JP6309867B2 - Imprint image generation program, imprint image generation method, and imprint image generation apparatus - Google Patents

Description

  The present invention relates to an imprint image generation program, an imprint image generation method, and an imprint image generation apparatus for generating a registered imprint to be registered in an imprint database or generating a checked imprint to be collated with a registered imprint.

  Conventionally, as a means for assuring the authenticity of a paper document, a seal image (a seal) obtained by stamping a seal stamp has been used. For example, when a customer requests a refund by a check or the like at a bank, the authenticity of the check is confirmed by checking whether or not the stamp imprinted on the check matches a registered seal imprinted in advance.

  The seal registration device reads an imprint of a seal stamped on a form with an image reader (scanner) in a financial institution such as a bank or an administrative institution such as a ward office, and registers the read seal image as a registered seal in the seal database. (For example, refer to Patent Document 1). Similarly, the seal collation device collates the imprint image as the collated imprint read by the image reading device with the registered imprint registered in the seal stamp database, and whether a registered imprint that matches at a predetermined ratio is registered. Judge whether or not.

  Registration of the registered seal is performed by binarizing the seal image. In addition, the registered seal stamp and the check target seal are collated using the binarized both imprint images. Therefore, it is necessary to convert (binarize) a multi-valued image (color image, 256 grayscale image, etc.) of a registered imprint and a check target imprint read by the image reading apparatus (for example, refer to Patent Document 2). .)

  For binarization of a multi-valued image, for example, there is a known technique using the Niblack method. In the binarization technique of the Niblack method, the threshold value for binarization can be adjusted by parameter setting. Therefore, the binarized image can be optimally adjusted by increasing the threshold value in the case of a dark imprint image and decreasing the threshold value in the case of a thin imprint image (see, for example, Patent Document 3).

JP 2012-48513 A JP-A-9-128537 JP 2005-100441 A

  As described above, it is necessary to binarize the multi-valued image read by the image reading device into a binary image for the registered seal and the check target seal. The conventional binarization technique is effective when the entire imprint is dark or the entire imprint is thin because the binarization parameter settings are applied to the entire multi-value imprint.

  However, there is a case where the density of the imprint is uneven due to the imbalance of the stamp, that is, one side of the imprint imprints heavily and the other side is rubbed lightly. When binarizing such an imprint, if a binarization threshold matched to a dark part is used, a thin part will rub, and if a binarization threshold matched to a thin part is used, a dark part will blur.

FIG. 1 is a diagram for explaining a problem in binarizing an imprint with uneven shading.
In FIG. 1, an image (A) is a multi-value imprint image read by an image reading apparatus. The imprint image of (A) is unevenly shaded. When this imprint image is binarized using a predetermined binarization threshold, an image of (B) is obtained. The image of (B) is characterized in that the imprint density is uneven. That is, the imprint on the left side of the drawing is dark and the right side, which is the opposite side, is thin. It can be said that a gradation of density occurs in one direction.

  Therefore, when the binarization threshold value is increased to match the darker portion, the thin portion is rubbed and the rubbed portion 11 is generated as in the image of (C). On the other hand, when the binarization threshold value is decreased and matched to a thin portion, the dark portion is blurred and the blurred portion 12 is generated as in the image of (D). In other words, when binarizing an imprint that is uneven in density due to the imbalance of the stamp, it is difficult to realize uniform binarization over the entire imprint.

  Therefore, it is difficult to adjust the binarized image to the optimum appearance, and the operation efficiency is reduced, or the accuracy of the seal stamp is lowered because the entire imprint image is not binarized uniformly. There was a problem.

  The present invention has been made in view of the above-described situation, and an imprint image that can automatically correct the shading of an imprint image with a bias in the imprint density and realize uniform binarization over the entire imprint image. It is an object to provide a generation program, an imprint image generation method, and an imprint image generation apparatus.

The present invention employs the following configuration in order to solve the above problems.
That is, according to one aspect of the present invention, the imprint image generation program of the present invention causes the computer of the imprint image generation apparatus to acquire an imprint multi-value image of a seal stamp stamped on a medium read by the image reading apparatus. Multi-valued image acquisition means, a first binary for generating an imprint binary image by binarizing an imprinted multi-valued image acquired by the imprinted multi-valued image acquiring means using a first binarization threshold value determined in advance A binarizing unit; an imprint frame detecting unit for detecting an outer frame of the imprint binary image generated by the first binarizing unit; and detecting the intensity at a plurality of locations of the outer frame detected by the imprint frame detecting unit. Each of the imprint frame grayscale detecting means, the image dividing means for dividing the impression multivalued image to generate a plurality of partial impression multivalued images, and the plurality of partial impression multivalued images divided by the image dividing means, Imprint frame density inspection Second binarization that generates a plurality of partial imprint binary images by binarization using a second binarization threshold different from the first binarization threshold based on the intensity detected by the means Generated by the binarizing means, the image synthesizing means for synthesizing a plurality of partial impression binary images generated by the second binarizing means and generating one synthesized binary image, and generated by the first binarizing means The imprinted frame binary image is detected by the imprinted frame detecting unit and the straight line passing through the center detected by the imprinted image center detecting unit. An imprint image generation program characterized by detecting the gray level of the outer peripheral frame based on the number of black pixels in a predetermined area centered on an intersection with the outer peripheral frame .

Further , according to one aspect of the present invention, the imprint image generation program of the present invention causes the computer of the imprint image generation apparatus to acquire an imprint multi-value image of a seal stamp on a medium read by the image reading apparatus. Multi-valued image acquisition means, a first binary for generating an imprint binary image by binarizing an imprinted multi-valued image acquired by the imprinted multi-valued image acquiring means using a first binarization threshold value determined in advance. A binarizing unit; an imprint frame detecting unit for detecting an outer frame of the imprint binary image generated by the first binarizing unit; and detecting the intensity at a plurality of locations of the outer frame detected by the imprint frame detecting unit. Each of the imprint frame grayscale detecting means, the image dividing means for dividing the impression multivalued image to generate a plurality of partial impression multivalued images, and the plurality of partial impression multivalued images divided by the image dividing means, Imprint frame shading detection hand A second binary value that is binarized using a second binarization threshold value different from the first binarization threshold value to generate a plurality of partial imprint binary images based on the intensity detected by Generating means, image synthesizing means for synthesizing a plurality of partial impression binary images generated by the second binarizing means to generate one synthesized binary image, and generated by the first binarizing means. The imprinted image binary detection unit functions as an imprinted image center detecting unit that detects the center of the imprinted binary image, and the imprinted frame grayscale detecting unit detects a straight line passing through the center detected by the imprinted image center detecting unit and an outer periphery detected by the imprinted frame detecting unit. An imprint image generation program characterized by detecting the shade of the outer peripheral frame based on the number of black pixels (imprint frame width) in the linear direction at the intersection with the frame.

  Further, the imprint image generation program of the present invention is characterized in that the image dividing means uses the straight line orthogonal to the straight line obtained by obtaining the intersection where the number of black pixels detected by the imprint frame density detection means is the largest. It is desirable to divide the value image to generate a partial impression multi-value image.

  In the imprint image generation program of the present invention, the image dividing unit divides the imprint multivalue image by a straight line passing through the center detected by the imprint center detection unit to generate a partial imprint multivalue image. Is desirable.

Further, according to one aspect of the present invention, the imprint image generation method of the present invention is an imprint image generation method executed in an imprint image generation apparatus, wherein a seal stamp stamped on a medium read by an image reading apparatus is used. The imprinted multi-valued image is binarized using a predetermined first binarization threshold value to generate an imprinted binary image, an outer peripheral frame of the generated imprinted binary image is detected, and the generated A plurality of the detected outer peripheral frames are detected on the basis of the number of black pixels in a predetermined area centered on an intersection between the detected straight line and the detected outer peripheral frame. Detecting the intensity of the location, dividing the impression multi-value image to generate a plurality of partial impression multi-value images, and each of the divided partial impression multi-value images to the detected intensity Based on a second binarization threshold different from the first binarization threshold By binarizing it was used to generate a plurality of partial imprint binary image, and generating one composite binary image by combining a plurality of partial imprint binary image that is generated.

Further, according to one aspect of the present invention, an imprint image generation apparatus according to the present invention is an imprint image multi-apparatus that acquires an imprint multi-value image of a seal stamped on a medium read by an image reading apparatus in the imprint image generation apparatus. A first binary that generates an imprint binary image by binarizing the image multi-value image acquired by the image acquisition unit and the imprint multi-value image acquisition unit using a predetermined first binarization threshold value An imprinting unit, an imprinted frame detecting unit for detecting an outer frame of the imprinted binary image generated by the first binarizing unit, and a center of the imprinted binary image generated by the first binarizing unit Based on the number of black pixels in a predetermined region centered on the intersection of the straight line passing through the center detected by the imprint center detecting unit and the outer peripheral frame detected by the imprint frame detecting unit. a plurality of outer peripheral frame detected by the seal impression frame detector An imprint frame density detecting unit for detecting the intensity of the place, an image dividing unit for dividing the impression multi-valued image to generate a plurality of partial imprinted multi-valued images, and a plurality of partial impressions divided by the image dividing unit Each of the multi-valued images is binarized using a second binarization threshold different from the first binarization threshold based on the shade detected by the imprint frame shade detection unit, and a plurality of images A second binarization unit that generates a partial imprint binary image and an image that generates a single composite binary image by combining a plurality of partial imprint binary images generated by the second binarization unit And a combining unit.

  According to the present invention, there is an effect that it is possible to automatically correct the density of an imprint image in which the density of the imprint is uneven and to realize uniform binarization in the entire imprint image.

