JPS60500188A - Document processing device and method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Document processing device and method Technical field This invention encodes the encoded (encoded) data field A transporter that moves documents sequentially along the feeding path in the feeding direction. and means for reading data from said document while said document is being fed. the reading means made and the predetermined criteria predetermined for said data met; and conversion means adapted to convert the failed document to the rejection area. It relates to a document processing device.

This invention also provides a feeding path: a feeding direction (τ) in which the documents are sequentially moved; reads data from said document as said document is being fed; documents that fail to meet predetermined criteria predetermined for the purpose of It relates to a type of document processing method including each step of processing.

This invention is particularly applicable to the type of check used in banking, where the document is a check of the type used in banking; If the device is a Me machine (reader)/sorter (sorter) for the check applied to.

Apparatus of the type herein specified can be learned from U.S. Pat. No. 3,988,571. , it will turn incompletely read documents into a reject pocket. Discloses a device for classifying financial documents that have been transferred to other countries. Reading documents For example: 1. Bad encoding or printing, dirt, or folding. It may become incomplete due to the existence of such things. a bundle of documents (also called a batch) After sorting the documents in the rejection pocket, manually The rejected documents were taken to Odai (Odai), where Oreta observed the rejected documents one by one with her eyes. Try to insert correct data from the keyboard and complete incomplete data. Ru. The rejected data is then put through the device again and the correct data is routed through the device. The ink is drawn from the buffer and printed onto the document. child The known system has the disadvantage of slow operation.

Disclosure of invention The object of the invention is, inter alia, to provide a device of a kind which eliminates the above-mentioned disadvantages.

Accordingly, according to one aspect of the invention, the invention provides a Receive the rejected @ category and add the missing data on the document. The present invention provides an apparatus including an encoding means for encoding.

3 Furthermore, this invention, according to other aspects, is effective against missing documents ( Mitsong) Create data for the data field and turn into the rejected area positioning an encoding means adjacent the rejected fine sand for receiving a rejected document; Completed data corresponding to each rejected document obtained from the data creation process. Provided is a method of the kind comprising each step of encoding in said encoding means. It is something.

The device according to the invention operates on-line and concurrently to achieve rapid operation. It is clear that there is a benefit in that In a preferred embodiment, the saw The first /F pass (processing operation of putting documents through a sorter) or to capture an image of the document. The shadow @ is used for re-enrolling (re-inputting) data. It will be done.

Physical repair of the rejected missing data field is next to the rejected area. This is accomplished by an adjacent encoder. data in the field to be repaired It is best to use a low cost or inexpensive encoder to print I'll come. Additionally, the encoding means can be added to a conventional sorter as an 'at-on'' device. It has the advantage that it can be added.

Brief description of the drawing An embodiment of the invention will now be described by way of example with reference to the accompanying drawings, in which: FIG.

1A and 1B are connected together to illustrate one embodiment of the invention in block form. Represents the circuit diagram of the drawn system.

FIG. 2 illustrates the types of checks processed by the system of FIGS. 1A and 1B. FIG. 3 is a plan view of a document.

Figure 3 shows further details of the imaging (jmaglng) device shown in Figure 1A. FIG. 1A is a plan view similar to FIG.

Figure 4 is a diagram of Figure 1A showing the organization of data bundles in a data capture file. Figure 2 is a schematic diagram of a data capture file.

FIG. 5 is a schematic diagram similar to FIG. 4 representing re-entered data.

Figure 6 is a document re-entry representing the organization of the document re-entry Phi JV in Figure 1A. 1-1) is a schematic diagram of a file.

Figure 7 shows the missing part of Figure 1B representing the document storage or buffer system. - It is a top view of a part of field module.

FIG. 8 is a schematic diagram of the remaining portions of the missing field module of FIG. 1B. It is.

FIG. 9 shows a belt cut along line 9-9 in FIG. 7 and treated as shown in FIG. FIG.

’Best form for implementing JIMp Figures 1A and 1B are connected together and refer to this output system as O, hereinafter system]. Online, simultaneous operation, image-based sorter reject, 2 depicts a preferred embodiment of the Ent IJ and repair system.

The system 10 is represented by a dotted line in the figure and is a conventional reader, hereinafter referred to as sorter 2. /sorter and the missing field (hereinafter referred to as MFE13+4) (missing) encoder.

Sorter 12 (Figures 1A and 1B) has pockets 16° 18, 20, 22 and 23. Multiple sorting (classification) like 7I? Includes document truck 14 with quesoto nothing.

7 In stem 10, the rock of 1 "Z Sochi" to be processed (e.g. For example, from a bundle of 300 documents as 1 batch (or (-) 1 bundle) ) is placed in the stack 24 next to the trunk 14 and a conventional feeding mechanism 26 is used. is used, it bincuffs 28 documents from the stack 24, and then A transport mechanism 30 connected to the feeding mechanism 26 transports documents along the track 14. The data is transferred to the trunk 14, which is configured to move at a constant speed.

