JP2016184359A - Character recognition device, character recognition processing system, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To specify the type of a character box in an image without defining the type of the character box in advance.SOLUTION: A character recognition device 10 to which the present invention is applied comprises: a document image acquisition part 21 that acquires an image including a character box to which characters are entered; a writing area specification part 22 that specifies information on the feature of the character box from the image to be acquired, or a rectangle information extraction part 23 and a writing area extraction part 24 that extract the information on the feature of the character box; and a character box type specification part 25 that specifies the type of the character box from the set information on the feature of the character box.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a character recognition device, a character recognition processing system, and a program.

Patent Document 1 discloses an image of a form in order to create appropriate format information without depending on the skill of the operator, and to simplify the operator's work at the time of creating and registering the format information to reduce the load on the operator. On the screen, a selection area is specified so as to surround at least one field frame, a field frame and a character frame are detected from within this selection area, field frame information relating to the detected field frame is created, and a character frame is detected. In such a case, there is disclosed a technique for creating character frame information relating to the character frame and registering field recognition information including the created field frame information and character frame information as format information.
In Patent Document 2, in order to accurately and efficiently associate an image extracted from a form image with a predetermined processing target area in the template, character recognition or the like is performed on the area extracted from the form image. Prior to aligning this area with the processing target area in the template confirmed to correspond to that area, if the height of the area is greater than the threshold of the processing target area, these areas Compare the size relationship between the distance between the upper side and the lower side of the area, and select the side with the shorter distance among the upper and lower sides that make up the extraction area as the side that was accurately extracted at the time of extraction. Yes.
Further, in Patent Document 3, in order to provide a format control method capable of reading OCR characters in a form without using a format definition body of an input form in an OCR device, a ruled line that partitions lines and a ruled line that partitions digits A character entry frame with non-dropout color printed on a sheet of characters is used to scan the image as an input form, and the number of black pixels in the scanned image data is accumulated in the horizontal and vertical directions, respectively. Creates a histogram, detects the position of the ruled line that separates the line of the character entry frame and the ruled line that separates the digit from the horizontal and vertical histogram distribution, and the ruled line that separates the line of the detected character entry frame and the ruled line Each character entry area is cut out from each position in FIG.

JP 2005-108062 A JP 2001-109826 A Japanese Patent Laid-Open No. 07-210625

Generally, for an image having a character frame in which characters are entered, for example, forms, when the character entered in the character frame is recognized, the type of the character frame is predefined for the character frame. For example, the type of character frame is specified from the definition. The character can be easily identified by referring to the identified character frame type, and the character can be recognized well. However, when the type of character frame is defined in advance for the character frame, it takes time for the definition work, so it is desirable not to define the type of the character frame in advance for the character frame.
An object of the present invention is to specify the type of the character frame without previously defining the type of the character frame for the character frame in the image.

According to the first aspect of the present invention, an acquisition means for acquiring an image having a character frame in which characters are entered, and character frame feature information is set by designation or extraction from the image acquired by the acquisition means. And a specifying unit that specifies the type of the character frame from the feature information of the character frame set by the setting unit.
The invention described in claim 2 is characterized in that the character frame feature information used for specifying the character frame type by the specifying means is information on the height of the frame and the width of the frame. It is a character recognition device of description.
The invention described in claim 3 is characterized in that the type of character frame specified by the specifying means is one character frame, one line frame, or multiple line frames. It is a recognition device.
The invention described in claim 4 is characterized in that the specifying means compares the height of the character frame with the upper limit of the character size and / or compares the width of the character frame with the upper limit of the character size, The character recognition device according to claim 1, wherein information is set.
The invention described in claim 5 further comprises display means for displaying information relating to the type of character frame specified by the specifying means. It is a recognition device.
The invention described in claim 6 includes a character recognition device, an image forming device, and a user entry information acquisition device, wherein the character recognition device is a character frame printed by the image forming device and filled with characters. The character frame feature information is set by designation or extraction from the acquired image, the character frame type is specified from the set character frame feature information, and the image forming apparatus Acquiring entry information entered by the user in the character frame of the printed image from the user entry information acquisition device, and performing character recognition on the acquired entry information using the type of the specified character frame Character recognition processing system characterized by
According to the seventh aspect of the present invention, there is provided a function of setting, by designating or extracting character frame feature information in which characters are entered from an acquired image, to a computer, and character from the set character frame feature information. This is a program for realizing the function of specifying the type of frame.

