JP2005234790A - Handwritten slip processing system and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a handwritten slip processing system and method suitable for the processing of a slip having choices. <P>SOLUTION: This handwritten slip processing system for acquiring information from a handwritten slip for processing the handwritten slip is provided with an image information acquiring means which acquires slip definition information to be used for acquiring the information from the slip and the image information of a region which is estimated to be handwritten in the slip based on the slip definition information, a handwritten information detecting means which detects the handwritten information from image information acquired in the region belonging to group elements constituted of a plurality of regions decided based on the slip definition information by the image information acquiring means, a converting means which converts the handwritten information detected by the handwritten information detecting means into a predetermined value based on the slip definition information and a group information acquiring means which acquires group information being information as the group elements from the value converted by the converting means based on the slip definition information. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a handwritten form processing system and a handwritten form processing method that acquire information from a handwritten form and perform processing.

  In the field of document exchange such as submission of various applications to public institutions, general companies, etc. represented by government agencies, the trend toward paperless is generalized, and movement to replace conventional paper media information with electronic information Is being promoted everywhere. For example, various electronic information transmission methods such as a method using various groupware, e-mail, WEB browser and ASP (Active Server Page) have been proposed and put into practical use.

  On the other hand, conventional paper-based applications and document transmissions are still widely performed. Possible reasons for this are as follows. In other words, it is easy to fill in paper without being restricted by location or equipment, and paper generally has a higher display resolution than electronic equipment. The point where information entry is required is high, and the visual searchability is high. In general, handwritten entry does not require familiarity with the operation method of electronic devices or application software. It is a point that can be easily taken by doing.

  Many methods, apparatuses, and systems for processing such paper entry have been proposed as follows.

  Patent Document 1 discloses an information processing system. This information processing system includes a means for interactively creating a form having a frame structure, a means for defining input / output data of the form, a means for printing a form, a means for inputting a written form image, and an input image. Processing of written forms, mainly having means for cutting out areas according to input / output data, means for taking out written contents from the cut out area data, and means for performing conditional branching and character recognition on the extracted contents Compared with the corresponding input / output data, the system grasps the written contents, performs conditional branching and information conversion from the contents as necessary, and automates the processing of the information written in the form. Provided is an apparatus for detecting a character / mark / image region described in a handwritten / printed document and outputting the document in an output form different from a form layout. In addition, the processing of a form whose output changes depending on the form entry form is automated to eliminate human input errors.

  Patent Document 2 discloses an information processing apparatus and a storage medium storing the program. This patent document discloses an information processing apparatus that can easily digitize printed matter, handwritten documents, forms, and the like. By comparing the image data read from the input material with multiple selection criteria image data as selection criteria, automatic format selection means is provided to automatically select the relevant selection criteria image data from this selection criteria image data, Based on the selected selection standard image data, image data processing of the input object is performed. Further, image data processing is performed from a form definition body that defines input / output data of the form corresponding to the selected selection reference image data.

  Patent Document 3 discloses a character recognition method and apparatus. The accuracy of handwriting / type discrimination in character recognition processing is improved, the processing time for character recognition is shortened, and the character recognition accuracy is improved. A character string is extracted, handwriting / type discrimination processing is performed, and handwritten character recognition processing or type recognition processing is performed based on the result. In the handwriting / type discrimination process, a character is extracted from the character string without including a cloud point, the center position of all the extracted characters is obtained, and the discrimination between the handwritten character and the type is performed based on the regularity. In addition, the center positions of all characters extracted without including lowercase letters included in the character string are obtained, and handwritten characters and printed characters are discriminated from their regularity. As a result, it is possible to solve the problem that the variation in the center position is large in characters including dakuten and lowercase letters, and handwriting / character discrimination processing is possible. Moreover, the accuracy of the handwriting / type discrimination process can be improved by performing the handwriting / type discrimination process using layout information such as a form.

  Patent Document 4 discloses an optical character reader. An optical character reader that detects the color printed in a designated area of a form and identifies the type of form is provided. The form 1 is converted into a color image, and a color image in which a rectangular area wider than the color ID area is thinned out, and a more accurate color image in the color ID area is extracted from the form position information extracted from the converted binary image. Among the extracted color information of each pixel, those having saturation S and luminance V equal to or lower than the threshold value are excluded as achromatic colors, and a feature amount is extracted from the remaining color information to identify a color ID.

  Patent Document 5 discloses a character recognition method and apparatus. A cell structure is appropriately recognized from a form format in which one item is represented by a plurality of cells or a set of data is represented by a plurality of cells, and character recognition processing corresponding to the cell structure is performed. The character recognition system performs cell extraction processing to extract the minimum unit of the area surrounded by ruled lines as cells from the image data obtained by reading the image written on the form, etc., and aligns the cells in the row direction. The processing and column alignment process specify item rows and data rows, and table cell structure data representing the structure of each cell and the relationship between cells is created. In the recognition process, this table cell structure data Based on this, recognition processing for each cell is performed.

  Patent Document 6 discloses a device for detecting a character recording area in a form. The present invention provides a character recording area detection device in a form for detecting a character recording area in an existing form, a method for detecting a character recording area in a form, a recording medium, and a form format creation device. An image of a plurality of completed forms in which different characters are entered in an existing form is input, and a common image and a difference image are created based on the input plurality of images. A character recording area is detected based on the created common image and difference image, and an attribute is set in the character recording area. The format of the completed form is set by associating the detected character recording area with the set attribute.

