JPH10232926A - Image processor and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the image processor and its method which process an image by setting a proper quantization threshold between object density and background density in the input image. SOLUTION: When a multilevel image is quantized and processed, a luminance frequency accumulation part 103 calculates the luminance frequency of the multilevel image, a quantization threshold calculation part 103 specifies a quantization threshold according to the calculated luminance frequency, and a quantization part 105 calculates a representative value used for the quantization of the multilevel image according to the specified quantization threshold and luminance frequency and quantizes the multilevel image by using the calculated representative values.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to, for example, an image processing apparatus and method for determining a quantization threshold of a multi-valued image and performing quantization.

[0002]

2. Description of the Related Art In recent years, image processing techniques have been remarkably developed, and image processing apparatuses capable of processing multi-valued images such as full-color images and character recognition processing in multi-valued images have become widespread. In such an image processing technique, the multivalued image 2
Value processing is an indispensable technology.

[0003] Conventional binarization methods include a simple binarization method using a fixed threshold value set in advance, and a method when the inter-class variance becomes maximum when a histogram is divided into two classes by a certain threshold value. Otsu method with threshold as binarization threshold (“Automatic threshold selection method based on discrimination and least square criterion” (Otsu), IEICE Transactions, Vo
l. J63-D, No. 4. pp. 349-356, 1
980) Alternatively, there is a binarization method of setting a threshold value for an image having a gradation according to the local density.

[0004]

However, the above-described binarization method in the conventional image processing apparatus has the following problems.

That is, in the simple binarization method using a fixed threshold value, it is difficult to set an appropriate threshold value between the object density and the background density in an image, and as a result, the entire image is crushed black or the reverse. It turns white. Further, in the Otsu method, when the distributions of the two classes are extremely different, the threshold value is closer to the larger class, and a binary image with much noise is generated. Furthermore, in the binarization method of setting a threshold according to the local density,
Since the image is divided locally, block distortion is likely to occur. Further, even if the optimum threshold value can be specified, there is a problem that the grayscale information such as the base of the original image and characters is lost due to the binarization.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an image processing apparatus for performing image processing by setting an appropriate quantization threshold between an object density and a background density in an input image, and It is intended to provide such a method.

Further, in order to solve the above-mentioned problems, the present invention provides an image processing apparatus and method which can specify an optimal quantization threshold and perform area separation without losing grayscale information such as the background of an original image or characters. The purpose is to provide.

[0008]

In order to achieve the above object, the present invention provides an image processing apparatus for quantizing a multi-valued image and performing image processing, the first method comprising calculating a luminance frequency of the multi-valued image. A calculating unit, a specifying unit that specifies a quantization threshold value for quantization based on the calculated luminance frequency, and a quantization unit that quantizes the multi-valued image based on the specified quantization threshold value and the luminance frequency. It is characterized by comprising a second calculating means for calculating a representative value, and a quantizing means for quantizing the multi-valued image using the calculated representative value.

According to another aspect of the present invention, there is provided an image processing method for performing image processing by quantizing a multi-valued image, comprising: calculating a luminance frequency of the multi-valued image; Based on the, specify a quantization threshold of quantization, based on the specified quantization threshold and the luminance frequency, calculate a representative value used for quantization of the multi-valued image, the calculated representative value The multi-valued image is quantized using the quantization.

[0010]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing a configuration of an image processing system for executing a quantization process according to an embodiment of the present invention. In FIG. 1, reference numeral 1 denotes an image processing device for performing a character recognition process, 2 denotes an image input device such as a scanner for inputting an image, and 3 denotes an image display device for displaying a processed image.

In the image processing apparatus 1 described above, reference numeral 101 denotes an input unit serving as an interface with the image input apparatus 2;
Reference numeral 02 denotes a storage unit such as a memory for storing data being processed;
Reference numeral 03 denotes a luminance frequency accumulator that accumulates the luminance frequency (histogram) of the input image. Reference numeral 104 denotes a quantization threshold calculation unit that calculates a quantization threshold of the input image, and reference numeral 105 denotes a quantization unit that creates a quantization image using the threshold calculated by the quantization threshold calculation unit 104.

