Image processing device, image processing method, and image processing program

Abstract

To provide an image processing device, an image processing method and an image processing program that correctly set a recognition region even when characters included in an image are handwritten.SOLUTION: An image processing function of an operation part 30 of MFP comprises: a character determination part 303 which determines whether characters entered into a recognition region where characters entered into an image are to be recognized are handwritten or printed; a first correction part 305 which varies the size of the recognition region when the character determination part 303 determines that the characters entered into the recognition region are handwritten; and a second correction part 308 which moves the position of the recognition region when the character determination part 303 determines that the characters entered into the recognition region are printed.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to an image processing device, an image processing method, and an image processing program. [Background technology]

[0002] Optical Character Recognition (OCR) is a technology that extracts areas of printed text from images of forms, documents, and other documents scanned by an MFP (Multifunction Peripheral / Printer / Product) or scanner, and recognizes the type of text. To accurately extract areas of printed text, this technology sometimes extracts distinctive areas, such as table borders, from the original document to set the recognition area where character recognition is performed, or sometimes the recognition area is set manually. However, in the case of original documents where only necessary information is added within pre-printed borders, characters may appear outside the pre-set recognition area due to misalignment of the text. Therefore, a technology is known that detects misalignment of the text in the scanned original document and corrects the recognition area accordingly.

[0003] As an example of such a technology, for example, Patent Document 1 discloses a method for reading print data in an image of a document containing frame lines that has been digitized by a scanner, which detects print-misaligned data that has been mixed in with areas where data is not printed, calculates the degree of overlap between the frame area and the circumscribed rectangle of the print-misaligned data, and determines which frame the print-misaligned data is protruding from by using the relationship between the position of the frame line separating the two frames, the position of the overlapping character, and the center position of the frame, the size of the circumscribed rectangle of the print-misaligned data, the size of the frame, and the overall direction of the print-misalignment. Summary of the Invention [Problem to be solved by the invention]

[0004] However, when the object of character recognition is a type character with a uniform deviation, it is possible to detect the deviation from a global perspective and associate the character with the recognition area. However, when the object of character recognition is a handwritten character with an inconsistent deviation, the recognition area cannot be set correctly, which makes it easy for misrecognition to occur.

[0005] The present application has been made in view of the above, and aims to correctly set a recognition area even when characters included in an image are handwritten. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems and achieve the object, the present invention is characterized by comprising a character determination unit that determines whether a character written in a recognition area, which is an area where recognition of characters written in an image is performed, is handwritten or printed, a first correction unit that changes the size of the recognition area when the determination unit determines that the character written in the recognition area is handwritten, and a second correction unit that moves the position of the recognition area when the determination unit determines that the character written in the recognition area is printed. [Effects of the Invention]

[0007] According to the present invention, it is possible to correctly set a recognition area even when characters included in an image are handwritten. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram showing an example of the configuration of an image processing system 1 according to an embodiment. [Figure 2] FIG. 2 is a diagram showing an example of the hardware configuration of the MFP 10 according to the embodiment. [Figure 3] FIG. 3 is a block diagram showing a module configuration related to the image processing function of the operation unit 30. As shown in FIG. [Figure 4] FIG. 4 is a flowchart showing the flow of image processing by the MFP 10. [Figure 5] FIG. 5 is a diagram showing an example of an image showing a document including handwritten characters. [Figure 6] FIG. 6 is a diagram showing an example of the determination made by the protrusion determination unit 304. In FIG. [Figure 7]FIG. 7 is a diagram showing an example of correction of the recognition area by the first correction unit 305. In FIG. [Figure 8] FIG. 8 is a diagram showing an example of the determination made by the additional write determination unit 306. In FIG. [Figure 9] FIG. 9 is a diagram showing an example of extraction of a rectangle of a character string printed outside the recognition area. [Figure 10] FIG. 10 is a diagram showing an example of calculation of the amount of deviation by the calculation unit 307. In FIG. [Figure 11] FIG. 11 is a diagram showing an example of correction of the recognition area by the second correction unit 308. In FIG. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, embodiments of an image processing device, an image processing method, and an image processing program will be described in detail with reference to the accompanying drawings. In the embodiments described below, the image processing device of the present invention will be described as a multifunction peripheral (MFP), but the present invention is not limited to this. Note that a multifunction peripheral is a device that has at least two of the following functions: printing, copying, scanning, and facsimile.

