Palm eye pupil payment charging pile without payment tool for payment

Abstract

The invention discloses a palm eye pupil payment charging pile without payment tool payment, which comprises a biological characteristic acquisition module, a data processing unit, a payment execution module, a charging control module, a charging pile shell, an operation area and a charging pile control cabinet, the palm print recognition device comprises a palm print recognition device, a palm vein recognition device and an iris recognition device, and the palm print recognition device adopts a 1200 dpi CMOS image sensor and is vertically installed in the center of an operation area; the palm vein recognition device is provided with a near-infrared light source array with the wavelength of 850nm + / -10nm. Therefore, card-free and code-scanning-free payment is realized, through palm print, palm vein and iris verification, the problem of payment medium dependence is solved, operation is simplified, efficiency is improved, triple verification is resistant to counterfeiting, dynamic weight adapts to the environment, block chain encryption prevents data leakage, double-factor verification prevents fraud, the charging process is monitored in real time, and whole-process protection is realized.

Description

Technical Field

[0001] This application relates to the technical field of electric vehicle charging, and particularly to a palm-eye-pupil payment charging pile that does not require payment with a payment tool. Background Art

[0002] The current payment system for electric vehicle charging piles mainly relies on the following technical solutions, which have significant defects:

[0003] 1. Dependence on physical cards / mobile phone payments: Traditional charging piles use physical IC cards or mobile phone scanning codes for payment, which have problems such as card loss, dependence on mobile phone signals (e.g., weak signals in underground parking lots), and cumbersome payment processes (requiring multiple operations).

[0004] 2. Risks of single biometric recognition: Some charging piles attempt to use fingerprint or face recognition technology, but single biometric features are vulnerable to forgery attacks (such as 3D printed fingerprints, photo spoofing), and have poor environmental adaptability (strong light / low temperature affects recognition).

[0005] 3. Hidden dangers of data storage security: Existing systems mostly use centralized databases to store user information, which has the risk of data leakage.

[0006] 4. Separation of identity verification and payment: Existing solutions require identity verification to be completed first before payment can be executed, which takes time (average ≥ 5 seconds), and there is a lack of continuous identity verification during the charging process. Summary of the Invention

[0007] This application aims to solve at least one of the technical problems in the related technologies to some extent.

[0008] To this end, the first object of this application is to provide a palm-eye-pupil payment charging pile that does not require payment with a payment tool. Without using a physical card or scanning code, charging can be directly completed through palmprint, palm vein, and iris verification, solving the problem of payment failure in scenarios such as card loss, dead mobile phone / weak signal. With FPGA parallel processing + multi-modal fusion, the efficiency is improved compared to traditional scanning code payment, and the operation process is simplified to two steps: "reach out for verification → insert the gun to charge".

[0009] The second object of this application is to provide a palm-eye-pupil payment charging pile that does not require payment with a payment tool. Triple verification of palmprint (surface), palm vein (deep blood vessels), and iris (living body dynamics) can resist attacks such as 3D printing and photo spoofing, and the false recognition rate is significantly reduced compared to traditional solutions. The dynamic weight algorithm automatically adjusts the feature weights according to light and distance (such as reducing the iris weight under strong light), improving the recognition accuracy in complex scenarios, and solving the recognition problems in environments such as low temperature and strong light.

[0010] The third object of the present application is to provide a palm-eye-pupil payment charging pile that does not require payment with a payment tool. Biometric features are hashed through the national secret SM4 algorithm, and the data is encrypted by AES-256 and stored locally. After the network is restored, it is fragmented and uploaded to the blockchain node to eliminate the risk of leakage of the centralized database. The biometric features and the GPS location of the charging pile are synchronously verified. When the positioning deviation exceeds 1 km, the payment is refused to prevent the charging pile from being illegally moved and stolen.

[0011] The fourth object of the present application is to provide a palm-eye-pupil payment charging pile that does not require payment with a payment tool. The charging gun grip sensor continuously detects the palm pressure distribution and blood flow signal. If the hand is removed for more than 5 seconds or there is no physiological signal, the charging is interrupted to prevent "remote charging theft after unauthorized start". When the recognition fails three times in a row, the start is made without contact, or the positioning is abnormal, an alarm is triggered to cover the risks in the whole process before, during, and after charging.

[0012] To achieve the above object, the first aspect embodiment of the present application proposes a palm-eye-pupil payment charging pile that does not require payment with a payment tool, including a biometric feature acquisition module, a data processing unit, a payment execution module, a charging control module, a charging pile housing, an operation area, and a charging pile control cabinet. Among them, the biometric feature acquisition module is fixedly installed above the operation area on the front of the charging pile housing and includes a palmprint recognition device, a palm vein recognition device, and an iris recognition device. Among them, the palmprint recognition device uses a 1200 dpi CMOS image sensor and is vertically installed at the center of the operation area; the palm vein recognition device is configured with a near-infrared light source array with a wavelength of 850 nm ± 10 nm and is horizontally installed 20 mm below the palmprint recognition device; the iris recognition device includes a binocular camera with a live detection function and is installed at a 30° tilt at the top edge of the operation area; the data processing unit is located inside the charging pile control cabinet and is signal-connected to the biometric feature acquisition module through a shielded cable, and includes a feature extraction module and a dynamic fusion module. Among them, the feature extraction module uses an FPGA chip to parallel-process palmprint, palm vein, and iris features; the dynamic fusion module implements a three-level feature fusion algorithm and outputs a comprehensive biometric code; the payment execution module is communicatively connected to the data processing unit through a CAN bus and includes a blockchain encryption unit and a two-factor verification unit. Among them, the blockchain encryption unit uses the national secret SM4 algorithm to generate a biometric template hash value; the two-factor verification unit synchronously verifies the biometric code and the GPS positioning data of the charging pile; the charging control module is electrically connected to the payment execution module and includes a charging gun and a real-time monitoring module, and the real-time monitoring module continuously detects the user's palm contact state during the charging process.