Reference numeral 106 denotes an area separating unit for separating an image into regions for each attribute, 107 denotes a character recognizing unit that performs a character recognition process on a region extracted as a character region by this region separation, and 108 denotes a region other than a character region. An image processing unit 109 that performs various types of image processing on the area, an output unit 109 serving as an interface with the image display device 3. Each of these components is a C that controls the entire device 1.
PU, RO storing the program of the CPU, etc.
M, a work area, a table, and the like used by the CPU at the time of execution of processing are generally controlled by a control unit (not shown) including a RAM in which a work area, a table, and the like are defined.

Hereinafter, the OCR processing executed in the image processing apparatus according to the present embodiment having the above configuration will be described.

FIG. 2 is a flowchart showing an image area separation OCR process using the quantization threshold value determination method according to the present embodiment.

First, in step S 201, the input unit 101 inputs image data from the image input device 2 such as a scanner, and stores the image data in the storage unit 102. The input of the image data here is performed as 8-bit multi-valued image data. Then, in step S202, step S
For the multi-valued image input in 201, the quantization threshold calculation unit 104 determines an optimal quantization threshold for image area separation described later, and the quantization unit 105 generates a quantized image based on the quantization threshold. I do. Then, in step S203, the area separation unit 106 performs image area separation of the quantized image generated in step S202, and outputs the area data to which the attribute has been added to the image processing unit 108. In the following step S204, the image processing unit 108 binarizes the area designated as "text" with respect to the area data separated in step S203, and then cuts out the area from the binary image. Then, the character recognizing unit 107 performs an OCR process on the binary image, and outputs a recognized character code.

<Description of Quantization Processing> The quantization processing in the present embodiment will be described.

FIG. 3 is a flowchart showing the procedure of the quantization process in the present embodiment. First, in step S301, an 8-bit multivalued image is input from the storage unit 102 in the image processing apparatus 1 to a memory (not shown) or the like, and in step S302, a processing block unit (64 ×
(64 pixels). It is assumed that the multi-valued image is read by the image input device 2 such as a scanner and stored in the storage unit 102 in advance. Then, in step S303, the luminance frequency accumulating unit 103
Calculates a histogram for each processing block. Here, the frequency for each digital value of 8 bits, that is, “0” to “255” is calculated using all the pixels of the processing block. Thereby, for example, a histogram as shown in FIG. 6 is obtained.

Next, in step S304, the parameters START and END are set to "0" and "25", respectively.
5 ”. These parameters STAET, E
ND corresponds to the start point and the end point of the statistic of the luminance value obtained in the subsequent step S305 or S306, respectively.

In step S305, START to E
An average value AV of pixels corresponding to digital values up to ND is calculated. For example, if START = 0 and END = 255, the average value AV of the pixels having values from “0” to “255” (in this case, all pixels) is calculated, and START = 0, END = 255.
If END = 177, the average value AV of pixels having values from “0” to “177” is calculated.

In step S306, START to E
The skew value Sk of the pixel corresponding to the luminance value up to ND is calculated. Here, the skew value is a statistic indicating the bias of the histogram distribution. For this skew calculation, the following equation (1) is used.

Sk = (Σ (Xi-AV) ＾ 3) / D (1) Here, “＾” means a power, and Xi is a luminance value of a pixel. D is a variance value of the entire image, and is calculated by the following equation (2).

D = {(Xi−AV)} 2 (2) In the above equation (1), the skew value is calculated by raising the difference between the luminance value of each pixel and its average value to the cube. However, if it is an odd power, it is not limited to the third power.

In subsequent steps S307 and S308, the direction of the bias of the histogram is determined. First, step S
In 307, the bias direction of the histogram is determined by the following equation (3). This is because the histogram bias is
It is determined whether or not the value is in a range smaller than the average value AV.

Sk <−1.0 (3) In step S307, regarding the calculated skew value, if equation (3) is “true”, the process proceeds to step 312. If equation (3) is “false”, Step S308
Proceed to. In this step S312, the average value AV is set in END without changing START. Then, the process returns to step S305, where the END value is again changed from the START value.
An average value AV up to the value is calculated.

On the other hand, in step S308, the bias direction of the histogram is determined by the following equation (4).
This is to determine whether or not the bias of the histogram is in a range of values larger than the average value AV.

