JP2021043500A - Information processing apparatus and program - Google Patents

Abstract

To cause validity of an imprint that exists at a specific stamped part in a document to be confirmed without opening a document file.SOLUTION: An image formation apparatus 10 includes a document type identification unit 13 that identifies a type of a document from a read image of the document, a key acquisition unit 14 that acquires a key corresponding to the type of the identified document from a key information storage unit 22, an image extraction unit 15 that extract an image linked to each key as an imprint, a reference character acquisition unit 16 that acquires a character included in a stamp to be stamped at a specified stamped portion in the document as the reference character, a verification character acquisition unit 171 that acquires a character included in an image of the imprint extracted from the image extraction unit 15 as a verification character, a determination unit 17 that determines the validity of the imprint existing at the specified stamped portion in the document to be processed by comparing the verification character and the reference character, and a determination result output unit 18 that presents a determination result of the validity of the imprint for each key by the determination unit 16.SELECTED DRAWING: Figure 1

Description

The present invention relates to an information processing device and a program.

Documents such as application forms, contracts, and approval documents are treated as official documents, with the original stamped at the designated stamped location. It may also be digitized and saved as a document file.

Some documents are required to be stamped with the correct seal, but to check in the document file whether the stamp is stamped with the correct stamp, open the document file and display it on the screen. It was necessary to confirm whether it was the correct imprint of the seal.

Japanese Unexamined Patent Publication No. 8-212343

However, it is troublesome and troublesome to open the document file and display it on the screen in order to confirm whether or not the imprint of the correct seal is present at a specific stamped part in the document.

An object of the present invention is to confirm the validity of an imprint existing at a specific imprinted portion in a document without opening the document file.

The information processing apparatus according to the present invention includes a processor, which extracts an image of an imprint existing at a specific imprinted portion in the document from a scanned image of the document to be processed and includes the image of the imprint. Characters are acquired as verification characters, characters included in the seal to be stamped on the specific stamped portion are acquired as reference characters, and are present at the specific stamped portion by comparing the verification character with the reference character. It is characterized in that the validity of the imprint to be imprinted is determined, and the determination result of the validity of the imprint existing at the specific imprinted portion is presented.

Further, when the character included in the image of the imprint cannot be acquired by using the character recognition function, the processor extracts the linear component included in the image of the imprint and stamps the document. The tilt angle of each extracted linear component with respect to the orientation of the power stamp is detected, the tilt angle of the image of the imprint is estimated by referring to the result of totaling the quantity of the extracted linear components for each tilt angle, and the image of the imprint is estimated. It is characterized in that the characters included in the image of the imprint are acquired by executing the character recognition function after correcting the inclination of the imprint.

Further, when the detected tilt angle has a set of orthogonal tilt angles, the processor includes a straight line included in the set of tilt angles that maximizes the sum of the aggregated results in the set of tilt angles. It is characterized in that the tilt angle of the linear component having a large aggregated result among the components is estimated as the tilt angle of the image of the imprint.

Further, the processor estimates that the tilt angle of the linear component that maximizes the aggregated result is the tilt angle of the imprint image when there is no set of orthogonal tilt angles in the detected tilt angles. It is a feature.

Further, the result of totaling the quantity of the linear components for each tilt angle is the sum of the number of linear components of the tilt angle or the length of the linear components of the tilt angle.

Further, the processor executes a character recognition function on four types of images obtained by rotating the image of the imprint after tilt correction based on the estimated tilt angle by 90 degrees, and among the four types of images. It is characterized in that the image having the maximum character recognition accuracy is extracted as an image of the imprint that stands upright with the seal.

Further, the processor identifies the type of the document to be processed by analyzing the scanned image of the document to be processed, and the specific characters associated with the specific image that should exist in the document are for each type of document. By referring to the specific character information set in, the specific character associated with the specific image that should exist in the document in the type of the document to be processed is acquired, and the acquired specific character is supported. An image that matches the preset search conditions is extracted as an image of the imprint.

Further, the processor is characterized in that the acquired specific character and the determination result of the validity of the imprint associated with the specific character are presented as a set.

Further, the processor is characterized in that the acquired specific character and the determination result of the validity of the imprint associated with the specific character are combined and included in the file name of the document to be processed. ..

Further, the processor is characterized in that it generates a file including a set of the acquired specific character and the determination result of the validity of the imprint associated with the specific character.

