Judgment system, judgment method, and program

Abstract

To provide a determination system capable of improving thickness determination accuracy of a card-like object.SOLUTION: A determination system according to the present embodiment comprise edge detection means, thickness determination means, and imaging state determination means. The edge determination means detects the edges accompanied by the object including two edges at the bottom or top of the object from the image taken of a card-like object. The thickness determination means determines whether or not the object thickness is appropriate from the distance of two edges. The imaging state determination means determines whether or not the object is imaged in a state suitable for the object thickness determination using the detected edge information by the edge detection means.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] Embodiments of the present invention relate to a determination system, a determination method, and a program. Background Art

[0002] In recent years, systems for capturing images of identity verification documents such as driver's licenses and My Number Cards using smartphones with camera functions, tablet PCs (Personal Computers) or the like and importing them as images have been widely used.

[0003] Recently, according to the Act on Prevention of Transfer of Criminal Proceeds, it is required to check the thickness of an identity verification document in a captured image of the identity verification document. A driver's license or a My Number Card has a thickness of about 0.76 mm, while altered versions of these documents produced using a printer or the like are often printed on thin paper of about 0.2 mm in thickness. Prior Art Documents Patent Documents

[0004] Patent Document 1 Japanese Unexamined Patent Application Publication No. 2020-77420 Patent Document 2 Japanese Unexamined Patent Application Publication No. 2020-77421 Patent Document 3 Japanese Unexamined Patent Application Publication No. 2019-8580 Summary of the Invention Problems to be Solved by the Invention

[0005] However, even for an altered identity verification document with a thickness of about 0.2 mm, when the document is placed on a table or the like and imaged from an oblique angle to enable determination of the thickness of the identity verification document, if a shadow is formed near the edge involved in thickness determination of the identity verification document on the table, in conventional systems, a length including the width of the shadow is calculated as the thickness of the identity verification document, which may result in an erroneous passing of the thickness determination.

[0006] One embodiment of the present invention provides a determination system, determination method, and program that can improve the accuracy of determining the thickness of a card-shaped object. [Means for solving the problem]

[0007] According to one embodiment, the determination system comprises edge detection means, thickness determination means, and imaging state determination means. The edge detection means detects edges relating to an object, including two edges at the bottom or top of the object, from an image of a card-shaped object. The thickness determination means determines whether the thickness of the object is appropriate based on the distance between the two edges. The imaging state determination means uses the edge information detected by the edge detection means to determine whether the object is being imaged in a state suitable for determining its thickness. [Brief explanation of the drawing]

[0008] [Figure 1] A diagram showing one example configuration of the identity verification system of the first embodiment. [Figure 2] Figure 1 shows an example of the configuration of the identity verification document thickness determination unit. [Figure 3] This diagram illustrates an example where the image quality of the identification document is not suitable for determining its thickness. [Figure 4] A diagram illustrating the method for capturing images of identity verification documents in the identity verification system of the first embodiment. [Figure 5] A diagram illustrating edge detection from an image by the edge detection unit of the identity verification system of the first embodiment. [Figure 6] Figure 1 illustrates the detection of a hand by the hand detection unit of the identity verification system of the first embodiment. [Figure 7] Figure 2 illustrates the detection of a hand by the hand detection unit of the identity verification system of the first embodiment. [Figure 8] A flowchart illustrating the procedure for detecting a hand by the hand detection unit of the identity verification system in the first embodiment. [Figure 9]FIG. 1 is a diagram for explaining thickness determination performed by a thickness determination unit 23 of the identity verification system according to the first embodiment. [Figure 10] FIG. 3 is a diagram for explaining edge detection performed by an edge detection unit of the identity verification system according to the first embodiment. [Figure 11] FIG. 6 is a diagram showing a configuration example of an identity verification document thickness determination unit in the identity verification system according to the second embodiment. [Figure 12] FIG. 9 is a diagram for explaining an imaging method for an identity verification document in the identity verification system according to the second embodiment. [Figure 13] FIG. 12 is a diagram for explaining monochrome stage detection performed by a monochrome stage detection unit of the identity verification system according to the second embodiment. [Figure 14] FIG. 15 is a flowchart showing a procedure of detecting a monochrome stage by monochrome stage detection performed by the monochrome stage detection unit of the identity verification system according to the second embodiment. [Figure 15] FIG. 18 is a diagram showing a configuration example of an identity verification document thickness determination unit in the identity verification system according to the third embodiment. [Figure 16] FIG. 21 is a diagram for explaining an imaging method for an identity verification document in the identity verification system according to the third embodiment. [Figure 17] FIG. 24 is a diagram for explaining stage edge detection performed by a stage edge detection unit of the identity verification system according to the third embodiment. [Figure 18] FIG. 27 is a flowchart showing a procedure of detecting a stage edge performed by the stage edge detection unit of the identity verification system according to the third embodiment. [Figure 19] FIG. 30 is a diagram showing a configuration example of an identity verification document thickness determination unit in the identity verification system according to the fourth embodiment. [Figure 20] FIG. 33 is a diagram for explaining shadow detection performed by a shadow detection unit of the identity verification system according to the fourth embodiment. [Figure 21] FIG. 36 is a flowchart showing a procedure of shadow detection performed by the shadow detection unit of the identity verification system according to the fourth embodiment. [Figure 22] FIG. 39 is a diagram showing a configuration example of an identity verification document thickness determination unit in the identity verification system according to the fifth embodiment. [Figure 23] FIG. 42 is a diagram for explaining an imaging method for an identity verification document in the identity verification system according to the fifth embodiment. [Figure 24] FIG. 12 is a diagram for explaining monochroid table detection by the monochroid table detection unit of the personal identification system according to the fifth embodiment. [Figure 25] FIG. 13 is a flowchart showing the procedure of upper monochrome platen detection by the upper monochrome platen detection unit of the personal identification system according to the fifth embodiment. MODE FOR CARRYING OUT THE INVENTION

