JPH11175742A - Character reader, layout information setting means, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily specify layout information even if a form shifts when an image is read by a scanner. SOLUTION: In this character reader which extracts a character pattern from an image according to attribute information and performs a recognizing process, an attribute palette where attribute information set for an area to be processed is registered and a layout palette for registering layout information are displayed. In the processing object area in the image, the attribute information is set by being selected out of the attribute palette. The layout information where the area and attribute information are related can be registered in an arbitrary place in the layout palette as a component for layout specification. The layout information registered in the layout palette can be specified for the image to be processed at the time of the recognizing process by selecting it out of the layout palette.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character reading apparatus, a layout information setting method, and a recording medium that extract and recognize a character pattern from an image of a document or the like based on information such as a layout and an area attribute.

[0002]

2. Description of the Related Art An optical character reading device is a device for reading an image of a document or the like by a scanner, extracting a character pattern from the image, executing a character recognition process, and converting the image into a character code string. When performing the character recognition process, the optical character reading device performs a layout (region information) indicating in which region in the image (document) the character is described, and a description of the character in the target region. Is required such as an area attribute indicating whether or not is described. Hereinafter, such information is referred to as layout information.

Conventionally, some optical character reading apparatuses are equipped with an automatic layout analyzing function for automatically extracting layout information by analyzing an image read from a document. However, the performance of the current automatic layout analysis is technically inadequate, and it is necessary to manually specify or correct the layout. Therefore, in order to execute the character recognition processing by the efficient optical character reading device, it is important how to easily designate or correct the layout information, that is, how easy to use the user interface.

Conventionally, when designating layout information in an optical character reading apparatus, the mainstream of a user interface is to display an image read from a document and draw a line representing an area to be processed in the image. It has become. Further, an optical character reading apparatus that can store a layout specified by drawing a line in an image in a file and execute character recognition processing by applying the layout to another document having the same layout has also been considered. .

[0005]

As described above, in the conventional optical character reading apparatus, a layout is set by drawing a line representing an area on a document image, and the layout is saved. By applying this method when processing is performed on another document having the same layout, it is not necessary to sequentially set the layout for each document.

However, in the conventional optical character reading apparatus, when the image of the document is read by the scanner, for example, if the document (paper) is displaced due to inclination or the like, the saved layout designation is not correctly applied. It has been necessary to manually change the layout specification by hand or to perform the same operation as when specifying the layout of another document.

[0007] In particular, when the object to be processed by the optical character reading apparatus is a form or the like and a plurality of forms of different types are arranged in a predetermined order and are treated as one set, each form is treated as a set. Different shifts may occur. In such a case, it is necessary to set the layout information for each of the different types of forms, and a large amount of work is required.

The present invention has been made in view of the above-mentioned circumstances, and a character reading apparatus, a layout information setting method, and a character reading apparatus capable of easily specifying layout information even when a sheet is misaligned when an image is read by a scanner. And a recording medium.

[0009]

SUMMARY OF THE INVENTION The present invention relates to a character reading device for extracting a character pattern from an image based on attribute information and executing a recognition process. Attribute information storage means for registering attribute information to be set, attribute palette display means for displaying an attribute palette for selecting attribute information registered in the attribute information storage means, and display by the attribute palette display means Image display means for displaying an image to be processed together with the attribute information thus set, and at least one area to be processed from the image displayed by the image display means is set, and for each of the set areas Layout setting means for associating attribute information designated from among those displayed by the attribute palette display means, and the layout setting means Accordingly, it characterized associated area and the attribute information that includes a layout information storing means for registering the layout information by.

According to such a configuration, the attribute palette in which the attribute information is registered is displayed together with the image to be processed, and the attribute information is designated from the attribute palette to the area to be recognized. Can be. Therefore, by specifying a common attribute for a plurality of areas to be processed, the result of the recognition processing can be classified and output according to the attribute.

[0011] Also, a layout palette for selecting layout information registered by the layout information storage means is displayed together with the image displayed by the image display means and the attribute palette displayed by the attribute palette display means. Layout palette display means, and layout information selected from layout information displayed by the layout palette display means,
Layout designation means for setting a specific area of the image displayed by the image display means,
A recognition process is performed based on the layout information set by the layout designating means.