[0010] (Embodiment) First, an overview of an entire image processing system 1 including an image processing device (MFP) 10 according to an embodiment will be described with reference to FIG. 1. FIG. 1 is a diagram showing an example of the configuration of the image processing system 1 according to an embodiment. As shown in FIG. 1, the image processing system 1 includes the MFP 10, a PC (Personal Computer) 11, a server 12, and a storage device 13. The MFP 10, the PC (Personal Computer) 11, the server 12, and the storage device 13 are connected via a network 14. Examples of the network 14 include a LAN (Local Area Network) and the Internet. Note that the image processing system 1 shown in FIG. 1 may include a plurality of MFPs 10, a plurality of PCs (Personal Computers) 11, a plurality of servers 12, and a plurality of storage devices 13.

[0011] The MFP 10 reads an original document and performs character recognition processing on the image of the original document. If information about a recognition area, which is an area where character recognition processing is performed, is not set for the original document to be read, the MFP 10 extracts the recognition area.

[0012] The PC 11 is an information processing device used by a user. For example, the PC 11 is used by the user to check the recognition results of the recognition process performed by the MFP 10. The PC 11 is also used by the user to modify the recognition area.

[0013] The character recognition process and the setting of the recognition area may be performed by PC 11 or server 12 receiving from MFP 10 an image showing the document read by MFP 10. Furthermore, the setting of the recognition area may not be performed automatically by MFP 10 at the same time as the document is read, but may be performed by the user manually operating MFP 10 while viewing the image.

[0014] Furthermore, the recognition results and information on the recognition area are basically stored in the PC 11 or the server 12, but can also be stored in an external storage device 13.

[0015] Next, the configuration of the MFP 10 according to the embodiment will be described with reference to FIG. 2. FIG. 2 is a diagram showing an example of the hardware configuration of the MFP 10 according to the embodiment. As shown in FIG. 2, the MFP 10, which is a multifunction peripheral, includes a main body 20 capable of implementing various functions such as a copy function, a scanner function, a fax function, and a printer function, and an operation unit 30 that accepts user operations. Note that accepting user operations is a concept that includes accepting information (including signals indicating screen coordinate values, etc.) input in response to the user operations. The main body 20 and the operation unit 30 are connected to each other so as to be able to communicate with each other via a dedicated communication path 100. The communication path 100 may be, for example, one conforming to the USB (Universal Serial Bus) standard, but may be one conforming to any standard, whether wired or wireless.

[0016] The main body 20 can perform operations according to operations received by the operation unit 30. The main body 20 can also communicate with an external device such as a client PC (for example, PC 11), and can perform operations according to instructions received from the external device.

[0017] First, the hardware configuration of the main body 20 will be described. As shown in Fig. 2, the main body 20 includes a CPU 21, a ROM 22, a RAM 23, an HDD (hard disk drive) 24, a communication I / F (interface) 25, a connection I / F 26, and an engine unit 27. The CPU 21, ROM 22, RAM 23, HDD 24, communication I / F 25, connection I / F 26, and engine unit 27 are interconnected via a system bus 28.

[0018] The CPU 21 comprehensively controls the operation of the main body 20. The CPU 21 executes programs stored in the ROM 22, the HDD 24, or the like using the RAM 23 as a work area, thereby controlling the operation of the entire main body 20 and realizing various functions such as the copy function, scanner function, fax function, and printer function described above.

