JP2000155804A - Information processor and its method, and computer- readable memory - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the character recognition precision of image data composed of pieces of color information by performing binarization by image data of the color components of color image data and putting together and outputting binarized image data by binarized color components. SOLUTION: Appearance frequency graphs are found for color components R, G, and B of the color image data and thresholds are determined (S11). Here, the peak values of appearance frequencies of the density of character parts and the density of the background part are detected and their intermediate value is decided as a threshold. Then, binarization is performed (S12) for the color components R, G, and B by using the determined thresholds. In the binarization, component values exceeding the thresholds are set to '0' and component values less than the thresholds are set to '1'. The obtained binarized data by the color component are logically put together (S13). In this case, OR operation is performed by the pixel positions of the binarized data of the color components. Consequently, image data suitable for optical character recognition(OCR) can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an information processing apparatus and method for generating image data for character recognition from input color image data, and a computer readable memory.

[0002]

2. Description of the Related Art Conventionally, OCR (optical character recognition)
It is intended for monochrome images. To apply OCR to a color image, it is necessary to temporarily convert the color image into a monochrome image.

[0003] Conversion to a monochrome image is performed by color conversion from RGB components to YUV components or from RGB components to L *,
Generally, color conversion into a * and b * components is performed, and image data of a Y component or an L * component representing luminance is treated as monochrome image data, and a process such as binarization is generally performed.

[0004]

However, in the conventional conversion method for extracting a luminance component, a color having a high luminance component, such as green, yellow, and cyan (light blue) characters, has a background color when binarized.
That is, the character is easily converted to white, and the character is completely erased or broken off.
CR), there is a problem that color characters cannot be recognized well.

When the background is a dark color such as black and the character portion is a light color, the black and white are reversed or the background and the character such as a green character on a yellow background or a yellow character on a green background are reversed. In the latter case, when the luminance of the color image data is close to each other, the black-and-white image after the binarization is inverted, a problem specific to the binarization of the color image data has emerged.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has an information processing apparatus and method capable of improving the character recognition accuracy for image data composed of a plurality of pieces of color information, and a computer readable memory. The purpose is to provide.

[0007]

An information processing apparatus according to the present invention for achieving the above object has the following arrangement.
That is, an information processing apparatus for generating image data for character recognition from input color image data, comprising: a binarizing unit for binarizing image data of each color component of the color image data; A synthesizing unit for synthesizing the binarized image data for each color component binarized in the above, and an output unit for outputting the binarized image data synthesized by the synthesizing unit.

Preferably, the binarizing means sets a threshold value for binarizing the image data of each color component based on a frequency of appearance of a pixel value of the image data of each color component of the color image data. The image data of each color component is binarized based on the determining means for determining and the threshold value determined by the determining means.

Preferably, the determining means determines a background color in the color image data, and binarizes the image data of each color component based on the determined pixel value of the background color. Is determined.

[0010] Preferably, the image processing apparatus further comprises designating means for designating a background color in the color image data, wherein the deciding means is based on the pixel value of the background color designated by the designating means. The threshold for binarizing the image data of each color component is determined. Also, preferably,
The binarizing means includes: first binarizing means for binarizing image data of each color component of the color image data using a first threshold; and each color component of the color image data using a second threshold. And a second binarizing means for binarizing the image data.

Preferably, the output means includes: first binary image data obtained by synthesizing binary image data for each color component, which has been binarized by the first binary means; Comparing means for comparing second binarized image data obtained by synthesizing the binarized image data for each color component binarized by the second binarizing means, based on a comparison result of the comparing means hand,
Either the first binarized image data or the second binarized image data is output.

An information processing method according to the present invention for achieving the above object has the following configuration. That is, an information processing method for generating image data for character recognition from input color image data, comprising: a binarization step of binarizing image data of each color component of the color image data; And a outputting step of outputting the binarized image data synthesized in the synthesizing step.

[0013] A computer readable memory according to the present invention for achieving the above object has the following configuration. That is,
A computer-readable memory storing a program code for information processing for generating image data for character recognition from input color image data, wherein the binarization step binarizes each image data of each color component of the color image data. , A program code of a synthesizing step for synthesizing the binarized image data for each color component binarized in the binarizing step, and an output for outputting the binarized image data synthesized in the synthesizing step Process code.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the drawings. [First Embodiment] FIG. 1 is a block diagram showing a schematic configuration of an information processing apparatus according to a first embodiment of the present invention.