[0009] Hereinafter, embodiments will be described with reference to the drawings.

[0010] (First Embodiment) First, the first embodiment will be described.

[0011] FIG. 1 is a diagram showing a configuration example of the personal identification system 1 according to the first embodiment.

[0012] The personal identification system 1 is a system that captures an image of an identification document such as a driver's license or a My Number Card. The personal identification system 1 can be implemented by, for example, an electronic device such as a smartphone or a tablet PC.

[0013] The personal identification system 1 includes a camera 11 and a personal identification document thickness determination unit (determination system) 12.

[0014] The camera 11 is a device that uses an image sensor such as a CMOS image sensor to form light from a subject incident on a lens into a digital image. The personal identification system 1 captures an image of the personal identification document with the camera 11. The image obtained by capturing the personal identification document is input to the personal identification document thickness determination unit 12.

[0015] The identity verification document thickness determination unit 12 is a module that determines whether the thickness of the identity verification document in the input image is appropriate or not. The identity verification document thickness determination unit 12 is implemented by executing a program on a processor that describes the procedure for determining whether the thickness of the identity verification document in the image is appropriate or not. The identity verification document thickness determination unit 12 may also be implemented as hardware, such as an electrical circuit.

[0016] Figure 2 shows an example of the configuration of the identity verification document thickness determination unit 12.

[0017] The identity verification document thickness determination unit 12 includes an edge detection unit 21, a hand detection unit 22, a thickness determination unit 23, and a character recognition unit 24. In Figure 2, the hatching of the hand detection unit 22 indicates that the hand detection unit 22 is a module (image state determination means) that determines whether the image state of the identity verification document is suitable for thickness determination, which is unique to the identity verification document thickness determination unit 12 in the identity verification system 1 of the first embodiment.

[0018] Here, as a comparative example, we will explain an example where the image quality of the identification document is not suitable for determining its thickness, referring to Figure 3.

[0019] Figure 3 shows an example of an image taken from a table 200 at an angle of approximately 45 degrees, of a (falsified) identity verification document 100, which is approximately 0.2 mm thick and is a modified health insurance card that can be used as an identity verification document, created using a printer or the like. Hereafter, regardless of whether it is genuine or not, the identity verification document that is captured by camera 11 will be referred to as identity verification document 100.

[0020] In the example shown in Figure 3, a shadow 150 is cast near the bottom edge of the identification document 100 on the table 200 where the document 100 is placed. In this case, the thickness should ideally be calculated as the length between the bottom edge of the front and bottom edge of the back of the identification document 100. However, there was a possibility that the thickness would be calculated by adding the width of the shadow 150 to this length. In other words, there was a possibility that the thickness of the identification document 100 in the image would be incorrectly judged as correct.

[0021] To prevent such misjudgments and improve the accuracy of thickness determination, it is preferable to prompt the user to re-image the identity verification document 100 if the document is being imaged in a state such as that shown in Figure 3, for example, as an error.

[0022] Based on this comparative example, we will now describe each part of the identity verification document thickness determination unit 12, including the hand detection unit 22.

