JP4853073B2 - Form classification processing system, electronic medium receipt creation method and delivery method - Google Patents

Description

  The present invention relates to a form classification processing system, and also relates to an electronic medium receipt creation method and a delivery method.

  2. Description of the Related Art A form classification processing system that performs form classification processing for monthly sales slips, billing slips, and the like in a business entity that provides products and services is known. For example, in Patent Document 1 below, a form such as a slip is accepted, characters written therein are read, sorted into a delivery destination pocket, and sorted and stored, and the billing amount described in the form is aggregated. A form classification processing system for executing processing is disclosed.

  By the way, in medical institutions such as hospitals and clinics, the above-described form classification processing system is applied to accept a receipt in which a billing amount (claimed point) etc. as a medical treatment fee issued monthly is described, The letters written there are read, and each receipt is sorted and stored in an insurer's pocket so that each receipt can be sent to a prescribed health insurance association (hereinafter referred to as an insurer). The billing amount (billing point) described in the receipt of each delivery destination is summed up and sent to the insurers in each region.

  FIG. 3 and FIG. 4 are a block diagram showing a conventional form classification processing system as described above and a schematic diagram showing a conventional classification receipt delivery method.

  As shown in FIG. 3, the conventional form classification processing system includes a form reading classification unit 1, a system control unit 2, a correction terminal 3, and a printer 4 mainly composed of an optical reader (Optical Character Reader). .

  The system control unit 2 controls the entire system, and first instructs the form reading classification unit 1 to read the receipt 40. Upon receipt of the instruction, the form reading and classifying unit 1 reads the description content of the receipt 40 as a form by the imaging mechanism, generates read data, and insurers A, B, C,. They are classified separately (see FIG. 4). In addition, the form reading classification unit 1 generates image data for correcting an item on the receipt in which unreading occurred during reading. Subsequently, the system control unit 2 holds the read data and image data of the receipt 40 read by the form reading classification unit 1 in a memory. When the read data and the image data are stored in the memory, the correction terminal 3 displays a correction screen based on the read data and the image data on the display unit, and an operator or the like visually confirms the correction screen as needed. Correct the read data. Subsequently, the system control unit 2 refers to the read data that has been confirmed to be corrected, and adds the billing sum data indicating the sum of the billing points described in each of the receipts classified by the delivery destinations A, B, C. Generated for each insurer, the printer 4 executes a printing process based on the generated combined charge data.

In this way, as shown in FIG. 4, the pre-classification receipt 40 is sorted into a plurality of insurers A, B, C... Receipt 41 and billing point summation data 42 are provided, insurer B is provided with receipt 43 and billing point summation data 44, and insurer C is provided with receipt 45 and billing point summation data 46. The receipt delivery service is executed.
Japanese Patent Application Laid-Open No. 8-27284

  However, the above-described form classification processing system has the following problems to be solved. That is, the sorted receipts 41, 43, 45 are sorted and delivered separately by insurers A, B, C,... Such as health insurance associations. , 43 and 45 have a problem of increasing management costs such as expenses for securing the storage space. In addition, it is necessary to examine the contents of the claim to determine whether the amount of the claim indicated as the claim point is appropriate. However, since such examination requires skill, if it is entrusted to a specialized examination body, the cost for transporting a large amount of receipts will be required. There was a problem that occurred. Further, even in the above-described examination organization, when examining a large number of brought-in receipts against a past receipt, there is a problem that it is very difficult to compare both receipts on paper.

  Therefore, in order to solve the above problem, from the received receipt, obtain the image data necessary for the examination, create an electronic medium receipt that records the obtained image data together with the read data and the billing point total data, It is conceivable to send the created electronic medium receipt to each insurer.

  However, as described above, only the control for acquiring image data necessary for examination from the received receipt, for example, when the received receipt has a paper break or heavy run, a part of the image is missing. There is a possibility that image data with poor visibility is acquired, and an electronic medium receipt in which the image data with poor visibility is recorded is created.

  Further, when the received receipt is meandering, there is a possibility that image data having a distorted image with poor visibility is acquired, and an electronic medium receipt in which the image data with poor visibility is recorded is created. In addition, since the density of characters, charts, and the like described in the received receipt is low, image data with poor color and shape and poor visibility is acquired, and the electronic medium on which the poor visibility image data is recorded There was a risk of creating a receipt.

  Further, since the density of the background color of the receptor is high, there is a possibility that image data including noise in the image having poor visibility is acquired, and an electronic medium reception in which the image data having poor visibility is recorded is created. . In addition, due to blurring of stamps or the like imprinted on the receipt, image data with poor visibility in which characters written on the receipt are crushed is acquired, and an electronic medium receipt in which the image data with poor visibility is recorded is created. There was a risk of being.

  In addition, when imaging a receipt in which a plurality of sheets are overlapped, image data with poor visibility in which an image of the lower receptor is reflected is acquired, and an electronic medium reception in which the image data with poor visibility is recorded is created. There was a fear.

  Further, when acquiring partial image data obtained by extracting a predetermined image area necessary for the examination of the receipt, since the characters and the like are described outside the image area, the described characters are difficult to identify. There is a possibility that bad image data is acquired and an electronic medium receipt in which the image data with poor visibility is recorded is created.

Thus, depending on the acquisition state of the image of the receipt, image data with poor visibility is acquired as described above. Further, in such image data with poor visibility, the accuracy of character recognition by the optical reading device is lowered, and there is a possibility that the classification may fail.

  In addition, conventionally, since image data with poor visibility including an area such as an unnecessary background other than the receipt is acquired, there is a possibility that an electronic medium reception in which the image data with poor visibility is recorded is created. . In addition, when the shape of the receipt is horizontally long due to the addition of a continuous sheet or the like, image data with poor visibility so that the entire screen cannot be displayed at the time of examination was acquired, and the image data with poor visibility was recorded. There was a risk that an electronic media receipt would be created.

The present invention adopts the following configuration in order to solve the above points.
<Configuration 1>
Read the contents of the received forms and classify them into pocket mechanisms for each delivery destination, and for each classified form, read the forms in a form classification processing system that holds the read contents data for each delivery destination. Based on the density change amount of each pixel constituting the obtained read image data including the background, the outline of the form is detected, the boundary coordinate data indicating the boundary between the background and the form, and the form angle coordinate indicating the corner of the form Creates partial image data necessary for reviewing the contents of the form based on the form outline detection unit that obtains each data and the area data indicating the area position of each item on the form registered in advance from the read image data a partial image creating section that, from the read image data, on the basis of the boundary coordinate data, and an entire image creation unit that creates a whole face image data, the partial image creating section Create partial image data obtained by extracting predetermined items necessary for examination of a form including an extended range indicated by previously registered extended value data from the read image data, and clarify the extended range for the extended range. The screen processing is performed accordingly.
<Configuration 2>
In the form classification processing system of configuration 1, the screen processing is translucent processing or dot pattern conversion processing.
<Configuration 3>
In the form classification processing system of configuration 1, the form is a receipt, the delivery destination is an insurer, the description content is the number of claim points described in the receipt, and further, the sum of the claim points is added to the claim point total data. It is characterized by holding by insurer.
<Configuration 4>
In the form classification processing system of Configuration 1, an abnormal shape pocket mechanism that accommodates abnormal forms, and a pocket control unit that discharges the form to the abnormal shape pocket mechanism when the shape determination unit determines that the form is abnormal. It is characterized by providing.
<Configuration 5>
In the form classification processing system having the configuration 1, when the shape determination unit determines that the form shape is abnormal, the form classification unit includes a numbering mechanism that executes print processing based on image determination result data indicating the shape abnormality for the form.
<Configuration 6>
In the form classification processing system of configuration 1, the full-surface image creation unit creates full-face image data of a form with the background removed, based on the acquired boundary coordinate data and form corner coordinate data.
<Configuration 7>
In the form classification processing system that reads the description contents of received forms and classifies them into pocket mechanisms for each delivery destination, and stores the read description data for each classified form for each delivery destination. Boundary coordinate data indicating the boundary between the background and the form, detecting the outer shape of the horizontally long form based on the density change amount of each pixel constituting the read image data including the background obtained by reading the horizontally long form made of paper, a boundary detecting unit for respectively acquiring a page split coordinate data indicating a boundary between form angle coordinate data and single form and continue the paper shows the corner of該帳votes, the lateral length form from the read image data, expansion of the single form and a connection sheet Single book extracted by expanding the width based on the width extension value data indicating the width of the width to be acquired, the boundary coordinate data to be acquired, the form corner coordinate data, and the page division coordinate data Performing a entire image creation unit, to the range of the width is expanded in the entire image data of a single form and a connection sheet is created, a screen processed to explicate the scope to create and continue the paper entirely image data, respectively A background editing unit is provided.
<Configuration 8>
In the form classification processing system of configuration 7, the screen processing is translucent processing or dot pattern conversion processing.
<Configuration 9>
In the form classification processing system having the configuration 7, when the entire image data of a single form and a follow-up sheet is created, the entire image data is synthesized to create one-file entire image data that can be referenced for each page. A page conversion unit is provided.

  In the form classification processing system of the present invention, as a specific example of the form, the medical institution is suitable for receiving a medical remuneration statement that is invoiced and issued to an insurer such as a health insurance association. In business, both medical institutions and customers can quickly process without delay.

  Further, based on the amount of change in density of the read image data, it is determined whether or not the receptor has a normal shape, and the read image data of the normal shape receptor is subjected to density correction to improve the visibility, and the corrected read Since the partial image data and the entire image data are extracted from the image data, an electronic medium receipt including the partial image data and the entire image data with excellent visibility can be created.

  In addition, when a shape abnormality or density abnormality in the read image data acquired from the receipt is detected, the pocket mechanism sorts and stores it in a pocket for each shape abnormality or density abnormality. It is possible to confirm whether the paper is broken or not, and it is possible to cope with whether the form classification process described above is executed again after the receipt is corrected, or returned to the receipt of the receipt.

  Also, partial image data obtained by extracting predetermined items necessary for examination of a receipt including an extended range indicated by previously registered extended value data is acquired from the read image data, and the extended range is recognized with respect to the extended range. Since image processing is performed as much as possible, characters and the like are not included in the items, and thus partial image data with poor visibility in which the written characters are difficult to identify is acquired. It is possible to prevent the creation of an electronic medium receipt in which partial image data with poor quality is recorded.

  Furthermore, since the entire image data is created by removing unnecessary outer areas such as the background from the area of the read image data divided based on each boundary coordinate data, the entire image data with poor visibility is acquired. Thus, it is possible to prevent creation of an electronic medium receipt in which the entire image data with poor visibility is recorded. Further, since printing can be performed based on the entire image data from which unnecessary outer areas such as the background are removed during printing, unnecessary toner can be prevented from being consumed by including unnecessary outer areas such as the background.

  In addition, when a receipt is attached to a single form and the shape of the receipt is horizontally long, the boundary between the receipt and each continuous sheet is detected from the read image data of the horizontally long form, and the detected boundary is Based on this, the horizontally read image data is divided and the entire image data of the receipt and each of the successive sheets is created, so that the visibility of the entire image at the time of examination can be improved.

  In addition, an unnecessary area such as the background of each full-surface image data is converted into white pixels, and image processing is performed to clarify the range with respect to the range whose width is expanded. The boundary between them can be easily identified, and at the time of printing, printing processing based on the entire image data from which unnecessary outer areas such as the background are removed can be executed. It is possible to prevent unnecessary toner from being consumed.

  Furthermore, since the entire image data created is synthesized to create one file of entire image data that can be referenced for each page, each image data can be managed in one file, which is convenient for data management. Can be improved.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

<Configuration of Example 1>
FIG. 1 is a block diagram showing a form classification processing system according to the present invention. FIG. 2 is a schematic diagram illustrating a method for delivering a categorized receipt according to the first embodiment.

  As shown in FIG. 1, the form classification processing system according to the first embodiment includes a form reading classification unit 1b, a system control unit 2b, a correction terminal 3, a printer 4, and an electronic medium creation unit 5.

