Document reading device and document reading method

Abstract

To easily secondarily utilize a document by efficiently and accurately reading and classifying the document.SOLUTION: In an image forming apparatus 10, an abnormality determination section 32 determines the presence / absence of abnormality in an original based on an image of the original which is read by an image reading section 11 and in a case where the absence of abnormality in the original is determined, a document classification section 33 generated attribute data of a document of the original based on the image of the original. In a case where the absence of abnormality in the original is determined, an image correction section 34 corrects the image of the original and improves recognizability of characters included in the corrected image of the original, and an output section outputs in an associated manner the image of the original which is read by the image reading section 11, the attribute data of the document of the original which are generated by the document classification section 33, and the image of the original which is corrected by the image correction section 34.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a document reading device and a document reading method for reading and storing a document contained in an image of a document, and more particularly to a technique for facilitating secondary use of the image of the read document. [Background technology]

[0002] In recent years, with the promotion of DX (Digital Transformation), there are more and more opportunities to digitize documents that were previously managed and operated on paper using image reading devices such as scanners. When reading and storing documents contained in original images, it is desirable that the reading and classification of documents be performed efficiently and accurately, and that secondary use of documents be easy.

[0003] In the document management system described in Patent Document 1, the document form and the corresponding document type are registered in advance, the document form is identified from the input document image, the document type corresponding to this form is selected, and this document type is associated with the input document image and stored. Also, if the corresponding document form is not registered, the form is newly registered. By adding the document type to the input document image, the work of adding the document type by the operator is reduced.

[0004] In addition, the document classification device described in Patent Document 2 analyzes a document image to obtain layout information that represents the layout of the components that make up each page of the document, extracts a text area in which text is continuous within the page based on this layout information, recognizes character strings contained in the text area, extracts visually emphasized character strings from the recognized character strings, and sets the extracted character strings as keywords. Then, for each page of the document, structure data that represents a hierarchical structure on the layout of the text area is generated, and the logical structure of the document is extracted using this structure data and keywords, and the document is classified and stored using the extracted logical structure. This allows documents that are not in a strict standard form to be properly classified. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2002-269126 A [Patent Document 2] JP 2005-122550 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, in Patent Document 1, when the document forms are different even for the same document type, the number of registered document types increases, making secondary use of the documents difficult. Also, in Patent Document 2, visually emphasized character strings are extracted as keywords, but such keywords are not necessarily extracted.

[0007] The present invention has been made in consideration of the above circumstances, and has an object to read and classify documents efficiently and accurately, and to make the documents easily available for secondary use. [Means for solving the problem]

[0008] A document reading device according to one aspect of the present invention comprises an image reading unit which reads an image of a document, an abnormality judgment unit which judges whether or not there is an abnormality in the document based on the image of the document read by the image reading unit, a document classification unit which generates attribute data indicating the type of document contained in the image of the document when the abnormality judgment unit judges that there is no abnormality in the document, an image correction unit which corrects the image of the document to improve the recognizability of characters contained in the image of the document when the abnormality judgment unit determines that there is no abnormality in the document, and an output unit which outputs the image of the document read by the image reading unit, the document attribute data of the document generated by the document classification unit, and the image of the document corrected by the image correction unit in association with each other.

[0009] Moreover, a document reading method according to one aspect of the present invention includes an image reading step of reading an image of a document, an abnormality determination step of determining whether or not there is an abnormality in the document based on the image of the document read in the image reading step, a document classification step of generating document attribute data contained in the image of the document if it is determined in the abnormality detection step that there is no abnormality in the document, an image correction step of correcting the image of the document to improve the recognizability of characters contained in the corrected image of the document if it is determined in the abnormality determination step that there is no abnormality in the document, and an output step of outputting the image of the document read in the image reading step, the document attribute data of the document generated in the document classification step, and the image of the document corrected in the correction step in association with each other. Effect of the Invention

[0010] According to the present invention, documents can be efficiently and accurately read and classified, and the documents can be easily reused. [Brief description of the drawings]

