Tablet dynamic permission management method and system based on biometrics and behavior patterns

Abstract

The application provides a tablet dynamic permission management method and system based on biological characteristics and behavior patterns, constructs a synchronous original biological characteristic vector, acquires a behavior sequence and constructs an extended initial behavior vector, acquires filling index data through the extended initial behavior vector, performs a correction judgment operation, and acquires a corrected behavior vector; acquires multiple tablet computer signal data, performs overlap detection, and determines an identity characteristic vector; classifies the identity characteristic vector, and acquires a permission update basis weight based on identity classification result data; acquires a user permission data matrix, adjusts the user permission data matrix based on the permission update basis weight, adjusts the tablet computer based on the most optimal permission result, and provides a right management method which can identify a dominant user in a multi-user scenario in real time, combines biological characteristics and behavior patterns for double verification, and dynamically adjusts permissions in a scenario where multiple users unlock and multiple people operate the same tablet in a short time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of permission management, in particular to a tablet dynamic permission management method and system based on biological characteristics and behavior patterns. BACKGROUND

[0002] With the popularity of tablet computers in home, office, education and other scenarios, its use mode has evolved rapidly from the early single-user exclusive to multi-user sharing. For example, in a home scenario, parents and children may alternately use the same tablet to handle work files, watch videos, and complete homework in a short period of time. In an office scenario, colleagues may temporarily borrow a tablet to review meeting materials and demonstrate solutions. In an education scenario, teachers and students may take turns using a tablet for teaching interaction and submitting homework.

[0003] The current mainstream tablet computer permission management still uses single-user authentication and fixed permission allocation to manage permissions. In a scenario where multiple users unlock and operate the same tablet in a short period of time, this traditional permission management method still has shortcomings, such as lag in identity recognition and permission switching. Existing permission management methods often require users to actively trigger identity switching, such as clicking to switch users after manual unlocking. They cannot identify the dominant user who is actually operating in real time. For example, in a multi-user home scenario, father Zhang handles work after unlocking the tablet without logging out of the account. Child Li picks up the tablet and unlocks it, but still maintains the administrator permission of father Zhang, which may cause child Li to mistakenly delete work files or delete work files due to accidental touch of the tablet. The traditional permission management method relies on a single static biological feature, which is vulnerable to biological feature theft and may lead to misuse of permissions.

[0004] In summary, in a scenario where multiple users unlock and operate the same tablet in a short period of time, there is an urgent need for a permission management method that can identify the dominant user in a multi-user scenario in real time, combine biological features and behavior patterns for dual verification, and dynamically adjust permissions to adapt to the current mainstream scenario demand of tablet sharing. SUMMARY

[0005] In view of the above-mentioned problems, in combination with the first aspect of the present application, the embodiments of the present application provide a tablet dynamic permission management method based on biological characteristics and behavior patterns, which comprises:

[0006] Collecting fingerprint texture data, facial contour data, and voiceprint spectrum signals and constructing a synchronous original biological feature vector;

[0007] Obtaining a behavior sequence and constructing an extended initial behavior vector, obtaining filling index data through the extended initial behavior vector, performing a correction judgment operation according to the filling index data, obtaining a corrected behavior vector, and simultaneously performing interference filtering on the corrected behavior vector;

[0008] Acquire signal data from multiple tablet computers, perform overlap detection on the multiple tablet computer signal data, determine the identity feature vector, and associate the identity feature vector with the corrected behavior vector accordingly;

[0009] The identity feature vectors are classified, and the permission update criteria weights are obtained based on the identity classification results data.

[0010] Obtain the user permission data matrix, adjust the user permission data matrix based on the permission update criteria weight, confirm the final permission result, and adjust the permissions of the tablet computer based on the final permission result.

[0011] As a further aspect of the present invention, a behavior sequence is obtained and an extended initial behavior vector is constructed. Filling index data is obtained through the extended initial behavior vector. A correction judgment operation is performed based on the filling index data to obtain a corrected behavior vector. Simultaneously, interference filtering is applied to the corrected behavior vector, including:

[0012] The behavioral sequence includes a pressure intensity sequence and a sliding trajectory coordinate sequence;

[0013] The pressure intensity sequence is linearly filled into the synchronous original biofeature vector to obtain an initial behavior vector. Based on the initial behavior vector and combined with the sliding trajectory coordinate sequence, the initial behavior vector is expanded to obtain an extended initial behavior vector.

[0014] Based on the extended initial behavior vector, continuous feature data is obtained, and the extended initial behavior vector is filled with data according to the continuous feature data, and filling index data is obtained.

[0015] The correction judgment operation is performed based on the filling index data. If the filling index data is greater than or equal to the preset correction threshold, the expansion initial behavior vector is corrected based on the curvature correction mechanism. If the filling index data is less than the preset correction threshold, the expansion initial behavior vector is corrected based on the backup correction mechanism to obtain the corrected behavior vector.

[0016] The modified behavior vector is noise filtered using the Kalman filter algorithm to obtain the filtered modified behavior vector. At the same time, the signal strength of the filtered modified behavior vector is judged to obtain a stable modified behavior vector.

[0017] As a further aspect of the present invention, if the filling index data is greater than or equal to a preset correction threshold, the initial behavior vector of the expansion is corrected based on the curvature correction mechanism to obtain the corrected behavior vector, including:

[0018] The filling index data is traversed and queried with a preset association mapping table to obtain the trajectory curvature parameter deviation;

[0019] The relative influence coefficient is obtained based on the deviation between the filling index data and the trajectory curvature parameter, and the trajectory curvature parameter is corrected according to the relative influence coefficient.

[0020] The sliding trajectory coordinate sequence in the extended initial behavior vector is corrected by modifying the trajectory curvature parameters, and the modified extended initial behavior vector is defined as the modified behavior vector.

[0021] As a further aspect of the present invention, if the filling index data is less than a preset correction threshold, the initial behavior vector of the expansion is corrected based on a backup correction mechanism to obtain a corrected behavior vector, including:

[0022] Obtain an activation signal, and activate the tablet computer buffer based on the activation signal;

[0023] Behavioral trajectory fragment data is obtained by retrieving the buffer of the tablet computer, and the basic data for reconstruction is determined based on the behavioral trajectory fragment data;

[0024] Based on the reconstruction foundation, the data integrity of the extended initial behavior vector is verified, and the verification result data is obtained.

[0025] The data of the verification result is used to perform data recovery and correction on the extended initial behavior vector, and the corrected extended initial behavior vector is defined as the corrected behavior vector.

