JP6438369B2 - Form image determination program, form image determination method, and form image determination system - Google Patents

Description

  The present invention relates to a form image determination program, a form image determination method, and a form image determination system.

  An insurance company, a financial institution, or the like takes a form or slip filled in by a customer as an image image into a form reading device, recognizes characters, and processes the form. The operator reads the form with an image scanner, OCR (Optical Character Recognition), or the like, so that the contents described in the form can be taken into the form reading device faster than manual input.

  In the form reading device, a pixel group constituting a long character in the horizontal direction, which is a target of direction identification, is detected from the read image information of the form, and the detected pixel group is forward direction, reverse direction, right 90 degrees, left A technique for recognizing a plurality of characters by changing directions in four directions of 90 degrees and determining whether or not a form image has been read in the correct direction (upright determination) is known. In addition, a technique for performing character recognition from image information obtained by imaging a form is known.

JP 2005-242825 A JP 2013-30040 A

In the prior art, an effective erecting determination cannot be performed for a form image in which there is no character whose direction is to be identified.
Conventionally, since the operator visually determines whether or not the form image is blurred or blurred, since the judgment criteria for bleeding and blur are different depending on the operator, the quality of the form image has not been unified. Further, since the operator visually determines whether or not the form image has been read correctly, the labor of the operator has been expended.

  In one aspect, an object of the present invention is to provide a form image determination program, a form image determination method, and a form image determination system that can improve the process of determining a form image.

In order to achieve the above object, the following form image determination program is provided. The form image determination program acquires a form image, extracts a straight line from the form image, determines whether the form image is clear based on the result of the sharpness determination process for the straight line, extracts a character recognition area from the form image, When it is determined whether the form image is erect based on the erecting determination process for the character recognition area, and the form image is determined to be clear, and when the form image is determined to be erect Causes the computer to execute processing for determining that the reading of the form image is normal.

  According to one aspect, it is possible to provide a form image determination program, a form image determination method, and a form image determination system that can improve the process of determining a form image.

It is a figure which shows an example of the form image determination apparatus of 1st Embodiment. It is a figure which shows an example of the form image determination system of 2nd Embodiment. It is a figure which shows an example of the hardware constitutions of the form reading apparatus of 2nd Embodiment. It is a figure which shows an example of the form set of 2nd Embodiment. It is a figure which shows an example of the form management table of 2nd Embodiment. It is a figure which shows the flowchart of the form image determination process of 2nd Embodiment. It is a figure which shows the flowchart of the image clear determination process of 2nd Embodiment. It is a figure which shows an example of the image of the thickness clear determination process of 2nd Embodiment. It is a figure which shows the flowchart of the thickness clear determination process of 2nd Embodiment (the 1). It is a figure which shows the flowchart of the thickness clear determination process of 2nd Embodiment (the 2). It is a figure which shows the flowchart of the thickness clear determination process of 2nd Embodiment (the 3). It is a figure which shows an example of the image of the straight line vicinity clear determination process of 2nd Embodiment. It is a figure which shows the flowchart of the straight line vividness determination process of 2nd Embodiment (the 1). It is a figure which shows the flowchart of the straight line vicinity clear determination process of 2nd Embodiment (the 2). It is a figure which shows the flowchart of the straight line vicinity clear determination process of 2nd Embodiment (the 3). It is a figure which shows an example of the image of the image erecting determination process of 2nd Embodiment. It is a figure which shows the flowchart of the image erecting determination process of 2nd Embodiment (the 1). It is a figure which shows the flowchart of the image erecting determination process of 2nd Embodiment (the 2). It is a figure which shows the flowchart of the image erecting determination process of 2nd Embodiment (the 3).

Hereinafter, embodiments will be described in detail with reference to the drawings.
[First Embodiment]
First, the form image determination apparatus according to the first embodiment will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of a form image determination apparatus according to the first embodiment.

  The form image determination device 1 includes a storage unit 2 and a control unit 3. The form image determination apparatus 1 is an apparatus that acquires a form image 4 obtained by imaging a form and determines whether or not the form image 4 has been read correctly, and is, for example, a computer.

  The storage unit 2 is a unit that stores the form image 4 and is, for example, an HDD (Hard Disk Drive). The storage unit 2 can store a form image 4 received from another computer, or can store a form image 4 obtained by imaging a form with an image scanner or the like.

  Here, a form image determination process for determining whether or not the form image determination apparatus 1 has read the form image 4 correctly will be described. The form image determination process can be provided as a form image determination program process executed by the form image determination apparatus 1.

The control unit 3 acquires the form image 4 stored in the storage unit 2, extracts a straight line from the form image 4, and determines whether the form image 4 is clear based on the result of the sharpness determination process 5 for the straight line.
The sharpness determination process 5 includes a thickness clearness determination process and a straight line vicinity clearness determination process. The sharpness determination process sets points at predetermined intervals with respect to a straight line, calculates the thickness of the straight line at each point, and the straight line is clear when each thickness is within a predetermined range. Includes the process of determining that there is. The straight line neighborhood clearness determination process acquires density values obtained by converting the form image 4 to gray scale, sets points at predetermined intervals with respect to the straight lines, and averages the density values of white portions in predetermined areas surrounding the respective points. If the difference between the white average density value and the black average density value is greater than a predetermined density value, a straight line is calculated. Includes a process of determining that is clear.

  The control unit 3 determines that the form image 4 is clear when it is determined that the straight line is clear in the thickness sharpness determination process and when the straight line vicinity clearness determination process determines that the straight line is clear. . In other words, when the result of the sharpness determination process 5 is clear, the control unit 3 can determine that the form image 4 is clear and the form image 4 is clear.

  The control unit 3 extracts the character recognition areas 7a, 7b, and 7c from the form image 4, and determines whether or not the form image 4 is upright based on the upright determination processing 6 for the character recognition areas 7a, 7b, and 7c. The erecting determination process 6 acquires a predetermined number of characters in the order of the size of the character recognition areas 7a, 7b, and 7c, and recognizes the characters by changing the angle for each of the character recognition areas 7a, 7b, and 7c. , Including a process of determining that the form image 4 is upright when the number of characters that can be recognized at an angle of 0 degrees is equal to or greater than a predetermined ratio with respect to a predetermined number. Of the plurality of characters included in the form image 4, an example is shown in which the upper three regions having the character recognition regions 7 a, 7 b, and 7 c are targets of the erecting determination process.

  When it is determined that the form image 4 is clear and the form image 4 is determined to be upright, the control unit 3 captures the form image 4 upright without blurring or blurring. It is determined that the read image is normal. The control unit 3 stores a determination result on whether the reading of the form image 4 is normal in the storage unit 2.

  Thus, since the form image determination apparatus 1 can determine whether or not the reading of the form image 4 is normal, the operator can reduce labor required for checking the form. In addition, the form image determination apparatus 1 can maintain the quality of the form image 4 because the determination standard for bleeding and blurring of the form image 4 is kept constant. Thus, the form image determination apparatus 1 can improve the process of determining a form image.

