JP2013069150A - Character input device and character input method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a character input device and a character input method for efficiently inputting a correct character in the case of confirming and correcting a character recognition result.SOLUTION: A character input device includes a character recognition section, an acceptance section, a first matching section, a character string setting section and a storage section. The character recognition section performs character recognition processing to an image of a business form to acquire one or more character strings as a result of recognition. The acceptance section includes items for classifying the one or more character strings, and accepts the input of the input character string to each item. The first matching section performs the matching of the input character string accepted by the acceptance section with the one or more character strings obtained by the character recognition section. The character string setting section stores the character string which has matched with the input character string in association with the item to which the input character string has been input in the storage section.

Description

  Embodiments described herein relate generally to a character input device and a character input method for reading information described in a form and classifying a character string included in the read information for each item.

For example, there is a character input device that classifies information (such as a company name, name, address, job title, and contact phone number) described in a form such as a business card for each reading item.
In this type of character input device, the read results are automatically classified for each item and displayed on the display screen to prompt the operator to confirm.
Conventionally, there is a technique in which an operator confirms a character of a recognition result, and in some cases, reads several character candidates from a default dictionary and displays them as a predicted character string on a display screen to replace the character of the recognition result.

JP 2000-348141 A

The conventional automatic item classification technology is not very accurate, so the frequency of misclassification is high, and if misclassification occurs, it is necessary to delete the character string from the corresponding item and reclassify it. It will take time.
In the case of the prior art, since characters displayed as classification candidates for each item are limited to those registered in the dictionary, there are many cases where appropriate characters are not displayed as classification candidates. When an unsuitable character is classified as an item as a recognition result, reclassification work is required in the same manner as described above, resulting in a problem that work efficiency is lowered.

  An object of the present invention is to provide a character input device and a character input method that can classify character strings read from a form into appropriate items.

  The character input device of the embodiment includes a character recognition unit, a reception unit, a first matching unit, a character string setting unit, and a storage unit. The character recognition unit obtains one or more character strings as a recognition result by performing character recognition processing on an image obtained by optically reading a form. The reception unit includes items for classifying one or more character strings recognized by the character recognition unit, and receives input of an input character string for each item. The first matching unit matches the input character string received by the receiving unit with one or more character strings obtained by the character recognition unit. The character string setting unit stores the character string matched with the input character string in the storage unit in association with an item in which the input character string is input.

It is a figure which shows the structure of the character input system of 1st Embodiment. It is a figure which shows an example of the input image which read the business card. It is a flowchart which shows operation | movement of the character input system of 1st Embodiment. It is a figure which shows a mode that the character image was enclosed with the circumscribed rectangle calculated for every group of the character string image cut out from the input image. An example of a recognition result obtained by recognizing each group is shown. It is a figure which shows an example of an input window. It is a figure which shows the mode of matching of the input character string and recognition result optimized by DP matching. It is a figure which shows the mode of the matching which is not optimal. It is a figure which shows the list of the character strings enumerated by matching with input character string "recognition" among the recognition results shown in FIG. It is a figure which shows an example of a recognition result, an input character string, and a matching result. It is a figure which shows the example of a display of the list box displayed on an input window. It is a figure which shows the example which highlighted and displayed the list box on the input window. It is a flowchart which shows the matching operation | movement of a character input system. It is a flowchart which shows operation | movement of the matching result selection part when a "↓" key is pressed. It is a flowchart which shows operation | movement of the matching result selection part when an "↑" key is pressed. It is a flowchart which shows operation | movement of the matching result selection part when an "Enter" key is pressed. It is a flowchart which shows operation | movement of the matching result selection part when a "Tab" key is pressed. It is a block diagram which shows the structure of the character input system of 2nd Embodiment. It is a figure which shows an example of the character string stored in a word dictionary. It is a figure which shows the mode of the word matching and replacement in a 2nd matching part. It is a block diagram which shows the mode of operation | movement of the character string setting part in 3rd Embodiment. It is a block diagram which shows the structure of the character input system of 7th Embodiment. It is a figure which shows the mode of the setting of the input character string in a 3rd matching part. It is a block diagram which shows the structure of the character input system of 8th Embodiment.

Hereinafter, embodiments will be described in detail with reference to the drawings.
(First embodiment)
FIG. 1 is a diagram showing a configuration of a character input system according to the first embodiment.
As shown in FIG. 1, the character input system of this embodiment includes an image reading device such as a scanner 1, a key input device such as a keyboard 2, a display device such as a monitor 3, and these devices and a communication cable or network. And a computer 10 (hereinafter referred to as “PC 10”) connected via a computer.

  The PC 10 is a computer having hardware such as a CPU, a memory, a hard disk device, and a communication interface, and a program (such as an operating system (hereinafter referred to as “OS”) and image recognition software) installed in the hard disk device. The memory functions as the storage unit 13.

  The software (program) includes an image reception unit 11, a character recognition unit 12, a character reception unit 14, a first matching unit 15, a matching result selection unit 16, a character string setting unit 17, and the like.

  The image receiving unit 11 receives an image of a form reading result input from the scanner 1 (hereinafter referred to as “input image”). The character recognition unit 12 cuts out a character image from the input image and performs character recognition on the cut out character image, thereby combining a recognition result (text data (character) and character image (or character string and character string image). ) That is, the character recognition unit 12 obtains one or more character strings (see FIG. 5) as a recognition result by performing character recognition processing on an image obtained by optically reading a form. In the following description, a single character or character string is referred to as a “character string”.

The character accepting unit 14 accepts text data (referred to as an input character string) input by a user's key operation on the keyboard 2.
Specifically, the character accepting unit 14 displays a screen (see the input window 51 in FIG. 6) having items for classifying one or more character strings recognized by the character recognizing unit 12 on the monitor 3. , And functions as a reception unit that receives input of an input character string for each item.

  The first matching unit 15 matches the character string of the recognition result obtained by the character recognition unit 12 with the input character string received by the character receiving unit 14. When there are a plurality of character strings that match the input character string as a result of matching by the first matching unit 15, the matching result selection unit 16 presents the list box 52 as a candidate list (selection candidate) to prompt the user to select. That is, the matching result matched by the first matching unit 15 is displayed on the input window 51 as an input candidate character string and presented to the user, and the user is allowed to select one of the candidates.

The character string setting unit 17 stores the character string matched with the input character string in the storage unit 13 in association with the item in which the input character string is input.
When there are a plurality of character strings that match the input character string, the character string setting unit 17 replaces the candidate selected by the user from the candidate list with the input character string input to the item. That is, the character string setting unit 17 sets the candidate selected from the candidates displayed in the input window 51 (candidate list of the list box 52) to the item in which the input character string is input, and the item and the character of the recognition result Associate a column.

  Since this embodiment can be configured by incorporating a general-purpose OS and image recognition software into the PC 10, the following description will be made assuming such a configuration. However, the embodiment shown here is only an example. For example, each unit may be dedicated hardware, an aggregate thereof, or a computer network for distributed processing. That is, as long as the configuration includes the function of each unit described above, any configuration may be used regardless of the configuration given here.

  In the first embodiment, a memory is used as an example of the storage unit 13. Specifically, for example, a random access memory (RAM) that can store data such as images and variables, a hard disk drive device, a flash memory, Solid state drive (SSD) or the like. In addition, unless otherwise specified, in this embodiment, the storage unit 13 is described. However, the storage unit 13 is not limited to the above, and any other storage medium may be used.

Next, the operation of the character input system according to the first embodiment will be described with reference to FIGS.
In the case of the first embodiment, when a business card set on the reading stand of the scanner 1 is read by the scanner 1, a business card image (input image) as shown in FIG. The input image is received by the image receiving unit 11 and passed to the character recognition unit 12.

  The character recognition unit 12 detects an area (part) including a character string from the input image, recognizes the area in units of groups, and stores a recognition result (a combination of the character image and text data) in the storage unit 13.

  Specifically, the character string detection / recognition operation in the character recognition unit 12 is performed according to the procedure shown in the flowchart of FIG.

  That is, the character recognizing unit 12 detects a connected component of black pixels constituting the character from the input image (step S101 in FIG. 3).

  The character recognizing unit 12 creates a group in which connected components whose distances are equal to or less than a preset threshold among the connected components are collected (step S102).

  The character recognition unit 12 calculates a circumscribed rectangle for each created group (step S103).

  As shown in FIG. 4, the character recognition unit 12 performs character recognition processing on the character image for each group surrounded by each circumscribed rectangle, and stores the recognition results (a plurality of character strings) as shown in FIG. Store in the unit 13 (step S104).

  The character reception unit 14 displays the input window 51 (see FIG. 6), and when the user operates the keyboard 2 and an input character string is input in the field of each item in the input window 51, the input character The column is received and passed to the first matching unit 15.

  As shown in FIG. 6, the input window 51 includes items such as “first and last name”, “company name”, “department name”, “title”, “business name”, “zip code”, and “address”. A character string input field is provided. In this example, it is assumed that, for example, a character string “recognition” is input as an input character string in the input field (column) of the item “first and last name”.

  In addition, the character reception part 14 may acquire a character by detecting a key type directly and converting it into the said character. Furthermore, the character may be acquired through an external program such as a kana-kanji conversion front-end processor, or the character may be acquired by any method as long as the above-mentioned meaning is not removed.

  Next, the first matching unit 15 matches the input character string received by the character receiving unit 14 with the character string of the recognition result by the character recognition unit 12, and lists the matching results (stores in the storage unit 13).

  The first matching unit 15 matches the input character string stored in the character receiving unit 14 and each character string of the recognition result using a technique such as DP matching, and calculates the score (cost) of each character string. The candidates are enumerated (selected) in the order of the calculated score (cost) (up to the top number in the score (cost)).

  In this character input system, DP matching is performed under conditions where both ends of the input character string are fixed and both ends of the recognition result are free.

  In general, in association by DP matching between character strings, as shown in FIG. 7, a score (hereinafter referred to as “cost”) is defined according to the association situation between characters, and by summing them, Define the cost of mapping. Then, the association is optimized by minimizing this cost.

  In this embodiment, as shown in FIG. 7 and FIG. 8, the cost of a location where a matching character is matched between the input character string “recognition” and the character string “recognition Taro” of the recognition result is 0, The cost of the location where the character in the character string is associated with the unread character in the character string of the recognition result is C1 = 10, and the character in the input character string is different from the character in the character string of the recognition result The cost associated with the character is C2 = 20, the cost when the character in the recognition result character string is not associated with the character in the input character string is C3 = 100, the character in the recognition result is The cost when characters in the input character string are not associated is assumed to be C4 = 100.

  However, even if the character in the input character string is not associated with the character in the character string of the recognition result, the character is in front of the character associated with the first character in the input character string or already entered If the character is after the character associated with the last character of the character string, the cost of the character is zero. By doing so, DP matching can be performed in which both ends of the character string of the recognition result are free (characters can be added to the recognition result).

  The values such as the costs C1, C2, C3, and C4 shown here are examples, and any values may be set as long as natural matching can be expected.

  The first matching unit 15 enumerates (selects) the recognition results with the lowest DP matching cost within a predetermined number. Further, the first matching unit 15 enumerates (selects) recognition results whose costs are equal to or less than a predetermined constant (for example, 0), so that a character string “recognition” is displayed as shown in FIG. The character strings “Recognition Taro”, “Recognition Solution Co., Ltd.”, “Recognized prefecture recognition city recognition town 9-99-999”, etc., are selected in order of increasing score.

  The first matching unit 15 replaces each character associated with the character in the input character string in the recognition result by the above-described DP matching with the character associated with the input character string to obtain a matching result. . An example of the recognition result, the input character string, and the matching result is shown in FIG.

  The matching result selection unit 16 displays the matching result matched by the first matching unit 15 as an input candidate character string in the corresponding field of the input window 51 shown in FIG. 6 and presents it to the user. To choose.

  The matching result selection unit 16 displays a list of matching results immediately below the character string being input. FIG. 11 shows an example of a list display. The list display function is realized by a list box display function provided by the OS or the widget set.

  The matching result is registered as a display item in the list box, and is selected by clicking with a pointing device such as a mouse. The number (value) of the selected matching result is stored in the variable SPOS. The variable SPOS is stored in the storage unit 13.

  After setting a value in the variable SPOS, the matching result selection unit 16 performs control so that the SPOS-th matching result is set as the input character string in the character string setting unit 17 described later.

  The matching result selection unit 16 controls the display of the list box by monitoring the key input so that the matching result can be selected by a key operation.

  As shown in FIG. 12, in the list box 52, one of the displayed items is highlighted, and the matching result selection unit 16 selects the item, that is, the provisional result of the matching result by highlighting. Show.

  The temporarily selected state is managed by a variable KPOS stored in the storage unit 13. The variable KPOS is set to 0 in a state where provisional selection is not performed, and in the state where provisional selection is performed, the number of the matching result that is provisionally selected is set. If the first matching result is provisionally selected, 1 is set. Hereinafter, the number of matching results is n.

  The matching result selection unit 16 performs the process of the flowchart shown in FIG. 13 each time characters are matched by the first matching unit 15.

  That is, the matching result selection unit 16 first checks whether or not the list box is displayed on the monitor 3 (step S201 in FIG. 13), and if the list box is not displayed (No in step S201), the list box A box is displayed (step S202).

  Subsequently, the matching result selection unit 16 sets a list of matching results in the displayed list box (step S203).

  Finally, the matching result selection unit 16 sets the variable KPOS = 0 to a state where no provisional selection has been made (step S204).

The matching result selection unit 16 performs the processing of the flowchart shown in FIG. 14 each time a first temporary selection key (for example, a key marked with “↓”) is pressed, that is, the matching result selection unit 16 Confirms the value of the variable KPOS and determines whether the value of the variable KPOS is 0 or larger than 0 (step S301).

  As a result of this determination, if the variable KPOS> 0 (Yes in step S301), that is, if a temporary selection has been made, the matching result selection unit 16 sets the variable KPOS = min (KPOS + 1, n) (step S302). However, min (x, y) is a function that returns the smaller number of x and y.

  As a result of the above determination, if the variable KPOS = 0 (No in step S301), that is, if no temporary selection has been made, the matching result selection unit 16 selects the variable KPOS = 1, that is, the first matching result. (Step S303).

  Finally, the matching result selection unit 16 highlights the KPOS-th matching result (step S304).

  The matching result selection unit 16 performs the process of the flowchart shown in FIG. 15 each time a second temporary selection key (for example, a key marked with “↑”) on the keyboard 2 is pressed.

  That is, the matching result selection unit 16 confirms the value of the variable KPOS and determines whether the value of the variable KPOS is 0 or a value greater than 0 (step S401).

  As a result of this determination, if the variable KPOS> 0 (Yes in step S401), that is, if a temporary selection has been made, the matching result selection unit 16 sets KPOS = max (KPOS-1, 1) (step S402). . However, max (x, y) is a function that returns the larger number of x and y.

  As a result of the determination, if the variable KPOS = 0 (No in step S401), that is, if a temporary selection has not been made, the matching result selection unit 16 sets the variable KPOS = n (step S403). That is, the nth matching result is selected.

  Finally, the matching result selection unit 16 highlights the KPOS-th matching result (step S404).

  The matching result selection unit 16 performs the processing of the flowchart shown in FIG. 16 every time a selection determination key (for example, a key marked with “Enter”) on the keyboard 2 is pressed.

  That is, the matching result selection unit 16 checks the value of the variable KPOS and determines whether the value of the variable KPOS is 0 or a value greater than 0 (Step S501 in FIG. 16).

  As a result of this determination, if KPOS> 0 (Yes in step S501), that is, if a temporary selection has been made, the matching result selection unit 16 sets SPOS = KPOS (step S502), and the character string setting unit 17 described later. Control is performed so that the SPOS-th matching result is set in the input character string (step S503).

  Subsequently, the matching result selection unit 16 sets KPOS = 0 (step S504), and makes a state in which no provisional selection is made.

  Finally, the matching result selection unit 16 hides the list box 52 (step S505).

  The matching result selection unit 16 performs the processing of the flowchart shown in FIG. 17 each time a second selection determination key (for example, a key marked with “Tab”) on the keyboard 2 is pressed.

  That is, the matching result selection unit 16 sets the counter value n to 1 (step S601 in FIG. 17).

  When the counter value n = 1 (Yes in step S601), the matching result selection unit 16 sets SPOS = 1 (step S602), and the character string setting unit 17 (to be described later) sets the SPOS-th matching result as the input character string. Control is performed so as to be set to (step S603).

  Then, the matching result selection unit 16 sets KPOS = 0 (step S504), and makes a state where no provisional selection is made.

  Finally, the matching result selection unit 16 hides the list box 52 (step S605).

The character string setting unit 17 sets the selected matching result as the character string being input.
The character string setting unit 17 sets the matching result selected by the matching result selection unit 16 in the edit box, and sets the matching result selected by the matching result selection unit 16 to the input character string temporarily held in the edit box. Replace with a string.

As described above, according to the first embodiment, for example, information (one or more character strings) obtained by reading a form such as a business card with the scanner 1 is used as, for example, “first and last name”, “company name”,. In order to classify each item such as “address”, the input character string and the recognition result are input by keying the input character string of about the first few characters (about one or two characters or words) into the input field of each item. Matches one or more character strings and stores the matched character string in the storage unit 13 in association with the input character string input item, so that the character or character string read from the form is classified as an appropriate item. can do.
In addition, when there are multiple matching character strings, they are displayed in the list box as selection candidates to be replaced with the input character string input by the key. The look and feel can be realized with the same interface, and the user can input from the candidates obtained from the result of the character recognition, and the character input to the item can be made efficient.

(Second Embodiment)
Next, a character input system according to the second embodiment will be described with reference to FIG. FIG. 18 is a diagram illustrating a configuration of a character input system according to the second embodiment. In FIG. 18, the same components as those in the first embodiment are denoted by the same reference numerals, and the description thereof is omitted.

  As illustrated in FIG. 18, the PC 10 of the character input system according to the second embodiment includes a word dictionary storage unit 18 and a second matching unit 19. The word dictionary storage unit 18 is realized by a memory or the like. The word dictionary storage unit 18 stores a word dictionary that stores a plurality of words. As shown in FIG. 19, the word dictionary stores the words in the dictionary as an array of character strings.

  The second matching unit 19 matches the input character string received by the character receiving unit 14 with one or more character strings obtained by the character recognizing unit 12 and the words in the word dictionary, and classifies the corresponding items. Select a power string.

  Specifically, the second matching unit 19 searches the word dictionary using the character string of the recognition result and the input character string as keys, and reads the hit word as a selection candidate (input candidate character string).

  In this case, the matching result selection unit 16 displays the matching result by the second matching unit 19 as a selection candidate character string to be classified into the corresponding item in the input window 51, and presents the selection result to the user for selection. The matching result selection unit 16 presents a selection candidate when there are a plurality of words that match the character string of the recognition result and the input character string.

  That is, the matching result selection unit 16 in the present embodiment presents the matching results matched by the second matching unit 19 instead of the first matching unit 15 to the user as a candidate list.

Next, the operation of the character input system according to the second embodiment will be described.
In the second embodiment, the second matching unit 19 matches the recognition result of the character recognition unit 12, the input character string received by the character reception unit 14, and the word in the word dictionary.

  The second matching unit 19 first matches each character string of the recognition result with each of the words in the word dictionary, and replaces the character string of the recognition result with the word in the word dictionary, thereby misreading characters or unread characters. A corrected recognition result character string excluding characters is generated.

  The matching between the word in the word dictionary and the character string of the recognition result is performed by the same method as the association of the input result character string with the character string of the recognition result in the first matching unit 15 in the first embodiment.

  Next, the second matching unit 19 matches the input character string key-input by the user with each character string of the corrected recognition result to obtain a matching result. Matching of the corrected recognition result and the input character string is performed by the same method as the association of the recognition result and the input character string in the first matching unit 15 in the first embodiment.

  As shown in FIG. 20, when the character string of the recognition result is, for example, “Recognition Sonshon Co., Ltd.” and the input character string input by the key is “Solution”, the second matching unit 19 converts the input character string into The corrected recognition result “Recognition Solution Co., Ltd.” is obtained by preferentially replacing the corresponding part of the character string of the recognition result. That is, the unread or misread character portion of the recognition result is replaced with the input character string, and the correct character can be input.

  As described above, according to the second embodiment, the input character string key-inputted to the classification target item, the character string of the recognition result, and the word in the word dictionary are matched together to thereby classify the classification target item. In addition, more appropriate selection candidates can be listed (presented). To match the input character string entered in the item field, the character string of the recognition result, and the character string from three different information sources such as words in the word dictionary, first match the two, then the result and the rest However, all combinations of character string matching are included in the scope of the present invention.

(Third embodiment)
Next, a character input system according to a third embodiment will be described. Since the third embodiment is configured similarly to the first embodiment or the second embodiment, the configuration diagram and the description thereof are omitted.

This third embodiment relates to an operation when an unread character indicated by, for example, “?” Is included in the matching result. As shown in FIG. When the user inputs an input character string “”, recognition results such as “recognition Taro”, “recognition solution ??? company”, “recognized prefecture recognition city recognition town 9-99-999”, etc. are entered in the list box 52 as matching results. Is displayed. In this example, the unrecognized characters “?” And “?” Are included in the character string “recognition solution”.
In this case, when the user selects a character string “recognition solution ??? company” including unread characters “?” And “?”, The character string setting unit 17 displays “recognition solution ??? company”. The unrecognized characters “?” And “?”, Which are characters just before “?”, Are replaced with the input character string “recognition” to be edited entered in the item field. Other operations are the same as those in the first or second embodiment.

  As described above, according to the third embodiment, when the character string of the recognition result is matched with the character string being input and the word in the word dictionary, unrecognized characters and misread characters are included in the recognition result. Even if the character string has been read, the character string up to immediately before the unread character or the misread character in the character string of the recognition result is replaced with the input character string key-inputted to the input field of the input window 51, so that Without taking the trouble of reclassification from misclassification, it is possible to classify an appropriate character string into the corresponding item and improve work efficiency.

(Fourth embodiment)
Next, a character input system according to a fourth embodiment will be described. Since the fourth embodiment is configured in the same manner as the first or second embodiment, the configuration diagram and the description thereof are omitted.

  In the fourth embodiment, in the first matching unit 15 or the second matching unit 19, the matching result selection unit 16 has already selected the input image from one or more character strings (selection candidates) of the recognition result. A flag indicating non-presentation is attached to the character string (candidate), and the recognition result with the flag is excluded from matching. Other operations are the same as those in the first or second embodiment.

  As described above, according to the fourth embodiment, in the first matching unit 15 or the second matching unit 19, the matching result selection unit 16 out of one or more character strings (selection candidates) of the recognition result for the input image. By excluding the character string (candidate) already selected in step 3, unnecessary candidates are not presented, and the user can easily select candidates.

(Fifth embodiment)
Next, a character input system according to a fifth embodiment will be described. Since the fifth embodiment is configured similarly to the first embodiment or the second embodiment, the configuration diagram and the description thereof are omitted.

  The fifth embodiment differs from the first embodiment or the second embodiment in the operation of the matching result selection unit 16. That is, the matching result selection unit 16 in this example attaches a flag indicating that the input image is not to be presented to a character string (candidate) already selected from one or more character strings (selection candidates) of the recognition result. The recognition result with the flag is excluded from the presentation target. Other operations are the same as those in the first or second embodiment.

  As described above, according to the fifth embodiment, the matching result selection unit 16 presents a character string (candidate) already selected from one or more character strings (selection candidates) of the recognition result for the input image. By excluding this, unnecessary candidates are not presented, and the user can easily select candidates.

(Sixth embodiment)
Next, a character input system according to a sixth embodiment will be described. Since the sixth embodiment is configured in the same manner as the second embodiment, the configuration diagram and description thereof are omitted.

  In the case of the sixth embodiment, a plurality of word dictionaries or word groups are provided for each item to be input, and the second matching unit 19 matches the target word by switching the word dictionary or word group for each item. The word dictionary is, for example, an address dictionary, a postal code dictionary, a name dictionary, and the like, and the word group is different from the word dictionary, and refers to a group of words divided into groups such as katakana, numbers, symbols, and kanji.

  As described above, according to the sixth embodiment, a plurality of word dictionaries or word groups are provided for each input target item, and the word dictionary or word group to be matched by the second matching unit 19 is switched for each item. Processing time can be reduced by reducing matching with unnecessary words. In addition, it is possible to prevent the recognition result from being corrected incorrectly by matching with an incorrect word.

(Seventh embodiment)
Next, a character input system according to a seventh embodiment will be described with reference to FIG.
As shown in FIG. 22, the character input system according to the seventh embodiment includes a word dictionary storage unit 18, a third matching unit 20, and a spare in addition to the configuration of the first embodiment (the configuration having the first matching unit 15). A setting unit 21 is provided.
The word dictionary storage unit 18 stores a word for each item to be input (“first name”, “company name”, “department name”, “title”, “business name”, “zip code”, “address”). Is stored.

  The third matching unit 20 matches each recognition result with a word in the word dictionary and determines an item corresponding to the recognition result. At this time, prior to input by the user, the third matching unit 20 matches each of the recognition results with each of the words in the corresponding word dictionary for each of the items.

  That is, the third matching unit 20 uses the same method for matching the recognition result and the input character string stored in the character accepting unit 14 in the first matching unit 15 of the first embodiment, and compares the recognition result and each word. Matching is performed, and the matching cost of each combination of the recognition result and the word is calculated. The cost calculation method is the same as that described with reference to FIGS.

  The preliminary setting unit 21 selects the character string of the recognition result that minimizes the cost calculated by the third matching unit 20 and sets it as the input character string of the field corresponding to the matched item. Notify Upon receiving this notification, the matching result selection unit 16 sets an input character string in the field of the corresponding item.

  The preliminary setting unit 21 and the matching result selection unit 16 set the recognition result in the items matched by the third matching unit 20 prior to input by the user.

  In the case of the seventh embodiment, as shown in FIG. 23, the third matching unit 20 determines that the recognition result character is, for example, a character string that is set in the field of the item “first name and last name”, for example, “recognized fat?” , Match the words "Taro", "Ichiro", "Hanako" ... in the word dictionary of the item "Recognition thick?" And "Last name" for the character "thick" in front of the unread character "?" Select “Taro” with the lowest cost.

  Then, the preliminary setting unit 21 notifies the matching result selection unit 16 to set the input character string “recognition Taro” in the field of the item matched by the third matching unit 20 prior to input by the user, and selects the matching result. Upon receiving this notification, the unit 16 sets the selected input character string “Recognition Taro” in the field of the corresponding item of the input window 51.

  As described above, according to the seventh embodiment, the word dictionary storage unit 18 that stores the word dictionary storing the word for each input target item and the recognition result are recognized by matching each of the recognition results with the word in the word dictionary. A third matching unit 20 that determines an item corresponding to the result, and a preliminary setting unit 21 that sets a character string of a recognition result to an item matched by the third matching unit 20 prior to input by the user. A recognition result that can automatically determine an item to which a character string is to be input is automatically input to the field of the item. This automatic input can reduce the user's trouble of key input.

(Eighth embodiment)
Next, a character input system according to an eighth embodiment will be described with reference to FIG.
As shown in FIG. 24, the character input system according to the eighth embodiment includes a word dictionary storage unit 18, a third matching unit 20, and a spare in addition to the configuration of the second embodiment (the configuration having the second matching unit 19). A setting unit 21 is provided. The word dictionary storage unit 18, the third matching unit 20, and the preliminary setting unit 21 in the eighth embodiment are the same as those in the seventh embodiment.

  The word dictionary storage unit 18 stores a word for each item to be input (“first name”, “company name”, “department name”, “title”, “business name”, “zip code”, “address”). Is stored.

  Prior to input by the user, the third matching unit 20 matches each of the recognition results with each of the words in the corresponding word dictionary for each of the items.

  That is, the third matching unit 20 matches the recognition result of the character recognition unit 12, the input character string received by the character reception unit 14, and the word dictionary word in the second matching unit 19 of the second embodiment. The matching result, the input character string, and each word in the word dictionary are matched in the same way as selecting the matching result character string, and the matching cost of each combination of the recognition result, the input character string, and the word is matched. Is calculated. The cost calculation method is the same as that described with reference to FIGS.

  The preliminary setting unit 21 notifies the matching result selection unit 16 to set the recognition result that minimizes the cost calculated by the third matching unit 20 to the character string corresponding to the item of the input window 51.

  The matching result selection unit 16 receives the notification and sets the input character string selected as the recognition result in the field of the corresponding item in the input window 51.

  In the case of the eighth embodiment, the third matching unit 20 matches the character of the recognition result with the word in the word dictionary for each item, and the preliminary setting unit 21 selects the character string with the lowest cost.

  Then, the matching result selection unit 16 sets the input character string “Recognition Taro” in the field of the corresponding item of the input window 51 prior to input by the user.

  The second matching unit 19 matches the input character string selected by the preliminary setting unit 21 with the minimum cost, the input character string stored in the reception storage unit 13 by the character reception unit 14, and the word dictionary word. Then, an input character string to be set in the field as a matching result is selected, and the matching result character string is input to the field of the corresponding item in the input window 51. Thereby, the user can classify the character string of the recognition result into an appropriate item.

  As described above, according to the eighth embodiment, it is possible to reduce the effort of key input by the user as in the seventh embodiment.

Thus, according to the first to eighth embodiments described above, it is possible to provide a character input device and a character input method that can classify characters or character strings read from a form into appropriate items.
In addition, when the recognition result cannot be automatically matched with the input item, such as when the result of character recognition does not match the pre-listed words or formats, or when multiple items are listed as candidates, the user can enter a key. By adopting the input character string as a selection candidate, it becomes possible to smoothly associate the recognition result with the input item, and it is possible to efficiently correct the character string of the recognition result.

  That is, when confirming and correcting the character recognition result, correct characters can be efficiently input.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  Further, each component shown in the above embodiment may be realized by a program installed in a storage such as a hard disk device of a computer, and the program is stored in a computer-readable electronic medium: electronic media, The computer may realize the functions of the present invention by causing a computer to read a program from an electronic medium. Examples of the electronic medium include a recording medium such as a CD-ROM, flash memory, and removable media. Further, the configuration may be realized by distributing and storing components in different computers connected via a network, and communicating between computers in which the components are functioning.

  DESCRIPTION OF SYMBOLS 1 ... Scanner, 2 ... Keyboard, 3 ... Monitor, 10 ... Computer, 11 ... Image reception part, 12 ... Character recognition part, 13 ... Memory | storage part, 14 ... Character reception part, 15 ... 1st matching part, 16 ... Matching result Selection unit: 17 ... Character string setting unit, 18 ... Word dictionary storage unit, 19 ... Second matching unit, 20 ... Third matching unit, 21 ... Preliminary setting unit, 51 ... Input window, 52 ... List box.

Claims (10)

A character recognition unit that obtains a plurality of character strings as recognition results by performing character recognition processing on an image obtained by optically reading a form;
An accepting unit that has an item for classifying one or more character strings recognized by the character recognizing unit, and that accepts input of an input character string for each item;
A storage unit capable of storing the character string and the input character string in association with each other;
A first matching unit that matches the input character string received by the receiving unit with one or more character strings obtained by the character recognition unit;
A character input device comprising: a character string setting unit that stores the character string that matches the input character string in the storage unit in association with an item in which the input character string is input. A character recognition unit that obtains one or more character strings as a recognition result by performing character recognition processing on an image obtained by optically reading a form; and
A dictionary storage unit for storing a dictionary storing words;
An accepting unit that has an item for classifying one or more character strings recognized by the character recognizing unit, and that accepts input of an input character string for each item;
A storage unit capable of storing the character string and the input character string in association with each other;
A second matching unit that matches the input character string received by the receiving unit, one or more character strings obtained by the character recognition unit, and a word in the dictionary;
A character input device comprising: a character string setting unit that stores the character string that matches the input character string in the storage unit in association with an item in which the input character string is input. When a plurality of the character strings and / or words that match the input character string exist, a matching result selection unit that presents as a selection candidate,
The character string setting unit
3. The character input according to claim 1, wherein a candidate selected from the presented selection candidates is stored in the storage unit in association with an item to which the input character string is input. apparatus. The character string setting unit
If an unread character is included in the selected character string from among one or more character string selection candidates of the recognition result, the character string up to immediately before the unread character is replaced with the input character string. The character input device according to claim 1, wherein the character input device is a character input device. The first matching unit or the second matching unit is
4. The character input device according to claim 3, wherein character strings of selection candidates already selected for the same image are excluded from matching. The matching result selection unit
4. The character input device according to claim 3, wherein a character string of selection candidates already selected for the same image is excluded from the subject of presentation. A plurality of the dictionaries are provided for each input target item,
The second matching unit includes
The character input device according to claim 2, wherein a dictionary to be matched is switched and used for each group of images or for each item of a character string recognized from the image. A word dictionary storage unit for storing a word dictionary storing words for each input target item;
A third matching unit that performs matching with a word in the word dictionary for each character string of the recognition result recognized by the character recognition unit, and determines an item to which the character string of the recognition result belongs;
A preliminary setting unit that sets the character string of the recognition result in the item determined by the third matching unit prior to the input;
The character input device according to any one of claims 1 to 6, further comprising: In a character input method in a character input device having a character recognition unit, a reception unit, a first matching unit, a character string setting unit, and a storage unit,
The character recognition unit performs a character recognition process on an image obtained by optically reading a form to obtain one or more character strings as a recognition result,
The receiving unit having items for classifying one or more acquired character strings, respectively, accepts input of an input character string for each item;
The first matching unit matches the received input character string and one or more character strings obtained by character recognition,
The character input method, wherein the character string setting unit stores the character string matched with the input character string in the storage unit in association with an item to which the input character string is input. In a character input method in a character input device having a character recognition unit, a reception unit, a second matching unit, a character string setting unit, and a storage unit,
The character recognition unit performs character recognition processing on an image obtained by optically reading a form, thereby obtaining one or more character strings as a recognition result,
The accepting unit having items for classifying one or more character strings recognized by the character recognizing unit accepts input of an input character string for each item,
The second matching unit matches the received input character string with one or more character strings obtained by character recognition and a word in a dictionary storing words;
The character input method, wherein the character string setting unit stores the character string matched with the input character string in the storage unit in association with an item to which the input character string is input.

Images