[0026] As a further aspect of the present invention, acquiring multiple tablet computer signal data, performing overlap detection on the multiple tablet computer signal data, and determining an identity feature vector includes:

[0027] Acquire signal data from multiple tablet computers in a multi-user scenario, wherein the tablet computer signal data represents the signals generated by multiple users using the tablet computers;

[0028] The overlapping signals in the multiple tablet computer signal data are detected, multiple overlapping region identifiers are obtained, and the signal strength of each tablet computer signal data is confirmed.

[0029] Based on the signal strength of each tablet computer signal data, the multiple overlapping region identifiers are sorted to obtain sorting feature data;

[0030] Dominant component analysis is performed on the sorted feature data to obtain dominant feature data. Based on the dominant component analysis method, feature separation is performed on the dominant feature data to obtain separated feature data.

[0031] The separated feature data is associated with the sorted feature data to generate an identity feature vector.

[0032] As a further aspect of the present invention, the identity feature vector is classified, and the permission update criteria weight is obtained based on the identity classification result data obtained from the classification, including:

[0033] The identity feature vector and the corresponding modified behavior vector are imported into the support vector machine algorithm, and the identity feature vector is classified based on the support vector machine algorithm to obtain identity classification result data.

[0034] The identity classification result data is matched with a pre-established user pattern matching database to obtain identity matching degree data, and the permission update basis weight is confirmed based on the identity matching degree data.

[0035] As a further aspect of the present invention, obtaining a user permission data matrix, adjusting the user permission data matrix based on the permission update criterion weights, confirming the final permission result, and adjusting the permissions of the tablet computer based on the final permission result includes:

[0036] The system obtains a real-time user permission data matrix through a tablet computer system, and adjusts the user permission data matrix in real time based on the permission update criteria weight and using a row update method.

[0037] Specifically, when adjusting the user permission data matrix, adjustments are made based on a real-time queue management mechanism.

[0038] As a further aspect of the present invention, when adjusting the user permission data matrix, the adjustment is based on a real-time queue management mechanism, including:

[0039] Key update parameters are extracted from the user permission data and used as queue input. The key update parameters include user activity and request frequency. A user priority ranking sequence is determined based on the key update parameters.

[0040] The user permission data matrix is ​​updated based on the user priority ranking sequence and the permission update is performed according to the weight, and an update report is generated.

[0041] As a further aspect of the present invention, fingerprint texture data, facial contour data, and voiceprint spectrum signals are collected and a synchronous original biometric vector is constructed, including:

[0042] The system collects fingerprint texture data, facial contour data, and voiceprint spectrum signals based on a multimodal sensor module.

[0043] The acquisition timing is obtained based on the acquisition time points of the fingerprint texture data, the facial contour data, and the voiceprint spectrum signal, and the acquisition timing is processed based on the synchronization protocol.

[0044] Acquire the amplitude feature data corresponding to the fingerprint texture data, the facial contour data and the voiceprint spectrum signal, normalize the amplitude feature data to a uniform scale, and obtain a standard original biometric vector.

[0045] The acquisition timing sequence is imported into the standard original biological feature vector to obtain the synchronous original biological feature vector.

[0046] Furthermore, embodiments of the present invention also provide a tablet dynamic permission management system based on biometrics and behavioral patterns, the system comprising:

[0047] The acquisition module is used to acquire fingerprint texture data, facial contour data, voiceprint spectrum signal, behavior sequence, user permission data matrix and multiple tablet computer signal data, and simultaneously construct a synchronous original biometric vector from fingerprint texture data, facial contour data and voiceprint spectrum signal.

[0048] A correction module is used to construct an extended initial behavior vector and obtain a corrected behavior vector after interference filtering.

[0049] The confirmation module is used to perform overlap detection on multiple tablet computer signal data to determine the identity feature vector, and associate the identity feature vector with the corresponding correction behavior vector;

[0050] A classification module is used to classify identity feature vectors and obtain the weights for permission updates.

[0051] An adjustment module is used to adjust the user permission data matrix according to the permission update basis weight and confirm the final permission result.

[0052] An allocation module is used to adjust the permissions of the tablet computer based on the final permission result.

[0053] Based on the above, this application embodiment first acquires biometric data by collecting fingerprint texture data, facial contour data, and voiceprint spectrum signals and constructing a synchronous original biometric vector. Next, it acquires behavioral sequences and constructs an extended initial behavioral vector to acquire behavioral feature data. Filling index data is obtained through the extended initial behavioral vector, and a correction judgment operation is performed based on the filling index data to obtain a corrected behavioral vector. Simultaneously, interference filtering is applied to the corrected behavioral vector, forming a dual verification of static identity and dynamic behavior. Then, multiple tablet computer signal data are acquired, and overlap detection is performed on the multiple tablet computer signal data to determine the identity feature vector. The identity feature vector is correlated with the corrected behavioral vector, and the mixed signals from multiple users are separated to prevent misjudgment of identity. The identity feature vector is then classified, and the permission update basis weight is obtained based on the classification result data. Finally, a user permission data matrix is ​​obtained, and the user permission data matrix is ​​adjusted based on the permission update basis weight. The final permission result is confirmed, and the tablet computer permissions are adjusted based on the final permission result. This provides a rights management method that can identify the dominant user in a multi-user scenario, combine dual verification of biometrics and behavioral patterns, and dynamically adjust permissions in real time, especially in scenarios where multiple users unlock and operate the same tablet within a short period. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the execution flow of the tablet dynamic permission management method based on biometrics and behavioral patterns provided in the embodiments of the present invention.

[0055] Figure 2 This is a schematic diagram of a tablet dynamic permission management system based on biometrics and behavioral patterns provided in an embodiment of the present invention. Detailed Implementation

[0056] The accompanying drawings in the embodiments provide a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can typically be arranged and designed in various different configurations.

[0057] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0058] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1This is a schematic diagram of the execution flow of a tablet dynamic permission management method based on biometrics and behavioral patterns provided in an embodiment of the present invention. The following is a detailed description of the tablet dynamic permission management method based on biometrics and behavioral patterns.

[0059] Specifically, the tablet dynamic permission management method based on biometrics and behavioral patterns includes:

[0060] Step S1: Collect fingerprint texture data, facial contour data, and voiceprint spectrum signals, and construct a synchronous original biometric vector.

[0061] In this embodiment, step S1 includes:

[0062] Step S11: Collect fingerprint texture data, facial contour data and voiceprint spectrum signal based on the multimodal sensor module.

