JP2017175600A - Image processing device, image processing method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform image area separation with high accuracy. A predetermined feature element such as a ruled line or a background ruled line is extracted from input image data (S22, S32), and the condition that the feature element should satisfy for the predetermined feature element and in the image data With reference to the predetermined determination condition, the attribute of the region defined based on the characteristic element satisfying the condition is defined by the determination condition in the input image data based on the characteristic element and the determination condition. The attribute defined by the determination condition is set in the area (S24, S34). [Selection] Figure 6

Description

  The present invention relates to an image processing apparatus, an image processing method, and a program.

2. Description of the Related Art Conventionally, when processing such as compression is performed on an image, a technique for accurately reducing a file size by accurately determining a character area and increasing a compression rate for a document containing only characters is known. . Such a technique is described in Patent Document 1, for example.
Furthermore, a technique is known in which vertical and horizontal ruled lines are extracted from a form, cells are extracted based on the extracted ruled lines, and character recognition is performed for each cell. Such a technique is described in Patent Document 2, for example.

  In Patent Document 3, after detecting the background level of an image obtained by digitally scanning a document, a character edge of the image is detected using a plurality of parameters, and the plurality of parameters are changed according to the background level. The structure to perform is disclosed.

  However, when image area separation such as distinguishing a character area from others is performed by the conventional method described above, when applied to an image scanned from a paper document having a high background level such as a newspaper, the character area There is a problem that the area is erroneously detected as a picture area. The problem relating to the erroneous detection of the character area cannot be sufficiently solved even by using the technique of Patent Document 3.

  An object of the present invention is to solve such problems and to perform image area separation with high accuracy. The area to be detected in the image is not limited to the character area.

  In order to achieve the above object, an image processing apparatus according to the present invention includes an input unit for inputting image data, and an extraction unit for extracting predetermined feature elements from the image data input by the input unit. The extraction means with reference to a predetermined determination condition for the predetermined feature element, a condition to be satisfied by the feature element, and an attribute of a region defined on the basis of the feature element satisfying the condition in the image data Is provided with attribute setting means for setting an attribute defined by the determination condition in an area defined by the determination condition in the input image data based on the feature element extracted by .

  According to the above configuration, image area separation can be performed with high accuracy.

1 is a diagram showing a hardware configuration of an MFP which is an embodiment of an image processing apparatus according to the present invention. FIG. FIG. 2 is a diagram illustrating a functional configuration of an image processing unit 200 included in the MFP illustrated in FIG. 1. FIG. 3 is a diagram illustrating a functional configuration of a scanner correction unit 210 illustrated in FIG. 2. It is a figure which shows the function structure of the character area determination part 212 shown in FIG. 3 in detail. 3 is a basic flowchart of image processing by the image processing unit 200 shown in FIG. 2. It is a flowchart of the more detailed process sequence of character area determination (S11) shown in FIG. FIG. 7 is a flowchart continued from FIG. 6. FIG. 10 is a diagram for describing an example in which ruled line region setting processing is applied to image data obtained by reading a paper surface. It is another figure. It is another figure. It is a figure for demonstrating the example which applies the process of the area | region setting about a background ruled line with respect to the same image data as FIG. 8A. It is another figure. It is a figure for demonstrating the process of the area | region setting about a specific character. It is a figure for demonstrating the method of extracting a background ruled line using the determination conditions by a ruled line. It is another figure. It is a figure for demonstrating the method of extracting the candidate of a background ruled line from the result of a background extraction. It is another figure. It is a figure for demonstrating the method of totaling the setting result of the area | region based on a some feature element, and finally determining a character area.

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 shows a hardware configuration of an MFP (digital multifunction peripheral) which is an embodiment of an image processing apparatus according to the present invention.
An MFP 100 shown in FIG. 1 is an image processing apparatus having functions such as scanning, printing, copying, facsimile communication, and document storage.
As illustrated in FIG. 1, the MFP 100 includes a CPU 101, a ROM 102, a RAM 103, an HDD (hard disk drive) 104, a communication I / F (interface) 105, an operation unit 106, a display unit 107, and an engine I / F 108. These are connected via the system bus 109 to constitute the MFP 100. In addition, a scanner 120, a plotter 130, and an image processing unit 200 are connected to the engine I / F 108.

Then, the CPU 101 controls the entire MFP 100 and implements various functions by executing necessary programs stored in the ROM 102 or the HDD 104 using the RAM 103 as a work area.
The ROM 102 and the HDD 104 are non-volatile storage media (storage means) and store various programs executed by the CPU 101 and various data described later.
The communication I / F 105 is an interface for communicating with an external device. What is necessary is just to provide the thing according to the standard of the communication channel to be used.