An apparatus or system that performs processing on handwritten paper as described above is referred to as a handwritten form processing system.
Japanese Patent Publication No. 7-107695 JP 2000-132542 A JP 2000-331122 A JP 2001-109842 A JP 2001-109843 A JP 2001-243423 A

In a general handwritten form processing system, processing is performed in the following steps.
1. Creation of paper forms2. 2. Image reading and identification of handwritten paper forms 3. Acquisition of handwritten partial images 4. Handwritten image detection in partial image 5. Information conversion for the partial image (handwritten image) Storage of the information conversion result In such a handwritten form processing system, there are two problems particularly when the form has a choice. One is the allocation of data corresponding to the selected item, and the other is the number of man-hours of the form processing program related to the presentation and correction of the information conversion result to the user.

  Generally, in a conventional handwritten form processing system having a form definition body, the form definition body is defined in two forms as shown in FIG.

  In the first form, an input element b2 capable of defining an input processing attribute such as handwriting is arranged on an arbitrary image original b1 in a form a, and various attributes (position, size, outline) of the image original and the input element are arranged. (Shape, line type, color, ...) is saved in the form definition body.

  In the second form, an output element c1 that controls display on the form a and an input element c2 that can define input processing such as handwriting are arranged, and these attributes are stored in the form definition body.

In a handwritten form processing system having such a form definition body, there are two major problems especially when the form has a configuration with options.
A) Data allocation corresponding to the selected item
B) A large number of man-hours for the form processing program for presenting and correcting information conversion results to the user. A case where contents having options such as a check box are included in the form as shown in FIG. 18 is taken as an example. For paper forms having options, marks such as check marks, x, and fill are described in the corresponding check boxes. These are determined by mark recognition, which is a special form of character recognition.

  Mark recognition returns a value (Boolean) that is either selected (True) or not selected (False). On the other hand, the result necessary for the operator who processes the form is not “whether or not it is selected” but “man” when it is selected, and no value is required if it is not selected. That is, in the case of the prior art, in order for the operator to finally return the necessary result, it is necessary to create a program module that replaces the Boolean value with the necessary value inside the handwriting processing program.

  For this reason, since the choices on the form vary depending on the form, the program module develops a unique module for each form, which increases the number of development processes.

  On the other hand, in another conventional example, the above problem is solved by taking the attribute structure of the drawing element as shown in FIG. What is characteristic is that the drawing element has an attribute for outputting a value to the element and an attribute for inputting a value from the element, and further, in the input attribute, conversion manner data that is a definition of the information conversion manner of the partial image for input processing, In addition, it has a plurality of input value data, and the output attribute has output value data which is the data itself displayed in the element.

  As shown in FIG. 21, the conversion manner data is classified by being hierarchized according to the type of information that can be input, and the information conversion method is defined by a keyword indicating the classification for each hierarchy. In this prior art, as shown in FIG. 19, the input value data 1 (image of the entry mark) is stored as True in the input value data 2 using the Boolean of the input value conversion method, and further the output of the input value conversion method. Using the value (in the case of True, it means replacing with the output value data), the output value data of this element is stored as the input value data 3.

  That is, if a mark is included in a square indicating a man on the form, the value “male” can be returned. However, for example, when there are a plurality of options in the prior art, a value necessary for the operator cannot be obtained.

  Next, the problem B will be described. In general, a handwritten form processing system uses character recognition technology for information conversion. However, in character recognition, especially handwritten character recognition is difficult to recognize 100% automatically, and erroneous recognition occurs.

  As a countermeasure, the handwriting processing program has a step for the operator to determine whether the entered data is correctly converted during the form processing. An example of this step is shown in FIG.

  FIG. 22 shows an entry form a, a result b, and confirmation correction forms c, d, and e. The result b is the information obtained by converting the entry form a. The confirmation / correction forms c, d, and e are displayed in correspondence with items indicating whether or not the conversion result of the input data is correct. The display editing function for displaying the confirmation / correction form may have a function for editing a misrecognized one.

  First, the conventional technology cannot generalize the display editing function. In other words, a unique module is developed for each form, which increases the number of development processes. Further, in the prior art, the limit is to select an option for one drawing element. In other words, as shown in the confirmation correction forms c, d, and e, “registration” or “deletion” cannot be selected, and whether “deletion” is selected or not, “registration” is selected. Whether it is general or not.

  There are few problems when there are few choices in the form, but as the number of choices increases, the content displayed by the input / output value display editing means increases.

  In view of such problems, an object of the present invention is to provide a handwritten form processing system and a handwritten form processing method suitable for processing a form having options.

  In order to solve the above problems, the present invention obtains information from a handwritten form, and in the handwritten form processing system for processing, forms definition information used for obtaining information from the form, and the form definition An image information acquisition unit that acquires image information of an area that is assumed to be handwritten on the form based on information, and a group element that includes a plurality of areas defined by the form definition information by the image information acquisition unit Handwritten information detecting means for detecting handwritten information from image information acquired in a region belonging to the information, and converting means for converting the handwritten information detected by the handwritten information detecting means into a predetermined value based on the form definition information And a group for obtaining group information, which is information as the group element, based on the form definition information from the value converted by the conversion means. And having a flop information acquisition means.

  In order to solve the above-mentioned problem, the present invention is characterized in that the information acquisition means uses, as the group information, a value acquired from one of the areas belonging to the group element.

