A product anti-counterfeiting verification system based on biological information

Abstract

The application discloses a product anti-counterfeiting verification system based on biological information, which comprises an information acquisition unit, a key generation unit, a quantum coding unit, a blockchain storage unit and a verification and judgment unit; the information acquisition unit acquires user biological characteristics and product identification information; the key generation unit generates biological keys by using a topological entropy fusion algorithm and realizes time evolution; the quantum coding unit converts the keys into quantum dot physical identification which is difficult to copy and has a tamper-proof function; the blockchain storage unit ensures safe data storage through multi-stage salting processing; the verification and judgment unit provides adaptive true and false judgment by using a similarity function; the application solves the problems of easy copying, lack of dynamics and verification security risks of traditional anti-counterfeiting technology, has higher anti-counterfeiting level and adaptability, provides all-round anti-counterfeiting protection for high-value products, and balances security and user experience.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of biological anti-counterfeiting, and particularly relates to a product anti-counterfeiting verification system based on biological information. BACKGROUND

[0002] Although traditional anti-counterfeiting technologies such as two-dimensional codes, RFID tags, holographic tags, etc. have played a certain role in anti-counterfeiting, with the progress of replication technology, these static anti-counterfeiting methods are facing the risk of being easily cracked and replicated. Especially in the field of high-value products, counterfeiters invest a large amount of resources to crack anti-counterfeiting technologies, making it difficult for a single anti-counterfeiting means to effectively cope with increasingly complex counterfeit attacks.

[0003] In recent years, biological feature recognition technology has made significant progress in the field of security authentication, such as fingerprint recognition, face recognition and voiceprint recognition, which have been widely applied in mobile device unlocking, authentication for payment, etc. However, there are still many challenges in applying biological information to the field of product anti-counterfeiting, mainly including the security storage of biological feature information, the stability of feature extraction, and the real-time and reliability of the anti-counterfeiting verification process. Especially under different environmental conditions, the variability of biological features will lead to unstable verification results, thereby affecting the practicality of the anti-counterfeiting system. In addition, with the development of quantum technology and blockchain technology, these frontier technologies have shown unique advantages in the field of information security. Quantum dots, as a new type of nanomaterial, have unique optical and electrical properties, which can be used to construct physical identifiers that are difficult to replicate; and the decentralized and tamper-proof characteristics of blockchain technology provide new possibilities for the secure storage of anti-counterfeiting information.

[0004] However, there is currently a lack of a comprehensive anti-counterfeiting solution that integrates biological feature recognition, quantum coding and blockchain technology to cope with the increasingly complex challenges of product anti-counterfeiting. SUMMARY

[0005] The application provides a product anti-counterfeiting verification system based on biological information, which integrates multi-modal biological feature recognition, quantum dot coding and blockchain technology to construct a multi-level dynamic anti-counterfeiting mechanism, in order to improve the uniqueness of anti-counterfeiting identifiers, enhance the security of the verification process, realize tamper-proof storage of anti-counterfeiting information, and establish adaptive verification criteria. This system effectively solves the problems of traditional anti-counterfeiting technologies, such as being easily replicated, lacking dynamicity and having security risks in the verification process, and provides comprehensive anti-counterfeiting protection for high-value products.

[0006] A product anti-counterfeiting verification system based on biological information, the system comprising: an information acquisition unit, a key generation unit, a quantum coding unit, a blockchain storage unit and a verification determination unit;

[0007] The information acquisition unit is used to acquire user biological features and product identification information;

[0008] The key generation unit is configured to generate a biological key according to the collected biological features, product identification information and environmental parameters;

[0009] The quantum encoding unit is configured to encode the generated biological key into a quantum dot physical identification;

[0010] The blockchain storage unit is configured to register a hash value of the quantum dot physical identification to a blockchain network;

[0011] The verification determination unit is connected with the information collection unit, the key generation unit and the blockchain storage unit respectively, configured to compare a verification hash value generated according to the current collected information with a hash value stored in the blockchain, and determine the authenticity of the product according to the similarity.

[0012] The biological features collected by the information collection unit include fingerprint features, facial features, voiceprint features and behavior features; the product identification information includes a product unique serial number, a production batch code, a manufacturing date and time stamp, a product physical characteristic parameter set, material composition information and a manufacturer digital signature; and the environmental parameters include geographic position coordinates, environmental temperature, humidity value, atmospheric pressure and time stamp at the time of collection.

[0013] The key generation unit generates a biological key K by using a topological entropy fusion algorithm model, and a mathematical expression of the topological entropy fusion algorithm model is:

[0014] wherein F(B, P, E) represents a fusion function of the biological feature parameter set B, the product identification information set P and the environmental parameter set E;

[0015] B={b1, b2,...b i ,...}, b i is the i th biological feature parameter, P={p1, p2,...p i ,...}, p i is the i th product identification information, E={e1, e2,...e i ,...}, e i is the i th environmental parameter; τ(B) is a topological mapping function of the biological features, w i is the weight of the i th biological feature parameter, and m is the total number of the biological feature parameters; γ(P) is a linear transformation of the product features, γ(P)=M·P+v, wherein M is a transformation matrix, and v is a bias vector; δ(E) is a standardization function of the environmental parameters, μ E and σ E are the mean and standard deviation of the environmental parameters, respectively;

[0016] Φ(x) = tanh(x), Ψ(x) = sigmoid(x); Ω(x) = ReLU(x), and represents a weighted sum operation, represents a modulation operation.

