JP2004240888A - Documents identification and classification method, documents identification and classification processing system and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately and efficiently identify and classify document types by not disarranging a reading order when identifying and classifying. <P>SOLUTION: An OCR device 1 successively and optically reads written information of documents to be delivered and determines pockets of conveying/discharging destinations of the documents based on classification of identified document types. Written information of a document whose document type cannot be identified in the OCR device 1 is transferred to a control device 2 and displayed on a screen. Based on the written information, an operator identifies a document type thereof for classification and instructs a pocket of the conveying/discharging destination on the display screen. A currently read document is conveyed/discharged to a document separately holding device 3 through a conveying device 4. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention describes, for example, the contents of documents such as medical specifications and invoices (appropriately referred to as “recept”) as an optical character reader (OCR: Optical Character Reader, here appropriately referred to as “OCR device”). ) To identify the type of the document from the written information of the document (also referred to as analysis as appropriate), and a document identification and classification method, a document identification and classification processing system, and a program for classifying the storage unit of the transport / discharge destination. About.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, the following examples are known as examples in which the content of description of various documents (hereinafter, also referred to as document description information) is read by an OCR device and the type of the document is identified and classified.
[0003]
FIG. 15 is a diagram for explaining an example of identifying and classifying a document type in a conventional OCR device.
(A) The document is placed on an automatic feeder of the OCR device (referred to as "ADF device" as appropriate), and automatic reading of the written information is started (FIG. 15 (1)).
(B) The read document description information is read from the read document description information in order to identify the type of the document in units of one document, for example, in a server on a local area network (LAN). Is found out (FIG. 15 (1)).
(C) Next, for example, the type of the document is identified by comparing it with character data registered in advance (for example, characters of “invoice” and “invoice”) (FIG. 15 (2) ).
(D) Based on the identification of the type of the document, the document is sorted and transported and stored for each document in a storage device (denoted as a pocket as appropriate) outside the OCR device (FIG. 15 (3) pocket A, ( 4) Pocket B, (5) Pocket C / Reject pocket).
(E) If data for identifying the type of document cannot be found, the data is automatically transported and stored in a non-identifiable storage device outside the OCR device (FIG. 15 (5) pocket C / reject pocket).
(F) Documents stored in the indistinguishable storage device (pocket C / reject pocket) are manually classified and classified into one of pockets A and B, and stored in the storage device (pockets A and B). ), It is inserted and arranged at the end of the bundle of documents stored earlier.
[0004]
As described above, conventionally, the types of documents are identified and classified manually and manually.
[0005]
[Patent Document 1]
Japanese Patent Application Laid-Open No. 8-272883, "Form processing method and system" (Claim)
[0006]
[Problems to be solved by the invention]
However, in the above-described conventional example, the classification of a document whose type cannot be identified is manually confirmed, and the document is inserted and arranged at the end of the previously stored document bundle. That is, the type of document in the OCR device and the order of classification by manual classification are mixed. There has been a problem that the reading order of the documents in the OCR device and the order of the documents stored in the storage device (for example, a month order or a date order) change, and the reading continuation is different.
[0007]
The present invention has been made in view of the above circumstances, it is possible to read the written content of the document with an optical character reading device, to identify the type of the document, when sorting, so that the reading order is not different, As a result, an object of the present invention is to provide a document identification / classification method, a document identification / classification processing system, and a program, which enable accurate and efficient identification and classification of documents.
[0008]
[Means for Solving the Problems]
The document identification and classification method of the present invention that achieves the above object uses an optical character reader, sequentially reads the written information of the document, identifies the document type, and conveys and discharges the classified document to a predetermined storage unit. The process of automatically sending out documents in sequence and the process of optically reading the written information of the documents to be sent out, identifying the document type, and storing the documents to be transported and discharged according to the predetermined classification. In the step of deciding a copy and in determining the storage section of the transport / discharge destination of the document, the document description information of the document whose document type cannot be identified is displayed on a screen, and based on the identification of the document type from this screen information. The method includes a step of instructing a storage unit to transport and discharge the classified documents, and a step of transporting and discharging the currently read document to the determined storage unit or the specified storage unit.
[0009]
According to the document identification and classification method of the present invention, the contents of the document are read by an optical character reader, and the data of the document that cannot be identified when the type of the document is identified and classified (image data or optical character recognition). The character data) is displayed on the screen when the identification (analysis) is impossible, and the document is discharged and conveyed to the sorting destination pocket by an instruction operation based on the sorting judgment of the operator. That is, the reading order of the documents is not changed.
[0010]
In this way, the reading order of the documents placed for reading and the order of transport and discharge of the classifications match, and as a result, accurate and efficient identification and classification of the documents becomes possible.