[0063] Specifically, data can be collected through various sensors built into the tablet. For example, the fingerprint sensor can generate grayscale matrix data of fingerprint texture by detecting the capacitance difference or light reflection difference between the ridges and valleys. The front-facing camera can capture the user's facial contour data. The built-in microphone can convert the user's voice into an electrical signal and then generate a voiceprint spectrum map through Fourier transform, extracting characteristic signals such as spectral peaks and frequency bandwidth to form a voiceprint spectrum signal. During the collection, the three types of sensors can be triggered and started synchronously under the tablet's system, thereby ensuring that the real-time biometric features collected are those of the same user.

[0064] For example, suppose that after Zhang San unlocks the tablet, the tablet's built-in sensors begin to collect data. The fingerprint sensor is placed against Zhang San's index finger to collect the fingerprint texture into a 256*256 grayscale matrix, where the ridge area has a grayscale value of 200-230 and the valley area has a grayscale value of 50-80. The front-facing camera captures an image of Zhang San's face and calculates the three-dimensional contour parameters, thereby obtaining data such as the brow bone height of 8mm and the chin contour curvature of 0.15. The built-in microphone collects the sound signal of unlocking the tablet for 2-3 seconds, and the voiceprint spectrum is obtained through Fourier transform. The main peak frequency of 1500Hz and the secondary peak frequency of 800Hz are extracted from the voiceprint frequency.

[0065] Step S12: Obtain the acquisition timing based on the acquisition time points of the fingerprint texture data, the facial contour data and the voiceprint spectrum signal, and process the acquisition timing based on the synchronization protocol.

[0066] Understandably, due to the different response speeds of different sensors (e.g., fingerprint sensors respond quickly while front-facing cameras respond slowly), there may be a time difference in data acquisition. If the three types of data are directly stitched together, there is a probability that the fingerprint data will be the current user's data, while the facial data will be the previous user's data. Therefore, it is necessary to extract the acquisition time point of each sensor, which can be obtained from the sensor's hardware clock, typically in the millisecond range. Then, through the clock synchronization mechanism of the I2C bus, all acquisition time points are aligned to the same base time, generating time-series labels and marking the unified timestamp corresponding to each of the three types of data to ensure the time consistency of subsequent feature vectors.

[0067] For example, when Zhang San unlocks the tablet, the data collection timestamps of the three sensors are as follows: fingerprint sensor completes collection at 10:00:00.1, front camera completes collection at 10:00:00.2, and built-in microphone completes collection at 10:00:00.03. The tablet uses the I2C clock synchronization protocol to correct the timestamp of the fingerprint data to 10:00:00.2, with a delay of 0.1 seconds before marking, and to correct the timestamp of the microphone data to 10:00:00.2, with a advance of 0.1 seconds before marking. In the end, all three types of data will carry a unified timestamp of 10:00:00.2. If Li Si touches the screen at 10:00:00.4, his facial data will be marked as 10:00:00.4, which can be clearly distinguished from Zhang San's data and avoid confusion.

[0068] Step S13: Obtain the amplitude feature data corresponding to the fingerprint texture data, the facial contour data, and the voiceprint spectrum signal; normalize the amplitude feature data to a uniform scale; and obtain a standard original biometric vector.

[0069] It is understandable that the units of the three types of data mentioned above are different. For example, fingerprint data is in grayscale, facial data is in millimeters, and voiceprint data is in Hertz. If they are used directly for calculation, the difference in units will lead to voiceprint feature weight being too high and fingerprint feature weight being too low. Therefore, it is necessary to normalize the three types of data to a unified scale.

[0070] Specifically, amplitude feature data of the three types of data are extracted. These data can be the average gray value of fingerprints, the average contour of faces, the average frequency of voiceprints, etc. Then, the min-max normalization algorithm can be used to uniformly map all amplitude feature data to the 0-1 interval. The fingerprints, faces and voiceprints normalized to a uniform scale are then spliced ​​in sequence to form a standardized original biometric vector, i.e., the standard original biometric vector, which can ensure the weight balance of the three types of data in subsequent calculations.

[0071] For example, suppose Zhang San's three types of data have the following feature amplitude values: fingerprint average gray value 180, with a minimum value of 50 and a maximum value of 230, normalized to (180-50) / (230-50) = 130 / 180 ≈ 0.72; facial average contour distance (distance between eyes + width of nose) / 2 = (65+30) / 2 = 47.5mm, with a minimum value of 40mm and a maximum value of 80mm, normalized to (47.5-40) / (80) / (230-50) ≈ 0.72; and average facial contour distance (distance between eyes + width of nose) / 2 = (65+30) / 2 = 47.5mm, with a minimum value of 40mm and a maximum value of 80mm, normalized to (47.5-40) / (80) / (230-50) ≈ 0.72. -40) = 7.5 / 40 = 0.19, the average frequency of the voiceprint is (1500+800) / 2 = 1150Hz, with a minimum value of 20Hz and a maximum value of 20000Hz. The normalized calculation is (1150-20) / (20000-20) = 1130 / 19980≈0.06. Then, the fingerprint, facial and voiceprint can be stitched together to obtain the standard original biometric vector [0.72, 0.19, 0.06].

[0072] Step S14: Import the acquisition timing sequence into the standard original biological feature vector to obtain the synchronous original biological feature vector.

[0073] Understandably, after the standard original biometric vector is obtained, it only includes feature data and does not include relevant data in the time dimension. This makes it impossible to determine whether the features are from real-time collection at a unified time. At this time, it is necessary to import the aligned unified timestamp into the standard original biometric vector as a new dimension. After importing, four dimensions are formed: fingerprint feature, facial feature, voiceprint feature, and collection time. Together, they form the synchronized original biometric vector. The synchronized original biometric vector can reflect the user's biometric features and can also verify the real-time nature of the features through the timestamp.

[0074] Furthermore, using the timestamp 10:00:00.2 from step S12 and the standard original biometric vector [0.72, 0.19, 0.06] from step S13 as an example, the unified timestamp 10:00:00.2 is used as the fourth dimension and imported into the standard original biometric vector [0.72, 0.19, 0.06] to generate a synchronized original biometric vector [0.72, 0.19, 0.06, 10:00:00.2]. If, in subsequent steps, the timestamp of a vector is found to differ from the current time by more than N minutes, the vector will be determined to be non-real-time data and rejected for use in identity recognition, thus preventing expired biometric features from being misused.

