JP2012048513A - Seal registration device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a seal registration device that automatically determines and records whether or not a seal impression of made a seal on a business form at the time of seal registration is "patchy" or "smudgy".SOLUTION: The seal registration device samples a midpoint and an outer peripheral frame center line of a seal impression on a business form from its image data, determines intersection points between a plurality of straight lines passing through the midpoint and radiating at an equal angle and the outer peripheral frame center line 12, sets inspection regions 14 in which the intersection point is the center pixel and each of which includes the same odd number of pixels n vertically and horizontally, and sequentially increases n to sequentially expand the inspection regions 14. At the point when white pixels 17 come to exist outside the outer peripheral frame beyond the outer peripheral frame center line 12 in all of the inspection regions 14, the seal registration device counts black pixels 18 positioned outside including the outer peripheral frame center line 12, compares the number of black pixels 18 with upper and lower limit values from a determination table. The device determines the image is "smudgy" when the number of black pixels exceeds the upper limit value and "patchy" when the number is smaller than the lower limit value.

Description

  The present invention relates to a seal registration apparatus in a financial institution, and more particularly to a seal registration apparatus that automatically determines whether there is any “blur” or “smear” in the seal imprinted on a form at the time of seal registration.

  Conventionally, in a financial institution such as a bank or a securities company, when a seal stamp of a seal stamped on a form is read with a scanner and the read seal stamp is stored as image data in a seal stamp database, the operator's Check if the seal is correct by visual judgment.

  In general, the seal stamped on a form may of course be stamped in the correct shape, but because it is affected by the shades of used vermilion and the shade of the form, it is a shape with `` shading '' or `` bleeding ''. Often imprints also occur.

  If the operator does not notice this and registers a seal with faintness or blurring, the seal imprinted for identity verification will be compared with the seal of the image data registered in the seal database at a later date. When performing seal stamp verification, the accuracy of verification decreases, and inconveniences such as the operator determining that the match mark is a difference mark occur.

  Regarding the verification of the stamped image data, the type of the image included in the partial area extracted from the image of the image data including a plurality of pixels is detected, and a predetermined pixel value among the pixels constituting the partial area is detected. There has been proposed an image processing apparatus that calculates the ratio of pixels and detects the type of image with reference to a table that defines a correspondence relationship with the ratio of pixels having pixel values corresponding to the type of image to be detected. (For example, refer to Patent Document 1.)

  In addition, the seal stamp matching device calculates the stamp matching rate based on the density gradient vector distance formed by the density gradient and the density gradient direction of each of the registered seal stamp image and the matching seal stamp image. There has been proposed a seal verification device that is not easily affected by fluctuations in the thickness or blurring. (For example, see Patent Document 2.)

JP 2006-148578 A JP 2001-202514 A

  However, the prior art disclosed in Patent Document 1 merely detects the type of image by a simple method of collating the number of black pixels in a predetermined area with the number of pixels of image information prepared in advance. However, since it does not detect blurring or blurring of the seal imprinted at the time of registering the seal stamp, there is a problem to be solved in order to apply it to the verification of the seal stamp.

  Further, the prior art disclosed in Patent Document 2 uses the distance of the density gradient vector in order to perform seal verification that is not easily affected by fluctuations in thickness or blurring of the line segment of the seal. Since it does not detect blurring or blurring of the stamp imprinted at times, it can be applied to automatic verification, but there is a problem that it does not solve verification error due to visual recognition by the operator.

  The present invention solves the above-described conventional problems, and automatically determines whether the seal imprint imprinted on the form at the time of registering the seal has “blur” or “smear”. It is an object of the present invention to provide a seal registration apparatus that records information that enables re-registration of the seal at a later date together with registered seal image data when there is “smear”.

  In order to solve the above-mentioned problems, the seal registration apparatus of the present invention includes a midpoint and outer peripheral frame centerline extraction means for extracting the midpoint of the image data of the seal stamp imprinted on the form and the centerline of the outer frame. The intersection acquisition means for acquiring the intersections of the straight lines that pass through the midpoint and radiate at an equal angle and the center line of the outer peripheral frame, and the odd-numbered pixels n that are the same vertically and horizontally with the intersection as the central pixel for each intersection. Inspection area setting means for setting an inspection area, determination means for determining whether or not there is a white pixel in the outer direction of the outer peripheral frame from the outer peripheral frame center line in all of the inspection areas, and the white pixel When there is the inspection area that does not exist, the inspection area expanding means for increasing the value of the odd pixel n until the white pixel is present in all of the inspection area, and the white pixel is present in all of the inspection area. When the inspection area Black pixel counting means for counting the number of black pixels in the outer peripheral frame center line and the outer direction, and upper and lower limit values indicating whether the black pixel number is a value indicating blur or a blur. The table provided corresponding to the inspection area of the value of the odd pixel n is compared with the upper and lower limit values of the table corresponding to the inspection area where the number of black pixels is counted. Determining means for determining whether the inspection area is blurred or blurred.