Reference numeral 1 denotes a CPU which executes the processing of the present invention and performs various controls. Reference numeral 2 denotes a CD-ROM drive. 3 is a floppy disk drive (FD)
D). 6 is a hard disk drive (HD
D). Each drive device stores a storage medium such as a CD-ROM, a floppy disk, or a hard disk, and a program recorded on any one of the storage media is activated to perform the processing of the present invention. A communication device 8 enables connection with a network or the like. The network may be a LAN or an interface such as SCSI. Reference numeral 9 denotes a printer for recording the recording data specified by the application. 3 is a mouse. 4 is a keyboard. The operator can interactively set or change the processing content from the mouse 3 or the keyboard 4 while referring to the monitor 5.

Next, an execution window displayed on the monitor 5 in each processing including the OCR processing executed in the first embodiment will be described with reference to FIG.

FIG. 2 is a diagram showing an example of an execution window according to the first embodiment of the present invention.

In FIG. 2, reference numeral 21 denotes a document display unit.
This area displays the document image data to be processed. The menu display unit 23 displays the document image displayed in this area.
The OCR instruction and the OCR pre-processing can be executed by clicking with the mouse 3 from each menu or pull-down menu. 22 is a background part,
This area displays the background of the document image data.

Reference numeral 23 denotes a menu display unit, which is an area for displaying a group of processing menus executed on the document image data. Reference numeral 24 denotes a toolbar.
Is an area for displaying an icon obtained by converting the processing menu group displayed by the icon into an icon. Reference numeral 26 denotes a status display section, which is an area for displaying the current state of the execution window.

The following describes the information processing apparatus described above and the processing contents that operate according to the instruction of the operator through the execution window.

FIGS. 3 to 5 show the appearance frequencies of all or one pixel point of the document image data to be subjected to OCR or the pixel points sampled at appropriate intervals in the entire density range for each of the R, G, and B color components. It is a figure showing an example of a (frequency) graph.

In general, the appearance frequency of the density of the background portion and the character portion increases, and the intermediate point in the appearance frequency graph of each color component is set as threshold values Rt, Gt, and Bt to perform binarization, thereby performing character recognition (OCR). Binary image data suitable for the above can be obtained. Also, an intermediate density portion in the color image data that does not include characters may become noise as a result of binarization or an outline of the image may be extracted, but as a result of the OCR, that portion is usually ignored.

FIG. 6 shows color image data in which green (green) characters are drawn. An R (Red) component obtained by color separation of this color image data,
The image data of the n) component and the B (Blue) component are as shown in FIGS. 7, 8, and 9, respectively. In FIG. 8, all G components are present, and in FIGS. 7 and 9, the background has all R components and B components. When the luminance component is binarized according to the related art, only unclear binary image data in which characters are not clear as shown in FIG. 10 is obtained, and it is difficult to obtain an effective OCR result. This is because the G component contributes to the luminance component at a high rate, and the characters on the luminance component are bright and pale. It goes without saying that a similar phenomenon occurs even in a case other than the green characters.

Therefore, in the first embodiment, the following processing is performed to improve the above problem. Hereinafter, the first embodiment
Will be described with reference to FIG.

FIG. 12 is a flowchart showing the processing executed in the first embodiment of the present invention. In step S11, as shown in FIGS.
An appearance frequency graph is obtained for each of the color components R, G, and B, and a threshold R
Determine t, Gt, and Bt. As the threshold value, there is a method of detecting the peak value of the appearance frequency of the density of the character portion and the density of the background portion, and using the intermediate value as the threshold value, but is not limited to this method.

In step S12, the threshold values Rt, G determined in step S11 are determined for each of the R, G, B color components.
Binarization is performed using t and Bt. In the case of binarization, a component value exceeding the threshold is “0”, and a component value below the threshold is “1”.
To This is because the so-called black portion is set to “1” and the white portion of the background is set to “0”. It is needless to say that, when the final monochrome image data is obtained after step S13, "0" and "1" may be inverted according to the input image format of the OCR.

The 2 for each color component obtained in step S12
The valued image data is logically synthesized in step S13.
In the first embodiment, a logical sum is calculated for each pixel position of the binary image data of each color component. For example, FIG.
, The binary image data of each color is roughly as shown in FIG. 7, FIG. 8, and FIG. The logical sum of such binarized image data 2
The binarized image data is as shown in FIG.
Image data suitable for CR) processing can be obtained. Actually, one of the three color components in step S12 is a reliable 2
If binarization is performed, appropriate binarized data can be obtained,
Even if the character portion of the binarization result in step S12 is truncated, the logical sum of the binarized image data of each color component is obtained in step S13. Can be eliminated, and the accuracy of character recognition can be improved.