The information processing apparatus according to the present invention includes a processor, and the processor extracts an image of an imprint existing at a specific imprinted portion in the document from a scanned image of the document to be processed and includes the image of the imprint. When the character is not acquired by using the character recognition function, the linear component contained in the image of the imprint is extracted, and the inclination angle of each extracted linear component with respect to the direction of the stamp to be stamped on the document. Is detected, the tilt angle of the imprint image is estimated by referring to the result of totaling the quantity of the extracted linear components for each tilt angle, and the character recognition function is executed after the tilt correction is performed on the imprint image. By doing so, it is characterized in that the characters included in the image of the imprint are acquired.

The program according to the present invention has a function of extracting an image of an imprint existing at a specific stamped portion in the document from a scanned image of the document to be processed by a computer, and character recognition of characters included in the image of the imprint. If it cannot be obtained by using the function, the linear component contained in the image of the imprint is extracted, and the tilt angle of each extracted linear component with respect to the orientation of the stamp to be stamped on the document is detected. The function of estimating the tilt angle of the image of the imprint by referring to the result of totaling the quantity of the linear components for each tilt angle, and the character recognition function after correcting the tilt of the image of the imprint. The function of acquiring the characters contained in the image of the imprint is realized.

According to the invention of claim 1, it is possible to confirm the validity of the imprint existing at a specific stamped portion in the document without opening the document file.

According to the invention of claim 2, most of the linear components constituting the character can be tilt-corrected on the assumption that the image of the imprint is related to the orientation of the seal to be imprinted on the document.

According to the invention of claim 3, most of the linear components constituting the character are oriented in a direction parallel to or orthogonal to the direction of the seal to be stamped on the document, and the image of the imprint is formed. Tilt correction can be performed.

According to the invention of claim 4, most of the linear components constituting the character can be tilt-corrected on the assumption that the image of the imprint is parallel to the direction of the seal to be imprinted on the document.

According to the fifth aspect of the present invention, it can be estimated that the deviation between the direction in which the number of linear components is large and the reference direction is the inclination angle of the image of the imprint.

According to the invention of claim 6, the characters included in the image of the imprint can be acquired with high accuracy.

According to the invention of claim 7, the image of the imprint contained in the document can be automatically extracted.

According to the invention of claim 8, it is possible to confirm the determination result of the validity of the imprint associated with a specific character.

According to the invention of claim 9, it is possible to confirm the determination result of the validity of a specific character and the imprint associated with the specific character without opening the document file.

According to the invention of claim 10, when there are a plurality of specific characters corresponding to the type of the document, or when there are a plurality of documents to be processed, the determination result of the validity of the imprint is collectively confirmed. Can be made to.

According to the invention of claim 11, it is possible to confirm the validity of the imprint existing at a specific imprinted portion in the document without opening the document file.

According to the invention of claim 12, it is possible to confirm the validity of the imprint existing at a specific imprinted portion in the document without opening the document file.

It is a block block diagram which shows one Embodiment of the information processing apparatus which concerns on this invention. It is a hardware block diagram of the image forming apparatus in this embodiment. It is a figure which shows the data structure example of the key information stored in the key information storage part in this embodiment. It is a figure which shows the data structure example of the search condition information stored in the search condition information storage part in this embodiment. It is a flowchart which shows the imprint justification determination processing in this embodiment. It is a figure which shows the schematic layout of the document which the document type is classified into a drawing in this embodiment. It is a figure which shows the example of the imprint which was tilted and imprinted. It is a figure which shows only the linear component extracted from the imprint shown in FIG. 7A. It is a figure which shows the inclination angle of the linear component in FIG. 7B. It is a figure which shows the aggregation result for each inclination angle of the number of linear components and the length shown in FIG. 7B. It is a figure which shows the imprint which rotated the imprint after the inclination correction of the imprint shown in FIG. 7A by 90 degrees, and the result of character recognition with respect to the imprint.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block configuration diagram showing an embodiment of an information processing device according to the present invention. In the present embodiment, an image forming apparatus having a built-in computer, which is an information processing apparatus, will be described as an example.

FIG. 2 is a hardware configuration diagram of the image forming apparatus 10 according to the present embodiment. The image forming apparatus 10 can be formed by a multifunction device equipped with various functions such as a copy function and a scanner function. In FIG. 2, the CPU 31 controls the operation of various mechanisms mounted on the present device, such as the scanner 34 and the printer engine 36, according to the program stored in the ROM 39. The address data bus 32 communicates data by connecting to various mechanisms controlled by the CPU 31. The operation panel 33 accepts instructions from the user and displays information. The scanner 34 reads a document set by the user. The HDD 35 stores an electronic document or the like read by using the scanner 34. The printer engine 36 prints an image on the output paper according to an instruction from the control program executed by the CPU 31. The network interface (I / F) 37 connects the network 1 to transmit electronic data generated by the device, receive e-mails sent to the device, access the device via a browser, and the like. It will be used. The network interface 37 in the present embodiment receives a reference character used for determining the validity of the imprint from an external system via the network 1. The RAM 38 is used as a work memory when executing a program and as a communication buffer when transmitting and receiving electronic data. The ROM 39 stores various programs related to the control of the present device and the transmission / reception of electronic data. When various programs are executed, each component described later exerts a predetermined processing function. The external media interface (I / F) 40 is an interface with an external memory device such as a USB memory or a flash memory. The hardware configuration of the image forming apparatus 10 in the present embodiment may be the same as the conventional configuration.