[0013] The palm-eye-pupil payment charging pile in the embodiment of the present application that does not require payment with a payment tool realizes cardless and code-scanning-free payment. Through palmprint, palm vein, and iris verification, it solves the problem of relying on payment media, simplifies operations, improves efficiency, has triple verification to resist forgery, dynamic weights to adapt to the environment, blockchain encryption to prevent data leakage, two-factor verification to prevent fraud, real-time monitoring during the charging process, and full-process protection.

[0014] In addition, the palm-eye-pupil payment charging pile that does not require payment with a payment tool proposed above according to the present application may further have the following additional technical features:

[0015] In an embodiment of the present application, the near-infrared light source array of the palm vein recognition device includes 8 groups of symmetrically distributed LED light sources, with the power of each group of light sources being 3W ± 0.2W, and the included angle between the light source and the image sensor being 45° ± 5°.

[0016] In an embodiment of the present application, the dynamic fusion module implements an improved DS evidence theory algorithm, and sets the dynamic weight coefficients α, β, γ to satisfy:

[0017] α (palm vein weight) = 0.5 × e^(-0.1t) + 0.3 × I (ambient light intensity);

[0018] β (palmprint weight) = 1 - α - γ;

[0019] γ (iris weight) = 0.2 × D (user distance).

[0020] In an embodiment of the present application, the blockchain encryption unit includes a local encrypted storage subunit and a distributed ledger update subunit. Among them, the local encrypted storage subunit stores the transaction data of the recent 7 days using the AES-256 algorithm; when the network resumes, the distributed ledger update subunit uploads the encrypted data slices to at least 3 blockchain nodes.

[0021] In an embodiment of the present application, the real-time monitoring module includes a micro-piezoelectric sensor array and a near-infrared spectrometer. Among them, the micro-piezoelectric sensor array is embedded on the surface of the charging gun grip to detect the palm contact pressure distribution; the near-infrared spectrometer collects the palm vein blood flow pulse signal every 5 seconds.

[0022] In an embodiment of the present application, the live detection function of the iris recognition device includes a dynamic iris texture analysis unit and an eyelid micro-motion monitoring unit. Among them, the dynamic iris texture analysis unit detects that the change rate of the pupil diameter ≥ 0.5mm / s; the eyelid micro-motion monitoring unit captures the blinking frequency through a 30fps high-speed camera.

[0023] In one embodiment of the present application, the system further includes an exception handling module, which triggers an alarm when any of the following situations is detected: the biometric matching degree is continuously lower than 98% for three times; no palm contact is detected within 5 seconds after the charging gun is connected to the vehicle; the deviation between the GPS positioning data and the registered address of the charging pile exceeds 1 kilometer.

[0024] In one embodiment of the present application, the three-level feature fusion algorithm includes:

[0025] The first-level fusion: extracting the topological structure features of the bifurcation points and end points of the palmar vein blood vessels;

[0026] The second-level fusion: calculating the direction field consistency coefficient of the palmprint ridge lines;

[0027] The third-level fusion: matching the polar coordinate distribution features of the iris crypts and iris folds.

[0028] The advantages of the present application compared with the existing technologies are as follows:

[0029] (1) There is no need for a physical card or scanning a code. Charging can be directly completed through palmprint, palmar vein, and iris verification, solving the problem of payment failure in scenarios such as card loss, dead mobile phone / weak signal. FPGA parallel processing + multi-modal fusion improves the efficiency compared with traditional code scanning payment, and the operation process is simplified to two steps: "reach out for verification → insert the gun to charge".

[0030] (2) Triple verification of palmprint (surface), palmar vein (deep blood vessels), and iris (living body dynamics) resists attacks such as 3D printing and photo deception. The false recognition rate is significantly reduced compared with traditional solutions. The dynamic weight algorithm automatically adjusts the feature weights according to light and distance (such as reducing the iris weight under strong light), improving the recognition accuracy in complex scenarios and solving the recognition problems in environments such as low temperature and strong light.

[0031] (3) Biometric features are hashed by the national secret SM4 algorithm, and the data is encrypted by AES-256 and stored locally. After the network is restored, it is sharded and uploaded to the blockchain node, eliminating the risk of leakage of the centralized database. The biometric features are synchronously verified with the GPS location of the charging pile, and payment is refused when the positioning deviation exceeds 1 kilometer, preventing theft after the charging pile is illegally moved.

[0032] (4) The charging gun grip sensor continuously detects the palm pressure distribution and blood flow signal. Charging is interrupted if the hand is removed for more than 5 seconds or there is no physiological signal, eliminating "remote theft charging after unauthorized start". An alarm is triggered when there are three consecutive recognition failures, contactless start, or positioning anomalies, covering the risks in the whole process before, during, and after authentication.

[0033] Additional aspects and advantages of the present application will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present application. Description of the Drawings

[0034] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of embodiments in conjunction with the accompanying drawings, where:

[0035] Figure 1 Stereogram of a palm-eye-pupil payment charging pile that does not require payment with a payment tool according to an embodiment of the present application;

[0036] Figure 2 Stereogram of a palm-eye-pupil payment charging pile that does not require payment with a payment tool according to another embodiment of the present application;

[0037] Figure 3 Schematic diagram of the connection relationship between modules of a palm-eye-pupil payment charging pile that does not require payment with a payment tool according to an embodiment of the present application;

[0038] Figure 4 Schematic diagram of the control connection between modules of a palm-eye-pupil payment charging pile that does not require payment with a payment tool according to an embodiment of the present application;

[0039] Figure 5 Schematic diagram of the physical installation positions of devices of a palm-eye-pupil payment charging pile that does not require payment with a payment tool according to an embodiment of the present application;

[0040] Figure 6 Flowchart of a three-level feature fusion algorithm for a palm-eye-pupil payment charging pile that does not require payment with a payment tool according to an embodiment of the present application.