The operation unit 106 is an operation means for accepting an operation from a user, and includes various keys, buttons, a touch panel, and the like.
Display unit 107 is a presentation unit for presenting the operating state of MFP 100, setting contents, a message, and the like to the user, and includes a liquid crystal display, a lamp, and the like.

  Note that the operation unit 106 and the display unit 107 may be externally attached. Further, it is not always necessary for MFP 100 to directly receive an operation from the user. The MFP 100 may accept an operation by an external device connected via the communication I / F 105 or may present information to the external device. In that case, the operation unit 106 and the display unit 107 may not be provided.

The engine I / F 108 is an interface for connecting the scanner 120, the plotter 130, and the image processing unit 200 to the system bus 109 so that the CPU 101 can control them.
The scanner 120 is an image reading device that reads an image of a document and outputs the image data, and sends the read image data to the image processing unit 200.
The image processing unit 200 is an image processing unit that determines a document type of a document read by the scanner 120, selects an image resolution compression rate suitable for the document according to the document type, and converts image data according to the compression rate. Compress and output compressed image data.
The plotter 130 is an image forming unit that forms an image on a sheet according to the image data, and can form an image based on the compressed image data output from the image processing unit 200.

FIG. 2 shows a functional configuration of the image processing unit 200. The functions of the units shown in FIG. 2 may be realized by dedicated hardware, realized by causing a processor to execute software, or a combination thereof. It is also conceivable that the CPU 101 assumes some functions.
As shown in FIG. 2, the image processing unit 200 includes a scanner correction unit 210, a compression processing unit 220, a data interface unit 230, an expansion processing unit 240, and a printer correction unit 250.

Among these, the scanner correction unit 210 has a function of classifying image data read by the scanner 120 and performing image processing. Details of the function will be described later with reference to FIG.
The compression processing unit 220 has a function of compressing the image data corrected by the scanner correction unit 210 for internal processing.
The data interface unit 230 is an HDD management interface used when the image data compressed by the compression processing unit 220 is temporarily stored in the HDD 104.

The decompression processing unit 240 has a function of decompressing the compressed image data for internal processing for use in image formation by the plotter 130.
The printer correction unit 250 has a function of sending the image data expanded by the expansion processing unit 240 to the plotter 130 after correcting it as necessary.

Next, FIG. 3 shows a functional configuration of the scanner correction unit 210 shown in FIG.
As shown in FIG. 3, the scanner correction unit 210 includes a document type determination unit 211, a character region determination unit 212, an image region separation unit 213, a scanner γ unit 214, a filter processing unit 215, a color correction processing unit 216, and a character γ. A portion 217 is provided.
Among these, the document type determination unit 211 determines the image data of a document with only image data to be processed from the characteristics of a document having characters, the characteristics of a chromatic document, the characteristics of a photographic paper photograph, and the characteristics of a printed photograph. Or a color original image data.

The character area determination unit 212 has a function of determining the character area in the image indicated by the image data from the features of the image, in order to determine image data that cannot be determined by the document type determination unit 211, or to prevent erroneous determination. Plays the role of re-determination.
The image area separation unit 213 has a function of separating the image data to be processed into a black character edge area, a color character edge area, and other three areas while referring to the determination result of the character area by the character area determination unit 212. Yes.

The scanner γ unit 214 has a function of converting image data from reflectance linear to density linear data.
The filter processing unit 215 has a function of performing different filter processing on three regions separated by the image region separation unit 213 in the image data to be processed while switching a plurality of filters.

The color correction processing unit 216 is means for converting RGB image data into CMYK image data outside the black character edge region.
The character γ part 217 has a function of emphasizing a character by raising γ with respect to a color character and a black character in the character area. The image data processed by the character γ unit 217 is supplied to the compression processing unit 220.

Next, FIG. 4 shows the functional configuration of the character area determination unit 212 shown in FIG. 3 in more detail.
The character region determination unit 212 includes a ruled line region determination unit 261, a background ruled line region determination unit 262, a specific character region determination unit 263, a determination condition acquisition unit 264, a determination condition 265, a range extraction unit 266, and a determination result setting unit 267. Is done.
The ruled line area determination unit 261 has a function of setting an area and an attribute for image data based on a ruled line when an item related to a ruled line exists in the determination condition acquired by the determination condition acquiring unit 264.

  The ruled line area determination unit 261 extracts, for example, vertical ruled lines and horizontal ruled lines from image data read from form data or a form printed on paper, and calculates parameters for comparing each extracted ruled line with a determination condition. To do. Then, the parameter of each ruled line extracted is compared with the determination condition, and the setting of the region and the attribute are set according to the determination condition for each matched ruled line. As the attributes set in each area, a character area, a picture area, or the like can be considered.

