maximum and minimum gray values are replaced with the new maximum and minimum gray values.

The threshold calculator 220 receives the maximum gray value and the minimum gray value from the maximum gray value storing unit 112 and the minimum gray value storing unit 113, respectively, and subtracts the maximum gray value from the maximum gray value storing unit 112. Compute even the maximum gradation even between the minimum gradation value, and calculates the first threshold value by averaging the maximum gradation value and the minimum gradation value through the adder 122 and the divider 123 as shown in equation (1). The multiplier 124 calculates a second threshold value obtained by multiplying the maximum gray value by a proportional constant as shown in Equation 2 below.

In this case, although the threshold calculation unit 220 according to another embodiment of the present invention calculates the second threshold value through Equation 2, the second threshold value used in the present invention generally has a global threshold characteristic. In addition to setting it as half of the maximum gray scale value, the total average value obtained by summing all gray scale values of the binarization target image may be obtained, and then multiplying the total average value by a proportional constant to determine a global threshold value. The pixel values of the pixel group may be determined as an average value.

In this case, the overall average value is for reflecting the statistical characteristics of the image to the threshold value, and the proportional constant is a parameter that can add or subtract a ratio of black or white in the photo area. In addition to using only the overall mean to reflect the statistical properties of the image in the determination of the global threshold, the standard deviation and variance of the data may be compared to the overall mean to reflect more accurate statistical properties. It may be considered together.

Accordingly, the threshold calculator 220 outputs one of a threshold value proportional to an average value of pixel values of the adjacent pixel group and a threshold value proportional to an average value of all pixel values of the binarization target image. It may also include a second threshold (not shown).

On the other hand, even the comparison unit 130 receives even the maximum system from the subtraction unit 121 of the threshold calculation unit 220 and outputs a comparison result by comparing even the maximum system and the predetermined system threshold. For example, as an example of the output format of the comparator 130, even if the maximum system is larger than the threshold even, the system outputs a logic high state. Otherwise, the region of interest pixel is generated by generating a logic low state output. You can identify whether each is a text area or a picture area.

The first block-based binarization unit 140 receives a block of interest from the block selector 102, receives the first threshold value from the divider 123 of the threshold calculator 220, and then internalizes the block of interest. Binarize the pixels included in the. That is, it is determined whether each pixel included in the block of interest is greater than the first threshold, and if the gray value of each pixel is greater than the first threshold, the pixel is binarized to white (ie, binary 1). Otherwise, the pixel is binarized to black (that is, binary 0) and output.

Meanwhile, the second block-based binarizer 250 receives the block of interest from the block selector 102 and receives the second threshold value from the divider 124 of the threshold calculator 220. Binarize the pixels included in the.

That is, it is determined whether each pixel included in the block of interest is greater than the second threshold, and if the gray value of each pixel is greater than the second threshold, the pixel is binarized to white (i.e., binary 1). Otherwise, the pixel is binarized to black (that is, binary 0) and output.

Finally, the output multiplexer 260 outputs a comparison result even when the comparator 130 compares the system even with the maximum system even greater than the threshold (that is, even the comparator 130 is in a logic high state). Outputting the output), outputs the output of the first block-based binarization unit 140 otherwise (i.e., even if the comparator 130 outputs a logic low state output), the second block-based binarization By outputting the output of the display unit 250, the block overlap based image binarization according to the present invention is completed.

By repeatedly applying the binarization process described above to all blocks belonging to the block group of the binarization target image, the binarization result appropriately reflecting the object of the present invention as described above through the apparatus according to the present invention or the method according to the present invention is obtained. You can get it.

In the apparatus and method of the present invention described above, the first threshold value is determined as a value obtained by averaging the maximum gray value and the minimum gray value of the block pixel group, but similarly to the second threshold, the maximum of the block pixel group A value obtained by multiplying the average value of the gradation value and the minimum gradation value by a proportional constant may be used, and the maximum gradation value and the minimum gradation value may be determined by using different weights.

There is no compulsory condition that you must use only the average of the two values. However, the second threshold value must be determined to a value that can properly reflect the specification of the maximum gray value and the minimum gray value of the block pixel group to obtain a desirable result.

Typically, the first threshold generally has a relatively large value compared to the second threshold, but of course this is also not mandatory.

In the preferred embodiment of the present invention, the overlapping point is determined to be half of the horizontal and vertical directions, but more portions may be overlapped and less portions may be overlapped. Most preferred. In this case, the pixels included in the target block may include 25% of pixels included in a square block close to the center of the target block among the pixels included in the target block.