[0041] As shown in the figure: 1. Biometric acquisition module; 2. Data processing unit; 3. Payment execution module; 4. Charging control module; 5. Charging pile housing; 6. Operation area; 7. Charging pile control cabinet; 8. Exception handling module; 11. Palmprint recognition device; 12. Palm vein recognition device; 13. Iris recognition device; 21. Feature extraction module; 22. Dynamic fusion module; 31. Blockchain encryption unit; 32. Two-factor verification unit; 41. Charging gun; 42. Real-time monitoring module; 121. Near-infrared light source array; 131. Binocular camera; 311. Local encrypted storage subunit; 312. Distributed ledger update subunit; 421. Micro-piezoelectric sensor array; 422. Near-infrared spectrometer; 1311. Dynamic iris texture analysis unit; 1312. Eyelid micro-motion monitoring unit; 221. Palm vein vascular bifurcation point; 222. Palm vein vascular endpoint; 223. Palmprint ridge line; 224. Iris crypt; 225. Iris fold. Detailed implementation manners

[0042] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where like or similar reference numerals denote like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as limiting the present application. On the contrary, the embodiments of the present application include all variations, modifications, and equivalents falling within the spirit and scope of the appended claims.

[0043] The palm-eye-pupil payment charging pile of the embodiments of the present application that does not require payment with a payment tool will be described below with reference to the accompanying drawings.

[0044] As Figures 1 - 6 shown, the palm-eye-pupil payment charging pile of the embodiments of the present application that does not require payment with a payment tool may include a biometric collection module 1, a data processing unit 2, a payment execution module 3, a charging control module 4, a charging pile housing 5, an operation area 6, and a charging pile control cabinet 7.

[0045] It can be understood that the operation area 6 is provided on the front of the charging pile housing 5, and the biometric collection module 1 is fixedly installed above it; the data processing unit 2 is accommodated inside through the charging pile control cabinet 7 to form a hierarchical architecture of "front-end collection - back-end processing".

[0046] Palmprint recognition device 11: An IMOS image sensor with 1200 dpi is used, which is vertically aligned with the center of the operation area 6 to ensure that when the user's palm is placed naturally, the sensor can clearly collect palmprint texture images with a resolution of more than 500 ppi.

[0047] Palm vein recognition device 12: It is horizontally installed 20 mm directly below the palmprint recognition device 11. Its near-infrared light source array 121 emits light beams with a wavelength of 850 nm ± 10 nm, penetrates the palm epidermis to illuminate the vein blood vessels, and cooperates with the image sensor to collect grayscale images of the palm vein blood vessels at a 45° angle.

[0048] Iris recognition device 13: A binocular camera 131 with a live detection function is fixed at a 30° inclination on the top edge of the operation area 6, and the optical axis of the lens is aligned with the iris area when the user is looking straight ahead naturally, ensuring that the collection range covers the pupil, iris texture, and eyelid dynamics.

[0049] Biometric collection and preprocessing process:

[0050] The user stands in front of the charging pile and places the palm at the center of the operation area 6:

[0051] The palmprint recognition device 11 obtains palmprint images in real time;

[0052] The palm vein recognition device 12 projects a uniform light beam through the near-infrared light source array 121 to trigger the formation of a high-contrast blood vessel network image of the palm vein blood vessels on the image sensor;

[0053] The binocular camera 131 of the iris recognition device 13 synchronously collects binocular iris images and synchronously starts live detection (such as pupil scaling and blink action monitoring).

[0054] The three types of biometric raw data are transmitted to the data processing unit 2 through a shielded cable. The metal braided layer of the shielded cable effectively isolates electromagnetic interference to ensure data integrity.

[0055] Payment verification and charging control process:

[0056] (1) Payment execution module 3

[0057] Blockchain encryption unit 31: Uses the national cryptography SM4 algorithm to generate a hash value for the comprehensive biometric code to ensure that the biometric template cannot be reversely restored;

[0058] Two-factor verification unit 32: Synchronously verifies the biometric code and the positioning data of the built-in GPS module of the charging pile, and triggers the charging permission only when the two match.

[0059] (2) Charging control module 4

[0060] Charging gun 41: After two-factor verification is passed, the user inserts it into the charging port of the electric vehicle;

[0061] Real-time monitoring module 42: Continuously monitors the palm contact state through the built-in sensor of the charging gun grip:

[0062] The micro-pressure sensor array detects the pressure distribution of the grip to ensure continuous contact of the user's palm;

[0063] The near-infrared spectroscopy analyzer collects the palm vein blood flow pulse signal every 5 seconds to verify the liveness authenticity of the biometric.

[0064] In this embodiment, through the optimization of the hardware layout of the biometric acquisition module 1 (such as the 45° angle of the palm vein light source and the 30° inclination of the iris camera), high-quality acquisition of multi-modal data is ensured; the FPGA parallel processing and three-stage fusion algorithm of the data processing unit 2 achieve millisecond-level feature extraction; the two-factor verification and blockchain encryption of the payment execution module 3 ensure transaction security; the real-time monitoring mechanism of the charging control module 4 realizes continuous identity verification during the charging process. Each module forms a closed-loop control through a shielded cable, CAN bus, and electrical connection to implement the core technical solution of the cardless payment charging pile.

[0065] It should be noted that the binocular camera 131 integrates an automatic aperture adjustment module, which dynamically adjusts the exposure time according to the ambient light intensity I (unit: lux):

[0066] $$

[0067] t_{exp}=

[0068] \begin{cases}

[0069] \displaystyle\frac{1}{60}\,\text{s}&I\geq 1000\,\text{lux}\\

[0070] \displaystyle\frac{1}{30}\cdot\ln\left(\frac{1000}{I}\right)\,\text{s}&100\,\text{lux}<I<1000\,\text{lux}\\

[0071] \displaystyle\frac{1}{30}\,\text{s}&I\leq 100\,\text{lux}

[0072] \end{cases}

[0073] $$

[0074] Ensure that the signal-to-noise ratio of the iris image is ≥ 40 dB.

[0075] In an embodiment of the present application, as Figures 1 - 6 shown, the hardware design and working process of the palm vein recognition device 12 and its near-infrared light source array 121 are described in detail:

[0076] It can be understood that the palm vein recognition device 12 is horizontally installed 20 mm directly below the palmprint recognition device 11, and the two form a hierarchical acquisition structure in the vertical direction:

[0077] The palmprint recognition device 11 is located at the center of the operation area 6 and is used to collect the texture on the palm surface;

[0078] The palm vein recognition device 12 is horizontally arranged to ensure that when the user's palm is placed naturally, the palm center is facing the detection areas of the light source array 121 and the image sensor.