  In order to solve the above-mentioned problem, the present invention is characterized in that the information acquisition means uses, as the group information, values acquired from a plurality of areas among the areas belonging to the group element.

  In order to solve the above-mentioned problems, the present invention is characterized in that the information acquisition means uses as a group information a character string that is a value acquired from an area belonging to the group element.

  In order to solve the above problems, the present invention includes a display unit that displays the plurality of group information, and further includes a confirmation correction unit that allows an operator to confirm or correct the group information displayed on the display unit. And

  In order to solve the above problems, the present invention provides a handwritten form processing method for obtaining information from a handwritten form and performing processing based on form definition information used for obtaining information from the form. Acquired in an area belonging to a group element composed of a plurality of areas defined in the form definition information in an image information acquisition stage for acquiring image information of an area assumed to be handwritten, and the image information acquisition stage. From the image information, the handwritten information detection stage for detecting handwritten information, the conversion stage for converting the handwritten information detected in the handwriting information detection stage to a predetermined value, and the value converted in the conversion stage, the group And a group information acquisition step of acquiring group information which is information as an element.

  In order to solve the above-described problem, the present invention is characterized in that, in the information acquisition stage, a value acquired from one of the areas belonging to the group element is used as the group information.

  In order to solve the above-described problem, the present invention is characterized in that, in the information acquisition stage, values acquired from a plurality of areas among the areas belonging to the group element are used as the group information.

  In order to solve the above-described problem, the present invention is characterized in that, in the information acquisition step, a group of character strings that are values acquired from regions belonging to the group element is used as the group information.

  In order to solve the above problems, the present invention includes a display stage for displaying the plurality of group information, and a confirmation correction stage for allowing the operator to confirm or correct the group information displayed in the display stage. And

  As described above, according to the present invention, a handwritten form processing system and a handwritten form processing method suitable for processing a form having options can be obtained.

  Embodiments of the present invention will be described below with reference to the drawings.

  An outline of a handwritten form processing system as a base of the present invention will be described with reference to FIGS. 1 and 2. FIG. 1 is an example of a hardware configuration diagram of the handwritten form processing system 31. The handwritten form processing system 31 includes a form creation device 104, a form processing device 105, a storage device 106, and a paper form 2 that is an information medium. In the following description, a handwritten entry on the paper form 2 is expressed as an entry form 20.

  The form creation device 104 in the handwritten form processing system 31 includes a form creation PC 102, a printing device 107, and a form creation PC 102. The form creation PC 102 has a form creation program installed with form creation means.

  Here, the printing apparatus 107 generally means a device that creates a printed matter, such as a printer, a printing machine, or a plotter.

  Further, the form creation program displays a GUI (Graphical User Interface) on the screen of the form creation PC 102, and interactively interacts with the operator on an electronic manuscript of the paper form 2 (which corresponds to the form definition information in the form definition body described later). It is created and stored as an electronic file in a computer storage device 106 represented by a hard disk or the like.

  Further, the form creation program outputs a paper form 2 that is a paper representation of the electronic document from the printing apparatus 107. At this time, the form creation program creates a code image 4 in which a unique ID that can uniquely refer to the form definition that has already been stored is coded, and superimposes it on the paper form for printing. When the user fills in the necessary items on this paper form 2, the entry form 20 is completed.

  The form processing apparatus 105 includes a form processing PC 103 and a reading apparatus 108. The reading device 108 is a scanner in FIG.

  The form processing PC 103 has a handwriting processing program installed with form processing means installed therein. The handwriting processing program 101 captures an entry form as an image by the reading device 108. Then, the code image 4 is detected and decoded from the captured image, and the reference information of the form definition body existing in the storage device 106 is restored.

  Then, the handwriting processing program 101 takes in the form definition from the reference information, and processes the read image according to the form processing manner described in the form definition.

  The storage device 106 includes a data server 109. A database program is installed in the data server 109.

  The form creation PC 102, form processing PC 103, and data server 109 described above can communicate with each other via a network.

  Next, an example of processing by the handwritten form processing system 31 will be described with reference to FIG. FIG. 2 shows an entry form 20 as a withdrawal / enrollment application form, a handwritten form processing system 31, and an information table 22.

  The entry form 20 is a paper form created in advance by the form creation device 104 and is handwritten on the paper form. This entry form 20 has entry portions A to E and a code image 4, and necessary items have already been handwritten in each entry portion.

  When such an entry form 20 is applied to the handwritten form processing system 31, more specifically, the form processing apparatus 105, the handwritten part is extracted by the handwriting processing program, and the entry data is appropriately converted into the specific cell of the information table 22 (character recognition). ) And saved.

  Here, it is assumed that the information table 22 exists on the data server of the storage device 106, and a specific example of the information table 22 and association with the storage destination will be described later.

  In this way, the handwritten form processing system automates the processing of forms whose output changes depending on the form entry form, so that handwritten information can be reused as digital information.

  Next, details of the handwritten form processing system will be described.

  First, the form creation program will be described. The form creation program literally has a function of electronically creating forms. The GUI of this form creation program will be described with reference to FIG.

  In FIG. 3, three GUIs 7a and 7b and a property 501 are shown. Of these, the GUI will be described first using the GUI 7a.

  The GUI 7 a includes a layout area 200 and a tool box 201. The tool box 201 includes a label tool 202, a text box tool 203, a check box tool 204, and a barcode tool 205.