It is a figure for demonstrating the problem at the time of binarizing the impression stamp with which the shading was biased. It is a figure for demonstrating the outline | summary of this Embodiment. It is a system block diagram provided with the imprint control apparatus in this Embodiment. It is a hardware block diagram of the imprint control apparatus in this Embodiment. It is a functional block diagram of an imprint image generating device in this embodiment. It is a figure which shows the partial area | region division | segmentation of an imprint binary image. It is FIG. (1) for demonstrating a peripheral black pixel rate. It is FIG. (2) for demonstrating a peripheral black pixel rate. It is a figure which shows the table which illustrates the list of the difference of an outer periphery black pixel rate and an outer periphery black pixel rate. It is a figure (the 1) which shows the example of a division | segmentation of an imprint multi-value image. It is FIG. (2) which shows the example of a division | segmentation of an imprint multi-value image. It is a flowchart which shows the flow of an imprint image production | generation process. It is a figure which shows the example of a binarization parameter table. It is a flowchart which shows the flow of a subroutine "imprinting information detection process." It is a figure which shows the example of an imprint information table. It is a flowchart which shows the flow of the subroutine "imprint shadow detection process" in 1st Example. It is a flowchart which shows the flow of a subroutine "outer periphery black pixel calculation process". It is a figure which shows the example of the environment definition table in 1st Example. It is a figure which shows the example of an outer periphery black pixel rate table. It is a flowchart which shows the flow of the subroutine "shading correction judgment process" in 1st Example. It is a figure which shows the example of an outer periphery black pixel rate difference table. It is a flowchart which shows the flow of the subroutine "imprinted multi-value image division | segmentation process" in 1st Example. It is a figure which shows the example of an imprint division | segmentation image table. It is a flowchart which shows the flow of the subroutine "divided image binarization process" in 1st Example. It is a figure which shows the example of a correction | amendment binarization parameter table. It is a figure which shows the example of a binarization image table. It is a figure for demonstrating calculation of an outer periphery width | variety. It is a figure which shows the calculation result of an outer periphery width | variety and its difference. It is a figure for demonstrating calculation of the outer periphery width | variety containing a part of character of a seal stamp. It is a figure which shows the outer peripheral width after correction | amendment, and the calculation result of the difference. It is a flowchart which shows the flow of the subroutine "imprint shadow detection process" in 2nd Example. It is a flowchart which shows the flow of a subroutine "edge data extraction process." It is a figure which shows the example of an edge data table. It is a flowchart which shows the flow of a subroutine "outer periphery width calculation process." It is a figure which shows the example of the environment definition table in 2nd Example. It is a figure which shows the example of an outer periphery width table. It is a flowchart which shows the flow of the subroutine "shading correction judgment process" in 2nd Example. It is a figure which shows the example of an outer periphery width difference table. It is a flowchart which shows the flow of the subroutine "imprinted multi-valued image division | segmentation process" in 2nd Example. It is a flowchart which shows the flow of the subroutine "divided image binarization process" in 2nd Example. It is an example of a display screen of an imprint registration apparatus. It is a display screen example (part 1) of an imprint registration apparatus that displays a message indicating that density correction has been performed. It is a display screen example (part 2) of the seal imprint registration apparatus that displays a message indicating that density correction has been performed. It is a display screen example (part 3) of the seal imprint registration apparatus that displays a message indicating that the shading correction has been performed. It is a display screen example (the 4) of the seal imprint registration apparatus which displays the message which shows that the shading correction was performed. It is an example of a display screen of an imprint verification apparatus. It is a display screen example (the 1) of the seal impression collation apparatus which displays the message which shows that the shading correction was performed. It is a display screen example (the 2) of the stamp collation apparatus which displays the message which shows that the shading correction was performed. It is a display screen example (the 3) of the stamp collation apparatus which displays the message which shows that the shading correction was performed. It is a display screen example (the 4) of the seal stamp collation apparatus which displays the message which shows that the shading correction was performed.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 2 is a diagram for explaining the outline of the present embodiment.

  In FIG. 2, an image (A) is a multi-value imprint image read by an image reading apparatus. The imprint image of (A) is unevenly shaded. Therefore, the imprinted image of (A) is divided by, for example, the divided areas 21a, 21b, and 21c according to the shade. The divided image of the divided area 21a is a relatively dark image, the divided image of the divided area 21b is a medium image, and the divided image of the divided area 21c is a relatively thin (light) image.

  Then, each divided image is binarized using a binarization threshold value corresponding to the intensity. That is, the divided image of the divided region 21a, which is a relatively dark image, is binarized using a binarization threshold value that is larger than a predetermined binarization threshold value. The results are shown in (B). The divided image of the divided region 21b, which is an intermediate image, is binarized using a binarization threshold that is approximately the same as a predetermined binarization threshold. The result is shown in (C). The divided image of the divided region 21c, which is a relatively thin image, is binarized using a binarization threshold value that is smaller than a predetermined binarization threshold value. The result is shown in (D).

  Next, by combining these binarized divided images, an imprint image binarized uniformly as a whole is generated as shown in (E).

FIG. 3 is a system configuration diagram including the imprint control apparatus according to the present embodiment.
In FIG. 3, the imprint control system 30 includes an imprint control device 40, an image reading device 50 connected to the imprint control device 40, and an imprint database 60. The imprint control device 40 includes an imprint image generation device 41, an imprint registration device 42, and an imprint collation device 43. The imprint control device 40 is connected integrally with the display device 70. Here, the seal imprint control system 30 according to the present embodiment is used, for example, for registration and inquiry of a registered seal of a financial institution.

  The imprint control device 40 has functions of a computer such as a workstation, a server, and a personal computer, for example, and executes an imprint image generation process, an imprint registration process, and an imprint collation process. The seal image generation process is executed by the seal image generation device 41. The seal registration process is executed by the seal registration apparatus 42. The imprint verification process is executed by the imprint verification device 43.

  The image reading device 50 is, for example, an image scanner, and optically reads a seal stamp or the like on which a seal is formed by stamping the seal, and sends it to the seal control device 40.

  The imprint database 60 is, for example, a hard disk device, and stores image data processed as a registered imprint by the imprint registration processing of the imprint registration device 42 in association with an account number or the like.

  The display device 70 displays the read stamp, the seal registered in the seal database 60, the customer number associated with the seal, and the matching result.

  The imprint image generation process executed by the imprint image generation apparatus 41 binarizes the multivalued imprint image read by the image reading apparatus 50 to generate a binary imprint image. Details will be described later.

  The seal registration processing executed by the seal registration apparatus 42 includes a binary seal image generated by the seal image generation process, an input customer number (CIF: Customer's Information Files), a branch number, a store number, and an account The type of subject, the account number of the account, and the like are registered in the seal impression database 60 in association with identification information corresponding to the seal.

  In the seal collation processing executed by the seal collation apparatus 43, for example, a registered seal stamp associated with the input customer number (CIF) is read from the seal stamp database 60. On the other hand, an imprint imprinted for collation is read by the image reading device 50, and a binarized imprint image is generated by the imprint image generating device 41. Then, these two images are collated. The verification of the imprint image uses, for example, feature data in consideration of blurring and blurring of the imprint in addition to pattern matching in pixel units over all the imprint pixels.

FIG. 4 is a hardware configuration diagram of the imprint control apparatus according to the present embodiment.
In FIG. 4, the imprint control device 40 includes a CPU (Central Processing Unit) 401, an input device 402, an output device 403, a ROM (Read Only Memory) 404, a RAM (Random Access Memory) 405, and an interface (I / F) 406. It is configured to be connected to the bus 407. The input device 402 inputs various data and signals such as a keyboard, a joystick, a light pen, a mouse, a touch pad, a touch panel, and a trackball. The output device 403 outputs images and other information, for example, various displays such as an LCD (Liquid Cristal Display), a printer, and the like. The output device 403 corresponds to the display device 70 described above. The ROM 404 stores a control program and table data for controlling and executing each function of the imprint control device 40 in addition to a program for executing processing executed in the imprint control device 40. The RAM 405 stores a frame buffer for the output device 403, some application programs, and the like. The interface 406 is a unit for connecting to an external device such as a serial interface such as USB (Universal Serial Bus) or a parallel interface such as Ethernet (registered trademark). The CPU 401 controls these units.

FIG. 5 is a functional block diagram of the seal image generation apparatus according to the present embodiment.
In FIG. 5, an imprint image generation device 41 includes an imprint multi-value image acquisition unit 411, a first binarization unit 412, an imprint information detection unit 413, an imprint information detection unit 413, an imprint frame grayscale detection unit 416, and an image division unit. 417, a second binarization unit 418 and an image composition unit 419 are provided. The imprint information detection unit 413 includes an imprint frame detection unit 414 and an imprint center detection unit 415.

  The seal impression multivalued image acquisition unit 411 acquires the seal stamp multivalue image of the seal stamp stamped on the medium read by the image reading device 50 from the seal stamp registration device 42 in the case of seal registration, and in the case of seal check. Acquired from the seal collation device 43.

  The first binarization unit 412 binarizes the imprint multi-value image acquired by the imprint multi-value image acquisition unit 411 using a predetermined first binarization threshold to generate an imprint binary image. To do. For example, when each pixel value of the seal impression multi-valued image is equal to or greater than the first binarization threshold, it is converted to the maximum value (255 in the case of 256 gradations), and when it is less than the first binarization threshold Converts to the minimum value (0).

  The imprint information detection unit 413 detects the outer peripheral frame and the center of the imprint binary image generated by the first binarization unit 412. That is, the imprint frame detection unit 414 detects the outer peripheral frame of the imprint binary image generated by the first binarization unit 412. For example, the outer frame is extracted from the imprint binary image by performing Hough transform on the imprint binary image. Also, the seal impression center detection unit 415 detects the center of the seal binary image generated by the first binarization unit 412. For example, an area of a predetermined size surrounded by white pixels is recognized as an imprinted binary image area, and the intersection of a vertical straight line that bisects the area and a horizontal straight line that bisects the area is determined. It is detected as the center of the imprint binary image. Instead of an area of a predetermined size surrounded by white pixels, an area surrounded by an outer peripheral frame detected by the imprint frame detection unit 414 may be used. The imprint information detection unit 413 is not limited to the detection of the outer peripheral frame and the center by the imprint frame detection unit 414 and the detection of the center by the imprint center detection unit 415, but the imprint generated by the first binarization unit 412. It suffices if the outer peripheral frame and center of the binary image can be detected.

