JP2000020640A - Classification system, retrieval system, classification method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a classification system and the like, which requires a small amount of manual labor and can classify many documents. SOLUTION: A reading part 21 reads a slip and generates image contents data. A control part 22 applies character recognition to the image contents data, generates character contents data, prepares slip kind data and a key item and transmits those to a control server 11. On the other hand, a printing part 23 prints a two-dimensional code for indicating the key item on the slip concerned. The image contents data received by the control server 11 have the character recognition applied to by an OCR engine 12 and insufficient character contents data, and the slip kind data are generated. Also, the character contents data and the slip kind data are supplied by an entry terminal 14. The character contents data are stored in a data base 13 together with the key item and the image contents data. A sorter 15 reads the two-dimensional code printed on the slip, acquires the key item and assorts the slip according to the result retrieved from the data base 13 by using the key item.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a classification system,
More particularly, the present invention relates to a classification system, a retrieval system, and a classification method for sorting documents.

[0002]

2. Description of the Related Art As a method of classifying documents such as slips,
In recent years, a technique using an OCR (Optical Character Reader) has been adopted. The OCR has a function of reading an image on a printing surface of a document and discriminating characters and graphics printed on the printing surface (that is, performing character recognition).

[0003] However, in general, OCR is a method in which a character to be treated as a key for classification of a document is a character (OCR) having a predetermined characteristic.
Characters) or a document to be subjected to character recognition needs to follow a predetermined format.

[0004] For this reason, for example, when a form is classified using OCR, a method shown in FIG. 10 has been conventionally used. That is, first, the form to be classified is manually classified according to the size of the form or the like (step S1). Next, each of the classified forms is manually classified into a form (OCR form) that can be subjected to character recognition by OCR and another form (non-OCR form) for each of the forms belonging to the same classification. (Step S
2). Then, the non-OCR form is further classified manually by the operator after reading the contents of the non-OCR form, and then manually sorted into shelves or the like (step S3).

[0005] On the other hand, the OCR forms are manually classified according to a format or the like for each of those classified in the same classification in step S1 (step S4). Then, the OCR forms classified in step S4 are set in the OCR for each of those classified as belonging to the same classification in step S4, and the OCR receives character recognition for the OCR characters printed on the form (step S5). According to the result of character recognition, the data is further sorted by a sorter provided in the OCR (step S6).

[0006]

However, in the above-mentioned method, most of the steps are manually performed and the amount of work is enormous. Therefore, when the number of documents is large, the above method requires a long time to execute. For this reason, there is a demand for a method that requires a small amount of manual work and can classify a large amount of documents.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has as its object to provide a classification system, a retrieval system, and a classification method capable of classifying a large amount of documents with a small amount of manual work. I do.

[0008]

In order to achieve the above object, a classification system according to a first aspect of the present invention includes a reception center and a classification center connected to each other by a communication line, and the reception center includes a document center. A printing unit that prints the identification graphic indicating the document, a document reading unit that reads the document and creates image data representing an image of the document, and acquires the image data from the document reading unit. A character extracting unit that specifies a character included in the image of the document represented by the image data and creates character data representing the specified character, a code indicating the identification graphic, and the code identified by the code. Transmitting means for transmitting the image data representing the image of the document and the character data representing the characters contained in the image of the document to the classification center; Receiving means for receiving the code, the image data and the character data transmitted by the transmitting means from the transmitting means, and associating the code, the image data and the character data received by the receiving means with each other. A database that reads and stores the identification graphic printed on the document and obtains the code indicating the identification graphic; and a code reading unit that is associated with the code obtained by the code reading unit. Search means for searching for at least one of the character data and the image data from the database;
Classifying means for classifying the document on which the identification graphic read by the code reading means is printed, based on at least one of the character data and the image data retrieved by the search means. , Characterized in that.

According to such a classification system, image data representing an image of a document is obtained without preliminarily classifying the document according to a format or the like. Thereafter, the classification of the document is performed by comparing the identification graphic printed on each document with image data representing an image of the document on which the identification graphic is printed and character data extracted from the image data. Done. Therefore, the amount of manual work is reduced and a large amount of documents is classified.

[0010] The reception center includes means for generating a source identification code indicating that the code, the image data, and the character data transmitted by the transmission means are generated by itself. The means comprises means for transmitting the generator identification code to the classification center,
The receiving means includes means for receiving the generator identification code transmitted by the transmitting means, the database,
The receiving means includes means for storing the code received from the transmitting means, the image data, the character data, and the generation source identification code in association with each other, and the search means is obtained by the code reading means. Means for searching the database for the generator identification code associated with the code, wherein the classifying means is configured to read the code based on the generator identification code searched for by the search means. And means for classifying the document on which the identification graphic read by the method is printed. Thereby, the document is also classified based on the difference between the reception centers where the identification codes are printed.

Further, the classification system according to a second aspect of the present invention includes a code generating means for generating a code for identifying a document, a printing means for printing an identification graphic representing the code on the document, Reading, the document reading means to create image data representing the image of the document, and obtain the image data from the document reading means, to identify the characters contained in the image of the document represented by the image data, Character extracting means for creating character data representing the specified character; the code; the image data representing the image of the document identified by the code; and the character data representing a character included in the image of the document And a database for reading the identification graphic printed on the document and acquiring the code represented by the identification graphic. Retrieval means, retrieval means for retrieving at least one of the character data and the image data associated with the code acquired by the code reading means from the database, and retrieval means retrieved by the retrieval means. A classification unit that classifies the document on which the identification graphic read by the code reading unit is printed, based on at least one of the character data and the image data.