[0011]
The document identification and classification method of the present invention that achieves the above object uses an optical character reader, sequentially reads the written information of the document, identifies the document type, and conveys and discharges the classified document to a predetermined storage unit. The process of automatically sending out the documents sequentially, and the optically reading the written information of the documents to be sent out sequentially to identify the type of the document, and the destination of the document to be transported and discharged according to the predetermined classification. A step of deciding the storage unit, a step of automatically re-reading the description items of the unidentifiable document of the document type when the document type cannot be identified by the determination, and a step of automatically re-executing Automatically instructing the storage unit of the transport / discharge destination of the documents classified based on the identification of the document type by the appropriate reading process, and transporting / reading the currently read document to the determined storage unit or the specified storage unit. Exhaustion Characterized by a step of.
[0012]
According to the document identification / classification method of the present invention, the content of a document is read by an optical character reader, and when the document is identified and classified, the information items of the unidentifiable document are automatically read again. Has been implemented.
[0013]
For example, when the type of a document cannot be identified, it is displayed on the screen, and the operator does not discharge or convey the document to the sorting destination pocket by the instruction operation based on the classification determination (identification determination). The reading order is the same as the transport / discharge order of the classification. In other words, as a result, accurate and more efficient classification by identification of document types becomes possible.
[0014]
The document identification and classification processing system of the present invention that achieves the above object sequentially reads the written information of the document, identifies the document type, and transports and discharges the classified document to a predetermined storage destination. The automatic sending means for automatically sending out the document automatically, and the document description information of the document sent out by the automatic sending means is optically read in order to identify the type of the document, and the storing / receiving section of the document transfer / discharge destination corresponding to the predetermined classification is provided. The reading identification / classification means to be determined, the display means for displaying on the screen the written information whose document type cannot be identified by the reading identification / classification means, and the transport of the document by the classification in which the document type is identified based on the written information on the display means. Instruction means for instructing the storage unit to be discharged, and conveyance for discharging / discharging the currently read document to the storage unit determined by the read identification and classification means or the storage unit designated by the instruction means. Characterized in that it comprises left and means.
[0015]
The document identification and classification processing system of the present invention that achieves the above object sequentially reads the written information of the document, identifies the document type, and transports and discharges the classified document to a predetermined storage destination. An automatic sending means for automatically sending out the document sequentially, and a storage unit for a document carrying / discharging destination corresponding to a predetermined classification by identifying the document type by optically reading the document description information of the document sent out by the automatic sending means. And automatically performs the process of automatically reading the items described in the document whose document type has become indistinguishable by the read identification and classification means, and determines the transport / discharge destination based on the classification based on this process. Instruction means for automatically instructing the storage unit, and transport / discharge means for transporting / discharging the currently read document to the storage unit determined by the read identification / classification means or the storage unit designated by the instruction means. Characterized in that it comprises.
[0016]
In the document identification / classification processing system according to the present invention, similarly to the above-described document identification / classification method, the reading order can be prevented from being different, and as a result, accurate and efficient classification by document identification becomes possible. .
[0017]
In order to achieve the above object, the program according to the present invention includes a step of automatically sending out a document sequentially, a step of sequentially reading optically the written information of the document to be sent, identifying a document type, and transporting a document corresponding to a predetermined classification. A step of determining a storage section to be discharged;
When deciding the storage section to which the document is transported and discharged, the document description information of the document whose document type is unidentifiable is displayed on the screen, and the document type is identified based on this screen information, and the classified document is transported and discharged. This causes the computer to execute the control of the step of instructing the storage unit to be executed and the step of transporting and discharging the currently read document to the determined storage unit or the specified storage unit.
[0018]
Further, the program of the present invention includes a step of automatically sending out a document sequentially, a step of sequentially reading optically the written information of the document to be sent, identifying the type of the document, and carrying the document in accordance with a predetermined classification. A step of deciding a storage section of a discharge destination, and, if the document type is unidentifiable by the determination, a step of automatically re-reading the items described in the document whose document type is unidentifiable; and Automatically instructing the storage unit of the transport / discharge destination of the classified and identified document type based on the automatic reading process performed, and reading the storage unit to the determined storage unit or the specified storage unit this time. The control of the steps of transporting and discharging the document is performed by a computer.
[0019]
According to the program of the present invention, the present invention can be provided through download / installation from an information recording medium (eg, package software) or a communication network. Therefore, the present invention can be easily implemented by microcomputers mounted on various devices, and the versatility is improved.
[0020]
The following is a preferred embodiment of the technical concept of the document identification and classification method of the present invention.
[0021]
The screen display is a part or all of the read written information.
[0022]
Further, as a process of automatically reading the items described in the document of the above-mentioned document type that cannot be identified again, a process by a plurality of character recognition software is performed.
[0023]
Further, as a process of automatically reading the items described in the document whose document type is unidentifiable again, a process of retrieving all data of the written document reading information and identifying the document type is performed.
[0024]
In addition, as a process for automatically reading the items of the document whose document type is indistinguishable again, all the data of the written document reading information is taken in, the characters in this data are compared with the characters registered in advance, and the document is compared. Perform the process of identifying the type.
[0025]
The following is a preferred embodiment of the technical concept of the document identification and classification processing system of the present invention.