  Using such a GUI, the operator selectively places appropriate layout elements from the toolbox in the layout area. For example, a field for inputting a name is created using the text box tool 203 as shown in the GUI 7b. Each created layout element has a property that is a set of attribute values of the layout element. A property 501 indicates a property of the created text box.

  The property 501 includes, as attribute values, an ID, a type, a character string, a font, a font size, a position X1, a position Y1, a storage destination, and the like. The ID is uniquely assigned to each layout element. The type indicates the type of layout element such as a text box or a check box. The character string indicates the character string of the label when the layout element is a label. The font indicates the font type of characters input / output to / from the layout element. The font size indicates the size of characters input / output to / from the layout element. The positions X1 and Y1 indicate the XY coordinates of the layout element. For example, when the layout element is rectangular, two XY coordinates are used. The storage destination indicates a place where information such as input characters is stored.

  The operator can set these attribute values. The set property is evaluated for the attribute value.

  The layout elements can be set in various ways as shown in FIG. Reference numeral 8 shown in FIG. 4 indicates a drawing element. Reference numeral 14 indicates an output area. Reference numeral 15 indicates an input area.

  Among the layout elements shown in FIG. 4, a layout element 502 indicates a general layout element including content. The layout element 502 has an output area 14 and a drawing area 8 as shown in the layout element 503.

  The layout element 504 has the recording area inside. The layout element 505 has the recording area outside. The layout element 506 has the recording area on the inner right. The layout element 507 has a recording area on the left and right inner right and upper and lower outer sides.

  Next, the configuration of the form creation program will be described with reference to FIG. The form creation program is a program having a three-layer structure of a database layer 512, a logic layer 511, and an I / F layer 510.

  The I / F layer 510 that interactively designs a form with an operator includes a display unit 213, a pointing unit 214, and a key input unit 215. Specifically, the display means is a display, the pointing means is a mouse, and the key input means performs data input / output with a keyboard.

  The logic layer 511 includes an input area automatic adjustment unit 21, an offset amount calculation unit 12, a layout unit 211, an information table management unit 212, a creation program logic 210, a printing unit 219, and a barcode encoding unit 220. , Definition data storage means 216, form definition body storage means 218, and virtual form definition table A217.

  The input area automatic adjustment means 21 automatically adjusts the position of the input area described above using the offset amount calculation means 12.

  The layout unit 211 displays a GUI such as the GUI 7a (see FIG. 3) on the screen by the display unit 213. In addition, the layout unit 211 acquires mouse or keyboard input via the pointing unit 214 or the key input unit 215. Then, the layout unit 211 displays an attribute value input GUI related to the layout element and causes the operator to input the attribute of the layout element in order to cause the operator to set the property.

  The information table management unit 212 communicates the attribute value of the property set by the operator with the information table 22, specifically, the database program that manages the information table 22.

  The creation program logic 210 performs flow control in a series of form creation. This creation program logic 210 accesses the I / F layer via the layout means 211 and interactively designs a form with the operator. The creation program logic 210 holds the layout elements and their attribute values laid out in the virtual form definition table A 217 via the definition data storage means 216 in parallel with this work. Further, the creation program logic 210 transmits a print command to the printing apparatus 107 via the printing unit 219.

  The printing unit 219 superimposes a code image that can specify the location of the form definition body together with the paper image of the form to generate a print command to the printing apparatus 107.

  The barcode encoding means 220 receives an identifier for specifying the location of the form definition body 1 from the form definition body storage means 218, and creates a code image.

  The definition data storage unit 216 stores the definition data in the virtual form definition table A 217 according to an instruction from the creation program logic 210.

  The virtual form definition table A217 manages information that can be set by the operator displayed by the attribute input GUI and management information that does not allow the operator to be set in correspondence with the layout elements.

  The form definition body storage unit 216 creates the form definition body 1 in the database layer in accordance with an instruction from the creation program logic 210.

  Next, the database layer 512 will be described. The database layer 512 includes a form definition body 1 and an information table 22. Of these, the information table 22 has already been described, and a description thereof will be omitted.

  The form definition body 1 is a file in which information including the contents of the virtual form definition table and bibliographic items related to the form is described in a predetermined format. That is, it is a data file of an electronic form.

  When the interactive layout creation work with the operator is completed, the form definition body 1 is created by the creation program logic 210.

  The form definition body 1 will be described with reference to FIG. FIG. 6 shows a virtual form definition table 521 and a form definition body description example 522.

  The virtual form definition table 521 is a virtual form definition table corresponding to the unfilled form 20 (see FIG. 2), that is, the “name” and “entry part A” of the paper form. A form definition description example 522 is an example of an excerpt of the “entry part A” portion of the form definition in a paper form. That is, the data in the column of ID = 0002 in the virtual form definition table 521 is filed.

  As described above, the attribute value of each layout element of the form can be input / output as a file. This form definition example is the one in this embodiment, and a general form definition body is an attribute (hereinafter referred to as an output attribute) to be displayed (output) on a layout element for one layout element, or from a layout element. One of the attributes that accepts input (hereinafter referred to as input attribute) exists, and both exist in the present embodiment.

  In FIG. 6, attributes belonging to the attribute groups 523 and 526 are output attributes, and attributes belonging to the attribute groups 524 and 527 are input attributes. The input attribute stores an operation attribute value that is indispensable for performing handwriting processing, such as recognition knowledge.

