METHOD AND SYSTEM FOR VERIFYING AN IDENTITY OF AN ITEM, AND, COMPUTER-READABLE STORAGE MEDIA

专利摘要:
METHOD AND SYSTEM FOR VERIFYING AN ITEM IDENTITY AND COMPUTER-READABLE STORAGE MEDIA. Natural variations in tagged features on an item are used as a way to establish a specific identifier for that item, store the identifier separately from the item, and subsequently access the identifier to validate the identity of the item.
公开号:BR112014021578A2
申请号:R112014021578-2
申请日:2013-03-01
公开日:2021-05-25
发明作者:Michael L. Soborski
申请人:Sys-Tech Solutions, Inc.；
IPC主号:

专利说明:

[0001] [0001] This application relates to US Provisional Patent Applications 61/605,369, filed March 1, 2012, and 61/676,113 filed July 26, 2012, for a "Method for extracting unique identi/cation information fom marked features", and 61/717.711, filed October 24, 2012, for "Unique identi/ication information ji'om marked features", all by Soborski. All such orders are hereby incorporated by reference in their entirety. STATEMENT RELATED TO RESEARCH OR DEVELOPMENT WITH FEDERAL SPONSORSHIP
[0002] [0002] Not Applicable REFERENCE TO SEQUENCE LISTING, A TABLE OR A COMPUTER PROGRAM THAT LISTS DISK COMPACT ATTACHED
[0003] [0003] Not Applicable FUNDAMENTALS OF THE INVENTION
[0004] [0004] Aspects of the present description are in the technical field of Machine Vision.
[0005] [0005] Aspects of this description are also in the technical field of Anti-counterfeiting and Item Serialization for the purposes of security of Tracking and Monitoring of the supply chain.
[0006] [0006] There is prior technology in this field for the purpose of uniquely identifying items. Some methods are based on deliberately visible or disguised marks applied to an item, usually through printing. Other methods rely on natural variations in a material's substrate (fiber orientation in paper, for example) to be used as a unique identifier. There are significant deficiencies in the prior technology. These include the need to deliberately add visible or disguised tags to the item, in addition to any tags already present on the item for other purposes. In the case of the substrate variation method, an expert system that senses the variations is needed; also, for substrates that do not have a readily identifiable unique feature (some plastic films, for example) this method cannot be employed. These deficiencies seriously reduce the usefulness of these methods in the technical fields considered here. SUMMARY OF THE INVENTION
[0007] [0007] This description prescribes the use of natural variations in features marked on an item as a way to establish specific information or data for that item, which may be referred to as a "signature" or an "original item identifier", storage of the information separately from the item and subsequent access to stored information to validate the identity of an item claimed to be the original item. Deliberate application of disguised or visible identification marks on the item is not required, although it may be used in some modalities. Rather, the natural variations inherent in many manufacturing, marking, or printing processes can be exploited to extract identifying features from an item or a brand, such as one of many types of marks applied to items. Additionally, this approach easily integrates into existing reader systems for applied marks, such as barcode readers or machine vision systems; no specialized system is needed to perceive variations in a material substrate sufficient to identify an item.
[0008] [0008] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a tangible non-temporary storage media) for verifying a
[0009] [0009] Artifacts can be features of the item that were produced when the item was produced. At least some of the artifacts may be non-controllably producible in item production. The characteristic by which original artifacts are ranked can be a magnitude, for example, the size of an artifact. Information from the original ranked artifacts can be encoded into computer readable data corresponding to the original item to form a signature.
[00010] [00010] Embodiments of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) to verify an identity of an item, comprising: examining a mark comprising an identifier and an artifact , where the identifier is associated with an original item and the anephate does not change the association; extract information associated with the artifact; and storing the information on a non-temporary computer-readable storage device separate from the original item to be at least partially locatable using the identifier.
[00011] [00011] Respective information coming from a plurality of said marks can be stored in a storage device, for example, in the form of a database, and, using the identifier coming from one of said marks, the respective information coming from a number of marks smaller than said plurality of marks and comprising said one mark may be recoverable. In one example, the identifier might identify a group or category of items. O
[00012] [00012] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a tangible non-temporary storage media) to verify an identity of an item, comprising: examining an original item in relation to specific original artifacts of the original item; storing information associated with the original artifacts and information associated with at least one of the apparatus involved in creating the original artifacts and the apparatus involved in examining the original item on a non-temporary computer readable storage device separate from the original item.
[00013] [00013] The stored information may include information indicative of a type of apparatus involved in creating the original artifacts. The information stored may include information indicative of a resolution of the device involved in examining the original item.
[00014] [00014] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) for verifying the identity of an item, comprising: examining an unverified item for unverified artifacts. checked item-specific checks not checked; extract information representing unverified artifacts; retrieving, from a storage device, stored data that contains information representing original artifacts of an original item; retrieve the information from the original artifacts from the stored data; to compare
[00015] [00015] The stored data may include information relating to at least one of the apparatus involved in creating the original artifacts and the apparatus involved in examining the original item.
[00016] [00016] The correction may comprise comparing resolutions or other properties of the device involved in the examination of the original item and the device involved in the examination of the unverified item, and discounting artifacts detected by one of these devices that would not be reliably detected by the other of these devices. When the two fixtures have different resolutions, artifacts that are greater than the resolution threshold of one fixture, and are detected by this fixture, but less than the resolution threshold of the other fixture, can be discounted. The weighting may be based on a characteristic that resolves energy and imaging fidelity of the verification device against corresponding characteristics of the original imaging device.
[00017] [00017] When artifacts are from different categories, determining whether the information from unverified artifacts matches the information from the original artifacts can comprise comparing the detected artifacts in each category and combining the results of the comparisons, and the correction can then understand weight the combination according to a known tendency for the apparatus that created the original artifacts to produce artifacts in different categories with different Hequences or different magnitudes.
[00018] [00018] Modalities of the present invention provide methods, apparatus and computer programs (which can be stored on a non-temporary tangible storage media) to verify the identity of an item, comprising: examining an original item that has a symbol in it original comprising an array of differently colored printed cells for original artifacts specific to the original symbol, where: artifacts are features of at least some of the cells that were produced when the original symbol was produced; and at least some of the artifacts were not controllably producible in producing the original symbol; and the artifacts comprise at least one artifact category selected from the group consisting of deviation in the average color of a cell from an average derived from the mark, which may be an average for neighboring cells of the same nominal color, displacement in the position of a cell relative to a better suited grid of neighboring cells, areas of a color other than at least two colors from a nominal cell color, and deviation from a nominal shape of a long continuous edge; extract information representing the original artifacts for each cell; encode information from the original artifacts into computer-readable data corresponding to the original item; and store the data on a non-temporary computer-readable storage device separate from the original item.
[00019] [00019] In general, different "colors" may differ in clarity, hue, or both, and may be distinguished by differences in clarity, hue, or both. For example, when the symbol is printed on ink or other media of a single first color on a substrate of a single second color, . any measurement that distinguishes the first color from the second color can be used. In the most common case, commonly called "black and white" or
[00020] [00020] Embodiments of the present invention provide methods, apparatus and . computer programs (which can be stored on tangible, non-temporary storage media) to verify the identity of an item, which include: examining an original item against the original item's specific artifacts; extract information associated with the original artifacts; mnking the information of the original artifacts according to a characteristic of the artifact; calculate an autocorrelation series of information from the original ranked artifacts; and storing the data related to the autocorrelation series on a non-temporary computer readable storage device separate from the original item.
[00021] [00021] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a tangible, non-temporary storage media) to verify the identity of an item, comprising: examining an original item against specific original artifacts of the original item, where artifacts are features of the item that were produced when the item was produced, and at least some of the artifacts were not controllably producible in producing the item; extract information that represents the original artifacts; encode information from the original artifacts into computer-readable data corresponding to the original item; and storing the data on a non-temporary computer-readable storage device separate from the original item.
[00022] [00022] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) to verify the identity of an item, comprising: defining a plurality of modules in an original item and an order of modules; examining the modules in the original item against a plurality of original item-specific artifact categories, where artifacts are features of the item that were produced when the item was produced, and at least some of the artifacts were not controllably producible in production of item; extract information that represents the original artifacts; code, for each module in order, which artifact categories are present and which artifact categories are absent to form computer-readable data; and storing the data on a non-temporary computer-readable storage device separate from the original item.
[00023] [00023] Embodiments of the present invention provide methods, apparatus and computer programs (which may be stored on a tangible, non-temporary storage media) for verifying the identity of an item, comprising: examining an unverified item against one or more unverified artifacts specific to the unverified item; extract information representing unverified artifacts; retrieve stored data related to one or more artifacts of an original item from a storage device; retrieve the information of the original artifacts from the retrieved stored data; comparing information from unverified and original artifacts to determine if information from unverified artifacts matches information from original artifacts; and, in case the unverified artifacts information matches the original artifacts information, verify the unverified item as a verified original item. The processing of the unverified item, or the processing of the original item, is in accordance with any of the aspects and modalities of the present invention. It may be preferable to process both the unverified item and the original item by reasonably similar processes, to reduce the level of error and uncertainty introduced by differences between the processes used.
[00024] [00024] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) for verifying the identity of an item, comprising: examining an unverified item for non-verified artifacts. checked item-specific checks not checked; extract information representing unverified artifacts; rank the information of unverified artifacts according to an artifact characteristic; calculate an autocorrelation series of ranked unverified artifact information; retrieve an autocorrelation series representing artifacts of an original item from a storage device; compare the unverified and original autocorrelation series to determine if the information from the unverified artifacts matches the information from the original artifacts; and, in case the unverified artifacts information matches the original artifacts information, verify the unverified item as a verified original item.
[00025] [00025] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) to verify the identity of an item, comprising: examining an unverified item for unverified artifacts. checked item-specific checks not checked; extract information representing unverified artifacts; retrieving data comprising information from the original artifacts from a storage device; retrieve information from the original artifacts from the retrieved data; compare information from unverified and original artifacts; and, in case the unverified artifacts information matches the original artifacts information, verify the unverified item as a verified original item.
[00026] [00026] Modalities of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) to verify the identity · of an item, by successively carrying out any of the above processes to generate and classifying data or information and any appropriate exposed process for comparing an unverified item with stored data or information.
[00027] [00027] Embodiments of the present invention provide methods, apparatus and computer programs (which may be stored on a non-temporary tangible storage media) for verifying the identity of an item by combining features of any two or more of the methods, apparatus and programs of computer exposed.
[00028] [00028] Ranking of original artifact information may include treating artifacts with a characteristic below a threshold value differently than artifacts above the threshold. For example, artifacts smaller than the threshold might not be ranked, or might be grouped together with locations where no artifacts are detected, or might be discounted. The threshold can be chosen with consideration to a threshold for noise of artifacts and the apparatus used to detect them, below which artifacts cannot be reliably detected or cannot be reliably quantified. In one modality, ranking may simply consist of separating above-threshold artifacts from below-threshold or entirely absent artifacts. However, in many modalities, it is preferable that the trait be quantifiable and the ranking comprises ordering artifacts according to a magnitude or quantity of the trait.
[00029] [00029] The method can comprise extracting information that represents a plurality of different artifact categories and the ranking can then comprise ranking the information from the original artifacts separately for each artifact category.
[00030] [00030] The method may comprise defining a plurality of predetermined locations in the original item, and extracting the information representing the artifacts may then comprise associating each artifact with one of the predetermined locations. Where the original item bears a printed symbol comprising a plurality of cells, the predetermined locations can be at least some of the plurality of cells, and the artifacts can then be print artifacts of the cells. When practical, it is usually preferable to use the full symbol in order to maximize the number of available artifacts. However, this is not always necessary. For example, if the symbol has a large number of cells and a high incidence of usable artifacts, a smaller group of cells can be used. In one modality, six artifact categories, with 100 artifacts from each category, ranked by magnitude in each category, were found to provide a robust result. [0003 1] The artifacts may comprise at least one artifact category selected from the group of categories consisting of deviation in the mean color of a cell from the mean for neighboring cells of the same nominal color; displacement in the position of a cell relative to a better suited grid of neighboring cells; areas of a color other than a nominal color of the cell in which they appear; and deviation from a nominal shape of a long continuous edge.
[00032] [00032] The production of the original item may comprise applying a brand to the original item, and the artifacts may then be assets of the brand. "Original item production" can include each stage before the exam begins, and the mark can be applied in a separate step at any time between when the original item production begins and immediately before the exam.
[00033] [00033] When the item or mark is printed, artifacts may comprise imperfections or other variations in the print. When
[00034] [00034] Other information related to the original item can be incorporated into the stored data, in addition to information representing the original artifacts. Other original item information may include a serial number specific to the original item. Such other information can then be retrieved from the retrieved stored data in addition to the information representing the original artifacts.
[00035] [00035] When at least some of the artifacts are artifacts of a symbol encoding data, and the encoded data includes a Unique Identifier (UID) for an individual instance of the symbol or other identifying data, the stored data can be stored to be retrievable under an identifier derivable from the UID or other identifying data. When the other identification data only partially identifies the symbol, for example, it identifies a category or group of items smaller than all the items for which data is
[00036] [00036] The determination may comprise evaluating a statistical probability that the information from the unverified artifacts matches the information from the original artifacts. Then, it can be determined that the information from unverified artifacts matches the information from the original artifacts when the information from the unverified artifacts and the information from the original artifacts are at a predetermined percentage relative to each other.
[00037] [00037] In case the statistical probability exceeds a first threshold, it can be determined that the unverified item is an original verified item. In case the statistical probability is below a second threshold lower than the first threshold, it can be determined that the unverified item is not an original item. In case the statistical probability is between the first and second thresholds, then it can be reported that it cannot be determined whether the unchecked item is an original item.
[00038] [00038] In the assessment of statistical probability, greater weight can be given to artifacts of greater magnitude.
[00039] [00039] Comparison of artifact information may include detecting artifacts that are present in one of the original item and unverified item, and absent in the other of the original item and unverified item. The presence of an artifact in the unverified item that was not present in the original item, absent an indication that the item was damaged in the meantime, can be as significant as the presence of an artifact in the original item that is not present in the non-item checked.
[00040] [00040] In general, "discounting" an artifact includes considering this
[00041] [00041] When at least some of the artifacts are artifacts of a symbol that encodes data and supports error detection, extracting information representing the unverified artifacts can include determining an error state of the symbol containing the unverified artifacts. When the error state indicates that part of the symbol is damaged, the comparison can then comprise discounting artifacts on the damaged part of the symbol.
[00042] [00042] Prior to the storage step, the original item can be partitioned into a plurality of original zones. Each of at least a portion of the original artifacts can then be associated with the original zone in which it is located. Information representing the associated original artifacts and their respective original zones in the stored data can be preserved. The unverified item can then be partitioned into at least one available unverified zone corresponding to less than all original zones. Each of at least a portion of the unverified artifacts can be associated with the available unverified zone in which it is located. Information representing the original artifacts and associated original zones that correspond to the available unverified zones can be retrieved from the retrieved stored data. In the comparison step, only information representing the original artifacts and associated original zones that correspond to the available unverified zones can be used.
[00043] [00043] The original item can be attached to an object to form
[00044] [00044] The magnitude of a deviation in the mean color can be normalized by referring to a difference between mean colors for neighboring cells of at least two nominal colors. The magnitude of displacement in a cell's position relative to a better suited grid of neighboring cells can be normalized by referring to cell sizes. The magnitude of the areas of the color opposite a nominal cell color can be determined by the size of the areas, normalized by reference to the cell size. The magnitude of the deviation from a nominal shape of a long continuous edge can be normalized by referring to a better suited straight line or other smooth curve.