[0011] [Figure 1] 1 is a cross-sectional view showing an image forming apparatus which is an embodiment of a document reading apparatus according to the present invention. [Diagram 2] 1 is a block diagram showing a main internal configuration of an image forming apparatus according to an embodiment of the present invention; [Diagram 3] 4 is a diagram illustrating a first embodiment of a processing procedure in a document reading mode by the image forming apparatus of the present embodiment. FIG. [Figure 4] 11 is a flowchart showing a first embodiment of a process for efficiently and accurately reading and classifying documents in a document reading mode, and facilitating secondary use of the documents. [Diagram 5] 13 is a diagram showing an initial screen displayed on a display unit. [Figure 6] FIG. 1A is a diagram showing a fold in a document, and FIG. 1B is a diagram showing the document after the fold has been corrected. [Figure 7]FIG. 1A is a diagram showing a document with a sticky note attached thereto, and FIG. 1B is a diagram showing the document with the sticky note removed. [Figure 8] FIG. 13 is a diagram showing a message displayed on a display unit. [Figure 9] FIG. 1A is a diagram showing straight lines overlapping characters in an original image, and FIG. 1B is a diagram showing the original image from which the straight lines have been erased. [Figure 10] FIG. 1A is a diagram showing the background of characters in an image of an original document, and FIG. 1B is a diagram showing a state in which the density of the background has been reduced. [Figure 11] 11 is a diagram illustrating a second embodiment of a processing procedure in a document reading mode by the image forming apparatus according to the present embodiment. FIG. [Figure 12] 13 is a flowchart showing a second embodiment of the process for efficiently and accurately reading and classifying documents in a document reading mode, and facilitating secondary use of the documents. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] A document reading device and a document reading method according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a cross-sectional view showing an image forming device which is an embodiment of the document reading device according to the present invention. As shown in FIG. 1, the image forming device 10 of this embodiment is an MFP (multi-function peripheral) having multiple functions such as a copy function and a scanner function. The image forming device 10 includes an image reading unit 11 and an image forming unit 12.

[0013] The image reading unit 11 has an image pickup element that optically reads an image of a document, and the output of the image pickup element is converted into image data that represents the image of the document.

[0014] The image forming section 12 forms an image represented by the image data on a recording sheet, and includes an image forming unit 3M for magenta, an image forming unit 3C for cyan, an image forming unit 3Y for yellow, and an image forming unit 3Bk for black. In each of the image forming units 3M, 3C, 3Y, and 3Bk, the surface of the photosensitive drum 4 is uniformly charged, the surface of the photosensitive drum 4 is exposed to light, an electrostatic latent image is formed on the surface of the photosensitive drum 4, the electrostatic latent image on the surface of the photosensitive drum 4 is developed into a toner image, and the toner image on the surface of the photosensitive drum 4 is primarily transferred to the intermediate transfer belt 5. As a result, a color toner image is formed on the intermediate transfer belt 5. This color toner image is secondarily transferred to the recording sheet P conveyed from the paper feed section 14 through the conveying path 8 in the nip area N between the intermediate transfer belt 5 and the secondary transfer roller 6.

[0015] Thereafter, the recording paper P is heated and pressed in the fixing device 15 so that the toner image on the recording paper P is fixed by thermocompression, and the recording paper P is then discharged onto a discharge tray 17 via discharge rollers 16 .

[0016] Fig. 2 is a block diagram showing the main internal configuration of the image forming apparatus 10 of this embodiment. As shown in Fig. 2, the image forming apparatus 10 of this embodiment includes an image reading unit 11, an image forming unit 12, a display unit 21, an operation unit 22, a touch panel 23, a network communication unit (NW communication unit) 24, an image memory 26, a storage unit 28, and a control unit 29. These components are capable of transmitting and receiving data or signals to and from each other via a bus.

[0017] The display unit 21 is configured with a liquid crystal display (LCD) or an organic light-emitting diode (OLED) display, etc. The operation unit 22 includes physical keys such as a numeric keypad, a decision key, and a start key.