[0012] Even with such a classification system, image data representing an image of a document is obtained without preliminarily classifying the document according to a format or the like. Thereafter, the classification of the document is performed by comparing the identification graphic printed on each document with image data representing an image of the document on which the identification graphic is printed and character data extracted from the image data. Done. Therefore, the amount of manual work is reduced and a large amount of documents is classified.

The classification system obtains the image data and the character data representing the characters included in the image data, determines whether the character data includes a predetermined item, and determines whether the character data includes the character data. When it is determined that the character data is not included, a character included in the image of the document represented by the acquired image data is specified, and the character data representing the specified character is re-extracted. Alternatively, the classification system may acquire the image data and the character data representing the characters included in the image data, and determine whether the character data includes a predetermined item. And when it is determined that the image data is not included, an image display unit that displays the image of the document represented by the image data, and an input unit that inputs the predetermined item according to an instruction of an operator. It may be obtain things. As a result, even if the extraction of the character data has failed once, the character data is extracted again from the image data, so that the character data can be more reliably extracted.

[0014] The predetermined item includes information indicating the type of the document, and the classifying unit converts the identification graphic read by the code reading unit based on the information indicating the type of the document. There may be provided a means for classifying the printed document for each type. As a result, documents are also classified based on the type difference.

The code reading means includes means for scanning the identification graphic along one of a plurality of predetermined scanning directions to obtain the code, and the identification graphic is read by the code reading means. It may be composed of a two-dimensional code that causes the code reading means to acquire the code, even if it scans itself in any of the predetermined plurality of scanning directions. As a result, the identification graphic is read without aligning the direction of the document with the predetermined direction, and the code is obtained.
For this reason, the amount of work required for the classification of documents is further reduced, and the classification of a large amount of documents is performed efficiently.

The code reading means scans the two-dimensional code printed substantially at a predetermined distance from the edge of the rectangular document by scanning the code in a direction substantially parallel to the edge. Means for acquiring the, the printing means,
The two-dimensional code exists at a position substantially at the predetermined distance from each edge of the rectangular document, and substantially at the edge at the predetermined distance from the two-dimensional code. Means may be provided for printing a plurality of two-dimensional codes representing substantially the same code as each other so that the code is obtained by scanning the two-dimensional code in a parallel direction. Alternatively, the code reading means may place the two-dimensional code printed at a substantially predetermined distance from each of two edges contacting at a corner of the rectangular document on one of the two edges. Means for scanning in a substantially parallel direction to obtain the code, wherein the printing means comprises, for each corner of the rectangular writing, substantially from each of the two edges contacting at the corner. The two-dimensional code exists at a position separated by a predetermined distance, and the two-dimensional code is positioned in a direction substantially parallel to the two-dimensional code and one of the two edges separated by the predetermined distance. Means for printing a plurality of the two-dimensional codes representing substantially the same code as each other so that the code is obtained by scanning. As a result, the two-dimensional code is read along any edge or corner of the rectangular document, and the code is obtained. For this reason, the amount of work required for the classification of documents is further reduced, and the classification of a large amount of documents is performed efficiently.

If the printing means is provided with means for printing the identification figure on the document which does not substantially obscure the contents described in the document, visible information of the document is recorded. Even if the identification figure is printed over the existing portion, there is no possibility that the portion becomes invisible. Therefore, it is not necessary to avoid the part when printing the identification graphic. Therefore, the work of aligning the direction of the document to a predetermined direction to determine the position where the graphic for identification is printed becomes unnecessary, and the amount of work for classifying the document is further reduced. Done

Further, a retrieval system according to a third aspect of the present invention comprises a code generating means for generating a code for identifying an object to be classified, and acquiring information displayed by the object from the object. Content data creating means for creating content data representing the information, printing means for printing the identification graphic representing the code on the object, and information displayed on the object identified by the code and the code A database for acquiring the content data representing the identification data and storing them in association with each other; code reading means for reading the identification graphic printed on the object to obtain the code represented by the identification graphic; Searching means for searching the content data associated with the code obtained by the reading means from the database.

According to such a search system, content data representing information displayed by each object is obtained without classifying each object in advance. Then, the object is classified by comparing the identification graphic printed on each object with the content data representing the information displayed on the object on which the identification graphic is printed. Therefore, the amount of manual work is reduced and a large number of objects are classified.

The classification method according to a fourth aspect of the present invention includes a code generating step of generating a code for identifying a document, a printing step of printing an identification figure representing the code on the document, Reading a document to create image data representing the image of the document,
Acquiring the image data from the document reading step, specifying a character included in the image of the document represented by the image data, a character extracting step of creating character data representing the specified character, the code, The image data representing the image of the document identified by the code, and the character data representing the characters included in the image of the document are obtained, and a database stored in association with each other, and the database printed on the document A code reading step of reading the identification graphic and obtaining the code represented by the identification graphic; and at least one of the character data and the image data associated with the code obtained in the code reading step A search step of searching for the character data and the character data searched out by the search step. Based on at least one of the serial image data, the identification graphic read by the code reading step is characterized in that it comprises, a classification step of classifying the documents that have been printed.