  Next, group elements will be described with reference to FIG. FIG. 7 shows a drawing element 540 and a group element 541. The drawing element 540 is the check box described above. A group of several drawing elements 540 is called a group element 541 as shown in FIG. In this way, several drawing elements can be handled together.

  Next, the configuration of the form processing program will be described with reference to FIG. The form processing program is a program having a three-layer structure of a database layer 533, a logic layer 532, and an I / F layer 531.

  This program processes the read image according to the operation attribute (processing manner) of the form described in the form definition body from the form definition body corresponding to the scanned image. That is, it has a function of appropriately converting handwritten input data and storing it at a predetermined position. The form processing program corresponds to image information acquisition means, information detection means, conversion means, group information acquisition means, display means, and confirmation correction means.

  The I / F layer 531 has an image input unit 231. The image input unit 231 has a function of holding an acquired image read by the reading device 108 in a memory.

  The logic layer 532 includes an area extraction unit 232, an input value aggregation unit 44, a control selection unit 45, a processing program logic 230, a definition data acquisition unit 233, a form definition body acquisition unit 234, and an input / output value display unit. 26, barcode decoding means 236, input data conversion means 237, information table setting means 238, and virtual form definition table B235.

  The area cutout unit 232 acquires a partial image cut out from the acquired image. On the paper form, the input value aggregating unit 44 converts the item selected by the mark into a value that should be output from the form, and processes the grouped drawing elements described later.

  The control selection means 45 automatically selects a control having options in the input / output value display editing means 26. The input / output value display editing means 26 allows the operator to determine whether the entered data is correctly converted.

  The barcode decoding unit 236 receives the acquired image from the processing program logic 230, detects the code image, and decodes the code image. Since this code image stores an identifier that can refer to the form definition body when the form is created, the processing program logic 230 eventually acquires the reference information of the form definition body 1 from the code image.

  The input data conversion means 237 performs character recognition and determination of entry in a check box. The information table setting means 238 stores the data conversion result obtained from the information handwritten on the entry form at a predetermined position in the information table 22.

  The definition data acquisition unit 233 acquires each attribute of the form from the virtual form definition table B235. The form definition body obtaining unit 234 reads the form definition body from the form definition body 1 into the memory.

  The processing program logic 230 is a part that performs flow control in a series of form processing.

  First, the processing program logic 230 acquires a form image entered from the reading device 108 via the image input means 231. Next, the processing program logic 230 transmits the acquired image to the barcode decoding unit 236 to detect and decode the code image. Then, the processing program logic 230 loads the entity of the form definition body 1 into the memory via the form definition body acquisition means 234, and at the same time holds each attribute of the form in the virtual form definition table B235. Here, the form definition body acquisition means 233 has a function of taking out a file from the storage device based on the reference information of the form definition body 1.

  Next, the processing program logic 230 recognizes position information, input attributes, storage information, and the like for each layout element of the virtual form definition table 235. Among these, the input attribute indicates whether or not to use as input data at the time of form processing, and how to convert information when input. The storage information indicates information on where to store the information.

  Then, the processing program logic 230 acquires a partial image of the input data frame via the region cutout unit 232 based on the position attribute (image information acquisition stage). Further, the processing program logic 230 transmits this to the input data conversion means 237 and obtains a predetermined information conversion result (handwritten information detection stage, conversion stage).

  Finally, the processing program logic 230 stores the data conversion result at a predetermined position in the appropriate information table 22 via the information table setting means 238 based on the storage location information (group information acquisition stage).

  Thus, in this embodiment, the form processing program recognizes the characteristics of the group elements obtained from the form structure, and automatically determines the values in the group elements. In this embodiment, the input / output value display means 26 automatically selects a control having options, and at the same time, automatically determines the options.

  Next, the form structure will be described with reference to FIG. The form structure includes bibliographic data 601, a drawing element 8, and a group element 40. This form structure is included in the form definition body.

  Bibliographic data 601 indicates a title, the number of pages, a paper size, and the like. A drawing element 8 indicates an element to be drawn. The group element 40 includes operation data 41, active element data 42, group value data 43, a plurality of drawing elements 8, and a plurality of group elements 40.

  The operation data 41 is used in the form processing program and indicates how to process the values of the elements in the group. The group value data 43 is data for extracting group elements. The active element data 42 is an identifier that can be referred to a drawing element that exists directly under the group element and whose input permission data (see FIG. 20) is True, and an identifier that can refer to the group element that exists immediately below the group element. It is an enumeration. For example, if all drawing elements and group elements are managed with unique IDs in the form, the IDs are used as identifiers. If these are managed with unique names, the names are used. be able to.

  By having such a form structure, the form creation program can handle a plurality of elements existing in a group as one element. Therefore, for example, when moving or deleting a group element having a plurality of drawing elements as shown in FIG. 7, it is possible to assemble software to link the internal drawing elements by instructing the group elements to move or delete. Therefore, the operability for the operator who is creating the layout is improved.

  Next, a means by which the form processing program can determine the handling of drawing elements or group elements existing in the group elements will be described. This means can be realized by having the operation data described above in the group element. The operation data can be arbitrarily set according to the system.

  An example is shown in FIG. FIG. 10 shows a questionnaire form 610 and an entry form 611. The questionnaire form 610 includes a drawing element 614 surrounded by a solid line and group elements 550, 551, and 552 surrounded by a dotted line. These solid and dotted lines are not actually displayed. The entry form 611 is handwritten on the printed questionnaire form 610.

