JP2009011874A - Business form sorting method, and optical character reading system using the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a business form sorting method without recognition of characters or numbering. <P>SOLUTION: In the business form sorting method, a plurality of business forms 1 layered in a hopper part 2 are conveyed to a scanning part 6 and image information of the business forms 1 is read; logo and arrangement feature patterns of frames included in the read image information are determined; the image information is stored in an image DB for every determined arrangement feature pattern; the business forms 1 are read again by the scanning part 6; the re-read image information is compared to the stored image information by pattern matching; and the business forms of the image information with the patterns matching are discharged to a given stacker of a sorter stacker part 3 corresponding to the arrangement feature patterns, to thereby sort the plurality of business forms without recognition of characters or numbering. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a form sorting method capable of automatically sorting various forms in an optical character reading system, and an optical character reading system using the form sorting method.

  In general, an optical character reading system optically scans characters written on a form, cuts out characters from the scanned image information, and recognizes the characters. There is a demand for an automatic sorting technique in which a bundle of forms including a plurality of forms is set, the form is sequentially read from the set form bundle, and the read forms are sorted by type.

  The following patent document is cited as a document that describes an automatic sorting technique for sorting this form by type. Patent Document 1 includes a reading unit that reads data written in a form, and data of the read form. Recognizing recognition means, data storage means for storing the read form data and the character data of the form recognized by the recognition means, temporary holding means for temporarily holding the data read form, and the data storage Correction means for confirming and correcting the form data stored in the means, a plurality of storage means for sorting and storing the forms held in the temporary holding means, and the temporary holding means based on the data checked and corrected by the correcting means And a sorting control means for controlling sorting and storing of the held forms in the storing means, and temporarily storing all the read forms in the temporary holding means. Distillate City, based on the recognition result of the stored characters has been described to perform the sorting of the form in the storage means.

Patent Document 2 automatically reads images of a wide variety of account transfer request documents in order, and numbers each stored file data such as reception date and bundle serial number by invisible ink in each account transfer request document; Image data entry process for creating image data and attribute data for sorting account transfer request documents based on scanned image data of account transfer request documents, and attribute data for sorting account transfer request documents to a computer The process of creating an import sorting table and the invisible storage file data of a wide variety of account transfer request forms are read by the scanner of the form sort processing device. A wide variety of account transfer request forms are automatically created based on the read data and the sort table. And the process of sorting the direct debit request form .
JP 2006-143369 A JP 2002-318924 A

  However, in the technique described in Patent Document 1, when there are mixed forms that cannot be recognized, the administrator is resident in the apparatus, and each time an unreadable form is generated, the unrecognized form is manually sorted. There is a problem that the management work of the operator becomes complicated.

  Further, the technique described in Patent Document 2 requires that the storage file data be numbered in the form, and cannot be sorted for forms that are difficult to number in various forms. was there.

  An object of the present invention is to provide a form sorting method capable of sorting forms without recognizing characters and numbering, and an optical character reading system using the form sorting method.

In order to solve the above-described object, the present invention sorts a hopper unit that stacks a plurality of forms, a scanning unit that reads image information of a form conveyed from the hopper unit, and a form that has been read by the scanning unit. An optical character reading system comprising: a sorter stacker unit that is stored in a plurality of stackers; and a computer that conveys a form from the hopper unit to a scan unit, reads image information, and discharges the read form to a soft stacker unit Because
The computer is
A plurality of forms stacked on the hopper unit are conveyed to a scanning unit to read image information, a logo and a frame layout feature pattern included in the read image information are determined, and an image for each determined layout feature pattern A first step of storing information;
The plurality of forms are read again by a scanning unit, the read re-read image information and the image information stored in the first step are compared by pattern matching, and a form of image information having a matching pattern is obtained as the layout feature pattern. And a second step of discharging to a predetermined stacker of the sorter stacker unit according to the first feature.

  In the optical character reading system, the present invention further includes a temporary stacker unit that temporarily stores a form whose image information has been read by the scan unit, and the form whose image information has been read in the first step is stored in the temporary stacker. The second feature is that the form stored in the temporary stacker unit in the second step is transferred to the hopper unit and the form stored in the hopper unit is read.

Furthermore, the present invention provides a hopper unit that stacks a plurality of forms, a scan unit that reads image information of a form conveyed from the hopper unit, and a form that has been read image information by the scan unit. A form sorting method in an optical character reading system comprising: a sorter stacker section to be stored; and a computer that transports a form from the hopper section to a scan section to read image information, and discharges the read form to a soft stacker section. And
The computer is
A plurality of forms stacked on the hopper unit are conveyed to a scanning unit to read image information, a logo and a frame layout feature pattern included in the read image information are determined, and an image for each determined layout feature pattern Remember information,
The plurality of forms are read again by a scanning unit, the re-read image information and the stored image information are compared by pattern matching, and the form of the image information whose pattern matches is determined according to the arrangement feature pattern A third feature is that the sheet is discharged to a predetermined stacker of the sorter stacker unit.