[0023] In the first embodiment of the identity verification system 1, it is assumed that the imaging conditions for the identity verification document 100 stipulate that the identity verification document 100 should be held in the hand and imaged so that a linear shadow does not form near the bottom edge of the identity verification document 100, or more specifically, so that the area near the bottom edge of the identity verification document 100 appears to be floating in the air. Under this premise, the identity verification document thickness determination unit 12 detects a hand from the area below the bottom edge of the identity verification document 100 in the image of the imaged identity verification document 100. If a hand cannot be detected, the identity verification document thickness determination unit 12 fails the thickness determination before starting the thickness determination. Only if a hand is detected, the identity verification document thickness determination unit 12 determines the thickness from the bottom edge of the identity verification document 100. This prevents the identity verification system 1 of the first embodiment from making a false determination that an identity verification document 100 that has been altered with thin paper has the correct thickness.

[0024] Figure 4 is a diagram illustrating the method for capturing images of identity verification documents 100 in the identity verification system 1 of the first embodiment.

[0025] In the first embodiment of the identity verification system 1, as shown in Figure 4(A), the identity verification document 100 is placed in the palm of the hand and photographed from an angle of approximately 45 degrees. At this time, the identity verification system 1 may display a trapezoidal guide frame a1 for aligning the position of the identity verification document 100, superimposed on the subject image displayed in real time on, for example, the touchscreen panel of a smartphone or tablet PC, as shown in Figure 4(B), in order to facilitate the imaging of the identity verification document 100 in the state shown in Figure 4(A). Note that although an example of imaging the front surface of the identity verification document 100 is shown here, the same procedure applies when imaging the back surface of the identity verification document 100.

[0026] Figure 5 is a diagram illustrating edge detection by the edge detection unit 21 from the image captured as described above.

[0027] There are several known edge detection methods, such as the Canny edge detection method, and the edge detection unit 21 uses these methods to detect one edge each from the top, left, and right sides of the identity verification document 100, and two edges from the bottom of the identity verification document 100 (the bottom edge of the front and the bottom edge of the back), as shown by the symbol b1 in Figure 5. If these five edges cannot be detected at this stage, or if the detected edge segments are far from the guide frame shown in Figure 4(B), the identity verification system 1 outputs a message such as "All four sides are unclear" to prompt the user to retake the photograph.

[0028] Figure 6 is a diagram illustrating the detection of a hand by the hand detection unit 22.

[0029] In conventional systems, after edge detection, the process proceeds to thickness determination. However, in the first embodiment of the identity verification system 1, a distinctive process is performed to detect a hand before proceeding to thickness determination.

[0030] The hand detection unit 22 first detects the skin-colored area from the region below the lower edge of the identification document 100 detected by the edge detection unit 21. There are various methods for detecting the skin-colored area, but a simple method such as selecting a color in the range of (0,30,60) to (20,150,255) in the HSV color space is also acceptable. Then, as shown in Figure 6, the hand detection unit 22 can select the hand area c1. At this point, since the hand area (skin-colored area) c1 is selected using a simple function, there is a possibility that areas such as a wood-grain table may also be selected due to misjudgment.

[0031] Then, the hand detection unit 22 checks whether the area detected as skin is sufficiently wide. Specifically, for example, it calculates the range of the rectangle labeled c2 shown in Figure 7 based on its relative position from the bottom of the identification document 100, and checks whether the hand area (skin color area) c1 is larger than the range of this rectangle c2. If the rectangle c2 extends beyond the hand area (skin color area) c1, the hand detection unit 22 terminates the hand detection abnormality.

[0032] If the hand area (skin color area) c1 is larger, the hand detection unit 22 extracts the area of ​​rectangle c2 and determines whether the area of ​​rectangle c2 is a hand, for example, by using a neural network. This determination uses a pre-trained dictionary using images of hands and wood grain. Simple methods such as selecting a color in the range of (0,30,60) to (20,150,255) in the HSV color space may incorrectly detect wood grain tables as hands (skin). Therefore, a neural network is used to prevent incorrect detection of wood grain tables as hands and to detect only hands. Only if hand detection is successful using this method does the identity verification system 1 proceed to the next process. If hand detection fails, the identity verification system 1 declares the thickness determination a failure. If hand detection fails, the identity verification system 1 prompts the user to take a picture again by outputting a message such as "Please place it in your palm and take a picture."

[0033] In the explanation above, a neural network is used to distinguish between a hand and wood grain. However, other methods can be used to distinguish between a hand and wood grain, such as checking whether wrinkles intersect, or using OpenCV's cascade classifier.