Returning to FIG. 1, the image forming apparatus 10 in the present embodiment includes a scan data acquisition unit 11, a character recognition processing unit 12, a document type identification unit 13, a key acquisition unit 14, an image extraction unit 15, and a reference character acquisition unit 16. It has a determination unit 17, a determination result output unit 18, a document type information storage unit 21, a key information storage unit 22, a search condition information storage unit 23, and an output format information storage unit 24. The components not used in the description in the present embodiment are omitted from the drawings.

The scan data acquisition unit 11 acquires scan data (hereinafter, also referred to as “scanned image”) read by using the scanner 34. The document handled in the present embodiment is a document in which text characters including specific characters and specific images are mixed. The "specific character" is a text character that is set for each type of document and is associated with a specific image that should exist in the document. In this embodiment, a specific character is referred to as a "key". Further, the "specific image" in the present embodiment is an image of an imprint. Strictly speaking, what is in the paper document is the imprint, and what is included in the scanned image of the paper document is the image of the imprint. In some cases.

The character recognition processing unit 12 performs character recognition using an OCR (Optical Character Reader) function (also referred to as a “character recognition function”) to perform character recognition, and thus a character string (that is, a text character) described in a read document. Is extracted. The document type specifying unit 13 specifies the type of the document from the scanned image of the read document. The key acquisition unit 14 links to the imprint existing in the document in the type of the document by referring to the specific character information which is the information about the specific character (that is, the key) associated with the imprint existing in the document. Get the attached key. The specific character information is stored in the key information storage unit 22 as key information.

The image extraction unit 15 extracts an image that matches the search conditions, that is, an image of the imprint, corresponding to the key acquired by the key acquisition unit 14 from the scanned image. The search condition information regarding the search condition is preset in the search condition information storage unit 23. The reference character acquisition unit 16 acquires characters included in the seal to be stamped at a specific stamped portion of the document to be processed as reference characters.

The determination unit 17 determines the validity of the imprint existing at a specific stamped portion of the document to be processed by comparing the verification character with the reference character. The verification character acquisition unit 171 included in the determination unit 17 acquires the characters included in the imprint image extracted by the image extraction unit 15 as verification characters. The determination result output unit 18 outputs a determination result according to a designated output format. The output format information regarding the output format is preset in the output format information storage unit 24.

FIG. 3 is a diagram showing a data configuration example of key information stored in the key information storage unit 22 in the present embodiment. The key information includes one or more keys corresponding to the document type for each document type.

FIG. 4 is a diagram showing a data configuration example of the search condition information stored in the search condition information storage unit 23 in the present embodiment. The search condition information is set for each type of document. FIG. 4 shows an example of setting search condition information when the document type is a drawing. The search condition is set in association with each key corresponding to the type of the document in the search condition information. The key when the document type is drawing is the approver, drafting, structure, design and inspection as illustrated in FIG. 3, but as illustrated in FIG. 4, the key is when the document type is drawing. Search conditions are set for each. The search conditions corresponding to each key include "conditions for images", which are conditions for identifying the imprint that exists associated with the acquired key in the document to be processed, and the key and the document to be processed. The "positional relationship", which is a condition indicating the positional relationship with the imprint existing associated with the machine key, is included. By setting to "positional relationship", a specific imprint location corresponding to the imprint is specified. In the setting example shown in FIG. 4, the same item value is set for each key, but different search conditions may be specified according to the layout of the document. The "condition for an image" may include a condition relating to at least one of the colors or shapes contained in the imprint. In the setting example shown in FIG. 4, both the "red circle" and the red and round shapes are included.

The document type information storage unit 21 and the output format information storage unit 24 will be described together with the description of the operation.

Each component 11 to 18 in the image forming apparatus 10 is realized by a cooperative operation of a computer built in the image forming apparatus 10 and a program operated by a CPU 31 mounted on the computer. Further, each of the storage units 21 to 24 is realized by the HDD 35 mounted on the image forming apparatus 10. Alternatively, the RAM 38 or an external storage means may be used via the network.