  FIG. 5 is a block configuration diagram illustrating the form reading classification unit 1b according to the first embodiment. As shown in FIG. 5, the form reading and classifying unit 1b is a part that reads the contents of the pre-classified form (= recepto 40) one by one and classifies them separately by the insurers A, B, C,. Control unit 10, pocket control unit 11, pocket mechanism 12, numbering control unit 13, numbering mechanism 14, conveyance control unit 15, conveyance mechanism 16, imaging control unit 17, imaging mechanism 18, page number storage unit 19, recognition unit 20, A setting table 21 is included.

  The page number storage unit 19 is a memory composed of an HDD or the like, and holds page number data indicating the page number of the read form.

  The recognizing unit 20 is a unit that reads the received form and generates image data of the form. As shown in FIG. 6, the form outline detecting unit 401, the shape determining unit 402, the density distribution measuring unit 403, the density determining unit A unit 404, a density correction unit 405, a correction image storage unit 406, a reading unit 407, a partial image creation unit 408, and an entire image creation unit 409.

  The corrected image storage unit 406 is a memory such as an HDD and holds corrected read image data as will be described later.

  The system control unit 2 b is a unit that controls the entire system, and includes a business AP unit 50. As shown in FIG. 7, the business AP unit 50 includes a business setting table 51, a program unit 52, a reading result storage unit 53, a partial image storage unit 54, a full image storage unit 55, a total amount storage unit 56, and an electronic medium image. The storage unit 57, the printer channel 70, and the electronic medium channel 71 are configured.

  The job setting table 51 is a table that is referred to for controlling the reading of a form. As shown in FIG. 8, the pocket setting table 80, the reading setting table 81, the partial image setting table 82, and the full image setting table 83. , A service setting table 84 and an invalid form pocket setting table 85.

  The pocket setting table 80 is a table for identifying a pocket for an insurer as a form delivery destination, and pocket number data indicating a pocket number corresponds to each insurer identification data including an insurer name and an insurer number. It is a table attached. In this embodiment, as shown in FIG. 9, the pocket number data indicating the pocket number “1” for the insurer identification data (= insurer A, insurer number 001), the insurer identification data (= insurer B). Pocket number data indicating pocket number “2” for insurer number 002), pocket number data indicating pocket number “3” for insurer identification data (= insurer C, insurer number 003), insurance It is assumed that pocket number data indicating the pocket number “99” is associated with the person identification data (= other than the above).

  The reading setting table 81 is a table for identifying items on the form to be read for each insurer. For each insurer identification data including the insurer name and the insurer number, a billing point, a partial contribution, a symbol Of the items such as the number, the actual number of medical treatment days, the medical treatment start date, and the name of the wound, the reading necessity data indicating the items that need to be read are indicated by “◯”. In this embodiment, as shown in FIG. 10, read necessity data, insurer identification data, in which only the billing point needs to be read with respect to the insurer identification data (= insurer A, insurer number 001). (= Insurer B, Insurer No. 002) All items for reading necessary / unnecessary data and Insurer identification data (= Insurer C, Insurer No. 003) It is assumed that reading necessity data that makes an item an item that needs to be read is associated with each other.

  The partial image setting table 82 is a table for identifying items on a form to be acquired as a partial image for each insurer. For each insurer identification data including an insurer name and an insurer number, a billing point, Of the items such as the contribution, symbol number, actual medical treatment date, medical treatment start date, injury and illness name, etc., the partial acquisition necessity data indicating the items that need to be acquired are indicated by “◯”. In this embodiment, as shown in FIG. 11, partial acquisition necessity data, insurer identification having items that require acquisition of only billing points for insurer identification data (= insurer A, insurer number 001). For partial acquisition necessity data and insurer identification data (= insurer C, insurer number 003) that require acquisition of all items for data (= insurer B, insurer number 002) It is assumed that partial acquisition necessity data that sets all items as items that need to be acquired are associated with each other.

  The full image setting table 83 is a table for determining whether or not a full image of a form should be acquired for each insurer. The full image setting table 83 is provided for each insurer identification data including an insurer name and an insurer number. FIG. 12 is a table in which full acquisition necessity data indicating the insurer for which an image is to be acquired is indicated by “O”, and in this embodiment, as shown in FIG. 12, insurer identification data (= insurer B, insurance And the insurer identification data (= insurer C, insurer number 003) are associated with the full acquisition necessity data indicated by “◯” so as to acquire the full image of the form. And

  The service setting table 84 is a table for identifying a service method to be provided for each insurer, and should be provided among forms, list data, and electronic media for each insurer identification data indicating the insurer name. Service method data whose target is indicated by “◯” is associated. In this embodiment, as shown in FIG. 13, service method data for which forms and list data are provided for insurer identification data (= insurer A), and insurer identification data (= insurer B). Service method data that targets only electronic media and insurer identification data (= insurer C) are associated with service method data that targets all forms, list data, and electronic media, respectively. I will do it.

  The invalid form pocket setting table 85 is a table for identifying a pocket serving as a form delivery destination determined as an invalid form, and indicates a pocket number for each image determination result data indicating either shape abnormality or density abnormality. It is a table in which pocket number data is associated. In this embodiment, as shown in FIG. 14, the image number determination result data indicating “shape abnormality” is associated with the pocket number data indicating pocket number “91”, and the image determination result data indicating “density abnormality” is pocketed. It is assumed that pocket number data indicating the number “92” is associated.

  The correction terminal 3 refers to a correction screen displayed on a display unit (not shown) on the basis of the read data read by the form reading classification unit 1b and the corrected image data generated at the time of reading failure, as in the prior art. And a terminal for correcting the read data while visually confirming the read result and the corrected image.

  As will be described later, the printer 4 is a printing apparatus that executes a printing process based on print data respectively indicating the sum of billing points, the reading list, and the partial image list generated by the system control unit 2b.

  As will be described later, the electronic medium creation unit 5 writes the read list data, partial image list data, and full image list data generated by the system control unit 2b on an electronic medium such as a CD-ROM. This is the part that creates the receipt.

  Hereinafter, the interrelationship of the form reading and classifying unit 1b, the system control unit 2b, the correction terminal 3, the printer 4, and the electronic medium creating unit 5 of the form classification processing system of this embodiment will be described.

  For example, when a bundle of the receptors 40 is arranged on the transport mechanism 16 (see FIG. 15) and a reading instruction signal for instructing the system controller 2b to start reading the receptors 40 is output from an input unit (not shown), the system control is performed. First, the unit 2b transfers the pocket setting table 80, the reading setting table 81, the partial image setting table 82, the full image setting table 83, and the invalid form pocket setting table 85 constituting the business setting table 51 to the form reading classification part 1b. At the same time, the page number data in the page number storage unit 19 is initialized to “0”. Further, the contents of the work setting table 51 are transferred to the setting table 21.

  The control unit 10 of the form reading and classifying unit 1b is a unit that performs overall control of the form reading and classifying unit 1b. Upon receipt of each table such as the pocket setting table 80, the received table is temporarily stored in a memory (not shown). Then, a reading start signal for instructing the reading start of the receipt 40 is transferred to the transport control unit 15.

  The conveyance control unit 15 is a unit that controls driving of the conveyance mechanism 16. When the reading start signal is received, the conveyance control unit 15 drives the conveyance mechanism 16, and bundles the receptors 40 arranged on the conveyance mechanism 16 one by one. To transport.

  When the transported receipt passes through the imaging mechanism 18 (see FIG. 15), the control unit 10 outputs an imaging instruction signal for instructing imaging of the transported receptor to the imaging control unit 17.

  When receiving an imaging instruction signal, the imaging control unit 17 controls the imaging mechanism 18 and optically reads the transported receipt to obtain read image data. In this embodiment, the read image data includes at least a name part, a sex part, a date of birth part, a sickness name field, a summary field, a billing point, a partial contribution, a symbol number, an actual medical treatment date, a medical treatment start date, The image data indicates the contents of the injury / illness name. Further, it is assumed that the read image data of this embodiment is image data including a black background. Further, the read image data is composed of a plurality of pixels, and the density of each pixel is image data indicated by a numerical value from 0 to 255 (hereinafter, density value).

  When the read image data is acquired, the control unit 10 transfers the read image data to the recognition unit 20. At this time, the control unit 10 adds “1” to the page number data in the page number storage unit 19 and updates the page number data.

  The recognizing unit 20 is a unit that recognizes the item contents described in the receipt from the read image data and generates each image data and the like, and holds the form format specifying table and the item area correspondence table in a memory (not shown). ing.

  The form specification table is a table for specifying the receipt format, and the form format corresponding to the combination of ruled line data and ruled line data, which is defined by information such as the length, thickness, and size of the ruled line. The form form data shown is associated.

  The item area correspondence table is a table for specifying the area of each item included in the form. For each form format data, the area data indicating the area where the item content of each item included in the form is described by coordinates. Are associated. In the present embodiment, the area data indicating the area in which the item contents of the name field, name part, gender part, date of birth part, injury name part, and summary field are described by coordinates are associated with each other. .

  The form outer shape detection unit 401 of the recognition unit 20 is a part that detects the outer shape of the form, and stores, in a memory (not shown), reference change amount data indicating a concentration change amount that is a detection reference for the form outer shape and the following change amount calculation formula. is doing.

Change calculation formula = 4 * f (x, y) + f (x, y-1) -f (x, y + 1) -f (x-1, y) -f (x + 1, y) (1)

  Where f (x, y): density value of the selected pixel, f (x, y-1): density value of the upper adjacent pixel, f (x, y + 1): density value of the lower adjacent pixel, f (x-1) , Y): density value of the left adjacent pixel, f (x + 1, y): density value of the right adjacent pixel.

  This change amount calculation formula is for a pixel selected from a plurality of pixels constituting the read image data (hereinafter referred to as a selected pixel) and pixels adjacent to the selected pixel above, below, left, and right (hereinafter referred to as an adjacent pixel). This is an equation for calculating a density change amount between a selected pixel and an adjacent pixel based on a density value.

  Upon receiving the read image data, the form outer shape detection unit 401 selects one pixel as a density change amount calculation target from a plurality of pixels constituting the read image data, and acquires the density value of the selected pixel. Next, the density values of adjacent pixels adjacent to the selected pixel above, below, left, and right are respectively acquired. Then, the obtained density value of each pixel is applied to the above equation (1), and the density change amount between the selected pixel and the adjacent pixel is calculated. When the density change amount is calculated, the reference change amount data is acquired from the memory, and the calculated density change amount and the reference change amount are compared. If the density change amount is larger than the reference change amount, the selection is performed. The pixel is determined as a form boundary region, and the coordinate value of the selected pixel is acquired as boundary coordinate data indicating the boundary between the form and the background. On the other hand, if the density change amount is smaller than the reference change amount, the selected pixel is determined as a non-form boundary region, and the boundary coordinate data of the selected pixel is not acquired. The form outer shape detection unit 401 subsequently performs the same detection process for adjacent pixels, and tracks the boundary between the form and the background, that is, the form side.

  Note that the density change amount may be calculated for each region divided by 100 × 100 pixels or the like with the selected pixel as the center.

  Subsequently, when the form outer shape detection unit 401 performs the detection process on all the pixels of the read image data, the form angle is set to a pixel that intersects the extended line of each form side specified based on each acquired boundary coordinate data. It is determined as an area, and the coordinate value of the intersecting pixel is acquired as form angle coordinate data indicating the form angle coordinates.

  In this way, the form outer shape detection unit 401 performs the above-described detection processing, and holds each acquired boundary coordinate data and each form corner coordinate data in a memory (not shown) (see FIG. 39).

  The shape determination unit 402 is a unit that determines whether or not the form indicated by the read image data is a normal shape. The form outline detection unit 401 acquires four form angle coordinate data, and the angle of each form angle is a right angle. If the sides of the form specified based on the form angle coordinate data are parallel, it is determined as normal (see FIG. 40A), and four form angle coordinate data are acquired. If the angle is not a right angle, it is determined that there is an abnormality due to meandering (see FIG. 40D), and if five or more form angle coordinate data are acquired, it is determined that there is an abnormality due to paper folding or heavy running (FIG. Or refer to FIG. 40 (c)).