[0017] The key generation unit further comprises a time evolution subunit for implementing a time evolution mechanism of the biometric key;

[0018] The mathematical expression of the time evolution mechanism is: K(t+Δt) = G(K(t), λ, Δt), wherein K(t) is the biometric key value at time point t, K(t+Δt) is the biometric key value after a time interval Δt, G(K(t), λ, Δt) is an evolution function, G(K(t), λ, Δt) = K(t)·(1-λ·Δt)+λ·Δt·η(K(t)); η(K(t) is a perturbation function, η(K(t) = K(t)+∈·sin(K(t)), ∈ is a perturbation intensity parameter, and λ is an evolution rate parameter.

[0019] The quantum encoding unit maps the biometric key K to the physical identity T through a quantum dot encoding function Q(K), and the expression of the quantum dot encoding function Q(K) is:

[0020] Q(K) = {(x i ,y i ,z i ,c i )|i = 1, 2,..., r}; r is the total number of generated quantum dots, (x i ,y i ,z i ) represents the coordinates of the i-th quantum dot in three-dimensional space, (x i ,y i ,z i ) = (sin(πK 3i ), cos(πK 3i+1 ), K 3i+2 )·R i , R i is a scaling factor for adjusting the size of the i-th quantum dot; K 3i ,K 3i+1 ,K 3i+2 are the values of the corresponding index positions in the biometric key; c i is the chemical composition of the i-th quantum dot, c i = ξ(K[imodlen(K)], α i), ξ(k, a) is a quantum dot material selection function, ξ(k, a) = {(CdSe, 1-k), (ZnS, k a), (InP, k (1-a))}, a is a material proportion parameter, CdSe, ZnS, InP are different quantum dot material types; the total number of quantum dots satisfies r >= 2 n , n is a security parameter.

[0021] The quantum encoding unit further comprises an anti-tampering subunit, which triggers a preset degradation reaction by changing the chemical environment of the quantum dots when detecting a tampering attempt on the physical identifier, so as to invalidate the physical identifier.

[0022] The method for changing the chemical environment of the quantum dots to trigger a preset degradation reaction to invalidate the physical identifier comprises one or more of pH value change, external electric field action, photocatalysis, chemical reagent reaction and electromagnetic radiation.

[0023] The blockchain storage unit adopts a multi-level salting processing mechanism for hash value registration: H = H salt (K||T, {σ i |i = 1, 2,..., u}), H is the hash value finally stored in the blockchain, K||T is the concatenation of the biological key K and the physical identifier T, H salt (x, {σ i}) represents a cascade of salting hash functions, H salt (x, {σ i}) = H u (H u-1 (...H1(x, σ1)...), σ u ); H i (x, σ) = SHA-256(x σ) represents a single-level salting hash, SHA-256 is a secure hash algorithm; σ i = Γ(B, i, μ i ) represents the i-th biological salt; Γ(B, i, μ) is a biological salt derivation function, Γ(B, i, μ) = HMAC(B[i μ:(i+1) μ], key i ), B[i μ:(i+1) μ] is a specific part of the biological feature parameter set, key i is the key used by the i-th hash; u represents the number of salting processing levels.

[0024] The verification determination unit compares the hash values using a similarity function, which is represented as:

[0025] S(H, H') = (1 - w1) J(H, H') + w1 BL(PreH(H), PreH(H')), wherein S(H, H') is a similarity score, H' is a hash value generated by a current verification process, J(H, H') represents a Jaccard similarity coefficient, PreH(H) represents a prefix set of a hash value, k and m are respectively a prefix length parameter and a prefix interval parameter, H[i:i+k] is a substring of the hash value H from index i to i+k; BL(PreH(H), PreH(H')) represents a normalized reciprocal of an edit distance;

[0026] D(PreH(H), PreH(H')) is an edit distance between two prefix sets, max(|PreH(H)|, |PreH(H)|) is a maximum value of lengths of the two prefix sets, w1 is a weight coefficient for balancing influences of the Jaccard similarity coefficient and the edit distance, and 0.2 <= w1 <= 0.5.