[0075] In summary, step S1 is the data acquisition foundational stage of the overall process of this method. It ensures the uniqueness and accuracy of subsequent identity recognition by collecting multi-dimensional biometric features. It solves the problem of single biometric features being easily interfered with by using three types of data from sensors. By handling time-series differences and aligning timestamps with synchronization protocols, it avoids the problem of recognition errors caused by splicing features across time dimensions. Through amplitude feature normalization, it unifies feature data of different units to a 0-1 scale, eliminating the impact of unit differences on subsequent calculations. Finally, it fuses time information into the standard original biometric feature vector to generate a synchronization vector containing the time dimension, ensuring the real-time nature of the data.

[0076] Step S2: Obtain the behavior sequence and construct an extended initial behavior vector. Obtain the filling index data through the extended initial behavior vector. Perform a correction judgment operation based on the filling index data to obtain the corrected behavior vector. At the same time, perform interference filtering on the corrected behavior vector.

[0077] The behavior sequence includes a pressure intensity sequence and a sliding trajectory coordinate sequence.

[0078] In this embodiment, step S2 includes:

[0079] Step S21: The pressure intensity sequence is linearly filled into the synchronous original biofeature vector to obtain an initial behavior vector. Based on the initial behavior vector and combined with the sliding trajectory coordinate sequence, the initial behavior vector is expanded to obtain an extended initial behavior vector.

[0080] Specifically, there may be a mismatch between the sampling frequency of the pressure intensity sequence and the dimension of the synchronous original biometric vector. If the pressure intensity sequence and the synchronous original biometric vector are forcibly concatenated, the data length will be inconsistent. A linear padding algorithm can be used to pad the length of the pressure intensity sequence to match that of the synchronous original biometric vector, thereby generating an initial behavior vector that includes both biometric and pressure features. Then, the sliding trajectory coordinate sequence is used as a new dimension and concatenated into the initial behavior vector in chronological order to expand the dimension of the vector and form an expanded initial behavior vector. This vector contains both biometric and behavioral features, which can improve the accuracy of identity verification.

[0081] Understandably, the linear fill algorithm calculates the missing value between two adjacent true pressure values. For example, between pressure values ​​15 and 12, the value filled in the middle is 13.5.

[0082] For example, when Zhang San swipes through the photo album on a tablet, assuming the pressure intensity sequence is [10, 15, 12, 8], and the original biometric vector is 10-dimensional, a linear padding algorithm is used to complete 10 pressure points [10, 11.25, 12.5, 13.75, 15, 13.5, 12, 10, 9, 8], which are then aligned with the synchronization vector to generate a 10-dimensional initial behavior vector [0.72, 0.19, 0.06, 10:10:00.2, 1]. [0, 11.25, 12.5, 13.75, 15, 13.5], and then add 10 sliding trajectory coordinate points [(100, 200), (120, 220), (140, 240), (160, 260), (180, 280), (200, 300), (220, 320), (240, 340), (250, 350), (260, 360)], thus expanding into a 20-dimensional extended initial behavior vector.

[0083] Step S22: Obtain continuous feature data based on the extended initial behavior vector, fill the extended initial behavior vector with data according to the continuous feature data, and obtain the filling index data.

[0084] Understandably, during user operation, pressure or coordinate data may be lost due to the finger briefly leaving the screen or temporary sensor malfunction of the tablet. For example, the coordinates of a certain point in the sliding trajectory may not be collected. Therefore, it is necessary to extract the continuity features of the extended initial behavior vector. If data loss is detected, cubic spline interpolation is used to fill the data. After filling, the filling index is calculated, which is the error rate between the filled data and the adjacent real data. This index can be used to judge the filling effect and provide a basis for subsequent corrections.

[0085] Step S23: Perform a correction judgment operation based on the filling index data. If the filling index data is greater than or equal to a preset correction threshold, then the initial expansion behavior vector is corrected based on the curvature correction mechanism. If the filling index data is less than the preset correction threshold, then the initial expansion behavior vector is corrected based on the backup correction mechanism to obtain the corrected behavior vector.

[0086] Understandably, assuming the preset correction threshold is 0.1, i.e. the error rate is 10%, if the filling index is greater than or equal to 0.1, it means that the filling data error is large and needs to be corrected by trajectory curvature correction to expand the initial behavior vector. If the filling index is less than 0.1, it means that the filling data error is small and can be corrected by backup correction. By correcting the data deviation through two different strategies, we can ensure that the behavior vector can truly reflect the user's operating habits and avoid misjudgment of identity due to data errors.

[0087] In this embodiment, step S23 includes step S231:

[0088] Step S231-1: Trajectory and query the filling index data with the preset association mapping table to obtain the trajectory curvature parameter deviation.

[0089] Understandably, the preset association mapping table is generated in advance through training on a large amount of user behavior data. It is used to store the correspondence between the filling index and the trajectory curvature parameter deviation. For example, if the filling index is 0.1, after querying the preset association mapping table, the trajectory curvature parameter deviation is 0.12. The larger the filling index, the more serious the data loss or deviation, and the larger the corresponding trajectory curvature parameter deviation. The trajectory curvature parameter deviation reflects the degree of curvature of the trajectory. The larger the deviation, the more the trajectory deviates from the real path. By clarifying the deviation between the current trajectory curvature and the real curvature through the trajectory curvature parameter, the deviation basis is provided for subsequent correction.

[0090] For example, in a certain preset association mapping table, the correspondence between the filling index and the curvature deviation is: [0.1-0.06, 0.12-0.07, 0.15-0.12]. Assuming that Zhang San's filling index is 0.12, the preset association mapping table is traversed and queried, and the corresponding trajectory curvature parameter deviation is found to be 0.07.

[0091] Step S231-2: Obtain the relative influence coefficient based on the deviation between the filling index data and the trajectory curvature parameter, and correct the trajectory curvature parameter according to the relative influence coefficient.

[0092] Understandably, the relative influence coefficient is used to measure the total influence of the fill index and curvature deviation on the trajectory. It can be calculated as follows: relative influence coefficient = fill index * 0.6 + curvature deviation * 0.4. In the specific implementation, the weights of the fill index and curvature deviation can be set according to their importance to the trajectory. After obtaining the relative influence coefficient, the current trajectory curvature parameter is adjusted by calling the relative influence coefficient. The corrected curvature can be corrected by the formula: corrected curvature = current curvature - relative influence coefficient * correction coefficient. The correction coefficient can be set to 0.5 to avoid excessive correction and make the corrected curvature parameter closer to the true value.