Thus, the layout information set for the image to be processed using the attribute palette (the area to be processed, the attribute information set for the area) is used to specify the layout. Registered as
It can be displayed as a layout palette for use in layout specification.

The layout storage means stores a plurality of layout information hierarchically, and the layout palette display means displays a layout palette in which layout information is arranged according to a hierarchical structure of the layout storage means. It is characterized by.

Thus, when designating a layout for an image to be processed, layout information to be used can be immediately selected. Also, the layout designating means may drag an item representing layout information displayed by the layout palette display means to a specific area in the image displayed by the image display means in response to an instruction using a pointing device. , And the layout information is set by being dropped.

[0015] Thus, the layout can be designated using the layout palette only by dragging and dropping the pointing device, so that the number of operations is reduced and the operability is improved.

When an item representing layout information displayed by the layout palette display means is dragged, a pattern of an area indicated by the layout information of the item is displayed.

This makes it possible to visually confirm the layout information selected from the layout pallet for what kind of layout, and to facilitate application to an image to be processed.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a system configuration of the character reading device according to the present embodiment.
The character reading device performs character recognition and image cutout processing for each area to be read from an image read by optically scanning a sheet such as a document or a form.

The character reading apparatus is, for example, a CD-ROM,
It is realized by a computer which reads a program recorded on a recording medium such as a DVD or a magnetic disk and whose operation is controlled by the program.

As shown in FIG. 1, the character reading device according to the present embodiment includes a CPU 10, a memory 12,
4, keyboard 16, mouse 18, scanner device 20,
And a storage device 22.

The CPU 10 controls the entire apparatus, and realizes various functions by executing based on programs and data stored in the memory 12. In the present embodiment, by executing the layout registration program 12a1 stored in the memory 12, the layout information (layout (area information) and area attributes) necessary for the character recognition processing is executed.
Is executed, and by executing the character reading program 12a2, a character reading process for performing character recognition using the layout information set by the layout registering process is executed.

The memory 12 is a read-only or readable / writable memory device by the CPU 10 or the like, and stores programs and data. The program area provided in the memory 12 includes a layout registration program 12a
1. Various programs such as a character reading program 12a2 are stored. Also, the memory 12 stores the scanner device 2
Image data of an image read from 0 or an image read from the storage device 22 is stored as a target of the character recognition processing.

The display device 14 includes a liquid crystal display and a CR.
Display at T, layout registration processing,
And a user interface in the character reading process.

The keyboard 16 is for inputting data to be processed, commands for the apparatus, and the like. The mouse 18 is a pointing device for inputting data to be processed and instructions to the device, and the like. In each of the layout registration process and the character reading process, an operation using the user interface screen displayed on the display device 14 is performed. Used for

The scanner device 20 reads an image to be subjected to a character recognition process by optically scanning a sheet such as a document or a form. The storage device 22 is a device that stores various data, programs, and the like. Various data include
An image file 22a to be subjected to the character recognition processing, a layout palette 22b and an attribute palette 22c used in the layout registration processing and the character reading processing (details will be described later), and an output file 22d which is a processing result of the character recognition processing are included. included. The storage device 18
Can read a program recorded on a recording medium such as a hard disk, a CD-ROM, a DVD, or a floppy disk, and can be executed by a character reading device.

Next, the operation of this embodiment will be described. In the character reading device according to the present embodiment, layout information is created in advance before reading characters written on a paper such as a document or a form, and registered in a database called a layout palette. As the layout information, not only the layout information of the entire sheet but also the layout information of a partial area of the sheet can be registered as a component.

The componentized layout information is arranged manually on an image to be processed in combination with each component when reading characters. In this embodiment,
A user interface is provided that allows the user to place a component by simply dragging the component from the layout palette using a pointing device such as a mouse, and the component can be designated by a simple operation with a small number of steps. Therefore, even when there is a displacement when the image of the sheet is read by the scanner device 20, the layout information can be easily specified, and even if the sheet is a sheet having a complicated layout, such as a form. By combining parts, layout information can be easily set by the same operation.