[0027] The process of determining the product authenticity according to the similarity includes: when the similarity S(H, H') >= tau1, determining as a genuine product;

[0028] When tau2 <= S(H, H') < tau1, entering a secondary verification process, the secondary verification process including: extracting an additional biological feature B A , calculating an enhanced similarity S E (H, H') = S(H, H') * (1-gamma) + gamma * g(B A ), wherein g(B A ) is a credibility score function of the additional biological feature, and gamma is a fusion factor; when S E (H, H') >= tau1, determining as a genuine product, otherwise determining as a suspected fake product;

[0029] When S(H, H') < tau2, directly determining as a fake product;

[0030] tau1 and tau2 are preset determination thresholds of the system, and satisfy 0 < tau2 < tau1 < 1.

[0031] Compared with the prior art, the present application has the beneficial effects that:

[0032] The present application fuses biological feature recognition, quantum dot coding and blockchain storage technology, and constructs a multi-level and dynamic product anti-counterfeiting verification system. The system not only uses the uniqueness and time evolution mechanism of biological features to enhance the security of anti-counterfeiting, but also realizes difficult-to-copy anti-counterfeiting marks through quantum dot physical identification, and ensures the reliability of verification data by means of the tamper-proof property of the blockchain technology.

[0033] Compared with the traditional anti-counterfeiting technology, the system has higher security level, stronger anti-counterfeiting ability and more flexible verification mechanism, effectively solves the complex anti-counterfeiting demand faced by the field of high-value products, and provides strong technical support for enterprise intellectual property protection and consumer rights protection. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 A product anti-counterfeiting verification system based on biological information is shown. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme in the present application is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0036] As shown in Figure 1 A product anti-counterfeiting verification system based on biological information is shown. The connection relationship and data flow between the information acquisition unit, the key generation unit, the quantum encoding unit, the blockchain storage unit and the verification judgment unit are shown, and the overall architecture and workflow of the system are intuitively presented. The system comprises: an information acquisition unit, a key generation unit, a quantum encoding unit, a blockchain storage unit and a verification judgment unit.

[0037] The information acquisition unit is used to acquire user biological characteristics and product identification information. The biological characteristics acquired by the information acquisition unit include: fingerprint characteristics, facial characteristics, voiceprint characteristics and behavior characteristics. The product identification information includes: product unique serial number, production batch code, manufacturing date and time stamp, product physical characteristic parameter set, material composition information and manufacturer digital signature. The environmental parameters include: geographic position coordinates, environmental temperature, humidity value, atmospheric pressure and time stamp at the time of acquisition.

[0038] The information collection unit adopts a multi-modal biological feature collection device, including but not limited to: a capacitive or optical fingerprint collector for collecting fingerprints, with a resolution not less than 500 dpi; a high-definition infrared camera for collecting facial features, supporting 3D structured light technology; a high-sensitivity microphone array for collecting voiceprints, with a signal-to-noise ratio not less than 70 dB; and a touch screen and an acceleration sensor for recording behavioral characteristics. In actual application, the system can flexibly configure the types of biological features collected according to security level requirements. For example, ordinary consumer goods can only collect fingerprint features, while high-value luxury goods can collect multiple biological features at the same time to improve the security of anti-counterfeiting. Product identification information collection can be achieved by scanning a two-dimensional code, reading an RFID tag or an NFC tag, etc., and environmental parameters are automatically obtained by built-in various sensors.

[0039] Taking the fingerprint feature set as an example, the system uses Gabor filtering and direction field analysis to extract minutiae points (such as end points and bifurcation points), and records the coordinates, types and direction angles of each minutiae point. Usually, 60-80 minutiae point information is extracted from a single fingerprint. The facial feature set uses a deep learning model to extract a 128-dimensional facial feature vector, which contains key information such as facial contour and inter-feature distance ratio. The voiceprint feature set extracts the frequency spectrum features of the sound through the Mel frequency cepstrum coefficient (MFCC), and combines the dynamic time warping (DTW) algorithm to capture the time-varying characteristics of the sound. The behavioral feature set records the user's habitual behavior parameters such as input rhythm, sliding mode and pressing force. The fusion of these multi-modal biological features greatly enhances the anti-attack ability of the system. Even if the attacker obtains a copy of one kind of biological feature, it is difficult to fake other types of biological features at the same time.

[0040] The biological feature parameter set is represented as: B = {B FP ,B FE ,B VE ,B BH}, where B FP ,B FE ,B VE ,B BH represent the fingerprint feature set, the facial feature set, the voiceprint feature set and the behavioral feature set respectively. B FP = {f1, f2,..., f j} contains minutiae point coordinates and ridge direction; B FE = {a1, a2,..., a k} contains feature point distance and angle relationship between feature points; B VE = {v1, v2,..., v p} contains frequency spectrum features and pitch variation; B BH = {h1, h2,..., h qcomprises input modes and interaction habits.