[0019] The communication I / F 25 is an interface for connecting to the network 14. The connection I / F 26 is an interface for communicating with the operation unit 30 via the communication path 100.

[0020] The engine unit 27 is hardware that performs general-purpose information processing and processing other than communication in order to realize the copy function, scanner function, fax function, and printer function. The engine unit 27 includes, for example, a scanner (image reading unit) that scans and reads images of documents, a plotter (image forming unit) that prints on sheet materials such as paper, and a fax unit that performs fax communication. Furthermore, the engine unit 27 may also include specific options such as a finisher that sorts printed sheet materials and an ADF (automatic document feeder) that automatically feeds documents.

[0021] Next, a description will be given of the hardware configuration of the operation unit 30. As shown in Fig. 2, the operation unit 30 includes a CPU 31, a ROM 32, a RAM 33, a flash memory 34, a communication I / F 35, a connection I / F 36, and an operation panel 37. The CPU 31, the ROM 32, the RAM 33, the flash memory 34, the communication I / F 35, the connection I / F 36, and the operation panel 37 are interconnected via a system bus 38.

[0022] The CPU 31 comprehensively controls the operation of the operation unit 30. The CPU 31 executes programs stored in the ROM 32, flash memory 34, or the like using the RAM 33 as a work area, thereby controlling the operation of the entire operation unit 30 and realizing various functions, which will be described later, such as displaying information (images) according to input received from the user.

[0023] The communication I / F 35 is an interface for connecting to the network 14. The connection I / F 36 is an interface for communicating with the main body 20 via the communication path 100.

[0024] Operation panel 37 accepts various inputs in response to user operations and displays various information (for example, information in response to the accepted operations, information indicating the operating status of MFP 10, information indicating the setting state, etc.). In this example, operation panel 37 is configured as a liquid crystal display (LCD) equipped with a touch panel function, but is not limited to this. For example, operation panel 37 may be configured as an organic EL display equipped with a touch panel function. Furthermore, operation panel 37 may also be provided with operation units such as hardware keys and display units such as lamps in addition to or instead of this.

[0025] The image processing program executed by the operation unit 30 of the MFP 10 of this embodiment is provided in a state that it is pre-installed in a ROM or the like.

[0026] The image processing program executed by the operation unit 30 of the MFP 10 of this embodiment may be configured to be provided by being recorded in an installable or executable file format on a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, or a DVD (Digital Versatile Disk).

[0027] The image processing program executed by operation unit 30 of MFP 10 of this embodiment may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. The image processing program executed by operation unit 30 of MFP 10 of this embodiment may be provided or distributed via a network such as the Internet.

[0028] The image processing program executed by the operation unit 30 of the MFP 10 of this embodiment has a modular configuration including each of the units described below (image receiving unit 301, recognition area acquisition unit 302, character determination unit 303, overflow determination unit 304, first correction unit 305, appending determination unit 306, calculation unit 307, second correction unit 308, and recognition area output unit 309), and in actual hardware, the CPU (processor) reads and executes the image processing program from the ROM, loading each of the units described above onto the main memory, and the image receiving unit 301, recognition area acquisition unit 302, character determination unit 303, overflow determination unit 304, first correction unit 305, appending determination unit 306, calculation unit 307, second correction unit 308, and recognition area output unit 309 are generated on the main memory.

[0029] Next, the voice assistance function executed by the operation unit 30 of the MFP 10 will be described with reference to Fig. 3. Fig. 3 is a block diagram showing the module configuration related to the image processing function of the operation unit 30.

[0030] As shown in FIG. 3, the program executed by the operation unit 30 of the MFP 10 of this embodiment has a modular configuration including each unit (image receiving unit 301, recognition area acquisition unit 302, character determination unit 303, overflow determination unit 304, first correction unit 305, appending determination unit 306, calculation unit 307, second correction unit 308, and recognition area output unit 309). In actual hardware, the CPU 31 reads and executes the program from the ROM 32 or flash memory 34, etc., thereby loading the above-mentioned units onto the RAM 33, and the image receiving unit 301, recognition area acquisition unit 302, character determination unit 303, overflow determination unit 304, first correction unit 305, appending determination unit 306, calculation unit 307, second correction unit 308, and recognition area output unit 309 are generated on the RAM 33.

