JP7838433B2 - Authenticity determination device and program - Google Patents

Description

This invention relates to an authenticity determination device and its program, etc.

When opening an account, it is common practice for users to visit a financial institution, submit their identification (ID card), and for the financial institution to verify their identity in person using the submitted ID. Furthermore, Patent Document 1 describes the use of identification for identity verification when issuing various media in a photo booth. Thus, identity verification to prevent impersonation is carried out in a variety of situations.

In this context, it is crucial to determine the authenticity of the identification document and whether it has been forged or altered. Patent documents 1 and 2 describe a technology for determining the authenticity of an identification document from its image. This technology allows for automatic determination of authenticity based on the characteristics and textual information of the image.

Japanese Patent Publication No. 2020-38442 Japanese Patent Publication No. 2011-34535

The above is an example of using the image features and textual information of an identification document for verification. However, a characteristic of a genuine identification document (ID card) is that each area of the card is printed using a specific printing method, and the ink and printing method used differ for each area. Therefore, it is possible to improve the accuracy of authenticity verification by utilizing the differences in ink used for each area.

In recent years, identity verification has become increasingly stringent, and there is a growing demand for technologies that can reliably determine the authenticity of identification documents. Technologies that utilize the differences mentioned above are expected to contribute to improving the accuracy of such authenticity determination.

This invention has been made in view of the above-mentioned problems, and aims to provide a genuineness determination device, etc., that can suitably determine the authenticity of identity verification documents.

The first invention for solving the aforementioned problems is a genuineness determination device for determining the authenticity of an identification document, comprising: acquisition means for acquiring a visible light image and an infrared image of the identification document; and genuineness determination means for determining, for each of a plurality of inspection areas of the identification document, the number of pixels within that inspection area whose density in the visible light image and the infrared image satisfies a predetermined standard as a discrimination value; determining whether the inspection area is genuine or not based on a comparison result between the discrimination value and a threshold; and determining the authenticity of the identification document based on the discrimination result, wherein at least one of the predetermined standard and the relationship between the comparison result and the authenticity of the inspection area differs among the plurality of inspection areas.

This invention utilizes the fact that genuine identification documents are manufactured by printing each area of the document surface using a predetermined printing method, and that the appearance of shading in these areas differs between visible light and infrared images depending on the ink used, to determine authenticity. For example, areas printed with ink containing infrared-absorbing components such as carbon appear as dark areas (high-density areas) in both visible light and infrared images, while areas printed with ink that does not absorb infrared light appear as dark areas in visible light images and as light areas (low-density areas) in infrared images. Therefore, cases of forgery or alteration using ink with different infrared absorption characteristics than that used in genuine identification documents can be effectively detected. Furthermore, the accuracy of authenticity determination is improved by changing the criteria for each of the multiple inspection areas according to the ink used, etc., while determining authenticity (whether the inspection area is from a genuine identification document or not).

Here, the grayscale in the visible light image refers to the grayscale of a predetermined color component of the visible light image, or the grayscale indicating the degree of brightness of each pixel in the visible light image. The authenticity determination means may also create a spectral image in which an image showing the grayscale of each color component of the visible light image and an image showing the degree of brightness of each pixel in the visible light image are represented by grayscale, and may use the grayscale of the pixels in the spectral image for the determination. In this case, it is also desirable that the spectral image used for the determination differs among the multiple inspection areas.
In this way, by using spectral images corresponding to the color of the printed area to determine whether each inspection area is genuine or not, the accuracy of authenticity determination can be improved.

Preferably, the authenticity determination means creates a binary image by binarizing the grayscale of the spectral image and the infrared image, and determines whether the inspection area is authentic or not based on the pixel values in the binary image.
This reduces the processing burden when determining authenticity.

For example, the predetermined criteria may differ among multiple inspection areas, such that in one inspection area, the predetermined criterion is that the area appears as a dense area in both the spectral image and the infrared image, while in another inspection area, the predetermined criterion is that the area appears as a dense area in the spectral image and as a pale area in the infrared image.
This allows for accurate determination of the authenticity of an identification document by distinguishing between genuine and counterfeit areas printed with inks that have different infrared absorption properties.