As described above, according to the first embodiment, binarization is performed for each color component of color image data, and image data for character recognition is generated using the binarized image data of each color component. I do. Thereby, since the image data for character recognition can be generated without losing the information of the image data in each color component, more optimal character recognition can be performed. [Embodiment 2] In Embodiment 2, when color image data as shown in FIG. 13 is used, that is, when the background has an intermediate density and characters with higher density and lower density than the background are mixed, the color image data is A method for generating optimal character recognition image data will be described.

The processing executed in the second embodiment will be described below with reference to FIG.

FIG. 14 is a flowchart showing processing executed in the second embodiment of the present invention. In step S21,
Determine the background color of the image data. Although the background color is determined for each color component, the background may be determined from the density histogram of each color component as shown in FIG. Further, the image data to be processed is displayed on the monitor 5, and the user is instructed to designate the background portion by using the mouse 3, and the R, G, and B color component values of the designated point may be used as the background color. .

In step S22, a first threshold value, which is a value brighter than the background color for separating the character and the background, is determined.
In step S22, a second threshold value that is darker than the background color for separating the character and the background is determined. As an algorithm for determining these two thresholds, first, a peak value of the appearance frequency other than the background color is obtained using an appearance frequency graph as shown in FIG. Since these indicate the colors of appearing characters, some are required. However, two densities on the bright density and the dark density side closest to the background color are detected, and the intermediate value between these and the density of the background color is determined as the first density. The threshold may be a second threshold.

In step S24, using the first threshold, 3
First binarization is performed to binarize image data of an arbitrary color component among the three color components. At this time, binarization is performed so that a portion brighter than the first threshold becomes “1”. Step S25
Then, using the second threshold value, a second binarization for binarizing image data of an arbitrary color component among the three color components is performed. At this time, binarization is performed so that a component darker than the second threshold becomes “1”. In step S26, the binary image data obtained as a result of the first binarization and the second binarization in steps S24 and S25 are logically synthesized. Here, the logical sum of the binary image data obtained by each binarization may be simply obtained. In this manner, the final binary image data for one color component among the color components of the binary image data is obtained.

In step S27, it is determined whether or not the processing for all color components has been completed. If not completed (NO in step S27), the process returns to step S22. on the other hand,
If the processing has been completed (YES in step S27), the process proceeds to step S28. In step S28, the logical sum of the final binary image data for each color component is calculated. Thereby, it is possible to prevent the characters from being interrupted, and to obtain the same effect as in the first embodiment. Also,
Even if the above processing is not performed on each of the R, G, and B color components but only one of the luminance components, a sufficient effect can be obtained. In that case, step S27 and step S28 become invalid.

As described above, according to the second embodiment, the threshold value for binarizing the binary image data of each color component is determined according to the density value of the background. It is possible to generate image data for character recognition that has been appropriately binarized without losing information. [Embodiment 3] In Embodiment 3, a method of generating optimal character recognition image data from color image data as shown in FIG. 16, that is, when characters are brighter than the background, will be described. .

Hereinafter, the processing executed in the third embodiment will be described with reference to FIG.

FIG. 17 is a flowchart showing processing executed in the third embodiment of the present invention. First, step S3
In step 1, a background instruction of image data to be processed is input. The background instruction is input, for example, by displaying the image data to be processed on the document display unit 21 of the execution window shown in FIG. 2 and providing the user with a high density such as black or a high brightness low density such as white or yellow in the character portion. The user is set using the menu display unit 23. For example, open the “Edit Menu” and select “High Density Characters” and “High Brightness Characters”
Is set by selecting and clicking one of them while using the mouse 3. In this case, a check mark indicating that the menu is set is attached to one of the menus.

In step S32, a binarization method for image data is determined based on the input contents of the background instruction input. In the binarization method, for example, when the setting of “high-density character” is performed, the luminance components of R, G, and B are set to “1” when the pixel value is smaller than the threshold, and otherwise, Binarization is performed so as to be “0”. When the setting of “low-density character” is performed, the luminance component of R, G, and B is set to “1” when the pixel value is larger than the threshold, and is set to “0” otherwise. Such binarization is performed. In step S33, binarization of the image data is performed using the determined binarization method, and character recognition is performed using the obtained binarized image data.