[0041] The key generation unit is configured to generate a biological key according to the collected biological features, product identification information and environmental parameters; the key generation unit generates the biological key K by using a topological entropy fusion algorithm model, and a mathematical expression of the topological entropy fusion algorithm model is:

[0042] wherein F(B, P, E) represents a fusion function of the biological feature parameter set B, the product identification information set P and the environmental parameter set E;

[0043] B = {b1, b2,... b i ,...}, b i is the i th biological feature parameter, P = {p1, p2,... p i ,...}, p i is the i th product identification information, E = {e1, e2,... e i ,...}, e i is the i th environmental parameter; τ(B) is a topological mapping function of the biological feature, w i is the weight of the i th biological feature parameter, and m is the total number of the biological feature parameters; γ(P) is a linear transformation of the product feature, γ(P) = M · P + v, wherein M is a transformation matrix, and v is a bias vector; δ(E) is a standardization function of the environmental parameter, μ E and σ E are the mean and standard deviation of the environmental parameter, respectively;

[0044] Φ(x) = tanh(x), Ψ(x) = sigmoid(x); Ω(x) = ReLU(x), and represents a weighted sum operation, represents a modulation operation; the expression of the weighted sum operation is a ⊕ b = αa + (1-α)b, and α ∈ [0, 1]; the expression of the modulation operation is: β ∈ [0, 1]. The relative importance of the biological feature and the product information is controlled by the parameter α, and in high security level applications, the value of α is usually set to be between 0.6 and 0.7, so that the biological feature occupies a dominant position in the fusion process. The modulation operation controls the degree of modulation of the environmental factor on the biological key by the parameter β, and the value of β is usually set to be between 0.3 and 0.5, so as to balance the contribution of the environmental information and the stability of the key; compared with the traditional simple splicing or exclusive or operation, the fusion algorithm reduces the misrecognition rate of the biological feature by about 40%.

[0045] It should be noted that the biometric features have natural variability, for example, the angle and pressure during fingerprint collection will cause differences in the extracted feature points. The topological mapping function τ(B) reduces the influence of such variability by extracting the topological structure information of the biometric features. Specifically, the function calculates the gradient of the feature parameters The product feature linear transformation γ(P) and the environmental parameter standardization function δ(E) further enhance the uniqueness and environmental adaptability of the key generation.

[0046] The key generation unit also includes a time evolution subunit for implementing a time evolution mechanism for the biometric key;

[0047] The mathematical expression of the time evolution mechanism is: K(t+Δt) = G(K(t), λ, Δt), where K(t) is the biometric key value at time point t, K(t+Δt) is the biometric key value after a time interval Δt, G(K(t), λ, Δt) is the evolution function, G(K(t), λ, Δt) = K(t)·(1-λ·Δt)+λ·Δt·η(K(t)); η(K(t) is a perturbation function, η(K(t) = K(t)+∈·sin(K(t)), ∈ is a perturbation intensity parameter, and λ is an evolution rate parameter.

[0048] It should be noted that once the traditional static key is stolen, it will be permanently invalid; while the biometric key in the system will change predictably over time, even if the attacker obtains complete key information at a certain moment, these information will be invalid in the future. In the evolution function G(K(t), λ, Δt), the λ parameter controls the evolution rate, which is usually set between 0.01-0.05, so that the key changes significantly within a few hours to a few days; the nonlinear change (through the sin function) introduced by the perturbation function η(K(t)) makes the key evolution trajectory more difficult to predict, and the perturbation intensity parameter ε is usually set between 0.1-0.2, which ensures that the evolution trajectory is unpredictable, while also ensuring that the verification algorithm can effectively track the key changes. In practical applications, the system will dynamically adjust the λ and ε parameters according to the product type and security requirements, for example, a lower evolution rate can be set for daily consumer goods, while a higher evolution rate and perturbation intensity can be set for high-value goods.

[0049] The quantum encoding unit is configured to encode the generated biological key into quantum dot physical identification; the quantum encoding unit maps the biological key K into physical identification T through a quantum dot encoding function Q(K), and an expression of the quantum dot encoding function Q(K) is as follows:

[0050] Q(K) = {(x i ,y i ,z i ,c i )|i = 1, 2,..., r}; r is a total number of generated quantum dots, (x i ,y i ,z i ) represents coordinates of the i-th quantum dot in a three-dimensional space, (x i ,y i ,z i ) = (sin(πK 3i ), cos(πK 3i+1 ), K 3i+2 )·R i , R i is a scaling factor for adjusting a size of the i-th quantum dot; K 3i , K 3i+1 , K 3i+2 are values of corresponding index positions in the biological key; c i is a chemical composition of the i-th quantum dot, c i = ξ(K[imodlen(K)], α i ), ξ(k, α) is a quantum dot material selection function, ξ(k, α) = {(CdSe, 1-k), (ZnS, k·α), (InP, k·(1-α))}, α is a material proportion parameter, CdSe, ZnS, InP are different quantum dot material types; the total number of quantum dots satisfies r≥2 n , n is a security parameter.

[0051] The quantum dot encoding function Q(K) realizes conversion from an abstract biological key to a specific physical identification, and creates a difficult-to-replicate anti-counterfeiting mark; the quantum dot is a nanoscale semiconductor material, has unique optical and electrical properties, and its light-emitting wavelength can be accurately controlled by adjusting the size and material composition.