Further, the program used in the present embodiment can be provided not only by communication means but also by storing it in a computer-readable recording medium such as a CD-ROM or a USB memory. The programs provided by the communication means and the recording medium are installed in the computer, and various processes are realized by sequentially executing the programs by the CPU of the computer.

The position to be stamped may differ depending on the type of document and the form of the document, but in the document handled in this embodiment, a stamp is formed by stamping a specific stamped portion with a specific stamp. .. However, in some cases, the user may mistakenly stamp an incorrect stamped portion. Alternatively, the user may stamp a specific stamped portion with an erroneous stamp. In other words, a specific stamp may be stamped with the wrong stamp. Therefore, in the present embodiment, it is not necessary to individually open the document file to confirm whether the imprint is formed by correctly imprinting the imprint on the predetermined imprinted portion with the predetermined seal. It is characterized by having done it.

In the present embodiment, the "correct seal" means a seal to be stamped on a specific stamped portion, and is synonymous with the above-mentioned "specific stamp". Only when the seal is stamped on the correct stamped part will it become the "correct stamp". In other words, the imprint is valid only when the imprint and the imprinted part are in the correct combination. "Incorrect seal" means a seal other than the correct seal.

Next, the operation in the present embodiment will be described. Hereinafter, the process for determining the validity of the characteristic imprint in the present embodiment will be described with reference to the flowchart shown in FIG.

First, when the user causes the scanner 34 to read the document to be processed, the scan data acquisition unit 11 acquires the scanned image of the document (step 101). Subsequently, the character recognition processing unit 12 extracts the character string described in the read document by performing character recognition using the OCR function (step 102). When executing OCR, preprocessing such as cleansing for removing the upright image and the background color of the scanned image may be performed.

Next, the document type specifying unit 13 specifies the type of the document (step 103). Specifically, the type of document is specified by using one of the methods exemplified below.

First, when a data code such as a QR code (registered trademark) that identifies a document form or document identification information is attached to the document, the data code is read. In this case, the document type information storage unit 21 associates the data code with the type of the document corresponding to the data code and stores it as the document type information, and the document type identification unit 13 obtains from the read image. The type of document is specified by collating the data code with the data code included in the document type information.

Second, the type of document is estimated and identified by analyzing the scanned image of the document, especially the layout. For example, the position of a ruled line on a document is detected, and the layout of the ruled line is acquired. The document type is estimated from the layout of this ruled line. Alternatively, when the ruled lines form a table, the names of the items included in the table are extracted. In this case, the document type information storage unit 21 stores the item names of the table in association with the document type as the document type information, and the document type identification unit 13 stores the table items obtained from the scanned image. By collating the list of names with the list of item names included in the document type information, the document type is identified by estimation with reference to the matching rate of the item names and the like.

Third, the document name in the document is extracted by analyzing the scanned image of the document. Generally, a document has a document name, and the position of the document is at the top of the document or in the upper center of the document. In addition, parentheses are attached and the character size is large. Therefore, a character string having such a characteristic is estimated as a document name and extracted. In this case, the document type information storage unit 21 associates the document type with the document type and stores the document type information, and the document type identification unit 13 stores the document name obtained from the scanned image as the document. The type of document is identified by estimation by collating it with the document name included in the type information.

Fourth, when the scanner 34 reads the document, the user is asked to specify the type of the document. More specifically, when the scan data acquisition unit 11 reads the document, the document type identification unit 13 displays the document type input screen on the operation panel 33, and causes the user to input the document type from the input screen. .. Alternatively, the document type information storage unit 21 stores the document type selection candidates, and the document type identification unit 13 displays the document type selection screen on the operation panel when the scan data acquisition unit 11 reads the document. Display on 33. On the selection screen, the types of documents read from the document type information storage unit 21 are displayed in a list. Then, the document type specifying unit 13 specifies the type of the document selected by the user.

When the document type specifying unit 13 specifies the document type by any method, the key acquisition unit 14 acquires the key set corresponding to the specified document type from the key information storage unit 22 ( Step 104). According to the setting example shown in FIG. 3, when the document type specified by the document type specifying unit 13 is a drawing, the key acquisition unit 14 acquires the approver, drafting, structure, design, and drawing from the key information storage unit 22. Will be done.

Next, the image extraction unit 15 repeatedly executes the following processing for each acquired key. First, one unprocessed key that has not been subjected to the following processing is selected (step 105). The order of the keys to be selected does not have to be limited. Subsequently, the image extraction unit 15 collates the selected key with the character string obtained in the character recognition process performed in step 102, and identifies the position of the key on the document. The character recognition process may be performed again at this point without using the result of the character recognition process performed in step 102. Subsequently, the image extraction unit 15 acquires the search condition corresponding to the key to be processed when the document type is a drawing from the search condition information storage unit 23. For example, when the key is "Approver", the image extraction unit 15 specifies the position of the document in which "Approver" is printed. Then, according to the setting example of the search condition shown in FIG. 4, the image is within 3 cm below the print position of the key "approver" of the document, and the color is red and the shape is a circle (that is, a circular shape). ) Is extracted (step 106).