Documents such as 28 in this example include, for example, checks and vouchers. Therefore, they are often processed in "batches" as described above. Each document 28 is When moving along the trunk 14, each document is moved along the trunk 14. A conventional magnetic ink character identification (MICR) reader 32 and/or a conventional Operational relationship with optical character recognition (OCR) IJ-da (reader) 34. leader 3 2 is the data on the MI CR line 36 when the document 28 moves on that side (Fig. 2). ) is arranged to read. The MI CR line 36 also has a transition indicated by an arrow 38. OCR' located downstream from the MICRIJ-G32 along the direction of movement. It can also be read optically by the J-da 34. OCR'J-da; U4 is also It is generally printed with ordinary or non-magnetic inserts, such as ζ′°remittance slips”. It can also be used to process rejected documents for ICR code reasons. Wear. In this regard, the MICR line 36 of the document 28 is e.g. It is also used to identify bank and issuing numbers, customer account numbers and check amounts. We should remember that One batch of documents 28 passes through sorter j2 Once processed, the MTCR line 36 of each document is read and the read data sorter pockets 18-22, e.g. The pocket 16 is used as a rejection area for the first type of rejection. area or pocket, and pocket 23 is a rejection area for the second type of rejection. Or a pocket. These points will be discussed later.

The soak 12 is connected to the f-fold pockets 16-23 as shown in 28 above. Conventional type book 7 for converting documents Similar turning mechanism (schematically represented as turning machine or diverter 40, 40-1゜40-2) I am an eagle). Diverter 40.40-1.4.0-2 is shown in Figure 1A. Data Application Processor (Data Application Processor), designated globally as 4 A pocket machine that operates in response to sorting (classification) commands from It is controlled by a controller 42. Application Zorosesa 44 is written Using the data read from Category 28, the application program At the same time, a sorting command is supplied to the pocket controller 42.

Sorter 12 has a conventional recognition circuit 46 (FIG. 1A) associated therewith, in which Circuit "A" or 46-1 can be used for MICRIJ-da 32, and circuit " B''' or 46-2 is used for the OCRl)-der 34. Recognition circuit 46 -1.46-2 is data from readers 32 and 34 connected to them respectively Translation or interpretation of a set of information or data as described below, respectively. Data group) 48 is created. Data applications from other sources / Kon Zorose processor 44, and the composite data set can be sent to processor 44 for use by processor 44. It can also be prepared by the recognition circuit 46 for transmission to 44. However, for ease of illustration In the end, only one batch of data output from the M engineering CR reader 32 will be explained below. will do.

Continuing with the fully described example, this data set is stored in a conventional The application Zorosesa 411 and Anno 0 Rikano via the communication node 50 connected to a conventional interface 52 in collaboration with John 70 Losenosa, 44; . The recognition circuits-1, 6-], 16-2 also recognize the ゛-When you decide whether the evening is completed (complete) or not, it comes to an end. For example, if the 30 MICR characters for each document that the data group passes along track 14 one or more characters are missing. In this case, the data set is missing a particular character and, for example, Traditional code tags that display the location of that particular character in a data set. A tag will be added.

Anno 0 application using the data group from the recognition circuits 46-1 and 46-2. Norosessa 44. (FIG. 1A) shows interfaces 52, 54, 56 and 58; Read-only memory (ROM) 60 and random access memory (RA) M) 62, a display like a cathode ray tube (CRT) 64, and a keyboard. board (KB) 66, processor (MP) 68, and control and interface board. gic 70. Processing routines used in application processor 44 Although the routine resides in ROM 60, the routine more generally As part of the start-up procedure, e.g. The load may be set to 2.

Insert data or instructions using CRT 64 or KB66. It is used to communicate with the incoming network. Control and interface The logic 70 provides connections between all of the various components of the application processor 44. The application processor is used as an application processor. so that it can function normally.

Now, at this point, sorter 12 typically stores the various components of sorter 12 as such. It has a conventional host-sorter controller 72 (FIG. 1B) that controls the It should be pointed out that To simplify the drawing, host-saw From the controller 72, components cooperating therewith, such as, for example, the feed mechanism 26. The individual control lines up to are not shown. data application zorose The processor 44 performs functions related to the present invention as described below.

At some point, the flow of documents flowing from the punch (bundle of documents) passes through the leaders 32 and 34. Verify that all data read by the application is complete and correct. each document 28 is placed on track 14, assuming that it is transferred to processor 44; For example, data related to the document 28 as it moves past the reader 32. The data group is used by the application processor 44 to Decide whether the document will be forwarded or noted to the notes (18-22). Set. For example, a data set containing account number, issuing bank number, amount, etc. (application programming) decisions are made. This classification is determined by the interface. through the face 54 to the pocket controller 42 where the appropriate die The converter 40 is actuated to divert the document to the appropriate pockets 18-22. Various books The various gold sides and certificate/balance status of the same group Nodame also match (reconcilia tion) is calculated and stored as conventionally done in processing. .

However, in actual practice, Document 28 is all formulaic as explained in 1. It is very difficult to read the information and follow the instructions in Note 1. That is, below, the first type of rejection There is a rejection document that cannot be read correctly or is not complete. There is also a new document 28 that is completely missing the document amount called a type rejection. Money The second type of rejection U, where the forehead is completely missing: MFE monorail 13 is passed through. amount will be encoded. This point will be discussed later.

System 10 to facilitate the recovery and reen-IJ of rejected documents. is an imaging device that includes the imaging device 73. equipment. The imaging device 73 is of a conventional type, and the first It is schematically illustrated in Figure A and represented in detail in Figure 3.