  In the entry form 611, options 553 and 554 and handwritten mixed options 555 for three questions are shown. In the case of a form having such options, there are generally “single selection” for selecting one of the options 553 and “multiple selection” for selecting a plurality of options such as the option 554. To do. Further, there may be a case where a single selected result and other values are added as in the handwritten mixed option 555.

  The operation data holds an identifier indicating the type or the like that can represent these variations as its value. In this sense, the identifier is preferably a method for aggregating input data of drawing elements existing in the group element. Aggregation is a function for converting the mark-selected items into values that should be output from the original form for the drawing elements grouped on the paper form.

  In this way, settings such as “none”, “single selection”, “multiple selection”, and “character string concatenation” can be made. Among these, an example of “single selection” will be described with reference to FIG.

  FIG. 11 shows an element configuration example 625, a drawing element process 623, and a group element process 624.

  In the element configuration example 625, four drawing elements A to D are set for one group element 622. These drawing elements A to C each have an input attribute 621 and an output attribute 620. The drawing element D has only the output attribute 620.

  In this group element 622, it is assumed that any one of "10's, 20's and 30's" is checked by the writer as the age. Therefore, the operation data of the group element is set by the operator as “single selection”.

  The drawing element processing 623 indicates processing for each drawing element. This drawing element processing 623 includes target data, input control, input value conversion method, input value (image), input value (primary conversion value), input value (final conversion value), Output value data.

  The input control is expressed by TRUE or FALSE. In the case of TRUE, input control is performed, and in the case of FALSE, input control is not performed. When the input value conversion method is marked as BOOL, a selection mark as shown in the input value (image) column is accepted, and when the selection mark is detected, True is set as the input value (primary conversion value). Is to do.

  For example, since the selection mark is detected in the drawing element A as shown in the input value (image), True is set as the input value (primary conversion value), and the output value as the input value (final conversion value). Data teenage is set.

  Next, the group element processing 624 will be described. The group element processing 624 includes an active element, operation data, a total result, and group value data. Since the processing 624 of the group element reflects the case of the group element 622, the active elements are the drawing element A, the drawing element B, and the drawing element C, the operation data is a single selection, and the aggregation result Is a drawing element A (TRUE), and the group value data is a teenager.

  These form structures actually generated by the form creation program are called by the form processing program on the processing side, and are grouped based on the defined data (form definition body and virtual form table that is the internal representation of the memory). The value of is calculated.

  Next, an example of “multiple selection” will be described with reference to FIG. In FIG. 12, the same reference numerals as those in FIG.

  A group element 630 shown in FIG. 12 selects a plurality of items from breakfast, lunch, and dinner. In the case of FIG. 12, since breakfast and dinner are selected, as shown in the drawing element processing 631, mark input is selected for the input values (images) of the drawing elements A and C corresponding to breakfast and dinner, respectively. It has been shown. Breakfast and dinner are shown as group value data in the group element processing 632.

  Such aggregation is realized by the input value aggregation means described with reference to FIG. A logic example of the input value aggregation means will be described with reference to FIG.

  First, it is assumed that the input value aggregation means is called recursively. That is, when another group element exists in the lower layer of the group element, a process described later is first performed on the lower layer group element, and thereafter, a process described later is performed on the upper layer group element. Hereinafter, processing performed for each layer will be described.

  In step S101, the input value aggregating unit calculates input values of all the drawing elements existing directly under the target group. Active element data is used to select a drawing element to be calculated.

  This active element data is data generated during form creation by the form creation program. If there is no active element data, the input value aggregating means searches for all drawing elements immediately below the group element and finds that the input permission data is True, and uses these and the group element immediately below the group element. .

  In the former and the latter, since there is no active element extraction calculation, the former has higher processing efficiency and the handwriting processing program operates at high speed. In the former method, it is indispensable to generate active element data in the process of creating a form. However, since form creation generally proceeds interactively with an operator through a GUI, there is a large margin in computer processing capacity. That is, the form creation program does not hinder interactive form creation with the operator while dynamically creating active element data.

  11 and FIG. 12 as examples of input value calculation for each drawing element, the form processing program acquires a partial image based on the input area of each drawing element and stores it in the input value (image).

  Next, the input value aggregating means refers to the first column of the input conversion manner and performs mark detection of the input value (image). As a result, the input value aggregating means stores True in the input value (primary conversion value) if there is a selection mark, and False otherwise. Further, the input value aggregating means uses the data in the second column of the input value conversion manner, and if the input value (primary conversion value) is True, the output value data of the element is used. ) In the input value (final conversion value).

  In step S102, the input value aggregating unit acquires group operation data. Based on the acquired operation data, operation determination is performed in step S103. Here, it is assumed that the operation data is one of “none”, “single selection”, “multiple selection”, and “character string concatenation”. If the operation is set to “none”, the input value aggregation means determines that aggregation is not necessary, and calculates aggregate value = “none” in step S104.

  When the operation is a single selection, the input value aggregation means determines that the value is set for one selected item, and first, in step S105, the drawing element selected as a mark, that is, the input value is selected. When the number of elements of (primary transformation value = True) is counted and the counted number of elements is not 1, in step S104, the aggregate value = “none” is set, and when the counted number of elements is 1, in step S106. Calculated as aggregated value = input value (final conversion value).