  In this seal registration apparatus, for example, the table includes a table corresponding to the inspection region in which a collection of pixels constituting the outer peripheral frame center line extends vertically or horizontally in the inspection region, and an outer periphery in the inspection region. A group of pixels constituting the frame center line is configured to be divided into a table corresponding to the inspection region extending in an oblique direction or L-shaped.

  Further, in this seal registration apparatus, for example, when the determination means determines that the inspection area is blurred or blurred, the image data of the seal stamp imprinting the inspection area with a mark indicating blur or blurred is given. In addition, the display device is configured to display on the display device an instruction display for instructing input from the outside so as to give summary information indicating that a new seal stamp should be recollected.

  In this case, for example, when there is an input from outside to give the summary information to the instruction display displayed on the display device by the display means, the summary information is given to the image data, The image data to which the summary information is added is registered in the predetermined seal database as the registration data of the seal, and should be given, for example, when there is an input from outside to register without giving the summary information Information indicating that the summary information has not been assigned is added to the image data, and the image data is registered in a predetermined seal database as registration data of the seal.

  The present invention sets a plurality of inspection areas based on the center point extracted from the image data of the seal stamp imprinted on the form using known imprint image data detection software and the center line of the outer peripheral frame. Since the blurring or blurring of the imprint is detected by a unique method of counting black pixels outside the outer peripheral frame center line of the area, there is an effect that the blurring or blurring of the imprint at the time of seal registration can be detected quickly.

  In addition, when there is blurring or blurring in the seal of the registered registration seal, the present invention adds the summary information indicating that to the registered image data of the seal and registers it as a seal. When collating with the seal stamp, it is possible for the person in charge of the window to know whether or not there is summary information and to perform correct seal collation.

It is a figure which shows typically the whole structure of the seal registration management system provided with the seal registration apparatus which concerns on Example 1 of this invention. (a), (b), (c) is a figure for demonstrating plainly the judgment criteria of the blurring and blurring of the seal stamp of the seal stamp of the stamp registration apparatus which concerns on Example 1. FIG. It is a figure which shows the image data of the imprint of the normal shape which was correctly stamped on the seal stamp in the seal registration apparatus which concerns on Example 1, and was read with the image reader. 6 is a flowchart showing a processing operation of the seal registration apparatus for determining whether the seal imprint has blur or blur from the image data of the seal stamp imprinted on the seal slip in the seal registration apparatus according to the first embodiment. 6 is a flowchart showing a processing operation of the seal registration apparatus for determining whether the seal imprint has blur or blur from the image data of the seal stamp imprinted on the seal slip in the seal registration apparatus according to the first embodiment. It is a figure which shows the example of a display of the seal registration screen displayed on the input display screen of the seal registration machine of the seal registration apparatus which concerns on Example 1. FIG. In the seal registration apparatus according to the first embodiment, (a) is a diagram illustrating an example of image data of a stamped impression, and (b) is a diagram illustrating a midpoint and an outer peripheral frame center line extracted from the image data. is there. (a) The middle point is the origin of coordinates, and the intersection of the x axis and the outer peripheral frame center line and the intersection of the y axis and the outer peripheral frame center line are the centers of the first, third, fifth and seventh odd-numbered inspection areas. FIG. 5B is a diagram showing an example extracted as a point. FIG. 5B shows the intersection of the straight line y = (n / m) x and the outer peripheral frame center line and the straight line y = (− n / m) x with the middle point as the origin of coordinates. It is a figure which shows the example which extracted the intersection of an outer periphery frame centerline as a center point of the 2, 4, 6, 8th even-numbered test area | region. It is a table which shows the inspection pixel size corresponding to the set value n and the set value n of the size of the inspection area. (a) is a figure which shows eight inspection area | region center points extracted by FIG. 7 (a), (b), (b) is a figure which shows the inspection area | region centering on the center point. It is a figure which expands and shows pixel distribution used as the inspection object of each inspection area when setting value n = 5 of an inspection area is set. It is a figure explaining the reason why the number of black pixels inside the outer peripheral frame center line of each inspection area is not counted. It is a figure which expands and divides the pixel distribution used as the test object of each inspection area when the set value n = 1, 3, 5, 7 of the inspection area is divided into even-numbered areas and odd-numbered areas. It is a figure which shows a state when the setting value n of the magnitude | size of a test | inspection area | region is enlarged sequentially and a white pixel is included in the test object pixel area | region of all the test | inspection areas. In the limit value judgment, (a) is a judgment table indicating normal, faint, and blur judgment criteria and the judgment results, (b) is a judgment factor setting table 1 for odd-numbered test areas, and (c) is for even-numbered test areas. The determination factor setting table 2 is shown. In the ratio determination, (a) is a determination table indicating the determination criteria of normal, blurring and blurring and the determination result, and (b) is a table for ratio determination group A for the odd-numbered inspection area and the even-numbered inspection area of the determination factor setting table. It is a table for ratio determination group B for use. It is a figure which shows the data structure of the determination result table which stores a determination result. It is a figure which shows the example when specific data are stored in the determination result table. It is a figure which shows the example of the guidance screen displayed with respect to an operator by the input display screen of a seal stamp registration machine, when "blur" or "smear" exists in the image data of the seal impression. It is a figure which shows the example of the message screen displayed with respect to an operator by the input display screen of a seal stamp registration machine, when the summary information of "stamp form recollection" is added to the seal stamp registration data searched from the seal stamp database. .