  The imprint frame shade detection unit 416 detects the shades at a plurality of locations on the outer peripheral frame detected by the imprint frame detection unit 414. For example, based on the number of black pixels in a predetermined outer peripheral area centering on the intersection of a straight line passing through the center detected by the imprint center detecting unit 415 and the outer peripheral frame detected by the imprint frame detecting unit 414, Detects the intensity of the frame. Alternatively, the imprint frame density detection unit 416 includes the number of black pixels in the linear direction at the intersection of the straight line passing through the center detected by the imprint center detection unit 415 and the outer frame detected by the imprint frame detection unit 414. That is, the shade of the outer peripheral frame is detected based on the imprint frame width.

  The image dividing unit 417 generates a plurality of partial impression multi-value images by dividing the impression multi-value image acquired by the impression multi-value image acquisition unit 411. For example, the imprint multi-valued image is divided by a straight line orthogonal to the straight line obtained from the intersection where the number of black pixels detected by the imprint frame grayscale detection unit 416 is the largest, thereby generating a partial imprint multi-valued image. . Alternatively, the image dividing unit 417 divides the impression multi-value image by a straight line passing through the center detected by the impression center detection unit 415 to generate a partial impression multi-value image.

  The second binarization unit 418 determines each of the plurality of partial impression multi-value images divided by the image division unit 417 based on the gray level detected by the imprint frame gray level detection unit 416. Binarization is performed using a second binarization threshold value different from the binarization threshold value, and a plurality of partial impression binary images are generated. For example, when the value of the lightness is high, that is, when the value is dark, the second binarization threshold value higher than the first binarization threshold value is used. On the other hand, when the value of shading is low, that is, when the value is light, the second binarization threshold value lower than the first binarization threshold value is used.

  The image synthesis unit 419 synthesizes a plurality of partial impression binary images generated by the second binarization unit 418 to generate one synthesized binary image.

  Next, a first example and a second example of the present embodiment will be described regarding the flow of the imprint image generation process executed by the above-described imprint image generation apparatus 41.

  First, the outline of the first embodiment will be described with reference to FIGS. 6, 7, 8, 9, 10, and 11.

  FIG. 6 is a diagram showing the partial area division of the imprint binary image. 7 and 8 are diagrams for explaining the peripheral black pixel ratio. FIG. 9 is a table illustrating a list of differences between the peripheral black pixel ratio and the peripheral black pixel ratio. 10 and 11 are diagrams illustrating an example of division of an imprint multi-value image.

  First, the center of the seal, the diameter of the seal, and the edge data of the seal are obtained by a known seal detection technique.

  The imprint image binarized with the binarization parameter of the environment definition is divided into n equal parts by a line segment passing through the center of the imprint, and the central angle (360 / n) is set as a unit grayscale detection angle.

In the example of FIG. 6, it is divided into eight equal parts, and the unit grayscale detection angle is 45 degrees (= 360/8).
A point whose distance from the center of the imprint is the radius of the imprint is calculated for each density detection angle, and the coordinates of the point on the outer periphery of the imprint are obtained.

  The black pixel ratio is calculated for an m × m outer peripheral area centered on the coordinates of a point on the outer periphery. In the example of FIGS. 7 and 8, m = 7 and a 7 × 7 outer peripheral area is shown.

  The difference between the peripheral black pixel rate at a certain gray level detection angle and the peripheral black pixel rate at an angle shifted by 180 degrees from the angle is calculated. The difference is calculated for each unit grayscale detection angle in the range of the central angle of 0 degree or more and less than 180 degrees.

The angle at which the difference in black pixel ratio is the maximum is the maximum gray angle.
As shown in FIG. 9, in the case of this example, since the difference in the black pixel ratio is the maximum at 0 degrees, the maximum shade angle is 0 degrees.

  Then, it is determined whether or not the absolute value of the difference in the black pixel ratio at the maximum shading angle is equal to or larger than the environment definition threshold. For example, it is determined whether the threshold is 25% or more.

  If the absolute value of the difference in the black pixel ratio is equal to or greater than the environment-defined threshold, it is determined that the imprint has shading, and shading correction processing is executed.

  On the other hand, if the absolute value of the difference in the black pixel ratio is less than the threshold value, it is determined that there is no shading in the imprint and the shading correction process is not executed.

  Since the absolute value of the difference in the black pixel ratio at the maximum gray angle is 49%, which is a value equal to or larger than the threshold value, the subsequent gray level correction process is executed.

  Then, the imprint image is divided into p equal parts by a line segment orthogonal to the maximum shading angle, and the imprint image is divided for each density. In the example of FIG. 10, p = 3, and it is divided into three equal parts.

  Binarization is performed by correcting the binarization parameter for each partial area with reference to the binarization parameter of the environment definition. The binarization parameter of each partial area is corrected using the difference in the peripheral black pixel ratio of the maximum gray angle as an input. A uniform binary image is generated by combining the binarized impression images for each partial area.

  In this example, for simplification of explanation, only the line segment orthogonal to the maximum gray angle is divided, but as shown in FIG. 11, the line is further divided by the maximum gray angle line, and the black pixel rate for each angle. Correction may be performed using the difference between the two as an input.

  In this case, for the density correction of the partial area 1104 and the partial area 1106, the difference of −49% of the black pixel ratio at the density detection angle of 0 degree is corrected as an input.

  For the density correction of the partial area 1103 and the partial area 1107, a black pixel rate difference of −4% at a density detection angle of 45 degrees is corrected as an input.

  For the density correction of the partial area 1102 and the partial area 1108, the difference of −12% in the black pixel rate at the density detection angle of 90 degrees is corrected as an input.

  For the density correction of the partial area 1101 and the partial area 1109, the black pixel rate difference of 27% at the density detection angle of 135 degrees is corrected as an input.

  Next, FIGS. 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22. The first embodiment will be described in detail with reference to FIGS. 23, 24, 25, and 26. FIG.

  FIG. 12 is a flowchart showing the flow of the seal image generation process. FIG. 13 is a diagram illustrating an example of a binarization parameter table.

  The imprint image generation processing in the first embodiment is executed by the imprint image generation device 41 in response to an instruction from the imprint registration device 42 when the imprint registration device 42 binarizes the registered imprint uniformly. In addition, when the imprint collation device 43 binarizes the collated imprint uniformly, the imprint image generation device 41 executes in response to an instruction from the imprint collation device 43.

  First, in step S1201 of FIG. 12, the stamp multi-valued image acquisition unit 411 reads the seal stamp multi-value image stamped on the medium read by the image reading device 50 from the seal stamp registration device 42 in the case of stamp registration. In the case of imprint verification, it is acquired from the imprint verification device 43.

  In step S1202, the first binarization unit 412 acquires the binarization parameter from the binarization parameter table as illustrated in FIG. 13, and the imprint multi-value image acquired in step S1201 is converted into the binarization parameter. Is used to generate a seal impression binary image. For example, each pixel of a 256-gradation imprint multi-value image is formed by using a binarization parameter (parameter 1) value “128” (binarization threshold value) to form an imprint binary composed of white pixels or black pixels. Generate an image. The binarization parameter table is stored in, for example, the ROM 404 or an external recording medium.

  In step S1203, the subroutine “imprint information detection processing” is executed to detect the center and outer peripheral frame of the binary image imprint binarized in step S1202. Details of the “imprint information detection process” will be described later with reference to FIGS. 14 and 15.

  In step S1204, the subroutine “imprint density detection processing” is executed to detect the density of a plurality of locations in the outer peripheral frame of the binary image imprint binarized in step S1202. Details of the “imprint density detection process” will be described later with reference to FIGS. 6, 7, 8, 16, 17, 18, 19, 20, and 21.

  In step S1205, it is determined whether or not it has been determined in a subroutine “imprint density detection process” described later that the density of the impression is necessary. That is, the shading correction presence / absence flag F set in the “shading correction determination process” to be described later with reference to FIG. 20 is “TRUE” indicating that correction is necessary, or “correction is not necessary”. It is determined whether it is “FALSE”. If the density correction presence / absence flag F is “TRUE” (step S1205: YES), the process proceeds to step S1206, and if it is “FALSE” (step S1205: NO), the imprint image generation process ends.

  In step S1206, the subroutine “imprinted multi-valued image dividing process” is executed to divide the imprinted multi-valued image acquired in step S1201 to generate a plurality of partial imprinted multi-valued images. Details of the “imprint multi-valued image dividing process” will be described later with reference to FIGS. 10, 11, 22, and 23.

  In step S1207, a subroutine “divided image binarization process” is executed, and the first binary data is determined for each of the plurality of partial impression multivalued images divided in step S1206 based on the gray level detected in step S1204. Binarization is performed using a second binarization threshold different from the binarization threshold to generate a plurality of partial imprint binary images. Details of the “divided image binarization process” will be described later with reference to FIGS. 24, 25, and 26.

  In step S1208, the image composition unit 419 generates a single composite binary image by combining the plurality of partial impression binary images generated in step S1207. For example, the partial imprint binary image recorded in step S2409 of FIG. 24 described later is read out from the print divided image table (see FIG. 26) and synthesized, and the “imprint image generation process” is completed. The binarized image table is stored in, for example, the ROM 404 or an external recording medium.

  As a result, it is possible to correct the shading of the imprint image with a bias in the imprint density, and to realize uniform binarization in the entire imprint image.

  Next, subroutines “imprint information detection processing” (step S1203), “imprint density detection processing” (step S1204), “imprint multi-value image division processing” (step S1206), and “divided image binarization processing” (step S1207). ) Will be described in detail.