[0018] A touch panel 23 is disposed on the screen of the display section 21. The touch panel 23 is a so-called resistive film type or capacitive type touch panel, which detects a contact (touch) made by a user's finger or the like on the touch panel 23 together with the contact position, and outputs a detection signal indicating the coordinates of the contact position to a control section 31 (described later) of the control unit 29, etc.

[0019] The network communication unit (NW communication unit) 24 is a communication interface including a communication module such as a LAN chip (not shown), and is connected to an external device via a wired or wireless LAN to transmit and receive data to and from the external device.

[0020] In the image memory 26, image data indicating the image of the document read by the image reading unit 11 is stored.

[0021] The storage unit 28 is a large-capacity storage device such as a solid state drive (SSD) or a hard disk drive (HDD), and stores various application programs and various data.

[0022] The control unit 29 is composed of a processor, a RAM (Random Access Memory), a ROM (Read Only Memory), etc. The processor is, for example, a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or an MPU (Micro Processing Unit), etc. The control unit 29 functions as a control unit 31, an abnormality determination unit 32, a document classification unit 33, an image correction unit 34, and an output unit 35 by the processor executing a control program stored in the ROM or the storage unit 28.

[0023] The control unit 31 comprehensively controls the image forming apparatus 10. The control unit 29 is connected to the image reading unit 11, the image forming unit 12, the display unit 21, the operation unit 22, the touch panel 23, the network communication unit 24, the image memory 26, and the storage unit 28. The control unit 31 controls the operation of these components and transmits and receives signals or data between these components.

[0024] The control unit 31 serves as a processing unit that executes various processes necessary for image formation by the image forming apparatus 10. The control unit 31 also accepts operation instructions input by a user based on a detection signal output from the touch panel 23 or an operation of a physical key of the operation unit 22. Furthermore, the control unit 31 has a function of controlling the display operation of the display unit 21 and a function of controlling the communication operation of the network communication unit 24.

[0025] The abnormality determination unit 32 determines whether or not there is an abnormality in the document based on the image of the document read by the image reading unit 11.

[0026] If it is determined that there is no abnormality in the document, the document classification unit 33 generates attribute data of the document included in the image of the document based on the image of the document.

[0027] When it is determined that there is no abnormality in the document, the image correction unit 34 corrects the image of the document to improve the recognizability of characters included in the corrected image of the document.

[0028] The output unit 35 outputs the image of the original document read by the image reading unit 11, the image of the original document corrected by the image correcting unit 34, and the document attribute data of the original document generated by the document classification unit 33 in association with each other.

[0029] In image forming apparatus 10 configured as above, for example, when a user places an original on image reading unit 11 and operates operation unit 22 or touch panel 23 to instruct execution of a copy, control unit 31, in accordance with the instruction, causes image reading unit 11 to read the image of the original and stores image data representing the image of the original in image memory 26. Control unit 31 further causes image forming unit 12 to form an image of the original represented by the image data stored in image memory 26 on recording paper.

[0030] In addition, the image forming apparatus 10 of this embodiment operates in a document reading mode in which the functions of the abnormality determination unit 32, document classification unit 33, image correction unit 34, and output unit 35 are used to efficiently and accurately read and classify documents contained in the image of the original document, making it easy to reuse the documents.

[0031] Next, a first embodiment of the processing procedure in the document reading mode will be described with reference to Fig. 3, which is a diagram showing an outline of the first embodiment of the processing procedure in the document reading mode.

[0032] The image reading unit 11 reads an image of a document. The abnormality determination unit 32 acquires image data showing the image of the document read by the image reading unit 11, and determines whether or not there is an abnormality in the document based on the image data. The abnormality in the document may be whether or not the document is folded or whether or not there is a sticky note attached to the document. The abnormality determination unit 32 learns images of various folded documents or images of various documents with sticky notes attached by deep learning using, for example, a CNN (Convolution Neural Network), creates an inference model for determining whether or not the document is folded or the document with a sticky note attached, and determines whether or not the document is folded or the document with a sticky note attached to the document by using this inference model. However, it is not intended to limit the abnormality determination by the abnormality determination unit 32 to the determination method.