  The background ruled line region determination unit 262 has a function of setting an area and an attribute for image data based on the background ruled line when an item related to the background ruled line exists in the determination condition acquired by the determination condition acquisition unit 264. The background ruled line is a ruled line that uses a background space instead of a line.

  The background ruled line area determination unit 262 extracts a background area from image data read from, for example, form data or a form printed on paper, and calculates a parameter for comparing the extracted background area with a determination condition. From the beginning, only a long and narrow area that can be a background ruled line candidate may be extracted. Then, the parameters of the extracted skin regions are compared with the determination conditions, and the matching background regions are set as the background ruled lines, and the settings and attributes of the regions are set according to the determination conditions for the local skin ruled lines. A method for extracting the background area will be described later with reference to FIGS. 12A and 12B.

The specific character area determination unit 263 has a function of setting an area and an attribute for image data based on a specific character when an item related to the specific character exists in the determination condition acquired by the determination condition acquisition unit 264.
The specific character area determination unit 263 searches for, for example, whether there is a specific character that matches the determination condition in the image data read from the form data or the form printed on paper. Then, for each extracted specific character, the setting of the region and the attribute are set according to the determination condition.

The determination condition acquisition unit 264 has a function of reading and acquiring information registered as the determination condition 265.
The determination condition 265 defines various conditions for determining various regions including a character region based on the characteristics of the document (image data) that is likely to be erroneously detected.
The determination is made based on feature elements such as a ruled line, a background ruled line, and a specific character that serve as a reference for the positions of areas such as a character area and a picture area. Although a specific example will be described later, the determination condition 265 is data that defines which attribute region is set at which position based on the condition that the feature element should satisfy and the feature element that satisfies the condition in the image data. is there.

Based on the results set by the ruled line region determining unit 261, the background ruled line region determining unit 262, and the specific character region determining unit 263, the range extracting unit 266 determines the range of various attribute regions such as character regions and design regions in the image data. Confirm.
The determination result setting unit 267 is a means for setting a compression rate different from that of the other portions for the range determined as the character area in the image data.
Note that whether or not the ruled line region determination unit 261, the background ruled line region determination unit 262, and the specific character region determination unit 263 actually perform the determination depends on which characteristic element is held as the determination condition 265. .

Next, Tables 1 to 4 show some specific examples of the determination condition 265.
The determination condition 265 includes a feature element category list shown in Table 1 and a detailed table of each category shown in Tables 2 to 4.

  Each table in Tables 1 to 4 can be prioritized among items in the table. As an example, when it is desired to perform the determination regarding the background ruled line before the determination regarding the ruled line, the characteristic elements in the table of Table 1 are rearranged in the desired execution order so that the determination regarding each characteristic element can be performed in the execution order. Can be considered. The same applies to the determination conditions in Tables 2 to 4. In Tables 1 to 4, ID is a sequential number including priority.

  Table 1 is a table composed of feature element category IDs and information on types of feature elements constituting the category. It is also conceivable to use types other than the ruled lines, specific characters, and background ruled lines in Table 1. In this case, the character region determination unit 212 is provided with a determination unit corresponding to the type of feature element.

Tables 2 to 4 are tables in which details of determination conditions regarding the feature elements of each category defined in Table 1 are registered. As many detailed tables as the number of categories defined in Table 1 are created. Here, Tables 2 to 4 are provided corresponding to the three categories defined in Table 1.
In the detailed tables in Tables 2 to 4, one line of data indicates the contents of one determination condition. Each determination condition includes information of “ID”, “element condition”, “range”, and “determination result”.

Among these, the “element condition” is a condition in which a characteristic element or a condition to be satisfied by a relationship between the characteristic element and another element is defined, and the number and contents differ depending on the category of the characteristic element.
“Range” indicates a position where an area is set based on a feature element. “Determination result” indicates an attribute to be set in the set area. In the examples of Tables 2 to 4, the determination result is only a character (character area), but it is needless to say that other attributes may be set.

The table in Table 2 is a table of determination conditions related to the ruled line with ID = A in the table in Table 1. For ruled lines, orientation, line color, thickness, range color, and interval are provided as element conditions.
Among these, “direction” is a condition of the direction of the ruled line, and any one of a vertical line, a horizontal line, and Don't care (arbitrary) is entered. “Line color” is a ruled line color condition and is shown as black in Table 2, but may be color information such as RGB ratio or amount. “Thickness” is a condition of the thickness of the ruled line, and is shown in millimeters (mm) in the figure, but an appropriate unit such as inch or pixel can be used. Even if it is not just the thickness, it may be within a predetermined error range.