According to the first aspect of the present invention, the type of character frame can be specified without defining the type of character frame in advance for the character frame in the image.
According to the invention described in claim 2, the type of the character frame can be automatically set.
According to the invention described in claim 3, it is possible to set in accordance with the actual definition of the character frame.
According to the fourth aspect of the present invention, the character frame type can be specified in consideration of the characteristics of the written character with respect to the entry frame.
According to the invention described in claim 5, it is possible to make the user confirm the result of specifying the character frame type.
According to the invention described in claim 6, it is possible to provide a system capable of character recognition according to the type of automatically recognized character frame.
According to the invention described in claim 7, it is possible to cause the computer to function so as to specify the character frame type without defining the character frame type in advance for the character frame in the image.

It is the schematic which shows the whole structure of the character recognition processing system to which this Embodiment is applied. It is a block diagram which shows the function structure of a character recognition apparatus. It is a figure which shows an example of the form image acquired, and an example of the result image in which character frame information was defined by this aspect. (A), (b) is a figure for demonstrating designation | designated of the writing area | region made by the writing area | region designation | designated part with respect to the unfilled form image. It is explanatory drawing which shows the result from which the rectangular information was extracted. (A), (b) is a figure for demonstrating the extraction operation | work of the writing area | region performed after rectangular information is extracted. It is a flowchart which shows the flow of the specific process of a character frame type. (A), (b) is a figure which shows the example of a result as which the kind of character frame was identified by the specific process of a character frame type. It is a figure for demonstrating the relationship of the character frame produced | generated.

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.
[Overall Configuration of Character Recognition Processing System 100]
FIG. 1 is a schematic diagram illustrating an overall configuration of a character recognition processing system 100 to which the exemplary embodiment is applied. The character recognition processing system 100 includes a character recognition device 10, a pen-type device 70 that inputs information such as handwritten characters to the character recognition device 10, and a network 90 such as a character recognition device 10 and a LAN (Local Area Network). And an image forming apparatus 80 that outputs a form image and the like.

  The character recognition device 10 is constituted by a personal computer (PC) and has various functions that a general personal computer has. More specifically, a CPU 11 that executes various software such as an OS (Operating System) and applications and executes various functions described later, a RAM 12 that is used as a working memory of the CPU 11, and various programs that are executed by the CPU 11 are stored. And a ROM 13 which is a memory to be used. The RAM 12 can also function as a primary storage area such as a form image or input information from the pen-type device 70. In addition, a storage unit 14 composed of a magnetic disk or the like for storing a form image, information on an entry frame to be described later, and the like is provided. Furthermore, a display 15 as a display means (output means) such as a liquid crystal display, a keyboard 16 as input means, and a pointing device 17 as auxiliary input means are provided. Furthermore, a network interface 18 that transmits and receives various types of information to and from the pen-type device 70 and the image forming device 80 is provided. The network interface 18 acquires information such as handwritten characters from the pen-type device 70 by wireless communication, for example. Acquisition of information from the pen-type device 70 may be performed via a connector conforming to the USB (Universal Serial Bus) standard, or a configuration performed via the network 90 may be employed. is there.

  The pen-type device 70 that functions as an example of a user entry information acquisition device is a writing device for writing characters, drawings, and the like on paper on which a form image has been formed by the image forming device 80, for example. The pen-type device 70 has a function as an electronic pen, and includes a reading unit 71 that reads information recorded on paper, a storage unit 72 that stores the read information, and transmits the stored information to the character recognition device 10. A communication unit 73 that communicates with the character recognition device 10 is provided. As another user entry information acquisition device, a scanner 82 provided in the image forming device 80 and various input devices (not shown) connected to the network 90 can be employed. In addition, when the reading device 71 reads information recorded on paper, the pen-type device 70 may read an identification image indicating an identifier for uniquely identifying a form image formed on the paper. Good. At least one identification image may be formed on the paper side by side with the form image, or a plurality of fine dot shapes may be scattered throughout the form image. In the case of forming a plurality of identification images, a unique corresponding form image may be specified by reading at least one of the identification images or the like, or may be uniquely supported by some combination A form image may be specified.