[0033] When the anomaly determination unit 32 determines that the document is normal, the document classification unit 33 acquires image data representing the document image, and generates attribute data representing the type of document included in the document image based on the image data. In other words, the attribute data represents the type of document included in the document image.

[0034] The document types are invoices, purchase orders, contracts, receipts, etc. The document classification unit 33 learns images showing various invoices, purchase orders, contracts, and receipts, for example, by deep learning using CNN, creates an inference model for determining the document type, and determines the document type by this inference model. In this case, the task of "determining the document type" is regarded as the task of "classifying document images by form," and an inference model (algorithm) that can determine the document type is created and used by deep learning using an image classification network.

[0035] CNN is a type of multi-layer neural network that contains layers specialized for image processing. Various CNN networks have been devised depending on the purpose, and there are networks that enable processing such as image classification, which identifies the type of image or object in an image and classifies the image into multiple classes, and semantic segmentation, which extracts specific regions from an image on a pixel-by-pixel basis.

[0036] Alternatively, the document classification unit 33 may identify the characteristics of a predefined form by known character recognition, ruled line recognition, etc. as in Patent Document 1 (JP Patent Publication No. 2002-269126), or may extract the logical structure of a document and classify the document using the extracted logical structure as in Patent Document 2 (JP Patent Publication No. 2005-122550). However, this is not intended to limit the method of determining the document type by the document classification unit 33.

[0037] Furthermore, when the abnormality determination unit 32 determines that there is no abnormality in the document, the image correction unit 34 inputs image data showing the image of the document, corrects the document image based on this image data, and improves the recognizability of characters in the document included in the corrected document image. This correction of the document image includes, for example, erasing line segments that overlap characters in the document image and reducing the density of the background of the characters in the document image.

[0038] The image correction unit 34 detects straight lines and the like in the image of the document using a Hough transform based on image data showing the image of the document, and for each intersection between the detected straight lines and the lines that depict each character, the image correction unit 34 maintains the density of each pixel at the intersection and erases the straight lines. As a result, the lines that depict each character are left as they are, and only the straight lines that overlap each character are erased, making the image of each character clearer. The image correction unit 34 also compares the density of each pixel that constitutes the image of the document with a preset density threshold, and does not change the density of pixels that constitute each character that are equal to or greater than the density threshold, and reduces the density of pixels that constitute the background of each character that are less than the density threshold. As a result, the density of the background of each character is reduced, making the image of each character clearer. As a result, the recognizability of characters included in the image of the document is improved, and character recognition by, for example, a known OCR (Optical Character Recognition) function is more accurately performed.

[0039] The output unit 35 stores image data showing the image of the original document read by the image reading unit 11, document attribute data of the original document generated by the document classification unit 33, and image data showing the image of the original document corrected by the image correction unit 34 together in a single file, associates these pieces of data with each other, and stores this file in the memory unit 28 or transmits it to an external memory device via the network communication unit 24 for storage.

[0040] Next, a first embodiment of a process for efficiently and accurately reading and classifying documents using the document reading mode described above and making the documents easily available for secondary use will be described in detail with reference to the flowchart shown in FIG.

[0041] The control unit 31 causes the display unit 21 to display an initial screen G1 as shown in Fig. 5, for example. On this initial screen G1, a plurality of function keys 41a to 41h, each associated with a different function, are displayed. At this time, when the user touches the function key 41h indicating the document reading mode, the control unit 31 receives an instruction corresponding to the touch operation on the function key 41h through the touch panel 23, and sets the image forming apparatus 10 to the document reading mode in accordance with this instruction (S101). After this setting, the control unit 31 causes the image forming apparatus 10 to operate in the document reading mode (S101).

[0042] Next, the user places one or more documents on which text is written in the image reading unit 11, and operates the start key of the operation unit 22. The image of the document may include figures, photographs, etc. in addition to the text.

[0043] The control unit 31 receives an instruction to read a document in response to the operation of the start key, and starts reading the images of one or more documents by the image reading unit 11 in accordance with the instruction to read a document. The images of one or more documents are successively read by the image reading unit 11, and each time an image of a document is read, the control unit 31 stores image data representing the image of the document in the image memory 26 (S102).