  The “range color” is assumed that the corresponding feature element satisfies the element condition or may include color only when the area set in accordance with the “range” condition based on the corresponding feature element is monochrome. Is specified. The “interval” is an interval between ruled lines, and is a condition for an interval between the extracted adjacent ruled lines. The “range color” and “interval” are not conditions related to the characteristics of the feature elements themselves.

  For example, in the determination condition of ID = A_001, a certain ruled line is a vertical black ruled line having a thickness of 0.1 mm and sandwiched between adjacent ruled lines (based on the “range” condition). In the case of black and white, the element condition is satisfied regardless of the interval between adjacent ruled lines. In this case, an area is set between the ruled line that satisfies the element condition and the adjacent ruled line (which also needs to satisfy the element condition of the same ID), and the attribute of the area is set to “ It is specified that it should be set to "character".

Next, the table of Table 3 is a table of determination conditions regarding the specific character of ID = B in the table of Table 1. For specific characters, element conditions include symbols, range colors, combinations, symbols included, or orientations.
“Symbol” indicates a specific character (may be a character string) used in the determination, and is indicated by image data, character code, font, size (width and height), and the like. The “range color” is the same as in Table 2. “Combination” is an item for specifying an ID when it is desired to define “range” using a feature element that satisfies an element condition of another ID. A plurality of IDs may be specified.
“Is symbol included” indicates whether or not “range” is set to include a specific character itself. “Direction” indicates the direction of the character string used in the determination, and designates either horizontal or vertical.
Of these, conditions other than “symbol” and “orientation” are not conditions relating to the characteristics of the characteristic elements themselves.

  For example, under the determination condition of ID = B_001, a specific character “Å” is arranged horizontally, includes the character “Å”, and from that character toward the arrangement direction of the character string, the next ID = A_001 If the area up to the ruled line (based on the “range” condition) is black and white, the character “Å” satisfies the element condition. In this case, it is stipulated that an area based on the above "range" condition should be set and the attribute of the area should be set to "character".

  Next, the table of Table 4 is a table of determination conditions regarding the background ruled line of ID = C in the table of Table 1. For the background ruled line, orientation, thickness, range color, and interval are defined as element conditions. The determination conditions related to the background ruled lines are the same as the determination conditions related to the ruled lines in Table 2 except that the element condition does not include “color”.

Next, a procedure of image processing including setting of areas and attributes according to the determination conditions described above by the image processing unit 200 will be described using a flowchart.
FIG. 5 shows a basic flow of the image processing.
When the image processing unit 200 acquires image data from the scanner 120, the image processing unit 200 starts processing shown in the flowchart of FIG. The process related to the acquisition of the image data is an input procedure process. This also corresponds to the function of the input means.

  First, the character area determination unit 212 of the scanner correction unit 210 determines a character area in the image data and other areas (S11). Details of the character region determination in step S11 will be described later with reference to FIGS. Next, each unit from the image area separation unit 213 to the character γ unit 217 of the scanner correction unit 210 sequentially performs image processing for the scanner image on the image data acquired from the scanner 120 (S12).

  Next, in accordance with the determination by the character area determination unit 212, the compression processing unit 220 performs compression processing on the image data acquired from the scanner 120 by an appropriate method for each of the character area and other areas (for example, the pattern area). (S13). Thereafter, the image processing unit 200 stores the compressed image data in the HDD 104 through the data interface unit 230 (S14).

Thereafter, when the plotter 130 prints out, the decompression processing unit 240 decompresses the image stored in step S14 (S15). Next, the printer correction unit 250 performs image processing according to the characteristics of the plotter 130 on the decompressed image data, and performs image processing that emphasizes characters in the character area according to the determination result of step S11. Then, the processed image data is output (S16).
Thus, the process of FIG. 5 ends. The reason why special image processing is performed on the character in step S16 is that the end of the character is thinned by performing the processing in steps S12 to S15, or the back-off occurs in a document having a high background level. This is to improve the difficulty in reading characters.

Next, FIG. 6 shows a flowchart of a more detailed processing procedure for character region determination in step S11 of FIG. The processing shown in this flowchart corresponds to the function of the character area determination unit 212 shown in FIG.
In the process of FIG. 6, first, the character region determination unit 212 refers to the category list (Table 1) of the determination condition 265 acquired by the determination condition acquisition unit 264, and whether or not the determination condition 265 has a rule line condition. Judgment is made (S21). Here, if there is any ruled line area, the ruled line area determination unit 261 extracts all the ruled lines from the image data (S22) and sequentially processes the determination conditions regarding the ruled lines (see Table 2). The process of steps S23 to S25 is repeated while targeting.