[0093] For example, if Zhang San's relative influence coefficient is 0.15, then the corrected curvature is 0.15 - 0.15 * 0.5 = 0.075, which is closer to the true value and ensures that the magnitude of the correction matches the degree of deviation.

[0094] Step S231-3: The sliding trajectory coordinate sequence in the extended initial behavior vector is corrected by the corrected trajectory curvature parameter, and the corrected extended initial behavior vector is defined as the corrected behavior vector.

[0095] Understandably, by using the corrected trajectory curvature to infer the reasonable position of each trajectory coordinate point, and adjusting the coordinate values ​​of the X and Y axes according to the corrected curvature to make them fit the curvature trend of the overall trajectory again, the sliding trajectory coordinate sequence in the original extended initial behavior vector is updated, and the new extended initial behavior vector can be defined as the corrected behavior vector.

[0096] For example, suppose Zhang San's corrected trajectory curvature is 0.1, and the 6th coordinate point (200, 300) deviates from the trajectory due to filling error. Since the curvature is 0.1, the X-axis value should be between 215 and 220. At this time, we can reverse the calculation based on the corrected curvature. After the reverse calculation, the X-axis value is 218 and the Y-axis value is 300. The Y-axis value conforms to the overall increasing trend and does not need to be adjusted. Based on the corrected coordinate point (218, 300), the sliding trajectory coordinate sequence in the original extended initial behavior vector is updated to form the corrected behavior vector. The trajectory of this vector is consistent with Zhang San's actual sliding habits.

[0097] In this embodiment, step S23 includes step S232:

[0098] Step S232-1: Obtain an activation signal and activate the tablet computer buffer based on the activation signal.

[0099] Specifically, the activation signal is a trigger signal automatically generated when the filling index is less than a preset threshold. For example, the level signal is a high level 1, which is used to wake up the behavior trajectory buffer in the tablet computer. This buffer is a built-in storage area of ​​the tablet computer used to store recent user behavior trajectory fragments, and is stored according to user identity. For example, the trajectory fragment of user Zhang San is stored under user Zhang San's target. When the activation signal is triggered, the buffer changes from a dormant state to a readable state, and then the historical trajectory data of the corresponding user can be retrieved.

[0100] For example, if Zhang San's fill index is 0.08, which is less than the preset threshold of 0.1, the tablet computer system automatically generates a high-level signal 1, triggering the activation of the tablet computer's behavior trajectory buffer. The buffer stores nearly N hours of user trajectory. Zhang San's trajectory fragments are stored under the user Zhang San target, including a total of 5 records, 3 slides in the photo album, and 2 clicks on the file. The buffer may also include Li Si's trajectory, which is stored under the user Li Si target, and Wang Wu's trajectory is stored in the user Wang Wu's directory. After activation, the tablet computer only has permission to read the trajectory in the user Zhang San's directory.

[0101] Step S232-2: Obtain behavioral trajectory fragment data by retrieving the tablet computer buffer, and determine the basic data for reconstruction based on the behavioral trajectory fragment data.

[0102] Specifically, the tablet system searches for similar trajectory segments in the active buffer based on key characteristics of the current behavior, such as swipe direction: upward to the right, swipe length: 500 pixels, and application: photo album. The criteria for similarity judgment can be that the direction is consistent, the length deviation is less than 10%, and the application is the same. After a similar segment is found, the complete data in the segment is extracted to confirm the basic data for reconstruction.

[0103] For example, based on Zhang San's current behavioral characteristics, the tablet system swipes to the upper right with a length of 500 pixels, using the photo album application. It finds a similar trajectory segment in Zhang San's directory, collected at 10:02:00. The segment is swipe to the upper right with a length of 480 pixels and a length deviation of 4%, which is less than 10%. This segment is then extracted to confirm the basic data for reconstruction.

[0104] Step S232-3: Based on the reconstruction foundation, perform data integrity verification on the extended initial behavior vector and obtain the verification result data.

[0105] Understandably, after extraction, it is necessary to compare the current expanded initial behavior vector with the reconstructed base data point by point to verify the data integrity of the vector.

[0106] Specifically, during the verification process, it is necessary to check for missing data, whether the data deviation is within a reasonable range, and whether the trajectory trend is consistent. After the verification is completed, verification result data is generated. The verification result data is used to clarify the number of missing data points, the number of data points with deviations exceeding the range, and the consistency of the trajectory trend.

[0107] Step S232-4: Perform data recovery correction on the extended initial behavior vector using the verification result data, and define the corrected extended initial behavior vector as the corrected behavior vector.

[0108] Specifically, targeted corrections are made based on the verification results data. If there are missing data points, the corresponding points in the reconstructed basic data can be directly used to fill them. If there are deviations that exceed the reasonable range, the average value of the basic data and the current data is used for correction. If the trajectory trend is inconsistent, multiple coordinates are adjusted to make the overall trend consistent with the basic data. After the correction is completed, the original extended initial behavior vector is updated, and the new extended initial behavior vector is defined as the correction behavior vector.

[0109] Step S24: Based on the Kalman filter algorithm, noise is filtered on the modified behavior vector to obtain the filtered modified behavior vector. At the same time, signal strength is judged on the filtered modified behavior vector to obtain a stable modified behavior vector.

[0110] It is understandable that during user operation, interference such as hand tremors and screen smudges may cause some random noise in the correction behavior vector. This noise can be eliminated by the Kalman filter algorithm.

[0111] Furthermore, the Kalman filter algorithm effectively filters the aforementioned noise through a prediction and update cycle. It predicts the reasonable value for the current moment based on the behavioral data from the previous moment and combines it with the currently collected actual value to calculate the optimal estimated value, thereby eliminating random interference. After filtering, it is also necessary to make a judgment based on the signal-to-noise ratio (SNR). The higher the SNR, the more stable the signal. If the SNR meets the condition, it can be directly determined as a stable corrected behavioral vector. However, if the SNR does not meet the condition, behavioral data needs to be re-collected to avoid unstable data affecting subsequent identity judgment.

[0112] For example, in Zhang San's corrected behavior vector, the third coordinate point is noisy due to hand tremors. Assuming the actual collected value is (140, 240), the hand tremors cause it to become (148, 245), and the pressure value changes from 12.5g to 15g. Using the Kalman filter algorithm, the coordinates of the third point (140, 240) and the pressure value of 12.5g are predicted based on the second point (120, 220) and the fourth point (160, 260). Combining this with the actual collected value, the optimal estimated value is calculated to be (142, 242) and the pressure value is 13g. The hand tremor noise is filtered out. After filtering, the signal-to-noise ratio is calculated to be 28dB. If the condition is greater than 20dB, the optimal estimated value meets the condition, and a stable corrected behavior vector is obtained, eliminating the instability of random interference.