As described above, according to the third embodiment, according to the background density value specified by the user,
Since the color image data is binarized, binarized character recognition image data intended by the user can be generated. Fourth Embodiment In a fourth embodiment, a method for generating character recognition image data from color image data as shown in FIG. 18, that is, when characters are dark relative to the background, will be described.

The processing executed in the fourth embodiment will be described with reference to FIG.

FIG. 19 is a flowchart showing processing executed in the fourth embodiment of the present invention. First, step S4
1, if the pixel value is smaller than the first threshold value for the R, G, and B luminance components of the color image data to be processed, "1"
In other cases, binarization is performed so as to be “0”. In step S42, R of the color image data to be processed,
For the luminance components of G and B, binarization is performed such that the pixel value is “1” when the pixel value is larger than a second threshold (<first threshold), and “0” otherwise. In step S43,
The ratio of “1” and “0” of the binary image data obtained in step S41 or S42 is checked. Step S
At 44, based on the inspection result, the binarized image data obtained by one of the binarization in step S41 and the binarization in step S42 is used as the final binarization result, and the obtained binarization is performed. Character recognition is performed using the valued image data.

For example, in step S43, step S4
If the frequency of occurrence of “1” in the binarized image data obtained by binarization of “1” is low, the process proceeds to step S41 in step S44.
Is selected as the binarization result. On the other hand, if the appearance frequency of “1” is high, it means that the background portion has been extracted. In this case, since the background is darker than the threshold and the characters are bright, the process proceeds to step S42 in step S44.
Is selected as a binarization result.

In the binarization, only the luminance component of the color image may be extracted, or the logical sum of the binarization for each color component may be calculated as in the first embodiment. Needless to say, it does not matter.

As described above, according to the fourth embodiment, the color image data is binarized using each of the first threshold and the second threshold for distinguishing the background and the character.
By adopting one of the binarization results according to the binarization result, the information on the color image data can be maintained more faithfully.
Valued character recognition image data can be generated.

As described above, according to the first to fourth embodiments, color image data is separately stored for each color component.
By performing binarization and logically synthesizing the binarization result, an appropriate binarization is performed even on an image that has not been accurately performed by binarization of a luminance component, and characters are lost or cut off. Can be prevented.

Further, it is possible to generate binarized character-recognition image data for color character image data in which characters lighter and darker than the background color are mixed without damaging both characters. .

Further, the user designates whether there is a dark character on a light background or a bright character on a dark background, or automatically judges it, and binarizes it in consideration of the designation result or judgment result. , Binarized character recognition image data that is not affected by the background and the character density can be generated.

By generating image data for character recognition by the above processing, character recognition for color image data can be performed more accurately and stably.

Even if the present invention is applied to a system including a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), a device including one device (for example, a copying machine, a facsimile machine) Etc.).

It is another object of the present invention to provide a system or an apparatus with a storage medium storing a program code of software for realizing the functions of the above-described embodiments, and to provide a computer (or CPU) of the system or the apparatus.
And MPU) read and execute the program code stored in the storage medium.

In this case, the program code itself read from the storage medium realizes 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 instructions 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 processing realizes the functions of the above-described embodiments.

[0054]

As described above, according to the present invention,
An information processing apparatus and method capable of improving the character recognition accuracy for image data composed of a plurality of pieces of color information, and a computer readable memory can be provided.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of an information processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating an example of an execution window according to the first embodiment of the present invention.

FIG. 3 is a diagram illustrating an appearance frequency graph of an R component according to the first embodiment of the present invention.

FIG. 4 is a diagram illustrating an appearance frequency graph of a G component according to the first embodiment of the present invention.

FIG. 5 is a diagram illustrating an appearance frequency graph of a B component according to the first embodiment of the present invention.

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

FIG. 7 is a diagram illustrating image data of an R component of the image data of FIG. 6 according to the first embodiment of the present invention.

8 is a diagram illustrating image data of a G component of the image data of FIG. 6 according to the first embodiment of the present invention.

FIG. 9 is a diagram illustrating image data of a B component of the image data of FIG. 6 according to the first embodiment of the present invention.

FIG. 10 is a diagram illustrating binary image data obtained when conventional binarization is performed on the image data of FIG. 6 according to the first embodiment of the present invention.

11 is a diagram illustrating binary image data obtained when the image data of FIG. 6 according to the first embodiment of the present invention is subjected to binarization according to the first embodiment.

FIG. 12 is a flowchart illustrating processing executed in the first embodiment of the present invention.