[0026]
As a process for automatically re-reading the description items of a document whose document type cannot be identified, (a) a process using a plurality of character recognition software is performed, and (b) all data of the read document written information is taken in. And (c) taking in all the data of the written and written information of the document, comparing the characters in the data with the characters registered in advance, and performing the process of identifying the document type. Further, there is provided means for selectively setting any one of (a), (b), and (c) for execution.
[0027]
In addition, as a process for automatically reading the items described in the document whose document type is unidentifiable again, (a) a process using a plurality of character recognition software is performed, and (b) all data of the read document written information is read. (C) Importing all data of the written and written information of the document, and comparing the characters in the data with the characters registered in advance to identify the document type. And a means for continuously processing (a), (b) and (c) to identify a document type in any of (a), (b) and (c).
[0028]
The following is a preferred embodiment of the technical concept of the program of the present invention.
[0029]
(A) performing a process using a plurality of character recognition software as a process for automatically re-reading the description items of a document whose document type is indistinguishable; (C) taking in all the data of the written and written information of the document and comparing the characters in the data with the characters registered in advance to identify the document type; And (a), (b), and (c) are selectively set and executed, or (a), (b), and (c) are continuously performed. Processing is performed to cause the computer to execute a control of identifying a document type in any one of (a), (b), and (c).
[0030]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings, but the configuration and signal processing are schematically shown to the extent that the present invention can be understood.
[0031]
Hereinafter, a preferred configuration example of the present invention will be described. However, each configuration and processing mode are merely preferred examples. Therefore, the present invention is not limited to the following embodiments, but is applicable to various forms based on the description of the present invention including the claims.
[0032]
(Configuration of First Embodiment and Individual Operation of Main Part)
FIG. 1 is a schematic external configuration diagram of a document identification and classification method, a document identification and classification processing system, and a program according to a first embodiment of the present invention.
[0033]
Referring to FIG. 1, this example shows an OCR device 1, a control device (for example, a small general-purpose computer and an interface connection device) 2 for executing control of the OCR device 1, and a document sorting / storing device (sorter) 3. And a transport device 4 for transporting the document from the OCR device 1 to the document sorting and storage device 3.
[0034]
When the number of classified documents is large, the number of installed document sorting / storing apparatuses (sorters) 3 is increased, and the following operations are performed in parallel.
[0035]
Further, the OCR device 1 to the transport device 4 are not limited to this configuration. For example, a local area network (LAN) accommodating a correction terminal or the like for confirming and correcting converted character data in the optical character reading in the OCR device 1 can be used.
[0036]
Next, individual operations of the above-described main parts will be described.
[0037]
Referring to FIG. 1, an OCR apparatus 1 reads written information of a document by using an optical system and a photoelectric conversion element by sub-scanning and main scanning, and outputs character data obtained by converting the image data by application software. I do.
[0038]
The control device 2 executes setting control regarding reading of the OCR device 1. For example, the reading accuracy of the document, the reading range (for example, the paper size of the document) and the like are set.
[0039]
The document sorting / storing device (sorter) 3 sorts documents read by the OCR device 1 according to a predetermined classification (for example, a medical statement and a bill, a medical department, a dentistry, a dispensing bureau, and a medical facility). Are discharged to a sorter shelf (which is appropriately described as a pocket) for transport and discharge.
[0040]
The transport device 4 transports the document read by the OCR device 1 to the document separation and storage device 3.
[0041]
FIG. 2 is a block diagram showing a configuration example of a control system for executing a program in the configuration of FIG.
[0042]
Referring to FIG. 2, in this configuration example, an I / F circuit 31, an MPU 32 (CPU, working RAM and ROM), an input / output (I / O) circuit 33, a display device 34, an input operation It has a device 35, a memory 36, an auxiliary storage device 37, and a driver 38, each of which is connected to a bus line for control, address and data transfer.
[0043]
The configuration example of this control system operates as follows.
[0044]
The I / F circuit 31 performs a data transfer process, and processes the transfer data so that the MPU 32 can capture the data. The MPU 32 executes a known process including a CPU, a working RAM and a ROM. The input / output (I / O) circuit 33 outputs various kinds of processing data from the input / output (I / O) circuit 33 to a display device 34 which is a CRT monitor or an LCD monitor. Data is exchanged with the operation device 35.
[0045]
In the configuration shown in FIG. 2 and the configuration in which the control device 2 in FIG. 1 executes control of the OCR device 1, the following signals (a) and (b) are input and output to the I / O circuit 33. You.
[0046]
(A) The document transport / storage unit 40a (indicated by a dotted line) is executed under the control of the MPU 32 (see the processing flow of FIG. 5), and the document read by the OCR device 1 is classified in a predetermined classification (for example, medical Each sorter shelf (see the pockets 41a to 41d of the classification pocket 41 shown in FIG. 3 in the document sorting and storing apparatus 3) is divided according to the contents specification and bill, medical department, dentistry, dispensing bureau, medical facilities, etc.). Output of control signal for discharging to.