According to such a classification method, image data representing an image of a document is obtained without preliminarily classifying the document according to a format or the like. Thereafter, the classification of the document is performed by comparing the identification graphic printed on each document with image data representing an image of the document on which the identification graphic is printed and character data extracted from the image data. Done. Therefore, the amount of manual work is reduced and a large amount of documents is classified.

Further, a computer-readable recording medium according to a fifth aspect of the present invention is a computer-readable recording medium comprising: printing means for printing, on a document, an identification graphic representing a code for identifying the document; A computer comprising: a document reading unit that creates image data representing a document image; and a code reading unit that reads the identification graphic printed on the document and acquires the code represented by the identification graphic. A code generation unit for generating the code, and the image data is obtained from the document reading unit, a character included in the image of the document represented by the image data is specified, and character data representing the specified character is obtained. The character extracting means to be created, the code, the image data representing the image of the document identified by the code, and the character representing the characters included in the image of the document. A database for acquiring character data and storing them in association with each other; and a search for searching at least one of the character data and the image data associated with the code acquired by the code reading means from the database. And means for recording a program for causing it to function.

A computer that executes a program recorded on such a recording medium acquires image data representing an image of a document without preliminarily classifying the document according to a format or the like. Then, the classification of the document is performed by comparing the identification graphic printed on each document with image data representing an image of the document on which the identification graphic is printed and character data extracted from the image data. Do. Therefore, the amount of manual work is reduced and a large amount of documents is classified.

According to a sixth aspect of the present invention, there is provided a computer-readable recording medium comprising: printing means for printing an identification figure indicating the document on a document; and an image representing the document image by reading the document. A document reading unit for creating data, the computer acquires the image data from the document reading unit, specifies characters included in the image of the document represented by the image data, and specifies the specified characters. Character extracting means for creating character data to be represented, a code indicating the figure for identification, the image data representing the image of the document identified by the code, and the character data representing a character included in the image of the document. A transmission means for transmitting to the outside and a program for functioning the transmission means are recorded.

A computer that executes a program recorded on such a recording medium acquires image data representing an image of a document without preliminarily classifying the document according to a format or the like. For this reason, the identification graphic printed on each document is later compared with the image data representing the image of the document on which the identification graphic is printed and the character data extracted from the image data to compare the identification graphic. If the classification is performed, the amount of manual work is reduced and a large amount of documents is classified.

According to a seventh aspect of the present invention, there is provided a computer-readable recording medium including a code for identifying a document, image data representing the image of the document identified by the code, and an image included in the image of the document. A computer having code reading means connected to transmitting means for transmitting character data representing the character to be read, reading the identification graphic printed on the document, and acquiring the code represented by the identification graphic. Means for receiving the code transmitted by the means, the image data and the character data from the transmitting means, and a database for storing the code, the image data and the character data received by the receiving means in association with each other Among the character data and the image data associated with the code acquired by the code reading unit. Search means for searches out from the database one even without a program for causing to function and characterized by recording.

A computer that executes a program recorded on such a recording medium acquires image data representing an image of a document to be classified and character data extracted from the image data. Then, the classification of the document is performed by comparing the identification graphic printed on each document with image data representing an image of the document on which the identification graphic is printed and character data extracted from the image data. Do. Therefore, the amount of manual work is reduced and a large amount of documents is classified.

[0028]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A classification system and a classification method according to an embodiment of the present invention will be described below with reference to the drawings, taking a sorting system for sorting forms as an example. FIG. 1 is a block diagram showing a physical configuration of the sorting system. As shown in the figure, the sorting system includes a centralized center 1 and a sales office system 2.

The centralized center 1 has a control server 11 and an OC
An R (Optical Character Reader) engine 12, a database 13, an entry terminal 14, and a sorter 15 are provided.

The control server 11 has an OCR engine 12,
It is connected to the database 13, the entry terminal 14, and the sorter 15, and is connected to an after-mentioned control unit 22 of the branch office system 2 by an INS (Information Network System).
They are connected to each other via a communication line such as a circuit network. The control server 11 acquires data to be described later from the branch office system 2 and stores the data acquired from the branch office system 2 and data supplied from the OCR engine 12 and the entry terminal 14 in the database 13 in accordance with the processing to be described later. I do.

The OCR engine 12 is connected to the control server 11. The OCR engine 12 controls the control server 1
1, character recognition is performed on image data described below supplied from the sales office system, a character string included in the image represented by the image data is extracted, and character data described below that represents the obtained character string is controlled. Supply it to the server 11.

The database 13 is connected to the control server 11 and stores data to be described later supplied from the control server 11 in accordance with an instruction from the control server 11. Then, the database 13 supplies the data stored therein to the sorting machine 15 in response to the access of the sorting machine 15.

The data stored in the database 13 is stored as a group for each data related to the same form. The data belonging to the same group is stored in the database 13 in the data structure shown in FIG. 2, for example. As illustrated, the data belonging to each group includes a key item, control information, and content data.

The key items are data for which the control server 11 searches the database 13 in the processing described later. In the key items, information shown as [1] to [4] below, that is, [1] date information D1 indicating the date on which the form was handled, and [2] the sales office system 2 handling the form are installed. Branch number D2 to identify the store
And [3] a device number D for identifying the branch system 2
3 and [4] after the branch system 2 is started,
And a serial number D4 indicating the order of the forms handled by the sales office system 2.

It should be noted that a plurality of sales office systems 2 may be installed in the same store. A group of data representing a form handled by each of the plurality of branch systems 2 installed in the same store includes the same branch number D2 and the different device numbers D3. .