[0113] Step S3: Acquire multiple tablet computer signal data, perform overlap detection on the multiple tablet computer signal data, determine the identity feature vector, and associate the identity feature vector with the corrected behavior vector.

[0114] In this embodiment, step S3 includes:

[0115] Step S31: Obtain signal data from multiple tablet computers in a multi-user scenario. The tablet computer signal data represents the signals generated by multiple users using the tablet computers.

[0116] Understandably, tablet signal data refers to various electrical signals generated by the system when a user operates the tablet. These signals contain multi-dimensional information, such as behavioral operation signals and device interaction signals. The tablet system captures these signals through the tablet's signal acquisition module and stores them according to the user's representation.

[0117] For example, the tablet computer detects three active users: user Zhang San, user Li Si, and user Wang Wu. It categorizes the different signals from these three users into three types and stores them in the signal directories for user Zhang San, user Li Si, and user Wang Wu, respectively. Each directory contains three fields: signal type, data content, and acquisition time.

[0118] Step S32: Detect overlapping signals in the multiple tablet computer signal data, obtain multiple overlapping region identifiers, and confirm the signal strength of each tablet computer signal data.

[0119] Specifically, overlapping signals refer to the overlap of signals from different users in the time or space dimensions. The time dimension refers to the same time period, while the space dimension refers to the same screen area. These overlapping signals are more likely to cause misjudgments by the tablet computer system.

[0120] Furthermore, overlapping areas are marked by comparing timestamps or spatial coordinates. A combination of time range and spatial region can be used as the identifier for overlapping areas, and then the intensity of each overlapping signal is detected.

[0121] In some possible embodiments, taking the behavioral operation signals in the tablet computer signal data as an example, Zhang San's swipe signal timestamp is 10:00:00-10:03:00, and Li Si's click signal timestamp is 10:01:00-10:02:00. By comparing the timestamps, the overlapping area is identified as 10:01:00-10:02:00 and the full screen area. After the overlapping area is identified, the signal strength is measured. If Zhang San's swipe signal amplitude is large and the strength is 80dB, and Li Si's click signal amplitude is medium and the strength is 65dB, the overlapping area identification is associated with the corresponding signal strength to generate a lookup table of overlapping areas and strengths.

[0122] Step S33: Sort the multiple overlapping region identifiers based on the signal strength of each tablet computer signal data to obtain sorting feature data.

[0123] Understandably, all signals corresponding to the overlapping area are sorted from highest to lowest intensity. If there are signals with the same intensity, they can be sorted based on the intensity of the remaining signals. In practice, users can choose the sorting order and the signal intensity basis for sorting according to the actual situation.

[0124] In some possible embodiments, it is assumed that the signal strength of the overlapping area 10:01:00-10:02:00 and the corresponding signal strength of the full screen area are sorted according to the user's action signal. The sorting order is: user Zhang San - strength 80dB, user Li Si - strength 65dB. The data is spliced ​​in the above order to generate sorting feature data. If the signal strength of user Li Si is also 80dB, the same as that of user Zhang San, then the strength of the device interaction signal can be compared next.

[0125] Step S34: Perform dominant component analysis on the sorted feature data to obtain dominant feature data, and perform feature separation on the dominant feature data based on the dominant component analysis method to obtain separated feature data.

[0126] Understandably, since the ranking feature data consists of tablet computer signal data, which includes multi-dimensional signals such as behavioral operation signals and device interaction signals, it is necessary to use Principal Component Analysis (PCA) to reduce the dimensionality of the high-dimensional ranking feature data. This maps the high-dimensional data to a low-dimensional space, retaining the dominant components that contribute the most to the data—that is, the user features with the highest intensity and the most active operations—while removing some redundant information. PCA is then performed on the ranking feature data to obtain the dominant feature data. For example, user Zhang San has a high contribution and is considered the dominant feature. Then, based on the feature separation algorithm of PCA, the dominant feature is separated from the features of other users. For example, after separation, user Li Si's feature is click, resulting in separated feature data [user Zhang San's dominant feature, user Li Si's click feature]. By separating the feature data, the features of different users can be clearly distinguished, avoiding feature confusion.

[0127] Step S35: Associate the separated feature data with the sorted feature data to generate an identity feature vector.

[0128] Understandably, by separating the feature data, the identity characteristics of the main user and other users of the current tablet can be confirmed. The classification feature data and the ranking feature data are associated to generate an identity feature vector, which provides an identity identifier for different users and avoids misjudgment of identity caused by interference from multiple user signals. Then, the identity feature vector can be associated with the correction vector, thereby providing an adjustment basis for the final permission adjustment.

[0129] Step S4: Classify the identity feature vector, and obtain the permission update basis weight based on the identity classification result data obtained from the classification.

[0130] In this embodiment, step S4 includes:

[0131] Step S41: Import the identity feature vector and the corresponding modified behavior vector into the support vector machine algorithm, classify the identity feature vector based on the support vector machine algorithm, and obtain identity classification result data.

[0132] Understandably, after obtaining the identity feature vector through step S3 and associating it with the corrected behavior vector, the support vector machine algorithm is used for classification.

[0133] Specifically, the Support Vector Machine (SVM) algorithm is an efficient classification algorithm that can distinguish different user types through a hyperplane. It uses a stable modified behavior vector as an auxiliary classification basis, merging the identity feature vector with the corresponding modified behavior vector to form a comprehensive feature vector of identity + biometric behavior. This comprehensive feature vector is then input into the SVM algorithm. The SVM algorithm determines the user type by calculating the distance between the comprehensive vector and the hyperplane. For example, if the hyperplane is closer to the administrator, the user is classified as an administrator. Finally, the SVM algorithm outputs the identity classification result data, which is represented as the user type, such as administrator, ordinary user, etc.

[0134] Step S42: Match the identity classification result data with the pre-established user pattern matching database to obtain identity matching degree data, and confirm the permission update basis weight based on the identity matching degree data.

[0135] Specifically, the user pattern matching database is a pre-built static database used to store relevant data in the user type-standard feature template-permission weight mapping table. The standard feature template is the typical feature of the corresponding user type, while the permission weight mapping table is the correspondence rule between identity matching degree and permission weight. The higher the matching degree, the greater the weight, indicating that the user's identity is more trustworthy. The user features in the identity classification results, such as the face, voiceprint, fingerprint, trajectory, and pressure mentioned above in this embodiment, are compared with the standard feature template of the corresponding user type in the database to calculate the overlap with the standard feature template, that is, the identity matching degree data. Then, the permission update basis weight corresponding to the matching degree is queried according to the mapping table.