[00045] [00045] When encoding the information of the ranked original artifacts comprises calculating an autocorrelation series of the information of the ranked original artifacts, the encoding may further comprise representing or approximating the autocorrelation series as a polynomial with respect to a fixed order and using the coefficients of the polynomial to form the stored data. The approximation can be relative to a polynomial of a predetermined order, and the coefficients can be approximated to a predetermined precision.
[00046] [00046] When coding the information of the ranked original artifacts comprises calculating an autocorrelation series of the information of the ranked original artifacts, the comparison may comprise calculating an autocorrelation series of the information of the unverified artifacts, and comparing the two autocorrelation series. The comparison may additionally or alternatively comprise comparing the Discrete Fourier Transform (DFT) energy series of the two autocorrelation series, and may then comprise comparing at least one of the Kurtosis and Distribution Shift functions of the DFT energy series .
[00047] [00047] According to embodiments of the invention, an apparatus or system is provided to verify the identity of an item, comprising: an operable original item scanner to examine an original item and extract information representing original artifacts of the original item, by the method of any one or more of the mentioned embodiments and aspects of the invention; an encoder operable to encode the extracted information into a computer-readable item identifier; and a computer-readable storage device operable to store the item identifier.
[00048] [00048] According to embodiments of the invention, an apparatus or a system is provided for verifying the identity of an item by the method of any one or more of the mentioned embodiments and aspects of the invention, comprising: an operable verification scanner for examining an unverified item and extract information representing unverified artifacts from the unverified item; and a processor operable to retrieve a stored item identifier from a storage device, retrieve original artifact information from the retrieved item identifier, compare information from unverified and original artifacts, and produce output dependent on the result of the comparison. .
[00049] [00049] According to embodiments of the invention, an apparatus or a system for verifying the identity of an item is provided, comprising, in combination, the above-described apparatus or system for creating and storing an item identifier and the above-described apparatus or system to examine and compare an unverified item.
[00050] [00050] Verification scanner can be attached to a point of sale device. The verification scanner can be incorporated into a cell phone.
[00051] [00051] The system may further comprise an original item producer operable to produce an original item, where artifacts are item resources that are produced when the original item producer produces the item, and at least some of the artifacts are not controllably producible by the original item producer.
[00052] [00052] The original item producer may be operative to intentionally produce or enhance at least some of the artifacts.
[00053] [00053] The original item producer may comprise an original brand applicator that applies a brand to the original item, with the artifacts then being assets of the brand.
[00054] [00054] The original item producer may comprise a printer, with at least some of the artifacts then comprising variations or imperfections in the print.
[00055] [00055] The system may further comprise at least one original item for which the item identifier is stored on the computer readable storage device.
[00056] [00056] In various modalities, artifacts can be features of the item itself or of a tag that has been applied to the item. The item can be the thing that ultimately needs to be checked, or it can be attached (typically, but not necessarily, in the form of a label) to an object that is to be checked. When the object, item, or brand involves printing, some or all of the artifacts may be variations or imperfections in the print. "Verifying the identity of an item" may include verifying that impression or other marking applied to an item, or an item attached to an object, has not been altered or replaced, even if the underlying item or object is original. For example, you may want to check whether an expiration date, serial number, or other tracking or identification data has been inappropriately altered.
[00057] [00057] In many embodiments, it is preferable that artifacts are resources that do not affect, or at least do not diminish, the function or commercial value of the brand, item, or object in which they appear.
[00058] [00058] Another aspect of the invention provides original items, including original objects comprising objects to which original items have been attached, to which signature data has been stored on the storage device of a system in accordance with another aspect of the invention. BRIEF DESCRIPTION OF THE DRAWINGS
[00059] [00059] The foregoing and still other aspects, features and advantages of the present invention may become more apparent from the following more particular description of the embodiments thereof, presented together with the following drawings. In the drawings:
[00060] [00060] Figure 1 is an illustration of an example of a printed mark that has taken advantage of methods embodying the present invention.
[00061] [00061] Figure 2 is an illustration of the brand in Figure 1 with the brand edge features extracted for clarity.
[00062] [00062] Figure 3 is an illustration of a second example of the same mark as in Figure 1, which may represent a counterfeit version of the mark in Figure 1.
[00063] [00063] Figure 4 is an illustration of the mark in figure 3 with the mark edge features extracted for clarity.
[00064] [00064] Figure 5 is a 2-D Data Matrix that illustrates some features that can be used in the present methods.
[00065] [00065] Figure 6 is an illustration comparing the features of the upper left sections of Figure 2 and Figure 4.
[00066] [00066] Figure 7 is a schematic diagram of a computer system.
[00067] [00067] Figure 8 is a block diagram of a computer operating system to carry out the process of the embodiments of the present invention.
[00068] [00068] Figure 9 is a flowchart of an embodiment of a method of recording a new brand.
[00069] [00069] Figure 10 is a diagram of the weighting of feature features.
[00070] [00070] Figure ll is a flowchart of a modality of a method of evaluating a brand.
[00071] [00071] Figure 12 is a 1-D barcode that illustrates some features that can be used in the present methods.
[00072] [00072] Figure 13 is a graph of a polynomial approximation of an autocorrelation series for a genuine item with a genuine "candidate" symbol.
[00073] [00073] Figure 14 is a graphical representation of an energy series for the genuine data in Figure 13.
[00074] [00074] Figure 15 is a graphical representation similar to Figure 14 for the "candidate" data in Figure 13.
[00075] [00075] Figure 16 is a graph similar to figure 14 for a fake "candidate" symbol.
[00076] [00076] Figure 17 is a graphical representation similar to Figure 14 for the spoofed data used in Figure 16.
[00077] [00077] Figure 18 is a diagram of part of a 2-D Data Matrix that illustrates an encoding process. DETAILED DESCRIPTION OF THE INVENTION
[00078] [00078] A better understanding of various features and advantages of the present methods and devices can be obtained by referring to the following detailed description of illustrative embodiments of the invention and the accompanying drawings. While these drawings represent embodiments of the methods and devices contemplated, they are not to be construed as limiting apparent alternative or equivalent embodiments to those skilled in the art.
[00079] [00079] The currently described modality of a method operates on marks that are applied to items. These tags can be for the purpose of uniquely identifying an item, such as with a serial number, for example, or they can be tags that are for other purposes, such as branding, labeling, or decoration. These marks can be printed, engraved, molded, formed, transferred or otherwise applied to the item using various processes. In order for the method of the present modalities to operate on the marks, the marks must be acquired in such a way that they can be processed electronically. Electronic acquisition methods are varied, and may include, but are not limited to, machine vision cameras, bar code readers, in-line scanning imaging devices, flat bed scanners, handheld handheld imaging devices or many other devices.
[00080] [00080] Now, referring to the drawings, in figure 1, there is shown an example of a printed mark indicated generally by the reference numeral 20 in which the present methods can be applied. In this example, the printed tag is a two-dimensional barcode. This barcode is an information data carrier, where the information is encoded as a pattern of light areas 22 and dark areas 24 on the mark. An ideal example of a 2-D barcode would consist of a rectangular grid, with each cell or "module" 22, 24 in the grid both black and white representing one bit of data.
[00081] [00081] Figure 2 provides an improved view of some of the variations present in the mark shown in figure 1. Figure 2 shows only the edges 26 between light and dark areas of the mark shown in figure 1. Features such as edge linearity, region discontinuities, and feature shape at the mark shown in Figure 1 are readily apparent. Numerous irregularities along the edges of the brand's printed features are clearly visible. Note that this illustration is provided for clarity.
[00082] [00082] Figure 3 shows an example of a second printed mark, indicated generally by the reference number 30, which may represent a forgery of the mark 20 shown in Figure 1, or may represent a second example exclusive of the mark for purposes of identification. This second printed mark 30 is also a two-dimensional bar code. This forged bar code 30, when read with a two-dimensional bar code reader, presents exactly the same decoded information as the mark 20 of figure 1. When the mark 30 of figure 3 is acquired, the present modality again identifies significant features and the capture as "signature" data that uniquely identifies the brand. As in the case of Figure 1, this signature data is derived from the physical and optical characteristics of the geometry and appearance of the mark and, furthermore, may include data that is encoded in the mark, if the mark is a data-carrying symbol, such as a two-dimensional barcode. The tag properties evaluated to create the signature data are usually the same properties used in evaluating the first tag example, so that the two signatures are directly comparable.
[00083] [00083] Figure 4 provides an improved view of some of the variations present in the mark 30 shown in figure 3. Figure 4 shows only the edges 32 of the mark shown in figure 3, similarly to figure 2.
[00084] [00084] Figure 6 shows an immediate comparison of the features in the upper left corner of Figure 2 and Figure 4. As can be seen more clearly in Figure 6, the two printed marks 20, 30 of Figures 1 and 3, even though identical in relation to their visibly encoded data, contain numerous differences on a finer scale, resulting from imperfections in the printing process used to apply the marks. These differences are durable, usually almost as durable as the brand itself, and are practically unique, especially when a large number of differences that can be discovered between the symbols in Figure 1 and Figure 3 are combined. Additionally, the differences are difficult, if not nearly impossible, to fake, as the original symbol needs to be image treated and reprinted at a much higher resolution than the original print, while not introducing new print imperfections. distinguishable. Although only the upper left corner section of the marks is shown here, differentiable features between the two marks shown in figures 1 and 3 run throughout the entirety of the marks and can be used by the present modality.
[00085] [00085] Referring to figure 7, an embodiment of a computing system indicated, in general, by the reference number 50 comprises, among other equipment, a processor or a CPU 52, input and output devices 54, 56, including a image acquisition device 58, random access memory (RAM) 60, read-only memory (ROM) 62, and magnetic disks or other long-term storage 64 for programs and data. The computing system 50 can have a printer 65 for generating marks 20, or the printer 65 can be a separate device. The computing system 50 can be connected through an interface 66 to an external network 68 or other communications media and, through the network 68, to a server 70 with long-term storage 72. Although not
[00086] [00086] Referring to Figure 8, in an embodiment of a computing system, the image acquisition device supplies image data to a signature extraction and encoding processor 74, which may be software running on the primary CPU 52 of the computer system 50, or it can be a dedicated coprocessor. The signature extraction and encoding processor 74 supplies signature data to the network accessible tag signature data storage 76, which may be the long-term storage 72 of the server 70. The tag signature search engine accessible by network 78, which may be software running on primary CPU 52 of computer system 50, or may be a dedicated coprocessor, receives signature data from signature extraction and encoding processor 74 and/or signature data storage 76. The signature comparison processor 80 usually compares a signature extracted by the signature extraction and encoding processor 74 from a freshly scanned token 30 with a signature previously stored in the signature data store 76 and associated with a genuine token 20 In the form shown symbolically by the separation between the upper part of figure 8, in relation to the capture and storage of the signature ture of the genuine mark, and the lower part of figure 8, in relation to capturing, comparing and verifying the signature of the candidate mark, the computer system 50 that scans the candidate mark 30 may be different from the computer system 50 that scans the mark 20. If they are different, then usually they each share access to the signature data storage 76, or a copy of the stored signature data is passed from storage 76 in the genuine tag capture system 50 to the candidate brand evaluation 50.
[00087] [00087] In more detail, and with respect to figure 9, in an embodiment of a method according to the invention, in step 102, a mark, which, in this example, is illustrated as a 2-D barcode similar to that shown in Figure 1, is applied to an object, or a label that is subsequently applied to an object, by printer 65. As already explained, a printer that applies a 2-D barcode typically introduces a significant amount of artifacts that are too small to affect the readability of the visible data encoded by the barcode, and are too small for their appearance to be controllable in the printing process, but are visible (possibly only in magnification) and durable. If a particular printer does not naturally produce a good supply of artifacts, some printers can be caused to include random or pseudo-random variations in their output, as discussed further below.
[00088] [00088] In step 104, the mark is acquired by an imaging device or other suitable data acquisition device 58. The imaging device can be of any expedient form, including conventional devices or devices to be developed to follow. The only real restriction in this modality is that the imaging device must gather data about the appearance of the mark at a level of detail considerably finer than the controllable output of the device that applied the mark. In the example shown in figures 1 - 4, the detail is the shape of the contours between the light and dark areas, at a resolution considerably finer than the size of the printed 2-D barcode modules. Other examples of suitable features are described below. If the mark is being used as an anti-counterfeiting measure, it is stronger if the imaging device gathers data at a finer level of detail than the controllable output of a device that is likely to be used to apply a counterfeit mark. However, this is not
[00089] [00089] In step 106, a Unique Identifier Number (UID) included in the visible data of token 20 is decoded. If the printer 65 is on the same computer system 50 as the image acquisition device 58, the UID can be passed from one to the other, avoiding the need to decode the UID of the image acquired by the image acquisition device 58. if tag 20 does not include a UID, some other information that uniquely identifies the specific example of tag 20 will usually be required in this step.
[00090] [00090] In steps 110 and 112, the image of the mark 20 is analyzed by the signature extraction and encoding processor 74 to identify significant features. In step 120, data related to these features will then be stored in the signature data store 76 as "signature" data that uniquely identifies the mark 20. This signature data is derived from the physical and optical characteristics of geometry and appearance of the mark and, furthermore, may include data which is encoded in the mark if the mark is a data bearing symbol, such as a two-dimensional bar code. Brand properties evaluated to create signature data may include, but are not limited to, feature shape, feature contrast, edge linearity, region discontinuities, extraneous marks, print defects, color, pigmentation, contrast variations, ratios feature appearance, feature locations, and feature size.
[00091] [00091] Now, also in relation to figure 5, in the following example, deviation in the modulus pigmentation or in the intensity of the mean marking 92, displacement of the modulus 94 position in relation to a better suitable grid, the presence or the location of odd or void marks 96 in the symbol, and the shape (linearity) of long continuous edges 98 in the symbol are used as exemplary variable resources. These act as the primary metrics that make up the unique symbol signature. Illustrations of some of these features are shown in Figure 5.
[00092] [00092] In case the mark is a data carrier symbol, such as a two-dimensional bar code, the present embodiment can take advantage of the additional information incorporated by the symbol, and encoded in it. The information itself that is encoded, for example a unique or non-unique serial number, can then be included as part of the signature data or used to index the signature data for easier retrieval.
[00093] [00093] Additionally, in the case of a two-dimensional barcode or other data carrier for which a measure of quality can be established, in step 108, information representing the quality of the symbol can optionally be extracted and included as part of the data of subscription.
[00094] [00094] The quality information can be used to detect changes in the mark 20 that can cause a false determination of the mark as counterfeit, as these changes can alter the signature data of the mark. Some of the quality measurements that can be used are, but are not limited to, Unused Error Correction and Fixed Standard Damage, as defined in ISO spec 15415 "Data Matrir Grading processes" or other comparable standard. These measures make it possible to detect areas that would contribute signature data that have been altered by the damage to the brand, and thereby discount them from consideration when comparing one of the brand's signature data against the stored signature data of the genuine brand.
[00095] [00095] Weighting of subscription metrics
[00096] [00096] In this example, the ease with which each of the four metrics illustrated in Figure 5 can be extracted depends on the imaging resolution, and the metrics can be arranged in the order of the required resolution to extract useful data related to each. of the four metrics, as shown in Figure 10. In order of lowest to highest resolution, these are modulus pigmentation, modulus position offset, void/mark location, and edge shape projection.
[00097] [00097] Increasing image fidelity and resolution allows increasingly accurate analysis, making use of progressively higher precision analytics. For example, in an image with low resolution, perhaps only average pigmentation from modulo 92 and position shift from modulo 94 can be extracted with significant security, so these results are given more weight in determining the match of a symbol's signature. candidate against genuine stored data. With a high resolution image, processing can continue all the way to the thin edge projection metric 98 and use this as the highest weight consideration in determining signature match. If there are disagreements between others (lower weight) measured in relation to the expected signature, these may be due to damage to the symbol or artifacts of the image capture device. However, it is highly unlikely that damage, alteration of symbol 20, or imaging device artifacts modify a spoof code 30 to coincidently match with high accuracy the signature metric of the edge projection 98 of valid item 20. Therefore, the projection edge, if highly correlated and exhibiting adequate magnitude in dynamic range, can outperform lower resolution metrics in support of high matching confidence.
[00098] [00098] Additionally, in the preferred embodiment, using a 2-D Data Matrix code as an example, the use of Error Correction information, provided by the standard decoding algorithms of this symbology, is made to further weight metric data. signature properly. If a data region in the symbol is corrupted by token damage and this region produces a mismatch with stored signature data, while other non-corrupted regions agree well, the corrupted region's voting weight should be decreased. This mechanism prevents detectable symbol corruptions from giving a false negative result in a metric comparison of the candidate symbol against the genuine symbol signature data. The ISO 16022 "Data Matrix Symbol" specification describes an example of how Error Correction Codes can be distributed in a 2-D Data Matrix, and how corrupted and uncorrupted regions in a Data Matrix can be identified.
[00099] [00099] Magnitude Filtering