For example, among multiple inspection areas, the predetermined criteria are the same, but the relationship between the comparison result and the correctness of the inspection area differs, and the predetermined criteria are that it appears as a dense area in the spectral image and as a pale area in the infrared image, and in one inspection area, the inspection area is considered correct when the discriminant value is less than or equal to the threshold, and in another inspection area, the inspection area is considered correct when the discriminant value is greater than or equal to the threshold.
This allows for the detection of forgery or alteration of genuine identification documents using inks with different infrared absorption properties (inks that do not absorb infrared rays) in areas printed with inks containing infrared-absorbing components, and also allows for the detection of cases where areas printed with non-infrared-absorbing ink on genuine identification documents are lost during the forgery or alteration process.

It is also desirable that the authenticity determination means perform a dilation process on the binary image obtained by binarizing the grayscale of the infrared image, and use the dilated binary image for the determination.
This makes it possible to suppress the decrease in judgment accuracy caused by the misalignment between visible light images and infrared images.

The second invention is a program for causing a computer to function as an authenticity determination device for determining the authenticity of an identity document, comprising: acquisition means for acquiring a visible light image and an infrared image of the identity document; and authenticity determination means for determining, for each of a plurality of inspection areas of the identity document, the number of pixels within the inspection area whose density in the visible light image and the infrared image satisfies a predetermined standard as a discrimination value, determining whether the inspection area is genuine or not based on a comparison result between the discrimination value and a threshold, and determining the authenticity of the identity document based on the discrimination result, wherein at least the predetermined standard differs among the plurality of inspection areas, or the relationship between the comparison result and the correctness of the inspection area differs.
The second invention is a program for the authenticity determination device of the first invention.

This invention makes it possible to provide a genuineness determination device, etc., that can suitably determine the authenticity of identification documents.

A diagram showing the authenticity determination system 1. A diagram showing the hardware configuration of the authenticity determination device 3. A diagram showing an example of ID card 10. A diagram illustrating the function of the authenticity determination device 3. A flowchart outlining the method for determining authenticity. A flowchart showing the procedure for determining authenticity. Figure 20 and 30 show binary images of the spectral and infrared images. An example of examination area 11. This diagram illustrates the method for determining whether the inspection area 11 is correct or incorrect (mode 0). A diagram illustrating the method (mode 1) for determining whether the inspection area 11 is correct or incorrect. This diagram illustrates the method for determining whether the inspection area 11 is correct or incorrect (mode 2). A diagram explaining setting information 40.

The following describes preferred embodiments of the present invention in detail with reference to the drawings.

(1. Authenticity Determination System 1)
Figure 1 shows a counterfeit detection system 1 having a counterfeit detection device 3 according to an embodiment of the present invention. As shown in Figure 1, the counterfeit detection system 1 is configured by connecting a reader 2 and a counterfeit detection device 3 to each other via wired or wireless communication. In the counterfeit detection system 1, an image of the ID card 10 is transmitted from the reader 2 to the counterfeit detection device 3, and the counterfeit detection device 3 performs a counterfeit detection of the ID card 10 from the image.

The reader 2 photographs the ID card 10 and acquires its image. In particular, in this embodiment, the reader 2 includes a visible light camera and visible light illumination, and an infrared camera and infrared illumination.

When using a visible light camera, the visible light emitted from the visible light illumination and reflected from the ID card 10 is received by the visible light camera. Similarly, when using an infrared camera, the infrared light emitted from the infrared illumination and reflected from the ID card 10 is received by the infrared camera. Hereinafter, the image of the ID card 10 captured by the visible light camera will be referred to as the visible light image, and the image of the ID card 10 captured by the infrared camera will be referred to as the infrared image.

Figure 2 shows the hardware configuration of the authenticity determination device 3. The authenticity determination device 3 can be implemented using a computer configured by connecting a control unit 31, a storage unit 32, a communication unit 33, a display unit 34, etc., via a bus or similar means. However, it is not limited to this configuration, and various configurations can be adopted as appropriate.

The control unit 31 consists of a CPU, ROM, RAM, etc. The CPU retrieves and executes programs related to the processing of the authenticity determination device 3, stored in storage media such as the memory unit 32 and ROM, into a work area on the RAM. ROM is non-volatile memory and permanently holds programs and data such as the boot program and BIOS. RAM is volatile memory and temporarily holds programs and data loaded from the memory unit 32 and ROM, and also includes a work area used by the control unit 31 for various processing tasks.