First, the layout registration processing will be described with reference to the flowchart shown in FIG. When the execution of the layout registration process is instructed, the CPU 10 executes the layout registration program 12 a stored in the memory 12.
1 to start the layout registration process.

In the layout registration processing, first, an image of a sheet to be processed for layout setting is input to the scanner device 20.
Alternatively, the image file 22 stored in the storage device 22
a and is stored in the memory 12 (step A1). The CPU 10 causes the display device 14 to display the image to be processed in the layout setting window in the layout setting screen (step A).
2) Display windows for the layout palette and the attribute palette (step A3).

FIG. 3 shows an example of the layout setting screen. As shown in FIG. 3, the layout setting screen includes a layout setting window on the left side on which an image to be processed is displayed, a layout palette window on the upper right, and an attribute palette window on the lower right.

The layout pallet is a database for hierarchically classifying and registering componentized layout information. The hierarchical structure can be arbitrarily set by the user. In the example illustrated in FIG. 3, items (holders) of “profile”, “receipt”, and “invoice” are created, and the layout information is already registered for each item. The item “new layout” is provided when execution of new creation of layout information is instructed.

The attribute palette is a database for hierarchically registering attribute information set for an area to be processed, which is referred to when performing character recognition processing or image cutting processing. The hierarchical structure can be arbitrarily set by the user. As the attribute information, whether the processing target is “text” (character) or “image”
In the case of text, the type of character string to be entered is specified. The type of the character string can be specified by the content to be described, for example, “address”, “name”, “telephone number”, “age”, “zip code”, “company name”, and the like. In this case, for example, in the case of "address", kanji, kana, and numbers are set as target character types, and in the case of "telephone number", numbers are set. In addition,
Instead of the specification based on the content to be described, it is possible to arbitrarily specify a character type such as “Kanji”, “Hiragana”, “Katakana”, “Numeric”, “English character”. In the example shown in FIG.
Items such as “name”, “telephone number”, and “age” are provided, and each attribute can be selected by designating this item.

First, in order to register layout information,
When new registration is selected from a layout registration processing menu (not shown), the CPU 10 provides a “new layout” item for registering new layout information in the layout palette, as shown in FIG. .

Here, when the "new layout" item is dragged by operating the mouse 18, moved to an arbitrary item for which layout information is to be registered in the layout palette, and dropped, the CPU 10 sets the destination item. A blank item for registering layout information is provided in (Step A4). Here, when a character string representing the name of the item is input from the keyboard 16, the CPU 10
Sets the item by the input character string.

FIG. 4 shows an example in which items for registering new layout information are set. In the example shown in FIG. 4, the item of “new layout” is “profile”.
Is dragged and dropped to the item of "", and "profile B" is input as the name of the item.

Next, layout setting for the image displayed in the layout setting window is executed (step A5). Here, as shown in FIG. 4, the areas to be set for the layout setting include an area in which each of a name, an address, a telephone, and an age is entered, and an area for a photograph (image). .

In the layout setting, first, an area to be designated as an area attribute is set (step A51).
For example, in the screen in the layout setting window, an area (a rectangular area) is operated by operating the mouse 18.
Is specified by the positions of two diagonal points. The area specified here is an area where characters to be subjected to the character recognition processing are written or an area where an image to be stored is recorded. In addition, for a region where characters are written, a range in which the character recognition result can be classified for each predetermined item is specified (for example, in a region where a name is described, the entire character string of the name is included). range).

When the range of the area is specified, the CPU 10 stores the area information indicating the area range and displays the area by changing the display color of the specified area. In addition,
A correction may be made to the specification of the area range by operating the mouse 18. In general, in forms, etc.,
The area in which characters are to be written is separated by a rectangular frame or the like. Therefore, the frame position is detected in advance from the image to be processed, and it is determined which of the areas indicated by the rectangular frame is specified for the area range, and the area indicated by the corresponding rectangular frame is determined. Treat as a specified area.

This makes it easy to operate because there is no need to strictly specify the range of the area. Even when the image to be processed is displaced from the original position due to inclination or the like, the area can be specified. Is easily done.