Subsequently, the reference character acquisition unit 16 acquires, as a reference character, a character included in the seal to be stamped at a specific stamped portion corresponding to the key to be processed (step 107). Acquisition of reference characters is carried out as follows.

Here, if the person's name is at least included in the imprint, for example, in the document to be processed (“design drawing” in the example shown in FIG. 6), who (or or) is imprinted corresponding to each key. In many cases, the person in which department) is unique or roughly specified. Therefore, when a person (or a plurality of people) who stamps each key is stored in a database in advance, the database is accessed and the name of the person corresponding to the key to be processed is acquired. Alternatively, the database of the company is accessed to identify the creator of the document and the department to which the document belongs, and the names of members of the department or the boss of the department are obtained. Further, depending on the document, the person's name of the person who stamps may be pre-printed corresponding to the stamped portion. In this case, the person's name printed on the document is acquired as the reference character. For convenience of explanation, it is assumed here that one person's name is acquired as a reference character.

When the reference character acquisition unit 16 acquires the reference character, the determination unit 17 subsequently acquires the character as a verification character by analyzing the image of the imprint extracted by the image extraction unit 15 (step 108). Specifically, the verification character acquisition unit 171 in the determination unit 17 extracts characters included in the imprint image by performing character recognition using the OCR function. Then, the determination unit 17 compares the reference character and the verification character, and if they match (Y in step 110), the imprint obtained from the document is imprinted and formed by the correct seal, that is, it is legitimate. It is determined that there is (step 111). On the other hand, if the reference character and the verification character do not match (N in step 110), the determination unit 17 determines that the imprint obtained from the document is not stamped with the correct seal, that is, is unreasonable (N). Step 112).

FIG. 6 is a diagram showing a schematic layout of documents in which the types of documents are classified into drawings. In FIG. 6, a stamping column provided as a predetermined stamping place is provided at the lower right of the document, and a person who should stamp the document will stamp a specific stamping column of the document. Since the key is printed on the upper part of the stamp field, the user can confirm the stamp field to be stamped by referring to the key. The stamp column 2 shown in FIG. 6 is a stamp column associated with the key “approver”, and the stamp column 2 shows an example in which the seal of the user “Kubo” is stamped. That is, the verification character is "Kubo". Here, if the reference character acquisition unit 16 has acquired "Kubo" as the reference character corresponding to the key "approver", it is determined that the imprint of the imprint column 2 is valid. If the reference character acquisition unit 16 has acquired a character other than "Kubo" as the reference character corresponding to the key "approver", it is determined that the imprint of the imprint column 2 is unreasonable.

When the same processing is performed on the keys that have not been processed as described above (N in step 113, steps 105 to 112), and the processing is performed on all the keys acquired by the key acquisition unit 14 (step 113). Y) in 113, the determination result output unit 18 presents the determination result of the validity of the imprint by the determination unit 17 (step 114).

By the way, since the imprint is not necessarily imprinted in the correct direction, it may not be possible to accurately extract the character string included in the imprint image even if character recognition is performed using the OCR function. In the stamp column 3 shown in FIG. 6, an example of a stamp imprint that was not stamped in the correct direction is shown. The "correct orientation" here means the "orientation of the stamp to be stamped on the document" as illustrated in the stamp column 2, and is usually the upright state of the stamp. "Upright" means that the upper and lower parts of the seal and the imprint are the same. The imprint in the stamp column 2 shown in FIG. 6 is a state in which the stamp is upright because the character string written horizontally is stamped in parallel along the x-axis direction (that is, the horizontal direction) of the document shown in FIG. It will be stamped. In the case of a vertically written character string, the correct orientation is when the documents shown in FIG. 6 are stamped in parallel along the y-axis direction (that is, the vertical direction). That is, when the stamp is stamped in the direction of the stamp to be stamped on the document, the determination unit 17 can accurately extract the character string from the stamp impression by executing the OCR function.

On the other hand, in the stamp column 3 shown in FIG. 6, the stamp is not stamped so that the person's name appears along the x-axis direction of the document shown in FIG. 6, that is, the direction of the stamp to be stamped on the document is not. An example of tilting and imprinting is shown. With this imprint, it can be assumed that the person's name cannot be accurately extracted using the OCR function. The present embodiment is characterized in that even if the image of the imprint is used as it is, it is possible to deal with the case where the characters included in the imprint cannot be acquired by using the OCR function. In this case, the verification character acquisition unit 171 processes as follows.