The memory unit 32 is a hard disk drive, solid-state drive, flash memory, etc., and stores the program executed by the authenticity determination device 3 during the processing described later, the data necessary for program execution, the OS, etc.

The communication unit 33 is a communication interface that mediates communication with the reader device 2. The display unit 34 is a liquid crystal display or the like, and displays various information related to the authenticity determination of the ID card 10.

(2. ID card 10)
In this embodiment, ID card 10 is an identification document that is subject to authenticity verification. An identification document is a medium used for administrative procedures and identity verification by various government agencies, financial institutions, mobile phone carriers, etc., and is, for example, a driver's license. However, ID card 10 is not limited to this and may also be a My Number card, residence card, special permanent resident certificate, etc.

Figure 3 is a schematic diagram of ID card 10 (driver's license), illustrating a schematic example of the card's surface. ID card 10 is printed on a roughly rectangular card base with the cardholder's name 101, date of birth 102, address 103, date of issuance 104, expiration date 105, issuance number 106, issuing authority 107, seal 108, and photograph 109 printed on it.

The name (101), date of birth (102), address (103), issuance date (104), expiration date (105), issuance number (106), and facial photograph (109) are all on-demand information that differs for each individual ID card (10). This on-demand information is formed by printing it at a fixed position on the card substrate using a thermal transfer printing method with a thermal head. Text other than the facial photograph (109) is printed using a fusion thermal transfer method, while the facial photograph (109) is printed using a sublimation thermal transfer method.

On the other hand, the text such as "Name," "Address," "Issued," and "Number" indicating the fields for name 101, address 103, issuance date 104, and issuance number 106, as well as the issuer 107, seal 108, the text "Driver's License" 110, and ruled lines, are used in common across all ID cards 10. These texts and ruled lines are pre-formed in fixed positions on the card substrate using printing methods such as offset printing or letterpress printing before the on-demand information is printed, and are referred to as pre-printed areas.

The ID card 10 is manufactured by printing the aforementioned on-demand information onto a card substrate that already has a pre-printed portion formed on it. The configuration of the on-demand information and the pre-printed portion varies depending on the type of ID card 10, such as a driver's license, My Number card, residence card, or special permanent resident certificate.

(3. Functions of the Authenticity Determination Device 3)
Figure 4 is a block diagram showing the functions of the authenticity determination device 3. The authenticity determination device 3 includes an acquisition means 301, an authenticity determination means 302, etc.

The acquisition means 301 involves the control unit 31 of the authenticity determination device 3 receiving and acquiring the visible light image and infrared image of the ID card 10 from the reader 2 via the communication unit 33.

The authenticity determination means 302 is performed by the control unit 31 of the authenticity determination device 3, which uses the visible light image and infrared image of the ID card 10 acquired from the reader 2 to determine the authenticity of the ID card 10.

(4. Outline of methods for determining authenticity)
Figure 5 is a flowchart illustrating the authenticity determination method performed by the authenticity determination system 1. In Figure 5, S1 to S2 are processes executed by the control unit (not shown) of the reading device 2, which controls each part of the reading device 2, and S3 to S5 are processes executed by the control unit 31 of the authenticity determination device 3, which controls each part of the authenticity determination device 3.

In this embodiment, the user first places the ID card 10 in a predetermined position on the reader 2, and the reader 2 photographs the ID card 10 using a visible light camera and an infrared camera, respectively, in response to the user's operation (S1).

The reader 2 transmits the visible light image and infrared image obtained by photographing the ID card 10 to the authenticity determination device 3 (S2). The authenticity determination device 3 receives the visible light image and infrared image of the ID card 10 (S3) and performs authenticity determination of the ID card 10 using the visible light image and infrared image (S4). Both the visible light image and the infrared image are assumed to be images in which the color of each pixel is represented by the gradation values of the R (red), G (green), and B (blue) components. In the infrared image, the R, G, and B gradation values of each pixel are approximately the same. However, in some cases, the infrared image obtained by photographing the ID card 10 with an infrared camera may be a grayscale image representing the degree of brightness of each pixel by gradation values; in this case, the grayscale conversion in S105 described later is unnecessary.

The authenticity determination device 3 displays the authenticity determination result of the ID card 10 on the display unit 34 (S5), and terminates the process. The person in charge who sees the display unit 34 can then take appropriate action. The authenticity determination device 3 may also display instructions corresponding to the authenticity determination result on the display unit 34, along with the authenticity determination result of the ID card 10, to prompt the person involved to take appropriate action.