[Second Embodiment]
Next, as a second embodiment, a form image determination system in which the form image determination apparatus is applied as a server apparatus will be described with reference to FIG. FIG. 2 is a diagram illustrating an example of a form image determination system according to the second embodiment.

  As shown in FIG. 2, the form image determination system 300 includes a plurality of form reading apparatuses 100 (one of which is shown in FIG. 2). The form reading apparatus 100 is connected to the server apparatus 200 and the network 10. It is connected. For example, the form image determination system 300 may be provided in an insurance company, a financial institution, a data center, or the like. In addition, an insurance company, a financial institution, a data center, etc. are examples, and other things may be sufficient.

  The form reading apparatus 100 is an information processing apparatus that reads forms and displays form images. The form reading apparatus 100 transmits image information (hereinafter also referred to as a form image) of the form acquired by imaging the form to the server apparatus 200. In addition, the form reading device 100 receives and displays the form image stored in the server device 200.

  The server apparatus 200 is an information processing apparatus that manages information related to form images and forms. The server apparatus 200 acquires the form image read by the form reading apparatus 100, determines whether the form image has been read correctly, and stores the form image according to the determined result. For example, the server apparatus 200 distinguishes and stores a form image in which defects such as blurring, blurring, or a reading direction error are present, and a form image that has been correctly read without defects. The server apparatus 200 may store the form image and the information related to the form in a storage unit such as an HDD provided in the server apparatus 200 or may store it in a database accessible by the server apparatus 200. The server device 200 stores a threshold value, an upper limit value, and the like used for processing for determining whether or not the form image is correctly read in the HDD as environment definition parameters.

Next, the hardware configuration of the form reading apparatus according to the second embodiment will be described with reference to FIG. FIG. 3 is a diagram illustrating an example of a hardware configuration of the form reading apparatus according to the second embodiment.
The form reading apparatus 100 includes a control unit 110. The control unit 110 includes a processor 111, a RAM (Random Access Memory) 112, an HDD 113, an input / output signal interface 114, a storage medium interface 115, and a communication interface 116. The form reading apparatus 100 is entirely controlled by a processor 111. A RAM 112 and a plurality of peripheral devices are connected to the processor 111 via a bus 117. The processor 111 may be a multiprocessor. The processor 111 is, for example, a central processing unit (CPU), a micro processing unit (MPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), or a programmable logic device (PLD). Further, the processor 111 may be a combination of two or more elements among CPU, MPU, DSP, ASIC, and PLD.

  The RAM 112 is used as a main storage device of the form reading device 100. The RAM 112 temporarily stores at least part of an OS program and application programs to be executed by the processor 111. The RAM 112 stores various data necessary for processing by the processor 111.

Peripheral devices connected to the bus 117 include an HDD 113, an input / output signal interface 114, a storage medium interface 115, and a communication interface 116.
The HDD 113 magnetically writes and reads data to and from the built-in disk. The HDD 113 is used as an auxiliary storage device of the form reading device 100. The HDD 113 stores an OS program, application programs, and various data. As the auxiliary storage device, a semiconductor storage device such as a flash memory can be used.

  An input device and an output device are connected to the input / output signal interface 114. Examples of the input device include a keyboard 119, a mouse 120, an image scanner 121, and a touch panel. Examples of the output device include a monitor 118, a liquid crystal display, and various panel display devices.

  The input / output signal interface 114 transmits signals sent from the keyboard 119, the mouse 120, and the like to the processor 111. Note that the mouse 120 is an example of a pointing device, and other pointing devices can also be used. Examples of other pointing devices include a touch panel, a tablet, a touch pad, and a trackball.

  The output device displays an image on the monitor screen according to a command from the processor 111. Examples of the monitor 118 include a display device using a CRT (Cathode Ray Tube) and a liquid crystal display device.

  The storage medium interface 115 reads and writes data recorded on the storage medium 122 using magnetism, laser, or the like. The storage medium interface 115 may read data recorded in a storage medium such as a semiconductor memory. The storage medium 122 includes, for example, an optical disk, a semiconductor memory such as a flash memory, and the like. An optical disc is a portable recording medium on which data is recorded so that it can be read by reflection of light. Optical discs include DVD (Digital Versatile Disk), DVD-RAM, CD-ROM (Compact Disc-Read Only Memory), CD-R (Recordable) / RW (ReWritable), and the like.

  The storage medium interface 115 can also be used as a communication interface for connecting peripheral devices to the form reading apparatus 100. For example, a memory device or a memory reader / writer can be connected to the storage medium interface 115. The memory device is a recording medium equipped with a communication function with the storage medium interface 115. The memory reader / writer is a device that writes data to the memory card or reads data from the memory card. A memory card is a card-type recording medium.

  The communication interface 116 is connected to the network 10. The communication interface 116 transmits / receives data to / from other computers or communication devices via the network 10. The network 10 may be a wired network or a wireless network.

  With the hardware configuration as described above, the function of the form reading apparatus 100 of the second embodiment can be realized. The form image determination apparatus 1 shown in the first embodiment can also be realized by the same hardware configuration as the form reading apparatus 100 shown in FIG. The server device 200 can also be realized by the same hardware configuration as the form reading device 100 shown in FIG.

  The server device 200 implements the processing functions of the second embodiment by executing a program recorded on a computer-readable recording medium, for example. A program describing the processing contents to be executed by the server device 200 can be recorded in various recording media. For example, a program to be executed by the server device 200 can be stored in the HDD 113. The processor 111 loads at least a part of the program in the HDD 113 into the RAM 112 and executes the program. In addition, a program to be executed by the server device 200 can be recorded on a portable recording medium such as an optical disk, a memory device, or a memory card. The program stored in the portable recording medium becomes executable after being installed in the HDD 113 under the control of the processor 111, for example. The processor 111 can also read and execute the program directly from the portable recording medium.

Next, the form set will be described with reference to FIG. FIG. 4 is a diagram illustrating an example of a form set according to the second embodiment.
The form set 400 is an example of a form that is read by the operator using the form reading apparatus 100. There are a plurality of types of forms read by the form reading apparatus 100. For example, there is a fixed form that is a form having a predetermined format, and a non-standard form that has no predetermined format.

  The form set 400 includes a standard form 401 and non-standard forms 402 and 403. Forms included in the form set 400 are read from the form reading apparatus 100, and the read form image is transmitted to the server apparatus 200. Note that the form set 400 shown in FIG. 4 forms a set of form sets with three forms, but this is only an example, and includes only one of the standard form and the non-standard form. Also good.

  The fixed form 401 is a form having a predetermined format. The fixed form 401 includes OCR recognition locations 404 and 405. The fixed form 401 is a form that can determine whether or not the form has been read upright by reading the OCR recognition locations 404 and 405 described at predetermined coordinate positions. Note that one OCR recognition location may be included in one form, or two or more locations may be included.