  Hereinafter, embodiments of the present invention will be described in detail.

  FIG. 1 is a diagram schematically illustrating an overall configuration of a seal registration management system including a seal registration apparatus according to Embodiment 1 of the present invention. As shown in FIG. 1, the seal registration management system of the present example includes a seal stamp management server 2 having a seal stamp database 1 composed of a large-capacity storage device and the like, and a wireless or wired communication network 3 connected to the seal stamp management server 2. It is comprised by the seal registration machine 4 connected through this.

  A display device 4-1, a keyboard 4-2, a mouse 4-3, and an image reader 5 are connected to the seal registration machine 4, and the seal registration device 6 of this example is configured as a pair. Yes. The seal registration device 6 is installed at a seal registration window of a bank or a securities company, for example.

  2 (a), 2 (b), and 2 (c) show imprints of seals that are stamped on forms (hereinafter referred to as seal slips) that are presented to a financial institution in order to register the seal by the seal registration device 6. It is a figure which shows the example of.

  FIGS. 2A, 2B, and 2C are diagrams for easily explaining the judgment criteria for blurring and blurring of a stamped seal. In the following description, the bank seal “Tanabe” is taken as an example.

  FIG. 2 (a) is an example of fading (thin), and shows a case where the outer frame line of the imprint is discontinuous and cut by a certain amount or more. The reason why the discontinuity is not a discontinuous and the discontinuity is a certain amount or more is that a seal with a missing edge may be used.

  Further, in the blurring judgment criterion, even a continuous outer frame line includes a thin outer frame line than an actual seal. In this case, it is determined that a small number of black pixels in the inspection region of the image data of the seal impression is faint in comparison with the statistical data.

  FIG. 2 (b) shows a properly imprinted imprint that is correctly imprinted with no blur or blur. As for the thickness of the actual outline of the seal, it has been found that the ratio between the length of the outline of the seal and the thickness of the line is statistically within a certain value. Therefore, if the thickness of the outer frame line of the imprint is within a certain value, it is determined that the imprint has a normal shape.

  FIG. 2 (c) is an example of blurring (dark), and the outer frame line of the imprint is thicker than the actual seal. Also in this case, if the thickness of the outer frame line exceeds a certain value in light of the above statistical data, it is determined that the image is blurred. That is, it is determined that the number of black pixels in the inspection area of the image data of the seal imprint is large in light of the statistical data.

  FIG. 3 is a diagram showing the imprint image data 10 when the image reader 5 reads the normal shape imprint shown in FIG. The imprint image data 10 is formed of a large number of black pixels. A large number of projections and depressions having an angle of 90 degrees in the imprint image data 10 indicate that each pixel has a square shape.

  FIG. 4 shows a seal registration machine 4 that determines whether there is any blur or blur in the seal stamp of the seal stamp stamped from the seal stamp image data shown in FIG. It is a flowchart which shows the processing operation of the seal | sticker registration apparatus 6 discriminate | determined by these.

  This process is a process performed by the control unit (not shown) of the seal registration machine 4 in the seal registration apparatus 6. In this process, it is assumed that a seal stamp with a seal imprinted in advance is presented to an operator at a window of the seal registration machine 4 and that the seal is read by the image reader 5 by the operator.