That is, the ruled line region determination unit 261 first acquires a ruled line extracted in step S22 that meets the element condition in the determination condition to be processed (S23). At this time, the ruled line region determination unit 261 also holds information on ruled lines that have not been acquired.
Then, the ruled line area determination unit 261 sets an area based on the ruled line and its attribute in the image data according to the ruled line acquired in step S23 and the determination condition of the processing target (S24). For example, when the determination condition of ID = A_001 is used, an area is set in a portion sandwiched between ruled lines that satisfy the element condition, and the attribute of the area is set to “character”. Finally, a ruled line that matches the element condition is stored in association with the determination condition to be processed (S25). This storage is performed for reference in processing related to another feature element.

  Further, after the end of the loop of steps S22 to S25, or if NO in step S21, the process proceeds to step S31. Here, the character region determination unit 212 determines whether or not the determination condition 265 acquired by the determination condition acquisition unit 264 includes a condition related to the background ruled line (S31). In the example of FIG. 6, unlike the data in Table 1, the determination regarding the background ruled line is performed second.

  If Yes in step S31, the background ruled line region determining unit 262 extracts all the background ruled lines from the image data in order to examine the setting of the area for the background ruled lines (S32), and each determination condition (table 4) is sequentially processed, and the processes in steps S33 to S35 are repeated.

That is, the background ruled line region determination unit 262 first acquires a ruled line extracted in step S32 that meets the element condition in the determination condition to be processed (S33). At this time, the background ruled line region determination unit 262 also holds information on the background ruled lines that have not been acquired.
Then, the background ruled line region determination unit 262 sets a region based on the background ruled line and its attribute in the image data according to the background ruled line acquired in step S33 and the determination condition of the processing target (S34). For example, when the determination condition of ID = C_001 is used, an area is set in a portion sandwiched between background ruled lines that satisfy the element condition, and the attribute of the area is set to “character”. Finally, a background ruled line that matches the element condition is stored in association with the determination condition to be processed (S35). This storage is performed for reference in processing related to another feature element.

Further, after the end of the loop of steps S32 to S35, or if NO in step S31, the process proceeds to step S41 of FIG. Here, the character region determination unit 212 determines whether or not the determination condition 265 acquired by the determination condition acquisition unit 264 includes a condition related to a specific character (S41).
If Yes in step S41, the specific character area determination unit 263 sequentially sets each determination condition (see Table 3) regarding the specific character as a processing target in order to examine setting of the area for the specific character. Repeat the process. In the case of a specific character, since the condition for extracting the specific character is included in the determination condition, the specific character is extracted in the loop unlike the case of the ruled line or the background ruled line.

In the loop processing, the specific character area determination unit 263 first generates specific character data for search according to the designation of “symbol” and “direction” from the determination condition to be processed (S42). Next, the specific character area determination unit 263 extracts all specific characters from the image data to be processed in accordance with the generated specific character data (S43).
Then, among the extracted specific characters, a character that matches the element condition in the determination condition to be processed is acquired (S44). At this time, the specific character area determination unit 263 also holds information on specific characters that have not been acquired. In addition, the specific character area determination unit 263 stores the feature elements stored in association with the determination condition specified as “combination” in the determination condition to be processed (here, the ruled line stored in step S25 or the step S35). The obtained background ruled line) is acquired (S45).

  Next, the specific character area determination unit 263 includes, in the image data, the area based on the specific character and the specific character acquired in step S44, the feature element acquired in step S45, and the determination target condition. The attribute is set (S46). For example, when the determination condition of ID = B_001 is used, an area is set in a range from a specific character that satisfies the element condition to the next ruled line of ID = A_001 in the character arrangement direction, and the attribute of the area is set to “character”. Set to.

After the end of the loop of steps S42 to S46, or when NO in step S41, the range extraction unit 266 combines the areas set based on the ruled lines, the background ruled lines, and the specific characters, and the characters in the image data. The attribute area is determined (S47). Although the character attribute is used here, other attributes may be used.
The process in FIG. 7 is thus completed. Details of the process in step S47 will be described later with reference to FIG.

In the processing of FIGS. 6 and 7 described above, the processing of steps S22, S32, and S43 is the processing of the extraction procedure, and corresponds to the function of the extraction means. The processing of steps S24, S34, S46, and S47 is processing of the attribute setting procedure, and corresponds to the function of the attribute setting means.
Note that the character area determination unit 212 in this embodiment is a feature element that considers the extraction of feature elements for a plurality of feature elements by the extraction unit and the description attribute setting by the attribute setting unit as described above. It is performed while selecting in a predetermined order.