[000100] [000100] In steps 114 and 116, signature candidate capabilities are evaluated to ensure they have adequate magnitude to act as a part of each signature metric. This step ensures that the features that make up each signature metric have an actual "signal" for encoding as a distinguishing feature of the brand. Failure to apply minimum thresholds to contributing signature candidates may allow a signature that is easily included by noise in any subsequent attempts to validate a mark against the stored genuine signature, making the validation process highly susceptible to quality and fidelity limitations of ( s) device(s) used to capture the tag data for signature analysis. By ensuring that signature metrics are made up exclusively of features that meet these magnitude minima, the ability to successfully verify brand signatures with a wide variety of acquisition devices (camera-equipped cell phones, machine vision cameras, low-quality or low-resolution imaging devices, etc.) and in a wide range of environments (varied, low or non-uniform lighting, etc.) can be guaranteed or greatly facilitated.
[000101] [000101] In the preferred embodiment, using a 2-D Data Matrix code as an example, in steps 110, 112 and 114, candidate features for the four signature metrics 92, 94, 96, 98 are extracted and sorted by magnitude . As previously described, unless these methods operate on the mark, the mark 20 must be purchased in such a way that the resources can be processed in electronic form, typically as a color or grayscale image. As a preliminary step, the 2-D Data Matrix is first analyzed as a whole, and a "best fit" grid defining the "ideal" positions of the contours between matrix cells is determined. Candidate features are then selected by finding features that are more deviant from the "normal" or "ideal" states of the tag attribute(s) for the particular metric being analyzed. Considering the example 2-D Data Matrix Code shown in Figure 5, some suitable attributes are:
[000102] [000102] 1. MODULES 92 whose average color, pigmentation or brand intensity are closer to the global mean threshold that differentiates dark modules from light modules, as determined by the Data Matrix reading algorithms; this is the "lighter" dark modules and the "darker" light modules.
[000103] [000103] 2. Modules 94 that are marked in a position that is most deviant from the idealized location, as defined by a better suited grid applied to the general symbol 20. Two methods of identifying these modules are currently preferred: (a) extract the candidate mark module edge positions and comparing these edge positions with their expected positions, as defined by a better suited grid idealized for the full symbol 20; (b) extract a histogram of the contour region between two adjacent modules of opposite polarity (light/dark or dark/light), with the sample region overlaying the same percentage of each module in relation to the most suitable grid, and evaluate the histogram deviation from a 50/50 bimodal distribution.
[000104] [000104] 3. Strange marks or voids 96 on the symbol modules, whether they are either light or dark, are defined as modules that have a wide range of luminance or pigment density; that is, a module that has pigmentation levels on both sides of the global mean threshold that differentiate dark modules from light modules, with the best signature candidates being those with bimodal luminance histograms that have the longest distance between the farthest dominant modes .
[000105] [000105] 4. The shape of long continuous edges 98 in the symbol, measuring either its continuity/linearity or degree of discontinuity/nonlinearity. The preferred method of extracting this data is a projection of the luminance value referring to the pixel, with a projection length of one modulus, offset from the most suitable grid by half a modulus, which runs perpendicular to the grid line that confines this edge to the grid. better suited is the symbol.
[000106] [000106] The 2-D Data Matrix makes a good example, in that it consists of black and white square cells, in which the above-described features are easily seen. However, the same principles can certainly be applied in other forms of visible tag data encoding or non-data encoding.
[000107] [000107] Once candidate resources that meet the above criteria have been identified, the candidate resources are ranked, in step 114, in a list in order of magnitude, and are then subject, in step 116, to filtering of the magnitude threshold finding the first resource in each list that does not meet the minimum magnitude established to qualify as a contributor to this metric. The threshold can be set at any convenient level low enough to include a reasonable number of features that cannot be easily reproduced and high enough to exclude features that are not reasonably durable, or are close to the noise floor of the acquisition device. image 58. In this modality, the low magnitude end of the sorted list is then truncated from this point and the remaining resources (higher magnitude) are stored, along with their locations in the tag, as the signature data for this metric . Preferably, all resources above the truncation threshold are stored, and this implicitly includes, in the signature, the information that there are no signature resources above the filter magnitude threshold elsewhere in the token.
[000108] [000108] As it is known in advance that different marking device technologies have higher or lower signature capabilities on different attributes for use in creating Metrics signature data, the marking device type can be used to pre-weight the metrics in what is referred to as a Weighting Profile. For example, if genuine marks are created using a thermal transfer printer, it is known that edge projections parallel to the direction of movement of the substrate material are unlikely to carry a sufficient signature magnitude to encode as part of the genuine signature data . This knowledge of various marking device behaviors can be used when capturing the original genuine signature data. If employed, all metrics used in creating the genuine brand signature are weighted as appropriate for the known behaviors of this particular type of branding device, and the resulting emphasis/de-emphasis mapping of the metrics becomes a Metric Weighting Profile . In step 118, this metric weighting profile, based on the type of marking device used to create the original tag, is stored as part of the signature data.
[000109] [000109] In step 120, the subscription metrics are stored as
[000110] [000110] In this mode, the record for each symbol is indexed under a unique identifier content (typically, a serial number) included in the data explicitly encoded in the symbol. The log can be stored on a server or network-accessible data storage device, or it can be stored locally where it is needed. Copies can be distributed for local storage in multiple locations.
[000111] [000111] Low Amplitude Signature Metrics
[000112] [000112] If the example of a symbol 20, or an identifiable region in the symbol, lacks any signature feature that satisfies the minimum magnitude for one or more of the signature metrics, in this modality, this fact itself is stored as part of the signature data, thereby using the lack of significant feature variation as part of the unique identifying information for this symbol. In this case, a symbol subject to verification against this data is considered genuine only if it also has zero signature features that satisfy the minimum magnitude for the metric(s) in question, or at least significant features few enough to pass a statistical test. In these cases, the weight for this particular metric is lowered, as a region without distinguishing features is a less robust identification resource than would be a region with significant distinguishing features. A symbol or region without significant signature feature is mostly negatively used. The absence of significant features from both the Genuine Brand 20 and the Candidate Brand 30 is only weak evidence that the Candidate Brand is genuine. The presence of a significant resource in a candidate brand 30, where the
[000113] [000113] An exception is made for features of appreciable signature magnitude that can be attributed to symbol damage in candidate symbol 30, revealed through the aforementioned use of symbol Eito Correction information from the decoding algorithms of this particular symbology , and subject to the principles of weighting signature metrics with the fidelity of the captured image, as previously described.
[000114] [000114] In the extreme case where both genuine brand 20 and candidate brand 30 contain ONLY subthreshold data (as in 2 "perfect" symbols), they will be indistinguishable by the process of the present example, as the process is based on some measurable variation in both genuine and counterfeit brands to act as a means of detection. This is not a problem in practice as none of the currently contemplated usage scenarios (typically high speed online printing) produce perfect symbols.
[000115] [000115] If necessary, for example, if the printing process used is very well controlled to produce a sufficiency of measurable variations, then, in step 102, marks 20 can be created with random or quasi-random variations deliberately introduced. Such variations, then, can be detected in conjunction with detection of variations that arise naturally from the branding process in the manner described previously. For example, if the marks are printed on labels, a printer and label substrate can be used which produce high quality marks such that naturally arising variations are uninformative to reliably distinguish individual brands from one another. In this case, the printing process can be modified to introduce random or quasi-random anomalies in the printed marks, so that randomly introduced anomalies and variations that
[000116] [000116] Unlike methods that rely solely on deliberately applied security features, the present process only needs to add a minimum of quasi-random features to fortify naturally occurring variation. In this way, it is possible to create the conditions under which the mark can then sufficiently satisfy the minimum limits of the magnitude filter to create a usable signature. Such artifacts can be introduced into the mark 20 using any appropriate method. For example, in exemplary modalities, the printer can be adapted to self-create the artifacts needed as part of the printing process, or the software that generates the marks before printing can be modified to introduce artifacts, or the like. Thus, deliberately introduced artifacts can boost the performance of the systems and methods described here when using low-variation markup technologies.
[000117] [000117] Referring to Figure 11, in the present embodiment, signature metrics are stored as a sorted list, in descending order of magnitude, and includes information that locates its position in the mark from which they were extracted. In the preferred embodiment, using a 2-D Data Matrix code as an example, the process by which a candidate mark or symbol is evaluated to determine whether they are genuine is as follows:
[000118] [000118] In step 152, an image of the candidate mark 30 is acquired by the image acquisition device 58.
[000119] [000119] In step 154, the explicit data in candidate tag 30 is decoded and its unique identifier (UID) content is extracted.
[000120] [000120] In step 156, the UID is used to fetch the signature metric data originally stored for the original symbol 20 that has
[000121] [000121] In step 158, in the case of a two-dimensional bar code or other data carrier for which a quality measure can be established, quality measurements 158 for candidate mark 30 can be obtained, similarly to those obtained in step 108 for the genuine brand 20. Quality measurements can be used in subsequent analysis steps to reduce the weight given to a brand, or parts of a brand, that appear to have been damaged since it was applied. Also, if the quality measurements of the original symbol 20 have been stored as part of the genuine signature data, the stored quality measurements can be checked against the extracted signature data of the candidate mark 30.
[000122] [000122] In step 160, significant signature features are extracted from the candidate brand image 30 that was acquired in step
[000123] [000123] In step 162, the signature features are encoded for analysis.
[000124] [000124] In step 164, the signature data extracted from candidate symbol 30 is sorted in the same order (e.g. sorted by magnitude) as the original list of original symbol 20.
[000125] [000125] In step 166, the candidate signature data is compared with the stored original signature data. The data is subjected to a statistical operation that reveals a numerical correlation between the two sets of data. Each metric is subjected to individual numerical analysis that produces a measure that reflects the individual safety of the candidate symbol as the genuine item for that metric. If the tag does not contain UID data, and no alternative identification data is available, it may be necessary to search through a database of similar tags using the procedures discussed in relation to Figure 13 below. For example, in the case of Figures 1 and 3, it may be necessary to search through all the genuine 20 brands that have the same visible pattern of black and white modules. The purpose of the search is to identify, or fail to identify, a single genuine brand 20 that is uniquely similar to the candidate brand 30.
[000126] [000126] In step 168, when the Metrics Weight Profile was stored as part of the genuine signature data, this information is used to emphasize and/or de-emphasize metrics in a manner appropriate to the type of marking device used to create the tags Genuine originals.
[000127] [000127] In step 170, when the image acquisition devices 58 used in steps 104 and 152 have different sensitivities, the contributions of signature data to the overall analysis result may need to be adjusted. For example, the minimum magnitude threshold used for significant features may need to be set at an appropriate level for the least sensitive image acquisition device 58, or a particular metric may need to be omitted from the analysis set as it is known. for not carrying adequate signature magnitude on marks produced by the original marking device. In some cases, a feature that is recognized in one of the higher resolution categories on the scale shown in Figure 10 may be mistaken by a lower resolution scanner for a feature in a different category. For example, a feature that is viewed at high resolution as a black module with a white blank might be viewed at low resolution as a "low pigment module". In general, the scan scanner resolution 58 is used in conjunction with the marking device's Metrics Weight Profile to determine which metrics to emphasize/de-emphasize. In this example, in the low resolution image, the resource may exist in the "low pigment" list, but there will be both "low pigment" and "empty" lists in the high resolution image. Since the methods used are ultimately subject to statistical-based analytics, the occasional occurrence of a minor mark that has dropped below the resolution of the original scan will be of negligible impact, especially since, even though not resolved as a "object", its effect will be captured on at least one of the metrics employed (such as reduced module gray level, as in this example). This has proven true in practical tests even when using scan resolutions up to 2x higher in the verification image than was used in the original signature scan.
[000128] [000128] If it is desired to explicitly correct the resolution of the original and/or verification scan, in many cases the resolution can be determined at the time of verification by detecting a comparatively abrupt drop in the number of artifacts at the scanner's resolution threshold. Alternatively, where the original scanner may be of lower resolution than the verification scanner, the scan resolution, or other information from which the resolution can be derived, may be included as metadata with the stored signature, similarly to the Weighted Profile of Metrics discussed earlier. Whichever procedure is used, sorting the signature data in order by the magnitude of the artifact makes it very easy to apply or change a threshold magnitude.
[000129] [000129] In step 172, by exclusion, all locations in a tag not represented in the ranked list of resource locations that meet the minimum magnitude threshold are expected to be devoid of significant signature features when parsing a genuine tag . This condition is evaluated by examining the signature magnitude resource at all locations in a candidate tag where subthreshold resources are expected, and by fitting the results to the appropriate metric in the negative direction when resources exceeding the minimum threshold are found. If significant features are found in a region determined to be damaged when evaluated against symbol error correction or other quality attributes, the adjustment is lessened or not performed depending on the location of the damage relative to the feature's extraction point and the nature of the particular metric involved. For example, if a discrepancy in a signature feature from the original tag 20 is extracted from a module of candidate tag 30 that is close to, but not the same as, the damaged module(s), the negative adjustment for the metric because this feature can be decreased by a proportion that reflects reduced security in the metric signature, because the previous module, which is close to a known damaged region, may well have suffered damage that affects the metric, but falls below the detectable threshold of the quality or ECC evaluation mechanism of the symbology. If the discrepancy is drawn directly from a damaged module, or if the metric is one of the types that span multiple modules and this span includes the damaged one, the adjustment will not be applied.
[000130] [000130] In step 174, these individual security values are then used to determine an overall security in candidate symbol 30 as genuine (or spoofed), with the individual security values being weighted appropriately, in the manner described above, using information from symbol image, resolution and damage fidelity.
[000131] [000131] In step 176, it is determined whether the result is sufficiently defined as acceptable. If the comparison of signature data yields an indeterminate result (for example, the individual metrics having contradictory indications unresolvable through the use of the data weighting mechanism), the user who submits the symbol for verification is encouraged to resubmit another image of the symbol for processing, and the process returns to step 152.
[000132] [000132] For practical reasons, the number of retries allowed is limited. In step 178, it is determined whether the retry limit has been exceeded. If so, further return for rescan is prevented.
[000133] [000133] Once the analysis has completed successfully, the comparison analysis results are reported in step 180. The report can be pass/fail, or it can indicate the level of security in the result. These results can be displayed locally or transferred to a networked computer system or other device for further action. If the result is still indeterminate when the retry threshold is reached, it also proceeds to step 178, where the indeterminate result can be reported as such.
[000134] [000134] By storing the signature data extracted from the trademark 20 shown in Figure 1, the present method is able to recognize this same trademark as genuine when presented as a candidate trademark 30 due to the fact that, when analyzed by it process, be determined to have the same signature data, at least up to a desired level of statistical security. Similarly, the present method is able to identify a counterfeit copy 30 of the trademark shown in Figure 1 or to distinguish a different unique example 30 of the trademark, by recognizing that the signature data, for example, extracted from the trademark example in figure 3, it does not correspond to that originally stored from when the genuine brand shown in figure 1 was originally processed.
[000135] [000135] In the development of signature metrics in the preferred modality, immunity to distortions of the substrate on which the analyzed marks are made can be important. Module luminance or color, displacement of module grid position, void or mark locations and edge profile shape are properties in which the extraction methods employed can become highly immune to the impacts of signature data caused by presentation on substrates distorted. This is accomplished by using feature extraction methods that dynamically scale on the geometry of the presented token, independent of changes in the token's aspect ratio. The primary mechanism for this in the preferred modality is to create the most suitable grid for the candidate tag at the start of extraction. This is especially important in the case where the genuine mark 20 is made on a label that runs on a flat label web, and the label is then applied to an object that is not flat, such as a bottle with a surface. curve. The candidate brands 30 submitted for analysis to verify their status as genuine or counterfeit will, of course and usually, be acquired for processing while on the non-flat surface (a rounded bottle in this example). The ability to verify symbols presented on multiple substrate geometries with minimal impact on reported signature metrics represents a significant advantage for the methods described herein.
[000136] [000136] To make extracting accurate signature data additionally robust, whenever possible, the methods of this invention use area location reference in the analyzed symbol to compose the signature data. This provides greater immunity to things such as the aforementioned substrate distortion, non-uniform illumination of the candidate symbol when acquired for processing, suboptimal or poor quality optical characteristics in the acquisition device, or many other environmental or systematic variables. For the preferred modality, the metric reference locations are:
[000137] [000137] 1. Average modulus color, pigmentation, or mark intensity reference the nearest neighbor(s) of the opposite modulus state (dark vs. light or light vs. dark). When a cell is identified as a significant resource 92 with deviating average pigment density, the cells for which it was a closest neighbor may need to be re-evaluated by discounting the identified deviant cell as a reference.