[0079] Core parameters of the near-infrared light source array 121:

[0080] It includes 8 groups of symmetrically distributed LED light sources, which are arranged in a circular / linear symmetry centered on the palm vein image sensor. The power of each group of light sources is strictly controlled within 3W ± 0.2W to ensure uniform and stable light emission intensity;

[0081] Optical angle: The emission direction of each light source forms an angle of 45°±5° with the optical axis of the image sensor. This angle has been optimized through optical simulation to ensure that near-infrared light effectively penetrates the palm epidermis (approximately 0.5-2mm) while avoiding image noise caused by direct light reflecting off the skin surface.

[0082] Palm vein feature collection workflow:

[0083] (1) Light source projection and biometric imaging

[0084] When the user's palm is placed on the operation area 6, the near-infrared light source array 121 of the palm vein recognition device 12 emits near-infrared light with a wavelength of 850nm±10nm (this wavelength has strong penetration into human tissue and a high hemoglobin absorption coefficient, which can highlight the contrast between the veins and the surrounding tissues);

[0085] Eight groups of symmetrical light sources form a uniform near-infrared light field. The light enters the palm from a 45° direction. After passing through the epidermis and dermis, the veins absorb the near-infrared light due to their rich hemoglobin, forming a dark-striped image of the vascular structure on the image sensor (the non-vascular area reflects light more strongly and appears as a bright area).

[0086] (2) Image sensor data acquisition

[0087] The image sensor and the light source array 121 are arranged at an angle of 45°, receive scattered light after being transmitted through the palm, convert the optical signal into an electrical signal, and generate a palm vein grayscale image (resolution ≥ 500dpi);

[0088] The symmetrically distributed light source design eliminates the shadow area caused by a single light source, ensuring that the veins at the edge of the palm (such as the veins at the base of the fingers) can also be clearly imaged.

[0089] (3) Signal transmission and anti-interference

[0090] The collected palm vein image data is transmitted to the feature extraction module 21 of the data processing unit 2 via a shielded cable. The metal shielding layer of the shielded cable effectively isolates the electromagnetic interference inside the charging pile (such as high-frequency noise of the charging module) to ensure the integrity of the image data.

[0091] It should be noted that the near-infrared light source array 121 has a built-in temperature sensor and a PID control circuit. When the ambient temperature is below 0°C, the light source power is adjusted according to ΔP = 0.05 × (25-T) 2 (T is Celsius temperature) to ensure that the luminous intensity is stable within the range of 2.8-3.2W at low temperatures.

[0092] It should be noted that the light source power of each group of 3W±0.2W described in this embodiment provides sufficient light intensity to ensure that near-infrared light penetrates the palm of different thicknesses (the average thickness of an adult palm is about 15-25mm). Even for users with darker skin tones (melanin has a weak absorption of near-infrared light), the blood vessel contours can still be clearly displayed;

[0093] By adjusting the duty cycle of the light source drive circuit, while ensuring power stability, the energy consumption is reduced (the average power consumption in the pulse drive mode ≤10W).

[0094] The light source and the image sensor form a 45° angle, which is the best transmission angle that conforms to the Lambert-Beer law:

[0095] Avoid surface reflection (specular reflection) caused by direct 0° irradiation;

[0096] Compared with 90° side irradiation, the scattering loss of light in the epidermis layer is reduced, enabling more light signals to reach the venous blood vessel layer;

[0097] At this angle, the contrast of the blood vessel image is increased by 30% compared with the traditional 30° or 60° layout, and the recognition rate of edge blood vessels is increased from 75% to over 95%.

[0098] 8 groups of light sources are symmetrically distributed (such as 4 groups on the top and bottom or 8 groups in a ring), forming a 360° uniform light field, eliminating the illumination difference in different areas of the palm (palm center, back of the hand), solving the problem of local overexposure or underexposure caused by traditional single-sided light sources, and ensuring the balanced acquisition of venous blood vessels across the entire palm.

[0099] Cooperative work with the palmprint recognition device:

[0100] The palm vein recognition device 12 and the palmprint recognition device 11 are spaced 20mm in the vertical direction, forming a three-dimensional acquisition structure of "surface texture + deep blood vessels":

[0101] The user's palm does not need to adjust the posture, and the palmprint and palm vein features can be collected simultaneously by placing it once;

[0102] The image sensors of both achieve data acquisition synchronization through a synchronous trigger circuit (not shown), and the time difference is controlled within 10ms to ensure the timeliness of subsequent feature fusion.

[0103] In an embodiment of the present application, as Figures 1 - 6 shown, in combination with the technical solutions in the claims below, the improved DS evidence theory algorithm and dynamic weight mechanism of the dynamic fusion module 22 will be described in detail:

[0104] It can be understood that the dynamic fusion module 22, as the core component of the data processing unit 2, works in cooperation with the feature extraction module 21:

[0105] The feature extraction module 21 outputs the pre - processed feature vectors of palmprint, palm vein, and iris in parallel through the FPGA chip (such as the palmprint ridge line direction matrix, palm vein blood vessel topology, iris polar coordinate features);

[0106] The dynamic fusion module 22, based on the improved DS evidence theory, adaptively adjusts the confidence levels of the three types of features through dynamic weight coefficients (α, β, γ), and finally generates a comprehensive biometric code.

[0107] (1) Palm vein weight α

[0108] Formula: α = 0.5×e -0.1t +0.3×l

[0109] t (number of consecutive authentications): Real - time recorded by the system counter, t = 0 at the first authentication, α = 0.5 + 0.3I (prioritize relying on palm vein features); when consecutive authentication fails, t increases, and α decays exponentially (such as when t = 3, α = 0.5×e -0·3 +0.3I≈0.37 + 0.3I), to avoid the accumulation of misjudgments caused by repeated authentication of a single feature;

[0110] I (ambient light intensity): Measured by integrating a photoresistor (0≤I≤1, 1 is strong light), under strong light (such as at noon outdoors, I = 1), α = 0.8, enhancing the palm vein feature weight (the penetrability of the near - infrared light source is not affected by ambient light).