Next, an area attribute (attribute information) prepared in the attribute palette is designated for the area setting.
For example, an item indicating a specific area attribute displayed in the attribute palette window is designated by being dragged by operating the mouse 18 and dropped at the area setting position (step A52). Thereby, the CPU 1
0 associates the area attribute selected from the attribute palette with the previously specified area setting.

FIG. 5 shows an example of a state where an area attribute is set for one area. The example shown in FIG.
The area in which the character string of the name is entered is set, and the item of "Name" of the attribute palette displayed in the attribute palette window is dragged and dropped to the position where the area is set. . When the area attribute is specified, as shown in FIG. 5, the item name of the specified area attribute is displayed at the position where the area is set in the layout setting window.

In the same manner, the setting of the area and the designation of the area attribute are executed for the other areas to be processed. In FIG. 6, in addition to the area in which the character string of the name is entered, the area, the area in which the character string of the address, the telephone number, and the age are entered, and the area of the photograph are similarly set, respectively. This indicates that the attribute has been specified.

It is assumed that the area attribute of "photo" is provided below "image" in the attribute palette.
When the area attribute of “photo” is used, an item of “image” is selected in the attribute palette window, so that lower-level items are displayed.

The range in which a plurality of areas are set is managed by the relative position of each area, not the absolute position in the image. For example, it is assumed that the range in which other areas are set is managed by a relative position based on the position where the area is set first.

In this way, for the image displayed in the layout setting window, the area range and area attribute are set for the area to be processed, and save is selected from the layout registration processing menu (not shown). When this is done (step A6), the CPU 10 registers the content set on the image as layout information in a predetermined location of the layout palette, in this case, "profile B" (step A7). The layout information registered in the layout pallet can be used as components used for layout specification at the time of character reading processing.

Next, the character reading process will be described with reference to the flowchart shown in FIG. When the execution of the character reading process is instructed, the CPU 10
Is started and the character reading process is started.

When executing the character reading process, an area for an image to be processed is manually designated by using a layout palette in which layout information is registered in the layout registration process.

In the character reading process, first, an image of a sheet to be processed is read from the image file 22a stored in the scanner device 20 or the storage device 22, and stored in the memory 12 (step B1). CPU10
Displays the image to be processed on the character recognition window in the character reading screen on the screen by the display device 14 (step B2) and displays the layout palette window (step B3).

On the character reading screen, a character recognition window for displaying an image to be processed is provided on the left side, and a layout palette window is provided on the upper right side. In the layout palette displayed in the layout palette window, layout information is hierarchically classified and registered by layout registration processing.

Here, when the mouse 18 is operated and an item in the layout palette is dragged, the CPU 10
Displays a grid pattern indicating the layout area at the cursor position based on the area information of the layout information registered in the dragged item (step B5).

That is, when an item in the layout palette is dragged, the CPU 10 reads out the layout information corresponding to the dragged item, and based on the area information managed by the relative positions of the plurality of areas. To generate a grid pattern and display it according to the cursor position (for example, the cursor position is the lower right vertex of the entire area).

FIG. 8 shows an example of a state in which a grid pattern indicating a layout is displayed by dragging an item on the layout palette. FIG. 8 shows a state in which a grid pattern is displayed based on the layout information (area information) registered in the item of “Profile B”.

The grid pattern represents a combination (layout) of a plurality of areas, and item names (such as “photograph” and “name”) indicating the area attributes added to each area are displayed.

When a lot of layout information is registered in the layout palette, it is possible to determine which necessary layout information is registered on the basis of the item name input during the layout registration process. Even if you forget the item name, just drag each item slightly and the grid pattern corresponding to the registered layout information will be displayed. I can find it.

When the mouse 18 is operated and the cursor is moved on the display screen, the CPU 10 displays a lattice pattern so as to follow the moved cursor. Then, when dropped at an arbitrary position (step B6),
The CPU 10 determines whether or not the dropped image has a layout area corresponding to the area information indicated by the lattice pattern (step B7).

For example, generally, in a form or the like, an area in which characters are to be written is separated by a rectangular frame or the like. Therefore, a frame position is detected in advance from an image to be processed, and the frame position and the grid pattern are detected. Is determined to match. It should be noted that it is not necessary to strictly determine the coincidence here, and an allowance is provided.