Next, an example in which the processing of FIGS. 6 and 7 is applied to specific image data will be described.
First, FIGS. 8A to 8C show an example in which the region setting processing for the ruled lines in steps S22 to S25 is applied to image data obtained by reading a paper surface.
FIG. 8A is a simplified representation of image data in the initial state. Thick lines 601 to 605 in the figure indicate ruled lines, and thin lines 610 in the figure simply represent characters.
FIG. 8B shows a state in which ruled lines that satisfy the element conditions are extracted from the thick lines 601 to 605. Assuming that the interval of one stage is 30 mm and the thickness of each ruled line is 0.1 mm, the ruled line is satisfied as a ruled line that satisfies the element condition by the processing in steps S22 and S23 of FIG. 701 to 704 are extracted.

FIG. 8C shows regions set according to the ruled lines. In the process of step S24 in FIG. 6, since areas are set in the portion sandwiched between the extracted ruled lines, areas 706 to 708 are set.
FIG. 8B also shows a ruled line 705 extracted when ID = A_001 in Table 2 is a processing target. However, the ruled line extracted under this condition is only the ruled line 705, and there is no portion between the plurality of ruled lines, so no area is set.

Next, FIG. 9A and FIG. 9B show an example in which the region setting processing for the background ruled lines in steps S32 to S35 of FIG. 6 is applied to the same image data as FIG. 8A.
FIG. 9A shows a state in which the background satisfying the element condition is extracted therefrom. As shown in FIG. 9A, as the background area in the image data, for example, areas indicated by reference numerals 801 to 804 can be extracted. However, for example, the background areas satisfying the element ID = C — 002 in Table 4 are only the background ruled lines 801 and 802, and these are extracted as the background ruled lines.

  FIG. 9B shows areas set according to the ruled lines. In the process of step S34, an area 811 is set because an area is set in a portion sandwiched between the extracted background ruled lines. Note that the region 812 is not sandwiched between the background ruled lines in FIG. 9A. However, if A_002 is defined as “combination” in the element condition so that an area can be set even in a portion sandwiched between the ruled line 703 and the background ruled line 802, the area 812 can also be set. The same applies to the “range” as “the part in contact with the background ruled line”.

Next, FIG. 10 shows an example in which the region setting process for specific characters in steps S42 to S46 in FIG. 7 is applied. Here, the determination condition of ID = B_001 in Table 3 will be described.
In this determination condition, since the ruled line A_001 is designated as “combination”, an area is set using the ruled lines 901 and 902 that satisfy the element condition A_001.
The specific character area determination unit 263 in FIG. 4 generates image data of specific characters from “横” of “symbol” and “horizontal” of “direction” in the element condition, and the image data generated in the image data Search for a match. As a result, the character “Å” indicated by reference numeral 95 is found. Then, a region sandwiched between the position of the character “と” and the next ruled line 902 is set as a region 903.
Here, since it takes time to search for matching character data from the image data, the position of the specific character with respect to the ruled line defined in “Combination” is defined, and only at the defined position. It is better to search for specific characters.

  Next, with reference to FIGS. 11A and 11B, a method for extracting the background ruled line using the determination condition based on the ruled line when the document includes the background ruled line and there is no condition for determining the background ruled line will be described. This process corresponds to a modified example of the background extraction in the processes of FIGS.

In FIG. 11A, ruled lines 701 to 704 are the same as those extracted in FIG. 8B. In this case, there are often ground ruled lines on the extended lines where the ruled lines are cut off in the middle of the image, such as the ruled lines 701 and 702. Therefore, by performing background detection on the extension lines 1001 and 1002 and applying a determination condition related to ruled lines to the detected background area, it is possible to extract a background ruled line area as a basis for setting the area.
Here, it is assumed that ruled lines 1003 and 1004 are extracted on the extended lines 1001 and 1002 as shown in FIG. 11B, and areas 1006 and 1007 sandwiched between the ruled lines are set according to the determination conditions regarding the ruled lines.

Next, a method for extracting background ruled line candidates from the background extraction result will be described with reference to FIGS. 12A and 12B.
In addition, it is possible to specify whether a character is written vertically or horizontally in image data by appropriately applying a known technique. In this specified state, as shown in FIG. 12A, histograms 1102 and 1103 of black portions in the vertical direction and horizontal direction of the image data 1101 of the original are acquired. Since the image data 1101 is written vertically, the line spacing of the histogram 1103 is acquired. However, the line spacing is not a candidate for the background ruled line. The background in the image data 1101 is extracted from the portions corresponding to the portions 1131 and 1132 having a width equal to or greater than the line spacing and the low histogram value, and as shown in FIG. 12B, regions 1133 and 1134 where the background is continuous are extracted as candidates for the background ruled line. it can.