[0111] (2) Iris weight γ

[0112] Formula: γ = 0.2×D

[0113] D (user distance): Measured in real - time by a TOF sensor (model VL53L1X) (unit: m, effective range 0.2 - 2m), at a short distance (such as D = 0.3m), γ = 0.06, reducing the iris weight (reducing the impact of near - distance imaging distortion on iris texture matching); at a long distance (D = 1m), γ = 0.2, reasonably distributing the iris feature confidence level.

[0114] (3) Palmprint weight β

[0115] Formula: β = 1 - α - γ

[0116] Dynamic balance: Automatically compensates for the weight changes of α and γ, ensuring that the sum of the weights of the three types of features is always 1 (β≥0.1, avoiding too high a weight of a single feature).

[0117] Feature fusion process based on DS evidence theory:

[0118] (1) Evidence matrix construction

[0119] Calculate the confidence levels for the three types of biometric features respectively:

[0120] Palm vein: Generate a confidence level m1 according to the topological matching degree between the blood vessel bifurcation point 221 and the end point 222;

[0121] Palm print: Generate a confidence level m2 according to the consistency coefficient of the ridge line 223 direction field;

[0122] Iris: Generate a confidence level m3 according to the polar coordinate matching degree between the crypt 224 and the fold 225.

[0123] (2) Dynamic weight assignment

[0124] Take α, β, γ as the weight factors of the evidence theory and adjust the original confidence level:

[0125]

[0126] where w i is the dynamic weight coefficient to ensure environmental adaptability (such as α increases in low light to enhance the confidence level of palm vein features).

[0127] (3) Evidence synthesis and feature code generation

[0128] Synthesize the adjusted confidence levels through the orthogonal sum operation of the DS evidence theory to generate a comprehensive confidence level matrix, and finally output a biometric code containing the confidence levels of three types of features for the two-factor verification of the payment execution module (3).

[0129] It should be noted that in the improved DS evidence theory algorithm, the basic probability assignment function m_i is defined as follows:

[0130]

[0131] where: F i is the palm print / palm vein / iris feature vector extracted currently

[0132] F DB is the pre-stored template feature vector

[0133] sim(·) is the cosine similarity function

[0134] W I ∈{α,β,γ} is the dynamic weight coefficient

[0135] The final comprehensive confidence level is calculated through the orthogonal sum formula:

[0136]

[0137] In an embodiment of the present application, such as Figures 1 - 6As shown below, in combination with the technical solution of the claims, the local storage and distributed update mechanism of the blockchain encryption unit 31 will be described in detail:

[0138] It can be understood that the blockchain encryption unit 31, as the core component of the payment execution module 3, works in coordination with the dynamic fusion module 22 and the two-factor authentication unit 32:

[0139] Data input: Receive the comprehensive biometric code and transaction data (charging duration, amount) output by the dynamic fusion module 22;

[0140] Encryption processing: Implement a two-layer mechanism of "local encrypted caching + distributed secure storage" through the local encrypted storage subunit 311 and the distributed ledger update subunit 312.

[0141] Workflow of the local encrypted storage subunit 311:

[0142] (1) Data encryption

[0143] Use the AES-256 algorithm (Advanced Encryption Standard, 256-bit key) to encrypt the transaction data for the past 7 days, including:

[0144] Biometric template hash value (pre-generated through the national cipher SM4 algorithm);

[0145] Charging transaction records (timestamp, charging pile ID, vehicle identification code, amount, etc.).

[0146] Encryption process: After segmenting the data, perform exclusive OR with a randomly generated initialization vector (IV), and then ensure the security of the ciphertext through 14 encryption rounds.

[0147] (2) Local storage

[0148] The encrypted data is stored in a secure storage chip (such as Secure Element) within the charging pile control cabinet 7, which supports power-off data continuation and has a built-in hardware firewall to prevent data leakage caused by physical intrusion;

[0149] The storage period is set to the past 7 days, and old data beyond the period is automatically overwritten to balance the storage capacity and data traceability requirements.

[0150] Workflow of the distributed ledger update subunit 312:

[0151] (1) Network status monitoring

[0152] Real-time monitor the network connection status of the charging pile (Wi-Fi / 4G). When the network is detected to be disconnected, suspend the distributed upload and switch to the local caching mode;

[0153] After the network is restored, trigger the data synchronization mechanism (supporting resume from breakpoint).

[0154] (2) Data Sharding and Upload

[0155] Data Sharding: The encrypted transaction data is split into at least 3 data fragments (such as using the Shamir secret sharing algorithm), and each fragment contains part of the original data and verification information;

[0156] Blockchain Node Communication: The sharded data is uploaded to at least 3 blockchain nodes (such as charging pile operator nodes, third-party authentication nodes) via the CAN bus, and each node stores different data fragments;

[0157] Hash Verification: Before uploading, generate the SM4 algorithm hash value for each shard, and after the node receives it, perform hash comparison to ensure data integrity.

[0158] (3) Distributed Ledger Update

[0159] After the node receives the sharded data, it verifies the data legality through a consensus mechanism (such as PBFT Practical Byzantine Fault Tolerance), and after passing the verification, writes it into the distributed ledger;

[0160] The local encryption storage subunit 311 deletes the uploaded old data fragments after confirming that at least 2 nodes have received them successfully to avoid duplicate storage.

[0161] In an embodiment of the present application, as Figures 1 - 6 shown, the following combines the technical solutions of the claims to elaborate in detail on the hardware design and working process of the real-time monitoring module 42:

[0162] It can be understood that the real-time monitoring module 42, as the core component of the charging control module 4, is deeply integrated with the charging gun 41:

[0163] Micro-piezoelectric sensor array 421: Uniformly embedded under the insulating layer on the surface of the charging gun grip 43, forming a 16×8 matrix distribution (covering more than 80% of the grip contact surface area), and the size of each sensor unit is 2mm×2mm, which can detect pressure changes above 0.1N;

[0164] Near-infrared spectroscopy analyzer 422: Integrated at the front end of the charging gun grip 43 near the plug, emitting near-infrared light with a wavelength of 760-940nm (the same as the light source wavelength of the palm vein recognition device 12), and the sampling frequency is 0.2Hz (that is, once every 5 seconds).