  The image forming apparatus 80 forms an image on a recording medium such as paper, communicates with the character recognition apparatus 10 via the network 90, and forms an image on paper based on information transmitted from the character recognition apparatus 10. . It also has functions such as reading and storing images. More specifically, for example, a printer 81 that outputs a form image or an identification image for identifying a form image, a scanner 82 that reads a form image, for example, a storage unit 83 that stores various image information such as a form image, and a network A communication unit 84 that transmits and receives information to and from the character recognition device 10 and the like is provided. The printer 81 can use, for example, an electrophotographic method, but other image forming methods such as an ink jet method can also be adopted.

[Functional Description of Character Recognition Device 10]
FIG. 2 is a block diagram illustrating a functional configuration of the character recognition device 10. As shown in FIG. 2, the character recognition device 10 includes a form image acquisition unit 21 that acquires a form image having a character frame (writing frame) in which characters are entered, and writing that specifies a writing area as an example of a setting unit. An area designating unit 22, a rectangular information extracting unit 23 for extracting rectangular information from the form image, and a writing area extracting unit 24 for extracting a writing area designated by the user. Further, the character frame type specifying unit 25 that estimates the type of the character frame that is the writing region based on the specification of the writing region specifying unit 22 or the extraction of the writing region extracting unit 24, and the relationship between the character frames based on the estimated type. And a character frame relationship generation unit 26 for generation. Furthermore, it has the character frame information output part 27 which outputs the information of the kind of estimated character frame, and the relationship information of the produced | generated character frame.
The form image acquisition unit 21 acquires, for example, an identification image that is read by the scanner 82 included in the image forming apparatus 80 and identifies a form image, a form image, or the like via the network 90. In some cases, the image is stored in the storage unit 83 of the image forming apparatus 80 and acquired via the network 90, or acquired from another apparatus (not shown) connected to the network 90. In some cases, the form image stored in the storage unit 14 of the character recognition device 10 is simply read and acquired.

[Get character frame]
FIG. 3 is a diagram illustrating an example of a form image acquired by the form image acquisition unit 21 and an example of a result image in which character frame information is defined according to the present embodiment. The character frame information includes, for example, character frame type and character frame ordering information. In FIG. 3, an unfilled form image 201 that is a form image in which no characters are filled in, and a result form in which the result of this embodiment is applied and the result of estimating the type of character frame is highlighted on the display 15. An image 202 is shown. The unfilled form image 201 may be an input scanned image or vector information such as pdf and xps. The result form image 202 is displayed after emphasis processing is performed on the defined character frame information, for example, the character frame is displayed in a different color. In FIG. 3, it is indicated by a bold line. A predetermined font character is printed on the result form image 202 shown in FIG.

[Setting by specifying character frame feature information]
FIGS. 4A and 4B are diagrams for explaining the designation of the writing area performed by the writing area designating unit 22 for the blank form image 201 shown in FIG. In the writing area designating unit 22, the character frame feature information is designated, and the character frame feature information includes information such as the height and width of the rectangular information frame. First, in the character recognition device 10, the form image acquired by the form image acquisition unit 21 is read by a dedicated application, and an image for area designation is displayed on the display 15. In FIG. 4A, area designation is performed for an area where a name is entered, and in FIG. 4B, area designation is performed for an area where a zip code is entered. The area designation is performed using the display 15, the keyboard 16, and a pointing device 17 such as a mouse.

  First, on the screen of the display 15, the mouse button as the pointing device 17 is pressed with the mouse pointer over the upper left corner of the frame for which the region is to be specified, and the mouse is moved (dragged) as it is, Release the mouse button at the lower right corner of the frame (drop). In FIG. 4A, the writing area is defined as a field having one rectangle by this drag and drop operation. In the example shown in FIG. 4A, Left 297, Top 594, Width 1123, and Height 71 are defined as coordinate information.