11 Imaging device 73 is downstream from MI CR and OCR reader 32.34 , the document 28 passes the imaging device 73 along the track 14. It is positioned so that it will come into motion with the document when it is pressed.

The imaging device 73 is placed on the wall 14-1 of the track 14. a light guide 80 that includes a glass window 74 that shares light from light sources 76 and 78, respectively; 82 to the scan line 84. scan line 8 In FIG. 3, 4 represents the track 14 in its plan view, so it appears as a dot.

In Figure 3, the upper long end of the document 28 is visible, and the honoro is in normal operation with the lower long end. It is moved with the edge □ down, and when it is moved like that, the front of the document 28 is facing the ging device 73. In this situation, the scan line is directed towards. As the document 28 moves past the scan line 84, the The rightmost end (when viewed in FIG. 3) is illuminated by a light source 76 * 78 and is illuminated through a window 74. The light reflected therefrom is transmitted to the imaging sensor by a suitable lens system 86. The light is focused on an array 88. Array 88 in this embodiment is along scan line 84. It is of the type that generates a fixed number of picture elements or pixels. Such an array 8 8, for example, RL-1 manufactured by Reticon Covporation. 024B, with 1024 pixels along scan line 84. 640 to meet the resolution requirements of this example. Only pixels are used. As the document 28 moves in the downstream direction indicated by the arrow 38, A new area of the document is then provided on scan line 84 along which a new 6 - Generate a set of 40 pixels. Each pixel from sensor array 88 is the corresponding analog gray scale signal that is converted or digitized by path 90. For example, each pixel generates 6 bits of data. , records 64 gray range shades (intermediate colors) from white to black. Scanning Upon completion, a stream of data bytes is output from processing circuitry 90. On that point Since this is a prior art, there is no need to describe it in further detail.

Document 280 ``Image'' (image) Bytes from C scan line 84 (Figure 3) The flow of data or pixels is from the output of the processing circuit 90 to the encoder 9. 2 for further processing by supplying the output to a compressor (pressure compressor) 94. Enhancer 92 is a conventional one, unnecessary background information and can be used, for example, to create relevant data that stands out from its background information. This is the circuit used. Compressor 94 is a conventional circuit. Receives the highlighted data from the non-nsa 92 and removes meaningless or redundant data. Remove the data and compress the remaining data to create compressed and dinotted image data. Sea urchin 3 Shorten the unnecessary transmission time width for data transmission related to the shadow (8) of 28, and Requires less memory storage. The output of compressor 94 is connected to conventional communication node 96 (sugar IB diagram), and its output is fed to the image data selector switch 100. It is connected to an interface 98, which is a part of the interface. Selector switch 100 is rejected is part of the means 102 (Figure 1B) for completing the data regarding the document 28 that has been Ru. This point will be discussed in detail later.

For the time being, the Imaging de Cuisine 73 captures the image of each document 28 passing through it. Suffice it to say that the MTCRIJ If the data group of document 28 read from 32 is complete, the information regarding the document 28 is complete. The image data that is included in the image data is purged or removed. Also, for example, MIC If the data group of the document 28 read from the RIJ-da 32 is incomplete, the document The image data related to the data is obtained by means 102 (FIG. 1B) for completing the data as described below. Retained to facilitate data completion by

Continuing with the example of completing the data read from each document 28, recognition circuit 46 Each of the related data groups output from the communication node 50 and the interface 5 2 to a data application processor 44 (FIG. 1B). sent. The data application processor 44 controls each data group. and the memory device through the interface Rononok 70 and the interface 58. data capture file 104. The data group is shown schematically in Figure 4. are sequentially arranged in the file 104 as follows. For example, solid line 106 is the first document The dotted line 108 represents the complete data group for 28, and the dotted line 108 is determined by the recognition circuit 16. the incomplete data set of the next document 28 moving along the defined track 14; represents.

Dotted line 108 (Figure 4) represents the first rejected rock, class 28, and dotted line 110 represents this represents the second rejected document removed from stack 24 in the example of FIG. in front As mentioned, data groups may be missing or have irremovable "characters" or by the recognition circuit 46 in such a way as to indicate the number character '' as a rejection of the first type. ``tagged'' or ``fog'' and the data set is missing the entire amount. ``tag'' or ``flag'' in such a way as to indicate a second type of rejection. be done.

The first type of rejection document was converted into rejection pocket 16 by Gui Gu-yu 40-1. Then, the second type of rejected mist is diverted to the pocket 23 by the diverter 40-2. Ru.

Data group for the second rejection (rejection≠2; REJ 2) document (dotted line 110 ), there are several lines enclosed in parentheses 112 (Figure 4) that are complete Hiko data. The dotted line 114 represents the third rejection document 28 or rejection≠3 (REJ3). represents an incomplete set of data. line 10615 The document 28 corresponding to the complete data group represented by the parentheses 112 is an application. Specific application programs used with the computer processor 44 The light-shielding pockets 18 to 22 (FIG. 1B) are classified by the ram. Also, Figure 4 of the first and second types as represented by dotted lines 108, 110 and 114 of The documents 28 corresponding to the rejected data group are routed to the rejection pockets 16 and 23, respectively. sent. The application program that controls the sorting of documents 28 is It can also be installed in the ROM 60 of the data anorization processor 44. but typically, e.g. disk units or tape cartridges (not shown) may be loaded onto the RAM 62 (Figure 1B). stomach. Data Application Processor 440KB66 has one or more It can be used to initiate and control the load of application programs.