FIG. 7A is a diagram in which only the imprint in the imprint column 3 shown in FIG. 6 is extracted. The verification character acquisition unit 171 first extracts the vector information of the line from the imprint and extracts the linear component. FIG. 7B shows an example of an image in which only the linear component is extracted from the imprint. Then, when the linear components are in contact with each other, for example, "field", each linear line segment is separated so that each linear component can be processed. Subsequently, the verification character acquisition unit 171 detects the inclination angle and the length of the line segment from the reference horizontal direction (direction along the x-axis direction) for each of the extracted linear components. FIG. 7C shows a case where the inclination angle θ of the linear component 4 is detected. When the inclination angle and the length of the line segment are detected for each linear component, the verification character acquisition unit 171 totals the number of linear components for each detected inclination angle. In addition, the length of the linear component is totaled for each detected angle. FIG. 8 is a diagram showing an example of the result of totaling the number and length of the linear components obtained for each angle.

Subsequently, the verification character acquisition unit 171 estimates the inclination angle of the imprint according to the following procedure in order to correct the inclination of the imprint.

First, from the table shown in FIG. 8, two tilt angles having a large number of linear components, a1 and a2 in the numerical example shown in FIG. 8, are extracted. Then, if the inclination angles a1 and a2 are orthogonal to each other, the inclination angle a1 having a large number is estimated as the inclination angle of the imprint.

When the corresponding set of tilt angles does not exist, it is determined whether the tilt angles having the largest number are orthogonal to each other in order from the tilt angle having the largest number. For example, when the tilt angle a1> a2> a3> a4> a5, the combination of a1 and a2, a1 and a3, a1 and a4, a1 and a5, then a2 and a3, a2 and a4, and a2 and a5. judge. Then, when there are a plurality of orthogonal pairs, the sum of the results aggregated in the tilt angle pairs, that is, the tilt angle pair that maximizes the sum of the number of linear components is specified, and the specified pair is specified. Of the tilt angles included in, the tilt angle with the largest number is estimated as the tilt angle of the imprint.

If the corresponding set still does not exist, the tilt angle at which the number of linear components is maximized (a1 in the above example) is estimated as the tilt angle of the imprint.

When the inclination angle of the imprint is estimated as described above, the verification character acquisition unit 171 corrects the inclination of the image of the imprint according to the estimated inclination angle. Actually, it is considered that the image of the imprint is upright due to this correction, but when the seal is stamped in the inverted state (when the stamp is stamped upside down), the image of the imprint is inverted. Since it is possible that the state is corrected, in the present embodiment, the following processing is performed so that this case can also be dealt with.

As shown in FIG. 9, the image of the imprint after tilt correction is rotated by 90 degrees to form four types of images, that is, images of imprints in four directions of up, down, left, and right. Then, the OCR function is executed for each image to extract the character string. FIG. 9 is a diagram showing an image of imprints in four directions and a result of character recognition for each image, that is, an extracted character string corresponding to each image. As is clear from this result, it is considered that the character recognition accuracy of the image of the imprint in the upright state surrounded by the thick line 5 is maximized. The verification character acquisition unit 171 acquires the characters included in the imprint, that is, the verification characters, by executing the OCR function after performing the tilt correction as described above.

When the seal is a data seal as stamped in the seal column 3, the set of the department name, the person's name, and the date is used as the reference character. In the case of a date, when comparing the reference character and the verification character, the match of the date indicated by the date may be obtained, or the date indicated by the verification character is in a predetermined range specified by the reference character. If it is included in, it may be regarded as a match. Alternatively, if the information of the date is included in the verification character, it may be regarded as a match.

In general, it is considered that a kanji contains a linear component more than a curved component, and the linear component is often oriented in the horizontal direction or the vertical direction. The "Tanaka" illustrated in the present embodiment is composed of only linear components, and each linear component is oriented only in the horizontal direction or the vertical direction. Therefore, most of the linear components of the character string extracted by executing the OCR function are considered to be orthogonal because their orientations are horizontal or vertical. Therefore, in the present embodiment, the inclination of the imprint is corrected by paying attention to this point so that the character string can be accurately extracted from the image of the imprint after the correction.

In the above description, the number of linear components is used as the number of linear components, and the order of the tilt angles is determined using the total result of the number of linear components. However, instead of the number of linear components, the linear components for each tilt angle are determined. You may use the sum of the lengths of. Alternatively, the tilt angle may be ranked by multiplying each of the number of linear components and the sum of the lengths by a coefficient such as a weight.