[0052] In the system, the biological key K determines the three-dimensional spatial distribution and material composition of the quantum dots, forming a unique quantum fingerprint. The three-dimensional coordinate calculation formula maps the key value to the three-dimensional space using trigonometric functions, creating a complex spatial structure; the material selection function ξ(k, α) dynamically adjusts the proportion of different quantum dot materials such as CdSe, ZnS, and InP according to the key value, further enhancing the complexity of the physical identifier; these quantum dots can be precisely deposited on the surface or embedded inside the product through microfluidic technology, forming an anti-counterfeiting mark that is invisible to the naked eye but presents a unique fluorescence pattern under specific excitation light.

[0053] The total number of quantum dots r is usually set to an exponential of the security parameter n, with n taking a value of 8-12 for ordinary consumer goods, generating 256-4096 quantum dots; for high-value products, n can take a value of 16 or more, generating more than 65536 quantum dots, making the difficulty of counterfeiting grow exponentially.

[0054] The quantum encoding unit also includes a tamper-proofing subunit that triggers a pre-set degradation reaction by changing the chemical environment of the quantum dots when a tampering attempt is detected, rendering the physical identifier invalid.

[0055] The method of changing the chemical environment of the quantum dots to trigger a pre-set degradation reaction to render the physical identifier invalid includes one or more of pH value change, external electric field action, photocatalysis, chemical reagent reaction, and electromagnetic radiation.

[0056] It should be noted that the quantum dots can be embedded in specially designed microcapsules that have properties sensitive to specific external stimuli. For example, pH-sensitive microcapsules can rupture when detecting acidic or basic solutions (commonly used to remove or transfer anti-counterfeiting labels), releasing chemicals that can react with the quantum dots, causing permanent changes in their fluorescence properties; the external electric field action mechanism can detect attempts to modify the distribution of quantum dots through electrolysis or electroplating, and trigger an oxidation reaction of the quantum dots; the photocatalysis is sensitive to strong light irradiation (commonly used to steal quantum dot pattern information), and when detecting light intensity exceeding a threshold, it starts a reaction between photosensitive substances and quantum dots, changing their chemical structure.

[0057] These self-destruction mechanisms form an additional physical line of defense, making it difficult to completely copy or transfer the quantum dot identifier without triggering the self-destruction mechanism.

[0058] The blockchain storage unit is used to register the hash value of the quantum dot physical identifier to the blockchain network; the blockchain storage unit uses a multi-level salting processing mechanism for hash value registration: H = H salt (K||T, {σ ii = 1, 2,..., u), H is the hash value stored in the blockchain finally, K||T is the concatenation of the biometric key K and the physical identity T, H salt (x, {σ i}) represents the salted hash function cascade, H salt (x, {σ i}) = H u (H u-1 (...H1(x, σ1)...), σ u ); H i (x, σ) = SHA-256(x σ) represents a single-level salted hash, SHA-256 is a secure hash algorithm; σ i = Γ(B, i, μ i ) represents the i-th level of biological salt; Γ(B, i, μ) is a biological salt derivation function, Γ(B, i, μ) = HMAC(B[i μ:(i+1) μ], key i ), B[i μ:(i+1) is a specific part of the biometric feature parameter set, and key i is the key used by the i-th level hash; u represents the number of salted levels.

[0059] The multi-level salted processing mechanism solves the two core problems faced by traditional hash storage: resistance to quantum computing attacks and adaptability to biometric feature variations.

[0060] Traditional single-level hash may become vulnerable in the face of the development of quantum computing technology, while the multi-level hash cascade used in the system can significantly increase the complexity of quantum computing attacks. Each level of hash is processed using a salt value σ i = Γ(B, i, μ i ) derived from the biometric feature, which itself has the uniqueness and variability of the biometric feature, making the final stored hash value both highly secure and tolerant to subtle changes in the biometric feature.

[0061] The system usually uses 3-5 levels of salted processing (u = 3-5), and for higher security level applications, it can be increased to more than 7 levels. The blockchain network uses a consortium chain structure, which is jointly maintained by product manufacturers, sellers and third-party certification agencies to ensure that the stored hash value is not tampered with. The smart contract mechanism is used to automatically execute the verification logic, and when a new verification request is submitted, the smart contract automatically compares the currently generated hash value with the historical hash value stored in the blockchain, and determines the authenticity of the product according to the similarity.

[0062] The verification determination unit is connected to the information collection unit, key generation unit and blockchain storage unit respectively, and is used to compare the verification hash value generated by the currently collected information with the hash value stored in the blockchain, and determine the authenticity of the product according to the similarity.