  FIG. 14 is a flowchart showing a flow of a subroutine “imprint information detection process”. FIG. 15 is a diagram illustrating an example of an imprint information table.

  The subroutine “imprint information detection processing” detects the center and outer peripheral frame of the binary image imprint binarized in step S1202 of FIG.

  First, in step S1401 in FIG. 14, labeling is executed on the imprint binary image generated in step S1202 in FIG. 12, and the obtained number of labels is substituted into a variable R. Labeling is a conventional technique that assigns the same number to pixels in which black portions (or white portions) are continuous in an image that has been binarized, and the area (number of pixels), width, and height for each same number. Are used for defect inspection and classification processing. Here, it is used for detection of the outer frame of the imprint binary image. For example, in the case of an imprint of a seal that encloses the word “Nakazawa” with a circular frame, “Middle”, “First stroke of Sawa no Suizu”, “Second stroke of Sawa no Suizu”, “ It is assumed that the label will be labeled as "Sansui's 3rd stroke", "Sawa no Sakai", and "Frame".

  In order to repeatedly execute steps S1404 to S1406 for the number of labels, 1 is substituted into the counter variable n in step S1402, and in step S1403, it is determined whether or not the number of labels R is n or more.

  If it is determined that R is greater than or equal to n (step S1403: YES), in step S1404, the smallest circle including all the black pixels (or white pixels) with the nth label number is obtained.

  In step S1405, the center and radius of the circle obtained in step S1404 are calculated.

  In step S1406, the number of black pixels on the circumference is obtained from the center and radius of the circle calculated in step S1405, and the imprint degree X is calculated as a ratio of the number of black pixels to the circumference (length). Here, the imprinting degree indicates the degree to which the image is an imprinting frame.

  In step S1407, the value of n is incremented by 1 (the variable value is incremented by 1), and the process returns to step S1403.

  If it is determined in step S1403 that R is not greater than or equal to n (step S1403: NO), in step S1408, the circle with the label number for which the maximum imprinting degree is calculated among the imprinting degrees X calculated in step S1406 is imprinted. Detect as a frame.

  In step S1409, the center coordinates O (xo, yo) of the imprint frame detected in step S1408 and the radius ro are recorded in the imprint information table as shown in FIG. The imprint information table is stored in, for example, the ROM 404 or an external recording medium.

  When the subroutine “imprint information detection processing” is terminated, step S1203 in FIG. 12 is terminated.

  Note that the subroutine “imprint information detection process” is not limited to the above-described process, and it is only necessary to detect the center and outer peripheral frame of the binary image imprint binarized in step S1202 of FIG. For example, the imprint frame detection unit 414 extracts the outer frame from the imprint binary image by performing the Hough transform on the binarized image that has been binarized in step S1202 of FIG. , A region of a predetermined size surrounded by white pixels is recognized as a region of the imprint binary image, and an intersection of a vertical straight line that bisects the region and a horizontal straight line that bisects the region, You may detect as the center of an imprint binary image.

  FIG. 16 is a flowchart showing the flow of the subroutine “imprint density detection process” in the first embodiment.

  The subroutine “imprint density detection process” in FIG. 16 is binarized in step S1202 in FIG. 12, and the density of the outer frame of the imprint binary image in which the center and outer frame are detected in the subroutine “imprint information detection process” in FIG. Is detected. For example, in step S1601, a subroutine “peripheral black pixel calculation process” is executed, and a straight line passing through the center coordinates O (xo, yo) of the imprint frame detected by the subroutine “imprint information detection process” described with reference to FIG. Based on the number of black pixels in the outer peripheral area of a predetermined size centered on the intersection with the outer peripheral frame, the density of the outer peripheral frame is detected. In step S1602, a subroutine “shading correction determination processing” is executed, and it is determined whether or not the shading correction of the seal is necessary based on the detected lightness.

  When the subroutine “imprint density detection processing” is terminated, step S1204 in FIG. 12 is terminated.

  Next, the subroutines “peripheral black pixel calculation process” and “shading correction determination process” will be described in detail.

  FIG. 17 is a flowchart showing the flow of the subroutine “peripheral black pixel calculation process”. FIG. 18 is a diagram illustrating an example of the environment definition table in the first embodiment. FIG. 19 is a diagram illustrating an example of the peripheral black pixel rate table.

  First, in step S1701 in FIG. 17, the center angle division number n and the peripheral black pixel rate calculation area size m are acquired from the environment definition table as shown in FIG. The environment definition table is stored in, for example, the ROM 404 or an external recording medium. The central angle division number n is preferably “8”, “16”, “32”, “64”, etc. For example, “8” is acquired as the central angle division number n. Also, the outer peripheral black pixel rate calculation area size m is preferably “5”, “7”, “9”, etc. For example, “7” is acquired as the outer peripheral black pixel ratio calculation area size m.

  In step S1702, 360 degrees is divided by the center angle division number n to calculate a unit grayscale detection angle α (α = 360 / n). When the center angle division number n is 8, the unit grayscale detection angle α is 45 degrees. As shown in FIG. 6, the unit grayscale detection angle α is an angle at which the imprint binary image is divided into partial areas in order to detect the grayscale of the imprint binary image.

  Then, in step S1703, the center coordinates O (xo, yo) of the imprint are acquired from the imprint information table as shown in FIG. 15 recorded in step S1409 of FIG. 14, and the radius of the imprint is acquired in step S1704. .

  In order to repeatedly execute Steps S1706 to S1721 for the center angle division number, 0 is substituted into the counter variable i in Step S1705, and in Step S1706, it is determined whether or not the center angle division number n is larger than i.

  When it is determined in step S1706 that the center angle division number n is larger than i (step S1706: YES), in step S1707, the intersection point between the straight line passing through the center coordinates O (xo, yo) of the imprint frame and the outer peripheral frame is the center. The initial value 0 is substituted into the peripheral black pixel counter zi for counting the number of black pixels in the peripheral area of the predetermined size.

  In step S 1708, xi = xo + ro · cos (α × i) and yi = yo + ro · sin (α × i) are calculated, and the pixel is on the imprint frame of the imprint binary image at the unit grayscale detection angle α. A certain peripheral black pixel ratio calculation center coordinate Bi (xi, xi) is obtained. When the center angle division number n is 8, eight peripheral black pixel rate calculation center coordinates Bi (xi, xi) are obtained every 45 degrees.

  In order to repeatedly execute steps S1711 to S1717 for the outer peripheral black pixel rate calculation area size for each of the eight outer peripheral black pixel rate calculation center coordinates Bi (xi, xi), 0 is substituted into the counter variable j in step S1709. In step S1710, it is determined whether or not the peripheral black pixel rate calculation area size m is larger than j.

  When it is determined in step S1710 that the peripheral black pixel rate calculation area size m is larger than j (step S1710: YES), in step S1711, xj = xi− (m−1) / 2 + j is calculated to determine the peripheral black pixel. The X coordinate xj is obtained.

  In order to repeatedly execute steps S1713 to S1716 for the outer peripheral black pixel rate calculation region size, 0 is substituted into the counter variable k in step S1712, and in step S1713, is the outer black pixel rate calculation region size m larger than k? Judge whether or not.

  If it is determined in step S1713 that the peripheral black pixel rate calculation area size m is larger than k (step S1713: YES), yk = yi− (m−1) / 2 + k is calculated in step S1714 to determine the peripheral black pixel. Y coordinate yk is obtained.

  In step S1715, it is determined whether or not the coordinates (xj, yk) are black pixels.

  If the coordinate (xj, yk) is a black pixel in step S1715 (step S1715: YES), the outer peripheral black pixel counter zi is incremented by 1 in step S1716, and the process proceeds to step S1717. On the other hand, if the coordinate (xj, yk) is not a black pixel in step S1715 (step S1715: NO), the process proceeds to step S1717.

  In step S1717, the value of k is incremented by 1, and the process returns to step S1713.

  If it is determined in step S1713 that the peripheral black pixel rate calculation area size m is not larger than k (step S1713: NO), the value of j is incremented by 1 in step S1718, and the process returns to step S1710.

  If it is determined in step S1710 that the peripheral black pixel rate calculation area size m is not larger than j (step S1710: NO), βi = 100 × zi / (m × m) is calculated in step S1719, and the peripheral black pixel is calculated. The rate βi is obtained. For example, in the example shown in FIG. 7, since there are 11 black pixels in the outer peripheral area 711 centered on the outer peripheral black pixel ratio calculation center coordinate B0 (x0, x0) 701, the outer peripheral black pixel ratio β0 is 22 Since there are 35 black pixels in the outer peripheral area 712 centered on the outer peripheral black pixel ratio calculation center coordinate B4 (x4, x4) 702, the outer peripheral black pixel ratio β4 is 71%. In the example shown in FIG. 8, there are 28 black pixels in the peripheral area 811 centered on the peripheral black pixel ratio calculation center coordinate B3 (x3, x3) 801, and therefore the peripheral black pixel ratio β3 is 57. Since there are 15 black pixels in the outer peripheral region 812 centered on the outer peripheral black pixel rate calculation center coordinate B7 (x7, x7) 802, the outer peripheral black pixel rate β7 is 30%.

  In step S1720, the peripheral black pixel rate βi calculated in step S is recorded in the peripheral black pixel rate table as shown in FIG. The peripheral black pixel rate table is stored in, for example, the ROM 404 or an external recording medium.

  In step S1721, the value of i is incremented by 1, and the process returns to step S1706.

  If it is determined in step S1706 that the center angle division number n is not greater than i (step S1706: NO), the subroutine “outer peripheral black pixel calculation process” is ended, thereby ending step S1601 in FIG.

  FIG. 20 is a flowchart showing the flow of a subroutine “shading correction determination process” in the first embodiment. FIG. 21 is a diagram illustrating an example of the peripheral black pixel rate difference table.