In this embodiment, in S4, the genuine ID card 10 is manufactured by printing each area of the card surface using a predetermined printing method. Authenticity is determined by utilizing the fact that the appearance of shading in these areas differs between visible light images and infrared images depending on the ink used. The details of the authenticity determination method are described below.

(5. Authenticity verification of ID card 10)
Figure 6 is a flowchart showing the procedure for determining the authenticity of ID card 10 in S4.

In this embodiment, in S4, the visible light image and infrared image received in S3 are first resized to a predetermined size (S101). This speeds up subsequent processing. The above size is the same for both the visible light image and the infrared image; for example, the original visible light image and infrared image are reduced to about half their original size, but this is not limited to this.

Next, the authenticity determination device 3 masks the face photograph 109 in the resized visible light image and infrared image (S102). This process is performed to prevent the grayscale values of the pixels in the face photograph 109 from affecting the subsequent binarization process. The masking process is performed by replacing the grayscale values of the pixels in the face photograph 109 with specific mask values.

Next, the authenticity determination device 3 creates a spectral image from the visible light image after masking (S103). In this process, the spectral image is created using the R, G, and B component tonal values of each pixel in the visible light image as the respective pixel tonal values. Hereinafter, the image based on the B component tonal value will be called the B image, the image based on the G component tonal value will be called the G image, and the image based on the R component tonal value will be called the R image. The B image, G image, and R image are spectral images that show the intensity of the R, G, and B components, respectively.

In step S103, a spectral image is created, where the highest grayscale value among the R, G, and B component grayscale values of each pixel in the visible light image is used as the grayscale value for each pixel. This image shows the degree of brightness of each pixel in the visible light image as shades of gray, and is hereinafter referred to as the V (brightness) image. For example, if the grayscale values of the R, G, and B components of a certain pixel in the visible light image are (R, G, B) = (10, 20, 30), then the grayscale value of the corresponding pixel in the V image will be 30.

As a result, in subsequent processing, the grayscale of the visible light image will be represented by the grayscale of each color component of the visible light image, or by the grayscale indicating the brightness level of each pixel in the visible light image.

The authenticity determination device 3 binarizes the spectral images created in S103 (S104). While the binarization method is not particularly limited, in this embodiment, binarization is performed using the Otsu method. The Otsu method is a technique that automatically determines the threshold value for binarization from the grayscale value histogram of the entire image, and is well-known in the field of image processing. However, it is also possible to perform binarization using a predetermined threshold value.

Figure 7(a) shows an example of a binary image 20 of a spectral image (B image, G image, R image, V image). These binary images 20 represent high-density areas (dark areas) in the spectral image with black pixels and low-density areas (light areas) with white pixels. Printed areas are basically represented by black pixels (dark areas).

However, in the binary images 30 of the B, G, and R images, the parts printed in these colors (for example, the parts printed in red in the R image) do not appear as black pixels, but as white pixels (light areas). On the other hand, the parts printed in colors close to the complementary color of these colors (for example, the parts printed in red in the B image) appear as black pixels.

The authenticity determination device 3 performs grayscale conversion and then binarization processing on the infrared image after masking (S105). Grayscale conversion is performed because the R, G, and B component tonal values of pixels in the infrared image may differ from each other; however, as mentioned above, this step can be omitted if the infrared image is already a grayscale image. Grayscale conversion can be performed using known methods.

Unlike the previous method, the binarization of infrared images is performed using a predetermined threshold. It is desirable to set an appropriate threshold value depending on the type of ID card 10. However, binarization may also be performed using the Otsu method, as described above. Figure 7(b) shows an example of a binarized image 30 of an infrared image. In the infrared image, areas printed with ink containing infrared-absorbing components become high-density areas (dark areas), and in the binarized image 30, these areas appear as black pixels.

Next, the authenticity determination device 3 performs a dark area expansion process on the binary image 30 of the infrared image (S106). This expansion process is a known technique in the field of image processing. For example, if black pixels exist within a certain size range, all pixels within that range are converted to black pixels, thereby expanding the dark areas. This allows the process described in S108 to be performed without problems even if there is a slight positional shift between the spectral image and the infrared image. Figure 7(c) shows an example of the binary image 30 after the expansion process.

