Authenticity determination system and authenticity determination program

Abstract

To provide an authenticity determination system that can correctly determine authenticity of a target article provided with a counterfeit feature area in which a feature area of a genuine article is reproduced in shades of grey by comparing a registration image with a collation image, and that reduces effort required for authenticity determination compared to a case in which a plurality of registered images is prepared in advance.SOLUTION: In an authenticity determination system 10, a registration image acquisition device 12 acquires a registration image by photographing a feature area FAt of a genuine article Ot under a first photographing condition. A photographing processing unit of a collation image acquisition device 14 photographs a feature area FAr of a target object Or under the first photographing condition to acquire a first collation image, and photographs the feature area FAr under a second photographing condition to acquire a second collation image. A registered image determination unit of the authenticity determination device 16 determines that the target item Or is a genuine item Ot when inter-collation image similarity between the first collation image and the second collation image is less than a first threshold value, and pair-registered image similarity between the first collation image and the registration image is greater than or equal to a second threshold value.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an authenticity determination system and an authenticity determination program. [Background technology]

[0002] Conventionally, an authenticity determination system has been proposed in which a registered image is obtained in advance by photographing a feature area, which is an area of ​​a genuine item that exhibits unique features of the item, and the authenticity of the target item is determined by comparing a collation image obtained by photographing the feature area of ​​the target item to be authenticated with the registered image. In this specification, "unique feature of an item" refers to a structure that is difficult to control and is randomly different for each item, and that has multiple mirror surfaces (light-reflecting surfaces) facing in various directions. Examples of unique feature of an item include an ink surface printed with ink in which metal particles are randomly dispersed, a matte surface, a grained surface, or a hammer-tone surface.

[0003] The content of the registered image or the match image changes depending on how light is applied to the feature region or the position of the camera lens relative to the feature region (these will be referred to as "photography conditions" in this specification, as will be described in detail later). More specifically, the luminance distribution in the registered image or the match image changes depending on the photography conditions. For example, the distribution of high-luminance pixels, which are significantly brighter than other pixels, changes in the registered image or the match image depending on the photography conditions. This is because, among the multiple mirror surfaces possessed by the unique features of the item, the mirror surface that reflects the specularly reflected light that enters the lens changes depending on the photography conditions. Therefore, in order to determine the authenticity of the target item by comparing the photography image with the registered image, it is necessary to photograph the match image under photography conditions equivalent to those of the registered image.

[0004] As an example of a conventional authenticity determination system, Patent Document 1 discloses an authenticity determination system that reads the surface state of a genuine article when light is irradiated onto the genuine article from a first direction to generate a first registered image, reads the surface state of the genuine article when light is irradiated onto the genuine article from a second direction to generate a second registered image, reads the surface state of a target article when light is irradiated onto the target article from the first direction to generate a first matching image, reads the surface state of the target article when light is irradiated onto the target article from the second direction to generate a second matching image, and determines that the target article is a genuine article if the normalized correlation value between the first registered image and the first matching image is greater than or equal to a predetermined threshold and the normalized correlation value between the second registered image and the second matching image is greater than or equal to a predetermined threshold.

[0005] Furthermore, Patent Document 2 discloses an authenticity determination system that photographs a target item using a smartphone camera or the like to obtain a match image, obtains the positions of pixels that produce specular reflection from the match image, obtains the brightness values ​​of pixels in a registered image that correspond to the obtained positions, and determines that the registered image and the match image match if the dispersion of the obtained brightness values ​​is within a certain range, i.e., that the target item is a genuine item. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent No. 4834968 [Patent Document 2] Patent No. 6864267 Summary of the Invention [Problem to be solved by the invention]

[0007] Now, consider a case where a counterfeit feature area, in which the feature area of ​​a genuine article is reproduced in shades of gray, is provided on a target article. In particular, consider a case where the counterfeit feature area expresses in shades the same luminance distribution as that of the registered image. In this case, when a match image obtained by photographing the counterfeit feature area is compared with the registered image, the similarity between the two images may be quite high. As a result, the authenticity of the target article may not be accurately determined by comparing the registered image with the match image.

[0008] As described above, the luminance distribution of the unique features of a genuine item changes depending on the imaging conditions. On the other hand, the counterfeit feature area, reproduced in shades of gray, does not have a structure composed of multiple mirror surfaces facing in various directions, as does the feature area of ​​a genuine item. Therefore, the luminance distribution in the verification image does not change significantly even when the imaging conditions are changed. Therefore, it is conceivable to first photograph the feature area of ​​the genuine item under a first imaging condition to obtain a first registered image, and then photograph the feature area of ​​the genuine item under a second imaging condition different from the first imaging condition to obtain a second registered image. Then, the authenticity of the target item can be determined by comparing the first registered image obtained by photographing the feature area of ​​the target item in the first photographed image with the first registered image, and comparing the second registered image obtained by photographing the feature area of ​​the target item in the second photographed image with the second registered image. According to this method, if the target item is genuine, the similarity between the first registered image and the first verification image will be high, and the similarity between the second registered image and the second verification image will be high. On the other hand, if the characteristic region of the target article is a forged characteristic region reproduced in shades of gray, the similarity between the first registered image and the first verification image or the similarity between the second registered image and the second verification image should be low, which allows for accurate authentication of the target article.

[0009] However, the above method requires the preparation of multiple registered images (first registered image and second registered image) in advance. The multiple registered images must be photographed under different photographing conditions, which makes it time-consuming to acquire the multiple registered images. Storing multiple registered images for each of a large number of genuine articles can also lead to a problem of straining the memory capacity for storing the registered images. Furthermore, when photographing a match image, not only is it necessary to adjust the photographing conditions to the first photographing conditions when photographing the first match image, but it is also necessary to adjust the photographing conditions to the second photographing conditions when photographing the second match image. Thus, when photographing a match image, not only is the photographing conditions changed, but the adjustment to the predetermined photographing conditions occurs twice, which also requires time and effort.

[0010] The object of the present invention is to ensure that the authenticity of a target item having a counterfeit characteristic area in which the characteristic area of ​​the genuine item is reproduced in shades of gray can be correctly determined by comparing a registered image with a matching image, while reducing the effort required for authenticity determination compared to when multiple registered images are prepared in advance. [Means for solving the problem]

[0011] The invention of claim 1 is an authenticity determination system that determines the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item that displays the item's unique features, and is equipped with a processor.The processor calculates a match image similarity between a first match image obtained by photographing the feature area of ​​the target item under first photographing conditions that are equivalent to the photographing conditions when the registered image was acquired, and a second match image obtained by photographing the feature area of ​​the target item under second photographing conditions that are different from the first photographing conditions, calculates a pair-of-registered image similarity between the first match image and the registered image, and determines that the target item is a genuine item if the match image similarity is less than a first threshold and the pair-of-registered image similarity is greater than or equal to a second threshold. The invention of claim 2 is the authenticity determination system described in claim 1, characterized in that if the similarity between the matched images is greater than or equal to the first threshold, the processor determines that the target item is not a genuine item without calculating the similarity between the registered images. The invention of claim 3 is the authenticity determination system described in claim 2, characterized in that the calculation of the inter-match image similarity is performed by a match image acquisition device that photographs the characteristic area of ​​the target item and acquires the first match image and the second match image, the calculation of the pair-registered image similarity is performed by an authenticity determination device different from the match image acquisition device that is communicatively connected to the match image acquisition device, and the match image acquisition device does not transmit the first match image to the authenticity determination device if the inter-match image similarity is greater than or equal to the first threshold. In the invention according to claim 4, when the inter-match image similarity is equal to or greater than a first threshold, the processor presents a notification to the user urging the user to visually check the characteristic region of the target article. 2. The authenticity determination system according to claim 1, wherein: The invention according to claim 5 is the authenticity determination system according to claim 4, characterized in that the processor presents the registered image to the user together with the notification. The invention of claim 6 is the authenticity determination system described in claim 1, characterized in that the processor displays a guide on the shooting screen of the matching image acquisition device that photographs the characteristic area of ​​the target item, indicating at least one of the position or posture of the matching image acquisition device when photographing the characteristic area of ​​the target item under the first shooting condition, and at least one of the position or posture of the matching image acquisition device when photographing the characteristic area of ​​the target item under the second shooting condition. The invention of claim 7 is an authenticity determination system that determines the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item that displays the item's unique features, and is equipped with a processor.The processor calculates a first pair of registered image similarity between the registered image and a first match image obtained by photographing the feature area of ​​the target item under first photographing conditions that are equivalent to the photographing conditions when the registered image was acquired, and calculates a second pair of registered image similarity between the registered image and a second match image obtained by photographing the feature area of ​​the target item under second photographing conditions that are different from the first photographing conditions, and determines that the target item is a genuine item if the first pair of registered image similarity is greater than or equal to a second threshold and the second pair of registered image similarity is less than a third threshold. The invention of claim 8 is an authenticity determination program that causes a computer to determine the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item that displays the item's unique features, to calculate a match-image similarity between a first match-image obtained by photographing the feature area of ​​the target item under first photographing conditions that are equivalent to the photographing conditions when the registered image was obtained, and a second match-image obtained by photographing the feature area of ​​the target item under second photographing conditions that are different from the first photographing conditions, calculate a pair-of-registered image similarity between the first match-image and the registered image, and determine that the target item is a genuine item if the match-image similarity is less than a first threshold and the pair-of-registered image similarity is greater than or equal to a second threshold. The invention of claim 9 is an authenticity determination program that causes a computer to determine the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item that displays the item's unique features, to calculate a first pair of registered image similarity between the registered image and a first match image obtained by photographing the feature area of ​​the target item under first photographing conditions that are equivalent to the photographing conditions when the registered image was obtained, and calculate a second pair of registered image similarity between the registered image and a second match image obtained by photographing the feature area of ​​the target item under second photographing conditions that are different from the first photographing conditions, and determines that the target item is a genuine item if the first pair of registered image similarity is greater than or equal to a second threshold and the second pair of registered image similarity is less than a third threshold. [Effects of the Invention]

[0012] According to the invention of claim 1, 7, 8, or 9, it is possible to ensure that the authenticity of a target item having a counterfeit characteristic area in which the characteristic area of ​​the genuine item is reproduced in shades by comparing a registered image with a matching image can be correctly determined, while reducing the effort required for authenticity determination compared to when multiple registered images are prepared in advance. According to the invention of claim 2, the amount of calculation required in the authenticity determination process can be reduced compared to when the similarity between registered images is calculated even when the similarity between the matched images is equal to or greater than the first threshold value. According to the invention of claim 3, the amount of communication data between the match image acquisition device and the authenticity determination device during the authenticity determination process can be reduced compared to when the first match image is sent to the authenticity determination device even when the similarity between the match images is equal to or greater than the first threshold value. According to the invention of claim 4, it is possible to prompt the user to visually determine whether or not the target article is a genuine article. According to the invention of claim 5, the user can more easily compare the characteristic region of the target article with the registered image, compared to when the registered image is not presented to the user. According to the invention of claim 6, it is possible to assist in photographing the characteristic region of the target article under the first photographing condition and the second photographing condition. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a schematic diagram illustrating the configuration of an authenticity determination system according to first to third embodiments. [Figure 2] FIG. 10 is a diagram showing an example of a label sticker including a feature region. [Figure 3] 1 is a schematic diagram illustrating the configuration of a match image acquisition device according to a first embodiment. [Figure 4] FIG. 2 is a diagram showing parameters indicating shooting conditions. [Figure 5A] 10A and 10B are diagrams illustrating an example of the relationship between the positions and orientations of a light source, a camera, and a feature region under a first photographing condition. [Figure 5B] FIG. 10 is a diagram showing an example of a guide displayed when capturing an image of a feature region under a first capturing condition. [Figure 6A] 10 is a diagram showing a first example of the relationship between the positions and orientations of a light source, a camera, and a feature region under a second shooting condition. FIG. [Figure 6B] FIG. 10 is a diagram showing a first example of a guide displayed when capturing an image of a feature region under a second capturing condition. [Figure 7] FIG. 10 is a diagram showing a second example of a guide displayed when capturing an image of a feature region under a second capturing condition. [Figure 8A] 10 is a diagram showing a second example of the relationship between the positions and orientations of the light source, the camera, and the feature region under the second shooting condition. FIG. [Figure 8B] FIG. 10 is a third diagram showing an example of a guide displayed when photographing a feature region under the second photographing condition. [Figure 9A] FIG. 10 is a diagram showing vectors indicating the luminance distribution in the first match image. [Figure 9B] FIG. 10 is a diagram showing vectors indicating the luminance distribution in the second match image. [Figure 10] 1 is a schematic diagram illustrating the configuration of an authenticity determination device according to a first embodiment and a third embodiment. [Figure 11] 1 is a conceptual diagram showing the contents of an authenticity determination process in the authenticity determination system according to the first embodiment. FIG. [Figure 12]3 is a flowchart showing the flow of processing in the authenticity determination system according to the first embodiment. [Figure 13] FIG. 10 is a schematic diagram illustrating the configuration of a match image acquisition device according to a second embodiment and a third embodiment. [Figure 14] FIG. 10 is a schematic diagram illustrating the configuration of an authenticity determination device according to a second embodiment. [Figure 15] 10 is a flowchart showing the flow of processing in the authenticity determination system according to the second embodiment. [Figure 16] FIG. 11 is a conceptual diagram showing the contents of the authenticity determination process according to the third embodiment. [Figure 17] 10 is a flowchart showing the flow of processing in the authenticity determination system according to the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] First Embodiment FIG. 1 is a schematic diagram of the configuration of an authenticity determination system 10 according to a first embodiment. The authenticity determination system 10 is a system that photographs a feature area FAt, which is an area in a genuine item Ot that exhibits the item's unique features, to obtain a registered image in advance, photographs a feature area FAr of a target item Or that is the subject of authenticity determination to obtain a match image, and determines the authenticity of the target item Or based on the similarity between the registered image and the match image. In other words, the genuine item Ot is a registered item whose registered image, obtained by photographing its feature area FAt, is registered in advance. Furthermore, authenticity determination refers to determining whether the target item Or is identical to the genuine item (registered item) Ot (in which case it is "genuine") or whether the target item Or is a different item from the genuine item Ot (in which case it is "fake").

[0015] In this specification, when there is no particular distinction between the feature area FAt of the real object Ot and the feature area FAr of the target object Or, they will be referred to as the feature area FA. First, the feature area FA will be described. FIG. 2 is a diagram showing an example of a label sticker L including the feature area FA. In this embodiment, the label sticker L including the feature area FA is attached to the real object Ot and the target object Or. As a result, the feature area FA is formed on the surfaces of the real object Ot and the target object Or.

[0016] The label sticker L includes hologram portions Ha and Hb, a two-dimensional code Q, and a feature area FA.

[0017] The hologram portions Ha and Hb are areas where hologram processing has been performed. Different types of hologram processing have been performed on the hologram portions Ha and Hb. Note that the hologram portions Ha and Hb do not necessarily have to be formed on the label sticker L.

[0018] The two-dimensional code Q is, for example, a QR code (registered trademark). When the two-dimensional code Q is photographed with a photographing device (the registration image acquisition device 12 or the match image acquisition device 14), the photographing device can easily identify the position of the two-dimensional code Q in the photographed image. In this embodiment, the relative positional relationship between the two-dimensional code Q and the characteristic area FA is predetermined. That is, the photographing device can determine the position of the characteristic area FA by identifying the position of the two-dimensional code Q. Furthermore, the two-dimensional code Q may be an encoded label identifier (e.g., a character string) that uniquely identifies the label sticker L (i.e., the characteristic area FA). That is, when the two-dimensional code Q is read with a camera, the label identifier can be obtained.

[0019] The feature area FA is an area where a feature specific to the article (that is, specific to the label sticker L in this embodiment) appears. As described above, the feature specific to the article means a structure that is difficult to form in a controlled manner, that is randomly different for each article, and that has a large number of mirror surfaces facing in various directions.

[0020] In this embodiment, the feature area FA is a printed area printed with gravure ink containing metal particles (e.g., silver particles). In such a printed area, the metal particles contained in the gravure ink form a mirror surface that randomly faces in various directions. However, the feature area FA is not limited to such a printed area as long as it expresses the unique characteristics of the item. For example, the feature area FA may be a matte surface, a textured surface, a hammer-tone surface, or the like. The feature area FA does not need to be formed on the label sticker L, but may be formed directly on the surface of the real item Ot and the target item Or.

[0021] The characteristic area FAr of the target object Or may be counterfeited by a malicious person. In particular, a counterfeit characteristic area FAr, which is a reproduction of the characteristic area FAt of the real object Ot in shades of gray, may be provided on the target object Or.

[0022] 1, the authenticity determination system 10 includes a registration image acquisition device 12, a match image acquisition device 14, and an authenticity determination device 16. The registration image acquisition device 12 and the authenticity determination device 16, and the match image acquisition device 14 and the authenticity determination device 16 are each communicatively connected via a communication line such as a WAN (Wide Area Network) or a LAN (Local Area Network).

[0023] The registered image acquisition device 12 is a device that captures a registered image by photographing a feature area FAt of a real item Ot. The registered image acquisition device 12 may be dedicated equipment for acquiring registered images, such as that used by a business. The registered image acquisition device 12 includes a light source 20 and a camera 22. The light source 20 includes, for example, an LED (Light Emitting Diode) and emits light toward the feature area FAt. The camera 22 is a digital camera that includes a lens, an image sensor, and the like, and receives light from the feature area FAt to form a registered image.

[0024] The registration image acquisition device 12 captures an image of the feature region FAt under predetermined capture conditions. In this specification, the capture conditions refer to the relative position and orientation relationship between the light source 20, the camera 22 (specifically, its lens), and the feature region FAt, as well as whether or not light is irradiated from the light source 20 to the feature region FAt. Details of the capture conditions will be described later together with the description of the match image acquisition device 14.

[0025] In this embodiment, the registered image acquisition device 12 photographs the feature area FAt under a single predetermined photographing condition, and acquires one registered image for one feature area FAt (i.e., one real object Ot). In this specification, the photographing condition when acquiring the registered image is referred to as the first photographing condition. Figure 1 shows, as an example of the first photographing condition, a photographing condition in which the camera 22 is positioned on a normal to the surface of the feature area FAt, and the angle between a line extending from the light source 20 toward the feature area FAt and a line extending from the feature area FAt toward the camera 22 is φ.

[0026] The camera 22 has a communication interface, and the camera 22 associates the formed registered image, information indicating the first photographing conditions, and a label identifier indicated by the two-dimensional code Q included in the label sticker L including the characteristic area FAt, and transmits them to the authenticity determination device 16. In the authenticity determination device 16, the registered image, information indicating the first photographing conditions, and the label identifier are stored in association with each other.

[0027] The match image acquisition device 14 is a device that captures a characteristic area FAr of the target item Or to acquire a match image. Note that in this specification, the true item Ot and the target item Or are described as being distinct conceptually, but in reality, there are cases where the target item Or and the true item Ot are the same (where the authenticity determination system 10 determines the target item Or as being "genuine").

[0028] 3 is a schematic diagram of the configuration of the match image acquisition device 14 according to the first embodiment. The registration image acquisition device 12 may be configured with dedicated equipment for acquiring registration images, while the match image acquisition device 14 may be a device used by general users to determine the authenticity of items. For example, the match image acquisition device 14 may be a mobile terminal such as a smartphone.

[0029] The light source 30 includes, for example, an LED, etc. The light source 30 emits light toward the feature region FAr.

[0030] The camera 32 is a digital camera that includes a lens, an image sensor, etc. The camera 32 receives light from the characteristic region FAr and forms a match image.

[0031] The communication interface 34 is configured by, for example, a network adapter, etc. The communication interface 34 performs the function of communicating with the authenticity determination device 16 via a communication line.

[0032] The input interface 36 is configured by, for example, buttons, a touch panel, etc. The input interface 36 is used when the user inputs instructions to the match image acquisition device 14.

[0033] The display 38 is configured by, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, etc. Various screens are displayed on the display 38 in response to instructions from a processor 44, which will be described later.

[0034] The memory 40 includes a hard disk drive (HDD), an embedded multi-media card (eMMC), a read-only memory (ROM), a random access memory (RAM), or the like. The memory 40 stores a match image acquisition program for operating each unit of the match image acquisition device 14. The match image acquisition program constitutes part of an authenticity determination program. The match image acquisition program may also be stored in a computer-readable non-transitory storage medium such as a universal serial bus (USB) memory or an SD card. The match image acquisition device 14 can read and execute the match image acquisition program from such a storage medium. As shown in FIG. 3, the memory 40 also stores a first match image 42a and a second match image 42b acquired by an image processing unit 46 (described later) by capturing an image of a characteristic region FAr.

[0035] The processor 44 is configured by, for example, a CPU (Central Processing Unit) etc. The processor 44 is communicably connected to the light source 30, the camera 32, the communication interface 34, the input interface 36, the display 38, and the memory 40 via a data bus. The processor 44 performs the functions of a photography processing unit 46, a match image similarity calculation unit 48, a match image determination unit 50, a transmission processing unit 52, and a notification processing unit 54 by using a match image acquisition program stored in the memory 40.

[0036] In response to an instruction from the user, the photography processing unit 46 photographs the characteristic region FAr of the target object Or using the light source 30 and the camera 32 to acquire a first match image 42a and a second match image 42b. In particular, the photography processing unit 46 photographs the characteristic region FAr of the target object Or under first photographing conditions that are the same photographing conditions as when the registered image acquisition device 12 acquired the registered image to acquire the first match image 42a, and photographs the characteristic region FAr of the target object Or under second photographing conditions that are different from the first photographing conditions to acquire the second match image 42b.

[0037] The details of the photographing conditions will be described with reference to Fig. 4. Fig. 4 is a diagram showing parameters indicating the photographing conditions. Note that, while the parameters indicating the photographing conditions for photographing using the match image acquisition device 14 will be described with reference to Fig. 4, the photographing conditions for photographing using the registered image acquisition device 12 can also be expressed using similar parameters. In that case, in the following description, the match image acquisition device 14 can be read as the registered image acquisition device 12, the light source 30 as the light source 20, the camera 32 as the camera 22, and the feature region FAr as the feature region FAt. Furthermore, equivalent photographing conditions are photographing conditions in which the parameters indicating the photographing conditions for the match image and the registered image fall within predetermined ranges.

[0038] Similar to the photographing conditions in the registration image acquisition device 12, the photographing conditions in the match image acquisition device 14 are a concept that includes the relative positional and pose relationships between the light source 30, the camera 32 (more specifically, its lens), and the feature region FAr, as well as whether or not light is irradiated from the light source 30 onto the feature region FAr. Whether or not light is irradiated from the light source 30 onto the feature region FAr differs only in whether the light source 30 is turned on or off, and therefore the following description will focus on the relative positional and pose relationships between the light source 30, the camera 32, and the feature region FAr.

[0039] In Fig. 4, a point on the feature region FAr is set as the origin O, the axis parallel to (the surface of) the feature region FAr is set as the x-axis, the axis parallel to the feature region FAr and perpendicular to the x-axis is set as the y-axis, and the axis perpendicular to the x-axis and y-axis and normal to the feature region FAr is set as the z-axis. As described above, in the feature region FAr, many mirror surfaces facing in various directions are formed, and the line N represents the normal to the mirror surface at the origin O. The angle formed by the xy plane and the line N is called the angle α.

[0040] When light emitted from light source 30 is specularly reflected by the mirror surface and this specularly reflected light enters camera 32, the angle formed by the line extending from light source 30 to origin O and line N is the same as the angle formed by the line extending from origin O to camera 32 and line N. Therefore, point C1, which is the intersection of line N and the line connecting light source 30 and camera 32, is the midpoint of the line connecting light source 30 and camera 32. Here, the foot of the perpendicular line dropped from point C1 onto the xy plane is defined as point C2. The angle formed by line S extending from origin O to point C and the x-axis is called angle β.

[0041] The relative positions and orientations among the light source 30, the camera 32, and the feature region FAr can be expressed by angles α and β. That is, the user can acquire the first match image 42a by photographing the feature region FAr after adjusting the position and orientation of the match image acquisition device 14 so that the angles α and β match the first photographing conditions. The user can also acquire the second match image 42b by photographing the feature region FAr after changing the position and orientation of the match image acquisition device 14 so as to change at least one of the angles α and β from the first photographing conditions.

[0042] Methods for changing at least one of the angles α and β include, for example, changing the position of the matching image acquisition device 14 while maintaining the orientation of the matching image acquisition device 14 relative to the feature area FAr, or changing the orientation of the matching image acquisition device 14 relative to the feature area FAr.

[0043] In this embodiment, the shooting processing unit 46 forms the first matching image 42a and the second matching image 42b by taking still images under the first shooting condition and the second shooting condition, respectively. However, the shooting processing unit 46 may also take a moving image using the camera 32 and extract the first matching image 42a taken under the first shooting condition and the second matching image 42b taken under the second shooting condition from the group of frames that make up the moving image.

[0044] If the user knows the first photographing condition, the user can photograph the feature region FAr under the first photographing condition. However, there are cases where the user does not know the first photographing condition. To prepare for such a case, when photographing the feature region FAr, the photographing processing unit 46 may display, on the photographing screen displayed on the display 38 of the match image acquisition device 14, a guide indicating at least one of the position and orientation of the match image acquisition device 14 when photographing the feature region FAr under the first photographing condition and at least one of the position and orientation of the match image acquisition device 14 when photographing the feature region FAr under the second photographing condition.

[0045] As described above, the registered image acquisition device 12 transmits information indicating the first photographing conditions (for example, the above-mentioned angles α and β) along with the registered image to the authenticity determination device 16. Therefore, the photographing processing unit 46 acquires information indicating the first photographing information from the authenticity determination device 16 and displays a guide based on the information. Note that the authenticity determination device 16 may store multiple registered images for multiple genuine articles Ot, but the first photographing conditions when photographing the characteristic areas FAt of each genuine article Ot may be the same.

[0046] Various guide display modes will be described with reference to FIGS. 5A to 8B. FIG. 5A illustrates the positional and pose relationships between the light source 30, the camera 32, and the feature region FAr under the first photographing condition. FIG. 5B illustrates an example of a guide displayed when photographing the feature region FAr under the first photographing condition. When the positional and pose relationships between the light source 30, the camera 32, and the feature region FAr under the first photographing condition are as shown in FIG. 5A, i.e., when the feature region FAr is on the optical axis of the camera 32, the photographing processing unit 46 displays a rectangular frame FL as a guide in the approximate center of the photographing screen displayed on the display 38, as shown in FIG. 5B. When the user adjusts the position and pose of the verification image acquisition device 14 so that the feature region image FArI, which is the image of the feature region FAr displayed on the photographing screen, fits exactly within the frame FL, the positional and pose relationships between the light source 30, the camera 32, and the feature region FAr become the same as those under the first photographing condition. In this way, the frame FL allows the user to easily photograph the feature region FAr under the first photographing condition.

[0047] 6A is a diagram showing a first example of the relationship between the positions and orientations of the light source 30, the camera 32, and the feature region FAr under a second photographing condition (i.e., a photographing condition different from the first photographing condition). FIG. 6B is a diagram showing a first example of a guide displayed when photographing the feature region FAr under the second photographing condition. When the relationship between the positions and orientations of the light source 30, the camera 32, and the feature region FAr under the second photographing condition is as shown in FIG. 6A, that is, when the position of the feature region FAr under the second photographing condition is located to the side of the match image acquisition device 14 (i.e., the photographing region of the camera 32) compared to the first photographing condition shown in FIG. 5A, the photographing processing unit 46 displays a rectangular frame FL as a guide, shifted to the side of the photographing screen displayed on the display 38, as shown in FIG. 6B. In this case, too, if the user adjusts the position and orientation of the match image acquisition device 14 so that the feature region image FArI fits exactly within the frame FL, the relationship in position and orientation between the light source 30, the camera 32, and the feature region FAr will become a relationship (i.e., a second photographing condition) different from that in the first photographing condition. In this way, the frame FL allows the user to easily photograph the feature region FAr under the second photographing condition.

[0048] Fig. 7 is a diagram showing a second example of a guide displayed when capturing an image of the characteristic region FAr under the second capturing condition. When the second capturing condition is achieved by rotating the match image acquisition device 14 about the z-axis (see Fig. 4) while maintaining the orientation of the characteristic region FAr and the match image acquisition device 14 (i.e., the light source 30 and the camera 32) from the first capturing condition shown in Fig. 5A, the capturing processing unit 46 rotates and displays the frame FL (see Fig. 5B) that was displayed when capturing an image of the characteristic region FAr under the first capturing condition, as shown in Fig. 7.

[0049] FIG. 8A illustrates a second example of the relationship between the positions and orientations of the light source 30, camera 32, and feature region FAr under the second photographing condition. FIG. 8B illustrates a third example of a guide displayed when photographing the feature region FAr under the second photographing condition. Under the second photographing condition illustrated in FIG. 8A, the orientation of the camera 32 is changed so that the optical axis of the camera 32 is not parallel to the normal to the feature region FAr, as compared to the first photographing condition illustrated in FIG. 5A. In this case, as illustrated in FIG. 8B, the photographing processing unit 46 displays a trapezoidal frame FL as a guide. In this case, the user can adjust the position and orientation of the verification image acquisition device 14 so that the trapezoidal feature region image FArI on the photographing screen fits exactly within the frame FL due to the optical axis of the camera 32 not being parallel to the normal to the feature region FAr. This allows the relationship between the positions and orientations of the light source 30, camera 32, and feature region FAr to be set to the second photographing condition. In this way, the frame FL allows the user to easily photograph the feature region FAr under the second photographing condition.

[0050] As described above, in this embodiment, a frame FL is displayed on the shooting screen as a guide, but the guide is not limited to the frame FL and may be any display as long as it indicates at least one of the position and orientation of the match image acquisition device 14 when the feature region FAr is photographed under each of the first and second shooting conditions. For example, a picture showing the relationship between the positions and orientations of the light source 30, camera 32, and feature region FAr under the first and second shooting conditions (for example, if the match image acquisition device 14 is a smartphone, a picture showing the position and orientation of the smartphone) may be displayed as a guide.

[0051] In this embodiment, the photographing processing unit 46 photographs the characteristic region FAr and the two-dimensional code Q included in the label sticker L to obtain a first match image 42a and a second match image 42b.

[0052] The inter-match image similarity calculation unit 48 calculates the similarity between the first match image 42a and the second match image 42b (referred to herein as "inter-match image similarity"). As described above, the first match image 42a and the second match image 42b may represent a characteristic region FAr and a two-dimensional code Q. When calculating the inter-match image similarity, the "first match image 42a" refers to only the portion of the first match image 42a that represents the characteristic region FAr, and the "second match image 42b" refers to only the portion of the second match image 42b that represents the characteristic region FAr. In other words, the inter-match image similarity refers to the similarity between the portion of the characteristic region FAr in the first match image 42a and the portion of the characteristic region FAr in the second match image 42b. Various methods can be used to calculate the inter-match image similarity.

[0053] For example, the inter-matching image similarity calculation unit 48 can calculate a correlation value between the pixel values ​​(e.g., brightness values) of each pixel constituting the first matching image 42a and the pixel values ​​of each pixel constituting the second matching image 42b, and use the correlation value as the inter-matching image similarity.

[0054] The correlation value can be expressed, for example, as the difference between a vector representing the luminance distribution in the first match image 42a and a vector representing the luminance distribution in the second match image 42b. FIG. 9A is a diagram showing a vector Va representing the luminance distribution in the first match image 42a. FIG. 9B is a diagram showing a vector Vb representing the luminance distribution in the second match image 42b. The directions of the vectors Va and Vb represent the direction of change in the luminance distribution, and the magnitudes of the vectors Va and Vb represent the magnitude of the change in the luminance distribution. For example, in FIG. 9A, vector Va points downward, which indicates that the lower portion of the first match image 42a as a whole has higher luminance than the upper portion. The larger the magnitude of vector Va, the higher the luminance of the lower portion than the upper portion. On the other hand, in FIG. 9B, vector Vb points upward and to the right, which indicates that the lower left portion of the second match image 42b as a whole has higher luminance than the lower left portion. The larger the magnitude of vector Vb, the higher the luminance of the lower left portion.

[0055] 9B represents the magnitude of the difference between the first match image 42a and the second match image 42b. Therefore, the smaller the magnitude of vector Vd, the larger the correlation value between the first match image 42a and the second match image 42b calculated by inter-match image similarity calculation unit 48. Conversely, the larger the magnitude of vector Vd, the smaller the correlation value between the first match image 42a and the second match image 42b calculated by inter-match image similarity calculation unit 48.

[0056] Alternatively, the match image similarity calculation unit 48 may obtain a difference image between the first match image 42a and the second match image 42b and calculate a correlation value based on statistics of pixel values ​​of each pixel in the difference image (for example, the difference between the luminance value of the first match image 42a and the luminance value of the second match image 42b). In this case, the smaller the pixel values ​​of the difference image as a whole, the larger the correlation value between the first match image 42a and the second match image 42b that the match image similarity calculation unit 48 calculates to be. Conversely, the larger the pixel values ​​of the difference image as a whole, the smaller the correlation value between the first match image 42a and the second match image 42b that the match image similarity calculation unit 48 calculates to be.

[0057] Inter-match image similarity calculation unit 48 may calculate, as the inter-match image similarity, the similarity between a part of first match image 42a and a corresponding part of second match image 42b, rather than the similarity between the entire first match image 42a and the entire second match image 42b. For example, the inter-match image similarity may be calculated as the similarity between a group of pixels that make up first match image 42a and have a pixel value equal to or greater than a predetermined pixel threshold (for example, a predetermined luminance value or greater) and a corresponding group of pixels in second match image 42b.

[0058] In addition, the inter-matching image similarity calculation unit 48 may input the first match image 42a and the second match image 42b into a learning model trained to calculate the similarity between two images, and calculate the inter-matching image similarity based on the output data of the learning model.

[0059] The match image determination unit 50 determines whether the inter-match image similarity calculated by the inter-match image similarity calculation unit 48 is less than a first threshold value. The first threshold value may be determined in advance by an administrator of the authenticity determination system 10, for example.

[0060] If the target object Or is a real object Ot, then the feature region FAr should be configured with a large number of mirror surfaces facing in various directions, just like the feature region FAr of the real object Ot. In this case, the first match image 42a and the second match image 42b acquired by photographing the feature region FAr under different photographing conditions should be different images (more specifically, images with different luminance distributions). In other words, the similarity between the match images should be low. This is because, for example, the mirror surface that causes specular reflection of light from the light source 30 to enter the camera 32 changes as the photographing conditions change.

[0061] On the other hand, if the feature region FAr is a fake feature region reproduced using shading, the feature region FAr does not have many mirror surfaces facing in various directions. Therefore, even if the photographing conditions change, the mirror surfaces that cause specularly reflected light from the light source 30 to enter the camera 32 do not change. Therefore, the first match image 42a and the second match image 42b obtained by photographing the feature region FAr under different photographing conditions should be similar. In other words, the similarity between the match images should be high.

[0062] Therefore, if the inter-match image similarity calculated by the inter-match image similarity calculation unit 48 is less than the first threshold, the match image determination unit 50 determines that there is a possibility that the target item Or is the true item Ot (it is not determined that the target item Or is the true item Ot based on the inter-match image similarity alone). On the other hand, if the inter-match image similarity calculated by the inter-match image similarity calculation unit 48 is equal to or greater than the first threshold, the match image determination unit 50 determines at this point that the target item Or is not the true item Ot, without using the registered image acquired by the above-mentioned registered image acquisition device 12.

[0063] The transmission processing unit 52 transmits the first match image 42a (in which the characteristic region FAr and the two-dimensional code Q are displayed) to the authenticity determination device 16. When calculating the inter-match image similarity, the first match image 42a refers to only the portion of the first match image 42a that includes the characteristic region FAr. Therefore, the transmission processing unit 52 may transmit the first match image 42a, which is an image of the characteristic region FAr, and the label identifier indicated by the two-dimensional code Q. Here, in this embodiment, if the match image determination unit 50 determines that the inter-match image similarity is less than the first threshold, that is, if it determines that there is a possibility that the target item Or is the true item Ot, the transmission processing unit 52 transmits the first match image 42a to the authenticity determination device 16. In other words, if the match image determination unit 50 determines that the inter-match image similarity is equal to or greater than the first threshold, that is, if it determines that the target item Or is not the true item Ot, the transmission processing unit 52 does not transmit the first match image 42a to the authenticity determination device 16.

[0064] When the match image determination unit 50 determines that the inter-match image similarity is equal to or greater than the first threshold, the notification processing unit 54 notifies the user that the target object Or has been determined to be not a genuine object Ot. For example, the notification processing unit 54 displays a message indicating the determination result by the match image determination unit 50 on the display 38. Furthermore, when the match image determination unit 50 determines that the inter-match image similarity is equal to or greater than the first threshold, the notification processing unit 54 presents the user with a notification urging the user to visually inspect the characteristic region FAr of the target object Or. This is because, when the inter-match image similarity is equal to or greater than the first threshold, the characteristic region FAr is likely to be a forgery. For example, the notification processing unit 54 displays a message urging the user to visually inspect the characteristic region FAr of the target object Or on the display 38 of the match image acquisition device 14. Furthermore, the notification processing unit 54 may provide an audio notification instead of or in addition to displaying the message. For example, when the characteristic region FAr is a forgery characteristic region reproduced in shades of gray, it is often possible to detect that it is a forgery by visual inspection.

[0065] In addition, the notification processing unit 54 may present the user with a registered image to be compared with the characteristic region FAr along with the above-mentioned notification. Here, the authenticity determination device 16 may store a large number of registered images corresponding to a large number of genuine articles Ot. The notification processing unit 54 first transmits a label identifier obtained by reading the two-dimensional code Q included in the first match image 42a or the second match image 42b with the camera 32 to the authenticity determination device 16. The authenticity determination device 16 then transmits the registered image associated with the label identifier to the match image acquisition device 14. The notification processing unit 54 then displays the received registered image on the display 38 for presentation to the user. This allows the user to easily compare the registered image with the characteristic region FAr.

[0066] 10 is a schematic diagram of the configuration of the authenticity determination device 16 according to the first embodiment. The authenticity determination device 16 is configured, for example, by a server computer. However, the authenticity determination device 16 may be any device as long as it can perform the functions described below. The functions described below may also be distributed and performed by multiple devices. In this case, the multiple devices are collectively referred to as an authenticity determination subsystem.

[0067] The communication interface 60 is configured by, for example, a network adapter, etc. The communication interface 60 performs the function of communicating with the registration image acquisition device 12 and the match image acquisition device 14 via a communication line.

[0068] The memory 62 includes an HDD, eMMC, ROM, RAM, or the like. The memory 62 stores a registered image determination program for operating each unit of the authenticity determination device 16. The registered image determination program constitutes part of the authenticity determination program. The registered image determination program may also be stored in a computer-readable non-transitory storage medium such as a USB memory or an SD card. The authenticity determination device 16 can read and execute the registered image determination program from such a storage medium. As shown in FIG. 10, the memory 62 also stores a registered image 64 received from the registered image acquisition device 12. The memory 62 may store multiple registered images 64 corresponding to multiple genuine items Ot, respectively. The memory 62 also stores information indicating a first photographing condition.

[0069] The processor 66 is configured by, for example, a CPU, etc. The processor 66 is communicably connected to the communication interface 60 and the memory 62 via a data bus. The processor 66 performs the functions of a paired registered image similarity calculation unit 68, a registered image determination unit 70, and a notification processing unit 72 by using a registered image determination program stored in the memory 62.

[0070] The paired registered image similarity calculation unit 68 calculates the similarity (herein referred to as "paired registered image similarity") between the first match image 42a received from the match image acquisition device 14 and the registered image 64 stored in the memory 62. As described above, the memory 62 can store a large number of registered images corresponding to a large number of real articles Ot, and the paired registered image similarity calculation unit 68 identifies the registered image 64 for which the similarity with the first match image 42a is to be calculated, based on the label identifier indicated by the two-dimensional code Q included in the first match image 42a or the second match image 42b received from the match image acquisition device 14.

[0071] The method of calculating the paired registered image similarity by the paired registered image similarity calculation section 68 may be the same as the method of calculating the inter-match image similarity by the inter-match image similarity calculation section 48 .

[0072] Furthermore, if the match image determination unit 50 of the match image acquisition device 14 determines that the target item Or is not the true item Ot, the first match image 42a is not transmitted from the match image acquisition device 14 to the authenticity determination device 16, and therefore the paired registered image similarity calculation unit 68 naturally does not calculate the paired registered image similarity.

[0073] The registered image determination unit 70 determines whether the paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is less than a second threshold. The second threshold may be determined in advance by, for example, an administrator of the authenticity determination system 10. The second threshold may be the same as or different from the first threshold that serves as the basis for determination by the match image determination unit 50 of the match image acquisition device 14.

[0074] Here, if the target article Or is the true article Ot, the first match image 42a was acquired under the same photographing conditions (i.e., the first photographing conditions) as the registered image 64, and therefore the pair-registered image similarity should be high. On the other hand, if the target article Or is not the true article Ot, the first match image 42a was acquired by photographing a feature area FAr that is different from the feature area FAt of the true article Ot, and therefore the pair-registered image similarity should be low.

[0075] Therefore, the registered image determination unit 70 determines that the target item Or is the true item Ot when the paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is equal to or greater than the second threshold. On the other hand, the registered image determination unit 70 determines that the target item Or is not the true item Ot when the paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is less than the second threshold.

[0076] FIG. 11 is a conceptual diagram illustrating the details of the authenticity determination process in the authenticity determination system 10 according to the first embodiment. In the first embodiment, the authenticity determination process is performed by the match image determination unit 50 and the registered image determination unit 70 working together. In the first embodiment, the match image determination unit 50 of the match image acquisition device 14 first determines, as Determination 1, whether the similarity between the first match image 42a and the second match image 42b, i.e., the inter-match image similarity, is above a first threshold. If, in Determination 1, it is determined that the first match image 42a and the second match image 42b are similar, i.e., the inter-match image similarity is above the first threshold, the match image determination unit 50 determines that the target item Or is not the genuine item Ot.

[0077] On the other hand, if it is determined in Determination 1 that the first match image 42a and the second match image 42b are dissimilar, i.e., the inter-match image similarity is less than the first threshold, the process proceeds to Determination 2. In Determination 2, the registered image determination unit 70 of the authenticity determination device 16 determines whether the similarity between the first match image 42a and the registered image 64, i.e., the pair-registered image similarity, is equal to or greater than a second threshold. If it is determined in Determination 2 that the first match image 42a and the registered image 64 are similar, i.e., the pair-registered image similarity is equal to or greater than the second threshold, the registered image determination unit 70 determines that the target item Or is the true item Ot. On the other hand, if it is determined in Determination 2 that the first match image 42a and the registered image 64 are dissimilar, i.e., the pair-registered image similarity is less than the second threshold, the registered image determination unit 70 determines that the target item Or is not the true item Ot.

[0078] As described above, when a counterfeit characteristic area in which the characteristic area FAt of the genuine article Ot is reproduced in shades is provided on the target article Or, it may be difficult to determine that the target article Or is not the genuine article Ot simply by comparing the first match image 42a with the registered image 64. However, according to the authenticity determination system 10 of this embodiment, not only the first match image 42a with the registered image 64 but also the first match image 42a with the second match image 42b are compared, so that even when a counterfeit characteristic area in which the characteristic area FAt of the genuine article Ot is reproduced in shades is provided on the target article Or, it is possible to automatically determine that the target article Or is not the genuine article Ot.

[0079] In the first embodiment, judgment 1 is performed by match image acquisition device 14, and judgment 2 is performed by authenticity determination device 16, and judgments 1 and 2 are performed by different devices. Here, in the first embodiment, if it is determined in judgment 1 that first match image 42a and second match image 42b are similar, first match image 42a is not transmitted from match image acquisition device 14 to authenticity determination device 16. This has the effect of reducing the amount of communication data between match image acquisition device 14 and authenticity determination device 16 in the authenticity determination process.

[0080] In this embodiment, one registered image 64 is acquired for one feature region FAt (i.e., one genuine article Ot), and authenticity is determined based on the paired registered image similarity between the registered image 64 and the first match image 42a. However, multiple registered images 64 may be acquired for one feature region FAt by photographing the feature region FAt under different photographing conditions. For example, a first registered image acquired under a first photographing condition and a second registered image acquired under a third photographing condition may be acquired. In this case, the paired registered image similarity calculation unit 68 calculates the paired registered image similarity between the first match image 42a and the first registered image, and the paired registered image similarity between the third match image acquired under the third photographing condition and the third registered image. The registered image determination unit 70 may then determine that the target item Or is the true item Ot if both of the two paired registered image similarities are greater than or equal to a second threshold, and may determine that the target item Or is not the true item Ot if at least one of the two paired registered image similarities is less than the second threshold.

[0081] The notification processing unit 72 notifies the user of the determination result by the registered image determination unit 70. For example, the notification processing unit 72 displays a message indicating the determination result by the registered image determination unit 70 on the display 38 of the match image acquisition device 14 used by the user.

[0082] 12 is a flowchart showing the processing flow of the authenticity determination system 10 according to the first embodiment. At the start of the flowchart shown in FIG. 12, it is assumed that the registered image 64 has already been acquired by the registered image acquisition device 12 and stored in the memory 62 of the authenticity determination device 16.

[0083] In step S10, the photographing processing unit 46 of the match image acquisition device 14 photographs the characteristic region FAr of the target object Or under the first photographing condition, thereby acquiring the first match image 42a.

[0084] In step S12, the photographing processing unit 46 photographs the characteristic region FAr of the target object Or under the second photographing conditions, thereby acquiring the second match image 42b.

[0085] In step S14, the inter-match image similarity calculation unit 48 calculates the similarity between the first match image 42a and the second match image 42b, that is, the inter-match image similarity.

[0086] In step S16, the match image determination unit 50 of the match image acquisition device 14 determines whether the inter-match image similarity calculated in step S14 is less than the first threshold value. If the inter-match image similarity is equal to or greater than the first threshold value ("No" in step S18), the process proceeds to step S18.

[0087] In step S18, the match image determination unit 50 determines that the target item Or is not the true item Ot. Then, the process proceeds to step S20, where the notification processing unit 54 of the match image acquisition device 14 notifies the user of the determination result in step S16.

[0088] In step S16, if the match image determination unit 50 determines that the inter-match image similarity is less than the first threshold value ("Yes" in step S18), the process proceeds to step S22.

[0089] In step S22, the transmission processing unit 52 of the match image acquisition device 14 transmits the first match image 42a to the authenticity determination device 16.

[0090] In step S24, the paired registered image similarity calculation unit 68 of the authenticity determination device 16 calculates the similarity between the first match image 42a received in step S22 and the registered image 64 stored in the memory 62, i.e., the paired registered image similarity.

[0091] In step S26, the registered image determination unit 70 of the authenticity determination device 16 determines whether the pair-registered image similarity calculated in step S24 is equal to or greater than a second threshold. If the pair-registered image similarity is equal to or greater than the second threshold ("Yes" in step S26), the process proceeds to step S28. In step S28, the registered image determination unit 70 determines that the target item Or is a genuine item Ot.

[0092] In step S26, if the registered image determination unit 70 determines that the paired registered image similarity is less than the second threshold ("Yes" in step S26), the process proceeds to step S30. In step S30, the registered image determination unit 70 determines that the target item Or is not the true item Ot.

[0093] In step S32, the notification processing unit 72 of the authenticity determination device 16 transmits information indicating the determination result in step S28 or step S30 to the match image acquisition device 14. Then, in the subsequent step S20, the notification processing unit 54 of the match image acquisition device 14 notifies the user of the determination result based on the information received in step S32.

[0094] Second Embodiment Fig. 13 is a schematic diagram of the configuration of a match image acquisition device 14-2 according to the second embodiment. Fig. 14 is a schematic diagram of the configuration of an authenticity determination device 16-2 according to the second embodiment. The schematic diagram of the configuration of the authenticity determination system 10 according to the second embodiment is similar to that of the first embodiment shown in Fig. 1. In the description of the second embodiment, descriptions of the same parts as in the first embodiment will be omitted.

[0095] In the first embodiment, the inter-match image similarity between first match image 42a and second match image 42b was calculated by match image acquisition device 14, but in the second embodiment, the inter-match image similarity is calculated by authenticity determination device 16-2. Therefore, as shown in Fig. 13, processor 44 of match image acquisition device 14-2 does not have the functions of inter-match image similarity calculation unit 48 and match image determination unit 50, and as shown in Fig. 14, processor 66 of authenticity determination device 16-2 has the functions of inter-match image similarity calculation unit 48 and match image determination unit 50.

[0096] In the second embodiment, the photographing processing unit 46 of the match image acquisition device 14-2 acquires the first match image 42a and the second match image 42b, and the transmission processing unit 52 transmits the first match image 42a and the second match image 42b to the authenticity determination device 16-2.

[0097] Thereafter, as in the first embodiment, the matching image similarity calculation unit 48 of the authenticity determination device 16-2 calculates the matching image similarity between the first matching image 42a and the second matching image 42b, and the matching image determination unit 50 of the authenticity determination device 16-2 determines whether the matching image similarity is less than the first threshold value.

[0098] In the second embodiment, the paired registered image similarity calculation unit 68 calculates the paired registered image similarity between the first match image 42a and the registered image 64, and the registered image determination unit 70 determines whether the paired registered image similarity is less than the second threshold value.

[0099] In the second embodiment, the determination by the match image determination unit 50 of whether the similarity between match images is less than the first threshold value or the determination by the registered image determination unit 70 of whether the similarity between paired registered images is less than the second threshold value may be performed first.

[0100] Then, the match image determination unit 50 or the registered image determination unit 70 determines that the target item Or is the true item Ot when the inter-match image similarity is less than the first threshold and the paired registered image similarity is equal to or greater than the second threshold. On the other hand, the match image determination unit 50 or the registered image determination unit 70 determines that the target item Or is not the true item Ot when the inter-match image similarity is equal to or greater than the first threshold or the paired registered image similarity is less than the second threshold. Thus, in the second embodiment, the authenticity determination process is executed by the registered image determination unit 70.

[0101] Fig. 15 is a flowchart showing the processing flow of the authenticity determination system 10 according to the second embodiment. At the start of the flowchart shown in Fig. 15, it is assumed that the registered image 64 has already been acquired by the registered image acquisition device 12 and stored in the memory 62 of the authenticity determination device 16-2.

[0102] In step S40, the photographing processing unit 46 of the match image acquisition device 14-2 photographs the characteristic region FAr of the target object Or under the first photographing condition, thereby acquiring the first match image 42a.

[0103] In step S42, the photographing processing unit 46 photographs the characteristic region FAr of the target object Or under the second photographing conditions, thereby acquiring the second match image 42b.

[0104] In step S44, the transmission processing unit 52 of the match image acquisition device 14-2 transmits the first match image 42a and the second match image 42b to the authenticity determination device 16-2.

[0105] In step S46, the inter-match image similarity calculation unit 48 of the authenticity determination device 16-2 calculates the similarity between the first match image 42a and the second match image 42b received in step S44, that is, the inter-match image similarity.

[0106] In step S48, the paired registered image similarity calculation unit 68 of the authenticity determination device 16-2 calculates the similarity between the first match image 42a received in step S44 and the registered image 64 stored in the memory 62, i.e., the paired registered image similarity.

[0107] Either step S46 or step S48 may be performed first.

[0108] In step S50, the match image determination unit 50 of the authenticity determination device 16-2 determines whether the match image similarity calculated in step S46 is less than a first threshold value. Also, the registered image determination unit 70 of the authenticity determination device 16-2 determines whether the paired registered image similarity calculated in step S48 is equal to or greater than a second threshold value.

[0109] If the inter-match image similarity is less than the first threshold and the paired registered image similarity is equal to or greater than the second threshold (YES in step S50), the process proceeds to step S52. In step S52, the registered image determination unit 70 determines that the target item Or is a true item Ot.

[0110] On the other hand, if the inter-match image similarity is equal to or greater than the first threshold value, or if the paired registered image similarity is less than the second threshold value ("No" in step S50), the process proceeds to step S54. In step S54, the registered image determination unit 70 determines that the target item Or is not the true item Ot.

[0111] In step S56, the notification processing unit 72 of the authenticity determination device 16-2 transmits information indicating the determination result in step S52 or step S54 to the match image acquisition device 14-2.

[0112] In step S58, the notification processing unit 54 of the match image acquisition device 14-2 notifies the user of the determination result based on the information received in step S56.

[0113] Third Embodiment FIG. 13 is a schematic diagram of the configuration of a match image acquisition device 14-3 according to the third embodiment (same as in the second embodiment). FIG. 10 is a schematic diagram of the configuration of an authenticity determination device 16-3 according to the third embodiment (same as in the first embodiment). The schematic diagram of the configuration of an authenticity determination system 10 according to the third embodiment is similar to that of the first embodiment shown in FIG. In the description of the third embodiment, descriptions of the same parts as in the first embodiment will be omitted.

[0114] In the first and second embodiments, whether the target item Or is the true item Ot is determined based on whether the inter-matching image similarity between the first match image 42a and the second match image 42b is less than a first threshold (determination 1 in Figure 11) and whether the pair-registered image similarity between the registered image 64 and the first match image 42a is greater than or equal to a second threshold (determination 2 in Figure 11).In the third embodiment, whether the target item Or is the true item Ot is determined based on the first pair-registered image similarity between the registered image 64 and the first match image 42a and the second pair-registered image similarity between the registered image 64 and the second match image 42b.

[0115] In the third embodiment, as in the second embodiment, the photographing processing unit 46 of the match image acquisition device 14-3 acquires the first match image 42a and the second match image 42b, and the transmission processing unit 52 transmits the first match image 42a and the second match image 42b to the authenticity determination device 16-3.

[0116] The paired-registered image similarity calculation unit 68 calculates a first paired-registered image similarity, which is the similarity between the first match image 42a received from the match image acquisition device 14-3 and the registered image 64. The paired-registered image similarity calculation unit 68 also calculates a second paired-registered image similarity, which is the similarity between the second match image 42b received from the match image acquisition device 14-3 and the registered image 64.

[0117] The registered image determination unit 70 determines whether the first paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is equal to or greater than a second threshold. The registered image determination unit 70 also determines whether the second paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is equal to or greater than a third threshold. The third threshold may be determined in advance by an administrator of the authenticity determination system 10, for example. The third threshold may be the same as or different from the first threshold or the second threshold.

[0118] Here, if the target item Or is the true item Ot, the first matching image 42a was acquired under the same shooting conditions as the registered image 64 (i.e., the first shooting conditions), so the first pair-registered image similarity should be high, and the first matching image 42a was acquired under shooting conditions different from the registered image 64 (i.e., the second shooting conditions), so the second pair-registered image similarity should be low.

[0119] Therefore, the registered image determination unit 70 determines that the target item Or is the true item Ot when the first paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is equal to or greater than the second threshold and the second paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is less than the third threshold. On the other hand, the registered image determination unit 70 determines that the target item Or is not the true item Ot when the first paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is less than the second threshold or when the second paired registered image similarity calculated by the paired registered image similarity calculation unit 68 is equal to or greater than the third threshold.

[0120] FIG. 16 is a conceptual diagram showing the contents of the authenticity determination process in the authenticity determination system 10 according to the third embodiment. In the third embodiment, the authenticity determination process is executed by the registered image determination unit 70. In the third embodiment, the registered image determination unit 70 determines, as Determination 2, whether the similarity between the first match image 42a and the registered image 64, i.e., whether the first pair-registered image similarity is equal to or greater than a second threshold. Furthermore, as Determination 3, the registered image determination unit 70 determines the similarity between the second match image 42b and the registered image 64, i.e., whether the second pair-registered image similarity is less than a third threshold. Note that either Determination 2 or Determination 3 may be executed first.

[0121] If, in judgment 1, it is determined that the first match image 42a and the registered image 64 are similar, i.e., the first pair-registered image similarity is equal to or greater than the second threshold, and it is determined that the second match image 42b and the registered image 64 are dissimilar, i.e., the second pair-registered image similarity is less than the third threshold, the registered image determination unit 70 determines that the target item Or is the true item Ot. On the other hand, if, in judgment 1, it is determined that the first match image 42a and the registered image 64 are dissimilar, i.e., the first pair-registered image similarity is less than the second threshold, the registered image determination unit 70 determines that the target item Or is not the true item Ot.

[0122] Fig. 17 is a flowchart showing the processing flow of the authenticity determination system 10 according to the second embodiment. At the start of the flowchart shown in Fig. 17, it is assumed that the registered image 64 has already been acquired by the registered image acquisition device 12 and stored in the memory 62 of the authenticity determination device 16-3.

[0123] In step S60, the photographing processing unit 46 of the match image acquisition device 14-3 photographs the characteristic region FAr of the target object Or under the first photographing condition, thereby acquiring the first match image 42a.

[0124] In step S62, the photographing processing unit 46 photographs the characteristic region FAr of the target object Or under the second photographing conditions, thereby acquiring the second match image 42b.

[0125] In step S64, the transmission processing unit 52 of the match image acquisition device 14-3 transmits the first match image 42a and the second match image 42b to the authenticity determination device 16-3.

[0126] In step S66, the paired registered image similarity calculation unit 68 of the authenticity determination device 16-3 calculates the similarity between the first match image 42a received in step S44 and the registered image 64 stored in the memory 62, i.e., the first paired registered image similarity. The paired registered image similarity calculation unit 68 also calculates the similarity between the second match image 42b received in step S44 and the registered image 64, i.e., the second paired registered image similarity.

[0127] In step S68, the registered image determination unit 70 of the authenticity determination device 16-3 determines whether the first pair of registered images similarity calculated in step S66 is greater than or equal to a second threshold, and whether the second pair of registered images similarity calculated in step S66 is less than a third threshold.

[0128] If the first pair-registration image similarity is equal to or greater than the second threshold and the second pair-registration image similarity is less than the third threshold (YES in step S68), the process proceeds to step S70. In step S70, the registered image determination unit 70 determines that the target item Or is a true item Ot.

[0129] On the other hand, if the first pair-registered image similarity is less than the second threshold value or the second pair-registered image similarity is equal to or greater than the third threshold value ("No" in step S68), the process proceeds to step S72. In step S72, the registered image determination unit 70 determines that the target item Or is not the true item Ot.

[0130] In step S74, the notification processing unit 72 of the authenticity determination device 16-3 transmits information indicating the determination result in step S70 or step S72 to the match image acquisition device 14-3.

[0131] In step S76, the notification processing unit 54 of the match image acquisition device 14-3 notifies the user of the determination result based on the information received in step S74.

[0132] Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various modifications are possible without departing from the spirit of the present invention.

[0133] In the above embodiment, the processor refers to a processing device in a broad sense, and includes at least one of a general-purpose processing device (e.g., a CPU) and a dedicated processing device (e.g., a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a programmable logic device). The processor may not be a single processing device, but may be configured by the cooperation of multiple processing devices located at physically separate locations.

[0134] (Addendum) (((1))) An authenticity determination system for determining the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item where an item-specific feature is expressed, comprising: a processor; The processor: calculating a similarity between match images between a first match image obtained by photographing the characteristic region of the target article under first photographing conditions that are equivalent to photographing conditions when the registered image was acquired, and a second match image obtained by photographing the characteristic region of the target article under second photographing conditions that are different from the first photographing conditions; calculating a pair-registration image similarity between the first match image and the registration image; If the inter-match image similarity is less than a first threshold and the paired registered image similarity is equal to or greater than a second threshold, the target item is determined to be a genuine item. An authenticity determination system characterized by: (((2))) The processor: If the inter-match image similarity is equal to or greater than the first threshold, the target item is determined to be not a genuine item without calculating the paired registered image similarity. The authenticity determination system according to (((1))) is characterized in that: (((3))) the calculation of the inter-match image similarity is performed by a match image acquisition device that photographs the characteristic region of the target article and acquires the first match image and the second match image, the calculation of the paired registered image similarity is executed by an authenticity determination device that is communicably connected to the match image acquisition device and is different from the match image acquisition device; When the inter-match image similarity is equal to or greater than the first threshold, the match image acquisition device does not transmit the first match image to the authenticity determination device. The authenticity determination system according to (((2))) is characterized in that: (((4))) When the inter-match image similarity is equal to or greater than a first threshold, the processor presents a notification to a user urging the user to visually inspect the characteristic region of the target article. The authenticity determination system according to (((1))) is characterized in that: (((5))) the processor presents the registration image to the user along with the notification. The authenticity determination system according to (((4))) is characterized in that: (((6))) The processor: displaying a guide on a photographing screen of a match image acquisition device that photographs the characteristic region of the target article, the guide indicating at least one of a position or an attitude of the match image acquisition device when photographing the characteristic region of the target article under the first photographing condition, and the guide indicating at least one of a position or an attitude of the match image acquisition device when photographing the characteristic region of the target article under the second photographing condition; The authenticity determination system according to (((1))) is characterized in that: (((7))) An authenticity determination system for determining the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item where an item-specific feature is expressed, comprising: a processor; The processor: calculating a first matching image similarity between the registered image and a first matching image obtained by photographing the characteristic region of the target object under first photographing conditions that are the same as those under which the registered image was acquired; calculating a second matching image similarity between the registered image and a second matching image obtained by photographing the characteristic region of the target article under a second photographing condition different from the first photographing condition; If the first pair-registration image similarity is equal to or greater than a second threshold and the second pair-registration image similarity is less than a third threshold, the target item is determined to be a genuine item. An authenticity determination system characterized by: (((8))) A computer that determines the authenticity of a target item using a registered image obtained by photographing a feature area that is an area in which the item's unique features appear in the real item, calculating a similarity between match images between a first match image obtained by photographing the characteristic region of the target article under first photographing conditions that are equivalent to photographing conditions when the registered image was acquired, and a second match image obtained by photographing the characteristic region of the target article under second photographing conditions that are different from the first photographing conditions; calculating a pair-registration image similarity between the first match image and the registration image; determining that the target item is a genuine item when the inter-match image similarity is less than a first threshold value and the paired registered image similarity is equal to or greater than a second threshold value; The authenticity determination program is characterized by: (((9))) A computer that determines the authenticity of a target item using a registered image obtained by photographing a feature area that is an area in which the item's unique features appear in the real item, calculating a first matching image similarity between the registered image and a first matching image obtained by photographing the characteristic region of the target object under first photographing conditions that are the same as those under which the registered image was acquired; calculating a second matching image similarity between the registered image and a second matching image obtained by photographing the characteristic region of the target article under a second photographing condition different from the first photographing condition; When the first pair-registered image similarity is equal to or greater than a second threshold and the second pair-registered image similarity is less than a third threshold, the target item is determined to be a genuine item. The authenticity determination program is characterized by:

[0135] According to the inventions of (((1))), (((7))), (((8))), or (((9))), it is possible to ensure that the authenticity of a target article having a counterfeit characteristic area in which the characteristic area of ​​a genuine article is reproduced in shades of gray can be correctly determined by comparing a registered image with a matching image, while reducing the effort required for authenticity determination compared to when multiple registered images are prepared in advance. According to the invention related to (((2))), the amount of calculation required in the authenticity determination process can be reduced compared to when the similarity between registered images is calculated even when the similarity between the matched images is equal to or greater than the first threshold value. According to the invention related to (((3))), the amount of communication data between the match image acquisition device and the authenticity determination device in the authenticity determination process can be reduced compared to when the first match image is sent to the authenticity determination device even when the similarity between the match images is equal to or greater than the first threshold value. According to the invention (((4))), the user can be prompted to visually determine whether the target item is a genuine item. According to the invention (((5))), the user can easily compare the characteristic area of ​​the target object with the registered image, compared to when the registered image is not presented to the user. According to the invention pertaining to (((6))), it is possible to assist in photographing the characteristic area of ​​the target article under the first photographing condition and the second photographing condition. [Explanation of symbols]

[0136] 10 Authentication determination system, 12 Registration image acquisition device, 14, 14-2, 14-3 Matching image acquisition device, 16, 16-2, 16-3 Authentication determination device, 20, 30 Light source, 22, 32 Camera, 34, 60 Communication interface, 36 Input interface, 38 Display, 40, 62 Memory, 42a First match image, 42b Second match image, 44, 66 Processor, 46 Photography processing unit, 48 Matching image similarity calculation unit, 50 Matching image determination unit, 52 Transmission processing unit, 54, 72 Notification processing unit, 64 Registration image, 68 Pair of registered image similarity calculation unit, 70 Registration image determination unit.

Claims

1. An authenticity determination system for determining the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item where an item-specific feature is expressed, comprising: a processor; The processor: calculating a similarity between match images between a first match image obtained by photographing the characteristic region of the target article under first photographing conditions that are equivalent to photographing conditions when the registered image was acquired, and a second match image obtained by photographing the characteristic region of the target article under second photographing conditions that are different from the first photographing conditions; calculating a pair-registration image similarity between the first match image and the registration image; determining that the target item is a genuine item when the inter-match image similarity is less than a first threshold value and the paired registered image similarity is equal to or greater than a second threshold value; An authenticity determination system characterized by:

2. The processor: If the inter-match image similarity is equal to or greater than the first threshold, the target item is determined to be not a genuine item without calculating the paired registered image similarity.

2. The authenticity determination system according to claim 1.

3. the calculation of the inter-match image similarity is performed by a match image acquisition device that captures the characteristic region of the target article and acquires the first match image and the second match image, the calculation of the paired registered image similarity is executed by an authenticity determination device that is communicably connected to the match image acquisition device and is different from the match image acquisition device; When the inter-match image similarity is equal to or greater than the first threshold, the match image acquisition device does not transmit the first match image to the authenticity determination device.

3. The authenticity determination system according to claim 2.

4. When the inter-match image similarity is equal to or greater than a first threshold, the processor presents a notification to a user urging the user to visually inspect the characteristic region of the target article.

2. The authenticity determination system according to claim 1.

5. the processor presents the registration image to the user along with the notification.

5. The authenticity determination system according to claim 4.

6. The processor: displaying a guide on a photographing screen of a match image acquisition device that photographs the characteristic region of the target article, the guide indicating at least one of a position or an attitude of the match image acquisition device when photographing the characteristic region of the target article under the first photographing condition, and the guide indicating at least one of a position or an attitude of the match image acquisition device when photographing the characteristic region of the target article under the second photographing condition; 2. The authenticity determination system according to claim 1.

7. An authenticity determination system for determining the authenticity of a target item using a registered image obtained by photographing a feature area, which is an area in a genuine item where an item-specific feature is expressed, comprising: a processor; The processor: calculating a first matching image similarity between the registered image and a first matching image obtained by photographing the characteristic region of the target object under first photographing conditions that are the same as those under which the registered image was acquired; calculating a second matching image similarity between the registered image and a second matching image obtained by photographing the characteristic region of the target article under a second photographing condition different from the first photographing condition; determining that the target item is a genuine item when the first pair-registration image similarity is equal to or greater than a second threshold and the second pair-registration image similarity is less than a third threshold; An authenticity determination system characterized by:

8. A computer that determines the authenticity of a target item using a registered image obtained by photographing a feature area that is an area in which the item's unique features appear in the real item, calculating a similarity between match images between a first match image obtained by photographing the characteristic region of the target article under first photographing conditions that are equivalent to photographing conditions when the registered image was acquired, and a second match image obtained by photographing the characteristic region of the target article under second photographing conditions that are different from the first photographing conditions; calculating a pair-registration image similarity between the first match image and the registration image; determining that the target item is a genuine item when the inter-match image similarity is less than a first threshold value and the paired registered image similarity is equal to or greater than a second threshold value; The authenticity determination program is characterized by:

9. A computer that determines the authenticity of a target item using a registered image obtained by photographing a feature area that is an area in which the item's unique features appear in the real item, calculating a first matching image similarity between the registered image and a first matching image obtained by photographing the characteristic region of the target article under first photographing conditions that are the same as those under which the registered image was acquired; calculating a second matching image similarity between the registered image and a second matching image obtained by photographing the characteristic region of the target article under a second photographing condition different from the first photographing condition; determining that the target item is a genuine item when the first pair-registration image similarity is equal to or greater than a second threshold value and the second pair-registration image similarity is less than a third threshold value; The authenticity determination program is characterized by:

Images