(B) Capture of a detection signal (a signal of a photoelectric conversion level corresponding to the amount of transmitted light) from an optical detector 40b (indicated by a dotted line) for determining the thickness of a document to be read by the OCR device 1.
[0047]
The configuration of the optical detector 40b is not applied in the first and second embodiments. Hereinafter, detailed description is omitted.
[0048]
Further, in the configuration of FIG. 2, the processing data of the MPU 32 and the like are stored in the memory 36 and the auxiliary storage device 37, and the driver 38 takes in the storage information from an information recording medium such as a CD-ROM.
[0049]
FIG. 3 is a schematic diagram for explaining the reading, conveying, and discharging states of a document in the OCR apparatus in FIG.
[0050]
Referring to FIG. 3, a classification pocket 41 for discharging (storing) a document P placed on an ADF device (not shown) is provided in the document transport storage unit 40a.
[0051]
In the classification pocket 41, the following pockets (a) and (b) are arranged.
[0052]
(A) Transport / discharge of a normally read document in accordance with a predetermined classification (for example, classification of the medical description and bill of the above-mentioned claim, medical department, dentistry, dispensing bureau, medical facilities, etc.) Pockets 41a to 41d...
[0053]
In the pockets 41a to 41d, documents whose document types cannot be identified are displayed on the screen at this time, and discharged / transported (denoted as compulsory classified transport / discharge as appropriate) according to the classification judgment of the operator. You.
[0054]
(B) In addition to this, a discharge pocket for documents that cannot be processed by “normal classification transport / discharge” and “compulsory classification transport / discharge” (for example, damaged, unplanned mixture of documents, etc.), and if necessary, Although a pocket for a double feed error document in which documents are overlapped is also provided, it is not shown in this example.
[0055]
A known “claw, roller” or the like for sending out a document is disposed between the sorting pocket 41 and a storage portion (not shown) of the document P. Rollers are also arranged in the transport path between them.
[0056]
Further, an optical detector 40b and an optical reading unit 43 for detecting the thickness of the conveyed document by the amount of transmitted light are provided between the classification pocket 41 and a storage unit (not shown) for the document P.
[0057]
(Overall operation of the first embodiment)
FIG. 4 is a functional block diagram when the whole operation of the first embodiment is executed by the MPU configuration shown in FIG. 2, and FIG. 5 is a flowchart showing a processing procedure of the whole operation of the first embodiment. FIG. 6 is a diagram showing an example of a screen display for explaining a pocket instruction (forced classification transport / discharge) of a transport / discharge destination of a document whose document type cannot be identified. FIG. It is a figure for explaining an example of a sorting process.
[0058]
Referring to FIGS. 1 to 7, first, a bundle of documents (recept) is placed on an ADF device (not shown) of the OCR device 1. The control device 2 instructs the OCR device 1 to read the written information of the document, and the OCR device 1 starts reading the documents one by one (step S1 in FIG. 5).
[0059]
Here, the reading of the written information in the document is to read not the entire surface but, for example, a serial number such as the top of the document. The serial number is a serial number indicating a medical department, a dentistry, a dispensing bureau, a medical care facility, and a bill, a bill, etc. The serial number may be a one-dimensional or two-dimensional bar code or the like irrespective of numbers or English characters.
[0060]
The OCR device 1 transfers the written information of the document (image data or character data optically recognized by the OCR device 1) to the control device 2 (steps S2 and S3). When the OCR device 1 transfers the image data read here to the control device 2, the control device 2 performs character data conversion by character recognition. This character recognition is a known optical character reading conversion process by a dedicated software analysis engine (step S4).
[0061]
When the control device 2 performs normal identification of the document type (for example, individual identification of the medical description and bill of the above-mentioned claim, medical department, dentistry, dispensing bureau, medical facility, etc.) (step S4: Yes), the MPU 32 in FIG. 2 determines the transport destination / discharge destination pocket (one of 41a to 41d) in the document sorting and storing apparatus 3 based on the classification of the document this time.
[0062]
The MPU 32 instructs the document transport storage unit 40a shown in FIGS. 2 and 3 (step S5), and in the document transport storage unit 40a of the document sorting and storage device 3, the classified pocket 41 is placed in one of the designated pockets 41a to 41d. It is conveyed and discharged (step S6).
[0063]
If the type of document cannot be identified in step S4 (No), the image data read here (or the character data with optical character recognition) is transferred from the OCR device 1 to the control device 2 (step S7). .
[0064]
In this case, a notification sound for notifying the operator of the transfer of a document for which character recognition could not be performed is sounded, for example, in the MPU configuration (sounding circuit and speaker are not shown) shown in FIG. It is preferred in terms of.
[0065]
The control device 2 displays a screen 43 of the image data (or the character data obtained by optical character recognition) shown in FIG. 6 from the OCR device 1 (step S7). In this case, on the screen 43, the above-described serial number 46a is displayed on the top of the image in the image data (or the character data of which the optical character has been recognized) 45 of the document.