Sorting instructions for application programs used therein The instruction is sent from the application processor 44 as described above. is routed through interface 54 to pocket controller 42 of sorter 12. Ru.

Because the document 28 moves along the document track 14 at a certain speed, Type 28 is the data read after passing through the MICR or OCR reader 32.34. The status as to whether the data is complete, a first type rejection, or a second type rejection is determined. that it will move past the imaging device 73 before being determined. It should be pointed out. Therefore, the shadow of every document passing along track 14 An image is taken, which is transmitted to the image data center via communication node 96, as described above. output to the director switch 100.

data selector switch], OO (Figure 1B), for example Standard CMOS integrated cross-point settings such as director switch. Also, the switch 100 is first-come-first-served. Contains a memo IJ- or image buffer 116 that operates on an output or FIFO basis, Sections ≠1 and ≠ used for storing data representing 28 compressed images Each section has 2 to N sections.

Switch 100 generally has a data re-enabler designated as 118 (FIG. 1A). Controlled by Tori Zorosessa. This processor 118 is a data application operatively with the application processor 44 and the data selector switch 100. data by means 102 (FIG. 1B) connected and completing the data as described below. Make it easier to complete the data.

The reentry processor 118 (FIG. 1A) A two-way interface through which the data application processor 44 communicates with the data application processor 44. 120 and reentry processor 118 there]7 Transferring data to the switch 100 via the connected interface 124 interface 122, reentry processor 118 and data. A plurality of data entry terminals such as 128, 130 are part of the means 102 for and an interface 126 that is a two-way interface with the machine. Lien The tri-processor 118 further includes a control and interface processor 132, RO M 134, RAM 136 and MP 138, as shown. Constructed as before.

Application processor 44 sends commands to pocket controller 42. , controlling the distribution of documents 28 to the various sorter pockets 16-23. We should remember that. The document 28 with these complete data groups is, for example, a Book 28, which was classified as kets 18 to 22 and was considered a rejection of the first type and the second type, is They are forwarded to rejection pockets 16 and 23, respectively. Sorting data in your pocket - At the same time as being sent to the controller 42, each so-called data group is complete or The status indicating whether it is a rejection of the first type or the second type is data. - re-enable from application processor 44 via interface 120; Sent to Tori Zorosessa 118. For example, data file 104 in FIG. Taking the example from , it shows that the data set (line 106) is complete. stator or reentry 700 M:I8J [shipped, 1st and 1st 2 partially complete data group (lines 108, 110) is rejected ≠ 1, c REJ ]) and Re-Entry Proceno’Th11 as Rejection +2 (REJ 2) 8 and stored in, for example, R."J, 4] 3.C. Image The digitized image (shadow@) data generated from the digitizing device 73 is as described above. It is sent to the shadow 0 buffer 116 (Figure 1B) as shown above, but the processing is done by the data access. This is accomplished by application Zoroceno 911.4. For example, stack 2 The compressed image data for the first document 28 from 4 is stored in the image buffer 116. The compressed image data for the second class 28 is located at position ≠ 1, and the compressed image data for the second class 28 is You can place it at position ≠ 2 of F116, and place it in the same way below. . In this example, the image data of Imaging Temple 2 in image buffer ] ], 6 is in the first rejection. Corresponding to the missing documents, the data group is stored in the data capture file 104 (Figure 4). is incomplete, and is represented by line 1.08 in the same figure, ! -1,ru. Document 28 passes through the MICR and OCR reader 32134 and the imaging device 73. Immediately after that, the data annolocation processor 44, for example, complete or incomplete status (and first type or second type rejection status) would have received the TASS). That status data is used by data applications. The above-mentioned interface 56, 120 is connected to the The data is sent to the reentry processor 118 via the reentry processor 118.

Using this status data, the re-entry 70 processor 118 discard the compressed image data in the image buffer 116 for the corresponding document 28; Make a decision on what to keep. Typical classification - according to rejection rate, approximately 2 documents 28 An "image" of the image is maintained in the image buffer 116 to correspond to the rejected document. Will.

Images of these rejected documents 28 are shown at the data entry end as 128, 130. The image buffer 116 (first B Figure). Since the terminal 130 is the same as the terminal 128, the terminal Only 128 will be explained.

Terminal 128 (Figure xm) is a high resolution cathode ray tube (CRT) 140 A conventional semi-input device including a display and an entry keyboard (KB) 142. It is an intelligent device. The Kamesue machine 128 further includes a selector switch 100. An interface 144 for receiving compressed image data and viewing it on the CRT 140. Deconstruct the image data so that it can be compressed. Receives data from plenosa 146 and reentry processor 118 and a two-way interface 148 for transmitting data thereto, and a ROM 150. RAM 152, MP 154, and various types included in the terminal 128 in the conventional method It includes interfaces and control logic 156 for interconnecting the components.

Further, continuing the example given, if terminal 128 is available, the color buffer Compressed image data at position 4#2 of file 116 (for first rejected document 28) ) is sent to the terminal 128 via the interface 144, and then The compression is restored (recompressed) by a decompressor 146 for viewing on a CRT 140. Ru.