When it is determined that there is no layout area corresponding to the area information indicated by the lattice pattern, the CPU 10 cancels the layout information specified from the layout palette.

On the other hand, as shown in FIG. 8, when it is determined that the corresponding layout area exists, the CPU 10
Sets the area attribute of the layout information specified from the layout palette for the corresponding area in the image displayed in the character recognition window (step B9).

In the example shown in FIG. 8, the image to be subjected to character recognition has three specific areas to which the layout information registered in “profile B” of the layout palette is applied.
One is included. Therefore, for this image, drag and drop the item of “Profile B” of the layout palette to each area, that is, three times,
By simply dragging and dropping, layout designation for all areas is completed (step B10).

When the execution of the character recognition processing is instructed after the layout designation is completed, the CPU 10 executes the character recognition processing for each area for which the area attribute is set (step B11).

For example, for an area where the area attribute of “address” is set, a character pattern included in this area is cut out, and for each character pattern, the character type (kanji, kana, , Numbers) with reference to a character recognition dictionary.
Similarly, a character recognition process corresponding to each area attribute is executed for each area.

Further, the region of the photograph in which the region attribute of “image” is set is cut out, for example, from the image.
A predetermined compression process (for example, JPEG) is performed to generate and save another image file.

In the example of the image shown in FIG. 8, since there are three specific areas in which the layout information registered in "profile B" is set, the character recognition processing and the image processing are similarly performed for each of these three areas. Is executed.

The CPU 10 generates an output file by classifying the result of the character recognition processing of each specific area by item (step B12). In this case, since a name, an address, a telephone number, an age, and a photograph are set for each area by the attribute palette, the character recognition result is stored by these items. For the item of the photograph, the file name of an image file generated separately is registered.

FIG. 9 shows an example of a state in which the contents of the output file as the recognition result are displayed. As shown in FIG. 9, the result of character recognition is displayed for each item.
For the photo item, the file name of the image file is displayed.

In this way, as shown in FIG. 8, since the layout information registered in the layout palette is provided as a component for specifying the layout, a shift or the like occurs when the image is read by the scanner device 20. In this case, the layout information as a component can be applied to each specific area.

In particular, as shown in FIG. 8, when a plurality of forms are arranged in a predetermined order and handled as one set, even if each form is shifted, the layout is designated by the same operation. Can do it.

As shown in FIG. 10, even if forms (second areas) of different types are mixed, if the forms are registered in the layout palette by the layout registration processing, the character reading processing can be performed. A predetermined item of the layout palette, for example, an item of “profile A” can be designated by the same operation of drag and drop.

If the layout information registered in the layout pallet has the same area attribute set in the attribute pallet, as shown in FIG. 10, even if forms of different types are mixed, a special operation is performed. Without performing, as shown in FIG. 9, the result of the character recognition process can be classified and output for each common item.

In the above description, a case is described in which layout information necessary for performing character recognition processing on a document, a form, or the like is described. If there is a situation that requires setting, it can be applied.

Further, the method described in the above-described embodiment can be executed by a computer as a program such as a magnetic disk (floppy disk, hard disk, etc.), an optical disk (CD-ROM,
DVD and the like, and can be written to a recording medium such as a semiconductor memory and provided to various devices. Further, it is also possible to transmit the data via a communication medium and provide it to various devices. A computer that implements the present apparatus reads the program recorded on a recording medium or receives the program via a communication medium, and executes the above-described processing by controlling the operation of the program.

[0072]

As described above in detail, according to the present invention, an attribute palette in which attribute information is registered is displayed together with an image to be processed, and an area to be subjected to a recognition process from this attribute palette is displayed. Attribute information can be specified. Also, the layout information set for the image to be processed using the attribute palette (the area to be processed, the attribute information set for the area) is stored in the layout palette as a component used for layout specification. This layout can be registered, the layout palette is displayed, and layout information can be selected from the layout palette and used for layout specification. Therefore, it is possible to easily specify the layout information even if the sheet is misaligned when the image is read by the scanner.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a system configuration of a character reading device according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating a layout registration process according to the embodiment;

FIG. 3 is a view showing an example of a layout setting screen.