The control information includes [5] processed display data D5, that is, information indicating whether or not the form represented by the set to which it belongs has already been processed by the sorter 15.

The content data is data obtained by the sales office system 2 from the printing surface of the form in the processing described later, and is data representing the content of the form. The content data further includes character data and image data.

The character data is information shown as [6] to [8] below, that is, [6] character data length data D6 representing the number of characters described in the form, and [7] the type of the form. Form data D7 to be represented and [8] character content data D8 representing characters described in the form.

It should be noted that the form type data D7 is, for example, when the form includes a delivery note, an invoice and a receipt, each form is any of a delivery note, an invoice and a receipt. Any information that indicates this is acceptable.

The image data is as follows [9] and [1]
0], that is, [9] image data length data D9 representing the size of image data representing the printing surface of the form
And [10] image content data D10 representing the image of the print surface of the form.

The entry terminal 14 is connected to the control server 11 and comprises a CRT (Cathode Ray Tube) or the like, and has a display device 14 connected to the control server 11.
1 and an input device 142 composed of a keyboard or the like and connected to the control server 11. The entry terminal 14 displays an image represented by the image content data D10 supplied from the control server 11 on the display screen of the display device 141, and displays information supplied to itself by an operator or the like operating the input device 142. To the control server 11.

The sorting machine 15 includes a two-dimensional code reading unit 1
51, a control unit 152, and a sorter 153. When the form is set on itself, the two-dimensional code reading unit 151 scans and reads a later-described two-dimensional code printed at a predetermined position on the form, and reads a key item represented by the two-dimensional code (ie, The above information [1] to [4]) is obtained and supplied to the control unit 152.

The control unit 152 has a CPU (Central Processes).
sing unit) and connected to the database 13. The control unit 152 controls the two-dimensional code reading unit 15
The database 13 is searched using the key item supplied from 1 as a search key, and the form set in the two-dimensional code reading unit 151 is classified according to a predetermined criterion based on the retrieved information. To sort the form according to the result of. The sorter 153 sorts the form set in the two-dimensional code reading unit 151 according to the instruction supplied by the control unit 152.

The office system 2 comprises a reading unit 21, a control unit 22, and a printing unit 23.

The reading section 21 is composed of a scanner or the like having a reading button. When the reading button is pressed while the form is set on the reading section 21, the printing surface of the set form is read and the read image is displayed. The image content data D10 is generated and supplied to the control unit 22.

The control unit 22 comprises a CPU and the like, and as described above, the control server 11 of the centralized center 1 via the communication line.
Are connected to the reading unit 21 and the printing unit 23. The control unit 22 performs character recognition on the image content data D10 supplied from the reading unit 21, extracts a character string included in the image represented by the image content data D10, and character content data D8 representing the obtained character string. Generate In addition, the control unit 22 performs the operations described in [1] to [10] above.
Among the information, the data other than the character content data D8 and the image content data D10 are also created, and the information [1] to [10] is supplied to the control server 11. Further, the control unit 22
The key items are derived from the date information D1, and the key items are supplied to the printing unit 23 to instruct printing of a two-dimensional code.

The printing unit 23 is composed of a printer or the like. In accordance with an instruction from the control unit 22, a two-dimensional code representing the key item supplied from the control unit 22 is provided at a predetermined position on the back of the form set in the reading unit 21. Is printed.

However, the printing unit 23 may perform printing on the front surface, not only on the back surface of the form. In this case, in order to prevent the form from being difficult to read due to printing of a two-dimensional code or the like, for example, printing (stealth printing) in a form substantially invisible to humans may be performed. Specifically, for example, printing is performed using an ink that transmits visible light of substantially all wavelengths (ie, is substantially transparent) while blocking ultraviolet light, or using ink that emits magnetism. Printing may be performed.

The two-dimensional code is a figure representing a key item as described above, and is scanned by the two-dimensional code reading unit 151 of the sorter 15, and the key item is read. The two-dimensional code reading unit 151 acquires a key item even when the two-dimensional code is scanned along any of a plurality of predetermined scanning directions.

For example, it is assumed that the two-dimensional code is such that the key item can be acquired when the two-dimensional code reader 151 scans in two scanning directions perpendicular to each other. In addition, the two-dimensional code reading unit 151
By scanning a form along any edge of a set rectangular form, a two-dimensional code printed at a position separated from the edge by a predetermined distance (hereinafter, referred to as distance A) is scanned. It is assumed that

In this case, the printing unit 23 of the office system 2
Prints a two-dimensional code on the upper right and lower left of the form as shown in FIG. 3, for example. The two-dimensional code printed on the upper right is separated from the upper edge and the right edge of the form by a distance A, and the two-dimensional code reading unit 151
Thus, when scanning is performed in a scanning direction parallel to either the upper edge or the right edge of the form, the two-dimensional code reading unit 151 acquires key items. Further, the two-dimensional code printed at the lower left is set at a distance A from both the left edge and the lower edge of the form, and the two-dimensional code reading unit 151 places the two-dimensional code on either the left edge or the lower edge of the form. When scanning is performed in the parallel scanning direction, the two-dimensional code reading unit 1
Let 51 obtain key items.

Alternatively, the printing unit 23 prints a two-dimensional code at the four corners of the form, for example, as shown in FIG. The two-dimensional code printed at each corner is separated from the two edges contacting at the corner by a distance A, and the two-dimensional code reading unit 151 outputs one of these two edges. When the scanning is performed in the scanning direction parallel to the above, the key item is acquired by the two-dimensional code reading unit 151.