[0165] Palm Contact State Monitoring Process:

[0166] (1) Pressure Distribution Detection (Micro-piezoelectric Sensor Array 421)

[0167] When the user holds the charging gun 41, the pressure exerted by the palm is conducted through the grip 43 to the micro piezoelectric sensor array 421, and the sensor converts the pressure signal into an electrical signal (the change in voltage is proportional to the pressure).

[0168] The real-time monitoring module 42 performs noise reduction processing (mean filtering + threshold judgment) on the electrical signal, generates a palm contact pressure matrix (resolution 16×8), and calculates the effective contact area (the threshold is set to the area where the pressure ≥ 1N).

[0169] Trigger condition: When the effective contact area < 20 cm 2 or the pressure distribution continues to be lower than the preset threshold for 5 seconds (such as the average pressure < 3N), it is determined as "abnormal palm removal", and an interruption signal is sent to the two-factor verification unit 32.

[0170] (2) Palm vein blood flow pulse acquisition (near-infrared spectroscopy analyzer 422)

[0171] The near-infrared light emitted by the near-infrared spectroscopy analyzer 422 penetrates the palm epidermis, and the absorption difference of hemoglobin for light of different wavelengths causes the change in the intensity of the reflected light, which is converted into a blood flow pulse signal by the photodetector;

[0172] The signal processing unit extracts the characteristic parameters of the pulse wave (such as peak time, waveform period), and compares them with the blood flow characteristics initially collected by the biometric acquisition module 1 to verify the living state of the palm (it is determined as non-living when there is no blood flow signal or waveform abnormality).

[0173] Data interaction and control logic

[0174] (1) Real-time data transmission

[0175] The pressure distribution matrix and the blood flow pulse signal are transmitted to the two-factor verification unit 32 of the payment execution module 3 through the electrical connection line of the charging control module 4, and the transmission rate is 10 Mbps to ensure data synchronization within 50 ms;

[0176] The two-factor verification unit 32 performs multi-dimensional verification on the real-time monitoring data, biometric code, and GPS positioning data (such as when the contact state is abnormal, even if the biometric code matches, the charging is refused to continue).

[0177] (2) Abnormal handling mechanism

[0178] When it is detected that "the palm has been removed for more than 5 seconds" or "there is no blood flow pulse signal", the real-time monitoring module 42 controls the power circuit of the charging gun 41 to be disconnected through the relay, and at the same time triggers the alarm of the abnormal handling module 8;

[0179] After returning to normal contact, biometric authentication needs to be passed again to resume charging to avoid unauthorized continued use.

[0180] In an embodiment of the present application, as Figures 1 - 6 shown, the following will combine with the technical solution of the claims to elaborate in detail on the live detection function and working process of the iris recognition device 13:

[0181] It can be understood that the core component of the iris recognition device 13 is a binocular camera 131 with a live detection function, which is installed at the top edge of the operation area 6 at an inclination of 30°, and its optical axis is aligned with the iris position when the user is looking straight ahead naturally (the pupil center height is about 1.5 m), ensuring that the acquisition range covers both eyes' irises and eyelid areas. The camera resolution is 1920×1080, and it is equipped with an infrared filter to eliminate environmental light interference and supports high-speed imaging at 30 fps.

[0182] Working process of the dynamic iris texture analysis unit 1311:

[0183] (1) Pupil dynamic detection

[0184] The camera 131 continuously captures iris images, extracts the pupil contour through image preprocessing (gray conversion, noise filtering), and uses the active shape model (ASM) to track the pupil edge;

[0185] In two consecutive frames of images, calculate the change amount Δd of the pupil diameter, and divide it by the time interval Δt (33 ms, corresponding to 30 fps) to obtain the change rate v = Δd / Δt;

[0186] Live determination condition: When v≥0.5 mm / s (that is, the pupil diameter changes ≥0.5 mm within 1 second, which conforms to the natural reaction of the live pupil to light and emotional changes), it is determined as a valid live iris feature; if v < 0.1 mm / s for 5 consecutive frames, a non-live warning is triggered.

[0187] (2) Texture authenticity verification

[0188] Synchronously analyze the dynamic changes of the iris texture (such as the relative position changes of crypts and wrinkles), exclude the consistent texture features of static images (such as printed photos), and cooperate with the pupil change rate detection to form a double dynamic verification.

[0189] Working process of the eyelid micro-motion monitoring unit 1312:

[0190] (1) Blink frequency capture

[0191] Based on the 30 fps high-speed video stream, use the Viola-Jones algorithm to detect the eyelid contour, and judge the opening and closing state of the eyelids through the gray projection method (the eyelid coverage rate is <30% when the eyes are open and >70% when the eyes are closed);

[0192] Count the number of blinks per unit time (normal range: 15 - 20 blinks per minute). Abnormal frequencies (< 5 blinks per minute or > 30 blinks per minute) are determined as non-natural blinks (such as mechanical blinks, video frame switching).

[0193] (2) Live body assisted verification

[0194] Combined with the detection result of pupil change rate, confirm that the iris comes from a live body only when the blink frequency is normal and there is dynamic pupil change;

[0195] For the situation where the closed-eye state lasts for more than 2 seconds, suspend iris feature collection and prompt the user "Please open your eyes".

[0196] It should be noted that the pupil diameter change rate threshold of 0.5 mm / s is set based on the IEEE 2790-2022 biometric liveness detection standard and is measured through preliminary experiments:

[0197] Average change rate of live body samples (n = 1000): 0.62 ± 0.15 mm / s

[0198] Change rate of forged samples (3D printed eyeballs): 0.08 ± 0.03 mm / s

[0199] Therefore, a threshold of 0.5 mm / s is set to distinguish between live bodies and non-live bodies.