[0063] The verification decision unit compares the hash values using a similarity function, which is expressed as:

[0064] S(H, H') = (1 - w1) · J(H, H') + w1 · BL(PreH(H), PreH(H')), where S(H, H') is the similarity score, H' is the hash value generated in the current verification process, J(H, H') represents the Jaccard similarity coefficient, PreH(H) represents the prefix set of the hash value, k and m are the prefix length parameter and the prefix interval parameter, respectively, H[i:i+k] is the substring of hash value H from index i to i+k; BL(PreH(H), PreH(H')) represents the normalized reciprocal of the edit distance;

[0065] D(PreH(H), PreH(H')) is the edit distance between the two prefix sets, max(|PreH(H)|, |PreH(H')|) is the maximum of the lengths of the two prefix sets, w1 is a weight coefficient for balancing the influence of the Jaccard similarity coefficient and the edit distance, and 0.2 ≤ w1 ≤ 0.5.

[0066] In particular, when the weight coefficient w1 is set between 0.3 and 0.4, it is verified through experiments that this configuration allows the system to maintain a low false rejection rate (FAR < 0.01%) while also maintaining a low false acceptance rate (FRR < 0.1%).

[0067] The prefix length parameter k and the prefix interval parameter m have a significant impact on the performance of the system. Experiments show that when k is set to 5%-10% of the total length of the hash value and m is set to 2-3 times k, the system performance is optimal. In addition, the system also dynamically adjusts the w1, k, and m parameters according to the product type and application scenario. For example, for critical items such as medicines, the system will tend to stricter verification standards (smaller w1 values), while for daily consumer goods, more lenient verification standards (larger w1 values) can be used.

[0068] The process of determining the authenticity of a product based on the similarity includes: when the similarity S(H, H') ≥ τ1, it is determined to be a genuine product;

[0069] When τ2 ≤ S(H, H') < τ1, a secondary verification process is entered, which includes: extracting an additional biological feature B A , calculating an enhanced similarity S E (H, H') = S(H, H') · (1 - γ) + γ · g(B1), where g(B A ) is a credibility scoring function of the additional biological feature, and γ is a fusion factor; when S EWhen (H, H') >= tau1, it is determined to be a genuine product, otherwise it is determined to be a suspected fake product.

[0070] When S(H, H') < tau2, it is directly determined to be a fake product.

[0071] Tau1 and tau2 are system preset determination thresholds, satisfying 0 < tau2 < tau1 < 1.

[0072] The additional biological characteristics include additional fingerprints, iris features or more complete facial images; the credibility score function of the additional biological characteristics uses a deep learning model to evaluate the quality and reliability of the features, and outputs a score between 0 and 1.

[0073] The fusion factor gamma is set between 0.3 and 0.5, controlling the weight of the additional biological characteristics in the enhanced similarity calculation, and in the case of poor environment or abnormal user state (such as injured fingers, hoarse voice, etc.), the overall misjudgment rate of the system is reduced by about 75%.

[0074] The setting of the determination thresholds tau1 and tau2 directly affects the security and user experience of the system, and in general, tau1 is set between 0.85 and 0.95, and tau2 is set between 0.6 and 0.7. For high-value products, the system will automatically increase the determination threshold to increase the verification strictness; while for low-value daily necessities, the threshold can be appropriately reduced to improve the verification pass rate and user experience.

[0075] To further illustrate the practical application effect of the present application, the working process, parameter setting and technical effect of the system are described in detail through a case of high-end watch anti-counterfeiting authentication: a certain high-end luxury watch brand applies the system in its limited edition series of products to solve the problem of rampant fake products in the market. Each limited edition watch is initialized and entered with anti-counterfeiting information before leaving the factory, and consumers can verify the authenticity through a special mobile application after purchase.

[0076] For high-value luxury goods, the system adopts the following configuration parameters: the information acquisition unit is configured to collect biological features including fingerprints (resolution 800 dpi), faces (3D structured light, resolution 1080p), and voiceprints (sampling rate 48 kHz); product information acquisition relies on the NFC chip (containing serial number, manufacturing batch, and production date) etched on the watch bottom cover; environmental parameter acquisition records GPS positioning (accuracy ± 5 m), environmental temperature (accuracy ± 0.5 °C), and environmental illumination (accuracy ± 50 lux). In terms of key generation unit configuration, the weight distribution of biological features is fingerprint features (w = 0.5), face features (w = 0.3), and voiceprint features (w = 0.2); the time evolution parameters are λ = 0.03 (the key changes significantly after about 72 hours) and ε = 0.15; the fusion algorithm parameters are α = 0.65 (biological features dominate in fusion) and β = 0.4 (environmental factors are moderately modulated). The quantum encoding unit configuration uses CdSe / ZnS core-shell structure quantum dots, with a quantum dot quantity of n = 16, r = 65536 quantum dots, a quantum dot size range of 2-8 nm, and a tamper-proof mechanism of photocatalytic degradation (triggered by strong light irradiation > 10000 lux). The blockchain storage unit configuration uses a consortium chain (maintained by watch manufacturers, authorized distributors, and third-party certification agencies), with a salt processing series of u = 5 and a block confirmation number of 6. The verification judgment unit configuration similarity calculation parameters are w1 = 0.35, k = 12, and m = 30, the judgment threshold values are τ1 = 0.92 and τ2 = 0.75, and the fusion factor is γ = 0.4. The reading of quantum dot physical identification uses non-invasive optical excitation technology. Under the irradiation of specific wavelength (usually ultraviolet light region 365-405 nm) excitation light, quantum dots will emit fluorescence of specific wavelength. The system can read the distribution pattern and fluorescence characteristics of internal quantum dots without damaging the appearance and structure of the product by equipping high-sensitivity optical sensors with specific optical filters.