After preprocessing the spectral and infrared images to create binary images 20 and 30, the authenticity determination device 3 extracts the inspection area from the binary images 20 and 30 of the spectral and infrared images (S107). The inspection area is predetermined for each type of ID card 10, and Figure 8 shows a portion of it on the visible light image of the ID card 10. As shown in Figure 8, in this embodiment, multiple inspection areas 11 are defined on the surface of the ID card 10, and these inspection areas 11 are extracted from both the binary images 20 and 30 of the spectral and infrared images. Note that the inspection area 11 is mainly set in the text portion of the surface of the ID card 10.

The authenticity determination device 3 then determines whether the inspection area 11 is genuine or not (whether the inspection area 11 belongs to a genuine ID card 10) based on the pixel values of the binary images 20 and 30 (S108).

In S108, based on the pixel values of the binary images 20 and 30, the number of pixels within the inspection area 11 whose grayscale in the spectral image and infrared image meets a predetermined standard is determined as a discrimination value. The pass/fail status of the inspection area 11 is then determined from the comparison result between the discrimination value and the threshold. In this embodiment, the predetermined standard described above, and the relationship between the comparison result and the pass/fail status of the inspection area 11, differ for each inspection area 11. More specifically, the main difference lies in whether the inspection area 11 is printed with ink containing an infrared-absorbing component or with ink that does not. The differences in the methods for determining the pass/fail status of the inspection area 11 (referred to as "modes") are described below.

<Mode 0; when the inspection area 11 is printed with ink containing an infrared absorbing component>
If the inspection area 11 is the portion of the regular ID card 10 printed with ink containing an infrared-absorbing component, then, as shown in Figure 9(a), the printed portion will appear as a dark area in both the spectral image (V image in this example) and the infrared image, and will appear as black pixels in the binary images 20 and 30.

On the other hand, Figure 9(b) shows the ID card 10 with writing 12 done using infrared-infrared-inhibiting ink in the inspection area 11. In this case, the area with writing 12 appears as a dark area in the spectral image and a light area in the infrared image. It appears as a black pixel in the binary image 20 of the spectral image, but as a white pixel in the binary image 30 of the infrared image.

Figure 9(c) shows pixels that appear black in the spectral binary image 20 and white in the infrared binary image 30 as white pixels, and all other pixels as black pixels. The authenticity determination device 3, in order to detect alterations caused by writing 12 with inks having different infrared absorption characteristics, counts the number of pixels (white pixels in Figure 9(c)) that appear black in the spectral binary image 20 and white in the infrared binary image 30 (white pixels in Figure 9(c)) in the inspection area 11 extracted from the spectral binary image 20 and the infrared binary image 30, and uses this count as the discrimination value. Then, the discrimination value is compared with a threshold value. If the discrimination value is less than or equal to the threshold value, the inspection area 11 is determined to be genuine; if the discrimination value exceeds the threshold value, the inspection area 11 is determined to be invalid.

In this case, the predetermined criterion is "that the area appears dark in the spectral image and light in the infrared image (that the pixel appears black in the binary image 20 and white in the binary image 30)," and the relationship between the comparison result and the correctness of the inspection area 11 is "that the inspection area 11 is correct if the discrimination value is below the threshold."

<Mode 1: In the inspection area 11, there is a part printed with ink that does not absorb infrared light, and that part disappears due to forgery, alteration, etc.>
Some parts of ID card 10 are printed with ink that does not absorb infrared rays, and in counterfeit or altered ID cards 10, these parts are sometimes missing.

For example, Figure 10(a) shows the inspection area 11 containing the ID card 10's issuance number 106 for the spectral image (image B in this example) and the infrared image binary images 20 and 30, respectively. The inspection area 11 also contains a shaded area 13 printed with red ink that does not absorb infrared light, along with the issuance number 106. This shaded area 13 appears as a dark area in the spectral image and a light area in the infrared image. It appears as a black pixel in the spectral image binary image 20 and as a white pixel in the infrared image binary image 30.

On the other hand, Figure 10(b) shows the ID card 10 with the shaded area 13 removed from the inspection area 11. In this case, the area where the shaded area was removed becomes a pale area in the spectral image, and a white pixel in the binary image 20 of the spectral image.