Further, the background in the image data 1101 of the low values 1121, 1122, and 1123 can be extracted from the histogram 1102, and the regions 1124, 1125, and 1126 in which the background is continuous can be extracted as candidates for the background ruled line.
Here, the criteria 1104 and 1105 for determining whether or not the value of the histogram is low are set in advance in the system.

Next, with reference to FIG. 13, a method of finally summing up the area setting results based on a plurality of feature elements and finalizing the character area will be described.
Assume that the setting result of the character region based on the ruled line is 1201, the setting result of the character region based on the background ruled line is 1202, and the setting result of the character region based on the specific character is 1203.
The state where 1201 and 1202 are combined is 1204, and the state where 1203 is combined with 1204 is 1205. In order to explain the overlapping portion in 1205, one line is picked up, and the portion where the character attribute area is set based on each feature element is indicated by a frame and “1” (1211-1213). In addition, as a size standard for one line, 1214 indicates data that is all set to “0”.

The results 1211 to 1213 are added for each coordinate (1215 and 1216), and the addition result can be used for final character region determination.
1215 is an example in which 2 is used as the threshold value, and an area having an addition result of 2 or more is a character area. In this case, the framed range is determined as the character area.
Reference numeral 1216 denotes a case where the priority of the setting result of the character area based on the ruled line is increased. Here, the frame-enclosed portion in which the addition result is an area of 2 or more and the determination result of 1213 is “1” is determined as the character area.

As described above, the character area determination unit 212 according to this embodiment performs the feature element extraction by the extraction unit and the attribute setting for the input image data with reference to the determination condition by the attribute setting unit for each of the plurality of feature elements. Do. The attribute setting means aggregates the attributes set according to the plurality of feature elements for the plurality of feature elements, and sets the attributes of each part in the input image data.
Furthermore, when there are many determination methods, the determination result of the method with high priority is multiplied by a large number corresponding to the priority, and the sum of the multiplied values is a reference such as 2 when 1215 The character area can be determined depending on whether or not the value is greater than or equal to the value. In any case, the threshold and priority are set in advance.

This is the end of the description of the embodiment. In this invention, the specific configuration of the apparatus, the specific processing procedure, the data format, the specific data content, the content and number of determination conditions, items, etc. The invention is not limited to that described in the embodiment.
For example, when a feature related to a character region is acquired from a plurality of feature element extractions and handled, the final character region is determined by considering the priority of the feature and conditioning the character region based on multiple types of results. It is possible to do. This enhances extensibility even when the character region determination is performed in a processing step other than the above-described embodiment. Further, if a UI is provided so that the user can perform conditioning, there is an effect that convenience can be improved.

In addition, when the characteristics related to the character region are acquired from the ruled line extraction, it may be determined that the ruled line to be determined is, for example, a newspaper ruled line. This makes it possible to place a condition on the discrimination attribute of the attribute information, specify the characteristics of the ruled line extracted by ruled line extraction, and improve the accuracy.
In addition, when a feature related to a character region is acquired from background ruled line extraction, an extension line of ruled line extraction is assumed, and it is preferable to determine whether or not the character range can be determined from image data on the extension line. In this way, the accuracy of character area determination can be maintained even in places where there are no ruled lines.

In addition, when a feature related to a character area is acquired from specific character extraction, the character area may be determined based on font data. In this way, it is possible to determine a special character that was easily erroneously determined by line extraction as a character area, and thus it is possible to improve accuracy.
In addition, when performing extraction based on a plurality of feature elements, the user may be able to determine the priority of which extraction is to be performed first. By determining the extraction order, it is possible to determine the character area and the pattern area intended by the user.

In addition, when the determination condition is held as information including the extraction range and the determination result, the extraction range and the determination result are acquired together with the determination condition, and the extraction range and the determination result are determined based on the acquired information. . As a result, it is possible to determine the extraction range according to the number of items of extraction criteria and hold the determination result.
In addition, when the determination condition is held as information including the image processing control method, the control method may be acquired together with the determination condition, and control may be performed based on the acquired information. This makes it possible to change the control according to the number of extraction criteria items.

Further, it is preferable to hold an extraction standard for extracting the type of document and to determine the type of document such as a character document or a photo document as a feature of the document. If the type of the original can be discriminated, it is possible to shorten the compression processing time for the original having a high character ratio and to perform image processing according to the original type.
In addition, when the characteristics of characters and patterns are distinguished from each other as image characteristics, it is preferable to switch the image processing between the character part and the picture part. This makes it easy to read the character by making the end of the character portion darker or the character itself darker for the image data where the end portion of the character becomes thin and difficult to read.