In the present invention, the blocks are divided to overlap each other in the horizontal and vertical directions, but it is well known that the blocks may be divided to overlap each other in either the horizontal and vertical directions.

When the blocks are divided so as to overlap only one of the horizontal direction and the vertical direction, the binarization target pixels of each block of interest can be easily modified and set in the present invention.

In other words, when the blocks are overlapped only with respect to the horizontal direction, the binarized pixels of each block of interest are pixels located inside the block of interest and pixels located at a boundary between the vertical direction of the target block and the blocks only in the vertical direction. When the pixels are divided so as to overlap each other, the pixels to be binarized in each block of interest are pixels located inside the block of interest and pixels located in a horizontal boundary of the object block.

The terminologies used herein are terms defined in consideration of functions in the present invention, which may vary according to the intention or customs of those skilled in the art, and the definitions thereof are based on the contents throughout the present application. Will have to be lowered.

In addition, since the present invention has been described through the preferred embodiment of the present invention, in view of the technical difficulty aspects of the present invention, those having ordinary skill in the art can easily be modified from other embodiments of the present invention. As can be added, it is apparent that all embodiments and modifications cited in the above description belong to the claims of the present invention.

As described in detail above, in the block-based image binarization method of dividing the binarization target image into block units and performing binarization, while dividing the binarization target image into blocks so as to overlap the horizontal and vertical directions. A process of performing binarization on the pixels located inside the block of interest without performing binarization on the pixels located at the boundary of the block of interest by performing region determination on the block of interest is performed. According to the block overlap-based image binarization apparatus and method according to the present invention, which can be binarized while repeatedly reflecting the distribution characteristics of adjacent pixels by repeatedly performing the same operation, the same region of pixels located at the boundary of the block or pixels located inside the block. Forced binarization by applying judgment result That can minimize uneven development and globalization that occurred along the Torah can achieve the effect of improving picture quality according to the remarkable.

However, the present invention divides the binarization target image so as to overlap a part of the block in the horizontal and vertical directions, and as a result, the number of blocks increases and the amount of calculation increases, but the recent trend of hardware technology development and the demand for high quality of users In view of the above, an increase in the amount of calculation may be devalued in a slight burden in view of the performance improvement of the present invention.

Claims (25)