  The non-standard forms 402 and 403 are forms whose formats are not defined. The non-standard forms 402 and 403 are forms that require character recognition of the form image to determine whether the form image is upright.

  Each form included in the form set 400 is read as a form image via the image scanner 121 connected to the form reading apparatus 100 for the characters in the print portion, the items in the item column, and the entered contents.

Next, the form management table will be described with reference to FIG. FIG. 5 is a diagram illustrating an example of a form management table according to the second embodiment.
The form management table 450 is a table that stores management information for managing form images. The form management table 450 is stored in a storage unit such as an HDD of the server device 200.

  The form management table 450 includes a case number, a security number, a form image file name, a form type, and a determination result. Each time the server apparatus 200 receives a form image from the form reading apparatus 100, the server apparatus 200 records a case number, a security number, and a form image file name for managing the received form image in the form management table 450.

  The case number is a number for identifying the form set. The security number is a number for identifying a form included in the form set. The form image file name is the file name of the form image obtained by reading the form. Although illustration is omitted, information (folder information, directory information, etc.) indicating the location where the form image is stored can be held together with the form image file name.

  The form type is information for identifying whether it is a standard form or an atypical form. The server apparatus 200 determines whether to read the form image, and records “standard” or “non-standard” as the form type. The initial value of the form type is “− (before determination)”. When the server apparatus 200 acquires the OCR recognition part from the form image and determines that the image is a standard form image, the server apparatus 200 cannot acquire the OCR recognition part from the form image and determines that the form image is an atypical form. Record “atypical”.

  The determination result is information indicating a result of determining whether or not the form image has been read correctly. The server apparatus 200 determines the bleeding and blurring of the form image and the reading direction (whether it is upright) and records “normal” or “abnormal” as the determination result. The initial value of the determination result is “− (before determination)”. The server device 200 records “normal” when the form image has no defect such as bleeding or reading direction mistake, and records “abnormal” when there is a defect such as bleeding or reading direction mistake. In addition, the server device 200 performs blurring, blurring, and reading direction determination on a form image whose determination result is “− (before determination)”.

  Since the information recorded in the form management table 450 can identify which form image is judged to be “abnormal”, the operator can check only the form image having the judgment result “abnormal”. . Further, since it is not necessary for the operator to check the form image of the determination result “normal”, it is possible to reduce the labor of the visual inspection manually.

Next, the form image determination process according to the second embodiment will be described with reference to FIG. FIG. 6 is a flowchart illustrating the form image determination process according to the second embodiment.
The form image determination process is a process in which the server apparatus 200 acquires a form image read by the form reading apparatus 100, performs image sharpness determination and image erecting determination on the acquired form image, and records the determination result. . The form image determination process is a process executed by the control unit 110 (processor 111) of the server device 200.

  [Step S11] The control unit 110 acquires preset environment definition parameters from the HDD 113. The environment definition parameter is a generic name of a preset threshold value and upper limit number used as determination conditions in the image sharpness determination process and the image erecting determination process described later. Details of the threshold value and the like included in the environment definition parameters will be described later in the image sharpness determination process and the image erecting determination process.

  [Step S12] The control unit 110 acquires a form image. More specifically, the control unit 110 refers to the form management table 450 and acquires a form image file having a determination result “− (before determination)”.

  [Step S13] The control unit 110 executes image sharpness determination processing. The image sharpness determination process is a process for determining whether or not a form image is clear based on the width of a straight line and the density value of the straight line included in the form image. The image sharpness determination process will be described later with reference to FIG.

  [Step S14] The control unit 110 determines whether the form image is clear based on the result of the image sharpness determination process. The control unit 110 proceeds to step S15 when the result of the image sharpness determination process is clear, and proceeds to step S20 when the result is not clear.

  [Step S15] The control unit 110 determines whether the form image is an image of a standard form. The control unit 110 determines that the form is a standard form when the form image includes an OCR recognition portion. The control unit 110 proceeds to step S19 when it is a fixed form, and proceeds to step S16 when it is not a fixed form.

  [Step S16] The controller 110 executes an image erecting determination process. The image erecting determination process is a process of determining whether or not the form image is erect by extracting characters included in the form image and performing character recognition at different angles. The image erecting determination process will be described later with reference to FIGS.

  [Step S17] The control unit 110 determines whether the form image is upright based on the result of the image upright determination processing. The control unit 110 proceeds to step S19 when the form image is upright, and proceeds to step S18 when the form image is not upright.

[Step S18] The control unit 110 determines that the form image acquired in step S12 and is the determination target is “abnormal”, and records “abnormal” in the determination result of the form management table 450.
[Step S19] The control unit 110 determines that the form image acquired in step S12 and is the determination target is “normal”, and records “normal” in the determination result of the form management table 450.

  [Step S20] The control unit 110 refers to the form management table 450 and determines whether or not there is a form image whose determination result is “-(before determination)”. The control unit 110 proceeds to step S12 when the determination target form image exists, and ends the form image determination process when it does not exist.

Next, image sharpness determination processing according to the second embodiment will be described with reference to FIG. FIG. 7 is a flowchart illustrating image sharpness determination processing according to the second embodiment.
The image sharpness determination process is a process for determining whether or not a form image is clear based on the width of a straight line and the density value of the straight line included in the form image.

The image clarity determination process is a process executed by the control unit 110 (processor 111) of the server apparatus 200 in step 13 of the form image determination process.
[Step S31] The control unit 110 acquires a grayscale conversion value (density value) for the form image. The grayscale conversion value is a value representing the acquired form image in black and white, and is a value represented by 8-bit brightness information for each pixel. The grayscale conversion value “0” indicates black, and the grace case conversion value “255” indicates white. The grayscale conversion value is closer to black as the value is closer to “0”.

[Step S32] The control unit 110 acquires the resolution of the form image. The resolution is expressed in units such as a value (ppi) indicating the number of pixels per inch.
[Step S33] The control unit 110 extracts a straight line from the form image. As a method for extracting a straight line from a form image, an existing technique such as a Hough transform method (Hough transform method) can be used.

  [Step S34] The control unit 110 determines whether a straight line has been extracted from the form image. If the straight line can be extracted, the control unit 110 proceeds to step S35, and if the straight line cannot be extracted, the control unit 110 proceeds to step S41.

  [Step S35] The control unit 110 calculates a binarization threshold value from the form image. As a method for calculating the binarization threshold, existing techniques such as binarization by Otsu can be used. The binarization threshold value is a value serving as a reference for determining whether a grayscale conversion value (density value) is white or black for a certain target point (pixel). When the value is smaller than the binarization threshold, the control unit 110 determines that the target point is black.