Figure 10(c) shows pixels that appear black in the spectral binary image 20 and white in the infrared binary image 30 as white pixels, and all other pixels as black pixels. The authenticity determination device 3 detects cases where parts that should be printed with ink that does not absorb infrared light have disappeared during the process of forgery or alteration. To do this, it counts the number of pixels (white pixels in Figure 10(c)) that appear black in the spectral binary image 20 and white in the infrared binary image 30 of the inspection area 11 extracted from the spectral binary image 20 and the infrared binary image 30, and uses this count as the discrimination value. The discrimination value is then compared to a threshold value. If the discrimination value is greater than or equal to the threshold value, the area is determined to be genuine; if the discrimination value is less than the threshold value, the area is determined to be invalid.

In this case, the predetermined criterion is "that the area appears dark in the spectral image and light in the infrared image (that the pixel appears black in the binary image 20 and white in the binary image 30)," and the relationship between the comparison result and the correctness of the inspection area 11 is "that the inspection area 11 is correct if the discrimination value is above the threshold."

<Mode 2; When the inspection area 11 is printed with ink that does not absorb infrared light>
If the inspection area 11 is a portion of the regular ID card 10 that is printed with ink that does not absorb infrared light, then, as shown in Figure 11(a), the printed portion will appear as a dark area in the spectral image (V image in this example), but as a light area in the infrared image, appearing as a black pixel in the binary image 20 of the spectral image and as a white pixel in the binary image 30 of the infrared image.

On the other hand, Figure 11(b) shows that in the inspection area 11 of the ID card 10, the portion that is normally printed with ink that does not absorb infrared light is printed with ink containing an infrared-absorbing component. In this case, the printed portion 14 appears as a dark area in both the spectral image and the infrared image, and appears as a black pixel in both the binary image 20 of the spectral image and the binary image 30 of the infrared image.

Figure 11(c) shows pixels that appear black in both the spectral binary image 20 and the infrared binary image 30 as white pixels, and all other pixels as black pixels. The authenticity determination device 3, in order to detect forgery or alteration caused by printing with inks having different infrared absorption characteristics, counts the number of pixels that appear black in both binary images 20 and 30 (white pixels in Figure 11(c)) in the inspection area 11 extracted from the spectral binary image 20 and the infrared binary image 30, and uses this count as the discrimination value. Then, the discrimination value is compared with a threshold value. If the discrimination value is less than or equal to the threshold value, the inspection area 11 is determined to be genuine; if the discrimination value exceeds the threshold value, the inspection area 11 is determined to be invalid.

In this case, the predetermined criterion is "that the area appears as a dark area in the spectral image and infrared image (that the pixel appears as black in the binary images 20 and 30)," and the relationship between the comparison result and the correctness of the inspection area 11 is "that the inspection area 11 is correct if the discrimination value is below the threshold."

Thus, in S108, a predetermined discrimination method (modes 0 to 2 described above) is performed for each inspection area 11 according to the differences in the ink used, etc., and it is determined whether the inspection area 11 extracted in S107 is positive or negative.

The authenticity determination device 3 repeats the processes S107 to S108 until it has made the above determination for all inspection areas 11 (S109; No). After determining whether all inspection areas 11 are true or false (S109; Yes), it then makes an authenticity determination of the ID card 10 based on the determination results for each inspection area 11 (S110).

In this embodiment, the ID card 10 is considered true if all inspection areas 11 are determined to be positive, and false otherwise. However, this is not limited to this configuration. For example, the ID card 10 can be considered true if the number of inspection areas 11 determined to be positive is equal to or greater than a predetermined number.

Furthermore, the processing in S107-S108 can be performed by the authenticity determination device 3, which reads setting information from the storage unit 32 that defines the position and discrimination method of the inspection area 11 for each type of ID card 10, and performs the processing based on this setting information.

Figures 12(a) and 12(b) illustrate the setting information 40 for different types of ID cards 10. In this embodiment, the setting information 40 includes data for the inspection area range 41, the spectral image 42 used for discrimination, the mode 43, and the discrimination threshold 44, and is determined for each type of ID card 10. The authenticity determination device 3 can read the setting information 40 corresponding to the type of ID card 10 to be judged and execute the processes S107 to S108.

The inspection area range 41 is defined in S107 as the range from which the inspection area 11 is extracted from the binary images 20 and 30 of the spectral and infrared images. For example, the positions of the upper left and lower right corners of the rectangular inspection area 11 are specified on the orthogonal two axes.