[0034] Figure 8 is a flowchart showing the procedure for detecting a hand by the hand detection unit 22.

[0035] The hand detection unit 22 extracts the skin color area from the bottom edge of the lower edge of the identity verification document 100 (S101). The hand detection unit 22 determines whether the area of ​​the extracted skin color area is sufficient (S102). If it is sufficient (S102: YES), the hand detection unit 22 determines whether the bottom edge is a hand, for example, by using a learned dictionary of hands (S103). If it is a hand (S104: YES), the hand detection unit 22 completes the hand detection successfully.

[0036] On the other hand, if the area of ​​the skin color range is insufficient (S102:NO) or if the lower part of the edge is not a hand (S104:NO), the hand detection unit 22 terminates the hand detection abnormally. Note that "the area of ​​the skin color range is insufficient" includes "the skin color range is not extracted".

[0037] If the hand detection unit 22 successfully completes hand detection, the identity verification document thickness determination unit 12 performs thickness determination of the identity verification document 100 using the thickness determination unit 23. On the other hand, if the hand detection unit 22 abnormally completes hand detection, the identity verification document thickness determination unit 12 does not perform thickness determination of the identity verification document 100 using the thickness determination unit 23, but instead notifies the user of an error and prompts them to re-image the identity verification document 100.

[0038] Next, with reference to Figure 9, the thickness determination by the thickness determination unit 23 will be explained.

[0039] The thickness determination unit 23 calculates the thickness using the following formula (1) based on the length of A (the length of the bottom edge of the front or back of the identity verification document 100) and the length of B (the length between the bottom edge of the front and the bottom edge of the back of the identity verification document 100) shown in Figure 9. Here, length refers to the number of pixels.

[0040] Thickness D = B × 85.6 × 1.41 / A …(1) Since the width of a driver's license or My Number card is fixed at 85.6 mm, the vertical width of the image can be determined by multiplying the relative ratio of the number of pixels in A to the number of pixels in B by 85.6 mm. As the identity verification document 100 is photographed from an angle of approximately 45 degrees, the actual thickness can be calculated by multiplying this by 1.41. Depending on the image resolution, the width of B is only about 5 pixels, so even a 1-pixel shift in the line detection position can cause the calculated thickness to change by as much as 20%. Therefore, the accuracy of the thickness calculation is not high, so the thickness determination unit 23 determines that the thickness is acceptable if it is, for example, 0.4 mm or more, and unacceptable if it is less than that.

[0041] Next, with reference to Figure 10, we will explain the character recognition performed by the character recognition unit 24.

[0042] For example, if the identity verification document 100 is a driver's license, the character recognition unit 24, as shown in Figure 10(A), calculates the position d1 of the driver's license number relative to the position of the edge detected by the edge detection unit 21, extracts the image, and performs character recognition to recognize the driver's license number. This process is not necessarily required if the sole purpose is thickness determination, but performing this process has advantages such as (a) allowing the driver's license number used for thickness determination to be recorded as evidence, and (b) providing a simple deterrent against using cards other than driver's licenses for thickness determination. In the case of a My Number Card, as shown in Figure 10(B), the character recognition unit 24 may recognize the four-digit number at position d2 in the lower left corner, or the alphanumeric characters to its left.

[0043] Of course, this process can recognize items such as name and address, but because the driver's license is photographed in a trapezoidal shape and the focus is on the bottom edge, the text in fields such as the name, which are far away, may appear small, potentially reducing recognition accuracy. Therefore, items that are relatively close to the bottom edge can be recognized with higher accuracy.

[0044] As explained in the example, in the general thickness determination process of conventional systems, if the identity verification document 100 is placed on a table 200 or the like and photographed, a shadow may be cast, potentially leading to a false determination that a thin identity verification document 100 passes the thickness determination. In contrast, in the identity verification system 1 of the first embodiment, the identity verification document thickness determination unit 12 is equipped with a hand detection unit 22 that detects a hand from the bottom of the identity verification document 100. Therefore, if the identity verification document 100 is placed on a table 200 or the like and photographed, the process terminates due to a hand detection error, thereby reducing false determinations and enabling highly accurate thickness determination.

[0045] Of course, this method cannot detect forged identification documents 100 created by professional forgery rings, but it can significantly detect forged identification documents 100 created on a whim by ordinary people using a printing device. Generally, impulsive forgery is smaller in scale and more dispersed than professional forgery, making it difficult to track through investigations. By having this method in place, there is an effect of deterring impulsive forgery, which is difficult to track through investigations.