[0031] The image receiving unit 301 receives, via the communication path 100, an image of a document read by the scanner function realized by the engine unit 27.

[0032] The recognition area acquisition unit 302 acquires information about the recognition area, which is an area where character recognition processing is performed, from the PC 11 via the network 14, within the image received by the image receiving unit 301. The recognition area acquisition unit 302 also acquires information about the recognition area that the user has set by manually operating the MFP 10 while viewing the image showing the document.

[0033] If information about the recognition area cannot be acquired or if the recognition area has not been set, the recognition area acquisition unit 302 may extract and set the recognition area from an image showing a document. For example, the recognition area acquisition unit 302 extracts ruled lines and a frame from the image and sets the extracted frame as the recognition area.

[0034] Here, ruled line extraction may be achieved by, for example, extracting pixel continuity in the horizontal and vertical directions, or by applying a filter to detect horizontal and vertical lines. Furthermore, box extraction may be achieved by finding intersections of ruled lines and extracting the position of each box from their positional relationship. Furthermore, the extracted recognition area may be configured so that the user can add, delete, or otherwise modify it while checking it.

[0035] The recognition area extraction process and setting process performed by the recognition area acquisition unit 302 are not limited to the above-mentioned method, and may be realized using any conventional technology.

[0036] The character determination unit 303 determines whether characters written in a recognition area, which is an area where character recognition is performed within an image, are handwritten or printed. For example, the character determination unit 303 extracts blocks of pixels of the same color from the image and connects adjacent blocks of pixels to extract multiple rectangles for each character. If the rectangles for each character are approximately the same size, the character determination unit 303 determines that these characters are printed. On the other hand, if the rectangles for each character are not uniform in size, the character determination unit 303 determines that these characters are handwritten.

[0037] The character determination unit 303 may also determine whether the characters in the image are handwritten or printed by using the variation in the center positions of the extracted characters or the linearity of the line segments used in the characters. The character determination process by the character determination unit 303 is not limited to the above-mentioned method, and may be realized using any conventional technology.

[0038] If it is determined that the characters written in the recognition area are handwritten, the protrusion determination unit 304 determines whether the characters protrude from the recognition area.

[0039] The first correction unit 305 changes the size of the recognition area when the character determination unit 303 determines that the characters written in the recognition area are handwritten. Also, the first correction unit 305 changes the size of the recognition area when the overflow determination unit 304 determines that the characters written in the recognition area overflow. Protrusion determination unit 304 When pixel data exists within a predetermined range from the boundary of the recognition area, it is determined that the characters written in the recognition area protrude.

[0040] When it is determined that the character written in the recognition area is a printed character, the additional writing determination unit 306 determines whether the character is a character that has been pre-printed on the image or a character that has been additionally written on the image. Furthermore, the additional writing determination unit 306 determines whether the character is a character that has been pre-printed on the image or a character that has been additionally written on the image based on at least one of the position of the character written in the recognition area within the recognition area, the character color of the character, the font type of the character, and the font size of the character.

[0041] When the addition determination unit 306 determines that the character written in the recognition area is a character added to the image, the calculation unit 307 calculates the amount of displacement of the character from the recognition area.

[0042] The second correction unit 308 moves the position of the recognition area when the character determination unit 303 determines that the character written in the recognition area is a printed character. The second correction unit 308 also moves the position of the recognition area according to the amount of deviation calculated by the calculation unit 307.

[0043] The recognition area output unit 309 associates a new recognition area that combines the correction results from the first correction unit 305 and the correction results from the second correction unit 308 with the image, and outputs it to a module that performs character recognition processing in the MFP 10.