The spectral image 42 used for discrimination is specified in the discrimination procedure described in modes 0 to 2 above, indicating which spectral image to use. The inspection area 11 is mainly set to the text portion, and the spectral image should basically be the V image. However, exceptionally, for areas printed in color, it is desirable to use a color close to the complementary color of the printed color in order to make that area appear as a dark area in the spectral image. For example, for inspection area 11 that includes areas printed in red, such as the hatched area 13 mentioned above, the B image should be selected. Thus, in this embodiment, the spectral image used for discrimination may differ between inspection areas 11.

Mode 43 determines which of the modes 0 to 2 described above will be implemented. In this embodiment, as described above, different modes are set for multiple inspection areas 11 within the ID card 10. The threshold value 44 is the threshold value used for comparison with the discrimination value described in modes 0 to 2, and this also differs depending on the inspection area 11.

Furthermore, in this embodiment, as shown in the dashed box in the setting information 40 of Figures 12(a) and (b), for the inspection area 11 where discrimination by mode 1 has been performed, discrimination by mode 0 or mode 2 is also performed simultaneously.

This is because, for example, in the Mode 1 example explained in Figure 10, the issuance number 106 itself, excluding the shaded area 13, is printed with ink containing infrared-absorbing components, making it useful to perform a separate discrimination using Mode 0. Furthermore, in the Mode 2 example explained in Figure 11, it is useful not only to determine whether the portion printed with infrared-absorbing ink is present in the inspection area 11, but also to perform discrimination using Mode 1 to detect whether the portion printed with non-infrared-absorbing ink has disappeared during the counterfeiting or alteration process.

As explained above, in this embodiment, a genuine ID card 10 is manufactured by printing each area of the card surface using a predetermined printing method. Authenticity is determined by utilizing the fact that the appearance of shading in these areas differs between visible light images and infrared images depending on the ink used. For example, areas printed with ink containing infrared-absorbing components such as carbon appear as dark areas (high-density areas) in both visible light and infrared images, while areas printed with ink that does not absorb infrared light appear as dark areas in visible light images and as light areas (low-density areas) in infrared images.

Therefore, in this embodiment, as shown in Figures 9 to 11, it is possible to suitably detect cases where counterfeiting or alteration has occurred using ink with different infrared absorption characteristics than that used in the genuine ID card 10, or cases where parts that should have been printed with ink that does not absorb infrared rays have disappeared during the counterfeiting or alteration process. Furthermore, by changing the criteria, etc., for each of the multiple inspection areas 11 according to the ink used, the accuracy of authenticity determination can be improved.

Furthermore, in this embodiment, the processing burden for authenticity determination can be reduced by performing the above discrimination based on the values of the binary images 20 and 30. However, it is also possible to perform the above discrimination using multi-level images and based on the level of grayscale values.

Furthermore, in this embodiment, by determining the validity of the inspection area 11 using Mode 0 and Mode 2, it is possible to determine the validity of multiple inspection areas 11 printed on a genuine ID card 10 using inks with different infrared absorption characteristics, thereby enabling accurate determination of the authenticity of the ID card 10.

Furthermore, in this embodiment, by determining the validity of the inspection area 11 using Mode 0 and Mode 1, it is possible to detect counterfeiting, alteration, etc., of portions printed on a genuine ID card 10 using ink containing infrared-absorbing components, using ink with different infrared absorption characteristics (ink that does not absorb infrared rays). It is also possible to detect cases where portions printed on a genuine ID card 10 using ink that does not absorb infrared rays disappear during the counterfeiting or alteration process.

Furthermore, in this embodiment, the aforementioned dilation process is performed on the binary image 30 of the infrared image, and the dilated binary image is used to determine whether the inspection area 11 is correct or incorrect. This suppresses the decrease in judgment accuracy due to the misalignment between the visible light image and the infrared image.

Furthermore, in this embodiment, the accuracy of authenticity determination can be improved by using spectral images corresponding to the color of the printed portion to determine the correctness of each inspection area 11.

However, the present invention is not limited to the embodiments described above. For example, in this embodiment, B, G, R, and V images are created as spectral images in S103 (see Figure 6). However, depending on the type of ID card 10, some spectral images may not be used. In such cases, instead of binarizing all spectral images in S104, only the spectral images to be used may be selected and binarized. This allows for faster processing.