[0046] Although this example describes how to determine the thickness of identification document 100 in an image, this method can be applied to determining the thickness of various card-shaped objects in an image, not just identification document 100.

[0047] (Second Embodiment) Next, a second embodiment will be described.

[0048] Figure 11 shows an example configuration of the identity verification document thickness determination unit 12-2 in the identity verification system 1 of the second embodiment.

[0049] The first embodiment of the identity verification system 1 primarily assumes a case where the user holds the identity verification document 100 in their left hand and takes a picture with the camera in their right hand. However, if the camera held in the right hand is a heavy device such as a tablet PC, it becomes difficult to hold with one hand, resulting in a problem in the first embodiment where taking a picture is difficult. In addition, in the first embodiment, there was also the problem that images were often taken in which part of the document was obscured by the user's fingers.

[0050] In the second embodiment of the identity verification system 1, in order to solve these problems, a function to detect a single-color table (single-color table detection unit 25) is provided instead of a hand detection function (hand detection unit 22).

[0051] Referring to Figure 12, the method for imaging the identity verification document 100, which is a prerequisite for the identity verification system 1 of the second embodiment, will be described.

[0052] In the second embodiment of the identity verification system 1, as shown in Figure 12(A), a stand 250 is prepared, which is a single color such as black, with a height of 5mm to 15mm, a width of 150mm, and a depth of approximately 70mm. The identity verification document 100 is placed on the stand 250, and an image of the identity verification document 100 is captured from an angle of approximately 45 degrees. Figure 12(B) is an example of an image captured of the identity verification document 100 in the state shown in Figure 12(A).

[0053] In the identity verification system 1 of the second embodiment, the monochrome base detection unit 25 of the identity verification document thickness determination unit 12-2 detects a monochrome base 250, such as black, from an image of the identity verification document 100, as shown in Figure 12(B). For example, if the base 250 is black, the detection of the black base 250 can be achieved by an extremely simple method. For example, by selecting a color in the range of black and HSV V <= 60 from near the lower edge of the identity verification document 100, the monochrome base detection unit 25 can extract the range e1 of the base 250, as shown in Figure 13. The monochrome base detection unit 25 only needs to confirm that this range e1 of the base 250 is occupied by black except for the range of the identity verification document 100, that the upper boundary of the range e1 of the base 250 is 60-80% of the Y coordinate (vertical) range of the identity verification document 100, and that the lower boundary reaches the bottom edge of the image. If the upper boundary is between 60-80% of the Y-coordinate range of identity verification document 100, then identity verification document 100 will tilt backward due to its own weight, and there should always be a gap between the bottom edge of identity verification document 100 and the base. In other words, the bottom edge of identity verification document 100 should be suspended in mid-air.

[0054] Figure 14 is a flowchart showing the procedure for detecting a single-color stage by the single-color stage detection unit 25.

[0055] The monochrome background detection unit 25 extracts a monochrome area from the bottom edge of the lower edge of the identity verification document 100 (S201). The monochrome background detection unit 25 determines whether the lower half of the image (the entire area below the boundary line) excluding the area of ​​the identity verification document 100 is monochrome or not (S202). If it is monochrome (S203: YES), the monochrome background detection unit 25 determines whether the boundary line above the monochrome area is 60-80% of the Y coordinate range of the identity verification document 100 (S203). If it is 60-80% (S203: YES), the monochrome background detection unit 25 completes the monochrome background detection successfully.

[0056] On the other hand, if the lower half is not a single color (S202:NO) or if the upper boundary is not within the range of 60-80% (S203:NO), the single-color base detection unit 25 terminates the single-color base detection abnormally.

[0057] If the monochrome background detection unit 25 successfully completes the monochrome background detection, the identity verification document thickness determination unit 12-2 performs a thickness determination of the identity verification document 100 using the thickness determination unit 23. On the other hand, if the monochrome background detection unit 25 abnormally completes the monochrome background detection, the identity verification document thickness determination unit 12-2 does not perform a thickness determination of the identity verification document 100 using the thickness determination unit 23, but instead notifies the user of an error and prompts them to re-image the identity verification document 100.

[0058] Note that the "Date 250" can be a smartphone or similar device. This method has the advantage that when contracting for a mobile phone at a mobile phone carrier's counter, a special "Date 250" does not need to be procured; a smartphone or dummy available at the store can be used as the "Date 250." In addition, because the lower half is a single color, the lower edge of the identification document 100 can be clearly detected, resulting in high detection accuracy. Furthermore, holding the camera with both hands helps to reduce camera shake. The color of the "Date 250" does not have to be black; any single color such as blue or red is acceptable.