  When the operation is a plurality of selections, the input value aggregation means determines to set a set of values for the selected items. In step S107, the input value aggregating means concatenates the drawing elements selected as marks, that is, the input values (primary conversion values = True) of the input values (final conversion values) by separating them with commas, and the result is connected. The aggregate value.

  When operation = character string concatenation, the input value aggregation means determines to add the values of the respective elements as character strings. In step S108, the input value aggregating unit converts the input value (final conversion value) of each element and the calculated value of the immediately below group into a character string, concatenates them, and uses the result as an aggregate value.

  By performing such processing, the group value = “10's” in the example of FIG. 11, the group value = “breakfast and dinner” in the example of FIG. 12, and Q1 = “30s”, Q2 in the example of FIG. = "Golf, Other", Q3 = "Showa 40.12.13" is calculated.

  With this logic, the data that the operator actually wants can be obtained. Aggregation is called recursively, and group elements existing in the lower layer can be easily extracted from the group value data. Therefore, it is desirable that the aggregate value is held in the group value data. If the group value data does not exist, the algorithm for substituting efficient aggregate calculation by recursive call becomes complicated, and the performance deteriorates.

  Next, input / output value display means (see FIG. 8) for automating erroneous recognition correction in information conversion will be described. This input / output value display means is effective especially when there are a plurality of options.

  Specifically, it is realized by having control selection means (see FIG. 8) and dynamically arranging the control obtained by the control selection means on the GUI displayed by the input / output value display means. The input / output value display means is a module having a GUI, and has a function of displaying and correcting the information conversion result through interaction with the operator. Controlled by the processing program logic, a GUI is created and displayed from the output value data and input value data of the virtual definition table B and the group operation data.

  This GUI will be described with reference to FIG. FIG. 14 shows a form configuration example 570, a display example 577, and a control selection list automatic setting example 578.

  The form configuration example 570 includes the group elements 571, 572, and 573 described above. Assume that this form configuration is printed and handwritten information is that Q1 is a teenager, Q2 is tennis, and Q3 is December 23, 1965.

  Display example 577 is displayed as a display stage when this information is read. The display example 577 includes a pull-down menu 574, a check box group 575, and a date of birth column 576 including radio buttons and a text box. As a result, the operator can easily confirm and correct. For example, as shown in a control selection list automatic setting example 578, correct information can be input from the pull-down menu (confirmation correction stage).

  In this way, the control selection means creates a GUI as shown in the display example 577 and sets the contents of the information-converted item as its initial value. In order to automatically generate a GUI, it is necessary to arrange controls that can be input and displayed on the screen. However, if these are drawing elements, they are displayed on the first layer data of the conversion manner and its controls. It can be uniquely determined by input value data or output value data. General controls for configuring the GUI are mainly provided by the software development environment, but generally there are those shown in FIG. FIG. 15 shows the control name and the corresponding function description. A label control displays text and does not have an editing function. The text control displays text and has an editing function. The combo box control has a text option and can select one option from a list. The radio button control can select one of a plurality of options by pressing a button. The check box control is a control that can select a plurality of choices by having □ and a label that describes the choices for each item and changing the selection / non-selection by clicking the □ of the item.

  Next, a flowchart showing the logic of this process will be described with reference to FIG. In step S201, the input / output value display means calculates the input values of all the drawing elements existing directly under the target group. Next, the input / output value display means acquires group operation data in step S202. Based on the acquired motion data, motion determination is performed in step S203.

  If the operation = “none”, the control type is set to none in step S204. When the operation = “single selection”, the control type is set to a combo list in step S205. In the next step S206, the option is set as the final conversion value of all elements. In step S207, the initial value of the combo list is set as the aggregated value.

  If operation = “multiple selection”, the control type is a check box in step S208. In step S209, the check box display value is set as the final conversion value of each element. In step S210, the initial value of the check box is an aggregated value.

  If operation = “character string concatenation”, the control type is a text box in step S211. In step S212, the initial values of the check boxes are aggregated values.

  The process ends here. For these, it is only necessary to register the final conversion values defined by the conversion method of the drawing elements existing inside the group element as a list or as a label. In addition, a control having options usually has its initial value set prior to GUI display, but this can be easily done by selecting the primary conversion value = the final conversion value of True.

  As a result, the conventional labor of creating one module for confirmation and correction in information conversion for one form can be reduced by automatically generating a general purpose, which helps to reduce the development cost.

  The effects of the above-described embodiment will be described below.

  Base configuration that solves many man-hours of the form processing program related to the allocation of data corresponding to the selected item and the presentation and correction of information conversion results to the user, which occurs when the form has a configuration with options Can be provided. Furthermore, on the form creation program, it is possible to assemble software so that internal drawing elements can be linked by issuing instructions to move or delete group elements, which makes it easier for operators who create layouts. improves.

  In addition, when the form has a configuration with options, it was necessary to develop a module that assigns data according to the form corresponding to the selected item. However, in this embodiment, these can be generalized and automated. These development man-hours can be minimized and development costs can be reduced.

  Furthermore, in the past, it was necessary to create a module for checking and correcting misrecognition in information conversion for each form, but in this embodiment, the form has options in the form. Even so, modules can be automatically generated and displayed in a small display area. For example, a control such as a combo box is automatically arranged, so it is possible to provide a module for checking and correcting misrecognition with a simple and easy-to-understand layout. . For this reason, it is possible to provide a handwritten form processing system and a handwritten form processing method in which not only the development cost can be reduced, but also the operator's confirmation and correction work can proceed smoothly.