FIG. 13 is a diagram illustrating an example of image data according to the second embodiment of the present invention.

FIG. 14 is a flowchart illustrating processing executed in a second embodiment of the present invention.

FIG. 15 is a diagram illustrating a density histogram of the image data in FIG. 13 according to the second embodiment of the present invention.

FIG. 16 is a diagram illustrating an example of image data according to the third embodiment of the present invention.

FIG. 17 is a flowchart illustrating processing executed in a third embodiment of the present invention.

FIG. 18 is a diagram illustrating an example of image data according to the fourth embodiment of the present invention.

FIG. 19 is a flowchart illustrating processing executed in a fourth embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 CPU 2 CD-ROM drive device 3 Mouse 4 Keyboard 5 Monitor 6 Hard disk drive device 7 Floppy drive 8 Communication device 9 Printer

Claims (13)

[Claims] 1. An information processing apparatus for generating image data for character recognition from input color image data, comprising: a binarizing means for binarizing image data of each color component of the color image data; Information comprising: a synthesizing unit that synthesizes binarized image data for each color component binarized by the binarizing unit; and an output unit that outputs the binarized image data synthesized by the synthesizing unit. Processing equipment. 2. The method according to claim 1, wherein the binarizing unit determines a threshold for binarizing the image data of each color component based on an appearance frequency of a pixel value of the image data of each color component of the color image data. And binarizing the image data of each color component based on a threshold determined by the determining unit.
An information processing apparatus according to claim 1. 3. The determining means determines a background color in the color image data, and sets a threshold for binarizing the image data of each color component based on the determined pixel value of the background color. The information processing apparatus according to claim 2, wherein the information is determined. 4. The image processing apparatus according to claim 1, further comprising a designation unit for designating a background color in the color image data, wherein the decision unit determines each color component of the color image data based on a pixel value of the background color designated by the designation unit. The information processing apparatus according to claim 3, wherein a threshold value for binarizing the image data is determined. 5. A binarizing unit, comprising: a first binarizing unit that binarizes image data of each color component of the color image data using a first threshold; and a color binarizing unit using a second threshold. 2. The information processing apparatus according to claim 1, further comprising a second binarizing unit that binarizes image data of each color component of the image data. 6. The first binarized image data obtained by synthesizing binarized image data for each color component binarized by the first binarizing unit, and the second binarized image data. Comparing means for comparing second binarized image data obtained by synthesizing the binarized image data for each color component binarized by the binarizing means, based on a comparison result of the comparing means, The information processing apparatus according to claim 5, wherein either one of the first binarized image data and the second binarized image data is output. 7. An information processing method for generating character recognition image data from input color image data, comprising: a binarization step of binarizing image data of each color component of the color image data; Information comprising: a combining step of combining binary image data for each color component that has been binarized in the value converting step; and an output step of outputting the binary image data combined in the combining step. Processing method. 8. The binarizing step includes: determining a threshold value for binarizing the image data of each color component based on an appearance frequency of a pixel value of the image data of each color component of the color image data. And binarizing the image data of each color component based on a threshold value determined in the determining step.
An information processing method according to claim 1. 9. The determining step includes determining a background color in the color image data, and setting a threshold value for binarizing the image data of each color component based on the determined pixel value of the background color. 9. The information processing method according to claim 8, wherein the information is determined. 10. The method according to claim 1, further comprising a designating step of designating a background color in the color image data, wherein the deciding step is based on a pixel value of the background color designated in the designating step. 10. The information processing method according to claim 9, wherein a threshold value for binarizing the image data is determined. 11. The binarizing step includes: a first binarizing step of binarizing image data of each color component of the color image data using a first threshold value; 8. The information processing method according to claim 7, further comprising a second binarizing step of binarizing image data of each color component of the image data. 12. The output step includes: first binarized image data obtained by synthesizing binarized image data for each color component binarized in the first binarization step; A comparing step of comparing second binarized image data obtained by synthesizing binarized image data of each color component binarized in the binarizing step, and based on a comparison result of the comparing step, The information processing method according to claim 11, wherein either one of the first binarized image data and the second binarized image data is output. 13. A computer-readable memory storing a program code for information processing for generating image data for character recognition from input color image data, wherein the computer-readable memory stores a program code for each color component of the color image data. Program code for a binarizing step to be performed, program code for a synthesizing step for synthesizing binarized image data for each color component binarized in the binarizing step, and a binarized image synthesized in the synthesizing step And a program code of an output step of outputting data.