[0059] As described above, in the identity verification system 1 of the second embodiment, the identity verification document thickness determination unit 12-2 is equipped with a monochrome base detection unit 25 that detects a monochrome base 250 from the bottom of the image. This allows for stable edge detection without being affected by shadows even when the identity verification document 100 is placed on the base 250. Furthermore, if the image is taken without the base 250, a monochrome base detection error occurs, thus reducing misjudgments and enabling highly accurate thickness determination. Additionally, the camera can be held with both hands for stable shooting. Moreover, by shooting with the base 250, it is not necessary to hold the identity verification document 100 in one's hands, thus reducing the number of cases in which images with the document surface obscured by the hand are taken.

[0060] (Third embodiment) Next, a third embodiment will be described.

[0061] Figure 15 shows an example configuration of the identity verification document thickness determination unit 12-3 in the identity verification system 1 of the third embodiment.

[0062] As shown in Figure 15, the identity verification system 1 of the third embodiment is equipped with a table edge detection unit 26 instead of the monochrome table detection unit 25 of the second embodiment.

[0063] In the second embodiment of the identity verification system 1, when the identity verification document 100 was imaged, the identity verification document 100 was propped up on a monochrome stand 250. However, in the third embodiment of the identity verification system 1, when the identity verification document 100 was imaged, the identity verification document 100 was placed on the edge of the stand 250, as shown in Figure 16(A). In the third embodiment of the identity verification system 1, the color of the stand 250 is arbitrary. Figure 16(B) is an example image obtained by imaging the identity verification document 100 in the state shown in Figure 16(A).

[0064] In the third embodiment of the identity verification system 1, the base edge detection unit 26 of the identity verification document thickness determination unit 12-3 confirms, as shown in Figure 17, that there are clear horizontal lines on both the left and right sides of the identity verification document 100, and that the colors above and below these lines are clearly different. For example, in Figure 17, the base edge detection unit 26 confirms that the colors of areas A and B, and areas C and D are clearly different. As a result of the confirmation, the base edge detection unit 26 determines that base edge detection is successful only if there are clear horizontal lines on both the left and right sides, and the colors above and below both sides are clearly different; otherwise, it determines that base edge detection is unsuccessful.

[0065] Figure 18 is a flowchart showing the procedure for detecting the table edge by the table edge detection unit 26.

[0066] The table edge detection unit 26 determines whether there are clear horizontal lines on both the left and right sides of the identity verification document 100 (S301). If clear horizontal lines exist (S301: YES), the table edge detection unit 26 then determines whether there are clearly different colors above and below the horizontal lines on both the left and right sides (S302). If there are clearly different colors (S302: YES), the table edge detection unit 26 completes the table edge detection successfully.

[0067] On the other hand, if there is no clear horizontal line (S301:NO) or if the colors are not clearly different (S302:NO), the base edge detection unit 26 terminates the base edge detection abnormally.

[0068] If the table edge detection unit 26 successfully completes table edge detection, the identity verification document thickness determination unit 12-3 performs thickness determination of the identity verification document 100 using the thickness determination unit 23. On the other hand, if the table edge detection unit 26 abnormally completes table edge detection, the identity verification document thickness determination unit 12-3 does not perform thickness determination of the identity verification document 100 using the thickness determination unit 23, but instead notifies the user of an error and prompts them to re-image the identity verification document 100.

[0069] As described above, in the third embodiment of the identity verification system 1, the identity verification document thickness determination unit 12-3 is equipped with a table edge detection unit 26 that detects the edges of the table 250 from the left and right sides of the identity verification document 100, thereby enabling the user to photograph the identity verification document 100 using a table 250 that is relatively easy to prepare.

[0070] (Fourth embodiment) Next, a fourth embodiment will be described.

[0071] Figure 19 shows an example configuration of the identity verification document thickness determination unit 12-4 in the identity verification system 1 of the fourth embodiment.

[0072] As shown in Figure 19, the fourth embodiment of the identity verification system 1 is equipped with a shadow detection unit 27 instead of the monochrome base detection unit 25 of the second embodiment.

[0073] The method for capturing images of the identity verification document 100 in the identity verification system 1 of the fourth embodiment is the same as the method for capturing images of the identity verification document 100 in the identity verification system 1 of the second embodiment, as shown in Figure 12.