[0044] The abnormality determination unit 32 reads image data showing the image of the first document from the image memory 26 (S103), and determines whether or not there is an abnormality in the document based on this image data (S104). The abnormality in the document refers to whether or not there is a fold in the document M as shown in Fig. 6(A) or whether or not there is a sticky note F attached to the document M as shown in Fig. 7(A), etc.

[0045] When abnormality determination section 32 determines that an original document has an abnormality (S104 "Yes"), control section 31 determines and stores the reading order (page order) of the original document by image reading section 11 (S105), erases image data representing the image of the original document from image memory 26 (S106), and determines whether image data representing the image of the next original document is stored in image memory 26 (S107). When control section 31 determines that image data representing the image of the next original document is stored (S107 "Yes"), it reads image data representing the image of the next original document from image memory 26 (S108), returns to S104, and determines whether or not there is an abnormality in the next original document based on the read image data (S104).

[0046] Furthermore, when the abnormality determination unit 32 determines that there is no abnormality in the original document (S104 "No"), the control unit 31 determines whether or not image data representing the image of the next original document is stored in the image memory 26 (S107) without erasing the image data representing the image of this original document from the image memory 26, and if it determines that image data representing the image of the next original document is stored (S107 "Yes"), it reads out the image data representing the image of the next original document from the image memory 26 (S108), returns to S104, and determines whether or not there is an abnormality in the next original document based on this image data (S104).

[0047] If the control unit 31 determines that image data indicating the image of the next document is not stored (S107 "No"), it ends the determination of whether or not there are any abnormalities in the image data indicating the images of all documents containing documents containing text.

[0048] Then, the control unit 31 judges whether or not to read the image of the document again based on the judgment result of the presence or absence of an abnormality in the document in S103 (S109). For example, if the document is judged to have an abnormality in S104, the reading order (page order) of the document by the image reading unit 11 is stored in S105, and the image data showing the image of the document is erased from the image memory 26 in S106, so the control unit 31 judges that at least one document judged to have an abnormality needs to be read again (S109 "Yes"). In this case, the control unit 31 displays a message E1 on the screen of the display unit 21 as shown in FIG. 8, prompting the user to correct the fold of the document in the page order judged in S105 or to remove the sticky note attached to the document and to read the document again by the image reading unit 11 (S110).

[0049] The user looks at the message E1, finds the original M in that page order, and either corrects the fold in the original M as shown in FIG. 6(B) or removes the sticky note F affixed to the original M as shown in FIG. 7(B), and then sets the original again in the image reading unit 11 and operates the start key on the operation unit 22.

[0050] The control unit 31 accepts an instruction to read a document in response to the operation of the start key, and in accordance with the instruction to read a document, causes the image reading unit 11 to read the image of the document in the page order (the document that has been set again), and stores image data representing the image of the document in the page order in the image memory 26 (S111). At this time, the control unit 31 rearranges and stores the image data representing the image of the document in the page order together with image data representing the image of another document already stored in the image memory 26, according to the page order of all the documents constituting the document.

[0051] In addition, if the control unit 31 does not accept the document reading instruction and determines not to read the document image again based on the result of determining whether or not there is an abnormality in the document in S103 (S109 ``No''), the process proceeds to S112 without performing S110 and S111.

[0052] As a result, all of the documents constituting the document are found to be normal, and image data showing the images of all of these documents are stored in the image memory 26 in page order of the documents.

[0053] The document classification unit 33 acquires image data representing the images of each document constituting the document from the image memory 26, and generates document attribute data included in each document image based on the image data (S112). The attribute data indicates the type of the document.

[0054] Furthermore, the image correction unit 34 acquires image data showing the image of the document, corrects the image of the document, and improves the recognizability of characters included in the corrected image of the document (S113). Correcting the image of the document means, for example, erasing lines that overlap characters included in the image of the document, and reducing the density of the background of the characters included in the image of the document. For example, the image correction unit 34 erases the straight lines 52 that overlap characters in the image of the document M shown in Fig. 9(A) as shown in Fig. 9(B). Furthermore, the image correction unit 34 reduces the density N of the background of the characters shown in Fig. 10(A) as shown in Fig. 10(B).