[0077] The authorized retail store salesperson (registered as an authorized verification personnel) enters the personal biological information into the system, extracts 78 feature points through fingerprint recognition, extracts 128-dimensional feature vectors through facial recognition, and extracts 42 MFCC coefficients through voiceprint recognition. Then, the NFC reads the built-in chip in the watch to obtain the serial number LM21934567, production batch 2024Q1-086, manufacturing date 2024-02-15T09:30:42, and other information. The system automatically records the geographic location 31.2304°N, 121.4737°E, the environmental temperature 23.4°C, the light condition 450 lux, the timestamp 2024-03-10T14:25:36, and other environmental data. Next, the system generates a 256-bit biological key K based on the collected biological features, product information, and environmental parameters through a topological entropy fusion algorithm. Subsequently, the system maps the biological key K to a three-dimensional space distribution composed of 65536 quantum dots, and uses microfluidic technology to encapsulate the corresponding quantum dot array in a specific location inside the watch bottom cover, forming a physical anti-counterfeiting mark. Finally, the system calculates the combined hash value of the quantum dot physical mark and the biological key, and after 5-level salting, the final hash value H is registered in the blockchain network with transaction ID 0x7a8b9c0d1e2f3g4h5i6j7k8l9m0n1o2p.

[0078] After the consumer purchases the watch, the verification request is initiated through the dedicated application. Next, the system guides the consumer to provide fingerprint, facial features, and voice samples, while reading the product information in the watch NFC chip and recording the current environmental parameters. Then, the system generates a verification key K' based on the newly collected information and generates the corresponding hash value H'. The system calculates the similarity between H' and the original hash value H stored in the blockchain, obtaining S(H, H') = 0.88, which is between τ2(0.75) and τ1(0.92), triggering a secondary verification process. The system prompts the consumer to provide additional iris features B_A, calculates the enhanced similarity S_E(H, H') = 0.88 × (1-0.4) + 0.4 × 0.98 = 0.92, which is equal to the decision threshold τ1. Finally, the system determines that the watch is genuine, and displays detailed product information, including the production date, warranty period, etc., on the application interface.

[0079] It establishes a complete anti-counterfeiting chain from sales to consumption, so even if someone obtains the same model product, they cannot pass the verification because each product is bound to the biological information of a specific salesperson; when the consumer verifies, the system core is to compare the authenticity of the product information, rather than requiring the consumer's biological information to completely match that of the salesperson; this mechanism ensures that each product has a unique identity, and the verification process relies more on the product's own information and physical mark. The system allows a certain degree of difference (the user's biological information is different from the initial registered personnel) through the similarity algorithm, but can still accurately determine the authenticity of the product.

[0080] The system adopts a double-layer verification mechanism, binds the product with the biological information of the authorized personnel (such as a salesperson) registered initially to form a basis to generate a quantum dot physical identifier. However, in the consumer verification link, the system mainly verifies the integrity of the product physical identifier and the authenticity of the product information, rather than requiring the consumer to provide biological information that completely matches the initial registration personnel. The biological information of the consumer is mainly used to establish a secure channel for the verification session and prevent automated fraud attacks, and at the same time serves as an optional additional verification factor. The system can accurately track the key evolution trajectory through the product history verification records stored on the blockchain and the time evolution parameters, ensuring the effectiveness of the verification even if the consumer verifies a long time after purchase. To solve the compatibility problem of biological key time evolution and long-term verification, the system not only stores the initial hash value in the blockchain, but also records the key evolution parameters and timestamp sequence. When the consumer verifies a long time after purchase, the system will calculate the theoretical key evolution trajectory according to the difference between the current time and the initial registration time, and compare it with the verification hash value generated at the current time.

[0081] The system has shown high security in this high-end watch anti-counterfeiting application. The physical identifier composed of 65536 quantum dots far exceeds the security of traditional anti-counterfeiting technology. In a 1000-hour continuous operation test, the system processed more than 100,000 verification requests without system crashes or data loss. The average blockchain confirmation time is 2.8 minutes, meeting the real-time verification requirements.

[0082] Through the implementation and testing of this actual application case, it is proved that the system of the present application has significant technical advantages and practical value in the field of high-value product anti-counterfeiting. Not only does it effectively solve the problems of easy copying and staticity faced by traditional anti-counterfeiting technology, but also establishes a new type of anti-counterfeiting verification system with high security, high reliability and good user experience through the fusion of multi-modal biological characteristics, quantum coding and blockchain storage. The system has begun to be popularized and applied in the field of high-value products such as luxury goods and pharmaceuticals, and has achieved significant economic and social benefits.