  In FIG. 4, first, as shown in FIG. 5, the control unit displays a seal registration screen 8 displaying a seal image 7 read from the seal slip by the image reader 5 on the display device 4-1 of the seal registration machine 4. Then, it is determined whether or not there is a seal registration instruction from the operator, that is, whether or not the registration button 9 is pressed (step S1).

  When there is no seal registration instruction, that is, when the registration button 9 is not pressed (No in S1), the process waits until there is a seal registration instruction. When there is a seal registration instruction (Yes in S1), the control unit detects the midpoint and outer peripheral frame center line of the registered seal image data (step S2).

  This process is performed from the image data of the seal image 7 displayed on the seal registration screen 8 in FIG. 5, that is, from the registered seal image data (hereinafter also referred to as seal image data) 10 shown in FIG. For example, as shown in FIG. 6B, an imprint image is obtained by using a part of the function of a seal collation program that is used in, for example, Japanese Unexamined Patent Application Publication No. 2003-178199 or Japanese Unexamined Patent Application Publication No. 2005-165383. This is processing for detecting the midpoint 11 of the data 10 and the center line 12 of the outer peripheral frame.

  Here, the middle point 11 is one pixel which is the minimum unit of the image data, and the center line 12 of the outer peripheral frame is an aggregate of one continuous pixel.

  Subsequently, the control unit extracts the center point of the inspection region (step S3). In this process, eight center points to be used as the inspection region are extracted from the midpoint 11 detected as described above and the center line 12 of the outer peripheral frame.

  In the extraction processing of the center points of the eight inspection regions, first, as shown in FIG. 7A, the middle point 11 is set as the origin of the xy coordinates, and the intersection 13− between the x axis and the center line 12 of the outer peripheral frame. 3 and 13-7, and the intersections 13-1 and 13-5 between the y axis and the center line 12 of the outer peripheral frame are extracted as center points.

  Subsequently, as shown in FIG. 7B, the midpoint 11 is the origin of the xy coordinates, the radius of the center line 12 of the outer peripheral frame in the x-axis direction from the midpoint 11 is m, and the midpoint 11 is in the y-axis direction. Where the radius of the center line 12 of the outer peripheral frame is n, the intersection points 13-2 and 13-6 of the straight line y = (n / m) x and the center line 12 of the outer peripheral frame, and the straight line y = (− n / m) Intersection points 13-4 and 13-8 between x and the center line 12 of the outer peripheral frame are extracted as center points.

  As a result, the center points 13 (13-1, 13-2,..., 13-8) of the eight inspection regions are extracted. Hereinafter, the center points 13 of these eight inspection areas are sometimes referred to as first center point, second center point,..., Eighth center point, respectively.

  Next, the control unit sets an initial value, where n is the inspection area size (step S4). In this process, as shown in the chart of FIG. 8, a table indicating the set value n of the size (size) of the inspection area and the number of inspection target pixels “n × (n + 1) / 2” corresponding to the set value n is used. Is done.

  The inspection area size n is a value that represents the length (number of pixels) of one side of the inspection area. Since the inspection area is composed of a square pixel area having the center point 13 as a central pixel, n is an odd number of 1 or more. Will be set.

  The reason for obtaining the number of pixels to be inspected (counting object) by “n × (n + 1) / 2” is that the number of inspected pixels in each inspection region is the center of the outer peripheral frame including the center point 13 in this example. This is because it is outside the line 12. The reason why the number of pixels to be inspected is outside the center line 12 of the outer peripheral frame including the center point 13 will be described later.

  Following the above, the control unit extracts an inspection region (step S5). In this process, an inspection area corresponding to the inspection area size n is extracted for each center point 13.

  FIG. 9 (a) shows the center points 13 (13-1, 13-2,..., 13-8) of the eight inspection regions extracted in FIGS. 7 (a) and (b). FIG. 9B shows eight inspection regions 14 (14-1, 14-2,..., 14-8) extracted based on the inspection region size n for each of the center points 13. ing.

  FIG. 10 shows an enlarged pixel distribution of each inspection region 14 when the inspection region size n = 5. One inspection region 14 is composed of n × n pixels (in this case, a total of 25 pixels of 5 × 5).

  As shown in FIG. 10, the inspection area 14 to be extracted is formed of a square in which n × n (in this case, 5 × 5) pixels with the center point 13 at the center of the area are gathered. A broken line 12 'indicates a pixel group of the center line 12 of the outer peripheral frame.

  For each inspection region 14, 15 colored pixels including one black color and 14 mouse colors are the center line 12 of the outer peripheral frame including the center point 13 and the inspection target pixel 15 outside the center line 12. The determination of blur or blur is performed by counting the number of black pixels of the center line 12 of the outer peripheral frame including the center point 13 and the inspection target pixel 15 outside the center line 12.