At this time, reentry processor 118 processes the corresponding incomplete data line 1 o8 (FIG. 4) is sent to the terminal 128 via the interface 148. this The incomplete data line 108 is further displayed on a CRT 140Kf, for example For example, what is obtained through the MICR reader 32 can be reported to the operator of the terminal 128. indicate. The re-entrant 70 processor 118 is, for example, a ROM 134 or R AM 136 either has a conventional enough routine and lines like line 108 The incomplete data line is transferred to the corresponding compressed image data (image buffer 116 position ≠ 2 (such as rejection ≠ 1), the operator can see them It is sent to the CRT140 as shown in the image. Additionally, reentry processor 118 has one 08 CR missing character or multiple missing characters in incomplete data It can be emphasized, for example, by pointing with an arrow on T140.

The operator looks at the image of document 28 on CRT 1401 insert the missing data from KB142. Once the missing data is fully inserted, the 4 Rake activated the transfer key 158 of KB142 to complete the data line. 108-1 to interface 148. Reentry processor 118. applique data capture file via the 7’o setter 44 and U interface 58, etc. file 104 and document reentry file 158 (FIG. 1A). No. Figure 5 shows the data capture file 104 after re-entry, with the pairs existing in Figure 4. The now complete data line is located at the location of the corresponding partially complete data line 108. 108-1. Multiple data lines within parentheses 112 that were complete from the beginning In and data lines 106 are returned to their original locations as complete data lines. It will remain as it is.

The document reentrant IJ file 158 having the configuration shown in FIG. first rejected as having an incomplete data line due to the second type of rejection. The data line 28 has a completed data line corresponding to the document 28 that was created. Examples in the entire explanation , only line 108-1, now completely f'L, is added to file 158 at this time. It will be done.

After successful data entry for data line 108-1 above, The data entry terminal 128 handling the processing of the document relinquishes data regarding the document. do. Further regarding discarding, the compressed data from position ≠ 2 of the image buffer 116 The image data entered at the entry terminal completes the aforementioned data line 108. As soon as it is sent to machine 128, position ≠2 of image buffer 116 discards the data. In other words, the device 2 is compressed and decoded for other rejected documents 28. override, i.e., rewrite, the data that has been knotted. You will be able to do this. Image buffer 116 has complete external data ≠ 1 Since it is for document 28, it would also be abandoned.

During the time one operator completes the commanded data line 108 , the next operator at entry terminal 1i 130 (Figure 1B) The compressed and denoted data is received from position +3 of 116. Current theory Position +-2 in the illustrated example is now the second rejection -42 as seen in FIG. Recall that the data line 110 includes a data line 110 for

The operator at the terminal 130 then explains about the terminal 128. Line 1] Complete the 0 data as shown above. incomplete data line 1 10 (FIG. 4) has been completed similar to that described for line 108, The completed data line 110-1 is shown in FIGS. 5 and 6, respectively. Transferred to data capture file 104 and @modify entry file 158. is recorded and memorized. Therefore, 23 Terminal 130 receives data regarding line 1.10 and image data associated therewith. After giving up, preparations are made to receive the data of the next rejected document 28. that's all The process described is such that all documents 28 in stack 24 are soaked 12 and unprocessed. is repeated with 1.

In summary, the complete, Salk 12 MICR and OCRIJ-da 32,3 The document 28 read normally by the data application processor According to the application program on 44 IEL pocket, 18~ It is classified into 22. The read data is represented by the rejection of the first type and the second type. Those documents 28, in the case of incompleteness as described above, will be included in the detailed explanation stage of this invention. are diverted to rejection pockets 16 and 23, respectively, as described above. rejected book Data that is re-entered or completed in the same category may e.g. l) The completed data obtained from the terminals 128 and 130 is data capture file 104. That completed data is actually rejected Create a document reentry file 158 (Figure 6), which is a list of documents 28 M The FE module 13 is utilized and the data is completed at the terminal 128.130. Encode or print the correct amount for those rejections as soon as possible. incomplete The first type of rejection document of 16 represents all the documents of the batch. 6 until the sample is completely processed through the soak 12.

The MFE module 13 is shown in more detail in FIGS. 7 and 8, and below the buffer 162, the documents 28 flowing from the sorter 12 to the encoder 160 are including buffering or buffering means. The buffer means (buffer) used is a conventional one. can be used.

Buffer 162 (FIG. 7) is arranged in a horizontal plane in this embodiment. It has a supporting plate]64.

A convergent track consisting of walls 166 and 168 and rising from support plate 164 is used. The document 28 from the rejection pocket 23 (FIG. 1B) is transferred to the entrance of the buffer 162. 170. The buffer 162 is further connected to a pulley 174 as shown in the figure. 176 * 178 t 180 t 182゜The wide blank hanging at 184 The end belt 172 includes, for example, a pulley 180 driven by a motor 186 (conventional drive means). 188) to move belt 172 in the direction of arrow 190. C IJs 174 and 184 are rotatably mounted on rods that stand up at right angles from the support plate 164. It gets worn, but boo! 7176.178 is attached to a generally rectangular movable plate 192 I get kicked. The movable plate 192 has one end 193 fixed by an arm or bracket 196. It is secured to a wire or sleeve bearing 194. sleeve 5 The bearing 194 has one end attached to a mounting block 200 fixed to the support plate 164. It is slidably attached to a fixed rod 198. of the movable plate 192 One end on the opposite side to the one end 193 has a rotating ring 202 hanging down from the lower side thereof, It supports a movable plate 192 for reciprocating motion in the direction represented by double-headed arrow 204. I'm trying to do what I can. The movable plate 192 displays documents stored in the document stack 206. It moves along the direction of the double arrow 204 so as to be able to correspond to the change in the quantity of class 28.