[000138] [000138] 2. Module grid position offset is referenced to the most suitable grid of the general symbol and as such has native adaptive reference location.
[000139] [000139] 3. The analysis of extraneous or empty marks in the symbol modules uses references of color, pigmentation or mark intensity at the module location - that is, the luminance histogram of the image in the analyzed module itself provides reference values for the methods applied.
[000140] [000140] 4. The projection methods used to extract the long continuous edge shapes in the symbol are differential in nature and have native immunity to typical impacting variables.
[000141] [000141] Now, in relation to figure 12, an alternative modality is similar to the process described in relation to figure 5, but it can use different types of marks than the 2-D symbol. For example, the symbol could be a linear I-D barcode, a company logo, etc. Figure 12 shows some features of a linear 1-D barcode 200 that can be used as signature metrics. These include: variations in bar width and/or spacing between bars 202; variations in average color, pigmentation, or intensity 204; voids on black bars 206 (or black dots on white bands); or irregularities in the shape of the edges of bars 208. Analysis by the autocorrelation method
[000142] [000142] In the modalities described above, the raw list of data for each metric is first matched by array index and subjected to normalized correlation in relation to an extracted metric of similar order defined from a candidate symbol. These correlation results are then used to arrive at a match/no match (genuine versus fake) decision. To do this, signature storage necessarily includes the sort order of the original genuine symbol modules as well as the trained metric values themselves, complete for each metric. In addition to the need for exhaustive storage, the raw data is not "normalized" as each metric has its own scale, sometimes unlimited, which complicates the selection of storage bit depths. A typical implementation of the above-described embodiments has a stored signature size of approximately 2 kilobytes.
[000143] [000143] Now, referring to figures 13 through 17, an alternative modality of the post-processing, storing and comparing metrics methods is applied after the original artifact metrics are extracted and made available as an associated list of the index array (associatable by the position of the module in the symbol). Based on autocorrelation, the application of this new post-processing method can, in at least some circumstances, produce several significant benefits, when compared to the subscriptions of previous modalities. More significant is a reduction in data packet size: a 75 °/) reduction in stored signature data has been realized. Even more (up to 90 °/) reduction is possible by applying some additional data compression methods
[000144] [000144] When, in the above-described modalities, the analysis of a particular set of metrics data takes the form of comparing the sorted raw metrics extracted from a candidate symbol against the equally ordered raw metrics extracted from the genuine symbol, the autocorrelation method compares the autocorrelation series of ranked candidate symbol metrics data with the autocorrelation series of ranked (stored) genuine symbol data—effectively, we now correlate the autocorrelations. For clarity, the well-known statistical operation nETiyi-EziEy8 "" = JnE2-(Ezi)' JnEy -(Eyi)' is the common Normalized Correlation Equation, where: r is the result of the correlation, n is the length of the metric data list, ex and y are the Genuine and Candidate metric data sets.
[000145] [000145] When the operation is implemented as an autocorrelation, both x and y data sets are equal.
[000146] [000146] To produce the autocorrelation series, the correlation is performed multiple times, each time shifting the x series by an additional index position relative to the y series (remembering that y is a copy of x). As the offset progresses, the dataset must "roll" back to the beginning, as the last index in data series y is exceeded due to the offset of index x; this is often accomplished more practically by duplicating the y data and "slipping" the x data from the
[000147] [000147] In implementing the autocorrelation approach, the first benefit observed is that it is not necessary to store the signature data values themselves as part of the stored data. In autocorrelation, a series of data is simply correlated with itself. So, where it was previously necessary to distribute both the order of extraction (sort) and values of the genuine signature data to the verification device for validation, it is now only necessary to provide the sort/emation order for the operation of the autocorrelation series.
[000148] [000148] The genuine autocorrelation signature required to compare how the candidate symbol results does not require storing or passing the genuine data to the verifier. Because the signature generation operation is always performed on classified metric data, the autocorrelation series for the original artifact information is always a simple polynomial curve. Therefore, rather than having to store the entire autocorrelation series of each genuine symbol metric, it is sufficient to store a set of polynomial coefficients that describe (in a predetermined order and precision) a better-fitting curve that matches the shape. of the genuine autocorrelation results for each metric.
[000149] [000149] In one modality, r,y is computed, where each term Xj is an artifact represented by its magnitude and location, and each term yi = x(i+j)3 where j is the displacement of two sets of given, for j = 0 to (n - 1). Because Xj is sorted by magnitude, and magnitude comprises the most significant digits of Xj, there is a very strong correlation at or near j = 0, decreasing rapidly towards j = n/2. Because y is a copy of x, j and n - j are interchangeable. Therefore, the autocorrelation series always forms the U-shaped curve shown in figure 13, which is necessarily symmetric around j = 0 and j = n/2. So, in fact, it is
[000150] [000150] In practice, it has been found that a 6'-order equation that uses 6-byte floating point values for the coefficients always matches the genuine data at 1 °/o curve fit error or "recognition fidelity". That is, if a candidate validation is done using the real autocorrelation numbers and then the validation is again done on the same token using the polynomial-modeled curve, the obtained match scores will be 1 °/) relative to each other . This is true of both the high match score for a genuine candidate brand and the low match score for a candidate counterfeit brand. This allows a complete autocorrelation series to be represented with just 7 numbers. Considering that 100 data points are obtained for each metric and that there are 6 metrics (which turned out to be reasonable practical numbers), that produces a reduction from 600 data values to just 42, with no loss of symbol differentiability or analysis fidelity . Even if the individual numbers are larger, for example if the 600 raw numbers are 4-byte integers and the 42 polynomial coefficients are 6-byte floating-point numbers, there is an almost 90 °/o reduction of the data. In an experimental prototype, 600 individual byte values became 42 4-byte floating points, reducing 600 bytes to 168 bytes, a 72 °/o reduction.
[000151] [000151] Additionally, the stored signature data is now explicitly limited and normalized. The polynomial coefficients are expressed in a fixed precision, the autocorrelation data itself is, by definition, always between -1 and +1, and the sort order list is simply the location of the modulo angel's index in the analyzed symbol. For a 2-D data array, the modulo array index is an index sorted by rasterizing the modulo position in a symbol, sorted from the conventional source data for this symbology and thus has a maximum size defined by definition of the symbology of the matrix. In a common type of 2-D data array, the origin is where two solid bars bordering the left and bottom sides of the grid meet. A default ranked list length of 100 data points for each metric is also established, providing a predictable, stable, and compact signature.
[000152] [000152] In one embodiment, comparing a genuine signature to a candidate now begins with "reconstituting" the genuine symbol's autocorrelation signature by using the stored polynomial coefficients. Then the raw metrics data is extracted from the candidate symbol, and is sorted in the same sort order, which can be indicated as part of the genuine subscription data if it is not pre-determined.
[000153] [000153] The candidate metrics data is then auto-correlated. The resulting autocorrelation series can then be correlated against the genuine autocorrelation curve reconstructed for this metric or, alternatively, the two curves can be compared by computing a curve fit error between the pair. This correlation is graphically illustrated in Figures 13 and 16. This final correlation score then takes the individual "matching" score for this particular metric. Once completed for all metrics, the "Match" scores are used to make the genuine/fake decision for the candidate symbol.
[000154] [000154] Additionally, additional use can be made of the autocorrelation curves by applying energy series analysis to the data through the discrete Fourier transform (DFT). For clarity, the well-known operation
[000155] [000155] The Energy Series of the DFT data is then calculated. Each hequence component, represented by a complex number in the DFT series, is then analyzed for magnitude, with the phase component discarded. The resulting data describes the spectral energy distribution of the metric data, from low to high hequence, and this forms the basis for further analysis. Examples of this energy series are shown graphically in figures 14, 15 and 17.
[000156] [000156] Two frequency domain analytics are employed - Kurtosis and a measure of energy distribution around the centerband frequency of the full spectrum, referred to as Distribution Offset.
[000157] [000157] The uniform polynomial curve of the genuine symbol metric signatures (which arises from the magnitude classification) produces recognizable characteristics in the spectral signature when analyzed in the Hequence domain. A candidate symbol, when the metrics data is extracted in the same order prescribed by the genuine signature data, will have a similar spectral energy distribution if the symbol is genuine; that is, the genuine rank order "agrees" with the candidate metric magnitudes. Discordance in ranked magnitudes, or other overlapping signals (such as photocopying artifacts), tend to appear as high Hequence components that are otherwise absent from the genuine symbol spectra, thus providing an additional measure of symbol authenticity. This addresses the possibility that a spoof autocorrelation series might still satisfy the minimum statistical matching threshold of the genuine symbol.
[000158] [000158] Along with the autocorrelation matching score, this energy series distribution information is used as a measure of "safety" in checking for a candidate symbol.
[000159] [000159] Figure 13 shows a comparison of the autocorrelation series for a single metric between a genuine item (polynomial approximation) and a candidate symbol (genuine in this case). Note the intimate agreement; here, the correlation between the 2 autocorrelation series exceeds 93 °/).
[000160] [000160] Figure 14 is an energy series from the original genuine autocorrelation data used for Figure 13. It can be clearly seen that the spectrum is dominated by low Hequences.
[000161] [000161] Figure 15 is a series of energy similar to Figure 14 from an image acquired by cell phone of the genuine item of Figure 14. Some image noise is present, but the overall energy spectrum closely matches the genuine spectrum, with the same dominance of the low-hequence components.
[000162] [000162] Figure 16 shows a comparison of the autocorrelation series for a single metric between the polynomial approximation for a genuine item and a candidate symbol (here, a fake). There is considerable disagreement, and the candidate autocorrelation is noticeably more irregular than in Figure 13. The numerical correlation between the two series is low (< 5 °/0), and the irregular shape of the data is also apparent in the DFT analysis (a follow).
[000163] [000163] Figure 17 shows the series of energy from the image acquired by cell phone of the forged symbol of the graphical representation
[000164] [000164] Now, referring to figure 18, in some implementations it is desirable to avoid the use of computationally intensive methods, such as numerical correlation or other statistical operations. In other instances, the trademark that is used for signature extraction may not be a data-carrying symbol, or it may be a symbol with limited data capacity that does not allow the association of the trademark's signature metrics with a unique identifier, such as as a serial number. In an alternative embodiment, the signature data for the token may be encoded as a sequence of bytes, which may be viewed as ASCII characters, rather than the numerical magnitude data used in the above example. This alternative data format provides the ability to use the signature data directly as a search device for a particular brand, as in a database, for example, as would normally be done using a serial number in the case of a pomdor symbol. of data. When encoding the tag data as a literal string of ASCII characters, the signature ASCII data itself becomes the unique identifier information for the tag, which acts as a serial number would, for example, as in the case of a symbol. data carrier.
[000165] [000165] In this modality, instead of storing the location and magnitude of each signature metric for a brand, what is stored is the presence (or absence) of the significant signature resources and each of the evaluated locations in a brand . For example, in the case of a 2-D Data Array symbol that does not carry/encode a unique identifier/serial number, the signature data can be stored as a string of characters, each encoding the presence/absence of a resource which exceeds the minimum magnitude threshold for each signature metric in a module, but does not further encode data about the magnitude or number of features in any one metric. In this example, each module in the symbol has 4 bits of data, one bit for each of the signature metrics, where a "1" indicates that the particular metric signature has a significant resource at that location in the module. Therefore, in this example, every possible combination of the 4 metrics extracted and tested against minimum magnitude limits can be encoded in half a byte per module; 0000 (0 hexadecimal), which means that none of the tested signature metrics are present to a degree greater than the minimum magnitude in this particular module, up to 1111 (F hexadecimal), which means that all four tested signature metrics are present in a degree greater than the minimum magnitude in this particular module.
[000166] [000166] In the example of a 2-D data matrix 250 shown in figure 18, the first six modules are encoded as follows. A first module 252 has no artifact for average luminance: it is satisfactorily black. It has no grid offset. It has a big white void. It has no edge shape artifact: its edges are straight and regular. So it is encoded at 0010. A second module 254 has a void and edge shape artifact, and is encoded at 0011. A third module 256 is noticeable gray instead of black, but has no other artifacts, and is encoded at 1000. A fourth module 258 has no artifacts, and is encoded at 0000. A fifth module 260 has a grid offset, but no other artifacts, and is encoded at 0100. A sixth module has no artifacts, and is encoded at 0000 Thus, the first six modules are encoded as 00100011 10000000 01000000 binary, or 238040 hexadecimal, or 35-128-64 decimal, or #€@ASCII (Some ASCII codes, especially those in the extended range from 128-255 decimal, have variable character assignments. This is not important for the present implementation, as they are never actually expressed as human-readable characters.).
[000167] [000167] Analysis under the Sequence Literal Coding Modality
[000168] [000168] Genuine brand signature metrics are stored as an ASCII string, encoding the signature data as stated. Using a 2-D Data Matrix code as an example, with a typical symbol size of 22 x 22 modules, the part of the ASCII string that contains the unique signature data will be 242 characters in length, assuming the data is packed. in 2 modules per character (byte). Genuine tag signature string is stored in a database, verified file, text document, or any other suitable construct to store populations of distinct character strings. The stored data can be on local storage where it is expected to be needed, or it can be searchable over a network on any connected data storage device or server.
[000169] [000169] In this example, the process by which a candidate brand is evaluated to determine whether it is genuine is as follows:
[000170] [000170] The advantages of the embodiments of the present invention include, without limitation, the ability to uniquely identify an item by the use of a mark that has been placed on the item for another purpose, without the need to specifically introduce visible or disguised elements for the purposes of anti-counterfeiting. An additional advantage is that such identification can be very difficult to falsify. Additional advantages include the ability to integrate the functions of the present invention into existing technologies commonly used to read bar code symbols, such as machine vision cameras, bar code readers and camera-equipped consumer "smart phones" without change the primary behavior, construction, or usability of devices. Another advantage, in the case of a two-dimensional bar code, for example, is the ability to use the signature data as a device for providing a redundant data carrier for the purpose of identifying an item.
[000171] [000171] In an example where damage to the candidate mark makes it only partially readable, or makes it impossible to read and/or decode a data carrier symbol, or the like, undamaged identification features of only a part of the mark may be sufficient to identify the brand. Once the candidate brand is thus identified with a genuine brand, the signature of the genuine brand can be retrieved from storage, and
[000172] [000172] Although the foregoing description of the invention enables those skilled in the art to make and use what is currently considered to be the best way of it, those skilled in the art will understand and perceive the existence of variations, combinations and equivalents of the modality, method and examples specifics exposed here. Therefore, the invention is not to be limited by the above-described embodiment, method and examples, but by all embodiments and methods within the scope and spirit of the invention.
[000173] [000173] For example, an example of features of a 2-D barcode is described in relation to figure 5. An example of features of a 1-D barcode is described in relation to figure 12. other symbols such as a company logo can be used as a target symbol. The features, and specific variations in these features, which are used as subscription metrics are almost unlimited, and it is up to those skilled in the art, with an understanding of this specification, to choose an appropriate or available symbol and choose appropriate metrics and features to make the gifts methods. In some embodiments, the tag need not be applied with a view to extract signature data according to the present methods. Instead, a tag that has already been created can be used as long as it contains adequate artifact resources.
[000174] [000174] When an original tag is applied to an original item and/or an original item is attached to an original object, the tag or item may contain information about the item or object. In this case, the methods and systems described above may include checking information about the item or object that is included in the tag or item, even when the underlying item or object is not physically replaced or changed. For example, when an object is marked with an expiration date, it may be desirable to reject an object with an expiration date changed to "not authentic" even if the object itself is the original object. Modalities of the present systems and methods will produce this result if the artifacts used for verification are discovered on the date of exi)tion, for example, as imprint imperfections.
[000175] [000175] The modalities have been described primarily in terms of acquiring the full 2-D barcode for signature data. However, the brand can be divided into smaller zones. When the original tag is enough gmnde and has enough artifacts that are potential signature data, only one zone, or less than all zones, can be acquired and processed. When more than one zone is acquired and processed, signature data from different zones can be recorded separately.
[000176] [000176] Although the modalities have been primarily described in terms of distinguishing an original mark (and by the implication of an original item to which this mark is applied or attached) from a counterfeit copy of the mark, the present methods, apparatus and products may be used for other purposes, including distinguishing between different instances of the original brand (and item).
[000177] [000177] In the interest of simplification, specific modalities were described in which artifacts are defects in the printing of a printed mark, applied either directly on the item that must be checked or on a label applied to an object that must be checked. However, as already mentioned, any resource that is sufficiently detectable and permanent, and sufficiently difficult to duplicate, can be used.
[000178] [000178] Some of the embodiments have been described as using a database of signature data for genuine items, in which a search is conducted for signature data that at least partially matches signature data extracted from a candidate tag. However, if the candidate item is identified as a specific genuine item in some other way, a search may be unnecessary and the signature data extracted from the candidate tag can be compared directly with the signature data stored for the specific genuine item.
[000179] [000179] Accordingly, reference should be made to the appended claims, rather than the stated specification to indicate the scope of the invention.