  [Step S36] The control unit 110 executes a clear thickness determination process. The thickness clearness determination process is a process for determining whether or not the form image is clear based on the thickness of the straight line extracted from the form image. The thickness clearness determination process will be described later with reference to FIGS.

  [Step S37] The control unit 110 determines whether or not the result of the thickness sharpness determination process is clear. The control unit 110 proceeds to step S38 when the result of the thickness sharpness determination process is clear, and proceeds to step S41 when the result is not clear.

  [Step S38] The control unit 110 executes straight line neighborhood sharpness determination processing. The straight line vicinity clear determination process is a process for determining whether or not the form image is clear based on the density value in the vicinity of the straight line extracted from the form image. The straight line vicinity clear determination process will be described later with reference to FIGS.

  [Step S39] The controller 110 determines whether or not the result of the straight line vicinity clear determination process is clear. The control unit 110 proceeds to step S40 if the result of the straight line vicinity clear determination process is clear, and proceeds to step S41 if not clear.

[Step S40] The control unit 110 determines that the form image is clear, and ends the image sharpness determination process.
[Step S41] The control unit 110 determines that the form image is unclear and ends the image clear determination process.

Next, an image of the thickness sharpness determination process of the second embodiment will be described with reference to FIG. FIG. 8 is a diagram illustrating an example of an image of a sharpness determination process according to the second embodiment.
An image 500 of the sharpness determination processing is a conceptual diagram of processing for determining whether or not the form image is clear based on the thickness of the straight line extracted from the form image. When the thickness of the straight line is within a predetermined range, the form image is clear. When the thickness of the straight line is different, it is determined that the form image is unclear due to distortion or blurring.

  The target line M501 is a line to be subjected to a sharpness determination among a plurality of straight lines (straight line La531, straight line Lb532, straight line Lc533, straight line Ld534, straight line Le535, etc.) included in the form image 520. The target point Q504 is a point to be subjected to a clear thickness determination among the points Pa502 to Pn503 provided on the target line M501 at a predetermined interval. The value of the thickness 508 of the target line M at each of the points Pa502 to Pn503 provided at predetermined intervals on the target line M501 is obtained. The straight line N505 is a virtual straight line used for obtaining the thickness 508 of the target line M. The straight line N505 is a straight line that passes through the target point Q504 and is perpendicular to the target line M501.

  First, in order to obtain the thickness 508 of the target line M, the thickness of the target line M501 in the Y-axis positive direction is obtained. The target point R507, which is a point located in the positive direction of one pixel Y-axis starting from the target point Q504 on the straight line N505, is compared with the density value of the target point R507 and the binarization threshold to determine white or black. Do (a). When the determination result is black, the target line M501 is counted as the Y-axis positive direction thickness 509, and a point positioned in the positive direction of one pixel Y-axis is set as a new target point R, and the same processing is performed (A). When the determination result for the target point R is white, the Y-axis positive direction thickness 509 of the target line M501 is not counted, and the Y-axis negative direction thickness is determined for the target line M501.

  Next, the thickness of the target line M501 in the negative Y-axis direction is obtained. As in the case of the positive direction, the target point Q504 on the straight line N is set as the starting point, and the density value of the target point R is compared with the binarization threshold as a new target point R for a point located in the negative direction of one pixel Y axis. Then, white or black is determined (c). When the determination result is black, the target line M501 is counted as the Y-axis negative direction thickness 510, and a point located in the negative direction of one pixel Y-axis is set as a new target point R, and the same processing is performed (D). When the determination result is white, the thickness 508 of the target line M is obtained without counting the Y-axis negative direction thickness 510 of the target line M501.

  Next, the thickness 508 of the target line M is obtained. The thickness 508 of the target line M can be obtained as a value obtained by adding the Y-axis positive direction thickness 509 and the Y-axis negative direction thickness 510 and subtracting the number of pixels of the overlapping target point Q. In FIG. 8, a value “3 (”) obtained by adding the Y axis positive direction thickness “2 (pixel)” and the Y axis negative direction thickness “2 (pixel)” and subtracting the target point Q “1 (pixel)”. Pixel) ”is obtained as the thickness 508 of the target line M.

  In this way, the thickness 508 of the target line M is obtained at each of the points Pa502 to Pn503 of the target line M501. When the value of the thickness of the target line M501 at each of the points Pa502 to Pn503 is within a predetermined range, the target line M501 is determined to be clear.

When it is determined that the straight line included in the form image 520 is clear, it is determined that the result of the thickness clear determination process for the form image 520 is clear.
Next, a clear thickness determination process according to the second embodiment will be described with reference to FIGS. 9 to 11. FIG. 9 is a diagram illustrating a flowchart of a thickness clearness determination process according to the second embodiment (part 1). FIG. 10 is a diagram illustrating a flowchart of a thickness clearness determination process according to the second embodiment (part 2). FIG. 11 is a view illustrating a flowchart of a thickness clearness determination process according to the second embodiment (part 3).

  The thickness clearness determination process is a process for determining whether or not the form image is clear based on the thickness of the straight line extracted from the form image. The thickness sharpness determination process is a process executed by the control unit 110 (processor 111) of the server device 200 in step 36 of the image sharpness determination process.

[Step S51] The control unit 110 selects a predetermined number of straight lines La to Ln in the longest order from the straight lines extracted from the form image in step S33.
The predetermined number is a value stored in the HDD 113 in advance. For example, when the predetermined number is “5”, the control unit 110 selects five straight lines La531 to Le535 in the longest order as illustrated in FIG. Further, the control unit 110 is not limited to the length as a condition for selecting the straight line La to the straight line Le. (1) The longest straight line among the extracted straight lines is defined as the straight line La, (2) parallel to the straight line La, and The straight line farthest from the straight line La is the straight line Lb, (3) the angle is orthogonal to the straight line La, the straight line farthest from the straight line La is the straight line Lc, (4) parallel to the straight line Lc, and The straight line farthest from the straight line Lc can be defined as the straight line Ld, and (5) the straight line located at the shortest distance from the center point of the form image can be defined as the straight line Le. In addition, when there are a plurality of straight lines having the same condition when selecting the straight lines La to Le, the control unit 110 selects (A) a long straight line and (B) a straight line close to the origin (upper left corner) of the form image. Priority can be selected.

Note that, even when the predetermined number of straight lines cannot be extracted, the control unit 110 can execute the sharpness determination process using only the extracted straight lines.
[Step S52] The control unit 110 determines the target line M from the straight lines La to Ln. For example, the control unit 110 can determine the target line M in the longest order among the straight lines La to Ln.

  [Step S53] The control unit 110 determines the coordinates of the points Pa to Pn for the target line M based on the point extraction interval. The point extraction interval is a value set in advance by an environment definition parameter, and is a value that specifies an interval (length) for extracting points on a straight line. For example, the point extraction interval can be designated as “30000” (μm).