Furthermore, in this embodiment, in S104, the threshold for binarization is determined from the grayscale histogram of the entire image using Otsu's method. However, for example, in the field for the expiration date 105 of a driver's license, there is a high-density background color. If the threshold for binarization is determined from the grayscale histogram of the entire image, there are cases where the expiration date 105 does not appear due to the density of the background color (both the background color and the expiration date 105 become black pixels). Therefore, in order to make the characters on such high-density backgrounds visible during binarization, it is also possible to extract the inspection region 11 containing characters on a high-density background from the spectral image, and then apply Otsu's method only to that inspection region 11 to perform binarization.

Furthermore, the purpose of authenticity verification is not particularly limited. It can be applied to identity verification during account opening at financial institutions, mobile device contracts, and procedures at various government agencies. The present invention is also applicable to identity verification when applying for various services or issuing media using photo booths.

Preferred embodiments of the present invention have been described above with reference to the attached drawings, but the present invention is not limited to these examples. It will be clear to those skilled in the art that various modifications and alterations can be conceived within the scope of the technical idea disclosed herein, and these will naturally also fall within the technical scope of the present invention.

1: Authenticity determination system 2: Reading device 3: Authenticity determination device 10: ID card 20: Binary image of spectral image 30: Binary image of infrared image 40: Setting information 301: Acquisition means 302: Authenticity determination means

Claims (8)

A device for determining the authenticity of an identification document,
An acquisition means for acquiring visible light images and infrared images of the aforementioned identification document,
Authenticity determination means that, for each of the multiple inspection areas of the aforementioned identification document, determines the number of pixels within the inspection area whose density in the visible light image and the infrared image satisfies a predetermined standard as a discrimination value, determines whether the inspection area is genuine or not based on the comparison result between the discrimination value and a threshold, and determines the authenticity of the identification document based on the determination result,
Equipped with,
Authenticity determination device characterized in that, among a plurality of inspection areas, at least one of the following differs: the predetermined criterion and the relationship between the comparison result and the correctness of the inspection area. The grayscale in the visible light image represents the grayscale of a predetermined color component of the visible light image, or the grayscale indicating the degree of brightness of each pixel in the visible light image.
The aforementioned authenticity determination means is
The authenticity determination device according to claim 1, characterized in that it creates a spectral image which is an image showing the intensity of each color component of the visible light image, and an image which shows the degree of brightness of each pixel of the visible light image as intensity, and uses the intensity of the pixels of the spectral image for the determination. The authenticity determination device according to claim 2, characterized in that the spectral image used for the determination differs between the multiple inspection regions. The aforementioned authenticity determination means is
The authenticity determination device according to claim 2, characterized in that a binary image is created by binarizing the grayscale of the spectral image and the infrared image, and the authenticity of the inspection area is determined based on the pixel values in the binary image. The predetermined criteria differ among the multiple inspection areas.
In one of the inspection areas, the predetermined criterion is that it appears as a dense area in both the spectral image and the infrared image.
The authenticity determination device according to claim 4, characterized in that, in another inspection area, the predetermined standard appears as a dense area in the spectral image and as a pale area in the infrared image. Among the multiple inspection areas, the predetermined criteria are the same, but the relationship between the comparison result and the correctness of the inspection area differs.
The predetermined criterion is that it appears as a dark area in the spectral image and as a light area in the infrared image.
In one of the aforementioned test areas, the test area is considered positive when the discriminant value is less than or equal to the threshold.
The authenticity determination device according to claim 4, characterized in that in another inspection area, the inspection area is deemed to be true when the discrimination value is equal to or greater than the threshold. The authenticity determination means is characterized by performing a dilation process on the binary image obtained by binarizing the grayscale of the infrared image, and using the dilated binary image for the determination, as described in claim 5 or 6. Computers,
A device for determining the authenticity of an identification document,
An acquisition means for acquiring visible light images and infrared images of the aforementioned identification document,
Authenticity determination means that, for each of the multiple inspection areas of the aforementioned identification document, determines the number of pixels within the inspection area whose density in the visible light image and the infrared image satisfies a predetermined standard as a discrimination value, determines whether the inspection area is genuine or not based on the comparison result between the discrimination value and a threshold, and determines the authenticity of the identification document based on the determination result,
Equipped with,
A program for functioning as a genuine/counterfeit determination device, characterized in that, among a plurality of inspection areas, at least one of the following differs: the predetermined criterion or the relationship between the comparison result and the correctness of the inspection area.