  Here, the angle of each form angle can be calculated by applying each specified form angle coordinate data to a trigonometric function such as arctangent. In addition, whether or not each form side is parallel is calculated by, for example, calculating the inclination of each form side specified based on the acquired form angular coordinate data, and calculating the gap between the form sides where the calculated inclinations match. It can be determined as parallel and it can be determined that the inconsistent form sides are not parallel.

  When the shape determination unit 402 determines that there is an abnormality, the control unit 10 detects the invalid read image data due to the shape abnormality (= meandering, paper folding or heavy running), and numbering instruction signal for instructing the numbering of the data indicating the shape abnormality Is output to the numbering control unit 13.

  The numbering control unit 13 is a unit that controls numbering on the form. Upon receiving a shape abnormality numbering instruction signal, the numbering control unit 13 generates image determination result data indicating the shape abnormality and sends the generated image determination result data to the numbering mechanism 14. Then, the numbering mechanism 14 performs a printing process of data indicating a shape abnormality indicated by the received image determination result data on the receipt 40 conveyed by the conveyance mechanism 16 (see FIG. 15).

  Note that the data indicating the shape abnormality to be printed may be either character data or numerical data (1: shape abnormality or the like) indicating the shape abnormality. In addition to the data indicating the shape abnormality, data indicating the content of the shape abnormality such as paper folding, meandering, and heavy running may be printed. Furthermore, the trace information indicating the processing process determined to be abnormal in shape may be printed.

  When the numbering mechanism 14 performs the printing process on the receptacle 40, the control unit 10 transfers the image determination result data to the pocket controller 11 so as to classify the abnormal receptor 40 into a pocket.

  Upon receiving the image determination result data, the pocket control unit 11 refers to the invalid form pocket setting table 85 based on the image determination result data, acquires the corresponding pocket number data, and stores the acquired pocket number data in the pocket mechanism. 12 for transfer.

  When the pocket mechanism 12 receives the pocket number data, the pocket mechanism 12 discharges and sorts the receptors transported to the pocket of the pocket number indicated by the pocket number data (see FIG. 15). In this embodiment, the abnormally shaped receptor 40 is discharged into the pocket 91.

  The control unit 10 repeatedly executes the reading classification process as described above until the receptacles 40 arranged in the transport mechanism 16 are eliminated, and ends the reading classification process when the receptors 40 are eliminated.

  On the other hand, when the shape determination unit 402 determines that the state is normal, the control unit 10 outputs a concentration measurement start instruction signal to the concentration distribution measurement unit 403.

  The density distribution measurement unit 403 is a unit that generates density histogram data in which the number of pixels of each density value (0 to 255) is counted. When receiving the density measurement start instruction signal, the density distribution measurement unit 403 receives the density of each pixel constituting the read image data. Based on the value, density histogram data indicating the number of pixels for each density value (0 to 255) is generated.

  The density determination unit 404 is a unit that determines whether or not the read form is a blank sheet or a form with excessive noise, and is for determining a form that can be used effectively (= a form other than a blank sheet or excessive noise). Reference density dispersion ratio data indicating a reference density dispersion ratio is held in a memory (not shown).

  When the density distribution measurement unit 403 generates density histogram data, the density determination unit 404 first calculates density dispersion ratio data indicating the density dispersion ratio based on the density histogram data, and the calculated density dispersion ratio data. Is larger than the reference density dispersion ratio data in the memory, it is determined that the form can be effectively used (= other than a blank sheet or excessive noise form). If the density dispersion ratio data is smaller than the reference density dispersion ratio data, a blank sheet or excessive noise is determined. Judged as a form. Here, the density dispersion ratio data can be calculated by applying the density histogram data to, for example, a dispersion ratio calculation formula used in a statistical test such as an F test, as in the prior art.

  When the density determination unit 404 determines that the form is blank or excessively noisy, the control unit 10 detects it as invalid read image data due to a density abnormality and outputs a numbering instruction signal for instructing the numbering of data indicating the density abnormality. Output to.

  When the numbering control unit 13 receives the numbering instruction signal for the density abnormality, the numbering control unit 13 generates image determination result data indicating the density abnormality, transfers the generated image determination result data to the numbering mechanism 14, and the numbering mechanism 14 includes the transport mechanism 16. The print processing of the data indicating the density abnormality indicated by the received image determination result data is performed on the receipt 40 conveyed by (see FIG. 15).

  The data indicating the density abnormality to be printed may be either character data or numerical data (2: density abnormality etc.) indicating the density abnormality. In addition to the data indicating the density abnormality, data indicating the content of the density abnormality such as a high density or a low density may be printed. Furthermore, the trace information indicating the processing process determined to be abnormal density may be printed.

  When the numbering mechanism 14 performs the printing process on the receptacle 40, the control unit 10 transfers the image determination result data to the pocket controller 11 so as to classify the abnormal receptor 40 into a pocket.

  Upon receiving the image determination result data, the pocket control unit 11 refers to the invalid form pocket setting table 85 based on the image determination result data, acquires the corresponding pocket number data, and stores the acquired pocket number data in the pocket mechanism. 12 for transfer.

  When the pocket mechanism 12 receives the pocket number data, the pocket mechanism 12 discharges and sorts the receptors transported to the pocket of the pocket number indicated by the pocket number data (see FIG. 15). In the present embodiment, the receptor 40 having an abnormal concentration is discharged into the pocket 92.

  The control unit 10 repeatedly executes the reading classification process as described above until the receptacles 40 arranged in the transport mechanism 16 are eliminated, and ends the reading classification process when the receptors 40 are eliminated.

  On the other hand, when the density determination unit 404 determines that the form can be effectively used (= other than a blank sheet or a form with excessive noise), the control unit 10 outputs a density correction start instruction signal to the density correction unit 405.

  The density correction unit 405 is a unit that performs outline enhancement correction of the read image data in order to improve the visibility of partial image data and the like, which will be described later. Upon receiving the density correction start instruction signal, the density correction unit 405 Perform enhancement correction processing. For example, a Laplacian method or the like similar to the prior art may be applied to the contour enhancement correction process. When the density emphasis correction process is completed, the density correction unit 405 stores the corrected read image data in the correction image storage unit 406.

  When the corrected read image data is stored in the corrected image storage unit 406, the reading unit 407 performs character recognition of the corrected read image data, and also performs memory recognition based on the insurer identification data that has been character-recognized. , The item content of the item whose corresponding read necessity data is indicated by “◯” is character-recognized, and the character-recognized insurer identification data and the read data indicating the item content are acquired.

  Subsequently, the partial image creation unit 408 of the recognition unit 20 is a unit that creates partial image data for speeding up the confirmation and examination of the read data. When the read unit 407 acquires the read data, the character recognition is performed. Based on the insurer identification data, the partial image setting table 82 of the memory is referred to, and the partial image data including the item content of the item whose corresponding partial acquisition necessity data is indicated by “O” is extracted from the read image data. At the same time, partial image data including the insurer identification data is created.

  Then, when the partial image creation unit 408 creates the partial image data, the full-screen image creation unit 409 of the recognition unit 20 refers to the full-screen image setting table 83 in the memory on the basis of the character-recognized insurer identification data. If the full acquisition necessity data indicates “○”, full image data including the contents of all items in the receipt is created. If the corresponding full acquisition necessity data does not indicate “○”, full image data Do not create.

  If the full image creation unit 409 does not create full image data, the control unit 10 refers to the page number storage unit 19 to obtain the page number data, and the read data obtained together with the page number data is created. When the partial image data is transferred to the system control unit 2b and the full image creation unit 409 creates the full image data, the read data, the created partial image data, and the full image data are transferred to the system control unit 2b together with the page number data. To do.

  Upon receiving the page number data, read data, partial image data, and full-screen image data, the business AP unit 50 of the system control unit 2b stores the read data in the read result storage unit 53 in association with the page number data. The partial image data is stored in the partial image storage unit 54 in association with the number data, and the full image data is stored in the full image storage unit 55 in association with the page number data.

  As described above, when the control unit 10 of the form reading classification unit 1b transfers the reading data to the system control unit 2b, the control unit 10 instructs the numbering of the page numbers to associate the reading data and the receipt 40 with each other. The numbering instruction signal to be transmitted is transferred to the numbering control unit 13.

  When the numbering control unit 13 receives the numbering instruction signal, the numbering control unit 13 refers to the page number data in the page number storage unit 19 and transfers the page number data to the numbering mechanism 14, and the numbering mechanism 14 displays the page indicated by the received page number data. The number is printed by the transport mechanism 16 on the transported receipt 40 (see FIG. 15).

  When the numbering mechanism 14 prints the receipt 40, the control unit 10 transfers a classification instruction signal to the pocket control unit 11 to classify the receipt 40 into a pocket for each insurer.

  Upon receiving the classification instruction signal, the pocket control unit 11 first transfers a request signal for insurer identification data to the system control unit 2b in order to identify the insurer to whom the receipt 40 is delivered.

  Upon receiving the insurer identification data request signal, the business AP unit 50 of the system control unit 2b receives the read data associated with the last page number data among the read data stored in the read result storage unit 53. The insurer identification data is acquired, and the acquired insurer identification data is transferred to the pocket control unit 11 of the form reading classification unit 1b.

  Upon receiving the insurer identification data, the pocket control unit 11 refers to the pocket setting table in the memory based on the insurer identification data, acquires corresponding pocket number data, and stores the acquired pocket number data in the pocket mechanism 12. Forward to.

  When the pocket mechanism 12 receives the pocket number data, the pocket mechanism 12 discharges and sorts the receptors transported to the pocket of the pocket number indicated by the pocket number data (see FIG. 15).

  The control unit 10 repeatedly executes the reading classification process as described above until the receptacles 40 arranged in the transport mechanism 16 are eliminated, and ends the reading classification process when the receptors 40 are eliminated.

  In the present embodiment, as described above, the form reading classification unit 1b completes the form classification reading process, and the reading result storage unit 53 is associated with the page number data as shown in FIG. The read data group is stored, the partial image storage unit 54 stores the partial image data group associated with the page number data, as shown in FIG. 17, and the full image storage unit 55 shows in FIG. As described above, the entire image data group associated with the page number data is stored.

  Subsequently, when the control unit 10 completes the reading classification process, the correction terminal 3 sends a request signal for the reading data and the partial image data in order to confirm whether each reading data accurately indicates the item content described in the receipt. The data is transferred to the system control unit 2b.

  When the business AP unit 50 of the system control unit 2b receives the request signal for the read data and the partial image data, the business AP unit 50 sets the read result storage unit 53 and the partial image storage unit 54 based on the page number data indicating the earliest page number. The read data and partial image data associated with the same page number data are respectively acquired with reference to each other, and the acquired read data and partial image data are transmitted to the correction terminal 3.

  When the correction terminal 3 receives the read data and the partial image data, the correction terminal 3 generates correction screen data capable of correcting each item content indicated by the read data based on the read data and the partial image data, and the generated correction screen data Based on the above, a correction screen is displayed on a display unit (not shown).

  The operator of the correction terminal 3 refers to the correction screen and corrects the item content and the like of the read data that has failed in character recognition via an input unit (not shown). When the operator confirms the correction of the read data and completes the correction confirmation of all the read data, the correction terminal 3 transmits a correction confirmation completion signal indicating the completion of the correction confirmation to the system control unit 2b.

  When receiving the correction confirmation completion signal, the business AP unit 50 of the system control unit 2b continues to refer to the read data group in the read result storage unit 53 to calculate the amount to be charged to each insurer. The billing points indicated by the read data having the insurer numbers are summed to generate billing point total data indicating the total billing points for each insurer, and the generated billing point summing data is stored in the total amount storage unit 56.