[0074] In the fourth embodiment of the identity verification system 1, the shadow detection unit 27 of the identity verification document thickness determination unit 12-4 detects, as shown in Figure 20, that in addition to two horizontal lines f1 and f2 near the lower edge of the identity verification document 100, there is also another horizontal line f3 indicating a shadow at a distance, for example, about five times the distance between the two lines f1 and f2. The shadow detection unit 27 considers shadow detection successful only if there are three lines f1 to f3, and fails to detect shadows otherwise.

[0075] Figure 21 is a flowchart showing the procedure for shadow detection by the shadow detection unit 27.

[0076] The shadow detection unit 27 determines whether or not there are two horizontal lines near the bottom edge of the identity verification document 100 (S401). If two horizontal lines are present (S401: YES), the shadow detection unit 27 then determines whether or not there is a third horizontal line at a position N times or more away from the two horizontal lines (S402). If a third horizontal line is present (S402: YES), the shadow detection unit 27 completes the shadow detection successfully.

[0077] On the other hand, if two horizontal lines are not present (S401:NO) or if a third horizontal line is not present (S402:NO), the shadow detection unit 27 terminates the shadow detection abnormally.

[0078] If the shadow detection unit 27 completes shadow detection successfully, the identity verification document thickness determination unit 12-4 performs thickness determination of the identity verification document 100 using the thickness determination unit 23. On the other hand, if the shadow detection unit 27 completes shadow detection abnormally, the identity verification document thickness determination unit 12-4 does not perform thickness determination of the identity verification document 100 using the thickness determination unit 23, but instead notifies the user of an error and prompts them to re-image the identity verification document 100.

[0079] As described above, in the fourth embodiment of the identity verification system 1, the identity verification document thickness determination unit 12-4 is equipped with a shadow detection unit 27 that detects the shadow of the identity verification document 100, which allows for confirmation that there is a gap between the identity verification document 100 and the base 250, thereby increasing the reliability that the distance between the upper two lines (near the bottom edge of the identity verification document 100) indicates the thickness.

[0080] (Fifth embodiment) Next, a fifth embodiment will be described.

[0081] Figure 22 shows an example configuration of the identity verification document thickness determination unit 12-5 in the identity verification system 1 of the fifth embodiment.

[0082] In the second embodiment of the identity verification system 1, when the identity verification document 100 was imaged, the identity verification document 100 was propped up against the back side of the stand 250. In contrast, in the fifth embodiment of the identity verification system 1, when the identity verification document 100 was imaged, as shown in Figure 23(A), the identity verification document 100 was propped up against the front side of the stand 250. In the fifth embodiment of the identity verification system 1, the thickness of the top edge of the identity verification document 100 is measured. Figure 23(B) is an example image of the identity verification document 100 imaged in the state shown in Figure 23(A).

[0083] Therefore, unlike the monochrome base detection unit 25 (see Figure 11) of the identity verification document thickness determination unit 12-2 in the second embodiment, the identity verification document thickness determination unit 12-5 includes an upper monochrome base detection unit 28 that detects the monochrome base 250 from the top of the image.

[0084] As shown in Figure 24, the upper monochrome background detection unit 28 detects a total of three edges: the two top edges of the identity verification document 100 and the horizontal lines appearing on the left and right sides of the identity verification document 100. Then, if the upper monochrome background detection unit 28 can detect a black background 250 at the top (g1) of the image using the method described in the second embodiment, it determines that the black background detection was successful; otherwise, it determines that the black background detection was unsuccessful.

[0085] Figure 25 is a flowchart showing the procedure for detecting the upper monocolor stage by the upper monocolor stage detection unit 28.

[0086] The upper monochrome background detection unit 28 first detects two edges near the top edge of the identity verification document 100 (S501). If two edges are detected (S501: YES), the upper monochrome background detection unit 28 then detects the edges of horizontal lines appearing on the left and right sides of the identity verification document 100 (S502). If edges are detected (S502: YES), the upper monochrome background detection unit 28 detects a monochrome background at the top of the image (S503). If a monochrome background is detected (S503: YES), the upper monochrome background detection unit 28 completes the upper monochrome background detection successfully.

[0087] On the other hand, if two edges cannot be detected near the top edge of the identity verification document 100 (S501: NO), if the edges of the horizontal lines appearing on the left and right sides of the identity verification document 100 cannot be detected (S502: NO), or if a single-color background cannot be detected (S503: YES), the upper single-color background detection unit 28 terminates the upper single-color background detection abnormally.