[0136] For example, in a user pattern matching database, the standard feature template for administrator-type users is: fingerprint 0.7-0.8, face 0.15-0.25, voiceprint 0.05-0.07, etc. The permission weight mapping table assigns a matching degree of 0.9 for 90%-100%, 0.8 for 80%-89%, and 0.7 for 70%-79%. If Zhang San's user features (fingerprint 0.72, face 0.19, voiceprint 0.06, etc.) are compared with the standard template and the overlap is calculated to be 95%, then the corresponding weight is 0.9.

[0137] Step S5: Obtain the user permission data matrix, adjust the user permission data matrix based on the permission update criteria weight, confirm the final permission result, and adjust the permissions of the tablet computer based on the final permission result.

[0138] Specifically, the user permission data matrix is ​​obtained in real time through the tablet computer system, and the user permission data matrix is ​​adjusted in real time based on the permission update criteria weight and using a row update method;

[0139] Specifically, when adjusting the user permission data matrix, adjustments are made based on a real-time queue management mechanism.

[0140] Furthermore, key update parameters are extracted based on the user permission data, and these key update parameters are used as queue inputs. The key update parameters include user activity and request frequency. Based on the key update parameters, a user priority ranking sequence is determined.

[0141] Furthermore, the user permission data matrix is ​​updated based on the user priority ranking sequence and the permission update is performed according to the weight, and an update report is generated.

[0142] Specifically, the process begins by obtaining a real-time user permission data matrix for the tablet computer. Rows in the matrix represent users, columns represent permission items, and each user value represents a permission level. For example, 0 represents no permission, 1 represents read-only permission, 2 represents edit permission, and 3 represents full control permission. Then, based on the permission update criteria and weights, the rows in the matrix are updated. A higher weight results in a greater increase in permission level; for example, a weight of 0.9 can increase permission from 2 to 3. During the adjustment, key update parameters are extracted, and the update priority for each user is determined according to these parameters. Finally, permissions are adjusted sequentially based on the user priority ranking sequence.

[0143] Specifically, the real-time user permission data matrix for tablets can be represented as follows:

[0144] ;

[0145] Taking the aforementioned real-time user permission data matrix for tablets as an example, let's assume the second column represents album permissions, the third column represents file permissions, the fourth column represents settings permissions, and the fifth column represents payment permissions. After obtaining the permission update criteria weights, Zhang San has a higher priority and is designated as user A. Based on Zhang San's permission update criteria weights, user A's (Zhang San's) permissions are adjusted. For example, album permissions are adjusted from 2 to 3, file permissions from 1 to 2, settings permissions from 2 to 3, and payment permissions from 1 to 2. After the adjustment is completed, the tablet generates an update report based on the adjustment log.

[0146] Figure 2The diagram shows some embodiments of a tablet dynamic permission management system based on biometrics and behavioral patterns that can implement the ideas of this application. The tablet dynamic permission management system based on biometrics and behavioral patterns will be described in detail below.

[0147] Specifically, the tablet dynamic permission management system based on biometrics and behavioral patterns includes:

[0148] The acquisition module is used to acquire fingerprint texture data, facial contour data, voiceprint spectrum signal, behavior sequence, user permission data matrix and multiple tablet computer signal data, and simultaneously construct a synchronous original biometric vector from fingerprint texture data, facial contour data and voiceprint spectrum signal.

[0149] A correction module is used to construct an extended initial behavior vector and obtain a corrected behavior vector after interference filtering.

[0150] The confirmation module is used to perform overlap detection on multiple tablet computer signal data to determine the identity feature vector, and associate the identity feature vector with the corresponding correction behavior vector;

[0151] A classification module is used to classify identity feature vectors and obtain the weights for permission updates.

[0152] An adjustment module is used to adjust the user permission data matrix according to the permission update basis weight and confirm the final permission result.

[0153] An allocation module is used to adjust the permissions of the tablet computer based on the final permission result.

[0154] The specific usage and function of this embodiment are explained below:

[0155] First, biometric data is obtained by collecting fingerprint texture data, facial contour data, and voiceprint spectrum signals to construct a synchronous original biometric vector. Next, behavioral sequences are acquired and extended initial behavioral vectors are constructed to obtain behavioral feature data. Filling index data is obtained through the extended initial behavioral vector, and a correction judgment operation is performed based on the filling index data to obtain a corrected behavioral vector. Simultaneously, interference filtering is applied to the corrected behavioral vector, forming a dual verification of static identity and dynamic behavior. Then, signal data from multiple tablet computers is acquired, and overlap detection is performed on the signal data to determine the identity feature vector. The identity feature vector is correlated with the corrected behavioral vector, and the mixed signals from multiple users are separated to prevent misjudgment of identity. The identity feature vector is then classified, and the permission update criterion weight is obtained based on the classification results. Finally, a user permission data matrix is ​​obtained, and the user permission data matrix is ​​adjusted based on the permission update criterion weight. The final permission result is confirmed, and the tablet computer permissions are adjusted based on the final permission result. This provides a rights management method that can identify the dominant user in a multi-user scenario with multiple users unlocking and operating the same tablet in a short period of time, combining dual verification of biometrics and behavioral patterns, and dynamically adjusting permissions.

[0156] Some embodiments of this application also provide schematic diagrams of an electronic device that can implement the ideas of this application. The following is a detailed description of this electronic device.

[0157] Specifically, an electronic device includes:

[0158] At least one processor; and at least one memory communicatively connected to the processor; wherein the memory stores instructions executable by at least one processor, the instructions being executed by at least one processor to enable at least one processor to perform the method proposed in Embodiment 1 of the present invention.

[0159] The following is a detailed introduction to the various components of the electronic device:

[0160] In this context, the processor is the control center of the electronic device. It can be a single processor or a collective term for multiple processing elements. For example, a processor can be one or more central processing units (CPUs), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement Embodiment 1 of this invention, such as one or more digital signal processors (DSPs) or one or more field-programmable gate arrays (FPGAs).

[0161] The processor can perform various functions of an electronic device by running or executing software programs stored in memory and by calling data stored in memory.

[0162] The memory is used to store the software program that executes the solution of the present invention, and the execution is controlled by the processor. For specific implementation methods, please refer to the above method embodiments, which will not be repeated here.