Thus, the two-dimensional code reading unit 15
No. 2 scans any two-dimensional code printed on the form, even when scanning is performed along any edge of the form.

Next, the operation of this sorting system will be described with reference to FIGS.
This will be described with reference to FIG. FIG. 5 is a flowchart showing the processing of the sales office system 2. FIG. 6 is a flowchart showing the processing of the control server 11. FIG.
5 is a flowchart illustrating a process of an R engine. FIG. 8 is a flowchart showing the processing of the entry terminal 14. FIG. 9 is a flowchart showing the processing of the sorter 15.

(Processing of the branch office system) First, when the branch office system 2 is started, the reading section 21 of the branch office system 2 is started.
Waits until a form to be classified is set and the read button is pressed (FIG. 5, step S101).
Then, the reading unit 2 is operated by the operator of the office system 2 or the like.
1 and the reading button is pressed (step S102), the reading unit 21 scans the set form and reads the image content data D1 representing the print surface of the form.
0 is generated and supplied to the control unit 22 (step S10).
3).

The control section 22 supplied with the image content data D10 determines the size of the supplied image content data D10 and generates image data length data D9 (step S1).
04) Then, the supplied image content data D10 is subjected to character recognition, and it is determined whether or not the image content data D10 includes a character string (step S105).

When it is determined that the data is not included, the control unit 2
2 moves the process to step S110. On the other hand, when determining that the image content data is included,
Then, character content data D8 representing a character extracted by performing character recognition on 10 is created (step S106).

Next, the control unit 22 creates character data length data D6 by counting the number of characters included in the character content data D8 obtained in step S106 (step S10).
7) Further, it is determined whether or not there is a character string representing the type of the form in the character content data D8 (step S10).
8).

When it is determined that the data is not included, the control unit 2
2 moves the process to step S110. On the other hand, when it is determined that the character content data is included,
8, a part representing the form type is extracted, and form type data D7 is created (step S109).

Next, the control unit 22 generates information belonging to the above-mentioned key items, that is, date information D1, branch office number D2, device number D3, and serial number D4 (step S11).
0). Then, the key items and the content data generated by the processing up to step S110 are organized into one group, and the organized data group is transmitted to the control server 11 via the communication line. (Step S111). The data transmitted to the control server 11 is stored in the database 13 according to the processing of the centralized center 1 described later.

Next, the control unit 22 supplies the key item generated in step S110 with the latest date information D1 to the printing unit 23, and instructs printing of a two-dimensional code (step S112). In response to the instruction, the printing unit 23 prints the two-dimensional code representing the key item supplied in step S112 on the set form, and discharges the form (step S113).

By the processing described above, the print surface of the form is read and character recognition is performed, and the obtained contents are sent to the centralized center 1 together with key items. A two-dimensional code representing a key item is printed on the form. The form printed with the two-dimensional code is sent to the place where the centralized center 1 is installed. It is not necessary to physically sort the form prior to sending.

(Processing of Centralized Center) When the centralized center 1 is started, the control server 11 of the centralized center 1 connects to the sales office system 2 via the communication line, and transmits from the sales office system 2 in the above step S111. Wait for the supplied data group to be supplied (FIG. 6, step S20)
1).

When the control server 11 receives the data group transmitted from the branch office system 2 in step S111 (step S202), the control server 11 includes both the form type data D7 and the character content data D8 in the received data. Is determined (step S20).
3). If it is determined that both are included, the control server 11 advances the processing to step S210.

Form type data D7 and character content data D
8 is missing, the control server 11 converts the received data into the OCR engine 12.
To instruct the OCR engine 12 to perform character recognition again (step S204).

The OCR engine 12 receives the instruction for character recognition from the control server 11 and, when the control server 11 receives the supply of the data received from the sales office system 2, the image content data D 10 included in the supplied data contains the character. Recognition is performed, and character content data D8 is extracted from the supplied data (FIG. 7, step S301).

Next, the OCR engine 12 proceeds to step S
At 301, it is determined whether or not the character content data D8 has been extracted (step S302). When it is determined that the character data could not be extracted, the OCR engine 12 supplies a message to the effect that the character content data D8 could not be extracted to the control server 11 (step S303). On the other hand, if it can be extracted,
The OCR engine 12 extracts the character content data D extracted by itself.
Then, it is determined whether or not 8 includes a part indicating the type of the form (step S304).

When it is determined that the data is not included, the OCR engine 12 supplies a message to the effect that the form type data D7 could not be created to the control server 11, and further controls the character content data D8 extracted in step S301. The data is supplied to the server 11 (step S305).

On the other hand, if the OCR engine 12 determines that a part indicating the form type is included,
By extracting the part representing the type of form from 8,
The form type data D7 is created and supplied to the control server 11 together with the character content data D8 (step S306).

On the other hand, the control server 11 determines in step S30
3. When the information is supplied from the OCR engine 12 in S305 or S306 (step S205), the form type data D7 and the character content data D8 supplied from the OCR engine 12 and the data transmitted from the sales office system 2 are combined. Then, the data is added to the group of data received from the office system 2 in step S202 (step S206).