权利要求:
Claims (25)
[1]
1. Method for verifying an item's identity, characterized by the fact that it comprises: using a scanner, examining the item in relation to item-specific artifacts, in which at least some of the artifacts were not controllably producible in the production of the item; use a processor to: extract information associated with artifacts; ranking the information associated with the non-controllably producible artifacts of an order according to a characteristic of the non-controllably producible artifacts; storing data representing the ranked information on a non-temporary computer readable storage device (76) separate from the item; calculate an autocorrelation series from the ranked information; and approximating the autocorrelation series to a polynomial, where the stored data comprises data representing the autocorrelation series and where the data representing the autocorrelation series comprises polynomial coefficients for the selected order.
[2]
2. Method according to claim 1, characterized in that the information extracted by the processor comprises information that represents a plurality of artifact categories, and in which the processor ranks the information separately for each artifact category.
[3]
3. Method according to claim 1, characterized in that: the item comprises a brand (30) comprising an identifier, the non-controllably producible artifacts are in the brand (30),
the identifier is associated with the item non-controllably producible artifacts do not change the association; and the processor performs storage by storing the information to be at least partially localizable using the identifier.
[4]
4. Method for verifying an identity of an unverified item, characterized in that it comprises: using a scanner, examining the unverified item against unverified artifacts specific to the unverified item, in which at least some of the artifacts have been not controllably producible in the production of the unverified item; use a processor to: extract information associated with unverified artifacts; retrieving stored data containing ranked information associated with original artifacts of an original item from a storage device (76); compare information associated with unverified artifacts with ranked information associated with original artifacts; and when the information associated with the unverified artifacts matches the ranked information associated with the original artifacts, verifying the unverified item as a verified original item; where: at least some of the artifacts are artifacts of a symbol that encodes data and supports error detection, the processor extracts information representing the unverified artifacts by determining an error state of a symbol having the unverified artifacts; and where the error state indicates that part of the symbol is damaged, the processor performs the comparison by discounting artifacts in the damaged part of the symbol.
[5]
5. Method according to claim 4, characterized in that the determination comprises evaluating a statistical probability that the information from the unverified artifacts corresponds to the information from the original artifacts.
[6]
6. Method according to claim 5, characterized in that it further comprises using the processor to: in case the statistical probability exceeds a first threshold, determine that the unverified item is a verified original item; in case the statistical probability is below a second threshold lower than the first threshold, determine that the unverified item is not an original item; and, in case the statistical probability is between the first and second thresholds, report that it cannot be determined whether the unverified item is an original item.
[7]
7. Method according to claim 4, characterized in that at least some of the artifacts are artifacts of a symbol that encodes data and supports error detection, in which the processor extracts information representing the unverified artifacts by including determining a error state of a symbol that has unchecked artifacts, and where the error state indicates that part of the symbol is damaged, and the processor, when performing the comparison, discounts artifacts on the damaged part of the symbol.
[8]
8. Method according to claim 4, characterized in that the processor, when performing the comparison, corrects properties of at least one of the device that created the original artifacts, of the device used in examining the original item for the information it represents the original and apparatus artifacts used in examining the unverified item for information representing the unverified artifacts.
[9]
9. Method according to claim 8, characterized by the fact that: artifacts are of different categories; the processor, in determining whether the information from unverified artifacts matches the information from the original artifacts, compares the unverified and original artifacts in each category and combines the results of the comparisons; and the processor, when performing the correction, weights the combination according to a known tendency for the apparatus that created the original artifacts to produce artifacts in different categories with different frequencies or different values of a characteristic.
[10]
10. Method according to claim 4, characterized in that it further comprises using the processor to: rank the information of unverified non-controllably producible artifacts in an order; and calculate an autocorrelation series of ranked unverified artifact information; where the stored data comprises data representing a series of autocorrelation of the ranked original item artifacts; and wherein the processor performs the comparison by comparing the unverified and original autocorrelation series to determine whether the information from the unverified artifacts matches the information from the original artifacts.
[11]
11. System to verify the identity of an item, by the method as defined in claim 1, characterized in that it comprises:
the scanner; an operable encoder to rank the information from the non-controllably producible artifacts in an order according to a characteristic of the non-controllably producible artifacts and to encode the extracted information into computer-readable data; and the storage device (76).
[12]
12. The system of claim 11, further comprising an original item producer operable to produce an original item, wherein artifacts are item resources that are produced when the original item producer produces the item, and at least some of the artifacts are not controllably producible by the original item producer.
[13]
13. System according to claim 11, characterized in that it further comprises at least one original item for which ranked artifact data is stored on the storage device.
[14]
14. System to verify the identity of an item, by the method as defined in claim 4, characterized in that it comprises: the scanner; and the processor.
[15]
15. Non-temporary computer-readable storage media, characterized in that it stores computer-readable instructions which, when executed on a suitable computing processor, verify the identity of an item, in accordance with the method as defined in claim 1.
[16]
16. Non-temporary computer-readable storage media, characterized in that it stores computer-readable instructions which, when executed on a suitable computing processor, verify the identity of an item, in accordance with the method as defined in claim 4.
[17]
17. Method for verifying the identity of an unverified item, characterized by the fact that it comprises: using a verification scanner to examine the unverified item for unverified artifacts specific to the unverified item, in which at least some of the artifacts have not verified were not controllably producible in the production of the unverified item; use a processor to: extract information associated with unverified artifacts; retrieve stored data containing ranked information associated with the original artifacts of an original item from a storage device, where both unverified artifacts and the original artifacts are of distinct categories; compare information associated with unverified artifacts with ranked information associated with original artifacts; correct properties of at least one of: a device that created the original artifacts, a device used in examining the original item for information representing the original artifacts, and a device used in examining the unverified item; when the information associated with the unverified artifacts matches the information associated with the original artifacts, verifying the unverified item as a verified original item; and determining whether unverified artifact information matches original artifact information by comparing unverified and original artifacts in each category and combining the results of the comparison; where the processor performs the correction by weighting the combination of results according to a known trend of the device that created the original artifacts to produce artifacts in different categories with different frequencies or different values of a characteristic.
[18]
18. Method according to claim 17, characterized in that the processor performs the comparison by analyzing a statistical probability that the unverified artifact information corresponds to the original artifact information.
[19]
19. Method according to claim 18, characterized in that it comprises: in case the statistical probability exceeds a first threshold, using the processor to determine that the unverified item is a verified original item; in case the statistical probability is below a second threshold lower than the first threshold, use the processor to determine that the unverified item is not an original item; and, in case the statistical probability is between the first and second thresholds, use the processor to report that it cannot be determined whether the unverified item is an original item.
[20]
20. Method according to claim 17, characterized in that it further comprises using the processor to: rank the information of unverified non-controllably producible artifacts in an order; and calculate an autocorrelation series of ranked unverified artifact information; where the stored data comprises data representing a series of autocorrelation of the ranked original item artifacts; and wherein the processor performs the comparison by comparing the unverified and original autocorrelation series to determine whether the information from the unverified artifacts matches the information from the original artifacts.
[21]
21. System for establishing the identity of the item with a method as defined in claim 20, characterized in that it comprises: an original item scanner operable to examine the original item and extract information representing artifacts of the original item that were controllably non-producible and in production with the original item; and an encoder operable to rank the information of original artifacts in an order according to the characteristic of the original artifacts and to encode the extracted information into the computer readable data; and the storage device.
[22]
22. System according to claim 21, characterized in that it further comprises a verification scanner and the processor.
[23]
23. The system of claim 21, further comprising an original item production apparatus operable to produce the original item, wherein the original artifacts are features of the item that are produced when the item production apparatus original produces the item.
[24]
24. System for verifying the identity of an item with a method as defined in claim 17, characterized in that it comprises a verification scanner and the processor.
[25]
25. Non-temporary computer-readable storage media, characterized in that it stores computer-readable instructions which, when executed on a suitable computing processor, verify the identity of an item, in accordance with the method as defined in claim 17.