[0044] Next, the flow of image processing by the MFP 10 will be described with reference to Fig. 4. Fig. 4 is a flowchart showing the flow of image processing by the MFP 10. As shown in Fig. 4, the MFP 10 reads an original document using its scanner function (step S401).

[0045] Next, the MFP 10 acquires the recognition areas set in the image obtained by reading the document (step S402). Next, the MFP 10 determines whether the characters written in each recognition area are handwritten or printed (step S403).

[0046] If the characters written in the recognition area are handwritten (step S404; Yes), the MFP 10 changes the size of the recognition area (step S405).

[0047] If the characters written in the recognition area are not handwritten (in other words, printed) (step S404; No), the MFP 10 moves the position of the recognition area (step S406).

[0048] Next, the MFP 10 integrates the corrected recognition areas (step S407).

[0049] According to this embodiment, after determining whether the type of characters written in the recognition area is typed or handwritten, the recognition area can be shifted if the characters are typed, and expanded if the characters are handwritten, so that the recognition area can be set correctly. In other words, according to this embodiment, the recognition area can be set correctly even if the characters included in the image are handwritten.

[0050] In the above embodiment, the image processing device of the present invention has been described as being applied to a multifunction device having at least two of the functions of a copy function, a printer function, a scanner function, and a facsimile function, but the image processing device can be applied to any image processing device such as a copier, printer, scanner device, or facsimile device.

[0051] (Regarding processing by the protrusion determination unit 304 and the first correction unit 305) Next, specific examples of processing by the overflow determination unit 304 and the first correction unit 305 will be described with reference to Figs. 5 to 7. Fig. 5 is a diagram showing an example of an image showing a document including handwritten characters. Fig. 6 is a diagram showing an example of determination by the overflow determination unit 304. Fig. 7 is a diagram showing an example of correction of the recognition area by the first correction unit 305.

[0052] As shown in Figure 5, many manuscripts have item names pre-printed in type and require the user to handwrite the corresponding details. For example, as shown in Figure 5(A), product names are written and the user writes the number of items they wish to order by hand, or as shown in Figure 5(B), name, address, phone number, etc.

[0053] When the recognition area acquisition unit 302 executes the recognition area extraction process based on the information on the ruled lines and intersections, it can determine that the user will not add any notes to the areas where the item names are pre-printed (i.e., they are not recognition areas), and therefore recognition areas 500-503 are extracted for the manuscript in FIG. 5(A), and recognition areas 504-506 are extracted for the manuscript in FIG. 5(B). When adding notes by hand to a pre-printed manuscript, the notes are written in individual boxes, so the deviation from the recognition area is not uniform, as is the case when adding notes in type to a pre-printed manuscript. On the other hand, because the user adds notes while looking at the boxes, large deviations such as characters being written so that they are positioned in the center of two recognition areas, as is the case with type, are unlikely to occur.

[0054] Therefore, the overflow determination unit 304 scans the pixels around the boundary of the recognition area, and if a black pixel is found, determines that the character is overflowing. For example, in the example of Fig. 6, the overflow determination unit 304 determines that the written character is overflowing because the black pixel in the recognition area 500 exists in the area 600 so as to straddle the boundary of the recognition area 500.

[0055] On the other hand, the protrusion determination unit 304 determines that the characters written within the recognition area 501 do not protrude because the black pixels within the recognition area 501 exist only within the recognition area 501.

[0056] 7, the first correction unit 305 changes the size of the recognition area 500 determined to have the written characters protruding therefrom, to create a new recognition area 510. For example, in order to deal with the characters protruding therefrom, the first correction unit 305 enlarges the recognition area 500 to create the recognition area 510. Note that the enlargement rate may be set uniformly in advance, or the size of the rectangle (i.e., the character) extracted by connecting the clusters of black pixels may be determined and enlarged accordingly.