FIG. 4 is a diagram showing an example in which items for registering new layout information in a layout palette are set.

FIG. 5 is a diagram showing an example of a state in which an area attribute is set for one area by a layout registration process.

FIG. 6 is a diagram showing an example of a state in which region attributes have been set for a plurality of regions by layout registration processing.

FIG. 7 is a flowchart illustrating a character reading process according to the embodiment.

FIG. 8 is a diagram showing an example of a state in which a grid pattern for designating a layout in a character reading process is displayed.

FIG. 9 is a diagram showing an example of a state in which the contents of an output file as a recognition result are displayed.

FIG. 10 is a view for explaining character reading processing for an image in which forms of different types are mixed.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... CPU 12 ... Memory 12a1 ... Layout registration program 12a2 ... Character reading program 14 ... Display device 16 ... Keyboard 18 ... Mouse 20 ... Scanner device 22 ... Storage device 22a ... Image file 22b ... Layout palette 22c ... Attribute palette

Claims (7)

[Claims] 1. A character reading device that extracts a character pattern from an image based on attribute information and executes a recognition process registers attribute information set for a region to be processed in the image. Attribute information storage means for displaying, an attribute palette display means for displaying an attribute palette for selecting attribute information registered in the attribute information storage means, and attribute information displayed by the attribute palette display means. Image display means for displaying an image to be set; and at least one area to be processed is set from the image displayed by the image display means, and each of the set areas is displayed by the attribute palette display means. Layout setting means for associating the specified attribute information from the selected ones, and Document creating apparatus characterized by comprising a layout information storing means for registering the layout information by the band and the attribute information. 2. A layout palette for selecting layout information registered by the layout information storage means, the image being displayed by the image display means,
Layout pallet display means for displaying together with the attribute pallet displayed by the attribute pallet display means; and layout information selected from the layout information displayed by the layout pallet display means, and an image displayed by the image display means. 2. The character reading apparatus according to claim 1, further comprising: a layout designating unit configured to set a specific region, wherein the recognition processing is performed based on the layout information set by the layout designating unit. 3. The layout storage means stores a plurality of layout information hierarchically, and the layout palette display means displays a layout palette in which layout information is arranged according to a hierarchical structure of the layout storage means. 3. The method according to claim 2, wherein
The character reading device according to the above. 4. The layout designating means, wherein the item representing the layout information displayed by the layout palette display means is a specific item in the image displayed by the image display means in response to an instruction using a pointing device. 3. The character reading device according to claim 2, wherein the layout information is set by being dragged and dropped on the area. 5. The character according to claim 4, wherein when an item representing layout information displayed by said layout palette display means is dragged, a pattern of an area indicated by the layout information of the item is displayed. Reader. 6. A layout information setting method for a character reader that extracts a character pattern from an image based on attribute information and executes a recognition process, wherein the layout information is set for an area to be processed in the image. Register the attribute information to be displayed, display an attribute palette for selecting the registered attribute information, display an image to be processed together with the displayed attribute information, and perform processing from the displayed image. At least one subject
One area is set, attribute information specified from the displayed area is associated with each of the set areas, layout information based on the associated area and the attribute information is registered, and the registered layout is registered. A layout palette for selecting information is displayed together with an image to be processed and an attribute palette, and layout information selected from the displayed layout information is set for a specific area of the displayed image. A layout information setting method for performing a recognition process based on the set layout information. 7. A recording medium on which a program for extracting a character pattern from an image based on attribute information and executing a recognition process is recorded. Registers the attribute information to be set, displays an attribute palette for selecting the registered attribute information, displays an image to be processed together with the displayed attribute information, and displays an image to be processed from among the displayed images. At least one to be processed
One area is set, attribute information specified from the displayed area is associated with each of the set areas, layout information based on the associated area and the attribute information is registered, and the registered layout is registered. A layout palette for selecting information is displayed together with an image to be processed and an attribute palette, and layout information selected from the displayed layout information is set for a specific area of the displayed image. A computer-readable recording medium on which a program for controlling a computer to execute a recognition process based on the set layout information is recorded.