  The subroutine “shading correction determination process” in FIG. 20 determines whether or not to execute the shading correction based on the outer peripheral black pixel ratio calculated in the “outer peripheral black pixel calculation process” described with reference to FIG.

  First, in step S2001 of FIG. 20, the density correction threshold value t is acquired from the environment definition table as shown in FIG. 18, and in step S2002, “FALSE” indicating that correction is not necessary for the density correction presence / absence flag F. Set and initialize.

  In order to repeatedly execute steps S2004 to S2011 by a half of the central angle division number, 0 is substituted into the counter variable i in step S2003, and in step S2004, one half of the central angle division number n is calculated. It is determined whether it is larger than i.

  If it is determined in step S2004 that half of the center angle division number n is larger than i (step S2004: YES), the outer periphery of the index i is determined in step S2005 from the outer peripheral black pixel ratio table as shown in FIG. The black pixel ratio βi is acquired, and in step S2006, the outer peripheral black pixel ratio βi ′ of the index i + n / 2 is acquired.

  In step S2007, [gamma] i = [beta] i- [beta] i 'is calculated to obtain the peripheral black pixel rate difference [gamma] i.

  In step S2008, the outer peripheral black pixel rate difference γi calculated in step S2007 is recorded in an outer peripheral black pixel rate difference table as shown in FIG. The peripheral black pixel rate difference table is stored in, for example, the ROM 404 or an external recording medium.

  In step S2009, it is determined whether the density correction is performed by determining whether the absolute value of γi is greater than the density correction threshold t (| γi |> t?).

  If the absolute value of γi is greater than the light / dark correction threshold t (step S2009: YES), the light / dark correction flag F is set to “TRUE” in step S2010. On the other hand, when the absolute value of γi is not larger than the density correction threshold t (step S2009: NO), the process proceeds to step S2011.

  In step S2011, the value of i is incremented by 1, and the process returns to step S2004.

  If it is determined in step S2004 that one half of the center angle division number n is not greater than i (step S2004: NO), the subroutine “shading correction determination processing” is ended, thereby ending step S1602 in FIG. To do.

  FIG. 22 is a flowchart showing a flow of a subroutine “imprinted multi-valued image dividing process” in the first embodiment. FIG. 23 is a diagram illustrating an example of the seal impression division image table.

  The subroutine “imprint multi-valued image dividing process” of FIG. 22 is executed when the density correction presence / absence flag F is “TRUE” in step S1205 of FIG. 12 (step S1205: YES). In the subroutine “imprinted multi-valued image dividing process”, the imprinted multi-valued image acquired in step S1201 in FIG. 12 is divided to generate a plurality of partial imprinted multi-valued images.

  First, in step S2201, the imprint division number p for binarization correction is acquired from the environment definition table as shown in FIG.

  Then, from the outer black pixel rate difference table as shown in FIG. 21, the outer black pixel rate difference γi is acquired in step S2202, and the gray level detection angle ω that maximizes the outer black pixel rate difference is acquired in step S2203. .

  In step S2204, the coordinates of the line segment of the shade detection angle ω that passes through the center of the seal impression are calculated. The length of the line segment at this time becomes the diameter of the seal impression.

In step S2205, a coordinate C for dividing the line segment of the density detection angle ω into p equal parts is calculated.
In step S2206, a line segment D having a light / dark detection angle ω + 90 degrees passing through the coordinates C calculated in step S2205 is calculated. The length of the line segment at this time becomes the diameter of the seal impression.

  In step S2207, the seal impression multivalued image acquired in step S1201 of FIG.

  In step S2208, the partial impression multivalued image divided in step S2207 is recorded in the impression divided image table as shown in FIG. The imprint division image table is stored in, for example, the ROM 404 or an external recording medium.

  By terminating the subroutine “imprinted multi-value image dividing process”, step S1206 in FIG. 12 is terminated.

  FIG. 24 is a flowchart showing a flow of a subroutine “divided image binarization process” in the first embodiment. FIG. 25 is a diagram illustrating an example of a correction binarization parameter table. FIG. 26 is a diagram illustrating an example of a binarized image table.

  The subroutine “divided image binarization processing” in FIG. 24 is performed by substituting each of the plurality of partial impression multi-value images divided by the subroutine “imprint impression multi-value image division processing” in FIG. Are binarized using a binarization parameter different from the binarization parameter used in step S1202 of FIG. 12 to generate a plurality of partial impression binary images. For example, when the value of shade is high, that is, when the value is dark, a higher value binarization parameter is used. On the other hand, when the value of the lightness is low, that is, when the lightness is low, a binarization parameter having a lower value is used.

  First, in step S2401, the maximum peripheral black pixel rate difference γmax is obtained from the peripheral black pixel rate difference table as shown in FIG.

  In step S2402, the binarization parameter is corrected based on γmax acquired in step S2401 and the imprint division number p acquired in step S2201 of FIG.

  In step S2403, the binarization parameter corrected in step S2402 is recorded in the correction binarization parameter table as shown in FIG. The corrected binarization parameter table is stored in, for example, the ROM 404 or an external recording medium.

  In order to repeatedly execute steps S2405 to S2410 for the number of imprint divisions p, 0 is substituted for the counter variable i in step S2404, and in step S2405, it is determined whether or not the number of imprint divisions p is greater than i. .

  If it is determined in step S2405 that the imprint division number p is greater than i (step S2405: YES), in step S2406, a partial imprint multi-value image with index i is obtained from the imprint division image table as shown in FIG. .

  In step S2407, the binarization parameter corrected by the index i is acquired from the correction binarization parameter table as shown in FIG.

  In step S2408, the partial impression multi-value image is binarized using the correction binarization parameter acquired in step S2407.

  In step S2409, the partial impression multilevel image binarized in step S2408 is recorded in the binarized image table as shown in FIG. The binarized image table is stored in, for example, the ROM 404 or an external recording medium.

  In step S2410, the value of i is incremented by 1, and the process returns to step S2405.

  If it is determined in step S2405 that the imprint division number p is not greater than i (step S2405: NO), the subroutine “divided image binarization processing” is terminated, thereby terminating step S1207 in FIG.

  6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 22. The first embodiment has been described in detail with reference to FIGS. 23, 24, 25 and 26.

  Next, the outline of the second embodiment will be described with reference to FIGS. 6, 10, 11, 27, 28, 29, and 30. FIG.

  FIG. 27 is a diagram for explaining calculation of the outer peripheral width. FIG. 28 is a diagram showing the calculation results of the outer peripheral width and the difference between them. FIG. 29 is a diagram for explaining the calculation of the outer peripheral width including a part of the character of the seal. FIG. 30 is a diagram illustrating a calculation result of the corrected outer peripheral width and the difference between the corrected outer peripheral width and the difference.

  First, the center of the seal, the diameter of the seal, and the edge data of the seal are obtained by a known seal detection technique.

  The imprint image binarized with the binarization parameter of the environment definition is divided into n equal parts by a line segment passing through the center of the imprint, and the central angle (360 / n) is set as a unit grayscale detection angle.

In the example of FIG. 6, it is divided into eight equal parts, and the unit grayscale detection angle is 45 degrees (= 360/8).
Two edge data farthest from the center of the imprint are obtained for each density detection angle, and the outer peripheral width of the imprint is calculated by calculating the distance between the two points. FIG. 27A shows the result of calculating the distance between the two edge data farthest from the center of the seal impression at each of the light / dark detection angle 0 degree and the light / dark detection angle 180 degrees. The outer peripheral width at a gray level detection angle of 0 degrees is 1 [px], and the outer peripheral width at a gray level detection angle of 180 degrees is 5 [px]. Similarly, FIG. 27B shows the outer peripheral width at each of the gray level detection angle of 135 degrees and the gray level detection angle of 315 degrees. The outer peripheral width at a density detection angle of 135 degrees is 3 [px], and the outer peripheral width at a density detection angle of 315 degrees is 1 [px].

  If the imprint is thin, there may be no peripheral edge data. In such a case, for example, the outer peripheral width is set to 0 pixel [px]. If there is no peripheral edge data, the edge data farthest from the center of the seal may be the edge of the character in the seal. In such a case, the distance from the center of the imprint of the edge data is a numerical value smaller than the radius of the imprint. Therefore, in order to detect the case where the outer peripheral width is 0 [px], the circumference detection correction value a is set. When the outer peripheral width is 1 [px], the edge data farthest from the center of the imprint is on the outer periphery of the imprint, and the second farthest edge data is the coordinates inside the outer periphery of the imprint. Therefore, the distance from the center of the imprint of the edge data farthest from the center of the imprint is within the range of the imprint radius ± a, and the distance from the center of the imprint of the edge data second farthest is the radius of the imprint ± a. If it is out of the range, the outer peripheral width is set to 1 [px].

  In this way, the outer peripheral width of the outer periphery of the imprint is calculated for each density detection angle. The result is shown in FIG.

  Next, the difference between the outer peripheral width of a certain shade detection angle and the outer peripheral width of an angle shifted by 180 degrees from the angle is calculated. The difference is calculated for each unit grayscale detection angle in the range of the central angle of 0 degree or more and less than 180 degrees.

  Depending on the density detection angle, the edge data that is the second most distant from the center of the imprint may be a part of the character in the imprint. In that case, the value of the outer peripheral width becomes a larger value than the others, and the shading detection cannot be performed correctly.

  For example, in the example shown in FIG. 29, the width including a part of “sawa” is calculated at an angle of 270 degrees, and the outer peripheral width is not calculated correctly.

  Therefore, in the second embodiment, a “non-peripheral width correction value” for detecting the non-peripheral width is provided. A value obtained by multiplying the calculated median value of the outer peripheral widths by the non-peripheral width correction value is set as a non-outer peripheral width threshold value, and an outer peripheral width exceeding the non-outer peripheral width threshold value is determined as a non-outer peripheral width.