[0057] As described above, the first correction unit 305 enlarges only the recognition area where the written characters extend beyond the recognition area, thereby preventing the recognition results from being mixed with noise due to the inclusion of frame lines or characters written in adjacent frames for characters that fit within the recognition area, and by enlarging the recognition area for characters that extend beyond the recognition area, it is expected that all characters will fit within the recognition area, thereby improving recognition accuracy.

[0058] Even if the character fits within the recognition area, the first correction unit 305 may reduce the recognition area in accordance with the size of the character in order to further improve the recognition rate.

[0059] (Regarding processing by the write determination unit 306, calculation unit 307, and second correction unit 308) Next, specific examples of processing by the append determination unit 306, calculation unit 307, and second correction unit 308 will be described with reference to Figs. 8 to 11. Fig. 8 is a diagram showing an example of determination by the append determination unit 306. Fig. 9 is a diagram showing an example of extraction of a rectangle of a character string printed offset from the recognition area. Fig. 10 is a diagram showing an example of calculation of the amount of offset by the calculation unit 307. Fig. 11 is a diagram showing an example of correction of the recognition area by the second correction unit 308.

[0060] Normally, item names are pre-printed in type on a manuscript, and when the user adds the corresponding content in type, even if the text is printed on top of the manuscript, there will be no misalignment in position or size because the characters are aligned in the PC data. However, if the paper is misaligned during printing or an error occurs during transport, the text may be printed outside the pre-printed frame, as shown in Figure 8. When the printed text is misaligned in this way, the size remains the same and the text is misaligned in the same direction.

[0061] 8, when the recognition area extraction process is performed by the recognition area acquisition unit 302, the recognition area is extracted based on information about the ruled lines and intersections, so both areas 703-705 where pre-printed characters are written and areas 700-702 where additional characters are written are extracted as recognition areas. Furthermore, since the pre-printed characters are printed together with their frames and there is no misalignment with the frames, if the printing misalignment is determined to include this area, the amount of printing misalignment may be calculated incorrectly.

[0062] Therefore, the appending determination unit 306 determines whether the characters in each area are pre-printed or added characters. A characteristic of areas with pre-printed characters is that because they contain item names, the characters are located at the top or left edge of the frame. Also, because pre-printed characters are printed together with ruled lines, they are reproduced in the same color. Furthermore, when comparing pre-printed characters with added characters, there are many cases where they are printed in different positions within the area, or where different character colors, font types, and font sizes are used. Using this information, the appending determination unit 306 determines that areas 700 to 702 are recognition areas, and areas 703 to 705 are pre-printed areas (i.e., not recognition areas).

[0063] After the recognition areas 700 to 702 are determined as shown in FIG. 9A, the calculation unit 307 cuts out the recognition areas 700 to 702 and extracts black pixels from each area as shown in FIG. 9B. Next, the calculation unit 307 extracts black pixels connected to the black pixels extracted in FIG. 9B as shown in FIG. 9C. Next, the calculation unit 307 identifies the cluster of black pixels extracted from the recognition area 700 as a character string rectangle 710, the cluster of black pixels extracted from the recognition area 701 as a character string rectangle 711, and the cluster of black pixels extracted from the recognition area 702 as a character string rectangle 712 as shown in FIG. 9D. Note that the recognition area 702 also contains a "0" corresponding to the character string rectangle 710 and a "10" corresponding to the character string rectangle 711, but the calculation unit 307 ignores the character string rectangles that have already been detected from the other recognition areas.

[0064] 10, the calculation unit 307 calculates a vector from each of the extracted character string rectangles 710 to 712 to each of the recognition areas 700 to 702. Here, the starting point of the vector is either the left end, center, or right end of each of the character string rectangles 710 to 712. The ending point of the vector is either the left end, center, or right end of each of the recognition areas 700 to 702.