[0055] The output unit 35 obtains image data indicating the image of the original document read in S102 from the image memory 26, obtains attribute data indicating the document type of the original document generated in S112 from the document classification unit 33, and obtains image data indicating the image of the original document corrected in S113 from the image correction unit 34. Then, the output unit 35 performs output processing (S114) in which these pieces of data are stored together in one file, these pieces of data are associated with each other, and the file is stored in the storage unit 28 or transmitted via the network communication unit 24 to an external storage device or a storage device on the network for storage therein.

[0056] The output unit 35 may also store each type of document in association with the output destination of each of the files in advance. In this case, the output unit 35 selects an output destination of the file associated with the type of document indicated by the attribute data, and outputs the file to the selected output destination. The output destination of the file may be the storage unit 28 as described above, an external storage device, or a storage on a network (storage of a server, provider, etc.).

[0057] In this embodiment, the image of the original is read by the image reading unit 11, and the presence or absence of abnormalities in the read original is determined. If it is determined that there are no abnormalities in the original, the image of the original is corrected to improve the recognizability of characters contained in the original image. Also, attribute data of the document contained in the original image is generated based on the original image, and the image of the original read by the image reading unit 11, the corrected original image, and the attribute data of the original document are output in association with each other.

[0058] Furthermore, if an abnormality is determined to exist in the scanned document, the user is notified of the abnormality and given an opportunity to correct the abnormality in the document before the image of the document is scanned again. This prevents image data showing an image of a document without an abnormality from being mixed with image data showing an image of a document with an abnormality, and results in only image data showing an image of a document without an abnormality. This mixing does not adversely affect document management of the document, and improves the quality and reliability of the image data showing the image of the document.

[0059] In addition, the recognition of characters contained in the image of the original document is improved, and character recognition using well-known OCR functions can be performed more accurately.

[0060] In addition, the document attribute data of the original indicates the type of document included in the image of the original, which makes it possible to efficiently and accurately search for the image of the original based on the document attribute data, facilitating secondary use of the document included in the image of the original.

[0061] In the above embodiment, the output unit 35 stores the image data indicating the image of the document read in S102, the attribute data indicating the document type of the document generated in S112, and the image data indicating the image of the document corrected in S113 together in one file and associates these data with each other, but the attribute data indicating the document type of the document may be added or embedded as incidental information in the image data indicating the image of the document read in S102 and the image data indicating the image of the document corrected in S113. In this case, too, it is possible to efficiently and accurately search for images of the document using the attribute data.

[0062] Next, a second embodiment of the processing procedure of the document reading mode and a second embodiment of the processing for efficiently and accurately reading and classifying documents in the document reading mode and easily allowing secondary use of the documents will be described. FIG. 11 is a diagram showing an outline of the second embodiment of the processing procedure of the document reading mode. FIG. 12 is a flowchart showing a second embodiment of the processing for efficiently and accurately reading and classifying documents in the document reading mode and easily allowing secondary use of the documents. Note that in the second embodiment, only the processing different from the first embodiment will be described. In other words, the contents of the processing that are not described in the second embodiment are the same as those described in the first embodiment.

[0063] In the above-described first embodiment, the document classification unit 33 generates document attribute data contained in each document image based on image data representing each document image that constitutes the document (S112), and then the image correction unit 34 corrects the document image based on the image data representing the document image, thereby performing a process to improve the recognizability of characters contained in the corrected document image (S113).

[0064] In contrast, in the second embodiment, as shown in Figures 11 and 12, the image correction unit 34 performs a process of correcting the document image based on image data representing the document image to improve the recognizability of characters contained in the corrected document image (S113), and then the document classification unit 33 performs a process of generating document attribute data contained in each document image based on the document image corrected by the image correction unit 34 (S112).