  According to the present invention, in the form sorting method, the optical character reading system is provided with a temporary stacker unit that temporarily stores the form obtained by reading the image information by the scanning unit, and the form storing the image information for each arrangement feature pattern. Is stored in the temporary stacker unit, the form stored in the temporary stacker unit is transferred to the hopper unit, and the form stored in the hopper unit is read.

  The present invention determines the type of a form using a different logo and frame arrangement characteristic pattern for each type of form, then reloads the image data of the form, and the image information of the form for which the type is determined Character information and numbering are recognized by comparing the image information that has been read again by pattern matching and discharging the form of the image information that matches the pattern to a predetermined stacker of the sorter stacker unit according to the arrangement feature pattern. The forms can be sorted without any problems.

  Hereinafter, an optical character reading system using a form sorting method according to the present invention will be described in detail with reference to the drawings. 1 to 4 are diagrams showing respective steps of an optical character reading system according to an embodiment of the present invention, FIG. 5 is a diagram for explaining an example of a form that is an object of the present invention, and FIG. 6 is a form according to a form type. FIG. 7 is a diagram for explaining grouping based on the form characteristics, FIG. 8 is a diagram showing a flow of the first step of the form sorting method according to the present embodiment, and FIG. It is a figure which shows the flow of the 2nd process of the form classification method by embodiment.

  First, as shown in FIG. 1, the optical character reading system according to the present embodiment scans a hopper unit 2 that stacks a plurality of forms 1, a sheet one by one from the hopper unit 2, and reads image information of the form 1. Section 6, sorter stacker section 3 including a plurality of stackers for storing and storing forms 1 scanned by scanning section 6 by an operation described later, and reject forms in which scanning errors are caused by scanning section 6 A reject stacker unit 4, a personal computer (PC) 5 that controls each of the parts, and an image database 7 that stores arrangement feature patterns and read image information, which will be described later. The document 1 stacked on the sheet is fed out one by one, the image information of the form is read by the scanning unit 6, and the text included in the image information is read. While it recognized as character data cut out, and is configured so as to be stored in sorter stacker 3 to sort the document to the reading has been completed.

  The PC 5 includes a hard disk device that stores an OS and various programs, a memory, a CPU, a keyboard, a display device, and the like. In particular, the hard disk device has an image database (DB) that stores image information in groups. Yes. The reject stacker unit 4 is generally a discharge destination of the unreadable form described above, but is used as a temporary storage place for the form in the present embodiment as will be described later.

  Here, the forms to be read in the present embodiment will be described. These forms are roughly divided into four groups A to D as shown in FIG. 5, for example. A form 51 in which a title logo (“enrollment application form”) and a two-stage frame for entering predetermined items are arranged on a sheet of paper, and a three-column, three-stage sheet on a 15 cm × 20 cm size sheet like Group B A form 52 with an entry frame and name and date entry fields arranged as a logo, and a logo ("Notice", "Campaign") with no frame on a 15cm x 15cm size paper, such as Group C, is placed. It is configured like a form 53. In the present application, the feature of the layout of these various groups of form sizes, logos, and frames is referred to as “placement feature pattern”, and this placement feature pattern is assumed to be registered in advance in the hard disk device or the image database 7.

Next, the operation of the optical character reading system configured as described above will be described with reference to the configuration diagrams of FIGS. 1 to 4 and the flowcharts of FIGS.
First, in the optical character reading system according to the present embodiment, as shown in FIGS. 1 and 8, as the first step, step S801 in which a plurality of forms 1 are stacked on the hopper 2 and the plurality of stacked forms 1 are combined. Step S802 that sequentially feeds and conveys one sheet at a time, reads the image information of the conveyed form 1 by the scanning unit 6, step S803 that temporarily discharges the read form 1 to the reject stacker unit, and Step S804 of storing the image information of the form read in step S802 in the image database (DB) 7 is executed.

  Next, the system compares the layout feature pattern of a plurality of forms stored in advance in the hard disk device with the image information of the plurality of forms read in step S804, and based on the form size / frame position / logo, the layout feature. Step S805 for calculating the similarity between the pattern and the read image information, Step S806 for grouping the calculated similarity for each group based on a predetermined fixed threshold value, and the grouping performed in Step S806 is correct. By executing step S807 for inputting user confirmation correction such as whether or not, and step S808 for designating a stacker for discharging the form grouped in the above steps, a plurality of read forms as shown in FIG. Image database (DB) by group (placement feature pattern) And stores it in the 7.

  As shown in FIG. 7, the storage in the image data DB 7 assigns numbers in the order of the read forms, and the form input numbers 8, 13, 14... Are group A, form input numbers 2, 4, 12 and so on. .. Are performed in groups B and form input numbers 1, 5, 6,.