[0088] If the upper monochrome stage detection unit 28 successfully completes the upper monochrome stage detection, the identity verification document thickness determination unit 12-5 performs a thickness determination of the identity verification document 100 using the thickness determination unit 23. On the other hand, if the upper monochrome stage detection unit 28 abnormally completes the upper monochrome stage detection, the identity verification document thickness determination unit 12-5 does not perform a thickness determination of the identity verification document 100 using the thickness determination unit 23, but instead notifies the user of an error and prompts them to re-image the identity verification document 100.

[0089] As described above, in the identity verification system 1 of the fifth embodiment, the identity verification document thickness determination unit 12-5 is equipped with an upper monochrome base detection unit 28 that detects the upper monochrome base of the identity verification document 100. This allows for confirmation that there is a gap between the identity verification document 100 and the base 250, thereby increasing the reliability of the distance between the two lines near the top edge of the identity verification document 100 indicating its thickness. Furthermore, because the document is placed on the base 250, the user can take an image of the identity verification document 100 with less camera shake, and since the user does not need to hold the identity verification document 100 in their hand, the number of cases in which images with the document surface obscured by the hand are reduced.

[0090] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. [Explanation of Symbols]

[0091] 1...Identity verification system, 11...Camera, 12...Identity verification document thickness determination unit, 21...Edge detection unit, 22...Hand detection unit, 23...Thickness determination unit, 24...Character recognition unit, 25...Monochrome table detection unit, 26...Table edge detection unit, 27...Shadow detection unit, 28...Upper monochrome table detection unit, 100...Identity verification document, 150...Shadow, 200...Table, 250...Table.

Claims

1. An edge detection means for detecting edges relating to a card-shaped object, including two edges at the lower or upper part of the object, from an image of the card-shaped object; A thickness determination means for determining whether the thickness of the object is appropriate based on the distance between the two edges, An imaging state determination means that determines whether the object is being imaged in a state suitable for determining the thickness of the object, using the edge information detected by the edge detection means, A judgment system equipped with the following features.

2. The imaging state determination means is A monochrome area of ​​a predetermined size or larger is detected in the region below the two edges at the bottom of the object in the aforementioned image. If, excluding the area in which the object is imaged, the lower region of the boundary dividing the image vertically is occupied by the single color, and the boundary falls within a predetermined range of the vertical coordinate axes for determining the coordinates on the object, then it is determined that the object is imaged in a state suitable for determining the thickness of the object. The determination system according to claim 1.

3. The imaging state determination means is From both the region to the left of the edge on the left side of the object and the region to the right of the edge on the right side of the object detected by the edge detection means in the image, a horizontal line segment tangent to the edge on the left side of the object or the edge on the right side of the object is detected. If horizontal line segments are detected from both of the aforementioned regions, and the color difference between the upper and lower parts of both line segments is greater than or equal to a threshold, it is determined that the object has been imaged in a state suitable for determining the thickness of the object. The determination system according to claim 1.

4. The imaging state determination means is From the region below the two edges at the bottom of the object in the aforementioned image, a horizontal line segment is detected in which the distance between it and the lower edge of the two edges is at least a predetermined multiple of the distance between the two edges. If a horizontal line segment is detected whose distance from the lower edge among the two edges is at least a predetermined multiple of the distance between the two edges, it is determined that the object is being imaged in a state suitable for determining the thickness of the object. The determination system according to claim 1.

5. The imaging state determination means is A monochrome area of ​​a predetermined size or larger is detected from the region above the two edges at the top of the object in the aforementioned image. If, excluding the area in which the object is imaged, the upper region of the boundary dividing the image vertically is occupied by the single color, and the boundary falls within a predetermined range of the vertical coordinate axes for determining the coordinates on the object, then it is determined that the object is imaged in a state suitable for determining the thickness of the object. The determination system according to claim 1.

6. A determination method performed by an electronic device capable of capturing or inputting an image, From an image of a card-shaped object, detect edges relating to the object, including two edges at the lower or upper part of the object. The distance between the two edges is used to determine whether the thickness of the object is appropriate, Using the detected edge information, it is determined whether the object is being imaged in a state suitable for determining the thickness of the object. A determination method comprising the following:

7. Computers, An edge detection means for detecting edges relating to a card-shaped object, including two edges at the lower or upper part of the object, from an image of the card-shaped object. A thickness determination means for determining whether the thickness of the object is appropriate based on the distance between the two edges. An imaging state determination means determines whether the object is being imaged in a state suitable for determining the thickness of the object, using the edge information detected by the edge detection means. A program designed to function as such.

Images