In the image binarization method of performing binarization by reflecting the pixel distribution characteristic of the binarization target image: A block overlap division step of dividing the binarization target image into blocks having a predetermined size so as to overlap at least one of a horizontal direction and a vertical direction to generate a block group composed of a plurality of blocks; A block selection step of sequentially selecting a block of interest to be binarized at this time in the block group according to a predetermined scanning order; A block classification step of classifying a pixel distribution characteristic in a block pixel group including pixels located in the block of interest and classifying an area to which the block of interest belongs as one of two or more areas; And And a block-based binarization step of binarizing pixels included in the block of interest according to a region to which the block of interest belongs, through different thresholds. The method of claim 1, wherein the block classification step, A step of calculating gray scale information by comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest; And And an area determining step of calculating an even maximum gradation even between the maximum gradation value and the minimum gradation value, and determining an area to which the block of interest belongs according to whether even the maximum gradation is greater than a preset threshold. An overlapping block based image binarization method characterized by the above-mentioned. The method of claim 2, wherein the block-based binarization step As a result of the determination of the region, if the maximum system is even greater than the threshold, a first threshold value proportional to the average of the maximum gray value and the minimum gray value of the pixels included in the block of interest is determined. And a first block-based binarization step of binarizing through the block. The method of claim 3, wherein the block-based binarization step, As a result of the determination of the region determination step, if the maximum system is even smaller than the threshold, the second block-based binarization step of binarizing pixels included in the block of interest through a dither method using a dither threshold matrix is further performed. Superimposed block-based image binarization method comprising a. The method of claim 3, wherein the block-based binarization step, As a result of the determination of the region, if the maximum system is even smaller than the threshold, the second block base is binarized through the second threshold, which is smaller than the first threshold, of pixels included in the block of interest. The overlapping block-based image binarization method further comprising a binarization step. The method of claim 5, wherein the second threshold, And the maximum gray value multiplied by a proportional constant. The method of claim 6, wherein the proportional constant, Block overlap-based image binarization method, characterized in that 0.5. The method of claim 5, wherein the second threshold, And a value proportional to an average of pixel values of the block pixel group and a value proportional to an average of all pixel values of the binarization target image. The method of claim 1, wherein the predetermined scanning order is: Superimposed block based image binarization method characterized in that the raster scanning order (raster scanning order). The method of claim 1, wherein the size of the block, 4 pixels An overlapping block based image binarization method characterized by being four pixels. The method of claim 1, wherein the block overlap division step And dividing the binarization target image into blocks having a predetermined size so that half of the blocks overlap each other in a horizontal direction and a vertical direction to generate a block group composed of a plurality of blocks. The pixel of claim 1, wherein the pixels included in the target block include: The superimposed block-based image binarization method of claim 10, wherein the pixel comprises blocks of 25% of pixels included in a square block close to the center of the block of interest. In the image binarizing apparatus which performs binarization by reflecting the pixel distribution characteristic of the binarization target image: A block overlapping division unit for dividing the binarization target image so as to overlap at least one of a horizontal direction and a vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks; A block selector for sequentially selecting a block of interest to be binarized at this time in the block group according to a predetermined scanning order; A gray scale information calculating unit for comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest; A threshold calculation unit for calculating a maximum threshold that is even between the maximum gray value and the minimum gray value, a first threshold proportional to an average of the maximum gray value and the minimum gray value; A comparison unit which receives the maximum system even from the threshold calculation unit and compares a threshold even the maximum system with a predetermined system and outputs a comparison result; A first block-based binarization unit which receives the first threshold value from the threshold calculator and binarizes pixels included in the block of interest; A dither binarizing unit for binarizing pixels included in the block of interest through a dither method using a dither threshold matrix; And An output multiplexer for outputting the output of the first block-based binarizer if the comparator even outputs a comparison result that the maximum system is even greater than the threshold, and otherwise outputs the output of the dither binarizer Block overlay based image binarization apparatus comprising a. The method of claim 13, wherein the predetermined scanning order is: An overlapping block based image binarization apparatus characterized in that it is a raster scanning order. The method of claim 13, wherein the size of the block, 4 pixels An overlapping block based image binarization apparatus, characterized in that four pixels. The method of claim 13, wherein the block overlapping partition, And dividing the binarization target image into blocks having a predetermined size such that half of the blocks overlap each other in a horizontal direction and a vertical direction to generate a block group composed of a plurality of blocks. The pixel of claim 13, wherein the pixels included in the target block include: The overlapping block-based image binarization apparatus of claim 11, wherein the pixel includes the 25% pixels included in the square block near the center of the block of interest. In the image binarizing apparatus which performs binarization by reflecting the pixel distribution characteristic of the binarization target image: A block overlapping division unit for dividing the binarization target image so as to overlap at least one of a horizontal direction and a vertical direction in units of blocks having a predetermined size to generate a block group composed of a plurality of blocks; A block selector for sequentially selecting a block of interest to be binarized at this time in the block group; A gray scale information calculating unit for comparing and calculating a maximum gray scale value and a minimum gray scale value in a block pixel group including pixels located in the block of interest; Even the maximum gradation, which is even a gradation between the maximum gradation value and the minimum gradation value, has a first threshold that is proportional to the average of the maximum gradation value and the minimum gradation value, and a second threshold that is relatively smaller than the first threshold value, respectively. A threshold calculation unit for calculating; A comparison unit which receives the maximum system even from the threshold calculation unit and compares a threshold even the maximum system with a predetermined system and outputs a comparison result; A first block-based binarization unit which receives the first threshold value from the threshold calculator and binarizes pixels included in the block of interest; A second block-based binarization unit which receives the second threshold value from the threshold calculator and binarizes pixels included in the block of interest; And The comparator outputs the output of the first block-based binarization unit if the comparator outputs a comparison result that even the maximum system is even greater than the threshold, otherwise outputs the output of the second block-based binarization unit. Block overlap based image binarization apparatus comprising an output multiplexer. The method of claim 18, wherein the predetermined scanning order, An overlapping block based image binarization apparatus characterized in that it is a raster scanning order. The method of claim 18, wherein the size of the block, 4 pixels An overlapping block based image binarization apparatus, characterized in that four pixels. The method of claim 18, wherein the block overlapping partition, And dividing the binarization target image into blocks having a predetermined size such that half of the blocks overlap each other in a horizontal direction and a vertical direction to generate a block group composed of a plurality of blocks. The pixel of claim 18, wherein the pixels included in the target block include: The overlapping block-based image binarization apparatus of claim 11, wherein the pixel includes the 25% pixels included in the square block near the center of the block of interest. The method of claim 18, wherein the second threshold value, And the maximum gray value multiplied by a proportional constant. The method of claim 23, wherein the proportional constant, Block overlay-based image binarization device, characterized in that 0.5. The method of claim 18, wherein the second threshold value, And a value obtained by averaging pixel values of the block pixel group and an average value of all pixel values of the binarization target image.