[0200] In an embodiment of the present application, as Figures 1 - 6 shown, the exception handling module 8, as a security guarantee component of the entire palm-eye-pupil payment charging pile system, monitors the system operation status in real time and triggers an alarm in a timely manner when an exception occurs. The following details its working process:

[0201] It can be understood that (1) Biometric matching degree monitoring

[0202] Data acquisition: The exception handling module 8 obtains the matching degree data of each biometric recognition from the dynamic fusion module 22 of the data processing unit 2.

[0203] Counting and judgment: The built-in counter counts the number of times the biometric matching degree is lower than 98%. When the situation where the matching degree is lower than 98% occurs continuously for 3 times, the alarm mechanism is triggered.

[0204] Alarm execution: An alarm is issued through an audible and visual alarm to remind the staff that there may be problems such as unauthorized use or biometric collection device failure.

[0205] (2) Charging gun contact monitoring

[0206] Connection detection: The exception handling module 8 monitors the connection status between the charging gun 41 and the vehicle in real time. When it detects that the charging gun 41 is successfully connected to the vehicle, it starts a 5-second timer.

[0207] Contact detection: During the timing, palm contact data is obtained from the real-time monitoring module 42. If no palm contact is detected within 5 seconds, an alarm is triggered.

[0208] Alarm execution: An alarm is also issued through the audible and visual alarm to indicate that there may be a misconnection or abnormal user operation.

[0209] (3) GPS positioning monitoring

[0210] Data acquisition: The anomaly handling module 8 obtains real-time GPS positioning data from the two-factor authentication unit 32 of the payment execution module 3, and at the same time obtains the registered address information of the charging pile from the system database.

[0211] Deviation calculation: Calculate the distance deviation between the current GPS positioning data and the registered address of the charging pile.

[0212] Judgment and alarm: When the deviation exceeds 1 kilometer, an alarm is triggered to remind the staff that the charging pile may have been moved or there is a malfunction in the positioning system.

[0213] (4) Processing after alarm

[0214] Once the anomaly handling module 8 triggers an alarm, the system will record detailed information such as the time and type of the anomaly simultaneously, so that the subsequent staff can conduct investigations and processing. Before the anomaly is resolved, the system may suspend the charging service or take other safety measures to ensure the normal operation of the system and the safety of users.

[0215] In an embodiment of the present application, as Figures 1 - 6 shown, the three-level feature fusion algorithm is the core of the dynamic fusion module 22, which is used to fuse the features of palm veins, palm prints, and irises to generate a comprehensive biometric code. The following is the detailed workflow:

[0216] It can be understood that (1) The first-level fusion: Extraction of palm vein topological structure features

[0217] Data input: The dynamic fusion module 22 obtains palm vein image data from the feature extraction module 21.

[0218] Feature extraction: Process the palm vein image to extract the position information of the palm vein bifurcation points 221 and the palm vein endpoints 222.

[0219] Topological structure construction: Based on the extracted bifurcation point and endpoint positions, construct the topological structure of the palm vein blood vessels, and record information such as the connection relationship and distance between each point.

[0220] (2) The second-level fusion: Calculation of the consistency coefficient of the palm print ridge direction field

[0221] Data input: Receive the palmprint image data output by the feature extraction module 21.

[0222] Orientation field calculation: Analyze the palmprint image, calculate the orientations of the palmprint ridges 223 in each local area, and form a palmprint orientation field.

[0223] Consistency coefficient calculation: Statistically analyze the consistency degree of the orientations in each local area of the orientation field to obtain the orientation field consistency coefficient, which reflects the stability and reliability of the palmprint features.

[0224] (3) Third-level fusion: Iris polar coordinate distribution feature matching

[0225] Data input: Obtain the iris image data processed by the feature extraction module 21.

[0226] Polar coordinate transformation: Transform the iris image into the polar coordinate system, and determine the polar coordinate positions of the iris crypts 224 and iris folds 225.

[0227] Feature matching: Match the current polar coordinate distribution feature with the pre-stored template and calculate the matching degree score.

[0228] It should be noted that the control method of this application can be automatically controlled by a controller. The control method of the controller can be realized by simple programming by those skilled in the art, which belongs to the common general knowledge in the art. And this application mainly aims to protect the mechanical structure, so the control method and circuit connection will not be explained in detail in this application.

[0229] Specifically, when the user brings an electric vehicle to the charging pile area, the following full-process operations are performed:

[0230] (1) Biometric collection and authentication

[0231] The user stands in front of the charging pile, places the palm naturally in the center of the operation area 6, with the palm facing the palmprint recognition device 11 (vertically centered and installed, and the 1200dpi sensor collects palmprint images above 500ppi in real time);

[0232] The near-infrared light source array 121 of the palm vein recognition device 12 emits an 850nm beam (8 groups of 3W ± 0.2W light sources, with a 45° angle layout), penetrates the palm, and generates a vein blood vessel grayscale image on the image sensor;

[0233] The binocular camera 131 of the iris recognition device 13 tilts 30° to capture the binocular iris images, and simultaneously starts the live detection (the dynamic iris texture analysis unit 1311 detects that the pupil change rate ≥ 0.5mm / s, and the eyelid micro-motion monitoring unit 1312 captures the blinking frequency of 30fps).

[0234] Three types of original data are transmitted to the data processing unit 2 through shielded cables. The feature extraction module 21 performs parallel processing through FPGA, and the dynamic fusion module 22 executes three-level feature fusion:

[0235] The first level: Extract the topological structure of the palmar vein bifurcations 221 and endpoints 222;

[0236] The second level: Calculate the consistency coefficient of the palmprint ridge direction field;

[0237] The third level: Match the polar coordinate distributions of the iris crypts 224 and wrinkles 225 to generate a comprehensive biometric code.

[0238] (2) Payment verification and charging start

[0239] The biometric code is hashed by the blockchain encryption unit 31 of the payment execution module 3 using the SM4 algorithm. The two-factor verification unit 32 synchronously verifies the feature code and the charging pile GPS positioning data (passing when the deviation ≤ 1 km);

[0240] After passing the verification, the user inserts the charging gun 41 into the charging port of the electric vehicle, and the charging control module 4 starts:

[0241] The micro-piezoelectric sensor array 421 of the real-time monitoring module 42 detects the grip pressure distribution (effective contact area ≥ 20 cm 2 ), and the near-infrared spectrometer 422 collects the palmar vein blood flow pulse signal every 5 seconds (to verify the living state).