类似技术:

---

公开号 | 公开日 | 专利标题

---

US10922699B2|2021-02-16|Method and system for determining whether a barcode is genuine using a deviation from a nominal shape

---

US10997385B2|2021-05-04|Methods and a system for verifying the authenticity of a mark using trimmed sets of metrics

---

AU2015223174B2|2017-08-31|Methods and a system for verifying the identity of a printed item

---

US10546171B2|2020-01-28|Method and system for determining an authenticity of a barcode using edge linearity

---

CA2960716C|2019-01-22|Methods and a system for verifying the authenticity of a mark

同族专利:

---

公开号 | 公开日

---

KR101581196B1|2015-12-30|

---

SG11201405180SA|2014-09-26|

---

PL2820592T3|2018-07-31|

---

US20170221076A1|2017-08-03|

---

HK1207453A1|2016-01-29|

---

PT2820592T|2018-03-22|

---

MY174606A|2020-04-29|

---

IN2014DN07110A|2015-04-24|

---

EP2820592B1|2018-01-31|

---

ES2664722T3|2018-04-23|

---

US20190354994A1|2019-11-21|

---

US20200151739A1|2020-05-14|

---

US20130228619A1|2013-09-05|

---

AU2013225800A1|2014-10-16|

---

AU2013225800B2|2015-04-09|

---

US10380601B2|2019-08-13|

---

EP2820592A1|2015-01-07|

---

US20150083801A1|2015-03-26|

---

CN104303192B|2016-11-23|

---

CN107146088A|2017-09-08|

---

EP2820592A4|2015-11-11|

---

DK2820592T3|2018-04-23|

---

KR20150093254A|2015-08-17|

---

WO2013130946A1|2013-09-06|

---

EP2820592B8|2018-03-07|

---

US10922699B2|2021-02-16|

---

CN107146088B|2021-09-17|

---

KR20140139530A|2014-12-05|

---

US8950662B2|2015-02-10|

---

US10552848B2|2020-02-04|

---

RS57033B1|2018-05-31|

---

CN104303192A|2015-01-21|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

GB8812890D0|1988-05-31|1988-07-06|De La Rue Co Plc|Security device & methods & apparatus for verification|

---

US5153418A|1990-10-30|1992-10-06|Omniplanar, Inc.|Multiple resolution machine readable symbols|

---

CH684590A5|1991-09-26|1994-10-31|Grapha Holding Ag|Process for uniformity testing of printed sheets and apparatus thereof for implementation.|

---

US5325167A|1992-05-11|1994-06-28|Canon Research Center America, Inc.|Record document authentication by microscopic grain structure and method|

---

US5583950A|1992-09-16|1996-12-10|Mikos, Ltd.|Method and apparatus for flash correlation|

---

US5521984A|1993-06-10|1996-05-28|Verification Technologies, Inc.|System for registration, identification and verification of items utilizing unique intrinsic features|

---

US5465303A|1993-11-12|1995-11-07|Aeroflex Systems Corporation|Automated fingerprint classification/identification system and method|