  Further, when the image characteristics are the original characteristics, the image processing may be switched according to the original characteristics. As a result, a manuscript with a high background density and a thin paper, such as a newspaper or an advertisement, is used, but when the background is too dark or double-sided printing is carried out, characters are difficult to read. For image data obtained by reading a document having such characteristics, it is possible to make the text easier to read by increasing the background correction or removing the set-off.

The embodiment of the program of the present invention is a program for causing a computer to control required hardware to realize all or part of the functions of MFP 100 in the above-described embodiment.
Such a program may be stored in a ROM or other nonvolatile storage medium (flash memory, EEPROM, etc.) provided in the computer from the beginning. However, it can also be provided by being recorded on an arbitrary nonvolatile recording medium such as a memory card, CD, DVD, or Blu-ray disc. Each procedure described above can be executed by installing the program recorded in the recording medium in a computer and executing the program.
Furthermore, it is also possible to download from an external device that is connected to a network and includes a recording medium that records the program, or an external device that stores the program in a storage unit, and install and execute the program on a computer.

  In addition, the configurations of the embodiments and the modified examples described above can be implemented in any combination as long as they do not contradict each other, and it is needless to say that only a part can be taken out.

100: MFP, 101: CPU, 102: ROM, 103: RAM, 104: HDD, 105: Communication I / F, 106: Operation unit, 107: Display unit, 108: Engine I / F, 120: Scanner, 130: Plotter, 200: Image processing unit, 210: Scanner correction unit, 220: Compression processing unit, 230: Data interface unit, 240: Decompression processing unit, 250: Printer correction unit, 261: Ruled line area determination unit, 262: Background ruled line area Determination unit, 263: specific character region determination unit, 264: determination condition acquisition unit, 265: determination condition, 266: range extraction unit, 267: determination result setting unit

JP 2007-189275 A Japanese Patent No. 3898645 Japanese Patent No. 3251119

Claims (10)

Input means for inputting image data;
Extraction means for extracting predetermined feature elements from the image data input by the input means;
With respect to the predetermined feature element, the extraction means refers to a determination condition that predetermines a condition that the feature element should satisfy and an attribute of a region that is defined based on the feature element that satisfies the condition in the image data. Image processing comprising: attribute setting means for setting an attribute defined by the determination condition in an area defined by the determination condition in the input image data based on the extracted feature element and the determination condition apparatus. The image processing apparatus according to claim 1,
The extraction of feature elements by the extraction unit and the setting of attributes for the input image data with reference to the determination condition by the attribute setting unit are performed for a plurality of feature elements, respectively.
The attribute setting means sets the attribute of each part in the input image data by aggregating the attributes set according to the plurality of feature elements for the plurality of feature elements. . The image processing apparatus according to claim 2,
Image processing characterized in that the extraction of the feature elements for the plurality of feature elements by the extraction means and the setting of the attributes by the attribute setting means are performed while selecting the feature elements to be considered in a predetermined order apparatus. An image processing apparatus according to any one of claims 1 to 3,
One of the predetermined characteristic elements is a ruled line. An image processing apparatus according to any one of claims 1 to 3,
One of the predetermined feature elements is a specific character, and a range in which the extraction unit searches for the specific character in the input image data is determined based on a position of another extracted feature element. An image processing apparatus characterized by being determined according to a determination condition. The image processing apparatus according to claim 2 or 3,
The predetermined characteristic element includes a ruled line and a background ruled line, and when the extracting unit extracts a ruled line from the input image data, the extraction unit tries to extract a background ruled line for an area on the extension of the extracted ruled line. An image processing apparatus. The image processing apparatus according to any one of claims 1 to 6,
An image processing apparatus comprising image processing means for processing each part of the input image data by a processing method according to the attribute set by the attribute setting means for the part. The image processing apparatus according to claim 7,
The processing for the image data performed by the image processing means includes processing for emphasizing a character for a portion in which a character region attribute is set. Input procedure to input image data,
An extraction procedure for extracting predetermined feature elements from the image data input in the input procedure;
With respect to the predetermined feature element, with reference to a predetermined determination condition, a condition to be satisfied by the feature element and an attribute of a region defined on the basis of the feature element satisfying the condition in the image data are determined in the extraction procedure. Image processing comprising: an attribute setting procedure for setting an attribute defined by the determination condition in an area defined by the determination condition in the input image data based on the extracted feature element and the determination condition Method.   The program for functioning a computer as an image processing apparatus as described in any one of Claims 1 thru | or 8.