Sk> 1.0 (4) In step S308, if equation (4) is “true” with respect to the obtained skew value, the process proceeds to step S313, and if it is “false”, the process proceeds to step S313. Proceed to S309. In step S313, the average value AV is set in START, and END is not changed. And step S
Returning to 305, the average value AV from the START value to the END value is calculated again.

On the other hand, in step S309, step S309
The average value AV when the conditions in 307 and S308 are both “false” is set as the quantization threshold TH. Then, in step S310, a quantization process using the quantization threshold TH is performed.

In step S311, it is determined whether or not the input image is the last processing block (64 × 64 pixels). If it is the last processing block, the processing is terminated. It returns to S302.

The state of the quantization will be described with reference to FIG.

In the histogram calculated in step S303, an area smaller than the quantization threshold value TH calculated in step S309 is denoted by BB.
A region larger than H is defined as WB. Normally, binarization is performed by setting the representative value of the BB area to 0 and the representative value of the WB area to 1. However, in this case, the gray information is lost.

Therefore, in the present embodiment, the average value BBV of the BB area and the average value WBV of the WB area are calculated, and the image is quantized using these two average values BBV and WBV.

As a result, as shown at 501 in FIG. 5, the multi-valued information of the image area is represented by only two types of multi-valued information. The data after quantization is represented by 5 in FIG.
As shown in 02, the header information 503 based on BBV and WBV may be attached to a bitmap in which the area represented by BBV is replaced by 0 and the area represented by WBV is replaced by 1. Also, BBV,
The WBV is not limited to the average value, but may be the median value of each of the BB area and the WB area.

As described above, the quantization processing according to the present embodiment is performed, but the range shown by equations (3) and (4) is not limited to this.

Hereinafter, the quantization processing according to the present embodiment will be described in more detail with reference to specific examples of images. Using the example of the histogram shown in FIG. 6, the process of determining the quantization threshold TH in the present embodiment will be described.

FIG. 6 shows a histogram of a certain image (8-bit input). In the figure, the horizontal axis is
The left end is “0”, that is, black, and the right end is “255”, that is, a digital value of luminance representing white, and the vertical axis represents the frequency of each digital value.

FIG. 7 shows an image having a histogram as shown in FIG. 6 in the quantization processing shown in FIG.
FIG. 14 is a diagram illustrating a change in the value of each parameter at the time of processing in steps 305 and S306. Each parameter value in FIG. 7 is indicated by the number of times of passing through step S305 and step S306 in FIG.

First, in the first process passing through steps S305 and S306, START = 0 and END = 25
5, the average value AV and the skew value Sk are calculated to obtain values of “177” and “−78.9”, respectively. In this case, since the skew value Sk is less than [-1.0], in step S312 in FIG. 3, START = 0, EN
D = 177 is set. Then, in the second process,
Average value AV at START = 0, END = 177,
The skew value Sk is calculated, and “91”, “−”
8.6 ". Also in this case, since the skew value Sk is less than "-1.0", START = 0 and END = 91 are set in step S312 in FIG.

In the third processing, START = 0, EN
The average value AV and the skew value Sk at D = 91 are calculated to obtain values "43" and "9.6", respectively.
In this case, since the skew value Sk exceeds “1.0”,
In step S313 of FIG. 3, START = 43,
END = 91 is set. In the subsequent fourth processing, S
The average value AV and skew value Sk at TART = 43 and END = 91 are calculated, and are respectively “72” and “−7.
0 "is obtained. The skew value Sk for this value also
Is less than “−1.0”, so that step S31 in FIG.
At 2, START = 43 and END = 72 are set.

In the fifth processing, START = 43, E
The average value AV and the skew value Sk at ND = 72 are calculated to obtain values “58” and “−2.2”, respectively. Since the skew value Sk is also less than “−1.0”, in step S312 in FIG.
= 43 and END = 58 are set. Then, in the sixth process, the average value AV and the skew value Sk at START = 43 and END = 58 are calculated, and each of them is set to “5”.
0 ”and“ −0.4 ”are obtained.

Here, the skew and Sk are not less than "-1.0" and not more than "1.0".
7, the condition of S308 is not satisfied (the determination there is N
O), the process proceeds to step S309, and “50” is set as the quantization threshold TH. Then, in the subsequent step S310, the quantization threshold T
The quantization process using H is performed, and the quantized image is stored in the storage unit 102 in the image processing apparatus l.