---

US6449377B1|1995-05-08|2002-09-10|Digimarc Corporation|Methods and systems for watermark processing of line art images|

---

HU211281B|1994-07-25|1996-11-28|Lajos Pikler|Method for identificating objects on the basis of the forming and checking their surface roughness, further an object having surface suitable to said identification|

---

US6032861A|1995-01-03|2000-03-07|Lemelson; Jerome H.|Method and apparatus for encoding and decoding bar codes with primary and secondary information and method of using such bar codes|

---

US5563401A|1995-01-03|1996-10-08|Lemelson; Jerome H.|Bar codes and methods|

---

US6543691B1|1995-01-03|2003-04-08|Jerome H. Lemelson|Method and apparatus for encoding and decoding bar codes with primary and secondary information and method of using such bar codes|

---

US5706091A|1995-04-28|1998-01-06|Nikon Corporation|Apparatus for detecting a mark pattern on a substrate|

---

US8144368B2|1998-01-20|2012-03-27|Digimarc Coporation|Automated methods for distinguishing copies from original printed objects|

---

AU4311896A|1995-12-29|1997-07-28|S. E. Axis Limited|Authentication of articles|

---

US5978495A|1996-07-17|1999-11-02|Intelnet Inc.|Method and apparatus for accurate determination of the identity of human beings|

---

US5883971A|1996-10-23|1999-03-16|International Business Machines Corporation|System and method for determining if a fingerprint image contains an image portion representing a smudged fingerprint impression|

---

US6460766B1|1996-10-28|2002-10-08|Francis Olschafskie|Graphic symbols and method and system for identification of same|

---

US6685095B2|1998-05-05|2004-02-03|Symagery Microsystems, Inc.|Apparatus and method for decoding damaged optical codes|

---

CH693693A5|1997-06-06|2003-12-15|Ovd Kinegram Ag|An apparatus for detecting optical diffraction markings.|

---

US6185318B1|1997-08-22|2001-02-06|International Business Machines Corporation|System and method for matching images an aligned string-based representation|

---

US5946103A|1998-01-29|1999-08-31|Xerox Corporation|Halftone patterns for trusted printing|

---

US6513714B1|1998-09-14|2003-02-04|Psc Scanning, Inc.|Character reconstruction and element level processing in bar code scanning system|

---

US6760472B1|1998-12-14|2004-07-06|Hitachi, Ltd.|Identification method for an article using crystal defects|

---

US7038766B2|1999-04-01|2006-05-02|Microtrace, Llc|Identification particles and system and method for retrospective identification using spectral codes|

---

AU6610300A|1999-07-28|2001-02-19|Terrance A. Tomkow|System and method for verifying delivery and integrity of electronic messages|

---

WO2001024107A2|1999-09-30|2001-04-05|Dna Technologies, Inc.|Product authentication system|

---

US6398117B1|2000-01-14|2002-06-04|Hewlett-Packard Co.|Method and system for combining bar codes of different encoding dimensions|

---

US7162035B1|2000-05-24|2007-01-09|Tracer Detection Technology Corp.|Authentication method and system|

---

US7152047B1|2000-05-24|2006-12-19|Esecure.Biz, Inc.|System and method for production and authentication of original documents|

---

US7089420B1|2000-05-24|2006-08-08|Tracer Detection Technology Corp.|Authentication method and system|

---

US7107453B2|2000-05-25|2006-09-12|Hewlett-Packard Development Company, L.P.|Authenticatable graphical bar codes|

---

US6865337B1|2000-08-08|2005-03-08|Conexant Systems, Inc.|System and method for detecting modifications of video signals designed to prevent copying by traditional video tape recorders|

---

EP1325464A1|2000-10-11|2003-07-09|Digimarc Corporation|Halftone watermarking and related applications|

---

KR20010025234A|2000-11-09|2001-04-06|김진삼|A certification method of credit of a financing card based on fingerprint and a certification system thereof|

---

WO2002065782A1|2001-02-12|2002-08-22|Koninklijke Philips Electronics N.V.|Generating and matching hashes of multimedia content|

---

JP3896827B2|2001-02-21|2007-03-22|株式会社デンソー|Information code decoding method and decoding system|

---

US7013024B2|2001-04-13|2006-03-14|Pitney Bowes Inc.|Method for reading information that has been embedded in an image|

---

JP2002344736A|2001-05-21|2002-11-29|Oki Electric Ind Co Ltd|Original warrant system for print paper, printer, and verification device|

---

US7184580B2|2001-07-24|2007-02-27|Laurence Hamid|Fingerprint scar recognition method and apparatus|

---

ES2290347T3|2001-11-09|2008-02-16|International Barcode Corporation|SYSTEM AND METHOD TO INSERT CHARACTERS IN A BAR OF A BAR CODE.|

---

JP3804012B2|2002-03-08|2006-08-02|沖電気工業株式会社|Document image alteration determination method and system, and control program therefor|

---

US6970236B1|2002-08-19|2005-11-29|Jds Uniphase Corporation|Methods and systems for verification of interference devices|

---

JP4245883B2|2002-09-13|2009-04-02|株式会社日立製作所|Authenticity management method and system for electronic document and print medium thereof, program, and recording medium|

---

AU2003282943A1|2002-10-11|2004-05-04|Digimarc Corporation|Systems and methods for recognition of individuals using multiple biometric searches|

---

US20040161153A1|2003-02-18|2004-08-19|Michael Lindenbaum|Context-based detection of structured defects in an image|

---

WO2004090801A1|2003-04-08|2004-10-21|Nippon Telegraph And Telephone Corporation|Information coding device, information decoding device, and method thereof, and program|

---

US7025269B2|2003-04-24|2006-04-11|Watson Label Products Corp.|Barcodes including embedded security features and space saving interleaved text|

---

EP1477940B1|2003-04-30|2007-08-01|Hewlett-Packard Development Company, L.P.|Authentication method and system|

---

DE60323941D1|2003-04-30|2008-11-20|Hewlett Packard Development Co|Authentication method and system|

---

US7006630B2|2003-06-03|2006-02-28|Matsushita Electric Industrial Co., Ltd.|Methods and apparatus for digital content protection|

---

US7532804B2|2003-06-23|2009-05-12|Seiko Epson Corporation|Method and apparatus for video copy detection|

---

US7576842B2|2003-07-04|2009-08-18|Kwang-Don Park|Random-type identifying material, 3-D identifying system and method using the same|

---

US7044376B2|2003-07-23|2006-05-16|Eastman Kodak Company|Authentication method and apparatus for use with compressed fluid printed swatches|

---

SK10072003A3|2003-08-08|2005-03-04|Dušan Kočiš|Data channel of the background on paper carrier or other carrier|

---

US7685428B2|2003-08-14|2010-03-23|Ricoh Company, Ltd.|Transmission of event markers to data stream recorder|

---

US20050053236A1|2003-09-08|2005-03-10|Samii Mohammad M.|Printed material verification|

---

JP4460903B2|2004-01-20|2010-05-12|株式会社東芝|ID mark recognition method for semiconductor wafer|

---

AR043357A1|2004-01-23|2005-07-27|Salva Calcagno Eduardo Luis|PROCEDURE OF IDENTIFICATION OF PERSONS THROUGH THE CONVERSION OF DACTILAR FOOTPRINTS AND GENETIC CODES IN BAR CODES AND DISPOSAL USED IN THIS PROCEDURE|

---

US7697729B2|2004-01-29|2010-04-13|Authentec, Inc.|System for and method of finger initiated actions|

---

US7497379B2|2004-02-27|2009-03-03|Microsoft Corporation|Counterfeit and tamper resistant labels with randomly occurring features|

---

CA2559271A1|2004-03-12|2005-09-22|Ingenia Technology Limited|Methods and apparatuses for creating authenticatable printed articles and subsequently verifying them|

---

CN100481117C|2004-03-15|2009-04-22|武汉矽感科技有限公司|A two dimensional bar code coding-decoding method|

---

US7577844B2|2004-03-17|2009-08-18|Microsoft Corporation|Systems and methods for encoding randomly distributed features in an object|

---

US7609872B2|2004-04-05|2009-10-27|Eastman Kodak Company|Method and apparatus for verifying the authenticity of documents|

---

FR2870376B1|2004-05-11|2006-09-22|Yann Boutant|METHOD FOR RECOGNIZING FIBROUS MEDIA, AND APPLICATIONS OF SUCH A METHOD IN THE COMPUTER FIELD, IN PARTICULAR|

---

US7229025B2|2004-06-07|2007-06-12|Pitney Bowes Inc.|Barcode with enhanced additional stored data|

---

US7616797B2|2004-08-23|2009-11-10|Bailey Kenneth S|Minutia detection from measurement of a human skull and identifying and profiling individuals from the human skull detection|

---

KR100979457B1|2005-08-23|2010-09-02|가부시키가이샤 리코|Method and system for image matching in a mixed media environment|

---

JP4340618B2|2004-10-08|2009-10-07|富士通株式会社|Biometric information authentication apparatus and method, biometric information authentication program, and computer-readable recording medium recording the biometric information authentication program|

---

EP1825419A4|2004-11-09|2008-08-06|Digimarc Corp|Authenticating identification and security documents|

---

US7752077B2|2005-01-21|2010-07-06|Amazon Technologies, Inc.|Method and system for automated comparison of items|

---

US7773812B2|2005-04-11|2010-08-10|Microsoft Corporation|Method and system for performing searches and returning results based on weighted criteria|

---

US7533062B2|2005-05-27|2009-05-12|Pitney Bowes Inc.|Method for creating self-authenticating documents|

---

JP2009508226A|2005-09-12|2009-02-26|ウルトラドッツ・インコーポレイテッド|Authentication and identification of objects using nanoparticles|

---

US8212651B2|2005-12-15|2012-07-03|International Business Machines Corporation|Use of radio frequency identifier tags for identification of an authentication status of an item|

---

WO2007113040A1|2006-03-31|2007-10-11|International Business Machines Corporation|Method and systems using identifier tags and authenticity certificates for detecting counterfeited or stolen brand objects|

---

GB2440386A|2006-06-12|2008-01-30|Ingenia Technology Ltd|Scanner authentication|

---

US8108176B2|2006-06-29|2012-01-31|Cognex Corporation|Method and apparatus for verifying two dimensional mark quality|

---

EP2054836B1|2006-07-19|2019-05-29|Advanced Track And Trace|Methods and devices for securing and authenticating documents|

---

RU2458395C2|2006-07-19|2012-08-10|Эдванст Трэк Энд Трэйс|Methods and apparatus for ensuring integrity and authenticity of documents|

---

US20080025555A1|2006-07-31|2008-01-31|Canadian Bank Note Company, Limited|Method and apparatus for comparing document features using pattern recognition|

---

EP2084655A2|2006-10-26|2009-08-05|MAN Roland Druckmaschinen AG|Apparatus, method and process for the stochastic marking and tracking of printed products|

---

DE102006053788A1|2006-11-15|2008-05-21|Giesecke & Devrient Gmbh|Method for detecting contamination in the area of color transitions on value documents and means for carrying out the method|

---

US7422147B2|2006-12-22|2008-09-09|Walter Steven Rosenbaum|System and method for detecting fraudulent transactions of items having item-identifying indicia|

---

AU2006252239B2|2006-12-22|2010-04-01|Canon Kabushiki Kaisha|Barcode pattern|

---

WO2008078009A1|2006-12-26|2008-07-03|Alphacode|Method for characterising random elements in the reproduction of graphic elements imparting the properties of a seal thereto|

---

GB2445966B|2007-01-25|2011-05-18|Hewlett Packard Development Co|Method of and system for authenticating an item|

---

US7327859B1|2007-02-14|2008-02-05|Lam Ko Chau|Methods and systems for automated fingerprint recognition|

---

US7673807B2|2007-02-21|2010-03-09|Hewlett-Packard Development Company, L.P.|Multiple resolution readable color array|

---

US7900837B2|2007-03-14|2011-03-08|Microsoft Corporation|Optical fiber paper reader|

---

WO2008143087A1|2007-05-14|2008-11-27|International Frontier Technology Laboratory, Inc.|Authenticity validation subject, authenticity validation chip reader, and authenticity judging method|

---

NZ581538A|2007-06-01|2012-08-31|Kba Notasys Sa|Authentication of security documents by analysis of features characteristic to printing methods|

---

FR2916678B1|2007-06-01|2021-07-16|Advanced Track & Trace|PROCEDURE AND DEVICE FOR SECURING DOCUMENTS|

---

US7938331B2|2007-06-29|2011-05-10|Symbol Technologies, Inc.|Method and system for anti-counterfeit barcode label|

---

EP2045783A1|2007-10-02|2009-04-08|Kba-Giori S.A.|Method and system for controlled production of security documents, especially banknotes|

---

JP2009109419A|2007-10-31|2009-05-21|Sony Corp|Product identification apparatus, product identification method, and program|