  Pixels on the inner side of the center line 12 of the outer peripheral frame of each inspection region 14 (shown in white in order to be distinguished from the pixels to be counted) are non-inspection pixels 16 that are not used for the above determination.

  The reason why the pixels inside the center line 12 are excluded from the inspection target is that the inner side of the center line 12 in the inspection area 14 is a portion surrounded by a circle a in the imprint image data 10 in FIG. The name and outer frame line overlap to increase the number of black pixels. That is, the inner side of the center line 12 of the outer peripheral frame may be in contact with the character of the seal, so it is not suitable for inspection and is excluded from the determination. It is.

  Following the above, the control unit counts the number of black pixels of the inspection target pixel 15 in each inspection region 14 (step S6). And it is discriminate | determined whether a white pixel exists in the test object pixel 15 of all the test | inspection areas 14 (step S7).

  If no white pixel is present in the region of the inspection target pixel 15 (No in S7), there is a possibility that the outer frame line of the seal is not fit in the inspection region 14, that is, the outer frame line is thicker. is there.

  Therefore, in order to expand the inspection region 14 so that the white pixel appears in the region of the inspection target pixel 15 in all the inspection regions 14, “n ← n + 2” is set (step S8), the process returns to step S5, and steps S5 to S5 are performed. The process of S7 is repeated until white pixels appear in all inspection areas.

  FIG. 12 typically shows the inspection areas 14-1 and 14-2 in the case where the inspection area size n is sequentially increased and expanded as 1, 3, 5, and 7, and The pixel distribution is shown. Also in this case, the broken line 12 ′ indicates the pixel group of the center line 12 of the outer peripheral frame, the inspection target pixel 15 is indicated by a colored pixel, and the non-inspection pixel 16 is indicated by a white pixel.

  As shown in FIG. 12, in expanding the inspection area 14, the center point 13 is placed at the center of the inspection area 14, and is expanded pixel by pixel vertically and horizontally. The inspection area 14 is expanded until the inspection area size n reaches a separately determined maximum value.

  In FIG. 13, when the inspection area size n for the imprint image data 10 is increased to n = 7, the white pixel 17 exists in the inspection area including the inspection target pixels 15 of all the inspection areas 14. This shows the state (Yes in S7).

  In FIG. 13, there are 14 black pixels 18 and 14 white pixels 17 in the inspection target pixel 15 in the inspection region 14-1. In addition, the number of black pixels 18 and the number of white pixels 17 in the inspection target pixel 15 in the inspection region 14-2 are six.

  Here, the inspection area 14-1 indicates that the white pixel 17 is present in the area when the inspection area size n reaches n = 5, and the inspection area 14-2 also has the inspection area size. This shows that the white pixel 17 is present in the region when n becomes n = 5.

  The inspection area 14-4 indicates that the white pixel 17 is present in the area when the inspection area size n reaches n = 7. That is, in this state, it is indicated that the white pixel 17 is present in the region of the inspection target pixel 15 in all the inspection regions 14 (hereinafter also referred to as the inspection target region 15).

  Here, the control unit performs limit value determination as blur / bleed determination 1 (step S9). In this process, the number of black pixels in the inspection target area 15 is counted, and limit value determination is performed based on the determination table.

  The registered seal image data (see seal image data 10 in FIGS. 3, 6A, 11 and 13) is binary data, and as shown in FIG. The area 15 is composed of pixels (dots) of black pixels 18 and white pixels 17.

  In the limit value determination in step S9, the number of black pixels 18 in the inspection target region 15 is counted, and limit value determination is performed based on the determination criterion. In this limit value determination, the inspection area is classified into two based on the characteristics of the seal stamp.

  One is that an aggregate of pixels constituting the center line 12 of the outer peripheral frame extends vertically or horizontally in the inspection region. This corresponds to the odd-numbered inspection areas 14 of the inspection areas 14-1, 14-3, 14-5, and 14-7.

  The other is that the collection of pixels constituting the center line 12 of the outer peripheral frame in the inspection region extends in an oblique direction (in the case of a circular or elliptical seal), or an L-shaped (square or rectangular shape). In the case of a seal). This corresponds to the even-numbered inspection areas 14 of the inspection areas 14-2, 14-4, 14-6, and 14-8.

  FIG. 14A shows a determination table referred to in limit value determination. In the determination table shown in FIG. 14A, the value obtained by counting the number of black pixels in the inspection target region 15 is indicated by x, the limit pixel number for blur determination is indicated by a, and the limit pixel number for blur determination is indicated by b. These a and b are called determination factors.