  In FIG. 4B, first, a rectangular area including eight rectangular areas is designated by this drag and drop operation. Thereafter, “division number 8” is designated using the keyboard 16, for example. In the example shown in FIG. 4B, the hyphen area is deleted from the eight designated rectangles and defined as a field having seven frames. For the seven frames, the numerical values of Left, Top, Width, and Height are shown and defined. The defined result is held in, for example, the RAM 12 or the storage unit 14.

[Setting by extracting character frame feature information]
Next, instead of “specifying feature information” by the writing area specifying unit 22, “extracting feature information” by extracting the writing area by the writing area extracting unit 24 after extracting the rectangular information by the rectangular information extracting unit 23. explain.
FIG. 5 is an explanatory diagram showing a result of extracting rectangular information by the rectangular information extracting unit 23. The extracted result is displayed on the display 15 by, for example, coloring a line forming rectangular information or making the line thick. Here, first, a rectangular area surrounded by lines on all sides is automatically searched from the form image acquired by the form image acquisition unit 21. If the acquired form image is, for example, a bitmap image, a linear component is extracted by, for example, Hough transform. On the other hand, when the acquired form image is vector information such as pdf, the existing rectangular information is used as it is. In the extraction stage shown in FIG. 5, for example, an area that is not used for writing in which “application date”, “name”, etc. at the left end are printed may be extracted. The extracted rectangular information is held in the RAM 12, for example, as numerical values (coordinate values) of Left, Top, Width, and Height. Note that at the rectangular information extraction stage shown in FIG. 5, it is not yet known which frame is to be recognized.

FIGS. 6A and 6B are diagrams for explaining a writing area extraction operation performed after rectangular information is extracted. The writing area extraction unit 24 acquires the result of the extraction work performed here.
In the example shown in FIG. 6 (a), after the rectangular information is extracted as described above, the pointing device 17 with the mouse pointer over the frame on which the region is to be extracted is displayed on the screen of the display 15. Press a mouse button. Since the rectangular area is extracted by the rectangular information extracting unit 23 as shown in FIG. 5, the frame on the right side of “NAME” is defined as a field having one rectangle only by this click operation. . In the example shown in FIG. 6A, Left 297, Top 594, Width 1123, and Height 71 are defined as coordinate values.
In FIG. 6B, eight rectangular areas are collectively specified by the drag and drop operation. After that, for example, the area of the hyphen is clicked with the mouse, and the extra rectangle is deleted later. As a result, seven frames are defined as fields, and the numerical values of Left, Top, Width, and Height are shown and defined. The defined result is held in, for example, the RAM 12 or the storage unit 14.

In the present embodiment, when the writing area is specified, the writing area specifying unit 22 acquires the area specification as it is by the user, and the writing area extracting unit 24 extracts the rectangular information after the rectangular information extracting unit 23 extracts the rectangular information. One of the methods of specifying a writing area by extraction can be adopted. The setting of character frame feature information by so-called “extraction” by the rectangular information extraction unit 23 and the writing region extraction unit 24 is compared with the setting of character frame feature information by so-called “designation” by the writing region specification unit 22. The user's operation is slightly lighter.
In addition, in setting character frame feature information by so-called “extraction” or “designation”, at least one of the rectangular information extraction unit 23, the writing region extraction unit 24, and the writing region designation unit 22 uniquely sets the character frame. You may make it produce | generate the character frame identifier to identify. The character frame identifier is held in, for example, the RAM 12 or the storage unit 14. The character frame identifier may be held in association with an identifier for identifying a form image including the character frame indicated by the character frame identifier.

[Specification of character frame type]
Next, the character frame type specifying process will be described.
FIG. 7 is a flowchart showing the flow of the character frame type specifying process performed by the character frame type specifying unit 25. The processing shown in FIG. 7 is processing performed for each (individual) character frame set by the writing region specifying unit 22 or the writing region extracting unit 24.
First, when a new character frame is acquired (step 101), the height of the frame is compared with the upper limit of the character size (step 102). Here, “the height of the frame <the upper limit of the character size” is determined. For example, there is a method in which the upper limit of the character size is specified in advance with a constant and calculated from the specified constant. Further, for example, a method of reading a form in which font characters are preprinted in advance, measuring the size of the character, and calculating the upper limit of the character size to be written from the measured character size is conceivable. In general, it is known that written characters do not become extremely large and do not become extremely small with respect to preprinted type characters. Further, since the type characters are smaller than the written characters, about several times (about 2-3 times) of the preprinted type characters can be determined in advance as the upper limit of the character size.