In that regard, pulley 208 is rotatably attached to one end of arm 208 and the remainder of the pulley 208 is One end is rotatably supported by a rod 210 rising from the support plate 164. One end is an arm The tension spring 212 fixed to the upper pulley 214 The end is hung on the bracket 196 so that, for example, the document 28 can be removed from the stack 206. The speed at which the class 28 enters the stack 206 is faster or slower than the speed at which it is output. , the tension on belt 172 remains relatively constant even when the size of document stack 206 changes. Make sure to keep it constant. a second puller having one end secured to the mounting block 200; The tension spring 216 is passed around the lower pulley 218 and its remaining end is attached to the bracket. 196, forcing the bracket 206 toward the bottom of FIG. to be biased. Pulleys 214 and 2]8 are rods that stand up vertically from the support plate 164. It is rotatably attached to the card 220. a third whose one end is fixed to the support plate 164; The tension spring 222 passes around the pulley 224 so that one end of the tension spring 222 is blocked. The block 226 is fixed to the movable plate 192, and the block 226 is fixed to the movable plate 192. Sleeve bearing 3, maintaining an equal bias to the left side of 92 (as seen in Figure 7) 94 is.

The base 224 is attached to the support plate 164 so that it can freely slide on the throat 8. It is rotatably attached to a rod 228 that stands up vertically from the top.

As previously mentioned, the movable plate 192 (FIG. 7) controls the changes in the documents 28 in the stack 206. It is used to adapt to the amount of The platform 176 stands up vertically from the movable plate]92. It is rotatably attached to a rod 230.

Boo IJ 1.78 (Fig. 7) is actually a cylindrical pulley fixed to the shaft 232. It is rotatable as can be clearly seen in FIG. One of the shafts 232 The end is rotatably attached to and held by a bearing 234 fixed to the movable plate 192. be done. Upper roller 236 and lower roller 238 are each mounted on the opposite side of roller 178. and the shaft 23 is rotated by the rotation of the roller 178. Fixed to 2. The overall configuration as shown in Figure 7 is ``0'' ring-shaped endless A belt 240 is placed over pulleys 136, 242, and 244. Pulley 242 is The soak 24 is rotatably attached to a rod 246 rising vertically from the movable plate 192. It is attached to the four-bar rod 247 in a similar manner. one 27 each Only F'-IJ242 and 244 are shown in FIG. As shown in Figure 9, pulleys are added to support the endless belts 240 and 248. can be used. Belt] 72 is an endless bell) 24'.

and 248, and the documents 28 of the stack 206 are placed between the side walls 250 (FIG. 7). is used to drive the documents 28 in the direction of movement or alignment.

Two rollers such as 242 and 244 and endless belts 240 and 248 Rollers 236 and 238 provided with a resilient bias to the nug 206 Squeeze the stack toward end wall 252. Appropriate time-out circuit (shown) after a predetermined elapsed time using Endless belt 24 for document 28 in 7 tack 206 after entering entrance part 170 0.248" to minimize scrubbing motion.

After the document 28 enters the entrance section 170 (seventh door), it passes through the wide belt 172 and the side wall 258. and idler rollers 254 and 256 which are supported for rotation. sent. A diagonal plate fixed to the block 226 for movement with the movable plate 192 260 is used to deflect the document 28 towards the endless belt 240, 248. Ru. Document bending member 262 is fixed to block 226 and extends approximately half the height of document 28. Hold the finger end 2i64 at a height equal to the height of acts to bend the document into a convex curve at that height over its length. Ru. When the material 28 is bent in such a manner, it remains "stiff" and does not collapse and the belt 240 and 248 so that it can be moved and inserted in the stack 206-J.

Documents 28 in stack 206 (FIG. 7) are attached to plate 26 under the control of solenoid 268. I'm guessing 6. The board 266 has holes drilled into it and rotates like a "'noyoga". '(jogger) 270 settles and adjusts the document 28 in the stack 206 For example, when solenoid 268 is energized, plate 2 66 and the stack 206 toward the end wall 252 so that the pickoff (pick off) The belt 272 picks up the first document 28 in the stack 206. Make it possible to take out or pick up. Solenoid 268 is energized. to separate plate 266 and stack 206 from pink-off belt 270 when to prevent documents 28 from being fed from the stack. Pickoff Bae The belt 272 is hung on rollers 274, 27'6, and the strips (roller 278 prevents more than one document 28 from being removed from stack 206 at a time.

The document 28 (FIG. 7) removed from the stack 206 is the document schematically represented in FIG. It is moved to track 280. There it is schematically represented as roller 282. A suitable document drive means is used to drive the track 280. Move the document along. Joga 2 +6 +8', 1 pickoff belt 27 2 and Stritz 1 roller: 27.8+, for example, motor:]! B It is also possible to use an individual motor (with a power of 1 as shown in the figure). 075 - The charger can be driven in the conventional manner through the threading means. Star It is fixed to the movable plate 192 when there are no more documents 28 on the rack 206. A light source and sensor combination 284 working in conjunction with a reflective surface 286 is used, as shown in FIG. encoder processor 288.