  FIG. 14B shows the limit pixel number a for blur determination and the limit pixel number b for blur determination corresponding to each inspection area size n used in the determination table of FIG. The determination factor setting table 1 which shows the setting value of is shown.

  FIG. 14C shows the limit pixel number a for blur determination and the limit pixel number b for blur determination corresponding to each inspection area size n used in the determination table of FIG. The determination factor setting table 2 which shows the setting value of is shown.

  In the determination of the odd-numbered inspection area 14, the values of the determination factors a and b of the determination factor setting table 1 shown in FIG. 14B are used in the determination table of FIG. The values of the determination factors a and b in the determination factor setting table 2 shown in FIG. 14C are used in the determination table shown in FIG.

  In any case, regarding the imprint of the inspection region 14, if “a ≦ x ≦ b”, normal, “x <a”, faint, “b <x”, It is determined that it is blurred.

  When all the inspection areas 14-1 to 14-8 are "normal", the whole determination is "normal" (no blurring / bleeding). In all other cases, the overall judgment is “abnormal”. In this limit value determination, the inspection area size is set to “n = 5” or more.

  Therefore, in the setting of the initial value of the inspection area size in step S4, “n ← 5” is used. Moreover, each value of the determination factors a and b is a set value based on the implementation result of the present invention. In actual operation, each value is selected based on actual data (statistical data).

  Following the above determination, the control unit performs ratio determination as blur / bleed determination 2 (step S10). In this process, the number of black pixels 18 in the inspection target region 15 is counted, the average value of the count values is compared with the count value of each inspection region, and the ratio is determined based on the determination criterion of the determination table.

  This ratio determination process is performed as a method to compensate for the case where both blurring and blurring are present in the seal stamp image data 10 of a single seal and cannot be determined only by the limit value determination.

  That is, generally, the outer peripheral frame of the seal has a constant thickness. If the number of black pixels 18 counted in the inspection varies greatly depending on the inspection area 14, there is a possibility that blur or blur exists.

  In some cases, for example, a portion determined to be “bleed” is actually an appropriate thickness, and a portion determined to be normal is actually faint. However, since the present invention finally notifies the operator of an abnormality (the presence of “fading” or “bleeding”) and prompts a countermeasure, the following ratio determination method is employed. This ratio determination is also made by classifying the inspection area into two based on the characteristics of the seal.

  One is that an aggregate of pixels constituting the center line 12 of the outer peripheral frame extends vertically or horizontally in the inspection region. This corresponds to the odd-numbered inspection areas 14 of the inspection areas 14-1, 14-3, 14-5, and 14-7.

  The other is that the collection of pixels constituting the center line 12 of the outer peripheral frame in the inspection region extends in an oblique direction (in the case of a circular or elliptical seal), or an L-shaped (square or rectangular shape). In the case of a seal). This corresponds to the even-numbered inspection areas 14 of the inspection areas 14-2, 14-4, 14-6, and 14-8.

  FIG. 15 (a) is a determination table showing normal, blurred, and blurring determination criteria used in the ratio determination and the determination result. FIG. 15 (b) is a ratio for the odd-numbered test region in the determination factor setting table. It is a determination group A table and a ratio determination group B table for even-numbered inspection areas.

  Note that “*” in the determination table of FIG. 15A represents x (multiplication). In this determination table, x is the value obtained by counting the number of black pixels 18 in the inspection target area 15, and the average value of the values obtained by counting the number of black pixels 18 in the inspection target area 15 (four in the inspection group). The average value of the number of black pixels 18 in the inspection region 14) is indicated by y, the reference for blur determination is indicated by c, and the reference for blur determination is indicated by d. These c and d are called determination factors.

  In this ratio determination method, as shown in FIG. 15 (b), the determination is made separately for the odd-numbered inspection area (ratio determination group A) and the even-numbered inspection area (ratio determination group B). Then, for each group, determination based on the determination table is performed for each of the four inspection regions.

  Even in this determination, if all the inspection areas 14-1 to 14-8 are “normal”, the overall determination is “normal” (no blur / bleed). In all other cases, the overall judgment is “abnormal”.

  Each value of the determination factors c and d is a set value based on the result of implementation of the present invention. In actual operation, each value is selected based on actual data (statistical data).

  Following the above determination, the control unit stores the determination result in the determination result table (step S11). In this process, the determination result indicating “normal” or “abnormal” and the details of the determination result of “normal”, “smear” or “smear” of each inspection region 14 are stored in the determination result table.