---

US8488834B2|2007-11-15|2013-07-16|Certifi-Media Inc.|Method for making an assured image|

---

JP5023992B2|2007-11-16|2012-09-12|富士ゼロックス株式会社|Counterfeit discrimination system and counterfeit discrimination device|

---

US8267321B2|2008-02-11|2012-09-18|International Business Machines Corporation|Enhanced-density barcode|

---

JP5521276B2|2008-03-13|2014-06-11|富士通株式会社|Authentication apparatus, authentication method, and authentication program|

---

HUE026760T2|2008-03-20|2016-07-28|Univ Geneve|Secure item identification and authentication system and method based on unclonable features|

---

US8027468B2|2008-04-08|2011-09-27|Honeywell International Inc.|Method and system for camera sensor fingerprinting|

---

DE102009016194A1|2008-04-23|2009-10-29|Heidelberger Druckmaschinen Ag|Method for producing a security feature on a flat substrate|

---

US8150106B2|2008-04-30|2012-04-03|Xerox Corporation|Printer characterization, monitoring and diagnosis using dynamic test patterns generated by sensing and analyzing customer documents|

---

US8793498B2|2008-08-11|2014-07-29|Nbcuniversal Media, Llc|System and method for forensic analysis of media works|

---

WO2010021965A1|2008-08-17|2010-02-25|Dolby Laboratories Licensing Corporation|Signature derivation for images|

---

EP2313847A4|2008-08-19|2015-12-09|Digimarc Corp|Methods and systems for content processing|

---

FR2936336A1|2008-09-23|2010-03-26|Advanced Track Trace|METHOD AND DEVICE FOR AUTHENTICATING GEOMETRIC CODES|

---

JP2012523598A|2009-04-08|2012-10-04|コンチネンタル・テベス・アーゲー・ウント・コンパニー・オーハーゲー|Two-dimensional symbol code and method for reading the symbol code|

---

DE102009037626A1|2009-08-14|2011-02-17|Giesecke & Devrient Gmbh|Safe labeling of a product|

---

US8430301B2|2009-11-23|2013-04-30|Konica Minolta Laboratory U.S.A., Inc.|Document authentication using hierarchical barcode stamps to detect alterations of barcode|

---

EP2348465A1|2009-12-22|2011-07-27|Philip Morris Products S.A.|Method and apparatus for storage of data for manufactured items|

---

WO2011077459A2|2009-12-23|2011-06-30|Alexia Technologies Private Limited|A method of extracting naturally encrypted features from a natural surface of a product|

---

JP5668358B2|2010-08-10|2015-02-12|富士ゼロックス株式会社|Image processing apparatus, image processing system, and program|

---

US8905314B2|2010-09-30|2014-12-09|Apple Inc.|Barcode recognition using data-driven classifier|

---

JP2012141729A|2010-12-28|2012-07-26|Jvc Kenwood Corp|Authenticity determination method, authenticity determination device, authenticity determination system, and color two-dimentional code|

---

US8542930B1|2010-12-30|2013-09-24|Cognex Corporation|Mark reader configured to prioritize images|

---

US9022280B2|2011-06-24|2015-05-05|Verisign, Inc.|Multi-mode barcode resolution system|

---

KR101883425B1|2011-08-01|2018-07-31|삼성전자주식회사|Method for differentiation forgery by portable terminal device|

---

US8903181B2|2011-12-28|2014-12-02|Venkatesh Gangadharan|Low cost unique identification generation using combination of patterns and unique code images|

---

US9165187B2|2012-01-12|2015-10-20|Kofax, Inc.|Systems and methods for mobile image capture and processing|

---

EP2812849B1|2012-02-09|2019-12-18|Hewlett-Packard Development Company, L.P.|Forensic verification from halftone images|

---

BR112017005769A2|2014-09-23|2017-12-12|Sys Tech Solutions Inc|method for verifying authenticity of a mark, and, computer system.|

---

US20150379321A1|2012-03-01|2015-12-31|Sys-Tech Solutions, Inc.|Methods and a system for verifying the authenticity of a mark|

---

US20150169928A1|2012-03-01|2015-06-18|Sys-Tech Solutions, Inc.|Methods and a system for verifying the identity of a printed item|

---

RU2621006C1|2014-02-28|2017-05-30|Сис-Тек Солюшнз, Инк.|Methods and system for printed object authentication|

---

WO2013130946A1|2012-03-01|2013-09-06|Sys-Tech Solutions, Inc.|Unique identification information from marked features|

---

US20130254896A1|2012-03-23|2013-09-26|Infineon Technologies Austria Ag|Method to Detect Tampering of Data|

---

US9305028B2|2012-04-11|2016-04-05|Zynga Inc.|Gaming platform utilizing a fraud detection platform|

---

AT513243A1|2012-06-18|2014-02-15|Thomas Dipl Ing Fh Dipl Ing Weiss|Method or system for the unique identification of an object|

---

US10565483B2|2012-08-31|2020-02-18|International Business Machines Corporation|Two-dimensional barcode to avoid unintentional scanning|

---

US20140078556A1|2012-09-19|2014-03-20|John T. Anderson, III|Method of forming security markings|

---

US9036916B2|2012-09-27|2015-05-19|Apple Inc.|Unique part identifiers|

---

JP6164899B2|2013-04-05|2017-07-19|キヤノン株式会社|Hash value generation device, system, determination method, program, storage medium|

---

CN103279731B|2013-06-06|2016-05-25|格科微电子（上海）有限公司|Dimension code anti-counterfeit method and anti-counterfeit authentication method thereof|

---

US9233572B2|2013-08-30|2016-01-12|International Business Machines Corporation|Anti-counterfeiting opto-thermal watermark for electronics|

---

WO2015054188A1|2013-10-07|2015-04-16|Apdn , Inc.|Multimode image and spectral reader|

---

TWI525317B|2013-10-08|2016-03-11|國立清華大學|Method of Optical Defect Detection through Image analysis and Data Mining Integrated|

---

WO2015117915A1|2014-02-04|2015-08-13|Vesdo Ltd.|Method and apparatus for proving an authentication of an original item and method and apparatus for determining an authentication status of a suspect item|

---

WO2015138820A1|2014-03-12|2015-09-17|ClearMark Systems, LLC|System and method for authentication|

---

CN103903039B|2014-03-26|2017-02-22|深圳大学|Hidden code holographic anti-counterfeiting film and manufacturing method and recognition system thereof|

---

JP6708973B2|2014-09-01|2020-06-10|日本電気株式会社|Judgment method, judgment system, judgment device, and program thereof|

---

WO2016047377A1|2014-09-22|2016-03-31|富士フイルム株式会社|Image-recording device, and image defect detection device and method|

---

US9607200B2|2014-10-09|2017-03-28|Cognex Corporation|Decoding barcodes|

---

WO2016084011A1|2014-11-25|2016-06-02|Lau Tak Wai|Information bearing devices|

---

US9940572B2|2015-02-17|2018-04-10|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

US9519942B2|2015-02-17|2016-12-13|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

CN104809622A|2015-05-01|2015-07-29|海南亚元防伪技术研究所|Ultra-limit texture anti-counterfeiting method|

---

CA2989349C|2015-06-16|2018-07-17|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|RU2621006C1|2014-02-28|2017-05-30|Сис-Тек Солюшнз, Инк.|Methods and system for printed object authentication|

---

WO2013130946A1|2012-03-01|2013-09-06|Sys-Tech Solutions, Inc.|Unique identification information from marked features|

---

US20150379321A1|2012-03-01|2015-12-31|Sys-Tech Solutions, Inc.|Methods and a system for verifying the authenticity of a mark|

---

BR112017005769A2|2014-09-23|2017-12-12|Sys Tech Solutions Inc|method for verifying authenticity of a mark, and, computer system.|

---

US20150169928A1|2012-03-01|2015-06-18|Sys-Tech Solutions, Inc.|Methods and a system for verifying the identity of a printed item|

---

US10805523B2|2012-05-30|2020-10-13|Easy Printing Network Limited|Article authentication apparatus having a built-in light emitting device and camera|

---

US20140351265A1|2013-05-23|2014-11-27|Here Global B.V.|Method and apparatus for managing weight of items for transport|

---

JP2015032213A|2013-08-05|2015-02-16|株式会社東芝|Information processing apparatus, information processing method, and information processing program|

---

EP3047422A4|2013-09-20|2017-08-02|Mobile Search Security LLC|Instrument and document authentication system|

---

DE102013018036A1|2013-12-02|2015-06-03|Carl Freudenberg Kg|Arrangement for detecting counterfeit products|

---

WO2015117915A1|2014-02-04|2015-08-13|Vesdo Ltd.|Method and apparatus for proving an authentication of an original item and method and apparatus for determining an authentication status of a suspect item|

---

US11126902B2|2014-06-03|2021-09-21|IE-9 Technology Corp.|Optically variable data storage device|

---

US9471846B2|2014-11-12|2016-10-18|Xerox Corporation|Method and system for visualizing a correlation mark using an electronic device|

---

JP2018511029A|2014-12-18|2018-04-19|スリーエム イノベイティブ プロパティズ カンパニー|Batch identity verification of materials for automated anti-counterfeiting|

---

GB201501342D0|2015-01-27|2015-03-11|Univ Lancaster|Improvements relating to the authentication of physical entities|

---

US9519942B2|2015-02-17|2016-12-13|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

CA3046448C|2016-12-09|2020-06-02|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

US9940572B2|2015-02-17|2018-04-10|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

CN107624061B|2015-04-20|2021-01-22|康奈尔大学|Machine vision with dimensional data reduction|

---

RU2608259C2|2015-04-27|2017-01-17|Открытое акционерное общество "Авангард"|Method of improving protective properties of identification saw-mark|

---

US10020264B2|2015-04-28|2018-07-10|Infineon Technologies Ag|Integrated circuit substrate and method for manufacturing the same|

---

CA2989349C|2015-06-16|2018-07-17|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

EP3311336B1|2015-06-18|2020-08-26|Hicof Inc.|Authentication feature in a barcode|

---

PT3353944T|2015-09-24|2021-12-28|Sicpa Holding Sa|Re-certification of documents|

---

WO2017053951A1|2015-09-25|2017-03-30|CertiRx, Inc.|Systems and methods for linking unique identifiers embedded in machine verifiable marks|

---

CN107135091B|2016-02-29|2020-01-17|华为技术有限公司|Application quality index mapping method, server and client|

---

US10061958B2|2016-03-14|2018-08-28|Sys-Tech Solutions, Inc.|Methods and a computing device for determining whether a mark is genuine|

---

CN105915225B|2016-04-12|2019-03-05|科络克电子科技（上海）有限公司|The method of the feature enhancing and data compression of signal processing two-dimensional data array|

---

CN107067363B|2017-04-18|2018-09-04|海南大学|A kind of non-blind robust Spatial Domain Image Watermark method constant based on pixel order relation|

---

US10614350B2|2017-11-07|2020-04-07|Verifyme, Inc.|Dual code authentication process|

---

JPWO2019102911A1|2017-11-27|2020-11-19|日本電信電話株式会社|Abnormal communication detection device, abnormal communication detection method, program|

---

US10854251B2|2017-12-15|2020-12-01|Google Llc|Physical identifiers for authenticating an identity of a semiconductor component|

---

US10915662B2|2017-12-15|2021-02-09|International Business Machines Corporation|Data de-identification based on detection of allowable configurations for data de-identification processes|

---

CN109968848A|2017-12-28|2019-07-05|战马（北京）饮料有限公司|A method of the printed patterns on beverage bottle|

---

CN108074110A|2018-01-17|2018-05-25|王登顺|Utilize method for anti-counterfeit of the random sprinkling figure with digital picture correlation technique and application|

---

DE102020206060B4|2019-06-01|2021-10-21|Heidelberger Druckmaschinen Aktiengesellschaft|Dynamic mask for authentication|

---

CN110728784B|2019-10-22|2021-09-28|中钞印制技术研究院有限公司|Banknote authenticity auxiliary identification method and device, user terminal and storage medium|

---

FR3109831A1|2020-04-30|2021-11-05|Surys|Method for verifying a barcode.|

法律状态:
2018-12-04| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|

---

2019-12-31| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|

---

2021-09-21| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|

---

2022-01-04| B11D| Dismissal acc. art. 38, par 2 of ipl - failure to pay fee after grant in time|

优先权:

---

申请号 | 申请日 | 专利标题

---

US201261605369P| true| 2012-03-01|2012-03-01|

---

US61/605369|2012-03-01|

---

US201261676113P| true| 2012-07-26|2012-07-26|

---

US61/676113|2012-07-26|

---

US201261717711P| true| 2012-10-24|2012-10-24|

---

US61/717711|2012-10-24|

---

PCT/US2013/028574|WO2013130946A1|2012-03-01|2013-03-01|Unique identification information from marked features|

---