Encoder processor 288 carries encoder 1 to buffer 162 and 60.

Encoder processor 288 (Figure 8) li:r-7:1-da pro Processor 288 is connected to the interface of unprocessor 44 (FIG. 1A). includes an interface 290 for connecting to a base 292; Interfuneyy, 294 (Figure 8) Id-[-encoder processor 288 as document buffer 162 (collective 1nte) rface), and its action is controlled by the motor 186, solenoid 268ET and pin. control of the pull-off belt 272 and the control of the Pants 7162 can be placed in various locations on buffer 162 to operate in a conventional manner. and receiving signals from various sensors (not shown) capable of nothing. The encoder 70 processor 288 further connects it to the encoder 160. interface 298 (!l:, 7') an interface 300 that connects to an fCR reader 302. additional interface The base 304 may encode various document position detectors 3016, 308, for example. It is used to connect to the main processor 288. Zorosessa 288 is further All interconnected as usual via control and interface 318 ROM 312, RAM 314 and Zorosesa (MP) 316 Including an encoder processor called ``Parintelligent'' terminal cherry. 288 to be able to operate.

The second type of rejection documents are essentially rejected in the order they appear in the stack 24 (Figure 1A). is diverted to the absolute pocket 23 (FIG. 1B) and accumulated in the stack 206 of the buffer 162. It is accumulated.

Means 102 for data completion (FIG. 1B) receives rejections of the first type and the second type. Completed data for rejected documents. data capture file 104 (as shown in Figure 5). (or after completion of the data) and/or recorded in the document reentry file 158. be remembered. Completed data in file 104 or 158 is written in stack 24. Since the overall order is the same as that of the class, the completed data and the buffer 162 Corresponding physical documents 28 and 31 from stack 206 Matching or matching is easy. File 104 or 158 (Figure 1A) The data is further tagged as sugar 1 type rejection or sugar type 2 rejection. It should also be noted that

The data on line 108-1 of reentry file 158 in FIG. For example, assuming that we are dealing with the second type of rejection, we will explain how to complete it below. I will clarify. Application processor 44 reads line 10 from file 158. 8-1 data through the application processor 44 (Fig. 1A). The data is sent to RAM 314 of encoder processor 288 (FIG. 8). Next, Processor 288 "actuates" or When energized, picker belt 272 removes the first document 28 from stack 206. The document is represented by document track 2800 and roller 282 (FIG. 8). Exit section 2 adapted to be picked up and moved by a document transport It will be moved towards 96 (Calculation 7 diagram).

A sensor 306 positioned along track 280 detects the leading edge of document 28. and displays it on encoder processor 288. According to selection (optional) If a MICRIJ-da 302 is utilized, the reader will be energized and the incense burner will open. Read the MI CR line and send the read data to the encoder processor 2. 88, where at least a portion of the MICR data just read and the RAM 14 The encoder 160 compares the data with the stored data and prints it on the document. the amount to be encoded or encoded will be printed on the correct document 28. make sure what you say. However, the document 28 is, for example, a document M IJ entry file. Since the data is stored in the same order as the corresponding data in file 158 (Figure 1A), The use of MICRIJ-da 302 is not absolutely necessary.

As document 28 moves along track 280, second sensor 308 (FIG. 8) If the leading edge of the document is the amount from the encoder processor 288, for example, the MIC Conventional encoders such as those encoded or printed on documents on RIC] 60 It shows that it is approaching. After the document 28 is encoded, the encoded Coda Prose, Nosa 2880 edition, complete with deflector 320 under C. It is deflected to the pocket 318 of documents that have been changed or "modified". For example, by way of example, the encoder processor 288 may: Data obtained from the reader 302 and the data obtained from the entry file 158 Because there was no match in the comparison of If you decide not to do so, what about that particular? F is encoder process An optional rejection pocket is provided by deflector 324 under the control of sensor 288. can be deflected to the cut 322. The remaining document 28 in the stack 206 is the same by law en 3 will be coded.

In order to complete the physical one entry of Anburi 28, the rejection pocket of sorter 12 is The first type of rejection document 28 in Figure 16 (Figure 1B) and Iv! F'E module 13 (Figure 8), 2nd queso) 318, 322, the second type of rejection document 28 is there. Sorter as a second sorting operation (but with the same order preserved) 12 is mounted on a stack 24. Re-read those rejected documents 28 Moves sequentially through MICRIJ-da 32 and OCR reader 34 in order to try to read be done. Although there is a high possibility that these rejection documents 28 will not be completely read again, No matter what part of the data is read, the data application process Comparison operations on that portion of the data not attempted by the processor 44 will result in the rejected document 2 8 in the same order as they appeared in the document reentry file 158? It cannot be done for the purpose of reading to check whether it is correct or not. These documents28 is usually incomplete because the data read from it is incomplete, e.g. rejection pocket 1 would have been classified as 6, but the data completed for that purpose is now ampdated. data capture file 104 (year 5) and/or document entry file 158 (Fig. 1A and Fig. 6), and this completed data is The application processor 44 uses the rejected document 28 to In the explanation, the pocket code should be transferred or classified to the appropriate pocket 18-22. command to the controller 42.