[0065] In this case, as shown in FIG. 11, the output unit 35 stores image data showing the image of the original document read by the image reading unit 11, document attribute data of the original document generated by the document classification unit 33 based on the image of the original document with improved character recognizability, and image data showing the image of the original document corrected by the image correction unit 34 together in a single file, associates these pieces of data with each other, and stores this file in the memory unit 28 or transmits it to an external memory device via the network communication unit 24 for storage.

[0066] As a result, in the second embodiment, document attribute data is generated based on an original image with improved character recognition, so when the document classification unit 33 performs the process of determining the document type described above, a more accurate type determination can be made, and the accuracy of the attribute data generated for the document can be improved. Therefore, according to the second embodiment, a more accurate document classification result can be obtained, making it even easier to perform secondary use, such as extracting information from such digitized document images.

[0067] In addition, in the above embodiment, one embodiment of the document reading device of the present invention is described as an image forming device (multifunction device), but this is merely one example, and the document reading device of the present invention may be a system that combines a scanner that reads images of a document and an information processing device such as a personal computer.

[0068] Moreover, the configurations and processes of the above-described embodiment described with reference to FIGS. 1 to 12 are merely examples of the present invention, and the present invention is not limited to these configurations and processes. [Explanation of symbols]

[0069] 10 Image forming device 11 Image reading unit 12 Image forming section 21 Display section 22 Control section 23 Touch Panel 24 Network Communication Department 26 Image Memory 28 Memory section 29 Control Unit 31 Control Unit 32 Abnormality determination section 33 Document Classification Department 34 Image correction section 35 Output section

Claims

1. an image reading unit that reads an image of a document; an abnormality determination unit that determines whether or not there is an abnormality in the document based on the image of the document read by the image reading unit; a document classification unit that generates attribute data indicating a type of document included in an image of the document when the abnormality determination unit determines that the document is normal; an image correction unit that performs correction on the image of the document to improve recognizability of characters included in the image of the document when the abnormality determination unit determines that there is no abnormality in the document; and an output unit that outputs the image of the original document read by the image reading unit, document attribute data of the original document generated by the document classification unit, and the image of the original document corrected by the image correction unit in association with each other.

2. The document reading device according to claim 1 , wherein the document classification unit generates attribute data indicating the type of document contained in the image of the document based on the image of the document when the abnormality determination unit determines that there is no abnormality in the document.

3. The document reading device according to claim 1, wherein when the anomaly determination unit determines that the document is free of anomalies, the document classification unit generates attribute data indicating the type of document contained in the image of the document based on the image of the document corrected by the image correction unit.

4. 4. A document reading device as described in any one of claims 1 to 3, wherein the output unit selects one of a plurality of output destinations based on attribute data of the original document generated by the document classification unit, and outputs to the selected output destination the image of the original document read by the image reading unit, the attribute data of the original document generated by the document classification unit, and the image of the original document corrected by the correction unit in association with each other.

5. 4. A document reading device according to claim 1, wherein the abnormality determination unit detects whether the document is folded or has a sticky note attached thereto based on the image of the document read by the image reading unit, and determines whether or not there is an abnormality in the document based on the detection result.

6. 4. The document reading apparatus according to claim 1, wherein the document classification section generates attribute data of the document, which indicates a type of the document included in the image of the document, based on the image of the document.

7. 4. The document reading device according to claim 1, wherein the image correction section erases line segments overlapping characters included in the document image, or reduces a density of a background of characters included in the document image.

8. 4. The document reading device according to claim 1, wherein the image reading section reads the image of the document again when the abnormality determination section determines that the document has an abnormality.

9. an image reading step of reading an image of a document; an abnormality determination step of determining whether or not there is an abnormality in the document based on the image of the document read in the image reading step; a document classification step of generating attribute data of a document included in an image of the document when the document is determined to have no abnormality in the abnormality detection step; an image correction step of correcting an image of the document to improve recognizability of characters included in the corrected image of the document when the document is determined to have no abnormality in the abnormality determination step; and an output step of outputting the image of the original document read in the image reading step, the document attribute data of the original document generated in the document classification step, and the image of the original document corrected in the correction step in association with each other.

Images