  Here, when the frame height is smaller than the upper limit of the character size (Yes in step 102), whether or not the frame width is smaller than twice the frame height (frame width / frame height < 2.0) is determined (step 103). If the width of the frame is smaller than twice the height of the frame (Yes in step 103), it is determined that the frame is one character frame from the idea that “two or more characters will not be drawn in the width direction” ( Step 104). If the width of the frame is larger than twice the height of the frame (No in step 103), it is determined that the characters to be written are one-line horizontal frames arranged in a line in the horizontal direction (step 105).

  On the other hand, returning to step 102, when the height of the frame is larger than the upper limit of the character size (No in step 102), whether or not the width of the frame is smaller than the upper limit of the character size (frame width <upper limit of character size). Is determined (step 106). The upper limit of the character size can be set similar to that used in step 102. When the width of the frame is smaller than the upper limit of the character size (Yes in step 106), it is considered that “there is no more than two characters in the width direction” and is determined to be a vertical frame of one line (step 107). . If the frame width is larger than the upper limit of the character size in step 106 (No in step 106), whether or not the frame width is smaller than three times the upper limit of the character size (frame width <character size upper limit × 3) is determined (step 108). If the width of the frame is smaller than three times the upper limit of the character size (Yes in step 108), it is determined that the frame is a multi-line vertical frame that has a plurality of vertical lines as used in vertically written Japanese characters (step 109). ). When the width of the frame is larger than three times the upper limit of the character size (No in step 108), it is determined that the frame is a multi-line horizontal frame that has a plurality of horizontal lines as used in English-American characters and horizontal Japanese characters. (Step 110).

  FIGS. 8A and 8B are diagrams illustrating an example of the result of specifying the character frame type by the character frame type specifying process described with reference to FIG. 7. In FIG. 8A, the aspect ratio 311 is calculated as “15.82” for the character frame specified in FIG. 4A and the character frame extracted in FIG. The frame type 312 which is larger than twice the height of the frame (corresponding to No in step 103 in FIG. 7) and is the character frame type is specified as “one line horizontal”. Further, in FIG. 8B, the aspect ratio 311 is “0.49” for the seven character frames specified in FIG. 4B and the seven character frames extracted in FIG. “0.52”, “0.51”, “0.52”, “0.50”, “0.53”, “0.52” are calculated, and all seven character frames have a vertical length. The width of the frame (horizontal length) is shorter than twice the height of the frame (vertical length) (corresponding to Yes in step 103 in FIG. 7). Therefore, the frame type 312 which is the type of the character frame is specified as “one character” in all the seven character frames. The information specified by the character frame type specifying unit 25 as described above is associated with the corresponding character frame and held in, for example, the RAM 12 or the storage unit 14. At this time, it may be held together with at least one of an identifier or a character frame identifier. Thereby, by acquiring an identifier or a character frame identifier, a corresponding character frame and specified information can be acquired. The identifier or the character frame identifier is formed as an identification image or a character frame identification image, respectively, when the corresponding form image or character frame is formed by the image forming apparatus 80.

FIG. 9 is a diagram for explaining the relationship between character frames generated by the character frame relationship generation unit 26. In the character frame relationship generation unit 26, for example, the character frame order is given by the coordinate value of the frame. For example, in a character frame in which a zip code is entered, the accuracy of the recognition process can be improved by assigning an order to the characters in consideration of the regularity of the zip code and performing the recognition process in that order. In the example shown in FIG. 9, in accordance with the coordinate value of the character frame, the order 313 is assigned from No. 1 to No. 7 corresponding to the 7 digits of the zip code. The order is given by a predetermined rule, for example, from left to right and from upper left to lower right. By storing the information of the specified character frame type (frame type 312) together with the order 313 from NO.1 to NO.7, the accuracy of the subsequent recognition processing is improved.
Note that the order is changed depending on the type of character frame specified, for example, in the case of vertical writing, the order is assigned from right to left.