  That is, in the present embodiment, as shown in FIG. 19, a billing point total data group indicating the sum of billing points for each of the insurer A, the insurer B, and the insurer C is stored in the total amount storage unit 56.

  Subsequently, the business AP unit 50 refers to the service setting table 84 and detects a target to be provided to each insurer from service method data corresponding to the insurer identification data of each insurer.

  That is, in this embodiment, since the service method data corresponding to the insurer identification data of the insurer A and the insurer C indicates “provide list” by “◯”, the insurer A and the insurer C Since it is detected that the “list” should be provided and the service method data corresponding to the insurer identification data of the insurer B and the insurer C indicates “provided electronic medium” by “○”, the insurer B And the insurer C is detected that the “electronic medium” should be provided.

  The business AP unit 50 first prints a list to be provided to each of the insurer A and the insurer C that has been detected to provide the “list”.

  Here, as an example, a description will be given until the list for the insurer C is printed.

  The business AP unit 50 refers to the read result storage unit 53 based on the insurer identification data of the insurer C, extracts read data having the same insurer number, and lists the extracted read data. Create reading list data. Similarly, based on the insurer identification data of the insurer C, the partial image storage unit 54 is referred to extract the partial image data having the same insurer number, and the extracted partial image data is listed to Create image list data. Further, the business AP unit 50 refers to the total amount storage unit 56 based on the insurer identification data of the insurer C, and acquires billing point sum data having the same insurer number from the billing point sum data group. .

  Then, the business AP unit 50 generates print data indicating the billing point summation, the reading list, and the partial image list based on the acquired billing point summation data, the created reading list data and the partial image list data, and prints the printer channel. The generated print data is transmitted to the printer 4 via 70.

  When the printer 4 receives the print data, the printer 4 executes print processing based on the received print data, and creates a list to be provided to the insurer C as shown in FIG.

  Similarly to the above, the business AP unit 50 executes the printing process for the insurer A and creates a list to be provided to the insurer A.

  When the list printing process for all the insurers to whom the list is to be provided is completed, the business AP unit 50 continues to each of the insurer B and the insurer C that have been detected that the “electronic medium” should be provided. Create an electronic media receipt to be provided.

  Here, as an example, a description will be given until an electronic medium receipt for the insurer C is created.

  The business AP unit 50 refers to the read result storage unit 53 based on the insurer identification data of the insurer C, extracts read data having the same insurer number, and lists the extracted read data. Create reading list data. Similarly, based on the insurer identification data of the insurer C, the partial image storage unit 54 is referred to extract the partial image data having the same insurer number, and the extracted partial image data is listed to Create image list data. Further, the business AP unit 50 refers to the total amount storage unit 56 based on the insurer identification data of the insurer C, and acquires billing point sum data having the same insurer number from the billing point sum data group. At the same time, based on the same page number data included in the created read list data and partial image list data, the entire image storage unit 55 is referred to extract the entire image data having the same page number and the extracted entire surface List the image data to create full image list data.

  Subsequently, the business AP unit 50 stores the billing point summation, the reading list, the partial image list, and the whole image list based on the extracted billing point summing data, the created reading list data, the partial image list data, and the whole image list data, respectively. The electronic medium image data shown is generated, and the generated electronic medium image data is stored in the electronic medium image storage unit 57.

  Then, the business AP unit 50 transmits the electronic medium image data stored in the electronic medium image storage unit 57 to the electronic medium creation unit 5 via the electronic medium channel 71.

  Upon receiving the electronic medium image data, the electronic medium creation unit 5 writes the received electronic medium image data to an electronic medium such as a CD-ROM and provides it to the insurer C as shown in FIG. Create an electronic media receipt to be used. At this time, index data for referring to each list data included in the electronic medium image data is also written in the electronic medium.

  Similarly to the above, the business AP unit 50 writes the received electronic medium image data to an insurer B on an electronic medium such as a CD-ROM, and creates an electronic medium receipt to be provided.

  As described above, the bundles of the receptors 40 are classified according to the insurers A, B, and C, and are once sorted and stored in the pocket mechanism 12 (see FIG. 15), and are sorted according to the insurers A, B, C,. Since the receipt performs the list printing process and the creation of the electronic medium receipt according to the setting of the service setting table 84, for example, the insurer A can be provided with the receipt 41 and the billing point total data 42, The insurer B can be provided with only the electronic media receipt 47, and the insurer C can be provided with the receipt 45, the billing point total data 46, and the electronic media receipt 90 (see FIG. 2).

  Further, since the partial image data and the entire image data are extracted from the read image data subjected to the contour emphasis correction process, an electronic medium receipt including the partial image data and the entire image data with excellent visibility can be created. .

  Furthermore, when a shape abnormality or density abnormality of the read image data acquired from the receipt 40 is detected, the pocket mechanism 12 sorts and stores the pockets 91 and 92 according to the shape abnormality and density abnormality. It is possible to confirm whether the sorted receptacles 40 are in a state of paper breakage, whether to correct the state of the receipts 40 and then execute the above-described form classification process again, or return to the source of the receipts 40, etc. Can be dealt with.

<Operation of Example 1>
Next, the operation of the form classification processing system according to the first embodiment will be described.

  FIG. 22 is an operation flowchart of the form classification processing system according to the first embodiment.

  For example, when a bundle of the receptors 40 is arranged on the transport mechanism 16 (see FIG. 15) and a reading instruction signal for instructing the system controller 2b to start reading the receptors 40 is output from an input unit (not shown), the system control is performed. First, the unit 2b transfers the pocket setting table 80, the reading setting table 81, the partial image setting table 82, the full image setting table 83, and the invalid form pocket setting table 85 constituting the business setting table 51 to the form reading classification part 1b. At the same time, the page number data in the page number storage unit 19 is initialized to “0” (S1).

  Upon receiving the above tables such as the pocket setting table 80, the control unit 10 of the form reading and classifying unit 1b subsequently executes the classification reading process (S2).

  FIG. 23 is an operation flowchart of the form reading classification unit 1b.

  The control unit 10 of the form reading classification unit 1b temporarily stores each received table in a memory (not shown) and transfers a reading start signal instructing the reading start of the receipt 40 to the conveyance control unit 15 (S101).

  Upon receiving the reading start signal, the conveyance control unit 15 drives the conveyance mechanism 16 to convey the bundle of receptors 40 arranged on the conveyance mechanism 16 one by one into the apparatus (S102).

  When the transported receipt passes through the imaging mechanism 18 (see FIG. 15), the control unit 10 outputs an imaging instruction signal for instructing imaging of the transported receptor to the imaging control unit 17.

  Upon receiving the imaging instruction signal, the imaging control unit 17 controls the imaging mechanism 18 to optically read the transported receipt and obtain read image data (S103). In this embodiment, the read image data includes at least a name part, a sex part, a date of birth part, a sickness name field, a summary field, a billing point, a partial contribution, a symbol number, an actual medical treatment date, a medical treatment start date, The image data indicates the contents of the injury / illness name. Further, it is assumed that the read image data of this embodiment is image data including a black background. Further, the read image data is composed of a plurality of pixels, and the density of each pixel is image data indicated by a numerical value from 0 to 255 (hereinafter, density value).

  When the read image data is acquired, the control unit 10 transfers the read image data to the recognition unit 20. At this time, the control unit 10 adds “1” to the page number data in the page number storage unit 19 and updates the page number data.

  Upon receiving the read image data, the form outline detection unit 401 of the recognition unit 20 selects one pixel as a density change amount calculation target from a plurality of pixels constituting the read image data, and acquires the density value of the selected pixel. Next, the density values of adjacent pixels adjacent to the selected pixel above, below, left, and right are respectively acquired. Then, the obtained density value of each pixel is applied to the above equation (1), and the density change amount between the selected pixel and the adjacent pixel is calculated. When the density change amount is calculated, the reference change amount data is acquired from the memory, and the calculated density change amount and the reference change amount are compared. If the density change amount is larger than the reference change amount, the selection is performed. The pixel is determined as a form boundary region, and the coordinate value of the selected pixel is acquired as boundary coordinate data indicating the boundary between the form and the background. On the other hand, if the density change amount is smaller than the reference change amount, the selected pixel is determined as a non-form boundary region, and the boundary coordinate data of the selected pixel is not acquired. The form outer shape detection unit 401 continues to perform the same detection process on the adjacent pixels to track the form side (S104).

  Subsequently, when the form outer shape detection unit 401 performs the detection process on all the pixels of the read image data, the form angle is set to a pixel that intersects the extended line of each form side specified based on each acquired boundary coordinate data. It is determined as an area, and the coordinate value of the intersecting pixel is acquired as form angle coordinate data indicating the form angle coordinates.

  As described above, the form outer shape detection unit 401 executes the above-described detection process, and holds each acquired boundary coordinate data and each form corner coordinate data in a memory (not shown).

  In the shape determination unit 402, four form corner coordinate data are acquired by the form outer shape detection unit 401, the angle of each form angle is a right angle, and each form side specified based on the form angle coordinate data is parallel. If the angle of each form angle is not a right angle, it is determined that there is an abnormality due to meandering, and if five or more form angle coordinate data are obtained, It is determined that there is an abnormality due to bending or heavy running (S105).

  Here, the angle of each form angle can be calculated by applying each specified form angle coordinate data to a trigonometric function such as arctangent. In addition, whether or not each form side is parallel is calculated by, for example, calculating the inclination of each form side specified based on the acquired form angular coordinate data, and calculating the gap between the form sides where the calculated inclinations match. It can be determined as parallel and it can be determined that the inconsistent form sides are not parallel.

  When the shape determination unit 402 determines that there is an abnormality, the control unit 10 detects the invalid read image data due to the shape abnormality (= meandering, paper folding or heavy running), and numbering instruction signal for instructing the numbering of the data indicating the shape abnormality Is output to the numbering control unit 13.

  Upon receiving the shape abnormality numbering instruction signal, the numbering control unit 13 generates image determination result data indicating the shape abnormality, transfers the generated image determination result data to the numbering mechanism 14, and the numbering mechanism 14 includes the transport mechanism 16. In step S106, the receipt 40 that has been conveyed by the above-described process is printed with data indicating the shape abnormality indicated by the received image determination result data.

  When the numbering mechanism 14 performs the printing process on the receptacle 40, the control unit 10 transfers the image determination result data to the pocket controller 11 so as to classify the abnormal receptor 40 into a pocket.

  Upon receiving the image determination result data, the pocket control unit 11 refers to the invalid form pocket setting table 85 based on the image determination result data, acquires the corresponding pocket number data, and stores the acquired pocket number data in the pocket mechanism. 12 for transfer.

  When the pocket mechanism 12 receives the pocket number data, the pocket mechanism 12 discharges and sorts the receptors conveyed to the pocket of the pocket number indicated by the pocket number data (S107). In this embodiment, the abnormally shaped receptor 40 is discharged into the pocket 91.

  The control unit 10 repeatedly executes the reading classification process as described above until the receptacle 40 arranged in the transport mechanism 16 is eliminated, and ends the reading classification process when the receptor 40 is eliminated (S117).

  On the other hand, when the shape determination unit 402 determines that the state is normal, the control unit 10 outputs a concentration measurement start instruction signal to the concentration distribution measurement unit 403.

  Upon receiving the density measurement start instruction signal, the density distribution measuring unit 403 generates density histogram data indicating the number of pixels for each density value (0 to 255) based on the density value of each pixel constituting the read image data. (S108).

  When the density distribution measurement unit 403 generates density histogram data, the density determination unit 404 first calculates density dispersion ratio data indicating the density dispersion ratio based on the density histogram data, and the calculated density dispersion ratio data. Is larger than the reference density dispersion ratio data in the memory, it is determined that the form can be effectively used (= other than a blank sheet or excessive noise form). If the density dispersion ratio data is smaller than the reference density dispersion ratio data, a blank sheet or excessive noise is determined. (S109). Here, the density dispersion ratio data can be calculated by applying the density histogram data to, for example, a dispersion ratio calculation formula used in a statistical test such as an F test, as in the prior art.