[Step S54] The controller 110 determines the target point Q from the points Pa to Pn.
[Step S55] The control unit 110 acquires the coordinates of the target point Q.
[Step S56] The control unit 110 determines whether or not the target point Q is a point intersecting with a line other than the target line M. In other words, it is determined whether or not the target point Q is appropriate as a point for obtaining the thickness of the target line M. This is because when the target point Q is a point (intersection, vertex, etc.) that intersects with a line other than the target line M, the thickness of the target line M cannot be determined appropriately.

If the target point Q is a point that intersects with a line other than the target line M, the control unit 110 proceeds to step S57, and otherwise proceeds to step S58.
[Step S57] The control unit 110 determines the next target point Q from the points Pa to Pn, excluding the points already set as the target point Q, and proceeds to step S55.

[Step S58] The control unit 110 determines a virtual straight line N that passes through the target point Q and is orthogonal to the target line M.
[Step S59] The control unit 110 initializes a thickness counter value in both positive and negative directions, which is a variable for obtaining the thickness of the target line M. In other words, the control unit 110 sets the values of the positive direction thickness counter value and the negative direction thickness counter value to “0”.

  [Step S60] The control unit 110 determines an axis for calculating a thickness according to the straight line N. For example, in the example shown in FIG. 8, since the straight line N is in the Y-axis direction, the axis for calculating the thickness is the Y-axis.

[Step S61] The control unit 110 sets the positive direction of the target axis as the target direction. For example, in the example shown in FIG. 8, first, the Y axis positive direction is set as the target direction.
[Step S62] The control unit 110 sets a point located in the one-pixel target direction on the straight line N from the target point Q as the target point R.

[Step S63] The control unit 110 compares the density value of the target point R with the binarization threshold.
[Step S64] The control unit 110 determines whether the target point R is black as a result of the comparison in step S63. For example, when the density value of the target point R is smaller than the binarization threshold, it is determined that the target point R is black.

As a result of the comparison, the control unit 110 proceeds to step S65 when the target point R is black, and proceeds to step S67 when the target point R is not black.
[Step S65] The control unit 110 adds the thickness counter value in the target direction. For example, in the point shown in FIG. 8, first, the Y-axis positive direction thickness counter value is added.

[Step S66] The control unit 110 sets a point in the one-pixel target direction on the straight line N from the target point R as the next target point R, and proceeds to step S63.
[Step S67] The control unit 110 determines whether or not the positive and negative directions of the target axis are set as target directions. The control unit 110 proceeds to step S68 when both the positive and negative directions of the target axis are not set as the target direction, and proceeds to step S69 when the target axis is set as the target direction.

  [Step S68] The control unit 110 sets the negative direction of the target axis as the target direction, and proceeds to step S62. For example, in the example shown in FIG. 8, the target axis is the Y axis, and first the Y axis positive direction is the target direction, and then the Y axis negative direction is the target direction.

  [Step S69] The controller 110 calculates the value of the thickness of the target line M at the target point Q based on the value obtained by adding the thickness counter values in both the positive and negative directions. For example, as shown in FIG. 8, the thickness value of the target line M is obtained by adding the Y-axis positive direction thickness counter value and the Y-axis negative direction thickness counter value and subtracting the number of pixels of the overlapping target point Q. Calculated as the value obtained.

  [Step S <b> 70] The control unit 110 determines whether or not all points Pa to Pn are set as target points Q. The control unit 110 proceeds to step S71 when all the points Pa to Pn are not set as the target points Q, and proceeds to step S72 when all the points Pa to Pn are set as the target points Q.

[Step S71] The control unit 110 determines the next target point Q from the points Pa to Pn, excluding the points already set as the target point Q, and proceeds to step S55.
[Step S72] The control unit 110 calculates an average thickness value of the points Pa to Pn based on the thickness value calculated in step S69.

[Step S73] The controller 110 calculates a difference between the calculated thickness average value and the thickness values of the points Pa to Pn.
[Step S74] As a result of the calculation in step S73, the control unit 110 sets a point having a difference equal to or larger than the thickness threshold as an out-threshold point. The thickness threshold is a value set in advance by an environment definition parameter, and is a value specified by a thickness value (μm). For example, the thickness threshold can be specified as “200” (μm).

  [Step S75] The controller 110 determines whether or not there are points outside the threshold at a rate equal to or greater than the straight line determination threshold. The straight line determination threshold is a value set in advance by an environment definition parameter. The straight line determination threshold is a value indicating the ratio of the number of points outside the threshold to the total number of points (points Pa to Pn) set for measuring the thickness of the target line M. For example, the straight line determination threshold value can be specified as “10” (%). When the out-of-threshold points are present at a ratio equal to or greater than the straight line determination threshold with respect to the total number of all points (points Pa to Pn) on the target line M, the control unit 110 does not have a uniform thickness of the target line M ( It can be determined that there is variation in the thickness of the target line M).

The control unit 110 proceeds to step S76 if the out-of-threshold points are present at a ratio equal to or greater than the straight line determination threshold value, and proceeds to step S77 when the non-threshold points are not present at a predetermined ratio or more.
[Step S76] The control unit 110 determines that the target line M is unclear.

[Step S77] The control unit 110 determines that the target line M is clear.
[Step S78] The control unit 110 determines whether or not all the straight lines La to Ln are set as the target line M. The control unit 110 proceeds to step S79 when all the straight lines are not set as the target line M, and proceeds to step S80 when all the straight lines are set as the target line M.

[Step S79] The control unit 110 determines the next target line M from the straight lines La to Ln, excluding the straight line already set as the target line M, and proceeds to step S53.
[Step S80] The controller 110 determines whether or not all the straight lines La to Ln are determined to be clear. The control unit 110 proceeds to step S82 when it is determined that all the straight lines La to Ln are clear, and proceeds to step S81 when it is not determined that all the straight lines La to Ln are clear.

[Step S81] The control unit 110 blurs the result of the sharpness determination process, and ends the clearness determination process.
[Step S82] The control unit 110 sets the result of the sharpness determination process as clear, and ends the clearness determination process.

  Next, an image of the straight line vividness determination process of the second embodiment will be described with reference to FIG. FIG. 12 is a diagram illustrating an example of an image of the straight line neighborhood sharpness determination process according to the second embodiment.

The image 600 of the straight line vicinity clear determination process is a conceptual diagram of the process for determining whether or not the form image is clear based on the density value in the vicinity of the straight line extracted from the form image.
With respect to the target line M601 included in the form image 605, a target point to be determined in the straight line vicinity clearness determination process among points provided at predetermined intervals is set as a target point Q602.

  A square F603 having a predetermined size with respect to the target point Q602 is set, and the density value of each point included in the square F603 is determined to be white or black compared with the binarization threshold. For example, the density value of the target point S604 included in the square F603 is compared with the binarization threshold value. If the density value is larger than the binarization threshold value, the target point S604 is determined to be white.