[0066]
Then, the serial number 46a is analyzed by a database engine (not shown) or the like, and the classification contents 46b (medical departments, dentistry, dispensing bureaus, medical facilities, bills, bills, etc.) are displayed on the screen.
[0067]
When the operator can visually read the classification contents 46b from this screen display (step S8: Yes), the operator refers to the classification correspondence contents table 48 displayed on the screen 43.
[0068]
The classification correspondence table 48 exemplifies “classification contents: medical / invoice, pocket 41a / dentist / invoice, pocket 41b...”.
[0069]
Referring to the classification correspondence table 48, the operator determines the classification destination (conveyance / discharge destination pocket) of the document whose document type cannot be identified this time, and sends this instruction through the display unit 49 at the lower right of the screen. Do. For example, the instruction is given by clicking on the display screen.
[0070]
With this instruction, the MPU 32 instructs the document transport / storage section 40a shown in FIGS. 2 and 3 (step S9), and transports the document to the designated pocket 41a of the classification pocket 41 in the document transport / storage section 40a of the document sorting / receiving apparatus 3. It is discharged (step S6).
[0071]
It is possible that the classification contents cannot be grasped in step S8. It is considered that this is because the OCR device 1 was unable to perform normal classification, and thus the operator often could not grasp the classification contents from the screen 43 obtained by the above transfer.
[0072]
Therefore, if the identification of the document type by the operator cannot be classified (No) in step S8, all the data 47 of the image data (or the character data recognized by the optical character) 45 of the document from the OCR device 1 (in step S7, the document The transfer of only the serial number 46a at the top or the like is performed under the instruction of the operator from the control device 2, and the entire image data 47 (or the character data of which the optical character is recognized) of the document taken in from the OCR device 1 is displayed on the screen. It is displayed (steps S10, S11).
[0073]
In particular, the operator visually recognizes the contents of the classification of the document from the image data 47 other than the serial number 46a such as the uppermost part. After this determination, the operator gives a screen instruction (see "OK" display in FIG. 8) in the same manner as described above, and the MPU 32 instructs the document transport storage section 40a shown in FIGS. 2 and 3 (step S9).
[0074]
With this instruction control, the current read document is transported and discharged to the designated pocket 41a in the classification pocket 41 in the document transport storage unit 40a of the document sorting and storing apparatus 3 (step S6).
[0075]
When such document classification is performed, documents 4, 10, and 13 in which the document type cannot be identified in the read classifications A and B of documents 1 to 16 shown in FIG. , B, they are conveyed / discharged (sorted) to the pockets of the document sorting / storing device 3 without different reading orders as shown in (b) and (c) of FIG.
[0076]
(Advantages of the first embodiment)
As described above, in the first embodiment, the reading order when reading the written contents of the document with the optical character reader and identifying and classifying the document type can be prevented from being different. As a result, accurate classification becomes possible.
[0077]
(Configuration of Second Embodiment and Individual Operation of Each Unit)
FIG. 8 is an external configuration diagram showing a schematic configuration of the second embodiment.
[0078]
Referring to FIG. 8, the configuration example of the second embodiment is configured such that a server 5 is additionally provided in the configuration of FIG. 1 (OCR device 1, control device 2, document sorting and storage device 3, transport device 4). Are connected to a transmission path in a local area network (LAN) configuration.
[0079]
Next, individual operations of the above-described units will be described.
[0080]
The configuration in FIG. 1 (from the OCR device 1 to the transport device 4) is basically the same as that of the first embodiment (only the operation corresponding to the control operation of the server 5 is different).
[0081]
The server 5 executes the following controls (a), (b), and (c) according to the second embodiment.
[0082]
(A) When classification by character recognition at the time of reading cannot be performed (when the type of document cannot be identified), character recognition software different from the character recognition software that has performed the current character recognition (for example, the manufacturing company is different) is used in the OCR apparatus. Transfer to 1. In other words, a plurality of types of character identification software are installed, and in cooperation with the control system of the OCR device 1, character identification software that can be converted into character data is applied to enable identification of the previous document type. I do. This description is appropriately described as (a) “executed by other character recognition software”.
[0083]
(B) If the image of the document read by the OCR device 1 is partly, for example, in the first embodiment, when the document cannot be identified by reading the document such as “invoice” or “medical institution name”, the entire document is read. And a program for transferring image data obtained by reading the written information of the document. This description will be appropriately described as (b) “Importing all character data”.
[0084]
(C) When the image of the document read by the OCR device 1 is partially, for example, in the first embodiment, when the document is "invoice" or "medical institution name", the document type cannot be identified by reading the document. A program for converting the image data obtained by reading the written information on the entire surface of the document into character data and transferring the character data is installed, and the server 5 transfers the program.