  For example, the average value of the outer peripheral widths of the front and rear grayscale detection angles is set as the value of the outer peripheral width determined to be the non-peripheral width. As the value to be set for the outer peripheral width determined as non-peripheral width, the average value, the median value, the value before and after the entire outer peripheral width may be set, and the optimum value is set according to the system environment be able to.

  In the example shown in FIG. 28, the non-peripheral width correction value is 2.5, and the median is 3. Therefore, the threshold value of the non-peripheral width is 3 × 2.5 = 7.5 [px]. In this example, the lightness / darkness detection angle exceeds the threshold value of the non-peripheral width at 225 degrees and 270 degrees. For this reason, the average value (5 + 1) / 2 = 3 of the 180 ° and 315 ° light and shade detection angles before and after each is set as the correction value for the outer peripheral widths of the light and shade detection angles of 225 ° and 270 °.

In the subsequent processing, the corrected outer peripheral width as shown in FIG. 30 is used.
Next, the angle with the largest difference in the outer peripheral width is set as the maximum shade angle. In the case of the above-described example, since the difference in the outer peripheral width at 0 degree is the maximum, the maximum gray angle is 0 degree.

  Then, it is determined whether the absolute value of the difference between the outer peripheral widths at the maximum gray angle is equal to or larger than the environment-defined threshold. If the absolute value of the difference between the outer peripheral widths is equal to or greater than the environment-defined threshold value, it is determined that the seal has shading, and shading correction processing is executed.

On the other hand, if it is less than the threshold value, it is determined that the seal has no shading, and the shading correction process is not executed.
In the above example, it is assumed that the environment definition threshold is 2 [px]. The absolute value of the difference between the outer peripheral widths at the maximum gray angle is 4 [px], which is a value equal to or larger than the threshold value.

  Next, the imprint image is divided into p equal parts by a line segment orthogonal to the maximum shading angle, and the imprint image is divided for each density. In the example of FIG. 10, p = 3, and it is divided into three equal parts.

  Binarization is performed by correcting the binarization parameter for each partial area with reference to the binarization parameter of the environment definition. The binarization parameter of each partial area is corrected using the difference in the outer peripheral width of the maximum gray angle as an input. A uniform binary image is generated by combining the binarized impression images for each partial area.

  In this example, for simplification of explanation, only the line segment orthogonal to the maximum gray angle is divided, but as shown in FIG. Correction may be performed using the difference as an input.

  In this case, for the density correction of the partial area 1104 and the partial area 1106, the outer peripheral width difference −4 [px] at the density detection angle of 0 degrees is corrected as an input.

  For the density correction of the partial area 1103 and the partial area 1107, the difference 1 [px] in the outer peripheral width at the density detection angle of 45 degrees is corrected as an input.

  For the density correction of the partial area 1102 and the partial area 1108, the difference 0 [px] in the outer peripheral width at the density detection angle of 90 degrees is corrected.

  For the density correction of the partial area 1101 and the partial area 1109, the difference 2 [px] in the outer peripheral width at the density detection angle of 135 degrees is corrected as an input.

  Next, FIGS. 12, 14, 15, 23, 25, 26, 31, 31, 32, 33, 34, 35, 36, 37, 38, 39, and 40. The second embodiment will be described in detail with reference to FIG.

  The imprint image generation process in the second embodiment is similar to the imprint image generation process in the first embodiment when the imprint registration apparatus 42 uniformly binarizes the registered imprint in accordance with an instruction from the imprint registration apparatus 42. The generation device 41 executes. In addition, when the imprint collation device 43 binarizes the collated imprint uniformly, the imprint image generation device 41 executes in response to an instruction from the imprint collation device 43.

12 and 14 are the same as those in the first embodiment, and a description thereof will be omitted.
FIG. 31 is a flowchart showing the flow of the subroutine “imprint density detection processing” in the second embodiment.

  The subroutine “imprint density detection process” in FIG. 31 is binarized in step S1202 in FIG. 12, and the density of the outer frame of the imprint binary image in which the center and outer frame are detected in the subroutine “imprint information detection process” in FIG. Is detected. For example, in step S 3101, the subroutine “edge data extraction process” is executed, and the edge data of the imprint binary image binarized in step S 1202 of FIG. 12 is extracted. Next, in step S3102, a subroutine “peripheral width calculation process” is executed, and the outer peripheral width of the imprint binary image is calculated using the edge data extracted in step S3101. In step S3103, the subroutine “shading correction determination process” in the second embodiment is executed, and it is determined whether or not the shading correction of the seal is necessary based on the detected outer peripheral width.

  When the subroutine “imprint density detection processing” is terminated, step S1204 in FIG. 12 is terminated.

  Next, the subroutine “edge data extraction process”, the subroutine “peripheral width calculation process”, and the subroutine “shading correction determination process” will be described in detail.

  FIG. 32 is a flowchart showing the flow of a subroutine “edge data extraction process”. FIG. 33 is a diagram illustrating an example of the edge data table.

  First, in step S3201 in FIG. 32, black pixel information including the black pixel number B is acquired from the imprint binary image binarized in step S1202 in FIG.

  In order to repeatedly execute Steps S3203 to S3207 for the number of black pixels, 0 is substituted for the counter variable i in Step S3202, and it is determined whether or not the number of black pixels B is larger than i in Step S3203.

  If it is determined in step S3203 that the number of black pixels B is greater than i (step S3203: YES), one black pixel of interest is selected from the black pixel information acquired in step S3201 in step S3204.

  In step S3205, it is determined whether a white pixel exists in the peripheral pixels of the target black pixel selected in step S3204.

  If there is a white pixel (step S3205: YES), in step S3206, the black pixel of interest is recorded as edge data in the edge data table as shown in FIG. 33, and the process proceeds to step S3207. On the other hand, when there is no white pixel (step S3205: NO), the process proceeds to step S3207.

  In step S3207, the value of i is incremented by 1, and the process returns to step S3203.

  If it is determined in step S3203 that the number of black pixels B is not greater than i (step S3203: NO), the subroutine “edge data extraction processing” is terminated, and step S3101 in FIG. 31 is terminated.

  FIG. 34 is a flowchart showing the flow of a subroutine “peripheral width calculation process”. FIG. 35 is a diagram illustrating an example of an environment definition table in the second embodiment. FIG. 36 is a diagram illustrating an example of the outer peripheral width table.

  First, in step S3401 of FIG. 34, the center angle division number n, the circumference detection correction value a, and the non-outer circumference width correction value b are acquired from the environment definition table as shown in FIG.

  In step S3402, 360 degrees is divided by the center angle division number n to calculate a unit grayscale detection angle α (α = 360 / n).

  In step S3403, the imprint radius ro and the imprint center coordinates O (xo, yo) are acquired from the imprint information table (FIG. 15) recorded in step S1409 of FIG.

  In order to repeatedly execute steps S3405 to S3415 for the center angle division number, 0 is substituted into the counter variable i in step S3404, and in step S3405, it is determined whether or not the center angle division number n is larger than i.

  If it is determined in step S3405 that the center angle division number n is larger than i (step S3405: YES), in step S3406, a line segment E having an angle α × i passing through the center coordinates O of the impression is calculated.

  In step S3407, coordinates of edge data on the line segment E calculated in step S3406 and in the angle α × i direction are acquired from the edge data table as shown in FIG.

  In step S3408, among the edge data acquired in step S3407, the edge data coordinate P1 farthest from the center of the impression and the edge data coordinate P2 farthest from the center are acquired.

  In step S3409, it is determined whether or not the distance between the center coordinates O of the imprint and the edge data coordinates P1 is within the range of the imprint radius ro ± a.

  If within the range (step S3409: NO), in step S3410, the outer peripheral width Wi is set to 0, and the process proceeds to step S3414.

  On the other hand, if the distance between the imprint center coordinate O and the edge data coordinate P1 is not within the imprint radius ro ± a (step S3409: YES), in step S3411, the imprint center coordinate O and the edge data coordinate P2 It is determined whether the distance is within the range of the imprint radius ro ± a.

  If it is not within the range (step S3411: NO), in step S3412, 1 is set to the outer peripheral width Wi, and the process proceeds to step S3414.

  On the other hand, if the distance between the center coordinates O of the imprint and the edge data coordinates P2 is within the imprint radius ro ± a (step S3411: YES), the distance between the edge data coordinates P1 and the edge data coordinates P2 in step S3413. Is set to the outer peripheral width Wi, and the process proceeds to step S3414.

  In step S3414, the outer peripheral width Wi set in step S3410, S3412 or S3413 is recorded in the outer peripheral width table as shown in FIG.

  In step S3415, the value of j is incremented by 1, and the process returns to step S3405.

  When it is determined in step S3405 that the center angle division number n is not larger than i (step S3405: NO), in order to repeatedly execute steps S3417 to S3421 by the center angle division number, the counter variable j is set in step S3416. 0 is substituted, and in step S3417, it is determined whether the center angle division number n is larger than j.

  If it is determined in step S3417 that the center angle division number n is greater than j (step S3417: YES), the median value s of the outer peripheral width W is calculated in step S3418.

  In step S3419, it is determined whether or not the outer peripheral width Wj of the index j is larger than s × b.

  If it is determined that the outer peripheral width Wj is larger than s × b (step S3419: YES), the outer peripheral width Wj of the index j is corrected in step S3420, and the process proceeds to step S3421.

  If it is determined that the outer peripheral width Wj is not larger than s × b (step S3419: NO), the outer peripheral width Wj of the index j is not corrected and the process proceeds to step S3421.

  In step S3421, the value of j is incremented by 1, and the process returns to step S3417.

  If it is determined in step S3417 that the center angle division number n is not greater than j (step S3417: NO), the subroutine “outer peripheral width calculation process” is terminated, thereby terminating step S3102 in FIG.

  FIG. 37 is a flowchart showing the flow of a subroutine “shading correction determination process” in the second embodiment. FIG. 38 is a diagram illustrating an example of the outer peripheral width difference table.