It is determined whether each point included in the square F603 is white or black, and the white average density value of the point determined to be white and the black average density value of the point determined to be black are obtained.
When the difference between the density value of the point determined to be black and the black average density value is a difference equal to or greater than the black threshold value, it is determined that the form image is not clear because the density value of the black portion is not within the predetermined range. The black threshold value is a value set in advance by the environment definition parameter, and is a value for determining whether or not the density value of the target point determined to be black is within a predetermined range. In addition, when the difference between the density value of the point determined to be white and the white average density value is equal to or greater than the white threshold value, the density value of the white portion is not within a predetermined range, and thus the form image is determined not to be clear. To do. The white threshold value is a value set in advance by the environment definition parameter, and is a value for determining whether or not the density value of the target point determined to be white is within a predetermined range. If the density value of the point determined to be black and the density value of the point determined to be white are not within the range of the black threshold value and the white threshold value (if the density value varies), the form image 605 is blurred. There is a high probability that blur is occurring. For this reason, it is determined that the form image 605 is unclear.

  If the difference between the black average density value and the white average density value is equal to or smaller than the black and white density threshold, it is determined that the form image is not clear because the contrast between white and black is small. The black-and-white density threshold value is a value set in advance by an environment definition parameter, and is a reference value for determining a difference in contrast between a portion determined to be white and a portion determined to be black.

In this way, it is possible to determine whether or not the form image is clear based on the density value near the straight line extracted from the form image.
Next, the straight line neighborhood sharpness determination process of the second embodiment will be described with reference to FIGS. 13 to 15. FIG. 13 is a diagram illustrating a flowchart of the straight line vicinity clear determination process according to the second embodiment (part 1). FIG. 14 is a diagram illustrating a flowchart of straight line vicinity clearness determination processing according to the second embodiment (part 2). FIG. 15 is a diagram illustrating a flowchart of straight line neighborhood sharpness determination processing according to the second embodiment (part 3).

  The straight line vicinity clear determination process is a process for determining whether or not the form image is clear based on the density value in the vicinity of the straight line extracted from the form image. The straight line neighborhood sharpness determination process is a process executed by the control unit 110 (processor 111) of the server device 200 in step S38 of the image sharpness determination process.

[Step S91] The control unit 110 determines the target line M from the straight lines La to Ln extracted in step S51.
[Step S92] The control unit 110 determines the target point Q from the points Pa to Pn on the target line M.

[Step S93] The control unit 110 acquires the value of the thickness of the target point Q obtained in step S70.
[Step S94] The control unit 110 sets a value of one side of the square F that surrounds the target point Q. More specifically, the control unit 110 sets the minimum odd number of pixels exceeding the thickness of the target point Q to be a value on one side of the square F. In the example of FIG. 12, the thickness of the target line M601 is 3 pixels, and the value of one side of the square F603 is set to 5 pixels.

  [Step S95] The control unit 110 extracts one pixel from the inner area of the square F and determines the target point S. For example, the control unit 110 can determine the target point S sequentially from the point located at the corner of the inner area of the square F.

[Step S96] The control unit 110 compares the density value of the target point S with the binarization threshold value.
[Step S97] The control unit 110 determines whether the target point S is black as a result of the comparison in step S96. For example, when the density value of the target point S is smaller than the binarization threshold, it is determined that the target point S is black.

As a result of the comparison, the control unit 110 proceeds to step S99 if the target point S is black, and proceeds to step S98 if the target point S is not black.
[Step S98] The control unit 110 sets the target point S as a white target point.

[Step S99] The control unit 110 sets the target point S as a black target point.
[Step S100] The control unit 110 determines whether or not the density values of all points in the inner region of the square F have been compared with the binarization threshold.

The control unit 110 proceeds to step S102 if the density values of all points have been compared, and proceeds to step S101 if not compared.
[Step S101] The control unit 110 determines a next target point S from the inner area of the square F except for points that have already been set as the target point S, and proceeds to step S96.

  [Step S102] The control unit 110 calculates a black average density value for the black target point. In other words, the control unit 110 adds the density values of all the target points S that are black target points in step S99, and calculates the black average density value by dividing the added value by the number of black target points.

[Step S103] The control unit 110 calculates the difference between the density value of each black target point and the average black density value.
[Step S104] As a result of the calculation in step S103, the control unit 110 determines whether there is a black target point having a difference equal to or greater than the black threshold. The black threshold value is a value set in advance by the environment definition parameter, and is a value for determining whether or not the density value of the target point determined to be black is within a predetermined range. The black threshold is a value designated by 256 gray scales, and is set to “10”, for example.

The control unit 110 proceeds to step S110 when there is a black target point having a difference equal to or larger than the black threshold value, and proceeds to step S105 when there is no black target point.
[Step S105] The control unit 110 calculates a white average density value for the white target point. In other words, the control unit 110 adds the density values of all the target points S that are white target points in step S98, and obtains the white average density value by dividing the added value by the number of white target points.

[Step S106] The controller 110 calculates the difference between the density value of each white target point and the white average density value.
[Step S107] As a result of the calculation in step S106, the control unit 110 determines whether there is a white target point having a difference equal to or greater than the white threshold. The white threshold value is a value set in advance by the environment definition parameter, and is a value for determining whether or not the density value of the target point determined to be white is within a predetermined range. The white threshold is a value designated by 256 gray scales, and is set to “20”, for example.

The control unit 110 proceeds to step S110 when there is a white target point having a difference equal to or greater than the white threshold, and proceeds to step S108 when there is no white target point.
[Step S108] The control unit 110 calculates the difference between the black average density value and the white average density value.

  [Step S109] As a result of the calculation in step S108, the control unit 110 determines whether or not the difference between the black average density value and the white average density value is equal to or smaller than the black and white density threshold. The black-and-white density threshold value is a value set in advance by an environment definition parameter, and is a reference value for determining a difference in contrast between a portion determined to be white and a portion determined to be black. When the difference between the black average density value and the white average density value is equal to or less than the black and white density threshold, the black and white difference is not clear in the form image, and the control unit 110 determines that it is unclear.

The control unit 110 proceeds to step S110 if the difference between the average density values is less than or equal to the black and white density threshold, and proceeds to step S111 if it is not less than or equal to the black and white density threshold.
[Step S110] The control unit 110 determines that the target point Q is unclear.

[Step S111] The control unit 110 determines that the target point Q is clear.
[Step S112] The control unit 110 determines whether or not all the points Pa to Pn are set as the target points Q. The control unit 110 proceeds to step S113 when all the points are not set as the target points Q, and proceeds to step S114 when all the points are set as the target points Q.

[Step S113] The control unit 110 determines the next target point Q from the points Pa to Pn, excluding the points already set as the target point Q, and proceeds to step S93.
[Step S114] The control unit 110 determines that the target line M is clear.