In this quantization, the average value of the region frequency smaller than the quantization threshold value TH is set as the representative value 1, the average value of the region frequency larger than the quantization threshold value TH is set as the representative value 2, and these two values are used. Performs quantization. However, any representative value may be used as long as it represents a feature value of a region frequency smaller than the quantization threshold value TH and a region frequency larger than the quantization threshold value TH. For example, a median value may be used instead of the average value. No problem.

<Description of Image Area Separation Processing> The image area separation processing (step S203 in FIG. 2) using the quantization result will be described in detail with reference to the flowchart shown in FIG.

First, in step S801 of FIG.
A quantized image is input and stored in the storage unit 102. In step S802, the input image is thinned such that m × n pixels become one pixel, and an image for image area separation is generated.
At this time, if at least one black pixel exists in the m × n pixels, this pixel is regarded as one black pixel. Then, in step S803, for all pixels of the image area separation image, area division is performed with a predetermined number of black pixels continuing vertically, horizontally, and diagonally as one area. At this time, labels are assigned to the respective regions by assigning numbers to the regions in the order of detection.

Next, in step S804, the regions are classified according to the width and height of each region and the density of black pixels in the area, and the attributes are labeled. Area attributes include, for example, “table”, “outer frame area”, “text”, and the like. Then, in step S805, the average of the width and height of all the regions labeled "text" is calculated,
When the obtained average width is larger than the average height, the character set is determined by regarding the processed image as horizontal writing, and conversely, regarding the processed image as vertical or lower. At the same time, the average height is used for horizontal writing, and the average width is used for horizontal writing, which is the size of one character.

Further, from the histograms of all the "text" areas in the vertical direction (when writing horizontally) or in the horizontal direction (when writing vertically) on the image for separating the image area, columns of sentences and line intervals are detected. . In step S806, "text"
In the area, an area having a large character size is referred to as a “title”.

However, in the conventional area determination based on the binarized image, even if a band indicating title emphasis exists in the background of the area determined as the title, the background information is lost. I can't do that. For the same reason, even if the title character itself is colored, it is simply determined as "black character". However, adding a band to the background of the title or coloring the title character is determined to be the same "Title" regardless of the intention that the document creator wants to distinguish the title from other titles It was a weak point of the conventional method.

The greatest advantage of using the present quantized image as an image for segmentation lies in the improvement of this point. For example, when the title in the original image is represented by 200 gray characters in an 8-bit expression and the background is represented by a gray band of 64, as shown in 901 in FIG. Regardless of whether there is a background color or not, the character color and the background color information are lost at the time of the binarization, as shown in 902 in FIG.

On the other hand, in this quantization, as shown by 903 in FIG. 9, for example, an area smaller than the value of the quantization threshold TH calculated in the left block of 64 × 64 pixel block unit, in this case, the character “TITLE” The pixel average value BBV of the “T” character area is calculated to be 64. Similarly, in the same block, the average value of an area larger than the value of the quantization threshold value TH, in this case, the average value WBV of the area corresponding to the background color is calculated as 200, and the BBV and. In order to attach each 64 × 64 pixel block in addition to the image,
The title text color and the background color can be distinguished considerably.

In this embodiment, the processing shown in FIG. 10 is performed on the area determined to be the title. First, step S
In step 1001, the area is determined using only the image information quantized as described above. Here, the process of step S1002 is performed on the area determined to be “title”. The symbols WBV, BBV, PW, PB in the figure represent the following.

WBV: Average value of an area larger than the quantization threshold value TH calculated in block units BBV: Average value of an area smaller than the quantization threshold value TH calculated in block units PW: Base of original image Area representative value PB: Representative value of character area of original picture WBV and BBV are attached in addition to the binarization information for each 64 × 64 pixel dormitory block at the time of main quantization.
On the other hand, PW is a representative value of the base of the original image, and means the white density of the paper when read by the scanner. The PW is set in advance based on several types of samples because there is a slight variation between scanner devices and paper types. P
B is a representative value of the character area of the original image, and means the density of the character printed on the paper when read by the scanner. PB is also PW
Similarly, since there is some variation between scanner devices and paper types, the setting is made in advance based on several types of samples.