[0065] In the case of character string rectangle 710, the three vectors shown in FIG. 10(A) are calculated. In the case of character string rectangle 711, the three vectors shown in FIG. 10(B) are calculated. In the case of character string rectangle 712, the three vectors shown in FIG. 10(C) are calculated. Since vector 800 is included in all of FIGS. 10(A) to 10(C), the calculation unit 307 calculates the amount of deviation of character string rectangles 710 to 712 (in other words, the added characters) as a vector in the opposite direction to vector 800. In this way, the calculation unit 307 can calculate the amount of deviation of the added characters by finding the most frequent value of the vector from each character string rectangle to the recognition area.

[0066] Then, as shown in FIG. 11, the second correction unit 308 moves the recognition areas 700 to 702, respectively, based on the amount of deviation calculated as described above, to create new recognition areas 720 to 722.

[0067] By performing the above process, it is expected that the characters will fit within the recognition area and be recognized correctly. As disclosed in Patent Document 1, by removing ruled lines and completing characters, it is possible to further improve recognition accuracy. [Explanation of symbols]

[0068] 10 MFP 20 Main Unit 30 Control section 301 Image receiving unit 302 Recognition area acquisition unit 303 Character judgment section 304 Protrusion detection section 305 First Correction Section 306 Append judgment unit 307 Calculation Unit 308 Second Correction Section 309 Recognition area output section [Prior art documents] [Patent documents]

[0069] [Patent Document 1] Japanese Patent Application Laid-Open No. 2006-146741

Claims

1. a character determination unit that determines whether a character written in a recognition area, which is an area where character recognition is performed on an image, is handwritten or printed; a first correction unit that changes the size of the recognition area when the character determination unit determines that the character written in the recognition area is handwritten; a second correction unit that moves a position of the recognition area when the character determination unit determines that the character written in the recognition area is a printed character; An image processing device comprising:

2. a protrusion determination unit that, when it is determined that a character written in the recognition area is handwritten, determines whether the character protrudes from the recognition area; Furthermore, The first correction unit When the overflow determination unit determines that the characters written in the recognition area overflow, the size of the recognition area is changed.

2. The image processing device according to claim 1, wherein:

3. The protrusion determination unit If pixel data exists within a predetermined range from the boundary of the recognition area, it is determined that the characters written in the recognition area protrude.

3. The image processing device according to claim 2.

4. an additional writing determination unit that, when it is determined that the character written in the recognition area is a printed character, determines whether the character is a character that has been pre-printed on the image or a character that has been additionally written on the image; a calculation unit that calculates a deviation amount of the character from the recognition area when the addition determination unit determines that the character written in the recognition area is a character added to the image; and Furthermore, The second correction unit The position of the recognition area is moved in accordance with the amount of deviation calculated by the calculation unit.

4. The image processing device according to claim 1, wherein the image processing device further comprises: a processor for processing the image;

5. The additional writing determination unit It is determined whether the character written in the recognition area is a character that has been pre-printed on the image or a character that has been added to the image, based on at least one of the position of the character within the recognition area, the character color of the character, the font type of the character, and the font size of the character.

5. The image processing device according to claim 4.

6. An image processing method executed by an image processing device, comprising: a character determination step of determining whether a character written in a recognition area, which is an area where character recognition is performed on the image, is handwritten or printed; a first correction step of changing the size of the recognition area when the character determination step determines that the character written in the recognition area is handwritten; a second correction step of moving the position of the recognition area when the character written in the recognition area is determined to be a printed character by the character determination step; An image processing method comprising:

7. a character determination step for determining whether a character written in a recognition area, which is an area where character recognition is performed on the image, is handwritten or printed; a first correction step of changing a size of the recognition area when the character determination step determines that the character written in the recognition area is handwritten; a second correction step of moving the position of the recognition area when the character written in the recognition area is determined to be a printed character by the character determination step; An image processing program characterized by causing an image processing device to execute the above.

Images