[Step S115] The control unit 110 determines that the target line M is unclear.
[Step S116] The control unit 110 determines whether or not all the straight lines La to Ln are the target lines M. The control unit 110 proceeds to step S117 when all the straight lines La to Ln are not set as the target line M, and proceeds to step S119 when all the straight lines La to Ln are set as the target line M.

[Step S117] The control unit 110 determines the next target line M from the straight lines La to Ln, excluding the straight line already set as the target line M, and proceeds to step S92.
[Step S118] The control unit 110 blurs the result of the straight line vicinity sharpness determination process, and ends the straight line vicinity clearness determination process.

[Step S119] The control unit 110 sharpens the result of the straight line vicinity clearness determination process, and ends the straight line vicinity clearness determination process.
Next, an image of the image erecting determination process of the second embodiment will be described with reference to FIG. FIG. 16 is a diagram illustrating an example of an image erecting determination process according to the second embodiment.

The image erecting determination processing image 700 is a diagram illustrating the concept of processing for determining whether or not a form image is upright by extracting characters included in the form image and performing character recognition at different angles. .
First, as a procedure 1, characters included in the form image 701 are extracted as a rectangular character recognition area (rectangular area) using a labeling technique. The extracted rectangular areas are defined as a rectangular area A702, a rectangular area B703, a rectangular area C704, a rectangular area D705, and a rectangular area E706 in descending order of size. Note that the form image 701 includes a plurality of characters, but a predetermined number of characters are extracted as a rectangular area in order from the largest size character to be a character recognition target. This is because a large size character is easy to read and thus has few misrecognitions. Moreover, because large characters are often the titles of forms, plain terms are used and the probability of character recognition is high. On the other hand, there is a possibility that characters having a small size are crushed when the form image is acquired by the image scanner 121. In addition, when the form is a contract, a small-sized character is a character of an explanation sentence of the contract content, and a difficult character with a large number of strokes is often used, which tends to be unsuitable for character recognition. For this reason, the efficiency of character recognition is improved by performing character recognition in order from the character with the largest character size.

  Next, as procedure 2, for each of the extracted rectangular area A702 to rectangular area E706, character recognition is performed by changing the rotation angles to 0 degrees, 90 degrees, 180 degrees, and 270 degrees. In the example of the procedure 2, an example in which the angle of the rectangular area A702 is changed is shown. Character recognition is performed at each rotation angle for each rectangular area, and if the angle at which the number of characters that can be recognized is large is 0 degrees, it is determined that the form image is upright.

  Next, an image erecting determination process according to the second embodiment will be described with reference to FIGS. FIG. 17 is a diagram illustrating a flowchart of image erecting determination processing according to the second embodiment (part 1). FIG. 18 is a diagram illustrating a flowchart of image erecting determination processing according to the second embodiment (part 2). FIG. 19 is a diagram illustrating a flowchart of image erecting determination processing according to the second embodiment (part 3).

  The image erecting determination process is a process of determining whether or not the form image is erect by extracting characters included in the form image and performing character recognition at different angles. The image erecting determination process is a process executed by the control unit 110 (processor 111) of the server device 200 in step S18 of the form image determination process.

[Step S131] The control unit 110 performs binarization processing on the form image.
[Step S132] The control unit 110 performs a labeling process on the form image binarized in step S131, and extracts a rectangular area surrounding the outline of the black pixel.

  [Step S133] Based on the rectangular region extracted in step S132, the control unit 110 extracts a predetermined number of rectangular regions including a character image from the upper size. The predetermined number to be extracted is a value that can be set in advance by an environment definition parameter, for example, “10”.

  [Step S134] The control unit 110 determines whether or not a predetermined number of rectangular regions have been extracted in step S133. The control unit 110 proceeds to step S135 when the rectangular area can be extracted, and proceeds to step S155 when the rectangular area cannot be extracted.

[Step S135] The control unit 110 acquires a character image to be recognized from the rectangular area.
[Step S136] The controller 110 recognizes the character image acquired in step S135 at a rotation angle of 0 degrees.

  [Step S137] The control unit 110 determines whether or not characters can be recognized at a rotation angle of 0 degrees. The control unit 110 proceeds to step S138 when the character can be recognized at the rotation angle of 0 degree, and proceeds to step S139 when the character cannot be recognized.

[Step S138] The control unit 110 adds 1 point to the 0 degree recognition counter. The 0 degree recognition counter is a value for counting the number of characters that can be recognized at a rotation angle of 0 degree.
[Step S139] The control unit 110 recognizes the character image acquired in step S135 at a rotation angle of 90 degrees.

  [Step S140] The control unit 110 determines whether or not characters can be recognized at a rotation angle of 90 degrees. The control unit 110 proceeds to step S141 when the character can be recognized at the rotation angle of 90 degrees, and proceeds to step S142 when the character cannot be recognized.

  [Step S141] The control unit 110 adds 1 point to the 90-degree recognition counter. The 90-degree recognition counter is a value for counting the number of characters that can be recognized at a rotation angle of 90 degrees.

[Step S142] The control unit 110 recognizes the character image acquired in step S135 at a rotation angle of 180 degrees.
[Step S143] The control unit 110 determines whether or not characters can be recognized at a rotation angle of 180 degrees. The control unit 110 proceeds to step S144 when the character can be recognized at the rotation angle of 180 degrees, and proceeds to step S145 when the character cannot be recognized.

  [Step S144] The control unit 110 adds 1 point to the 180 degree recognition counter. The 180 degree recognition counter is a value for counting the number of characters that can be recognized at a rotation angle of 180 degrees.

[Step S145] The control unit 110 recognizes the character image acquired in step S135 at a rotation angle of 270 degrees.
[Step S146] The control unit 110 determines whether or not characters can be recognized at a rotation angle of 270 degrees. The control unit 110 proceeds to step S147 when the character can be recognized at the rotation angle of 270 degrees, and proceeds to step S148 when the character cannot be recognized.

  [Step S147] The control unit 110 adds 1 point to the 270 degree recognition counter. The 270 degree recognition counter is a value for counting the number of characters that can be recognized at a rotation angle of 270 degrees.

  [Step S148] The control unit 110 determines whether there is an unrecognized rectangular area that has not yet been recognized. The control unit 110 proceeds to step S149 when there is an unrecognized rectangular area, and proceeds to step S150 when there is no unrecognized rectangular area.

[Step S149] The control unit 110 obtains a character image to be recognized from an unrecognized rectangular area except for a rectangular area that has already been recognized, and proceeds to step S135.
[Step S150] The control unit 110 selects a recognition counter having the highest number of points from the 0-degree recognition counter, the 90-degree recognition counter, the 180-degree recognition counter, and the 270-degree recognition counter.

  [Step S151] The control unit 110 determines whether or not the 0 degree recognition counter is the maximum number of points. The control unit 110 proceeds to step S152 when the 0-degree recognition counter has the maximum number of points, and proceeds to step S155 when the 0-degree recognition counter does not have the maximum number of points.