[0085]
Further, the server 5 captures the character data in the OCR device 1 that has executed this program, and each word in the character data is registered in the server 5 in advance (this is the medical content of the above-described receipt). (Corresponding to the specification and the bill, the medical department, the dentistry, the dispensing bureau, the medical facilities, etc.), and the classification (the classification of the pockets 41a to 41d) for identifying the document type is performed. This description is appropriately described as (c) “word matching”.
[0086]
(Overall operation of the second embodiment)
FIG. 9 is a functional block diagram when the overall operation of the second embodiment is executed by the MPU configuration, and FIG. 10 is a flowchart showing a processing procedure in the operation of the second embodiment.
[0087]
8 to 10, first, a bundle of documents (recept) is placed on an ADF device (not shown) of the OCR device 1. The control device 2 instructs the OCR device 1 to read the written information of the document, and the OCR device 1 starts reading the documents one by one (step S21 in FIG. 10).
[0088]
Here, the reading of the written information on the document is not for reading the entire surface, but for reading the above-mentioned serial number, for example, at the top of the document.
[0089]
The OCR device 1 transfers the read document description information (character data optically recognized by the OCR device 1) to the control device 2 (steps S22 and S23). When the OCR device 1 transfers the image data read here to the control device 2, the control device 2 generates character data by character recognition. Character recognition is a known optical character reading conversion process by a software analysis engine (step S24).
[0090]
If the document type is normally identified (step S24: Yes), the control device 2 classifies the document based on the current document identification, that is, the transport / discharge destination pocket (41a) in the document sorting and storing device 3. To 41d) is determined by the MPU 32 in FIG. The MPU 32 instructs the document transport storage section 40a shown in FIGS. 2 and 3 (step S25), and any one of the designated pockets 41a to 41d in the classification pocket 41 in the document transport storage section 40a of the document sorting and storage apparatus 3. (Step S26).
[0091]
Steps S21 to S26 so far are the same as the processing of the first embodiment (FIG. 5).
[0092]
If the document type cannot be identified in step S24 (No), the image data (or character data) read here is transferred from the OCR device 1 to the control device 2 (step S27).
[0093]
The selective sorting in step S27 is performed, for example, by (a) “executing with other character recognition software” and (b) “capturing all character data” set in advance through the display device 34 and the input operation device 35 in FIG. , (C) executes any one of “word matching” (corresponding to “selective setting” in the present invention).
[0094]
(A) If “execute by other character recognition software” cannot be classified by the identification of the document type at the time of reading (step S24: No), the character identification software that has identified the document type by the current character recognition Transfer different character identification software (for example, different production company) to the OCR device 1 (step S28). In other words, a plurality of types of character identification software are installed, and in cooperation with the control system of the OCR device 1, character identification software that can be converted into character data is applied to determine the previous document type inability to be identified. The identification is made possible (step S29).
[0095]
(B) “Import all character data” means that a part of the image of the document read by the OCR device 1 is read, for example, in the first embodiment when the document is “billed” or “medical institution name”. If the document type is not identifiable, a program for transferring image data obtained by reading the written description information of the entire document is installed, and this program is transferred to the server 5 (step S30). With this program, the previous document type can be identified (step S31).
[0096]
(C) "Word matching" means that a part of the image of the document read by the OCR device 1, for example, in the first embodiment, when the document "billing" or "medical institution name" is read, the document is read. If the type cannot be identified, a program for converting the image data obtained by reading the written information on the entire surface of the document into character data and transferring the character data is installed, and the server 5 transfers the program. Further, the server 5 fetches the character data in the OCR device 1 that has executed the program (step S32).
[0097]
"Words" in the character data are words registered in advance in the server 5 (this corresponds to the medical description and bill of the above-mentioned claim, medical department, dentistry, dispensing bureau, medical facilities, etc.) and The document type is identified by matching, and the classification (classification of the pockets 41a to 41d) is performed (step S33). This matching makes it possible to identify the previous document type that cannot be identified (step S34).
[0098]
If the document type can be identified by any of these processes, as described above, the MPU 32 instructs the document transport / storage unit 40a shown in FIGS. 2 and 3 (step S25), and separates and stores the document. The document is transported / discharged to one of the designated pockets 41a to 41d of the classification pocket 41 in the document transport storage section 40a of the apparatus 3 (step S26).
[0099]
FIG. 11 is a flowchart illustrating another processing procedure in the operation of the second embodiment.
[0100]
Referring to FIG. 11, in the processing procedure of FIG. 10 described above, one of (a) “execute by other character recognition software” to (c) “word matching” is executed by “selective setting”. I have.
[0101]
On the other hand, in the second embodiment, (a) “execute with other character recognition software” to (c) “word matching” are “executed until the document type can be identified” (the present invention). Corresponding to the process continuously).
[0102]
Therefore, if the document type cannot be identified (No) in step S24 as in FIG. 10, first, (a) "execute with other character recognition software" is processed (steps S41 and S42).
[0103]
If the document type is not identifiable, (b) "all character data import" is automatically processed (steps S43 and S44). If the document type cannot be identified, (c) "word matching" is automatically processed (steps S45, S46, S47). As described above, the processing is continuously performed, and the document type can be identified in any one of the processing.