  FIG. 16 is a diagram illustrating a data configuration of the determination result table. As shown in FIG. 16, the determination result table 20 includes records including a series of data areas, and includes an operation date / time area 21, a registration number area 22, a determination result area 23, a determination result detail 1 (limit value determination) area 24, The determination result detail 2 (ratio determination) area 25, the determination data area 26, and the summary setting area 27 are configured.

  The date column of the operation date / time area 21 stores the date when the registration operation was performed. The time column stores the time when the registration operation was performed.

  The store number column of the registration number area 22 stores the store number where the registration operation was performed. The subject for which the registration operation has been performed is stored in the subject column. In the number column, the account number for which the registration operation has been performed is stored. The CIF (customer number) column stores the CIF number for which the registration operation has been performed.

  In the result column of the determination result area 23, “0” is stored if normal, and “1” is stored if abnormal. When the result is normal (“0”), “0” (normal) is stored in the type column. When the result is abnormal (“1”), “1” (with blur), “2” (with blur), and “3” (with blur and blur) are stored.

  In the determination result detail 1 (limit value determination) area 24, “0” (normal), “1” (with fading), and “2” (with blur) are stored, respectively. In the determination result detail 2 (ratio determination) area 25, “0” (normal), “1” (with fading), and “2” (with blur) are stored, respectively.

  The size field of the determination data area 26 stores the area size (n) at the time of determination. In the first to eighth columns, the number of black pixels (counted number) of each inspection area 14 is stored. When the result is normal (“0”), “0” (normal) is stored in the summary setting area 27. When the result is abnormal (“1”), “1” (summary setting completed) and “2” (unset) are stored according to the operator's operation.

  FIG. 17 is a diagram illustrating an example of determination results stored in the determination result table 20 described above. Both the limit value determination and the ratio determination are determined to be normal in all the inspection regions 14, and the first to eighth columns of the determination result detail 1 (limit value determination) region 24 and the first to eighth columns of the determination result detail 2 (ratio determination) region 25. Shows an example in which all “0” are stored.

  In this case, is the judgment result “normal”? Since the determination of (Step S12) is Yes, the seal registration process is executed (Step S13). In this process, the determination result table 20 shown in FIG. 17 is stored in the seal stamp database 1 as seal stamp data.

  On the other hand, if it is determined in step S12 that the result is not “normal”, that is, “1” is stored in the result column of the determination result area 23 of the determination result table 20 (No in S12), the determination result table Referring to columns 1 to 8 of 20 determination result details 1 (limit value determination) region 24 and determination result details 2 (ratio determination) region 25 (step S14), the occurrence location of “blur” or “smear” is displayed. (Step S15).

  In this process, the guidance screen shown in FIG. 18 is output, a message 28 indicating that “blur” or “smear” is present on the screen, and the display 29 of the occurrence location are displayed, and further the setting of summary information is prompted. A message 31 is displayed to notify the operator.

  At this time, the operator can press (click, and so on) the “print” button 32 as necessary. When recognizing that the “print” button 32 has been pressed, the control unit prints out the occurrence location information (occurrence location display 29) displayed on the screen (step S 16). .

  Next, the control unit determines whether or not there is a summary information addition instruction from the operator (step S17). This process is a process for determining whether or not the “set” button 33 is pressed by the operator.

  If the “set” button 33 is pressed (Yes at S17), “recollection of seal stamp” is added to the registered data as summary information (step S18).

  Subsequently, the control unit updates the determination result table (step S19). This process is a process of storing “1” in the summary setting area 27 of the determination result table 20. Then, the seal stamp data and the determination result table 20 to which the summary information is added in the process of step S13 described above are stored in the seal stamp database 1.

  When the operator searches for the seal data, the summary information given together with the seal data is read out, and as shown in FIG. 19, a “seal stamp recollection” message 36 is displayed on the display screen 34 together with the seal image data 35. Is displayed. As a result, the operator can recognize that it is a seal with “smear” or “smear” and can deal with it accordingly.

  In addition, the determination result table 20 indicates that the operator has pressed the “set” button 33 on the guidance screen of FIG. 18 in the seal registration operation operation, or has completed the registration operation for the determination result of the registered seal data. Use it to check later.

  On the other hand, if the “not set” button 37 is pressed on the display screen of FIG. 18 (No in S17), the control unit updates the determination result table (step S20). In this process, “2” indicating that “summary information has not been added by the operator's judgment” is stored in the summary setting area 27 of the judgment result table 20.

  In this way, the registration operation of the seal image at the time of registering the seal stamp automatically checks for blurring or blurring before registering in the stamp database, and if blurring or blurring is detected, blurring or blurring occurs. The location is indicated to the operator to call attention, and the user is prompted to set incidental information (summary information) for the seal image to be registered.