  [Step S152] The control unit 110 determines whether or not the cumulative number of recognized characters is equal to or less than the upper limit number of character recognition. The upper limit number of character recognition is a value that can be set in advance using environment definition parameters. For example, the upper limit number of character recognition can be set to “100”.

The control unit 110 proceeds to step S153 when the cumulative number of characters is less than or equal to the upper limit number of character recognition, and proceeds to step S155 when the accumulated number of characters is not less than or equal to the upper limit number of character recognition.
[Step S153] The control unit 110 calculates a character recognition rate based on the maximum number of points. More specifically, the control unit 110 calculates the character recognition rate by dividing the number of points of the 0 degree recognition counter by the total number of characters targeted for character recognition. For example, when the number of recognized characters is “10” and the characters can be recognized for all 10 characters at a rotation angle of 0 degrees (when the number of points of the 0 degree recognition counter is 10), the character recognition rate is 100%.

  [Step S154] The controller 110 determines whether or not the character recognition rate calculated in step S153 is equal to or greater than a character recognition rate threshold. The character recognition rate threshold value is a value that can be set in advance using environment definition parameters. For example, the character recognition rate threshold value can be set to “80” (%).

If the character recognition rate is not equal to or higher than the character recognition rate threshold, the control unit 110 proceeds to step S155. If the character recognition rate is equal to or higher than the character recognition rate threshold, the control unit 110 proceeds to step S156.
[Step S155] The controller 110 determines that the result of the image erecting determination process is abnormal, and ends the image erecting determination process.

[Step S156] The control unit 110 sets the result of the image erecting determination process as erect, and ends the image erecting determination process.
In this way, the server device 200 can determine whether the form image is correctly read based on the determination whether the form image is clear and the determination whether the form image is upright.

  By executing the form image determination process by the server device 200, a certain image quality can be maintained for the form image. In addition, the server device 200 executes the form image determination process, thereby reducing the work that the operator has visually determined. Thus, the form image determination apparatus 1 can improve the process of determining a form image.

  The above processing functions can be realized by a computer. In that case, a program describing processing contents of functions that the form reading apparatus 100 and the server apparatus 200 should have is provided. By executing the program on a computer, the above processing functions are realized on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium. Examples of the computer-readable recording medium include a magnetic storage device, an optical disk, a magneto-optical recording medium, and a semiconductor memory. The magnetic storage device includes a hard disk device (HDD), a flexible disk (FD), a magnetic tape, and the like. Optical discs include DVD, DVD-RAM, CD-ROM / RW, and the like. Magneto-optical recording media include MO (Magneto-Optical disk).

  When distributing the program, for example, a portable recording medium such as a DVD or a CD-ROM in which the program is recorded is sold. It is also possible to store the program in a storage device of a server computer and transfer the program from the server computer to another computer via a network.

  The computer that executes the program stores, for example, the program recorded on the portable recording medium or the program transferred from the server computer in its own storage device. The computer reads the program from its own storage device and executes processing according to the program. The computer can also read the program directly from the portable recording medium and execute processing according to the program. In addition, each time a program is transferred from a server computer connected via a network, the computer can sequentially execute processing according to the received program.

  In addition, at least a part of the processing functions described above can be realized by an electronic circuit such as a DSP, ASIC, or PLD.

DESCRIPTION OF SYMBOLS 1 Form image determination apparatus 2 Storage part 3 Control part 4 Form image 5 Sharpness determination process 6 Erecting determination process 7a, 7b, 7c Character recognition area 10 Network 100 Form reading apparatus 110 Control part 111 Processor 112 RAM
113 HDD
114 I / O Signal Interface 115 Storage Medium Interface 116 Communication Interface 117 Bus 118 Monitor 119 Keyboard 120 Mouse 121 Image Scanner 122 Storage Medium 200 Server Device 300 Form Image Determination System

Claims (8)

Get a form image
Extracting a straight line from the form image,
Determine whether the form image is clear based on the result of the sharpness determination process for the straight line,
Extract a character recognition area from the form image,
Determine whether the form image is upright based on the upright determination processing for the character recognition area,
When it is determined that the form image is clear and when it is determined that the form image is upright, it is determined that the reading of the form image is normal.
A form image determination program for causing a computer to execute processing. The sharpness determination process includes
Including thickness sharpness determination processing and straight line neighborhood sharpness determination processing,
The form image determination program according to claim 1. The thickness clearness determination process includes
Set points at predetermined intervals with respect to the straight line,
Calculate the thickness of the straight line at each of the points;
Including a process of determining that the straight line is clear when each of the thicknesses is within a predetermined range;
The form image determination program according to claim 2. The straight line vividness determination process is:
Obtain the grayscale converted density value of the form image,
Set points at predetermined intervals with respect to the straight line,
A white average density value that is an average value of density values of white portions in a predetermined area surrounding each of the points and a black average density value that is an average value of density values of black portions are calculated,
Including a process of determining that the straight line is clear when a difference between the white average density value and the black average density value is larger than a predetermined density value.
The form image determination program according to claim 2. To determine whether the form image is clear,
A process for determining that the form image is clear when it is determined that the straight line is clear in the thickness sharpness determination process and when the straight line vicinity clear determination process determines that the straight line is clear including,
The form image determination program according to claim 3 or 4. The erecting determination process includes
Obtain a predetermined number of the character recognition areas in order of the size of the character recognition area,
Character recognition by changing the angle for each of the character recognition areas,
Including a process of determining that the form image is upright when the number of the character recognition areas that can be recognized at the angle of 0 degrees is equal to or greater than a predetermined ratio with respect to the predetermined number.
The form image determination program according to claim 1. A form image determination method in a form image determination apparatus including a storage unit and a control unit,
The storage unit
Memorize the form image,
The controller is
Obtaining the form image from the storage unit;
Extracting a straight line from the form image,
Determine whether the form image is clear based on the result of the sharpness determination process for the straight line,
Extract a character recognition area from the form image,
Determine whether the form image is upright based on the upright determination processing for the character recognition area,
When it is determined that the form image is clear and when it is determined that the form image is upright, it is determined that the reading of the form image is normal.
Form image determination method. In a form image determination system including a form image reading device and a form image determination device,
The form image reading device comprises:
A reading unit for reading a form image;
A transmission unit that transmits the form image to the form image determination device,
The form image determination device includes:
A storage unit for storing the form image transmitted from the form image reading device;
Obtaining the form image from the storage unit;
Extracting a straight line from the form image,
Determine whether the form image is clear based on the result of the sharpness determination process for the straight line,
Extract a character recognition area from the form image,
Determine whether the form image is upright based on the upright determination processing for the character recognition area ,
A control unit that determines that the form image is clear, and determines that the reading of the form image is normal when it is determined that the form image is upright.
Form image determination system.

Images