First, WBV and PW are compared, and WBV is
If it is smaller than W, it is assumed that the background color of this block exists on paper with a background color darker than the background. Next, BBV and P
B is compared, and if BBV is greater than PB, it is assumed that the character density printed in this block is a color character lighter than the normal black character density. Therefore, if WBV is smaller than PW or WBB is larger than PB, it is determined that the title in this area has a background color or a character color that emphasizes the title, and the process branches to step S1003. Otherwise, the process proceeds to step S1004. move on. On the other hand, in step S1003, labeling is performed as "highlighted title", and in step S1004, labeling is performed as "normal title".

Then, in step S807, the "title" area which has been present independently without any relation,
The “text” area is merged according to the interval with the surrounding area to form one integrated area.

Next, in step S808, area data such as an attribute and coordinates and size in the original image is output for each area. By performing the above processing, image area separation processing of the quantized image is performed, and area data is obtained.

FIG. 11 is a diagram showing an example of the above-mentioned area data. Each area data item shown in FIG. "No.": Indicates the detection order of the areas. “Attribute”: Indicates attribute information of the area, and the following nine patterns are prepared.

"Route" Indicates that the image is the input image itself.

"Text" Indicates a character.

"Emphasis title" indicates that the title area is emphasized.

"Normal title" indicates a normal heading area.

"Table" Indicates a table area.

"Noise area" Indicates an area that could not be determined as either a character or an image. "Outer frame area" Indicates an area such as a ruled line.

"Photo image" Indicates a photo area.

"Line image" Indicates a line image area. "Start point coordinates": X and Y coordinates of the area start in the original image. "End point coordinates": X and Y coordinates of the end of the area in the original image. "Number of pixels": Indicates the total number of pixels in the area. "Character set information": Indicates three types of character set information: vertical writing, horizontal writing, and unknown.

In the area data shown in FIG. 11, only the area whose “attribute” is indicated by “text” hierarchically holds the area data (row area data) relating to the row before the merging in step S807 in FIG. ing.

Here, the image area separation processing is performed as described above. However, the area data shown in FIG. 11 is merely an example to which the present embodiment is applied, and for example, depending on the image processing apparatus, another area data may be used. Information may be added or reduced as appropriate.

As described above, according to the present embodiment, the quantization threshold is determined such that the luminance frequency of the input multi-valued image and the skew value indicating the deviation converge to a predetermined value. By performing quantization based on the quantization threshold, an area where a threshold most suitable for separating a background and an object in an image is specified, and then the average brightness value of this specific area is used to perform quantization. This makes it possible to easily obtain the optimum value for classifying the luminance value of each pixel in the area on the multi-valued input image into two classes, the background and the object, and to achieve high definition. OCR processing can be executed.

In the above-described embodiment, the input image is 8-bit multi-valued image data. However, the present invention is not limited to this. Therefore, it is sufficient if there is information of a plurality of bits as image information. Further, the convergence condition of the skew value Sk, which is a statistic, is ± 1.0, but is not limited thereto. In other words, it is sufficient that the threshold value for binarization is determined using the skew value Sk.

The present invention can be applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), but can also be applied to a single device (for example, a copying machine, a facsimile machine). Etc.).

Further, an object of the present invention is to supply a storage medium storing a program code of software for realizing the functions of the above-described embodiments to a system or an apparatus, and to provide a computer (CPU or MP) of the system or apparatus.
It goes without saying that U) can also be achieved by reading and executing the program code stored in the storage medium.

In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention.

As a storage medium for supplying the program code, for example, a floppy disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD
-R, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also the OS (Operating System) running on the computer based on the instruction of the program code. ) May perform some or all of the actual processing, and the processing may realize the functions of the above-described embodiments.

Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, based on the instruction of the program code, It goes without saying that the CPU included in the function expansion board or the function expansion unit performs part or all of the actual processing, and the functions of the above-described embodiments are realized by the processing.

[0074]

As described above, according to the present invention,
Based on the luminance frequency of the multi-valued input image and the bias of its distribution, an area where the luminance frequency is minimal is specified, and quantization is performed using the average luminance value of the specified area as a quantization threshold, thereby obtaining an image. An appropriate threshold value can be set between the target object and the background density, and an image in which occurrence of block distortion can be suppressed can be obtained.

[0075]

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image processing system according to an embodiment.

FIG. 2 is a flowchart illustrating an image area separation OCR process according to the present embodiment.

FIG. 3 is a flowchart illustrating a binarization process in the embodiment.

FIG. 4 is a diagram for explaining a quantization process in step S9 shown in FIG. 3;

FIG. 5 is a diagram for explaining a quantization result.

FIG. 6 is a diagram illustrating an example of an image histogram according to the present embodiment.

FIG. 7 is a diagram illustrating an example of conversion of each variable value in a binarization process in the embodiment.

FIG. 8 is a flowchart illustrating an image area separation process according to the present embodiment.

FIG. 9 is a diagram for explaining emphasized titles in the embodiment.

FIG. 10 is a flowchart illustrating a title determination process in the embodiment.

FIG. 11 is a diagram illustrating an example of area data according to the present embodiment.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 image processing device 2 image input device 3 image display device 101 input unit 102 storage unit 103 luminance frequency accumulation unit 104 quantization threshold calculation unit 105 quantization unit 106 area separation unit 107 character recognition unit 108 image processing unit 109 output unit

Claims (13)

[Claims] 1. An image processing apparatus for performing image processing by quantizing a multi-valued image, comprising: first calculating means for calculating a luminance frequency of the multi-valued image; Specifying means for specifying a quantization threshold, based on the specified quantization threshold and the luminance frequency,
An image, comprising: second calculating means for calculating a representative value used for quantization of the multi-valued image; and quantizing means for quantizing the multi-valued image using the calculated representative value. Processing equipment. 2. The method according to claim 1, wherein the quantization threshold is an average luminance value when the histogram distribution is converged such that the bias of the luminance frequency histogram distribution falls within a predetermined range. The image processing apparatus according to any one of the preceding claims. 3. The image processing apparatus according to claim 1, wherein the representative value is an average luminance value in each distribution area of a luminance frequency histogram distribution divided by the quantization threshold. 4. The method according to claim 1, wherein the representative value is a central luminance value in each distribution area of a luminance frequency histogram distribution divided by the quantization threshold.
The image processing apparatus according to any one of the preceding claims. 5. The image processing apparatus according to claim 1, further comprising means for performing image area separation of the image quantized by said quantization means, and outputting area data including an attribute of the image area separated area. The image processing apparatus according to any one of the preceding claims. 6. The image processing apparatus according to claim 5, further comprising means for judging whether or not the attribute of the area is a title when the attribute is a title. 7. An image processing method for performing image processing by quantizing a multi-valued image, wherein a luminance frequency of the multi-valued image is calculated, and a quantization threshold value for quantization is set based on the calculated luminance frequency. Identifying, based on the identified quantization threshold and the luminance frequency,
An image processing method, comprising: calculating a representative value used for quantization of the multi-valued image; and quantizing the multi-valued image using the calculated representative value. 8. The method according to claim 7, wherein the quantization threshold is an average luminance value when the histogram distribution is converged such that the deviation of the histogram distribution of the luminance frequency is within a predetermined range. The image processing method described in the above. 9. The image processing method according to claim 7, wherein the representative value is an average luminance value in each distribution area of a luminance frequency histogram distribution divided by the quantization threshold. 10. The image processing method according to claim 7, wherein the representative value is a central luminance value in each distribution area of a luminance frequency histogram distribution divided by the quantization threshold. 11. The image processing method according to claim 7, further comprising the step of performing image area separation of the image quantized in the quantization step, and outputting area data including an attribute of the image area separated area. The image processing method described in the above. 12. The image processing method according to claim 11, further comprising the step of determining whether the attribute of the tack area is a title if the attribute is a title. 13. A computer-readable storage medium storing a program code of an image processing method, comprising: a code of a step of calculating a luminance frequency of the multi-valued image; Code of the step of specifying a quantization threshold, Based on the specified quantization threshold and the luminance frequency,
A storage medium, comprising: a code for calculating a representative value used for quantization of the multi-valued image; and a code for quantizing the multi-valued image using the calculated representative value.