  The subroutine “shading correction determination process” in FIG. 37 determines whether or not to execute the shading correction based on the outer peripheral width calculated in the “peripheral width calculation process” described with reference to FIG.

  First, in step S3701 of FIG. 37, the density correction threshold u is acquired from the environment definition table as shown in FIG. 35. In step S3702, “FALSE” indicating that correction is not necessary for the density correction presence / absence flag F. Set and initialize.

  In order to repeatedly execute steps S3704 to S3711 by one-half of the center angle division number, 0 is substituted into the counter variable i in step S3703, and in step S3704, one half of the center angle division number n is calculated. It is determined whether it is larger than i.

  If it is determined in step S3704 that one half of the center angle division number n is larger than i (step S3704: YES), the outer peripheral width Wi of the index i is determined in step S3705 from the outer peripheral width table shown in FIG. In step S3706, the outer peripheral width Wi ′ of the index i + n / 2 is acquired.

  In step S3707, δi = Wi−Wi ′ is calculated to obtain the outer peripheral width difference δi.

  In step S3708, the outer peripheral width difference δi calculated in step S3707 is recorded in an outer peripheral width difference table as shown in FIG. The outer peripheral width difference table is stored in, for example, the ROM 404 or an external recording medium.

  In step S3709, it is determined whether or not density correction is performed by determining whether or not the absolute value of δi is greater than the density correction threshold u (| δi |> u?).

  When the absolute value of δi is larger than the density correction threshold u (step S3709: YES), the density correction presence / absence flag F is set to “TRUE” in step S3710. On the other hand, when the absolute value of δi is not larger than the density correction threshold u (step S3709: NO), the process proceeds to step S3711.

  In step S3711, the value of i is incremented by 1, and the process returns to step S3704.

  If it is determined in step S3704 that ½ of the center angle division number n is not greater than i (step S3704: NO), the subroutine “shading correction determination process” is ended, thereby ending step S3103 in FIG. To do.

  FIG. 39 is a flowchart showing a flow of a subroutine “imprinted multi-valued image dividing process” in the second embodiment.

  The subroutine “imprint multi-valued image division process” of FIG. 39 is executed when the density correction presence / absence flag F is “TRUE” in step S1205 of FIG. 12 (step S1205: YES). In the subroutine “imprinted multi-valued image dividing process”, the imprinted multi-valued image acquired in step S1201 in FIG. 12 is divided to generate a plurality of partial imprinted multi-valued images.

  First, in step S3901, the imprint division number p for binarization correction is acquired from the environment definition table as shown in FIG.

  Then, from the outer peripheral width difference table as shown in FIG. 38, the outer peripheral width difference δi is acquired in step S3902, and in step S3903, the gray level detection angle ω that maximizes the outer peripheral width difference is acquired.

  In step S3904, the coordinates of the line segment of the shade detection angle ω that passes through the center of the seal impression are calculated. The length of the line segment at this time becomes the diameter of the seal impression.

In step S3905, a coordinate C that divides the line segment of the density detection angle ω into p equal parts is calculated.
In step S3906, a line segment D having a light / dark detection angle ω + 90 degrees passing through the coordinates C calculated in step S3905 is calculated. The length of the line segment at this time becomes the diameter of the seal impression.

  In step S3907, the seal impression multivalued image acquired in step S1201 of FIG.

  In step S3908, the partial impression multivalued image divided in step S3907 is recorded in the impression divided image table as shown in FIG. The imprint division image table is stored in, for example, the ROM 404 or an external recording medium.

  By terminating the subroutine “imprinted multi-value image dividing process”, step S1206 in FIG. 12 is terminated.

  FIG. 40 is a flowchart showing a flow of a subroutine “divided image binarization process” in the second embodiment.

  The subroutine “divided image binarization processing” in FIG. 40 is executed by substituting each of the plurality of partial impression multi-value images divided by the subroutine “imprint impression multi-value image division processing” in FIG. Are binarized using a binarization parameter different from the binarization parameter used in step S1202 of FIG. 12 to generate a plurality of partial impression binary images. For example, when the value of shade is high, that is, when the value is dark, a higher value binarization parameter is used. On the other hand, when the value of the lightness is low, that is, when the lightness is low, a binarization parameter having a lower value is used.

  First, in step S4001, the maximum outer peripheral width difference δmax is acquired from the outer peripheral width difference table as shown in FIG.

  In step S4002, the binarization parameter is corrected based on δmax acquired in step S4001 and the imprint division number p acquired in step S3901 of FIG.

  In step S4003, the binarization parameter corrected in step S4002 is recorded in a correction binarization parameter table as shown in FIG.

  In order to repeatedly execute steps S4005 to S4010 for the number of imprint divisions p, 0 is substituted for the counter variable i in step S4004, and in step S4005, it is determined whether or not the number of imprint divisions p is greater than i. .

  If it is determined in step S4005 that the imprint division number p is greater than i (step S4005: YES), in step S4006, a partial imprint multi-value image with index i is obtained from the imprint division image table as shown in FIG. .

  In step S4007, the binarization parameter corrected by the index i is acquired from the correction binarization parameter table as shown in FIG.

  In step S4008, the partially imprinted multi-valued image is binarized using the correction binarization parameter acquired in step S4007.

  In step S4009, the partial impression multilevel image binarized in step S4008 is recorded in the binarized image table as shown in FIG.

  In step S4010, the value of i is incremented by 1, and the process returns to step S4005.

  If it is determined in step S4005 that the imprint division number p is not greater than i (step S4005: NO), the subroutine “divided image binarization processing” is terminated, thereby ending step S1207 in FIG.

  12, 14, 15, 23, 25, 26, 31, 32, 33, 34, 35, 36, 37, 38, 39, and 40. In use, the second embodiment has been described in detail.

  As described above, the imprint image generation apparatus 41 in the embodiment including the first example and the second example executes the imprint image generation process as in the first example and the second example. Then, the multi-value imprint image read by the image reading device 50 is binarized to generate a binary imprint image. Then, the imprint registration device 42 identifies the generated binary imprint image, the input customer number, the branch number, the account number, the account type, the account number of the account, etc. The information is registered in the seal impression database 60 in association with the information.

FIG. 41 is a display screen example of the seal imprint registration apparatus.
The imprint registration device 42 causes the imprint image generation device 41 to execute an imprint image generation process, and displays the generated binary imprint image in the “seal stamp section” field, as shown in FIG. 41. Then, the registration mark is registered by pressing the “Register” button.

  FIG. 42 is a display screen example (part 1) of the seal imprint registration apparatus that displays a message indicating that the shading correction has been performed.

  The imprint registration device 42 automatically executes the imprint image generation processing by the imprint image generation device 41 and displays the generated binary imprint image in the “sealing part” field. At this time, as shown in FIG. 42, a message indicating that the density correction has been performed on the binarized seal image may be displayed in the remark information column.

  FIG. 43 is a display screen example (part 2) of the seal impression registration apparatus that displays a message indicating that the shading correction has been performed.

  The imprint registration device 42 automatically executes the imprint image generation processing by the imprint image generation device 41 and displays the generated binary imprint image in the “sealing part” field. At this time, as shown in FIG. 43, a message indicating that the density correction has been performed on the binarized impression image may be displayed in the message dialog.

  FIG. 44 is a display screen example (part 3) of the seal registration apparatus that displays a message indicating that the shading correction has been performed.

  As shown in FIG. 44, the imprint registration device 42 may execute the imprint image generation processing by the imprint image generation device 41 when the operator presses the “imprint density correction” button. Even in such a case, the imprint registration device 42 displays the generated binary imprint image in the “Seal” section. At that time, as shown in FIG. 44, a message indicating that the density correction has been performed on the binarized impression image may be displayed in the remark information column.

  FIG. 45 is a display screen example (part 4) of the seal registration apparatus that displays a message indicating that the shading correction has been performed.

  As in the example shown in FIG. 44, the imprint registration device 42 executes the imprint image generation processing by the imprint image generation device 41 when the operator presses the “imprint density correction” button as shown in FIG. . At that time, the imprint registration device 42 displays the generated binary imprint image in the “Seal” section, and displays a message in the message dialog indicating that the density correction has been performed on the binarized imprint image. You may make it do.

FIG. 46 is an example of a display screen of the seal collation apparatus.
The seal impression device 43 executes the seal image generation processing by the seal image generation device 41 and displays the generated binary seal image in the “read seal” field as shown in FIG.

  FIG. 47 is an example (part 1) of a display screen of the seal collation apparatus that displays a message indicating that the shading correction has been performed.

  The imprint verification device 43 automatically executes the imprint image generation processing by the imprint image generation device 41 and displays the generated binary imprint image in the “reading stamp” field. At this time, as shown in FIG. 47, a message indicating that the shading correction has been performed on the binarized imprint image may be displayed in the remark information column.

  FIG. 48 is a display screen example (part 2) of the stamp collation apparatus that displays a message indicating that the shading correction has been performed.

  The imprint verification device 43 automatically executes the imprint image generation processing by the imprint image generation device 41 and displays the generated binary imprint image in the “reading stamp” field. At this time, as shown in FIG. 48, a message indicating that the density correction has been performed on the binarized imprint image may be displayed in the message dialog.

  FIG. 49 is a display screen example (No. 3) of the seal collation apparatus that displays a message indicating that the shading correction has been performed.

  As shown in FIG. 49, the seal impression device 43 may execute the seal image generation process by the seal image generation device 41 when the operator presses the “imprint density correction” button. Even in such a case, the imprint verification device 43 displays the generated binary imprint image in the “reading stamp” field. At that time, as shown in FIG. 49, a message indicating that the shading correction has been performed on the binarized imprint image may be displayed in the remark information column.

  FIG. 50 is a display screen example (No. 4) of the seal collation apparatus that displays a message indicating that the shading correction has been performed.