  When the density determination unit 404 determines that the form is blank or excessively noisy, the control unit 10 detects it as invalid read image data due to a density abnormality and outputs a numbering instruction signal for instructing the numbering of data indicating the density abnormality. Output to.

  When the numbering control unit 13 receives the numbering instruction signal for the density abnormality, the numbering control unit 13 generates image determination result data indicating the density abnormality, transfers the generated image determination result data to the numbering mechanism 14, and the numbering mechanism 14 includes the transport mechanism 16. In step S <b> 106, the receipt 40 that has been transported by the above-described process is printed with data indicating density abnormality indicated by the received image determination result data.

  The data indicating the density abnormality to be printed may be either character data or numerical data (2: density abnormality etc.) indicating the density abnormality. In addition to the data indicating the density abnormality, data indicating the content of the density abnormality such as a high density or a low density may be printed. Furthermore, the trace information indicating the processing process determined to be abnormal density may be printed.

  When the numbering mechanism 14 performs the printing process on the receptacle 40, the control unit 10 transfers the image determination result data to the pocket controller 11 so as to classify the abnormal receptor 40 into a pocket.

  Upon receiving the image determination result data, the pocket control unit 11 refers to the invalid form pocket setting table 85 based on the image determination result data, acquires the corresponding pocket number data, and stores the acquired pocket number data in the pocket mechanism. 12 for transfer.

  When the pocket mechanism 12 receives the pocket number data, the pocket mechanism 12 discharges and sorts the receptors conveyed to the pocket of the pocket number indicated by the pocket number data (S107). In the present embodiment, the receptor 40 having an abnormal concentration is discharged into the pocket 92.

  The control unit 10 repeatedly executes the reading classification process as described above until the receptacle 40 arranged in the transport mechanism 16 is eliminated, and ends the reading classification process when the receptor 40 is eliminated (S117).

  On the other hand, when the density determination unit 404 determines that the form can be effectively used (= other than a blank sheet or a form with excessive noise), the control unit 10 outputs a density correction start instruction signal to the density correction unit 405.

  When the density correction unit 405 receives the density correction start instruction signal, the density correction unit 405 executes a contour enhancement correction process on the read image data (S110). For example, a Laplacian method or the like similar to the prior art may be applied to the contour enhancement correction process. When the density emphasis correction process is completed, the density correction unit 405 stores the corrected read image data in the correction image storage unit 406.

  When the corrected read image data is stored in the corrected image storage unit 406, the reading unit 407 performs character recognition of the corrected read image data, and also performs memory recognition based on the insurer identification data that has been character-recognized. The reading setting table is referred to, the item content of the item whose corresponding reading necessity data is indicated by “◯” is character-recognized, and the character-recognized insurer identification data and the reading data indicating the item content are obtained ( S111).

  Subsequently, when the reading unit 407 acquires the read data, the partial image creation unit 408 of the recognition unit 20 refers to the partial image setting table 82 in the memory based on the character recognition data that has been recognized, and acquires the corresponding partial acquisition. Partial image data including the item content of the item whose necessity data is indicated by “◯” is extracted from the read image data, and partial image data including the insurer identification data is created (S112).

  Then, when the partial image creation unit 408 creates the partial image data, the full-screen image creation unit 409 of the recognition unit 20 refers to the full-screen image setting table 83 in the memory on the basis of the character-recognized insurer identification data. If the full acquisition necessity data to be displayed indicates “◯”, full image data including the contents of all items of the receipt is created (S113 & S114), and if the corresponding full acquisition necessity data does not indicate “O”, Do not create full image data.

  If the full image creation unit 409 does not create full image data, the control unit 10 refers to the page number storage unit 19 to obtain the page number data, and the read data obtained together with the page number data is created. When the partial image data is transferred to the system control unit 2b and the full image creation unit 409 creates the full image data, the read data, the created partial image data, and the full image data are transferred to the system control unit 2b together with the page number data. To do.

  Upon receiving the page number data, read data, partial image data, and full-screen image data, the business AP unit 50 of the system control unit 2b stores the read data in the read result storage unit 53 in association with the page number data. The partial image data is stored in the partial image storage unit 54 in association with the number data, and the full image data is stored in the full image storage unit 55 in association with the page number data.

  As described above, when the control unit 10 of the form reading classification unit 1b transfers the reading data to the system control unit 2b, the control unit 10 instructs the numbering of the page numbers to associate the reading data and the receipt 40 with each other. The numbering instruction signal to be transmitted is transferred to the numbering control unit 13.

  When the numbering control unit 13 receives the numbering instruction signal, the numbering control unit 13 refers to the page number data in the page number storage unit 19 and transfers the page number data to the numbering mechanism 14, and the numbering mechanism 14 displays the page indicated by the received page number data. The number is printed by the transport mechanism 16 on the transported receipt 40 (S115).

  When the numbering mechanism 14 prints the receipt 40, the control unit 10 transfers a classification instruction signal to the pocket control unit 11 to classify the receipt 40 into a pocket for each insurer.

  Upon receiving the classification instruction signal, the pocket control unit 11 first transfers a request signal for insurer identification data to the system control unit 2b in order to identify the insurer to whom the receipt 40 is delivered.

  Upon receiving the insurer identification data request signal, the business AP unit 50 of the system control unit 2b receives the read data associated with the last page number data among the read data stored in the read result storage unit 53. The insurer identification data is acquired, and the acquired insurer identification data is transferred to the pocket control unit 11 of the form reading classification unit 1b.

  Upon receiving the insurer identification data, the pocket control unit 11 refers to the pocket setting table in the memory based on the insurer identification data, acquires corresponding pocket number data, and stores the acquired pocket number data in the pocket mechanism 12. Forward to.

  When the pocket mechanism 12 receives the pocket number data, the pocket mechanism 12 discharges and sorts the receptors conveyed to the pocket of the pocket number indicated by the pocket number data (S116).

  The control unit 10 repeatedly executes the reading classification process as described above until the receptacle 40 arranged in the transport mechanism 16 is eliminated, and ends the reading classification process when the receptor 40 is eliminated (S117).

  In the present embodiment, as described above, the form reading classification unit 1b completes the form classification reading process, and the reading result storage unit 53 is associated with the page number data as shown in FIG. The read data group is stored, the partial image storage unit 54 stores the partial image data group associated with the page number data, as shown in FIG. 17, and the full image storage unit 55 shows in FIG. As described above, the entire image data group associated with the page number data is stored.

  Subsequently, when the control unit 10 completes the reading classification process, the correction terminal 3 sends a request signal for the reading data and the partial image data in order to confirm whether each reading data accurately indicates the item content described in the receipt. The data is transferred to the system control unit 2b.

  When the business AP unit 50 of the system control unit 2b receives the request signal for the read data and the partial image data, the business AP unit 50 sets the read result storage unit 53 and the partial image storage unit 54 based on the page number data indicating the earliest page number. The read data and partial image data associated with the same page number data are respectively acquired with reference to each other, and the acquired read data and partial image data are transmitted to the correction terminal 3.

  When the correction terminal 3 receives the read data and the partial image data, the correction terminal 3 generates correction screen data capable of correcting each item content indicated by the read data based on the read data and the partial image data, and the generated correction screen data Based on the above, a correction screen is displayed on a display unit (not shown).

  The operator of the correction terminal 3 refers to the correction screen and corrects the item content of the read data that has failed in character recognition via an input unit (not shown) (S3). The operator confirms the correction of the read data described above, and when the correction confirmation of all the read data is completed, the correction terminal 3 transmits a correction confirmation completion signal indicating the completion of the correction confirmation to the system control unit 2b.

  When receiving the correction confirmation completion signal, the business AP unit 50 of the system control unit 2b continues to refer to the read data group in the read result storage unit 53 to calculate the amount to be charged to each insurer. Billing points indicated by the read data having the insurer numbers are summed to generate billing point total data indicating the total billing points for each insurer, and the generated billing point total data is stored in the total amount storage unit 56 ( S4).

  That is, in the present embodiment, as shown in FIG. 19, a billing point total data group indicating the sum of billing points for each of the insurer A, the insurer B, and the insurer C is stored in the total amount storage unit 56.

  Subsequently, the business AP unit 50 refers to the service setting table 84 and detects a target to be provided to each insurer from service method data corresponding to the insurer identification data of each insurer.

  That is, in this embodiment, since the service method data corresponding to the insurer identification data of the insurer A and the insurer C indicates “provide list” by “◯”, the insurer A and the insurer C Since it is detected that the “list” should be provided and the service method data corresponding to the insurer identification data of the insurer B and the insurer C indicates “provided electronic medium” by “○”, the insurer B And the insurer C is detected that the “electronic medium” should be provided.

  Then, the business AP unit 50 first prints a list to be provided to each of the insurer A and the insurer C that has been detected to provide the “list” (S5).

  Here, as an example, a description will be given until the list for the insurer C is printed.

  The business AP unit 50 refers to the read result storage unit 53 based on the insurer identification data of the insurer C, extracts read data having the same insurer number, and lists the extracted read data. Create reading list data. Similarly, based on the insurer identification data of the insurer C, the partial image storage unit 54 is referred to extract the partial image data having the same insurer number, and the extracted partial image data is listed to Create image list data. Further, the business AP unit 50 refers to the total amount storage unit 56 based on the insurer identification data of the insurer C, and acquires billing point sum data having the same insurer number from the billing point sum data group. .

  Then, the business AP unit 50 generates print data indicating the billing point summation, the reading list, and the partial image list based on the acquired billing point summation data, the created reading list data and the partial image list data, and prints the printer channel. The generated print data is transmitted to the printer 4 via 70.

  When the printer 4 receives the print data, the printer 4 executes print processing based on the received print data, and creates a list to be provided to the insurer C as shown in FIG.

  Similarly to the above, the business AP unit 50 executes the printing process for the insurer A and creates a list to be provided to the insurer A.

  When the list printing process for all the insurers to whom the list is to be provided is completed, the business AP unit 50 continues to each of the insurer B and the insurer C that have been detected that the “electronic medium” should be provided. An electronic medium receipt to be provided is created (S6).

  Here, as an example, a description will be given until an electronic medium receipt for the insurer C is created.

  The business AP unit 50 refers to the read result storage unit 53 based on the insurer identification data of the insurer C, extracts read data having the same insurer number, and lists the extracted read data. Create reading list data. Similarly, based on the insurer identification data of the insurer C, the partial image storage unit 54 is referred to extract the partial image data having the same insurer number, and the extracted partial image data is listed to Create image list data. Further, the business AP unit 50 refers to the total amount storage unit 56 based on the insurer identification data of the insurer C, and acquires billing point sum data having the same insurer number from the billing point sum data group. At the same time, based on the same page number data included in the created read list data and partial image list data, the entire image storage unit 55 is referred to extract the entire image data having the same page number and the extracted entire surface List the image data to create full image list data.

  Subsequently, the business AP unit 50 stores the billing point summation, the reading list, the partial image list, and the whole image list based on the extracted billing point summing data, the created reading list data, the partial image list data, and the whole image list data, respectively. The electronic medium image data shown is generated, and the generated electronic medium image data is stored in the electronic medium image storage unit 57.

  Then, the business AP unit 50 transmits the electronic medium image data stored in the electronic medium image storage unit 57 to the electronic medium creation unit 5 via the electronic medium channel 71.

  Upon receiving the electronic medium image data, the electronic medium creation unit 5 writes the received electronic medium image data to an electronic medium such as a CD-ROM and provides it to the insurer C as shown in FIG. Create an electronic media receipt to be used. At this time, index data for referring to each list data included in the electronic medium image data is also written in the electronic medium.

  Similarly to the above, the business AP unit 50 writes the received electronic medium image data to an insurer B on an electronic medium such as a CD-ROM, and creates an electronic medium receipt to be provided.

<Effect of Example 1>
In the form classification processing system according to the first embodiment, a paper form is provided to an insurer who needs a paper form as in the conventional case, and the insurer who does not want a paper form due to restrictions on storage locations or the like. An electronic media receipt can be provided. Insurers who want both paper forms and electronic media receipts will be able to provide both. As a specific example of the form, it is suitable for the receipt of medical remuneration statements that are issued and issued by medical institutions to insurers such as health insurance associations, and both medical institutions and customers in receipt processing operations that occur in large quantities every month Can be processed quickly without delay.

  Further, based on the density change amount of the read image data, it is determined whether or not the receptor has a normal shape, and the read image data of the normal shape receptor is corrected for contour enhancement, and the partial image data is corrected from the corrected read image data. Since the entire surface image data is extracted, it is possible to create a partial image data with excellent visibility and an electronic medium receipt including the entire surface image data.

  Furthermore, when a shape abnormality or density abnormality of the read image data acquired from the receipt 40 is detected, the pocket mechanism 12 sorts and stores the pockets 91 and 92 according to the shape abnormality and density abnormality. It is possible to confirm whether the sorted receptacles 40 are in a state of paper breakage, whether to correct the state of the receipts 40 and then execute the above-described form classification process again, or return to the source of the receipts 40, etc. Can be dealt with.

<Configuration of Example 2>
FIG. 24 is a block diagram illustrating a form classification processing system according to the second embodiment.

  As shown in FIG. 24, the form classification processing system according to the second embodiment includes a form reading / classifying unit 1c, a system control unit 2b, a correction terminal 3, a printer 4, and an electronic medium creation unit 5. The correction terminal 3 and the printer 4 The electronic medium creation unit 5 is the same as that in the first embodiment.

  FIG. 25 is a block diagram showing the form reading classification unit 1c.

  As shown in FIG. 25, the form reading / classifying unit 1c includes a control unit 10, a pocket control unit 11, a pocket mechanism 12, a numbering control unit 13, a numbering mechanism 14, a conveyance control unit 15, a conveyance mechanism 16, an imaging control unit 17, It has an imaging mechanism 18, a page number storage unit 19, and a recognition unit 20a, and includes a control unit 10, a pocket control unit 11, a pocket mechanism 12, a numbering control unit 13, a numbering mechanism 14, a conveyance control unit 15, a conveyance mechanism 16, The imaging control unit 17, the imaging mechanism 18, and the page number storage unit 19 are the same as in the first embodiment.

  FIG. 26 is a block configuration diagram showing a recognition unit 20a of a form reading / classifying unit 1c different from the first embodiment.

  As shown in FIG. 26, the recognition unit 20a includes a form outline detection unit 401, a shape determination unit 402, a density distribution measurement unit 403, a density determination unit 404, a density correction unit 405, a corrected image storage unit 406, a reading unit 407, a part. The image forming unit 408b and the entire surface image generating unit 409b are composed of a form outer shape detecting unit 401, a shape determining unit 402, a density distribution measuring unit 403, a density determining unit 404, a density correcting unit 405, a corrected image storing unit 406, and a reading unit 407. These are the same as in Example 1.

  The partial image creation unit 408b according to the second embodiment is a unit that creates partial image data for speeding up the confirmation of the read data, the examination, and the like. As illustrated in FIG. A section 511 and an extended area editing section 512.

  The partial region conversion unit 510 is a unit that expands the region of each item to be extracted as a partial image in order to improve the visibility of characters or the like partially written outside the item, and receives a receipt in a memory (not shown). It holds area data indicating the item area of each item constituting by coordinates and extension value data indicating the length of each item area to be expanded, such as 2 mm.

  As in the first embodiment, when the reading unit 407 acquires the read data, the partial area conversion unit 510 refers to the partial image setting table 82 in the memory based on the character identification data that has been recognized, and responds accordingly. Extended area data that specifies the item area to be expanded whose partial acquisition necessity data is indicated by “O”, and continues to refer to the memory and expands the area data of each identified item area based on the extension value data Is generated.

  The image acquisition unit 511 is a unit that extracts partial image data from the read image data. When the partial region conversion unit 510 generates extended region data, the partial image data including the extended region is generated based on the generated extended region data. Is extracted from the read image data, and partial image data including insurer identification data is created.

  The extension area editing unit 512 performs image processing on the extension range of each item constituting the partial image data in order to clarify the extension area. When the image acquisition unit 511 creates the partial image data, the extension value data in the memory Based on the above, the extended range of each item constituting the created partial image data is specified, and screen processing is performed on the specified extended range. For this screen process, for example, a translucent process, a dot pattern conversion process, and the like similar to the prior art can be applied (see FIG. 41).

  The full-surface image creation unit 409b according to the second embodiment is a unit that creates full-surface image data excluding unnecessary areas such as the background from the read image data. As illustrated in FIG. 28, the acquisition range establishment unit 522 and the image acquisition unit 523.

  The acquisition range establishing unit 522 is a unit that determines a range of an unnecessary area such as a background included in the read image data. When the extended area editing unit 512 completes the screen processing, the form outline detecting unit 401 stores the area in the memory. Each boundary coordinate data indicating the boundary between the form and the background is acquired, and each acquired boundary coordinate data is transferred to the image acquisition unit 523.

  The image acquisition unit 523 is a unit that acquires the entire image data from the read image data. Upon receiving each boundary coordinate data, the image acquisition unit 523 refers to the entire image setting table 83 in the memory based on the insurer identification data that has been character-recognized. If the corresponding full acquisition necessity data indicates “○”, the outer area such as the background is included in the read image data area including the contents of all items in the receipt and divided based on each boundary coordinate data. Is created (see FIG. 42), and if the corresponding whole surface acquisition necessity data does not indicate “◯”, the whole surface image data is not created.

  When the entire image creation unit 409b does not create the entire image data, the control unit 10 refers to the page number storage unit 19 to acquire the page number data and the read data acquired together with the page number data. When the partial image data is transferred to the system control unit 2b and the entire image creation unit 409b generates the entire image data, the read data, the created partial image data, and the entire image data are transferred to the system control unit 2b together with the page number data. Thereafter, the operation is the same as in the first embodiment.

<Operation of Example 2>
Next, the operation of the form classification processing system according to the second embodiment will be described.

  In the second embodiment, only the partial image creation unit 408b and the entire image creation unit 409b of the form reading classification unit 1c are different from those in the first embodiment. Therefore, only operations of the partial image creation unit 408b and the entire image creation unit 409b will be described below. Do.

  FIG. 29 is an operation flowchart of the partial image creation unit 408b and the entire image creation unit 409b.

  When the classification reading process is executed in the same manner as in the first embodiment and the reading data is acquired by the reading unit 407, the partial area conversion unit 510 of the partial image creation unit 408b is based on the insurer identification data that has been character-recognized. With reference to the partial image setting table 82 of the memory, the item area to be expanded whose corresponding partial acquisition necessity data is indicated by “◯” is specified, and the area of each specified item area is continuously referred to the memory. Extended area data obtained by extending the data based on the extended value data is generated (S101).

  When the partial region conversion unit 510 generates the extended region data, the image acquisition unit 511 extracts the partial image data including the extended region from the read image data based on the generated extended region data, and also outputs the insurer identification data. The included partial image data is created (S102).

  When the image acquisition unit 511 creates the partial image data, the extended area editing unit 512 specifies the extended range of each item constituting the created partial image data based on the extended value data of the memory and the specified Screen processing is performed on the extended range (S103). For this screen processing, for example, the same translucency processing, dot pattern conversion processing, and the like as in the prior art can be applied.

  When the extended area editing unit 512 completes the screen processing, the acquisition range establishment unit 522 of the entire image creation unit 409b acquires each boundary coordinate data indicating the boundary between the form and the background stored in the memory by the form outline detection unit 401. At the same time, the acquired boundary coordinate data is transferred to the image acquisition unit 523 (S104).

  When the image acquisition unit 523 receives each boundary coordinate data, the image acquisition unit 523 refers to the entire image setting table 83 in the memory on the basis of the character-recognized insurer identification data, and the corresponding entire acquisition necessity data indicates “O”. If this is the case, the entire image data including the contents of all items in the receipt and removing the outer area such as the background from the read image data area divided based on each boundary coordinate data is created and the corresponding entire image is acquired. If the necessity data does not indicate “◯”, the entire image data is not created (S105 & S106).

  When the entire image creation unit 409b does not create the entire image data, the control unit 10 refers to the page number storage unit 19 to acquire the page number data and the read data acquired together with the page number data. When the partial image data is transferred to the system control unit 2b and the entire image creation unit 409b generates the entire image data, the read data, the created partial image data, and the entire image data are transferred to the system control unit 2b together with the page number data. Thereafter, the operation is the same as in the first embodiment.

<Effect of Example 2>
In the form classification processing system according to the second embodiment, a paper form is provided to an insurer who needs a paper form as in the past, and for an insurer who does not want a paper form due to restrictions on storage locations and the like. An electronic media receipt can be provided. Insurers who want both paper forms and electronic media receipts will be able to provide both. As a specific example of the form, it is suitable for the receipt of medical remuneration statements that are issued and issued by medical institutions to insurers such as health insurance associations, and both medical institutions and customers in receipt processing operations that occur in large quantities every month Can be processed quickly without delay.

  Further, partial image data obtained by extracting predetermined items necessary for examination of a receipt including an extended range indicated by previously registered extended value data from the read image data is obtained, and screen processing is performed on the extended range. Therefore, since the characters etc. are described out of the item, partial image data with poor visibility in which the written characters etc. are difficult to identify is obtained, and the electronic image recording the partial image data with poor visibility is recorded. A media receipt can be prevented from being created.

  Furthermore, since the entire image data is created by removing unnecessary outer areas such as the background from the area of the read image data divided based on each boundary coordinate data, the entire image data with poor visibility is acquired. Thus, it is possible to prevent creation of an electronic medium receipt in which the entire image data with poor visibility is recorded. Further, since printing can be performed based on the entire image data from which unnecessary outer areas such as the background are removed during printing, unnecessary toner can be prevented from being consumed by including unnecessary outer areas such as the background.

<Configuration of Example 3>
FIG. 30 is a block diagram illustrating a form classification processing system according to the third embodiment.

  As shown in FIG. 30, the form classification processing system according to the third embodiment includes a form reading classification unit 1d, a system control unit 2b, a correction terminal 3, a printer 4, and an electronic medium creation unit 5. The system control unit 2b The terminal 3, the printer 4, and the electronic medium creation unit 5 are the same as those in the second embodiment.

  FIG. 31 is a block configuration diagram illustrating the form reading classification unit 1d according to the third embodiment.

  As shown in FIG. 31, the form reading / classifying unit 1d includes a control unit 10, a pocket control unit 11, a pocket mechanism 12, a numbering control unit 13, a numbering mechanism 14, a transport control unit 15, a transport mechanism 16, an imaging control unit 17, It has an imaging mechanism 18, a page number storage unit 19, a recognition unit 20b, and a setting table 21, and includes a control unit 10, a pocket control unit 11, a pocket mechanism 12, a numbering control unit 13, a numbering mechanism 14, a transport control unit 15, The transport mechanism 16, the imaging control unit 17, the imaging mechanism 18, the page number storage unit 19, and the setting table 21 are the same as those in the second embodiment.

  FIG. 32 is a block diagram showing the recognition unit 20b of the form reading / classifying unit 1d.

  As shown in FIG. 32, the recognition unit 20b according to the third embodiment includes a form outer shape detection unit 401a, a shape determination unit 402b, a density distribution measurement unit 403, a density determination unit 404, a density correction unit 405, a corrected image storage unit 406, and a reading. A density distribution measurement unit 403, a density determination unit 404, a density correction unit 405, a correction image storage unit 406, a reading unit 407, and a partial image generation unit 408b. This is the same as in Example 2.

  The form outer shape detection unit 401a holds, for example, reference form image amount data indicating the amount of data when an A4 size form is read as an image in a memory (not shown).

  As with the first embodiment, when receiving the read image data, the form outline detecting unit 401a first analyzes the read image data, acquires read image amount data calculated based on the resolution of the read image data, and the like. The read image amount data and the reference form image amount data are compared. If the read image amount data is equal to or less than the reference form image amount data, it is detected that the read image data of the A4 size form is received. The same detection process is executed.

  On the other hand, when the read image amount data is larger than the reference form image amount data, the form outer shape detection unit 401a reads, for example, the read image data of the horizontally long form in which the continuous sheets A and B are manually attached as shown in FIG. Is detected, the read image data is transferred to the density determination unit 404, and thereafter, the same operation as in the second embodiment is performed until partial image data is created.

  The entire image creation unit 409c is a unit that creates entire image data of a horizontally long form, and as illustrated in FIG. 34, a landscape determination unit 520, a boundary detection unit 521, an acquisition range establishment unit 522b, an image acquisition unit 523, a background editing Section 524 and multi-page conversion section 525.

  The landscape determination unit 520 of the full-surface image creation unit 409c is a unit for determining whether the document is a landscape document. For example, reference form image amount data indicating the amount of data when an A4 size form is read as an image. Is stored in a memory (not shown), and when the read image data is received, the read image data is first analyzed, and the read image amount data calculated based on the resolution of the read image data is obtained. And the reference form image amount data are compared, and if the read image amount data is equal to or less than the reference form image amount data, it is determined that it is not a horizontally long form, and thereafter the same operation as in the second embodiment is performed. On the other hand, if the read image amount data is larger than the reference form image amount data, it is determined as a horizontally long form.

  The boundary detection unit 521 is a unit that detects the outer shape of a horizontally long form, and similarly to the form outer shape detection unit 401 according to the first embodiment, reference change amount data and change amount calculation indicating a density change amount that is a detection reference for the form outer shape. Expression (1) is held in a memory (not shown).

  When the landscape determination unit 520 determines that the landscape is a landscape document, the boundary detection unit 521 selects one pixel as a density change calculation target from a plurality of pixels constituting the read image data, and acquires the density value of the selected pixel. Next, the density values of adjacent pixels adjacent to the selected pixel above, below, left, and right are respectively acquired. Then, the obtained density value of each pixel is applied to the above equation (1), and the density change amount between the selected pixel and the adjacent pixel is calculated. When the density change amount is calculated, the reference change amount data is acquired from the memory, and the calculated density change amount and the reference change amount are compared. If the density change amount is larger than the reference change amount, the selection is performed. The pixel is determined as a form boundary region, and the coordinate value of the selected pixel is acquired as boundary coordinate data indicating the boundary between the form and the background. On the other hand, if the density change amount is smaller than the reference change amount, the selected pixel is determined as a non-form boundary region, and the boundary coordinate data of the selected pixel is not acquired. The boundary detection unit 521 continues to perform similar detection processing on adjacent pixels to track the boundary between the form and the background, that is, the form side.

  Note that the density change amount may be calculated for each region divided by 100 × 100 pixels or the like with the selected pixel as the center.

  When the boundary detection unit 521 executes the detection process on all the pixels of the read image data, the boundary detection unit 521 continues to acquire each boundary coordinate data acquired in order to detect the boundary between the single form and the continuous sheet attached to the single form. The pixel intersecting on the extension line of each form side specified based on the above is determined as a form angle area, and the coordinate value of the intersecting pixel is acquired as form angle coordinate data indicating the form angle coordinate.

  Then, among the inner areas of the horizontally long form divided based on the obtained boundary coordinate data and each form corner coordinate data, the area having a smaller vertical width is determined as the boundary between the single form and the continuous sheet. Then, page division coordinate data indicating the area by coordinates is acquired. That is, regarding an image acquired in a state in which a continuous sheet narrower than the vertical width is pasted on a single form (recepto 40), the part where the vertical width of the form part is narrow is determined as the boundary between the receipt 40 and the continuous sheet. Then, the coordinates of the part are stored.

  Further, the boundary detection unit 521 detects an area other than a single form that is divided based on the acquired page division coordinate data in order to detect a boundary between continuous sheets when continuous sheets are attached. With reference to each pixel of the paper region, a region where white pixels continue in the vertical direction is detected.

  When a region in which white pixels are continuous in the vertical direction is detected, it is determined that the region is a boundary between consecutively attached sheets, and continuous page division coordinate data indicating the region by coordinates is acquired. Similarly to the above, the boundary detection process between continuously attached sheets is repeated.

  At this time, since the horizontal width of the single form (recepto 40) and the continuous sheet is usually the same, the horizontal width of the single form (the horizontal width from the single form to the continuous sheet) is moved from the boundary between the single form and the continuous sheet. For the portion, it is efficient to perform the above-described continuous detection of white pixels. In addition, since a certain amount of margin is provided at the edge of the continuous paper, a region in which white pixels are continuous in the vertical direction is further continued in the horizontal direction to some extent (for example, 2 cm or more in terms of length). It is better to determine that white pixels are continuous. In that case, a vertical line passing through the center of the area is determined as a boundary between the continuous sheets.

  As described above, the boundary detection unit 521 performs the above-described detection process, and stores the acquired boundary coordinate data, each form corner coordinate data, page division coordinate data, and continuous page page division coordinate data in a memory (not shown). .

  The acquisition range establishment unit 522b is a unit that determines a range to be extracted as full-scale image data from the read image data. For example, the acquisition range setting unit 522b has a width of a horizontally long form to be expanded with respect to the extracted range, such as 2 mm. The width extension value data indicating the length is held in a memory (not shown).

  When the boundary detection unit 521 completes the detection process, the acquisition range establishing unit 522b acquires boundary coordinate data, form corner coordinate data, page division coordinate data, continuous page division coordinate data, and width extension value data from the memory. Based on each acquired data, a range to be extracted as full-screen image data is determined from the read image data.

  That is, the acquisition range establishment unit 522b reads the boundary coordinate data, the form corner coordinate data, the page division coordinate data, the continuous page division coordinate data, and the width extension value data as shown in FIG. The range is determined so as to extract the entire image data divided into single form, sequel A, and sequel B from the image data, and the extracted range data indicating the determined extraction range is transferred to the image acquisition unit 523. .

  Here, in the present embodiment, the vertical length of the separated sheets A and B is the same as that of the single form, and the horizontal length of the divided forms is relative to the page division coordinate data. , The length indicated by the horizontal width extension value data is the extended length, and the horizontal length of the continuous paper A is the length indicated by the horizontal width extension value data for each of the page division coordinate data and the continuous page division coordinate data. The width of the continuous paper B in the horizontal direction is a length obtained by expanding the length indicated by the horizontal width extension value data with respect to the continuous paper page division coordinate data. That is, an image whose range has been expanded so as to include the outside of the detected boundary is cut out. As a result, overlapping portions are included in each image, so that when the images are referred to, the joints between the cut-out portions can be clearly confirmed. Note that the coordinates of the portion of each image whose range has been expanded are stored so that they can be referred to later.

  The image acquisition unit 523 is a unit that acquires the entire image data from the read image data. Upon receiving the extraction range data, the image acquisition unit 523 refers to the entire image setting table 83 in the memory based on the insurer identification data that has been character-recognized. When the corresponding entire surface acquisition necessity data indicates “◯”, the entire image obtained by dividing the read image data into a single form, a continuous sheet A, and a continuous sheet B as shown in FIG. 36 based on the extraction range data. Each data is created, and if the corresponding full-screen acquisition necessity data does not indicate “◯”, the full-screen image data is not created.

  When the entire image creation unit 409b does not create the entire image data, the control unit 10 refers to the page number storage unit 19 to acquire the page number data and the read data acquired together with the page number data. The partial image data is transferred to the system control unit 2b, and thereafter the same operation as in the first embodiment is performed.

  The background editing unit 524 is a unit that edits an unnecessary area such as a background included in the entire image data of the continuous paper A and the continuous paper B, and the image acquisition unit 523 is the entire image data of the continuous paper A and the continuous paper B. Then, the density of each whole surface image data is analyzed, the black pixel region is converted into the white pixel region, and the screen processing is performed on the region expanded based on the horizontal width extension value data (see FIG. 37). For this screen processing, for example, the same translucency processing, dot pattern conversion processing, and the like as in the prior art can be applied.

  The multi-page converting unit 525 is a unit that synthesizes the created full-screen image data. When the background editing unit 524 completes the screen processing, the full-screen image data of the created single form, sequel A, and sequel B are created. Are combined to create full-screen image data of one file that can be referenced for each page.

  When the entire image creation unit 409c creates the entire image data, the control unit 10 transfers the read data, the created partial image data, and the entire image data together with the page number data to the system control unit 2b. Works as well.

<Operation of Example 3>
Next, the operation of the form classification processing system according to the third embodiment will be described.

  In the third embodiment, since only the form outer shape detection unit 401a and the entire surface image creation unit 409c of the form reading classification unit 1d are different from the second embodiment, only the operations of the form outer shape detection unit 401a and the entire surface image creation unit 409c will be described below. Give an explanation.

  First, the operation of the form outline detecting unit 401a will be described.

  When receiving the read image data, the form outer shape detection unit 401a first analyzes the read image data and acquires read image amount data obtained by calculating the total number of pixels of the read image data. The read image amount data and the reference form image amount data are compared, and if the read image amount data is equal to or less than the reference form image amount data, it is detected that the read image data of the A4 size form is received. 2 is executed.

  On the other hand, when the read image amount data is larger than the reference form image amount data, the form outer shape detection unit 401a reads, for example, the read image data of the horizontally long form in which the continuous sheets A and B are manually attached as shown in FIG. Is detected, the read image data is transferred to the density determination unit 404, and thereafter, the same operation as in the second embodiment is performed until partial image data is created.

  Next, the operation of the entire image creation unit 409c will be described.

  FIG. 38 is an operation flowchart of the entire image creation unit 409c.

  Upon receipt of the read image data, the horizontal length determination unit 520 of the full-surface image creation unit 409c first analyzes the read image data, acquires read image amount data calculated based on the resolution of the read image data, and the acquired read image The amount data and the reference form image amount data are compared, and if the read image amount data is equal to or less than the reference form image amount data, it is determined that it is not a horizontally long form, and thereafter the same operation as in the first embodiment is performed (S108). On the other hand, if the read image amount data is larger than the reference form image amount data, it is determined as a horizontally long form (S101).

  When the landscape determination unit 520 determines that the landscape is a landscape document, the boundary detection unit 521 selects one pixel as a density change calculation target from a plurality of pixels constituting the read image data, and acquires the density value of the selected pixel. Next, the density values of adjacent pixels adjacent to the selected pixel above, below, left, and right are respectively acquired. Then, the obtained density value of each pixel is applied to the above equation (1), and the density change amount between the selected pixel and the adjacent pixel is calculated. When the density change amount is calculated, the reference change amount data is acquired from the memory, and the calculated density change amount and the reference change amount are compared. If the density change amount is larger than the reference change amount, the selection is performed. The pixel is determined to be a ruled line region, and the coordinate value of the selected pixel is acquired as boundary coordinate data indicating the boundary between the form and the background. On the other hand, if the density change amount is smaller than the reference change amount, the selected pixel is determined as a non-ruled line region, and the boundary coordinate data of the selected pixel is not acquired. The boundary detection unit 521 continues to perform the same detection process on all the pixels of the read image data.

  Subsequently, when the boundary detection unit 521 executes the detection process on all the pixels of the read image data, each boundary coordinate data acquired to detect the boundary between the single form and the continuous sheet attached to the single form. The pixel intersecting on the extension line of each form side specified based on the above is determined as a form angle area, and the coordinate value of the intersecting pixel is acquired as form angle coordinate data indicating the form angle coordinate.

  Then, among the inner areas of the horizontally long form divided based on the obtained boundary coordinate data and each form corner coordinate data, the area having a smaller vertical width is determined as the boundary between the single form and the continuous sheet. Then, page division coordinate data indicating the area by coordinates is acquired.