  In the character frame information output unit 27 shown in FIG. 2, information obtained by the character frame type specifying unit 25 and the character frame relationship generation unit 26 is output to the display 15, for example. Depending on the application, in the character frame specifying process, the result estimated by the character recognition device 10 can be output to the display 15 and confirmed by the user as the final specified result. In addition, the character frame information output unit 27 can output information on the specified character frame type to the display 15 as overall attribute information.

  As described above, in the character recognition device 10 to which the exemplary embodiment is applied, for example, an image having a character frame printed by the image forming device 80 and filled with characters is converted into a form image acquisition unit as electronic information before printing. 21, and character frame feature information is set by designation by the writing area designation unit 22 or extraction by the rectangular information extraction unit 23 and the writing area extraction unit 24 from the obtained image. Then, the character frame type specifying unit 25 specifies the character frame type from the set character frame characteristic information. Thereafter, entry information entered by the user using the pen-type device 70 in the character frame of the image printed by the printer 81 of the image forming apparatus 80 is acquired via the network interface 18. Alternatively, the entry information entered by the user on the paper printed by the printer 81 of the image forming apparatus 80 is acquired via the network interface 18 from another user entry information acquisition apparatus (such as an image reading apparatus). At this time, at least one of the identification image printed together with the character frame or the character frame identification image may be acquired. For example, the character frame included in the corresponding form image and its type (specified character frame) are acquired from the acquired identification image. For example, the corresponding character frame and its type (specified character frame) are acquired from the acquired character frame identification image. Then, character recognition is performed on the acquired entry information using the specified character frame type.

  In the character recognition processing system 100 shown in FIG. 1, the character recognition device 10 using the pen-type device 70 has been described. This embodiment can also be applied to other character recognition processing systems. For example, the character recognition device 10 can be arranged as a function of a form server as data management connected to a core business server that performs form data generation processing. The character recognition device 10 to which this exemplary embodiment is applied can specify the type of character frame without defining the type of character frame in advance for a form image or form image having a character frame.

  The processing of the present embodiment described above can be provided as a program that is realized by the character recognition device 10 that is a computer. This program is stored in, for example, the ROM 13, expanded on the RAM 12, which is a working memory, and executed by the CPU 11.

DESCRIPTION OF SYMBOLS 10 ... Character recognition apparatus, 15 ... Display, 16 ... Keyboard, 17 ... Pointing device, 21 ... Form image acquisition part, 22 ... Writing area designation part, 23 ... Rectangular information extraction part, 24 ... Writing area extraction part, 25 ... Character Frame type identification unit, 26 ... character frame relationship generation unit, 27 ... character frame information output unit, 70 ... pen-type device, 80 ... image forming device, 100 ... character recognition processing system

Claims (7)

Obtaining means for obtaining an image having a character frame in which characters are written;
Setting means for setting character frame feature information by designation or extraction from the image acquired by the acquisition means;
Specifying means for specifying the type of character frame from the character frame feature information set by the setting means;
A character recognition device.   2. The character recognition apparatus according to claim 1, wherein the character frame feature information used for specifying the character frame type by the specifying means is information on a frame height and a frame width.   The character recognition device according to claim 1 or 2, wherein the type of character frame specified by the specifying means is one character frame, one line frame, or a plurality of line frames.   The identifying means compares the height of the character frame with the upper limit of the character size and / or compares the width of the character frame with the upper limit of the character size, and sets the character frame feature information. Item 4. The character recognition device according to any one of Items 1 to 3.   5. The character recognition device according to claim 1, further comprising display means for displaying information relating to a type of character frame specified by the specifying means. Including a character recognition device, an image forming device, and a user entry information acquisition device,
The character recognition device acquires an image having a character frame printed by the image forming device and filled with characters, and sets or sets character frame feature information by designation or extraction from the acquired image. Identifying the type of character frame from the character frame feature information, obtaining the entry information entered by the user in the character frame of the image printed by the image forming apparatus from the user entry information acquisition device, A character recognition processing system that performs character recognition on the entered information using the specified type of the character frame. On the computer,
A function to set or extract characteristic information of a character frame in which characters are entered from the acquired image;
A function for identifying the type of character frame from the set character frame feature information,
A program that realizes

Images