  Next, the optical character reading system is set so that the plurality of forms 1 temporarily stored in the reject stacker unit 4 in step S803 are returned to the hopper unit 2 as in the second step shown in FIGS. Step S 901, step S 902 that sequentially conveys the set form 1 and reads the image information by the scanning unit 6, and the read image information and the image information stored in the image database 7 of the hard disk device by pattern matching. The comparison is performed, the group of the read image information is determined, and the step S904 of discharging to a predetermined stacker of the sorter stacker unit 3 designated in advance for each group is executed.

  As described above, the form sorting method and the optical character reading system according to the present embodiment group forms using the feature size pattern based on the form size, logo, and frame layout included in the image information of the form read for the first time. Pattern matching is performed between the first step of storing the grouped image information and the image information of the form read in the second time and the image information grouped in the first step. By performing the second step of discharging the form to the designated stacker according to the above, a plurality of forms can be sorted without recognizing the characters and numbering described in the form.

  In the above-described embodiment, the example in which the image information of the form is read twice has been described. However, the fact that the forms grouped in step S808 in FIG. 8 are stacked in a predetermined order is used. The form temporarily stored in the reject stacker unit 4 based on the form input number shown in FIG. 7 may be discharged to a predetermined stacker of the sorter stacker unit 3 with reference to the table of FIG. . Further, in the above-described embodiment, the image information and the like are described as being stored in a database or a hard disk device, but may be stored in a large capacity semiconductor memory. Further, in the embodiment, the example in which the form scanned in step S902 in FIG. 9 is directly discharged to the sorter unit has been described. However, the scanned form is temporarily stored in the reject stacker unit as described above. Based on the form input number, it may be discharged to a predetermined stacker of the sorter stacker unit according to the form type (arrangement feature pattern).

The figure which shows the 1st process of the optical character reading system by one Embodiment of this invention. The figure which shows the 2nd process of the optical character reading system by this embodiment. The figure which shows the 3rd process of the optical character reading system by this embodiment. The figure which shows the 4th process of the optical character reading system by this embodiment. The figure for demonstrating the example of a form used as the object of this invention. The figure for demonstrating the characteristic of the form by form classification. The figure for demonstrating the grouping by the characteristic of a form. The figure which shows the flow of the form classification method by this embodiment. The figure which shows the flow of the form classification method by this embodiment.

Explanation of symbols

  1: form, 2: hopper unit, 3: sorter stacker unit, 4: reject stacker unit, 5: personal computer, 6: scan unit, 7: image database.

Claims (4)

A hopper unit that stacks a plurality of forms, a scan unit that reads image information of a form conveyed from the hopper unit, and a sorter stacker unit that sorts the forms read image information by the scan unit and stores them in a plurality of stackers And an optical character reading system comprising a computer that conveys a form from the hopper part to a scanning part, reads image information, and discharges the read form to a soft stacker part,
The computer is
A plurality of forms stacked on the hopper unit are conveyed to a scanning unit to read image information, a logo and a frame layout feature pattern included in the read image information are determined, and an image for each determined layout feature pattern A first step of storing information;
The plurality of forms are read again by a scanning unit, the read re-read image information and the image information stored in the first step are compared by pattern matching, and a form of image information having a matching pattern is obtained as the layout feature pattern. And a second step of discharging to a predetermined stacker of the sorter stacker unit corresponding to the optical character reading system.   A temporary stacker unit for temporarily storing a form whose image information has been read by the scan unit; storing the form whose image information has been read in the first step in the temporary stacker unit; and the temporary stacker in the second step. The optical character reading system according to claim 1, wherein the form stored in the hopper is transferred to the hopper and the form stored in the hopper is read. A hopper unit that stacks a plurality of forms, a scan unit that reads image information of a form conveyed from the hopper unit, and a sorter stacker unit that sorts the forms read image information by the scan unit and stores them in a plurality of stackers And a form sorting method in an optical character reading system comprising a computer that conveys a form from the hopper part to a scanning part, reads image information, and discharges the read form to a soft stacker part,
The computer is
A plurality of forms stacked on the hopper unit are conveyed to a scanning unit to read image information, a logo and a frame layout feature pattern included in the read image information are determined, and an image for each determined layout feature pattern Remember information,
The plurality of forms are read again by a scanning unit, the read re-read image information and the stored image information are compared by pattern matching, and the form of the image information having the matched pattern is selected as a sorter corresponding to the arrangement feature pattern A form sorting method for discharging to a predetermined stacker in the stacker unit.   The optical character reading system is provided with a temporary stacker unit for temporarily storing a form whose image information has been read by a scanning unit, and a form storing the image information for each arrangement feature pattern is stored in the temporary stacker unit, 4. The form sorting method according to claim 3, wherein the form stored in the temporary stacker is transferred to the hopper, and the form stored in the hopper is read.

Images