[0242] (3) Real-time monitoring during charging

[0243] During charging, the anomaly handling module 8 monitors in real-time:

[0244] If the biometric matching degree is continuously < 98% for 3 times, an audible and visual alarm is triggered;

[0245] If no palm contact is detected within 5 seconds after the charging gun is connected (pressure < 3 N or no blood flow signal), the charging circuit is automatically disconnected;

[0246] If the GPS positioning deviation exceeds 1 km, continue charging is refused and the anomaly is recorded.

[0247] (4) Charging end and data storage

[0248] After charging is completed, the charging control module 4 stops power supply, and the real-time monitoring module 42 terminates data acquisition;

[0249] The transaction data is processed by the blockchain encryption unit 31: The local encryption storage subunit 311 stores the data of the recent 7 days using AES-256, and the distributed ledger update subunit 312 uploads the sharded data to ≥ 3 blockchain nodes when the network is restored to ensure the security and traceability of the data.

[0250] In summary, the palm-eye-pupil payment charging pile of the embodiment of the present application that does not require payment with a payment tool realizes cardless and code-free payment. Through palmprint, palm vein, and iris verification, it solves the problem of dependence on payment media, simplifies operations, improves efficiency, has triple verification to resist forgery, dynamic weights to adapt to the environment, blockchain encryption to prevent data leakage, two-factor verification to prevent fraud, real-time monitoring during the charging process, and full-process protection.

[0251] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A palm-eye-pupil payment charging pile that does not require payment with a payment tool, characterized in that, It includes a biometric collection module (1), a data processing unit (2), a payment execution module (3), a charging control module (4), a charging pile housing (5), an operation area (6), and a charging pile control cabinet (7). Among them, The biometric collection module (1) is fixedly installed above the operation area (6) on the front of the charging pile housing (5), and includes a palmprint recognition device (11), a palm vein recognition device (12), and an iris recognition device (13). Among them, The palmprint recognition device (11) uses a 1200dpi CMOS image sensor and is vertically installed at the center position of the operation area (6); The palm vein recognition device (12) is configured with a near-infrared light source array (121) with a wavelength of 850nm ± 10nm and is horizontally installed 20mm below the palmprint recognition device (11); The iris recognition device (13) includes a binocular camera (131) with a live detection function and is installed at a 30° inclination at the top edge of the operation area (6); The data processing unit (2) is located inside the charging pile control cabinet (7) and is signal-connected to the biometric collection module (1) through a shielded cable. It includes a feature extraction module (21) and a dynamic fusion module (22). Among them, The feature extraction module (21) uses an FPGA chip to parallel-process palmprint, palm vein, and iris features; The dynamic fusion module (22) implements a three-level feature fusion algorithm and outputs a comprehensive biometric code; The payment execution module (3) is communication-connected to the data processing unit (2) through a CAN bus and includes a blockchain encryption unit (31) and a two-factor verification unit (32). Among them, The blockchain encryption unit (31) uses the national secret SM4 algorithm to generate a biometric template hash value; The two-factor verification unit (32) synchronously verifies the biometric code and the charging pile GPS positioning data; The charging control module (4) is electrically connected to the payment execution module (3) and includes a charging gun (41) and a real-time monitoring module (42). The real-time monitoring module (42) continuously detects the user's palm contact status during the charging process.

2. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, wherein The near-infrared light source array (121) of the palm vein recognition device (12) includes 8 groups of symmetrically distributed LED light sources, and the power of each group of light sources is 3W ± 0.2W. The included angle between the light source and the image sensor is 45° ± 5°.

3. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, wherein The dynamic fusion module (22) implements an improved DS evidence theory algorithm, and sets the dynamic weight coefficients α, β, γ to satisfy: α (palm vein weight) = 0.5×e^(-0.1t) + 0.3×I (ambient light intensity); β (palmprint weight) = 1 - α - γ; γ (iris weight) = 0.2×D (user distance).

4. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, wherein The blockchain encryption unit (31) includes a local encrypted storage subunit (311) and a distributed ledger update subunit (312). Among them, The local encrypted storage subunit (311) uses the AES-256 algorithm to store transaction data for the past nearly 7 days; The distributed ledger update subunit (312) uploads the encrypted data in slices to at least 3 blockchain nodes when the network resumes.

5. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, wherein The real-time monitoring module (42) includes a micro piezoelectric sensor array (421) and a near-infrared spectroscopy analyzer (422), where, The micro piezoelectric sensor array (421) is embedded in the surface of the charging gun grip (43) to detect the palm contact pressure distribution; The near-infrared spectroscopy analyzer (422) collects palm vein blood flow pulse signals every 5 seconds.

6. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, characterized in that The liveness detection function of the iris recognition device (13) includes a dynamic iris texture analysis unit (1311) and an eyelid micro motion monitoring unit (1312), where, The dynamic iris texture analysis unit (1311) detects that the pupil diameter change rate ≥ 0.5 mm / s; The eyelid micro motion monitoring unit (1312) captures the blink frequency through a 30 fps high-speed camera.

7. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, wherein The system further includes an exception handling module (8), which triggers an alarm when any of the following situations is detected: The biometric matching degree is continuously lower than 98% for 3 consecutive times; No palm contact is detected within 5 seconds after the charging gun (41) is connected to the vehicle; The deviation between the GPS positioning data and the charging pile registration address exceeds 1 km.

8. The palm-eye-pupil payment charging pile that does not require payment with a payment tool according to claim 1, wherein The three-level feature fusion algorithm includes: The first-level fusion: Extract the topological structure features of the palm vein bifurcation points (221) and the palm vein endpoints (222); The second-level fusion: Calculate the direction field consistency coefficient of the palmprint ridge lines (223); The third-level fusion: Match the polar coordinate distribution features of the iris crypts (224) and the iris folds (225).

Images