  Further, the boundary detection unit 521 detects each boundary between the continuous sheets when the continuous sheets are attached to each area other than the single form divided based on the acquired page division coordinate data. If the pixel is not referred to and a region where white pixels continue in the vertical direction is not detected, the boundary detection unit 521 determines that no continuous paper is attached, and ends the boundary detection processing (S102).

  On the other hand, when detecting a region in which white pixels continue in the vertical direction, the region is determined as a boundary between consecutively attached sheets, and at the same time, continuous page page division coordinate data indicating the region by coordinates is acquired. Subsequently, similarly to the above, the boundary detection process between consecutively attached sheets is repeated.

  As described above, the boundary detection unit 521 performs the above-described detection process, and stores the acquired boundary coordinate data, each form corner coordinate data, page division coordinate data, and continuous page page division coordinate data in a memory (not shown). .

  When the boundary detection unit 521 completes the detection process, the acquisition range establishing unit 522b acquires boundary coordinate data, form corner coordinate data, page division coordinate data, continuous page division coordinate data, and width extension value data from the memory. Based on each acquired data, a range to be extracted as full-screen image data is determined from the read image data (S103).

  That is, the acquisition range establishment unit 522b reads the boundary coordinate data, the form corner coordinate data, the page division coordinate data, the continuous page division coordinate data, and the width extension value data as shown in FIG. The range is determined so as to extract the entire image data divided into single form, sequel A, and sequel B from the image data, and the extracted range data indicating the determined extraction range is transferred to the image acquisition unit 523. .

  Here, in the present embodiment, the vertical length of the separated sheets A and B is the same as that of the single form, and the horizontal length of the divided forms is relative to the page division coordinate data. , The length indicated by the horizontal width extension value data is the extended length, and the horizontal length of the continuous paper A is the length indicated by the horizontal width extension value data for each of the page division coordinate data and the continuous page division coordinate data. The width of the continuous paper B in the horizontal direction is a length obtained by expanding the length indicated by the horizontal width extension value data with respect to the continuous paper page division coordinate data.

  Upon receiving the extraction range data, the image acquisition unit 523 refers to the full-screen image setting table 83 in the memory based on the character-recognized insurer identification data, and the corresponding full-screen acquisition necessity data indicates “◯”. 36, based on the extraction range data, as shown in FIG. 36, the read image data is respectively created as full-screen image data divided into a single form, a continuous paper A, and a continuous paper B, and the corresponding full-surface acquisition necessity data is obtained. If “◯” is not indicated, the entire image data is not created (S104 & S105).

  When the entire image creation unit 409b does not create the entire image data, the control unit 10 refers to the page number storage unit 19 to acquire the page number data and the read data acquired together with the page number data. The partial image data is transferred to the system control unit 2b, and thereafter the same operation as in the first embodiment is performed.

  When the image acquisition unit 523 creates the entire image data of the continuous paper A and the continuous paper B, the background editing unit 524 continuously analyzes the density of the entire image data and converts the black pixel area into the white pixel area. At the same time, the screen processing is performed on the area expanded based on the horizontal width expansion value data (see FIG. 37). For this screen process, for example, a translucent process, a dot pattern conversion process, and the like similar to the prior art can be applied (S106).

  The multi-page converting unit 525 is a unit that synthesizes the created full-screen image data. When the background editing unit 524 completes the screen processing, the full-screen image data of the created single form, sequel A, and sequel B are created. Are created, and the entire image data of one file that can be referred to for each page is created (S107).

  When the entire image creation unit 409c creates the entire image data, the control unit 10 transfers the read data, the created partial image data, and the entire image data together with the page number data to the system control unit 2b. Works as well.

<Effect of Example 3>
In the form classification processing system of the third embodiment, a paper form is provided to an insurer who needs a paper form as in the conventional case, and for an insurer who does not want a paper form due to restrictions on storage locations and the like. An electronic media receipt can be provided. Insurers who want both paper forms and electronic media receipts will be able to provide both. As a specific example of the form, it is suitable for the receipt of medical remuneration statements that are issued and issued by medical institutions to insurers such as health insurance associations, and both medical institutions and customers in receipt processing operations that occur in large quantities every month Can be processed quickly without delay.

  Further, when a receipt is attached to a single form and the shape of the receipt is horizontally long, the boundary between the receipt and each continuous sheet is detected from the horizontally long read image data, and based on the detected boundary. In this way, the horizontally long read image data is divided and the entire image data of the receipt and each successive sheet is created, so that the visibility of the entire image at the time of examination can be improved. In addition, since unnecessary areas such as the background of each full face image data are removed and each adjacent area is screen-processed, the boundary between the full face image data can be easily identified. Since the printing process based on the entire image data from which unnecessary outer areas such as the background are removed can be executed, unnecessary toner can be prevented from being consumed by including unnecessary outer areas such as the background. In addition, the entire image data is synthesized to create one file of entire image data that can be referenced for each page, so each image data can be managed in one file, improving the convenience of data management. it can.

  In the above embodiment, the extension value data is stored in the memory in advance. However, the present invention is not limited to this, and a configuration that can be set by a management terminal or the like may be employed.

The block block diagram which shows the form classification processing system of this invention The schematic diagram which shows the classification receipt delivery method of Example 1 Block configuration diagram showing a conventional form classification processing system Schematic diagram showing the conventional method of delivering categorized receipts Block configuration diagram showing the form reading classification unit 1b Block configuration diagram showing the recognition unit 20 Block configuration diagram showing the system control unit 2b An exemplary diagram showing a business setting table 51 Example showing the pocket setting table 80 An exemplary view showing a reading setting table 81 An exemplary view showing a partial image setting table 82 An exemplary view showing the entire image setting table 83 Example showing the service setting table 84 An exemplary diagram showing an invalid form pocket setting table 85 Explanatory diagram showing the flow of receipt classification An exemplary diagram showing a read data group stored in the read result storage unit 53 An exemplary diagram showing a partial image data group stored in the partial image storage unit 54 An exemplary diagram showing a full image data group stored in the full image storage unit 55 An exemplary diagram showing a billing point total data group stored in the total amount storage unit 56 Example diagram showing list to be provided to insurer C Illustration showing electronic media receipt to be provided to insurer C Operation flow chart of form classification processing system of embodiment 1 Operation flow chart of form reading classification unit 1b Block configuration diagram showing a form classification processing system of Embodiment 2 Block configuration diagram showing the form reading classification unit 1c Block configuration diagram showing the recognition unit 20a The block block diagram which shows the partial image preparation part 408b Block configuration diagram showing the entire image creation unit 409b Operation flowchart of the partial image creation unit 408b and the entire image creation unit 409b Block configuration diagram showing a form classification processing system of Embodiment 3 Block configuration diagram showing the form reading classification unit 1d Block configuration diagram showing the recognition unit 20b Illustration showing a landscape document Block configuration diagram showing the entire image creation unit 409c Illustration showing a landscape report (border detection example) Illustration showing segmented full-screen image data Illustration showing full-screen image data after screen processing Operation Flowchart of Full Image Creation Unit 409c Illustration showing an example of outline detection of a form Illustration showing the form determination example of a form Illustration showing partial image data after expansion Illustration showing full-scale image data with background removed

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Form reading classification | category part 1b Form reading classification | category part 1c Form reading classification | category part 1d Form reading classification | category part 2 System control part 2b System control part 3 Correction terminal 4 Printer 5 Electronic medium creation part 10 Control unit 11 Pocket control part 12 Pocket mechanism 13 Numbering Control unit 14 Numbering mechanism 15 Transport control unit 16 Transport mechanism 17 Imaging control unit 18 Imaging mechanism 19 Page number storage unit 20 Recognizing unit 20a Recognizing unit 20b Recognizing unit 21 Setting table 40 Recto 41 Recto 42 Claiming point total data 43 Rect 44 Claiming point Total data 45 Receipt 46 Claimed point total data 47 Electronic media receipt 50 Business AP section 51 Business setting table 52 Program section 53 Reading result storage section 54 Partial image storage section 55 Full image storage section 56 Total amount storage section 57 Electronic medium Image storage unit 70 Printer channel 71 Electronic medium channel 80 Pocket setting table 81 Reading setting table 82 Partial image setting table 83 Whole image setting table 84 Service setting table 85 Invalid form pocket setting table 90 Electronic medium receipt 401 Form outline detection part 401a Form outline Detection unit 402 Shape determination unit 402b Shape determination unit 403 Density distribution measurement unit 404 Density determination unit 405 Density correction unit 406 Correction image storage unit 407 Reading unit 408 Partial image generation unit 408b Partial image generation unit 409 Full image generation unit 409b Full image generation Section 409c Whole image creation section 510 Partial area conversion section 511 Image acquisition section 512 Extended area editing section 520 Horizontal length determination section 521 Boundary detection section 522 Acquisition range establishment section 522b Acquisition range establishment section 523 Acquisition unit 524 background editing section 525 multi-page conversion unit

Claims (9)

In the form classification processing system that reads the description content of the received form and classifies it into a pocket mechanism for each delivery destination, and holds the read description data for each classified form for each delivery destination,
Based on the density change amount of each pixel constituting the read image data including the background obtained by reading the form, the outline of the form is detected, the boundary coordinate data indicating the boundary between the background and the form, and the corner of the form A form outer shape detection unit for obtaining form angle coordinate data indicating
From the read image data, based on area data indicating the area position of each item on the pre-registered form, a partial image creation unit that creates partial image data necessary for examination of the description content of the form,
From the read image data, on the basis of the boundary coordinate data, and an entire image creation unit that creates a whole face image data,
The partial image creation unit creates partial image data obtained by extracting predetermined items necessary for examination of a form including an extended range indicated by previously registered extended value data from the read image data, and for the extended range. A form classification processing system, wherein screen processing is performed to clarify the extended range.   The form classification processing system according to claim 1, wherein the screen processing is translucent processing or dot pattern conversion processing. The form is a receipt, the delivery destination is an insurer,
The form classification processing system according to claim 1, wherein the description content is a claim score described in a receipt, and the claim score total data obtained by totaling the claim scores is held for each insurer. An abnormal shape pocket mechanism that accommodates abnormal forms ,
When shape determination unit determines that abnormal form shape, form classification processing system according to claim 1, characterized in that it comprises a pocket controller to discharge該帳votes to the abnormal shape pocket mechanism. When shape determination unit determines that abnormal form shape, the form to the shape anomaly image determination result form classification process according to claim 1, characterized in that it comprises a numbering mechanism to perform printing processing based on data indicating system.   The form classification processing system according to claim 1, wherein the full-surface image creation unit creates full-face image data of the form from which the background is removed based on the acquired boundary coordinate data and form corner coordinate data. . In the form classification processing system that reads the description content of the received form and classifies it into a pocket mechanism for each delivery destination, and holds the read description data for each classified form for each delivery destination,
Based on the amount of change in the density of each pixel constituting the read image data including the background obtained by reading a horizontally long form consisting of a single form and a continuous form, the outline of the horizontally long form is detected, and the boundary between the background and the form is detected. A boundary detection unit that respectively obtains boundary coordinate data indicating, corner angle coordinate data indicating a corner of the form, and page division coordinate data indicating a boundary between a single form and a continuous sheet ;
From the read image data before Symbol Horizontal form, the width extension value data indicating the length of the extension to the width to a single form and a connection sheet, the acquired the boundary coordinate data, on the basis of the form angle coordinate data and page division coordinate data An entire image creation unit for creating the entire image data of single forms and extended sheets extracted by expanding the width;
A form classification processing system comprising a background editing unit that performs screen processing on a range in which the horizontal width of each full-screen image data of the created single form and continuous paper is expanded.   The form classification processing system according to claim 7, wherein the screen processing is translucent processing or dot pattern conversion processing.   A multi-page conversion unit is provided for synthesizing each full-surface image data and creating full-image data of one file that can be referred to for each page when the full-screen image data of a single form and a continuous sheet is created. The form classification processing system according to claim 7.

Images