[0104]
If the process cannot be performed in the last step S47, the process returns to the start of the process or one process ends.
[0105]
If the document type is recognized in any of the processes or the processes up to the end, the MPU 32 instructs the document transport storage unit 40a shown in FIGS. Then (step S25), the document is transported / discharged to the designated pockets 41a to 41d in the classification pocket 41 in the document transport / storage section 40a of the document sorting / receiving apparatus 3 (step S26).
[0106]
(Advantages of the second embodiment)
As described above, in the second embodiment, as in the first embodiment, the reading order of the documents placed for reading matches the transport / discharge order of the classification. At this time, the process automatically reads the items of the document for which the document type cannot be identified again (from “(a) Execute with other character recognition software” to “(c) Word matching”). ing.
[0107]
In other words, as in the first embodiment, when the document type cannot be identified, an artificial instruction operation based on the screen display and the classification judgment of the operator is not required. As a result, accurate and more efficient classification becomes possible.
[0108]
(Modification)
Hereinafter, modified examples of the configurations in FIGS. 1 and 8 will be described.
[0109]
In the following Modifications 1 to 3, it is possible to perform the same processing as in the first and second embodiments for matching the reading order of the documents placed for reading with the order of transport and discharge of the classification.
[0110]
FIG. 12 is a communication network diagram illustrating a configuration example of the first modification.
[0111]
Referring to FIG. 12, in the first modification, the correction unit 1Aa and the reading unit 1Ba are separately arranged. For example, the correction unit 1Aa and the reading unit 1Ba are respectively installed on different floors in the same building.
[0112]
For this reason, a switch 52 for, for example, 100BASE LAN transmission (IEEE802) is provided in the correction unit 1Aa, and a switch 50 is also provided in the reading unit 1Ba. The switches 50 and 52 are connected by a 100BASE LAN transmission line 51.
[0113]
The correction unit 1Aa of this example is provided with a printer 53 for business management, a management W / S 54, and a magnetic tape (MT) device 55 in addition to the correction W / Ss 3a, 3b,. In addition, these can implement this invention, without providing especially.
[0114]
In the first modification, for example, the same processing as that of the first embodiment is executed by a procedure (for example, RPC: Remote Procedure Call) by a server on another LAN connected to a line concentrator (hub).
[0115]
The first modification has an advantage that the correction unit 1Aa and the reading unit 1Ba can be arranged separately (for example, on different floors of the same building). Other operational advantages are the same as in the first and second embodiments.
[0116]
FIG. 13 is a communication network diagram showing the configuration of the second modification.
[0117]
Referring to FIG. 13, in the second modification, the correction unit 1Ab and the reading unit 1Bb are separated at a remote place. The same processing as in the first embodiment is executed through the router 60, the communication network 61, and the router 62.
[0118]
The second modification has an advantage that the correction unit 1Ab and the reading unit 1Bb can be separately arranged (for example, in a local area and an urban area) at remote locations. The other operational advantages are the same as those of the first modification.
[0119]
FIG. 14 is a communication network diagram illustrating a configuration example of the third modification.
[0120]
Referring to FIG. 14, in the third modification, an optical jukebox 72 is connected to the configuration shown in FIG. 1 through a host device 70 or a server 71, and a large amount of read document data is managed. Further, a router 73 is provided. For example, it is connected to a municipal receipt processing system 76, 77,. Each of the municipal receipt processing systems 76, 77,... Includes a router 80, a server 81, processing terminals 82, 84, a small OCR device 83, and the like.
[0121]
【The invention's effect】
As apparent from the above description, according to the document identification / classification method, the document identification / classification processing system and the program of the present invention, the contents described in the document are read by the optical character reader, and the document type is identified and classified. In this case, the reading order can be prevented from being different, and as a result, there is an effect that the classification by the accurate and reliable identification of the document type becomes possible.
[Brief description of the drawings]
FIG. 1 is an external configuration diagram schematically illustrating a document identification and classification method, a document identification and classification processing system, and a program according to a first embodiment of the present invention.
FIG. 2 is a block diagram illustrating a configuration example of a control system for executing a program in the configuration of FIG. 1;
FIG. 3 is a schematic diagram for explaining a reading state of a document in the OCR device in FIG. 1;
FIG. 4 is a functional block diagram when the overall operation of the first embodiment is executed by the MPU configuration.
FIG. 5 is a flowchart illustrating a processing procedure of an overall operation according to the first embodiment.
FIG. 6 is a diagram for explaining a pocket instruction of a destination for conveying and discharging a document in the first embodiment.
FIG. 7 is a diagram illustrating an example of a document sorting process in the first embodiment.
FIG. 8 is an external configuration diagram illustrating a schematic configuration of a second embodiment.
FIG. 9 is a functional block diagram when the overall operation of the second embodiment is executed by the MPU configuration.
FIG. 10 is a flowchart showing a processing procedure in the operation of the second embodiment.