However, the form type data D7 or the character content data D8 is supplied from the branch office system 2, and the form type data D7 or the character content data D8 is supplied from the OCR engine 12 as well. If the character content data D8 is duplicated, the one supplied from the branch system 2 is discarded. Specifically, for example, when character content data D8 is supplied from both the branch office system 2 and the OCR engine 12, the control server 11 discards the character content data D8 supplied from the branch office system 2. .

Next, the control server 11 proceeds to step S20
It is determined whether or not both the form type data D7 and the character content data D8 are included in the data group that has undergone the processing in S6 (step S207). And
If it is determined that both are included, the control server 11 advances the processing to step S210. On the other hand, if it is determined that at least one of the form type data D7 and the character content data D8 is missing, the control server 11 determines that the image content data D10 and the missing form type data D7 and character content data D8 are missing. The information indicating which one is used is supplied to the entry terminal 14 (step S20).
8).

Display device 141 of entry terminal 14
When the data is supplied from the control server 11, the image of the print surface of the form represented by the supplied image content data D10 is displayed (FIG. 8, step S401). In addition, the display device 141 displays an image (for example, “character content data” or “form document data”) indicating which of the form type data D7 and the character content data D8 is missing from the data group that has undergone the processing of step S206 and needs to be supplemented. Is displayed on the display screen (step S40).
2).

The operator of the entry terminal 14 visually recognizes the displayed image, determines the contents of the information to be supplemented, and transmits the determined information to the input device 142 of the entry terminal 14.
(Step S40).
3). The entry terminal 14 stores the supplied information. Specifically, for example, when it is displayed that the information to be supplemented is character content data D8, the entry terminal 1
4 reads a character string included in the displayed image of the form and converts the read character string into the entry terminal 14.
To supply. The information to be supplemented is the form type data D
When "7" is displayed, the operator of the entry terminal 14 determines the type of the form from the displayed form image, and supplies the determination result to the entry terminal 14.

When the operation of supplying the information to be supplemented to the entry terminal 14 is completed, the operator or the like of the entry terminal 14 uses the input device 142 of the entry terminal 14 to
The completion of the work is notified to the entry terminal 14 (step S
404). The entry terminal 14 to which the notification of the completion of the work has been supplied, inputs the information supplied to itself as information to be supplemented, into the form type data D7 or the character content data D8.
Control server 1 together with information for identifying
1 (step S405).

When the control server 11 is supplied with information from the entry terminal 14 (Step S209), the control server 11 executes Step S209.
The process proceeds to 210. In step S210, the control server 11 adds the processed display data D5 to the group of data received from the branch office system 2 in step S202 and processed through step S209, and the data included in this group is The data is organized so as to form the data structure shown in FIG.

Then, the control server 11 proceeds to step S2
The group reorganized in step 10 is supplied to the database 13, and the database 13 stores the supplied group (step S211).

On the other hand, when the operator or the like of the centralized center 1 is supplied with the form on which the two-dimensional code is printed by the sales office system 2, the operator sets the form on the two-dimensional code reader 151 of the sorting machine 15. (FIG. 9, step S50)
1). The two-dimensional code reading unit 151 reads the two-dimensional code printed on the set form,
The key item attached to the form is acquired, and the acquired key item is supplied to the control unit 152 (step S502). The control unit 152 searches the database 13 using the obtained key item as a search key, and searches for a group including the key item (step S503).

The control unit 152 instructs the sorter 153 to sort the set forms based on the information included in the group found in step S503 (step S504). The sorter 153 sorts the forms set in the two-dimensional code reading unit 151 according to the instruction of the control unit 152. Specifically, for example, forms having substantially the same contents of form type data are put into the same container (not shown).

By the processing described above, the centralized center 1 sorts the forms. The configuration of this sorting system is not limited to the above. For example, database 13
2 need not have the data structure shown in FIG. 2 as long as the data for identifying the form and the information representing the image of the form identified by the data are stored in association with each other. Is optional. Also, the form type data D7 may represent a type classified according to an arbitrary standard, and is not limited to information indicating whether each form is a delivery note, an invoice, or a receipt. .

The number of the branch office systems 2 is not limited to one, but may be plural. In this case, the centralized center 1
The data supply source may be identified by determining which sales office system 2 indicates the sales office number D2 included in the information transmitted by each sales office system 2 to the centralized center 1. Good. And the sorting machine 15
For example, it is determined that the forms indicated by the data whose supply sources are identical to each other are the ones in which the printing of the two-dimensional code is performed by the same branch system 2 and
These forms may be put in the same container.

The sorter 15 does not need to physically sort the forms, and may output data representing the result of classifying the forms according to a predetermined standard.

Further, the centralized center 1 and the sales office system 2 do not need to be arranged in separate stores or the like, and may be installed in the same building, for example. The centralized center 1 and the branch office system 2 need not be connected by a dedicated communication line, but may be connected via a network such as the Internet. In this case, the data transmitted from the branch system 2 to the centralized center 1 may be transmitted in the form of e-mail, for example.

Further, the centralized center 1 and the branch system 2 need not be connected to each other. In this case, for example, both the centralized center 1 and the branch system 2 are MO
(Magneto-Optical) A device for accessing a recording medium such as a disk is provided. The branch office system 2 writes the above information [1] to [10] on the recording medium.
The information [1] to [10] written on the recording medium may be read.

Further, the number of entry terminals 14 in the centralized center 1 is not limited to one, and there may be a plurality of them, and the centralized center 1 may not have the entry terminals 14. Also, the entry terminal 14 stores the missing form type data D
7 and the character content data D8, the information already supplied from the branch office system 2 or the OCR engine 12 may be changed, or the database 1 may be changed.
3 may be changed.