  As in the example shown in FIG. 49, the seal impression device 43 executes the seal image generation processing by the seal image generation device 41 when the operator presses the “imprint density correction” button as shown in FIG. . At that time, the imprint checking device 43 displays the generated binary imprint image in the “reading stamp” field, and displays a message indicating that the density correction has been performed on the binarized imprint image on the message dialog. You may make it do.

  As described above, the embodiments of the present invention have been described with reference to the drawings. However, the imprint image generating apparatus to which the present invention is applied is not limited to the above-described embodiments.

  For example, in the embodiment, a circular imprint is described as an example. However, an imprint can be applied if the shape is point-symmetric, and shading correction can be similarly applied to an elliptical imprint and a square imprint. is there.

  As described above, according to the embodiment of the present invention, uniform binarization is realized over the entire imprint even if the imprint is uneven in density. For example, in the seal registration, even when the seal stamp of the seal slip has a shading, it is automatically binarized evenly, so that the adjustment work of the binarization parameter by the operator can be reduced. Further, in the stamp verification, the reading stamp is uniformly binarized, so that the accuracy of the stamp verification is improved. By improving the collation accuracy, it is easy for the operator to visually check the collation and the operation efficiency is improved. In addition, since the uniform binarization is automatically performed, the adjustment work of the binarization parameter by the operator can be reduced, and the operation efficiency is improved. In addition, since the imprint image of the registration mark is uniformly binarized, the accuracy of the collation is improved during the collation operation, and the operation efficiency is improved by a comprehensive operation of the seal collation and the seal registration.

  The embodiment of the present invention described above can be realized by hardware or firmware or software on a DSP (Digital Signal Processor) board or CPU board as one function of the seal image generation apparatus.

  In addition, the imprint image generating apparatus to which the present invention is applied is not limited to the above-described embodiment as long as the function is executed, and a system composed of a plurality of apparatuses even if it is a single apparatus. Needless to say, the integrated device may be a system in which processing is performed via a network such as a LAN or a WAN.

  It can also be realized by a system including a CPU, a ROM or RAM memory connected to a bus, an input device, an output device, an external recording device, a medium driving device, and a network connection device. That is, a ROM or RAM memory, an external recording device, or a portable recording medium that records a software program for realizing the system of the above-described embodiment is supplied to the imprint image generating device, and the imprint image generating device computer Needless to say, this can also be achieved by reading and executing the program.

  In this case, the program itself read from the portable recording medium or the like realizes the novel function of the present invention, and the portable recording medium or the like on which the program is recorded constitutes the present invention.

  Examples of portable recording media for supplying the program include flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, CD-Rs, DVD-ROMs, DVD-RAMs, magnetic tapes, and nonvolatile memory cards. Various recording media recorded via a network connection device (in other words, a communication line) such as a ROM card, electronic mail or personal computer communication can be used.

  The computer (information processing apparatus) executes the program read out on the memory, thereby realizing the functions of the above-described embodiment, and an OS running on the computer based on the instructions of the program. Performs part or all of the actual processing, and the functions of the above-described embodiments are also realized by the processing.

  Furthermore, a program read from a portable recording medium or a program (data) provided by a program (data) provider is stored in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer. After being written, the CPU of the function expansion board or function expansion unit performs part or all of the actual processing based on the instructions of the program, and the functions of the above-described embodiment are realized by the processing. obtain.

  That is, the present invention is not limited to the embodiments described above, and can take various configurations or shapes without departing from the gist of the present invention.

DESCRIPTION OF SYMBOLS 11 Rub part 12 Bleeding part 21a, 21b, 21c Division | segmentation area | region 30 Imprint control system 40 Imprint control apparatus 41 Imprint image production | generation apparatus 42 Imprint registration apparatus 43 Impression collation apparatus 50 Image reading apparatus 60 Impression database 70 Display apparatus 401 CPU (Central Processing Unit) )
402 Input device 403 Output device 404 ROM (Read Only Memory)
405 RAM (Random Access Memory)
406 Interface (I / F)
407 Bus 411 Imprint multi-value image acquisition unit 412 First binarization unit 413 Imprint information detection unit 414 Imprint frame detection unit 415 Imprint center detection unit 416 Imprint frame density detection unit 417 Image division unit 418 Second binarization unit 419 Image composition unit 701 Peripheral black pixel ratio calculation center coordinate B0 (x0, x0)
702 Peripheral black pixel ratio calculation center coordinates B4 (x4, x4)
711, 712 Outer peripheral area 801 Outer peripheral black pixel ratio calculation center coordinates B3 (x3, x3)
802: Peripheral black pixel ratio calculation center coordinate B7 (x7, x7)
811, 812 Outer peripheral area 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109 Partial area

Claims (6)

The computer of the imprint image generating device,
An imprint multi-value image acquisition means for acquiring an imprint multi-value image of a seal stamp stamped on a medium read by an image reader;
First binarization means for binarizing the impression multivalue image acquired by the impression multivalue image acquisition means using a predetermined first binarization threshold to generate an impression binary image;
An imprint frame detection means for detecting an outer peripheral frame of the imprint binary image generated by the first binarization means;
An imprint frame shade detection means for detecting shades at a plurality of locations on the outer peripheral frame detected by the imprint frame detection means;
Image dividing means for dividing the impression multi-value image to generate a plurality of partial impression multi-value images;
Each of the plurality of partially imprinted multi-valued images divided by the image dividing means is converted into a second second value that is different from the first binarization threshold based on the shade detected by the imprint frame shade detection means. A second binarization unit that binarizes using a binarization threshold and generates a plurality of partial imprint binary images;
Image synthesizing means for synthesizing a plurality of partial impression binary images generated by the second binarizing means to generate one synthesized binary image;
An imprint center detection means for detecting the center of the imprint binary image generated by the first binarization means;
Function as
The imprint frame shade detection means is based on the number of black pixels in a predetermined area centered on the intersection of the straight line passing through the center detected by the imprint center detection means and the outer peripheral frame detected by the imprint frame detection means. wherein detecting the density of the peripheral frame, sign shadow image generation program that is characterized in that. The computer of the imprint image generating device,
An imprint multi-value image acquisition means for acquiring an imprint multi-value image of a seal stamp stamped on a medium read by an image reader;
First binarization means for binarizing the impression multivalue image acquired by the impression multivalue image acquisition means using a predetermined first binarization threshold to generate an impression binary image;
An imprint frame detection means for detecting an outer peripheral frame of the imprint binary image generated by the first binarization means;
An imprint frame shade detection means for detecting shades at a plurality of locations on the outer peripheral frame detected by the imprint frame detection means;
Image dividing means for dividing the impression multi-value image to generate a plurality of partial impression multi-value images;
Each of the plurality of partially imprinted multi-valued images divided by the image dividing means is converted into a second second value that is different from the first binarization threshold based on the shade detected by the imprint frame shade detection means. A second binarization unit that binarizes using a binarization threshold and generates a plurality of partial imprint binary images;
Image synthesizing means for synthesizing a plurality of partial impression binary images generated by the second binarizing means to generate one synthesized binary image;
An imprint center detection means for detecting the center of the imprint binary image generated by the first binarization means;
Function as
The imprint frame grayscale detection means is based on the number of black pixels in the linear direction at the intersection of the straight line passing through the center detected by the imprint center detection means and the outer peripheral frame detected by the imprint frame detection means. detecting the density of the frame, sign shadow image generation program that is characterized in that. The image dividing means divides the imprint multi-value image by a straight line orthogonal to the straight line obtained from the intersection where the number of black pixels detected by the imprint frame shade detection means is the largest, and a partial imprint multi-value image The imprint image generation program according to claim 2 , wherein: Said image splitting means by a straight line passing through the center detected by the seal impression center detection means, said print image to generate a divided and partial imprint multivalued image multivalued image, according to claim 2, characterized in that Imprint image generation program. An imprint image generation method executed in an imprint image generation apparatus,
A seal imprint multi-valued image of a seal stamp imprinted on a medium read by the image reading device is binarized using a predetermined first binarization threshold value to generate an imprint binary image.
Detecting an outer peripheral frame of the generated imprint binary image,
Detecting the center of the generated imprint binary image;
Based on the number of black pixels in a predetermined region centered on the intersection of the detected straight line and the detected outer peripheral frame, the intensity of a plurality of locations of the detected outer peripheral frame is detected,
Dividing the impression multi-value image to generate a plurality of partial impression multi-value images;
Each of the divided partial impression multi-value images is binarized using a second binarization threshold different from the first binarization threshold based on the detected gray level. Generate multiple partial imprint binary images,
Combining the plurality of generated partial imprint binary images to generate one composite binary image;
An imprint image generation method characterized by the above. In an imprint image generating device,
An imprint multi-value image acquisition unit for acquiring an imprint multi-value image of a seal stamp stamped on a medium read by an image reading device;
A first binarization unit that binarizes the impression multi-value image acquired by the impression multi-value image acquisition unit using a predetermined first binarization threshold to generate an imprint binary image;
An imprint frame detection unit for detecting an outer peripheral frame of the imprint binary image generated by the first binarization unit;
An imprint center detection unit for detecting a center of an imprint binary image generated by the first binarization unit;
Detected by the imprint frame detection unit based on the number of black pixels in a predetermined area centered on the intersection of the straight line passing through the center detected by the imprint center detection unit and the outer frame detected by the imprint frame detection unit. An imprint frame shade detection unit for detecting shades at a plurality of locations on the outer peripheral frame;
An image dividing unit that divides the impression multi-value image to generate a plurality of partial impression multi-value images;
Each of the plurality of partially imprinted multi-value images divided by the image dividing unit is converted into a second second value that is different from the first binarization threshold value based on the shade detected by the imprint frame shade detection unit. A second binarization unit that binarizes using a binarization threshold and generates a plurality of partial imprint binary images;
An image combining unit that combines a plurality of partial impression binary images generated by the second binarization unit to generate one combined binary image;
An imprint image generating apparatus comprising: