Individual certification for products using terahertz waves

Abstract

A method for authenticating an article is provided, the method including, in a reference signature generation stage, acquiring a reference density image of a volume beneath a reference region of the article using a near-field terahertz imager array, and generating and storing a reference signature associated with the article from the reference density image.

Description

[[Technical Field]]

[0001] The present invention relates to a system for authenticating products such as luxury goods, or more generally products that need to be individually tracked. [[Background Art]]

[0002] In the industry of luxury goods and products complying with strict safety or health standards, there is generally a demand to individually track products. To achieve this, each product is provided with a unique identifier such as a serial number, which is recorded in a manufacturer's database together with the characteristics of the product or even a photograph of the product. The unique identifier is usually marked on the product by means of a label or an engraving or the like.

[0003] To authenticate a product, by providing the unique identifier, the owner can compare the characteristics of the product in the manufacturer's database with the characteristics of the product to be verified. In the case of a luxury item, a user can compare the item with a photograph of the item in the database.

[0004] To more reliably authenticate a specific item, it is known to use a unique signature based on the physical characteristics of the item. During manufacturing, a reference signature is acquired and stored in a database. In a subsequent authentication step, the signature is measured under manufacturing conditions and compared with the reference signature.

[0005] Patent GB2097979 describes an authentication system that uses a signature based on the surface condition of a product within a predefined reference zone. This signature consists, for example, of a frequency spectrum derived from a surface profile.

[0006] Since surface finishing is a microscopic characteristic, the authenticity verification procedure requires advanced measuring equipment and very accurate positioning of the measurement system relative to a predefined area, and this procedure is not practical outside of a laboratory environment.

[0007] Patent US8497983 applies the principle of the GB2097979 patent mentioned above by analyzing the reference zone of an article using coherent light. This analysis system is aligned with the reference zone using a mechanical camera tracking system based on a reference image of the article including the reference zone. [Overview of the project] [Problems that the invention aims to solve]

[0008] Such verification systems remain complex and too costly for widespread market adoption. Furthermore, this technology can only certify specific categories of articles whose surface finish is inherently irregular, resistant to deformation, or resistant to wear. These are items that are inherently metallic, or at least made from hard materials, whose manufacturing process (casting, machining, forging) involves the surface finish. [Means for solving the problem]

[0009] A general method for authenticating an article is provided, comprising the steps of: obtaining a reference density image of the volume beneath the reference region of the article using a nearby terahertz imager array during the reference signature generation stage; and generating and storing a reference signature associated with the article from the reference density image.

[0010] This method may include, in a subsequent authentication stage, the steps of: obtaining a current density image of the volume beneath the reference area of ​​the article using a nearby terahertz imager array; searching for a reference signature associated with the article; and searching for the reference signature within the current density image.

[0011] The imager may have a pixel array larger than the reference zone.

[0012] A reference signature can be a subimage containing a set of pixels aligned within the reference region.

[0013] A reference signature can be a sub-image containing a reference region.

[0014] This method may include the step of segmenting the imager's pixel values ​​into several levels significantly smaller than the pixel resolution.

[0015] This method may include the step of segmenting the pixel values ​​into three levels.

[0016] The proximity terahertz imager can be integrated using CMOS technology.

[0017] A near-field terahertz imager could be a lensless reflective imager.

[0018] A terahertz proximity imager can be a near-field imager.

[0019] The following non-limiting description is provided in connection with the attached drawings. [Brief explanation of the drawing]

[0020] [Figure 1] This figure shows an authenticationable article according to one embodiment, having a visually marked reference zone for applying a proximity terahertz imager. [Figure 2] This is a magnified view of the image taken in the reference zone of the item shown in Figure 1. [Figure 3] This figure shows the curve of pixel values ​​along the axes of the acquired image in Figure 2. [Modes for carrying out the invention]

[0021] The main obstacle to the widespread adoption of surface-condition-based item authentication systems is the difficulty in providing low-cost, user-friendly, on-site measurement devices.

[0022] Another disadvantage is that since the signature is based on microscopic features that require high-accuracy and high-resolution images, the techniques used to create and compare the signature are complex and have low reliability.

[0023] To enhance the appeal of authentication systems based on the physical characteristics of items, the present disclosure identifies macroscopic features that can constitute a unique signature across a wide range of item categories including non-metallic flexible items. It has been found that terahertz waves reveal a macroscopic internal structure specific to the material or manufacturing technique being analyzed, and exhibit random properties suitable for generating a unique product signature.

[0024] In another field, terahertz waves are used to verify that products comply with material composition standards. For example, patent application US2010 / 0232715 discloses a system for verifying the content of a specific material having a characteristic frequency in the terahertz frequency band. A product to be analyzed is irradiated with terahertz waves, and the system measures the spectrum of frequencies radiated by the product using a single terahertz detector. The measured spectrum is then compared with a reference spectrum stored in a database, and a conformance score for the product to be analyzed is calculated. This approach only verifies that the product was manufactured in compliance with composition standards, and does not authenticate individual products.

[0025] According to a feature of the present authentication system, it is proposed to use a terahertz imager to map the density of a volume of material under a reference zone of an article to extract a unique signature. When using a lensless close-proximity terahertz imager, the mapping becomes macroscopic because the pitch of the pixel array of such an imager remains relatively large when considering the wavelength involved. For example, at a frequency of 600 GHz, the pixel pitch of the imager is on the order of 200 μm. In comparison, a single pixel is four times larger than the entire reference region in the case of surface state analysis according to patent GB2097979.

[0026] Preferably, the proximity imager is a lensless reflective imager, as described in Hani Sherry's patent application WO2019155156, or a near-field imager, as described in Hani Sherry's patent application WO2019186074. All of the imagers described in these patent applications are manufactured using CMOS technology, making them compact, portable in stylus form, and inexpensive. A 30x30 pixel imager with a 200 μm pitch can, for example, be 6x6 mm in size and map the density of a macroscopic reference area of ​​the same dimensions.

[0027] In practice, to account for the uncertainty in positioning the imager relative to the reference zone visually marked on the object, the reference zone is smaller than the imager array; for example, in the case of a 6x6mm imager, it is a 4x4mm square or a 4mm diameter disk.

[0028] Figure 1 illustrates the use of a proximity terahertz imager stylus 10 in the authentication or reference signature creation stage of an item, such as a leather bag 12. In practice, the imaging stylus is designed to be in contact with the area to be scanned, and the stylus end is dimensioned to ensure the necessary distance between the imager and the surface to be scanned.

[0029] The article is identified here by a signature reference zone, exemplified by a dotted circle 14 on the visible surface of the article. (Of course, for aesthetic reasons, the reference zone and its markings are actually placed in an inconspicuous area, but still in an area accessible by the imaging stylus.)

[0030] The reference zone marking 14 may also be traced to assist in the precise positioning of the stylus end. Relatively precise positioning is preferred when the reference signature of the article is recorded at the end of manufacturing. In the subsequent authentication stage, positioning is less critical because the imager is larger than the reference zone, and any deviation can be compensated for by image processing.

[0031] Therefore, the diameter of the marking may be substantially equal to the diameter of the stylus end, for example, 1 cm in the case of a 6 × 6 mm imager. The effective reference area is preferably smaller, for example, a 4 × 4 mm (or less) square or a disk with a diameter of 4 mm (or less). Image processing can compensate for rotation, but relatively accurate rotational positioning can be achieved by aligning the stylus end with some markings in the reference zone. The end of the stylus can also be a square that coincides with the square markings on the article.

[0032] The right side of Figure 1 shows an example of an image produced by the stylus in the reference zone of the leather bag 12 when the imager's pixels are set to detector mode. The pixels of the terahertz imager may have two operating modes: a detector mode that simply provides the amplitude or power of the wave the pixel receives, and a phase measurement mode that also provides the phase of the wave the pixel receives. The simplest configuration is the detector mode configuration, which is more than sufficient for the authentication system disclosed herein.

[0033] A proximity terahertz imager in detector mode performs absorption or density measurements across a specific thickness of material beneath the observation surface. The resulting image converts density levels to grayscale, revealing macroscopic textures specific to the observed material and, where applicable, to the manufacturing process used. These textures are random and well-suited for establishing unique signatures.

[0034] Figure 2 is a magnified view of the image shown in Figure 1, rendered with a 30x30 pixel imager, i.e., at a fairly low resolution. In fact, the image in Figure 1 is represented at high resolution to highlight the texture, but this texture is sampled at a low resolution by the actual imager. This is not a disadvantage, as even a low-resolution image is sufficient to establish a unique signature that is not overly complex. For typical applications, each pixel provides 8 to 10 bits of grayscale.

[0035] In the image shown in Figure 2, a smaller reference region is depicted, for example, as a circle in the center. There are many possibilities for establishing a signature from the reference zone.

[0036] In one embodiment, a series of gray levels located on the diameter of a circle, for example, in the central column or row of pixels, or along the diagonal of an array, may be stored as a reference signature.

[0037] Figure 3 shows an example of a grayscale sequence obtained in the form of a curve. This sequence is measured at the end of the item's manufacturing process and recorded in a database along with the item's serial number or other unique identifier. The measurement may be performed manually using the same type of stylus used for subsequent authentication. For a 30x30 pixel imager, the sequence or curve would contain approximately 20 samples, coded in 8-10 bits at the imager's precision.

[0038] During the authentication phase, the user takes an image of the item's reference region using a stylus with the same features used to record the reference signature. This image is sent to a central authentication system, for example, along with the item's serial number. Using the serial number, the central system attempts to retrieve the reference signature in the form of a sub-image and find that signature within the provided image. Known techniques for this type of operation, such as motion estimation used in MPEG video coding, provide a correspondence score with a motion vector or transformation matrix. For the purposes of this system, it is sufficient to simply compare the score to a threshold to determine if there is a match, and therefore the item being verified is genuine.

[0039] These sub-image search techniques tolerate a fairly wide range of offset and rotation errors, so the user performing the authentication does not need to exercise extreme care when aligning the stylus with the reference area. On the other hand, it is preferable that the reference zone is included within the acquired image.

[0040] In another embodiment, a subimage corresponding to a reference region may be stored as a reference signature. This subimage can be used in the same manner as the grayscale sequence described above, but requires more storage and processing resources. Therefore, this embodiment may be more suitable for materials where texture changes are not very noticeable, i.e., where a single pixel line is not sufficient to establish a unique signature.

[0041] In particular, when using the entire image of a reference region as a signature, the pixel precision of the imager may be excessive for the purpose of guaranteeing signature uniqueness. In this case, the measurement can be segmented to fewer levels than the pixel's bit resolution, for example, 8 levels instead of 256 levels in the case of 8 bits, or even as few as 3 levels (white, gray, black). Applying such segmentation during the manufacturing and authentication stages significantly reduces the resources required for signature storage and recognition.

[0042] Although a leather bag was used as an example of an item to be authenticated, this system can be applied to any material that has a specific degree of transparency to terahertz radiation, and even to materials with a uniform appearance, such as plastics. Plastics have internal textures and specific structural particles related to manufacturing processes such as injection molding, which become apparent under terahertz radiation and have random properties.

[0043] This system is particularly well-suited for items made from composite materials. In this case, terahertz radiation reveals fibers embedded in the resin with noticeable contrast, requiring very little gray level to ensure a unique signature.

Claims

1. A method for authenticating goods, At the end of the manufacturing stage of the aforementioned article, The steps include defining a reference zone on the article having a random internal texture resulting from the manufacturing process of the material of the article, The steps include: marking the reference zone visibly on the article; The steps include: a terahertz proximity imager acquiring a density image of the reference zone of the article; The process involves a processor generating a reference signature from the density image, The processor records the reference signature in a database along with a unique identifier associated with the article, A method that includes this.

2. In the subsequent certification stage of the aforementioned articles, The steps include: finding the reference zone on the article using the visible markings; The steps include: acquiring a current density image of a reference zone of the article using a terahertz proximity imager; The steps include: searching the database for the reference signature associated with the article; The steps include searching for the reference signature within the current density image, The method according to claim 1, including the method described in claim 1.

3. The method according to claim 1, wherein the terahertz proximity imager comprises a pixel array larger than the reference zone.

4. The method according to claim 3, wherein the reference signature is a subimage comprising a series of pixels aligned within the reference zone.

5. The method according to claim 3, wherein the reference signature is a subimage including the reference zone.

6. The method according to claim 4 or 5, wherein the processor includes the step of segmenting the pixel values ​​of the terahertz proximity imager into several levels significantly smaller than the pixel resolution.

7. The method according to claim 6, comprising the step of segmenting the pixel value into three levels.

8. The method according to claim 1, wherein the terahertz proximity imager is integrally mounted using CMOS technology.

9. The method according to claim 8, wherein the terahertz proximity imager is a lensless reflective imager.

10. The method according to claim 8, wherein the terahertz proximity imager is a near-field imager.

Images