As described above, the determination unit 17 corrects the inclination of the image of the imprint as necessary to determine the validity of the imprint. When the same processing is performed on the keys that have not been subjected to the processing described above (N in step 113, steps 105 to 112), and the processing is performed on all the keys acquired by the key acquisition unit 14 (step 113). Y) in 113, the determination result output unit 18 presents the key and the determination result of the validity of the imprint associated with the key by the determination unit 17 (step 114). Specifically, the determination result is output using any of the methods illustrated below.

First, the determination result is included in the file name of the document. A naming convention for a document file is defined in the output format information stored in the output format information storage unit 24. For example, a naming convention such as "original file name + key + judgment result" that includes the key and the judgment result of the validity of the imprint that should exist associated with the key in the file name as a set is output format information. If the original file name is "ABC", the key is "Approver", and the imprint is valid, the document file is named "ABC_Approver_valid" according to the naming convention. On the other hand, when the original file name is "ABC", the key is "drafting", and the imprint is improper, the document file is named "ABC_drafting_illegal". Note that "_" is a delimiter that separates each item value, but the delimiter does not have to be limited to this. Also, the delimiter does not necessarily have to be included in the file name. If there are multiple keys, the file name is a plurality of pairs of "ABC_approver_legitimate_drafting_illegal_..." and the result of determining the validity of the imprint associated with the key. Include in.

In this way, if the determination result is included in the file name of the document, the determination result of the validity of the imprint associated with each key can be confirmed without opening the document and referring to the contents of the document.

Second, a file containing the determination result is generated. The output format information stored in the output format information storage unit 24 is a file containing a set of a key and a judgment result of the validity of the imprint existing associated with the key (hereinafter, referred to as a "judgment result file"). Is defined to generate. As described above, when the pair of the key and the judgment result is included in the file name, the file name may become very long if the number of keys is large. Therefore, by adopting this second output format, it is possible to prevent the file name from becoming too long. In order to check the judgment result, it may be necessary to open the judgment result file. However, if the judgment result file includes the judgment results for a plurality of documents, only one judgment result file needs to be opened when confirming the judgment results for a plurality of documents, and it is not necessary to open the plurality of document files individually. Absent. In addition, it is convenient for managing the validity of the imprint if the judgment result file is generated separately from the document file. The determination result file is created, for example, as a CSV (Comma Separated Values) file. When generating a judgment result file common to a plurality of documents, it is preferable to associate the document name with the key and the judgment result and register the judgment result file in the judgment result file.

The third output format is almost the same as the second output format. However, in the second output format, "legitimate" or "illegal" is included as the determination result, whereas in the third output format, the imprint image itself is further included in the file. In this way, not only the validity of the imprint but also the imprint itself can be confirmed.

As described above, in the present embodiment, the type of the document is specified from the scanned image of the document, and the key corresponding to the specified type of the document is acquired to extract the imprint existing in the vicinity of the key. I made it possible. Then, in the present embodiment, the characters included in the correct seal are acquired as the reference characters, the characters obtained from the imprint extracted from the scanned image of the document are acquired as the verification characters, and the reference characters and the verification characters match. The validity of the imprint was judged depending on whether or not it was used, and the judgment result could be presented. As a result, the validity of the imprint can be confirmed without opening the document. As described above, in the present embodiment, if the type of the document can be specified and the key associated with the imprint can be specified, the imprint can be extracted from the scanned image of the document. It becomes possible to adapt. That is, since it is not affected by the form of the document, it is possible to obtain the judgment result of the validity of the imprint without individually dealing with various types of document forms.

It can be assumed that the document has a plurality of character strings that are the same as the key. In this case, a condition for specifying a key associated with the imprint from a plurality of character strings may be set in advance so that the key can be automatically selected, or the user may be allowed to select the key.

Further, in the present embodiment, since the scanned image of the document is used, the image forming apparatus 10 has been described as an example of the information processing device, but a computer such as a general-purpose PC that receives and processes the scanned image of the document may also be used. Good.

In the above embodiment, the processor refers to a processor in a broad sense, and refers to a general-purpose processor (for example, CPU: Central Processing Unit, etc.) or a dedicated processor (for example, GPU: Graphics Processing Unit, ASIC: Application Special Integrated Circuit, FPGA). : Field Programgable Gate Array, programmable logic device, etc.).

Further, the operation of the processor in the above embodiment may be performed not only by one processor but also by a plurality of processors existing at physically separated positions in cooperation with each other. Further, the order of each operation of the processor is not limited to the order described in the above-described embodiment, and may be changed as appropriate.