Completed data present in data capture file 104 (Figures 1A and 5) As mentioned above, the data is used for calculations such as total balance and update calculations encountered in normal banking operations. It can be used to perform various matching processes. Other system 10 As a percentage, the document re-entry file 158 is used for subsequent normal banking operations. The document 28 regarding the transfer (issued bank) part For banks that have to process can be used to provide a list of documents 28 that you may have. That's what I mean.

Next, certain general matters will be explained. Description of preferred embodiments of the invention The type of sorter 12 selected for Model ≠6780 manufactured by Poration Co., Ltd. can be used. This sorter has a length of approx. Approximately 1,400 sheets of 152.4+nmO paper can be passed per minute. The typical refusal The absolute rate is about 2% of documents28 with ``incomplete'' data, but the maximum expected Or the worst rejection rate is about 28 sheets per minute. This rejection rate is shown in the data entry in Figure 1B. The operator at the bird terminal 128.130 is fully operable.

For example, you can use the proof-ma machine to type the entire amount field. Chine)'s Olay 35 In burst (continuous) mode, a keystroke speed of 65 keys per minute was measured. death However, during a normal working day, it is possible to maintain a typing speed of 25 to 35 keys per minute. can. Therefore, the rejection rate of about 28 sheets per minute or 2 V is 128, so the data of 2 machines ・This seems to be a range that can be easily handled by entry terminals.

Approximately out of the total rejection rate of approximately 28 sheets per minute for the first type and second type explained above. Section 10 concerns the second type rejection where the amount is completely missing from the document. Ru. This means that the second type of rejection is 10% of the 28 sheets per minute, or about 3 sheets per minute. This means that This second type of rejection rate is significantly lower than that of a fairly cheap and slow encoder. This means that the printer 160 can be used for the processing.

FIG.4 FIG.8 international search report

Claims (1)

[Claims] 1. Transfer the document (28) with the encoded data field to the feed path (14). ) and a transport means arranged to move sequentially in the feeding direction along A reader adapted to read data from the document (28) while the document (28) is being fed. in accordance with the predetermined criteria for said data. Conversion means removed to divert failed documents to rejection area (23) (・4O-2 ), the document processing device is provided adjacent to the rejection area (23); Receive the rejected document from the rejection area (23) and send the rejected document (28) to the rejected document (28). ) includes an encoding means (13) adapted to encode the missing data. document processing equipment. 2 @The arrangement processing device (13) is provided in the encoding means (13) and Claim 1 comprising a printer (160) for gliding data onto an extinct document. Apparatus described in section. 3. Each said rejected document has a missing data field therein, and said device transmitting data for the missing data field to the printer (160); Claim 2 comprising data generating means (116, 128) for supplying data. As described! . 4 Which of the above-mentioned books 79 is the reading means (32, 34, 46) Check whether the class (28) has the missing field, and check if there is no missing field. Complete signal and for document (28) and document (28) with said missing fields. 4. A device as claimed in claim 3, adapted to provide an incomplete signal. 5. The encoding means (13) stores the received documents (28) and performs them on a first-in, first-out basis. document storage in which the document (28) is sent to the printer (160); Means (162, device according to claim 4). 6 The @ device stores the data read from the document and stores the data about the document (28). a data record that allows the read data to be identified as complete or incomplete; processing means (44) including storage means (104, 158); and said reading means (32; 4.46) located along and passing through the feed path (14) downstream from the 1-D (28) each denotized image data is generated. imaging means (73), and the data generation means (116, 128) A memory (116) and a booth configured to store the Dinott image data. delay means (140) and data entry means (142); The missing data for the document (28) with the incomplete signal is recorded in the memo IJ-(116). ) and the influence of the said document (28) extracted from C from the related image data read from Both the display means (14, O) and the display means (14, O) are displayed. - Entry of said missing data and said processing procedure via data entry means (142); is performed on the storage means (]]04., 158) via the stage (44). transfer of the missing data, and the stored missing data is transferred to the memory hand. The encoder is then transferred from the stage (104, 158) to the encoding means (]]3. Tried to encode a document (28) with missing data fields 6. The device according to claim 5, characterized in that: 7. Move the document (28) primarily in the feeding direction along the feeding path (1, 4) and Read the data from the document (28) that has been sent, and Each step of converting documents (28) that fail to meet the criteria to the rejection area (23) a document processing method comprising: a document (28) diverted to said rejection area (23); Generate data for missing data fields and divert to said rejected flight area (23) an encoding means (13) arranged adjacent to said rejection area to receive rejected documents; and the corresponding completed data of each said document obtained from said data generation step. A document processing method including each step of encoding in the encoding means (13) Law. 8 The document processing method is = document (28) in the storage means (104, 158) The data read from the book # (28) is stored and the data read from the book # (28) is stored. ) so that the data read from the feed path can be identified as complete or incomplete; While moving along (14) ^ff After the reading process, take a photo of the document (28) and send it to each person. Denoted shadow for IJ document (28) □□□Each process that supplies data The data generation step is started after the data conversion step 8, and the data generation step is started after the reading step. Using the related data and the related Dinott image data from the KJ photography process, generate the complete data for the corresponding document (28) and record the complete data in said The method according to claim 7, which includes each step of storing in the storage means (104, ]-58). Law. 9. Claim 7, wherein the encoding step is accomplished using magnetic ink. the method of.