FIG. 11 is a flowchart illustrating another processing procedure in the operation of the second embodiment.
FIG. 12 is a communication network diagram illustrating a configuration example of a modified example in the first and second embodiments.
FIG. 13 is a communication network diagram illustrating a configuration example of another modification example of the embodiment.
FIG. 14 is a communication network diagram showing a configuration example of still another modified example in the embodiment.
FIG. 15 is a view for explaining a pocket instruction of a document transport / discharge destination in a conventional example.
[Explanation of symbols]
1. OCR device
2 ... Control device
3. Document sorting storage device (sorter)
4: Transport device
32 ... MPU
33 ... input / output (I / O) circuit
34 Display device
35 ... Input operation device
40: Document transport storage section
40a: Document transport storage section
41… Classification pocket
41a-41d ... pocket
43 ... Optical reading unit

Claims (6)

In the document identification and classification method of using an optical character reader, sequentially reading the written information of the document, identifying and classifying the document type, and transporting / discharging the document to a predetermined storage unit,
Automatically sending out the documents sequentially;
A step of optically reading the written description information of the document to be sent out, identifying the document type, and determining a storage unit of the transport / discharge destination of the document corresponding to a predetermined classification,
When determining the storage section of the transport / discharge destination of the document, the document description information of the document whose document type cannot be identified is displayed on a screen, and the document classified from the screen information based on the identification of the document type is transported. A process of instructing a storage unit to be discharged;
Transporting and discharging the document read this time to the determined storage unit or the specified storage unit;
And a document identification and classification method. In the document identification and classification method of using an optical character reader, sequentially reading the written information of the document, identifying and classifying the document type, and transporting / discharging the document to a predetermined storage unit,
Automatically sending out the documents sequentially;
A step of determining the storage unit of the transport / discharge destination of the document, corresponding to a predetermined classification, by sequentially optically reading the written description information of the sent document and identifying the document type,
If the document type is indistinguishable in the determination, a step of automatically re-reading the description items of the indistinguishable document of this document type,
A step of automatically instructing a storage unit of a transport / discharge destination of the document classified based on the identification of the document type by the automatic reading process performed again,
Transporting and discharging the document read this time to the determined storage unit or the specified storage unit;
And a document identification and classification method. In a document identification and classification processing system for sequentially reading the written information of documents, identifying and classifying the document type, and transporting and discharging the documents to a predetermined storage destination,
Automatic sending means for automatically sending the documents sequentially,
Reading identification and classification means for sequentially and optically reading the written information of the document sent out by the automatic sending means to identify the type of the document, and determining a storage portion of the transport / discharge destination of the document corresponding to a predetermined classification; and ,
Display means for displaying on the screen the document description information whose document type is unidentifiable by the read identification and classification means,
Instructing means for instructing a storage unit for transporting / discharging a document according to a classification in which a document type is identified based on the written information in the display means,
Conveyance / discharge means for conveying / discharging the document read this time to the storage unit determined by the read identification and classification means or the storage unit instructed by the instruction means,
A document identification and classification processing system comprising: In a document identification and classification processing system for sequentially reading the written information of documents, identifying and classifying the document type, and transporting and discharging the documents to a predetermined storage destination,
Automatic sending means for automatically sending the documents sequentially,
Reading identification and classification means for sequentially and optically reading the written information of the document sent out by the automatic sending means to identify the type of the document, and determining the storage section of the transport / discharge destination of the document corresponding to a predetermined classification When,
The reading identification / classification means automatically performs the process of automatically reading the items described in the document whose document type has become indistinguishable, and automatically instructs the storage unit of the transport / discharge destination based on the classification based on this process. Instruction means for performing
Conveyance / discharge means for conveying / discharging the document read this time to the storage unit determined by the read identification and classification means or the storage unit instructed by the instruction means,
A document identification and classification processing system comprising: Automatically sending out the documents sequentially;
A step of optically reading the written description information of the document to be sent out, identifying the document type, and determining a storage unit of a transport / discharge destination of the document corresponding to a predetermined classification;
When determining the storage unit of the transport / discharge destination of the document, the document description information of the document whose document type cannot be identified is displayed on a screen, and the document classified based on the document type based on the screen information is identified. Indicating a storage unit to be discharged;
Conveying and discharging the document read this time to the determined storage unit or the specified storage unit,
A program that causes a computer to execute control of the computer. Automatically sending out the documents sequentially;
A step of determining the storage unit of the transport / discharge destination of the document, corresponding to a predetermined classification, by sequentially optically reading the written description information of the document to be sent and identifying the document type,
When the document type is indistinguishable in the determination, a step of automatically re-reading the description items of the document in which the document type is indistinguishable,
Automatically instructing the storage unit of the transport / discharge destination of the document identified and classified based on the document type based on the automatic reading process performed again;
Conveying and discharging the document read this time to the determined storage unit or the specified storage unit,
A program that causes a computer to execute control of the computer.

Images