It is a figure which shows a handwritten form processing system structure. It is a figure which shows the example of handwritten information processing. It is a figure which shows layout attribute data input GUI. It is recording area explanatory drawing. It is a figure which shows a form creation program structure. It is a figure which shows a form definition body. It is a figure which shows a form processing program structure. It is a figure which shows a form structure. It is a figure which shows a drawing element. It is a figure which shows the example of each form. It is a figure which shows a process of drawing and a group element. It is a figure which shows a process of drawing and a group element. It is a flowchart which shows the logic example of an input value aggregation means. It is a figure which shows GUI which displays and corrects an information conversion result via a dialogue. It is a figure which shows the general control for comprising GUI. It is a flowchart which shows the logic with respect to a group element. It is a figure which shows the form preserve | saved at a form definition body. It is a figure which shows the example in the case of having a check box in a form. It is a figure which shows the example in the case of having a check box in a form. It is a figure which shows the attribute structure of a drawing element. It is a figure which shows conversion manner data. It is a figure which shows the function example of an input-output value display edit means.

Explanation of symbols

1 Form definition body 2 Paper form 4 Code image 7a, 7b GUI
8 drawing element 14 output area 15 input area 20 entry form 21 input area automatic adjustment means 22 information table 26 input / output value display means 31 handwritten form processing system 40 group element 41 operation data 42 active element data 43 group value data 44 input value aggregation Means 45 Control selection means 102 Form creation PC
103 Form processing PC
DESCRIPTION OF SYMBOLS 104 Form preparation apparatus 105 Form processing apparatus 106 Storage apparatus 107 Printing apparatus 108 Reading apparatus 109 Data server 200 Layout area 201 Tool box 202 Label tool 203 Text box tool 204 Check box tool 205 Bar code tool 210 Creation program logic 211 Layout means 212 Information Table management means 213 Display means 214 Pointing means 215 Key input means 216 Definition data storage means 217 Virtual form definition table A
218 Form definition storage means 219 Printing means 220 Barcode encoding means 230 Processing program logic 232 Area extraction means 233 Definition data acquisition means 234 Form definition body acquisition means 235 Virtual form definition table B
236 Barcode decoding means 237 Input data conversion means 238 Information table setting means 501 Properties 502, 503, 504, 505, 506, 507 Layout elements 510, 531 I / F layer 511, 532 Logic layer 512, 533 Database layer 521 Virtual form Definition table 522 Form definition body description example 523, 524, 526, 527 Attribute group 540, 614 Drawing element 541, 550, 551, 552, 571, 572, 573, 622, 630 Group element 553, 554 Option 555 Handwritten mixed option 570 Form configuration example 574 Pull-down menu 575 Check box group 576 Date of birth column 577 Display example 578 Control selection list automatic setting example 601 Bibliographic data 610 Questionnaire form 611 Form entry 20 processing 625 component configuration example of processing 624,632 group element of the output attribute 621 input attributes 623,631 drawing elements

Claims (10)

In a handwritten form processing system that acquires information from a handwritten form and processes it,
Form definition information used to obtain information from the form;
Based on the form definition information, image information acquisition means for acquiring image information of an area assumed to be handwritten on the form;
Handwritten information detecting means for detecting handwritten information from image information acquired in an area belonging to a group element consisting of a plurality of areas defined by the form definition information by the image information acquiring means;
Conversion means for converting the handwritten information detected by the handwritten information detection means into a predetermined value based on the form definition information;
Group information acquisition means for acquiring group information, which is information as the group element, based on the form definition information from the value converted by the conversion means. The handwritten form processing system according to claim 1, wherein the information acquisition unit uses a value acquired from one of the areas belonging to the group element as the group information. 2. The handwritten form processing system according to claim 1, wherein the information acquisition unit uses, as the group information, values acquired from a plurality of areas among the areas belonging to the group element. 2. The handwritten form processing system according to claim 1, wherein the information acquisition unit uses, as the group information, a concatenation of character strings that are values acquired from an area belonging to the group element. Display means for displaying the plurality of group information;
2. The handwritten form processing system according to claim 1, further comprising a confirmation / correction unit that allows an operator to confirm or correct the group information displayed on the display unit. In a handwritten form processing method for obtaining information from a handwritten form and performing processing based on the form definition information used for obtaining information from the form,
An image information acquisition step of acquiring image information of an area assumed to be handwritten in the form;
In the image information acquisition step, handwriting information detection step of detecting handwriting information from image information acquired in an area belonging to a group element consisting of a plurality of areas defined in the form definition information;
A conversion step of converting the handwritten information detected in the handwritten information detection step into a predetermined value;
A group information acquisition step of acquiring group information, which is information as the group element, from the value converted in the conversion step. 7. The handwritten form processing method according to claim 6, wherein, in the information acquisition step, a value acquired from one of the areas belonging to the group element is used as the group information. The handwritten form processing method according to claim 6, wherein, in the information acquisition step, values acquired from a plurality of areas belonging to the group element are used as the group information. The handwritten form processing method according to claim 6, wherein in the information acquisition step, the group information is obtained by concatenating a character string that is a value acquired from an area belonging to the group element. A display step of displaying the plurality of group information;
7. The handwritten form processing method according to claim 6, further comprising a confirmation / correction step in which an operator confirms or corrects the group information displayed in the display step.

Images