  As described above, according to the present embodiment, it is automatically determined whether the stamp imprint of the seal registered in the seal stamp management system has “blur” or “smear”, and “smear” or “smear” exists. In such a case, the operator is informed so that the customer can re-register the seal slip from the customer.

  As a result, the imprint image data that has “blur” or “smear” in the seal at the time of stamping the seal is excluded from the seal database of the seal management system, and the registration quality of the seal image data registered in the seal management system is reduced. Can be improved.

  In this way, the registration quality of seal image data of seals registered in the seal management system is improved, so that misjudgment at the time of seal verification is prevented in financial institutions, etc. Efficiency is improved.

  Similarly, since the registration quality of seal stamp image data registered in the seal stamp management system is improved, it is possible to contribute to the improvement of collation accuracy of seal stamp automatic collation that has been spreading in recent years.

  The present invention can be used for a seal registration apparatus that automatically determines whether there is “blur” or “smear” in the seal of a seal stamped on a form at the time of seal registration.

DESCRIPTION OF SYMBOLS 1 Seal database 2 Seal management server 3 Communication network 4 Seal registration machine 4-1 Display device 4-2 Keyboard 4-3 Mouse 5 Image reader 6 Seal registration device 7 Seal image 8 Seal registration screen 9 Registration button 10 Registration seal image data ( Imprint image data)
11 Midpoint 12 Centerline of outer peripheral frame 13 (13-1, 13-2,..., 13-8) Inspection area center point 14 (14-1, 14-2,..., 14-8) Inspection Region 15 Inspection target pixel (Inspection target region)
16 Non-inspection pixel 17 White pixel in inspection object pixel 18 Black pixel in inspection object pixel 20 Judgment result table 21 Operation date / time area 22 Registration number area 23 Judgment result area 24 Judgment result details 1 (limit value judgment) area 25 Judgment Result detail 2 (ratio determination) area 26 Judgment data area 27 Summary setting area 28 Message of “hazing” or “blowing” 29 Display of the occurrence of “hazing” or “blotting” 31 Message for prompting setting of hailing information 32 “Print” button 33 “Set” button 34 Display screen 35 Imprint image data 36 “Recollect stamp” message 37 “Do not set” button

Claims (4)

A midpoint and outer peripheral frame centerline extracting means for extracting the midpoint of the image data of the seal stamp imprinted on the form and the centerline of the outer peripheral frame;
An intersection acquisition means for acquiring an intersection of a plurality of straight lines radiating at an equal angle through the midpoint and the outer peripheral frame centerline;
Inspection area setting means for setting an inspection area consisting of odd-numbered pixels n having the same intersection in the center pixel for each intersection;
Discriminating means for discriminating whether or not there is a white pixel in the outer direction of the outer peripheral frame from the outer peripheral frame center line in all of the inspection area;
Inspection area expanding means for increasing the value of the odd pixel n until the white pixels are present in all of the inspection areas when there is the inspection area in which the white pixels do not exist;
Black pixel counting means for counting the number of black pixels in the outer peripheral frame center line and the outer direction for each inspection region when the white pixels are present in all of the inspection region;
A table having upper and lower limit values indicating whether the number of black pixels is a value indicating blur or a value corresponding to blurring, corresponding to the inspection area of the value of the odd pixel n,
A determination means for comparing the black pixel count with the upper and lower limit values of the table corresponding to the inspection area where the black pixel count is counted to determine whether the inspection area is blurred or blurred;
A seal registration apparatus characterized by comprising:   The table includes a table corresponding to the inspection region in which an aggregate of pixels constituting the outer peripheral frame center line extends vertically or horizontally in the inspection region, and a pixel constituting the outer peripheral frame center line in the inspection region. 2. The seal registration apparatus according to claim 1, wherein the assembly is divided into a table corresponding to the inspection area extending in an oblique direction or having an L shape.   When it is determined by the determining means that the inspection area is blurred or blurred, a new imprint should be recollected together with the image data of the seal stamp with a mark indicating the bleeding or blurred in the inspection area. 2. The seal registration apparatus according to claim 1, further comprising display means for displaying on the display device an instruction display instructing input from the outside so as to give summary information indicating that. For the instruction display displayed on the display device by the display means,
When there is an external input to give the summary information, the summary information is given to the image data, and the image data to which the summary information is given is registered in a predetermined seal database as registration data of the seal And
When there is an input from the outside to register without adding the summary information, information indicating that the summary information to be added is not given is given to the image data, and the image data is added to the seal Register it as a registered data in a predetermined seal database,
4. The seal registration apparatus according to claim 3, further comprising registration means.