Further, all the data groups to be stored in the database 13 may be temporarily displayed by the entry terminal 14 and checked by the operator or the like. In this case, for example, after the operator or the like checks the data belonging to the group, the entry terminal 14 is operated to add the processed display data D5 to the data. The attached data group may be stored in the database 13.

The branch office system 2 does not need to perform character recognition on the image content data D10. In this case, for example, the OCR engine 12 may perform character recognition on all the image content data D10 supplied from the sales office system 2.

The form type data D7 does not need to be extracted from the character content data D8. For example, the control unit 22 of the office system 2 or the OCR engine 12 prints the form based on the image content data D10. The form type may be determined based on the shape of the frame in the description column on the surface, and the form type data D7 representing the determination result may be created.

The object to be sorted by the sorting system is not limited to a form, but may be any two-dimensional code, a bar code, or any other object as long as a figure for identifying the object can be printed. is there.

Although the embodiment of the present invention has been described above, the classification system of the present invention can be realized by using a general computer system without using a dedicated system. For example, a medium (floppy disk, CD-ROM, or the like) storing a program for executing the above operation in a server computer including a scanner and a printer
By installing the program from, a classification system that executes the above-described processing can be configured.

The medium for supplying the program to the computer may be a communication medium (a medium that temporarily and fluidly holds the program, such as a communication line, a communication network, or a communication system). For example, the program may be posted on a bulletin board (BBS) of a communication network and distributed via the network. Then, by starting this program and executing it in the same manner as other application programs under the control of the OS, the above-described processing can be executed.

If the OS shares part of the processing,
Alternatively, when the OS constitutes a part of one component of the present invention, the recording medium may store a program excluding the part. Also in this case, in the present invention, it is assumed that the recording medium stores a program for executing each function or step executed by the computer.

[0093]

As described above, according to the present invention, a classification system and a classification method capable of classifying a large amount of documents with a small amount of manual work are realized.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a physical configuration of a sorting system according to an embodiment of the present invention.

FIG. 2 is a diagram schematically showing a data structure of data stored in a database.

FIG. 3 is a diagram schematically showing a position where a two-dimensional code is printed.

FIG. 4 is a diagram schematically showing a position where a two-dimensional code is printed.

FIG. 5 is a flowchart illustrating a process of a sales office system.

FIG. 6 is a flowchart illustrating a process of a control server.

FIG. 7 is a flowchart illustrating a process of the OCR engine.

FIG. 8 is a flowchart illustrating a process of an entry terminal.

FIG. 9 is a flowchart showing processing of the sorting machine.

FIG. 10 is a flowchart showing a conventional method for classifying forms.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Centralized center 11 Control server 12 OCR engine 13 Database 14 Entry terminal 141 Display device 142 Input device 15 Sorting machine 151 Two-dimensional code reading part 152 Control part 153 Sorter 2 Sales office system 21 Reading part 22 Control part 23 Printing part

Claims (15)

[Claims] A reception center and a classification center connected to each other by a communication line, wherein the reception center prints on a document an identification graphic indicating the document, and reads the document to read the document. A document reading unit that creates image data representing the image of the image, acquiring the image data from the document reading unit, specifying a character included in the document image represented by the image data, and specifying the specified character Character extraction means for creating character data to be represented, a code indicating the identification graphic, the image data representing the image of the document identified by the code, and the character data representing a character included in the image of the document. A transmitting unit that transmits the code, the image data, and the character data transmitted by the transmitting unit to the classification center. A receiving unit that receives from the transmitting unit; a database that stores the code, the image data, and the character data received by the receiving unit in association with each other; and reads the identification graphic printed on the document, A code reading unit for obtaining the code indicating the identification graphic; and a search for searching at least one of the character data and the image data associated with the code obtained by the code reading unit from the database. Means for classifying the document on which the identification graphic read by the code reading means is printed, based on at least one of the character data and the image data retrieved by the search means; A classification system comprising: 2. The reception center includes means for generating a source identification code indicating that the code, the image data, and the character data transmitted by the transmission means are generated by itself. The transmitting unit includes a unit that transmits the source identification code to the classification center; the receiving unit includes a unit that receives the source identification code transmitted by the transmitting unit; Means for storing the code received from the transmitting means, the image data, the character data, and the generation source identification code in association with each other, and wherein the search means obtains the code obtained by the code reading means. Means for retrieving the generator identification code associated with the search means from the database; The apparatus according to claim 1, further comprising: a unit configured to classify the document on which the identification graphic read by the code reading unit is printed, based on the retrieved source identification code. Classification system. 3. A code generating means for generating a code for identifying a document, a printing means for printing an identification graphic representing the code on the document, reading the document and converting image data representing an image of the document. A document reading unit to be created; a character that acquires the image data from the document reading unit, specifies a character included in the image of the document represented by the image data, and creates character data representing the specified character. An extraction unit, the code, the image data representing the image of the document identified by the code, and the character data representing the characters included in the image of the document, and a database that stores the data in association with each other; A code reading unit that reads the identification graphic printed on the document and obtains the code represented by the identification graphic; Searching means for searching at least one of the character data and the image data associated with the obtained code from the database; and at least one of the character data and the image data searched for by the searching means. A classification unit that classifies the document on which the identification graphic read by the code reading unit is printed based on one of the two. 4. Acquiring the image data and the character data representing the character included in the image data, and determining whether or not a predetermined item is included in the character data. And determining the characters included in the image of the document represented by the acquired image data, and re-extracting the character data representing the identified characters. Item 4. The classification system according to Item 1, 2 or 3. 5. The image data and the character data representing the character included in the image data are obtained, and it is determined whether or not a predetermined item is included in the character data. 2. An image display unit that displays the image of the document represented by the image data when the image data is determined, and an input unit that inputs the predetermined item according to an instruction of an operator. The classification system according to any one of claims 1 to 4. 6. The identification information read by the code reading means based on the information indicating the type of the document, wherein the predetermined item includes information indicating the type of the document. The classification system according to claim 4, further comprising: a unit configured to classify the document on which a graphic is printed for each type. 7. The code reading means includes means for scanning the identification graphic along any one of a plurality of predetermined scanning directions to obtain the code, and wherein the identification graphic reads the code. 7. A two-dimensional code for causing the code reading means to acquire the code even if the means scans itself in any of the predetermined plurality of scanning directions. Classification system described in section. 8. The code reading means scans the two-dimensional code printed substantially at a predetermined distance from an edge of the rectangular document in a direction substantially parallel to the edge. Means for acquiring the code, the printing means, the two-dimensional code exists at a position substantially the predetermined distance from each edge of the rectangular writing, and the two-dimensional code and A plurality of the two symbols representing substantially the same code as each other, such that the code is obtained by scanning the two-dimensional code in a direction substantially parallel to the edge at the predetermined distance. The classification system according to claim 7, further comprising means for printing a dimension code. 9. The two-dimensional code which is printed at a substantially predetermined distance from each of two edges contacting each other at a corner of the rectangular writing surface, and wherein the code reading means reads the two-dimensional code. Means for scanning in a direction substantially parallel to one of the two sides to obtain the code, wherein the printing means comprises: for each corner of the rectangular writing, substantially from each of two edges contacting at the corner. The two-dimensional code exists at a position separated by the predetermined distance, and the two-dimensional code is positioned in a direction substantially parallel to one of the two edges separated by the predetermined distance. The apparatus according to claim 7, further comprising: means for printing a plurality of the two-dimensional codes representing substantially the same code as each other so that the code is obtained by scanning the two-dimensional code. Classification system. 10. The printing apparatus according to claim 1, wherein said printing means includes means for printing the identification graphic which does not substantially conceal the content written on the document on the document. The classification system according to claim 1. 11. Code generation means for generating a code for identifying an object to be classified, content information creation for acquiring information displayed by the object from the object, and creating content data representing the information. Means, a printing means for printing an identification figure representing the code on the object, acquiring the content data representing the information displayed on the object identified by the code and the code, and associated with each other A database to be stored; code reading means for reading the identification graphic printed on the object to obtain the code represented by the identification graphic; and a code reading means corresponding to the code obtained by the code reading means. A search unit for searching the content data from the database. 12. A code generating step of generating a code for identifying a document, a printing step of printing an identification figure representing the code on the document, reading the document and converting image data representing an image of the document. A document reading step to be created, and the image data is obtained from the document reading step, a character included in the image of the document represented by the image data is specified, and a character is generated to generate character data representing the specified character. An extraction step, the code, the image data representing the image of the document identified by the code, and the character data representing the characters included in the image of the document, and a database for storing them in association with each other; A code reading step of reading the identification graphic printed on the document and acquiring the code represented by the identification graphic; A retrieval step of retrieving at least one of the character data and the image data associated with the code acquired by the code reading step from the database; and the character data retrieved by the retrieval step and the A classification step of classifying the document on which the identification graphic read by the code reading step is printed, based on at least one of the image data. 13. A printing means for printing, on a document, an identification figure representing a code for identifying the document, a document reading means for reading the document and creating image data representing an image of the document, A code reading unit that reads the identification graphic printed on the computer and obtains the code represented by the identification graphic. A code generation unit that generates the code. Character extraction means for acquiring data, identifying characters included in the image of the document represented by the image data, and creating character data representing the identified characters, the code, the code identified by the code A database configured to acquire the image data representing the image of the document and the character data representing the characters included in the image of the document, and store them in association with each other; Searching means for searching at least one of the character data and the image data associated with the code obtained by the code reading means from the database; and reading by a computer recording a program for causing the computer to function. Possible recording medium. 14. A computer comprising: a printing unit that prints an identification graphic indicating the document on a document; and a document reading unit that reads the document and creates image data representing an image of the document. Means for acquiring the image data from the means, identifying characters included in the image of the document represented by the image data, and creating character data representing the identified characters; and indicating the identification graphic. A code for transmitting a code, the image data representing the image of the document identified by the code, and the character data representing the characters included in the image of the document to the outside; Computer readable recording medium. 15. The document connected to a transmitting means for transmitting a code for identifying a document, image data representing the image of the document identified by the code, and character data representing characters included in the image of the document. A computer having code reading means for reading the identification graphic printed on the computer and obtaining the code represented by the identification graphic, transmitting the code, the image data and the character data transmitted by the transmission means to the computer; Receiving means for receiving from the means, a database storing the code, the image data and the character data received by the receiving means in association with each other, and being associated with the code obtained by the code reading means A search unit for searching at least one of the character data and the image data from the database; A computer-readable recording medium that stores a program for causing