1 network, 10 image forming device, 11 scan data acquisition unit, 12 character recognition processing unit, 13 document type identification unit, 14 key acquisition unit, 15 image extraction unit, 16 reference character acquisition unit, 17 judgment unit, 18 judgment result output Unit, 21 Document type information storage unit, 22 Key information storage unit, 23 Search condition information storage unit, 24 Output format information storage unit, 31 CPU, 32 Address data bus, 33 Operation panel, 34 Scanner, 35 Hard disk drive (HDD) , 36 Printer engine, 37 Network interface (I / F), 38 RAM, 39 ROM, 40 External media interface (I / F), 171 Verification character acquisition unit.

Claims (12)

Equipped with a processor
The processor
An image of the imprint existing at a specific imprinted portion in the document is extracted from the scanned image of the document to be processed.
The characters included in the image of the imprint are acquired as verification characters, and the characters are obtained.
The characters included in the seal to be stamped on the specific stamped part are acquired as reference characters.
The validity of the imprint existing at the specific stamped portion is determined by comparing the verification character with the reference character.
The result of determining the validity of the imprint existing at the specific imprint location is presented.
An information processing device characterized by this. When the processor cannot acquire the characters contained in the image of the imprint by using the character recognition function, the processor
The linear components included in the image of the imprint are extracted, and the inclination angle of each extracted linear component with respect to the orientation of the seal to be stamped on the document is detected.
The tilt angle of the imprint image is estimated by referring to the result of totaling the quantity of the extracted linear components for each tilt angle.
It is characterized in that the characters included in the image of the imprint are acquired by executing the character recognition function after correcting the inclination of the image of the imprint.
The information processing device according to claim 1. The processor
When there is a set of orthogonal tilt angles in the detected tilt angles, the totalized linear components included in the set of tilt angles that maximizes the sum of the aggregated results in the set of tilt angles. It is characterized in that the tilt angle of the linear component having a large result is estimated as the tilt angle of the image of the imprint.
The information processing device according to claim 2. The processor
When there is no set of orthogonal tilt angles among the detected tilt angles, the tilt angle of the linear component that maximizes the aggregated result is estimated as the tilt angle of the imprint image.
The information processing device according to claim 2. The result of totaling the quantity of the linear components for each tilt angle is the sum of the number of linear components of the tilt angle or the lengths of the linear components of the tilt angle.
The information processing device according to claim 2. The processor
A character recognition function is executed on four types of images obtained by rotating the image of the imprint after tilt correction based on the estimated tilt angle by 90 degrees.
The image having the maximum character recognition accuracy is extracted from the four types of images as an image of the imprint that stands upright with the seal.
The information processing device according to claim 2. The processor
By analyzing the scanned image of the document to be processed, the type of the document to be processed is specified.
Specific characters associated with a specific image that should exist in the document can be specified to exist in the document in the type of the document to be processed by referring to the specific character information set for each document type. Get the specific characters associated with the image of
An image that matches the acquired search conditions set in advance corresponding to the specific character is extracted as an image of the imprint.
The information processing device according to claim 1. The processor
It is characterized in that the acquired specific character and the determination result of the validity of the imprint associated with the specific character are presented as a set.
The information processing device according to claim 7. The processor
It is characterized in that the acquired specific character and the determination result of the validity of the imprint associated with the specific character are combined and included in the file name of the document to be processed.
The information processing device according to claim 8. The processor
It is characterized in that a file including a set of the acquired specific character and the determination result of the validity of the imprint associated with the specific character is generated.
The information processing device according to claim 8. Equipped with a processor
The processor
An image of the imprint existing at a specific imprinted portion in the document is extracted from the scanned image of the document to be processed.
When the character contained in the image of the imprint cannot be acquired by using the character recognition function, the linear component contained in the image of the imprint is extracted and the direction of the seal to be stamped on the document is extracted. Detects the tilt angle of each linear component
The tilt angle of the imprint image is estimated by referring to the result of totaling the quantity of the extracted linear components for each tilt angle.
By executing the character recognition function after correcting the inclination of the imprint image, the characters included in the imprint image are acquired.
An information processing device characterized by this. On the computer
A function of extracting an image of an imprint existing at a specific imprinted portion in the document from the scanned image of the document to be processed.
When the character contained in the image of the imprint cannot be acquired by using the character recognition function, the linear component contained in the image of the imprint is extracted and the direction of the seal to be stamped on the document is extracted. A function to detect the tilt angle of each linear component,
A function to estimate the tilt angle of the imprint image by referring to the result of totaling the number of extracted linear components for each tilt angle.
A function of acquiring characters included in the imprint image by executing a character recognition function after correcting the inclination of the imprint image.
A program to realize.

Images