[0083] The above specific embodiments further detail the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and does not limit the scope of protection of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A bio-information-based product anti-counterfeiting verification system, characterized in that, The system comprises an information collection unit, a key generation unit, a quantum encoding unit, a blockchain storage unit and a verification and determination unit. The information collection unit is configured to collect user biological features and product identification information. The key generation unit is configured to generate a biometric key according to the collected biometric feature, product identification information, and environmental parameters; and the key generation unit is configured to generate the biometric key by using a topological entropy fusion algorithm model , and a mathematical expression of the topological entropy fusion algorithm model is wherein, denotes a fusion function of a set of biometric parameters , a set of product identification information and a set of environmental parameters ; , is the nth biological characteristic parameter, , is the nth product identification information, , is the nth environmental parameter; is a topological mapping function of the biological characteristics, , is the weight of the nth biological characteristic parameter, is the total number of biological characteristic parameters; is a linear transformation of product characteristics, wherein, is a transformation matrix, is a bias vector; is a standardization function of the environmental parameters, , and are the mean and standard deviation of the environmental parameters, respectively;​​​​ , ; , denotes a weighted sum operation, denotes a modulation operation; The key generation unit further comprises a time evolution subunit configured to implement a time evolution mechanism of the biological key. The mathematical expression of the time evolution mechanism is: wherein, is the biometric key value at time point , is the biometric key value after a time interval , is the evolution function, ; , , is the perturbation intensity parameter, is the evolution rate parameter; The quantum encoding unit is configured to encode the generated biological key into a quantum dot physical identification. The blockchain storage unit is configured to register a hash value of the quantum dot physical identification to a blockchain network. The verification and determination unit is connected to the information collection unit, the key generation unit and the blockchain storage unit respectively, configured to compare a verification hash value generated based on the current collected information with a hash value stored in the blockchain, and determine the authenticity of the product according to the similarity.

2. The system of claim 1, wherein, The biological features collected by the information collection unit include fingerprint features, facial features, voiceprint features and behavior features; the product identification information includes a product unique serial number, a production batch code, a manufacturing date and time stamp, a product physical characteristic parameter set, material composition information and a manufacturer digital signature; and the environmental parameters include geographic location coordinates, environmental temperature, humidity value, atmospheric pressure and a time stamp at the time of collection.

3. The system of claim 2, wherein, The quantum encoding unit encodes the biological key by a quantum dot encoding function mapping to a physical identity mapping to a physical identity The quantum dot encoding function is expressed as ; is the total number of quantum dots generated, represents the coordinate of the th quantum dot in three-dimensional space, , is a scaling factor that adjusts the size of the th quantum dot; , , is the value of the corresponding index position in the biological key; is the chemical composition of the th quantum dot, , is the quantum dot material selection function, , is the material ratio parameter, CdSe, ZnS, InP are different quantum dot material types; the total number of quantum dots satisfies , is a security parameter.

4. The system of claim 3, wherein, The quantum encoding unit further comprises a tamper-proofing subunit configured to trigger a preset degradation reaction by changing the chemical environment of the quantum dot to disable the physical identification when a tampering attempt is detected.

5. The system of claim 4, wherein, The method for changing the chemical environment of the quantum dot to trigger the preset degradation reaction to disable the physical identification includes one or more of pH value change, external electric field action, photocatalysis, chemical reagent reaction and electromagnetic radiation.

6. The system of claim 5, wherein, The blockchain storage unit adopts a multi-level salting processing mechanism for hash value registration: , is the hash value finally stored on the blockchain, is the biological key and the connection with the physical identification, represents the cascade of the salting hash function, ; ; represents a single-level salting hash, is a secure hash algorithm; represents the level biological salt; Functions derived from biological salts. , For a specific part of the set of biometric parameters, For the first The key used for level hashing; This indicates the number of stages of salting treatment.

7. The system of claim 6, wherein, The verification and determination unit compares the hash values by using a similarity function, and the similarity function is represented as: ; is a similarity score, is a hash value generated for the current verification process, , ; denotes a set of prefixes of the hash value, , and are a prefix length parameter and a prefix interval parameter, respectively, is a hash value is a substring of from index to ; denotes a normalized inverse of the edit distance; ; is the edit distance between two prefix sets, is the maximum of the lengths of the two prefix sets, is a weight coefficient to balance the influence of the Jaccard similarity coefficient and the edit distance, .

8. The system of claim 7, wherein, The process of determining the authenticity of the product according to the similarity includes: when the similarity is determined as a genuine product; When , the secondary verification procedure is entered, which comprises extracting an additional biometric feature , calculating an enhanced similarity , wherein is a trustworthiness score function of the additional biometric feature, is a fusion factor; when , the product is determined to be authentic, otherwise it is determined to be a suspected counterfeit. When directly determined to be a fake; and a decision threshold preset for the system, which is satisfied .

Images