[0163] The memory can be a real-only memory (ROM) or other type of static storage device capable of storing static information and instructions, a random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only (CD-ROM), or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto. The memory can be integrated with the processor or exist independently and coupled to the processor through an interface circuit of an electronic device; this embodiment of the invention does not specifically limit this.

[0164] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via limited means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0165] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0166] It should be understood that, in the embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0167] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for dynamic permission management of tablets based on biometrics and behavioral patterns, characterized in that, The method includes: Collect fingerprint texture data, facial contour data, and voiceprint spectrum signals, and construct a synchronous original biometric vector; Obtain the behavior sequence and construct an extended initial behavior vector. Obtain the filling index data through the extended initial behavior vector. Perform a correction judgment operation based on the filling index data to obtain the corrected behavior vector. At the same time, perform interference filtering on the corrected behavior vector. The behavior sequence includes a pressure intensity sequence and a sliding trajectory coordinate sequence. The pressure intensity sequence is linearly filled into the synchronous original biometric vector to obtain an initial behavior vector. Based on the initial behavior vector and the sliding trajectory coordinate sequence, the initial behavior vector is expanded to obtain an extended initial behavior vector. Continuous feature data is obtained based on the extended initial behavior vector. Data is filled into the extended initial behavior vector according to the continuous feature data, and filling index data is obtained. A correction judgment operation is performed based on the filling index data. If the filling index data is greater than or equal to a preset correction threshold, the extended initial behavior vector is corrected based on a curvature correction mechanism. If the filling index data is less than the preset correction threshold, the extended initial behavior vector is corrected based on a backup correction mechanism to obtain a corrected behavior vector. Noise filtering is performed on the corrected behavior vector using a Kalman filter algorithm to obtain a filtered corrected behavior vector. Simultaneously, signal strength discrimination is performed on the filtered corrected behavior vector to obtain a stable corrected behavior vector. Acquire signal data from multiple tablet computers, perform overlap detection on the multiple tablet computer signal data, determine the identity feature vector, and associate the identity feature vector with the corrected behavior vector accordingly; The identity feature vectors are classified, and the permission update criteria weights are obtained based on the identity classification results data. Obtain the user permission data matrix, adjust the user permission data matrix based on the permission update criteria weight, confirm the final permission result, and adjust the permissions of the tablet computer based on the final permission result.

2. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 1, characterized in that, If the filling index data is greater than or equal to a preset correction threshold, the initial behavior vector of the expansion is corrected based on the curvature correction mechanism to obtain the corrected behavior vector, including: The filling index data is traversed and queried with a preset association mapping table to obtain the trajectory curvature parameter deviation; The relative influence coefficient is obtained based on the deviation between the filling index data and the trajectory curvature parameter, and the trajectory curvature parameter is corrected according to the relative influence coefficient. The sliding trajectory coordinate sequence in the extended initial behavior vector is corrected by modifying the trajectory curvature parameters, and the modified extended initial behavior vector is defined as the modified behavior vector.

3. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 1, characterized in that, If the filling index data is less than a preset correction threshold, the initial behavior vector of the expansion is corrected based on the backup correction mechanism to obtain the corrected behavior vector, including: Obtain an activation signal, and activate the tablet computer buffer based on the activation signal; Behavioral trajectory fragment data is obtained by retrieving the buffer of the tablet computer, and the basic data for reconstruction is determined based on the behavioral trajectory fragment data; Based on the reconstruction foundation, the data integrity of the extended initial behavior vector is verified, and the verification result data is obtained. The data of the verification result is used to perform data recovery and correction on the extended initial behavior vector, and the corrected extended initial behavior vector is defined as the corrected behavior vector.

4. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 1, characterized in that, Acquire signal data from multiple tablet computers, perform overlap detection on the multiple tablet computer signal data, and determine the identity feature vector, including: Acquire signal data from multiple tablet computers in a multi-user scenario, wherein the tablet computer signal data represents the signals generated by multiple users using the tablet computers; The overlapping signals in the multiple tablet computer signal data are detected, multiple overlapping region identifiers are obtained, and the signal strength of each tablet computer signal data is confirmed. Based on the signal strength of each tablet computer signal data, the multiple overlapping region identifiers are sorted to obtain sorting feature data; Dominant component analysis is performed on the sorted feature data to obtain dominant feature data. Based on the dominant component analysis method, feature separation is performed on the dominant feature data to obtain separated feature data. The separated feature data is associated with the sorted feature data to generate an identity feature vector.

5. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 1, characterized in that, The identity feature vector is classified, and the weights for updating permissions are obtained based on the identity classification results, including: The identity feature vector and the corresponding modified behavior vector are imported into the support vector machine algorithm, and the identity feature vector is classified based on the support vector machine algorithm to obtain identity classification result data. The identity classification result data is matched with a pre-established user pattern matching database to obtain identity matching degree data, and the permission update basis weight is confirmed based on the identity matching degree data.

6. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 1, characterized in that, Obtain a user permission data matrix, adjust the user permission data matrix based on the permission update criteria weights, confirm the final permission result, and adjust the tablet computer's permissions based on the final permission result, including: The system obtains a real-time user permission data matrix through a tablet computer system, and adjusts the user permission data matrix in real time based on the permission update criteria weight and using a row update method. Specifically, when adjusting the user permission data matrix, adjustments are made based on a real-time queue management mechanism.

7. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 6, characterized in that, When adjusting the user permission data matrix, adjustments are made based on a real-time queue management mechanism, including: Key update parameters are extracted from the user permission data and used as queue input. The key update parameters include user activity and request frequency. A user priority ranking sequence is determined based on the key update parameters. The user permission data matrix is ​​updated based on the user priority ranking sequence and the permission update is performed according to the weight, and an update report is generated.

8. The tablet dynamic permission management method based on biometrics and behavioral patterns according to claim 1, characterized in that, Collect fingerprint texture data, facial contour data, and voiceprint spectrum signals to construct a synchronized original biometric vector, including: The system collects fingerprint texture data, facial contour data, and voiceprint spectrum signals based on a multimodal sensor module. The acquisition timing is obtained based on the acquisition time points of the fingerprint texture data, the facial contour data, and the voiceprint spectrum signal, and the acquisition timing is processed based on the synchronization protocol. Acquire the amplitude feature data corresponding to the fingerprint texture data, the facial contour data and the voiceprint spectrum signal, normalize the amplitude feature data to a uniform scale, and obtain a standard original biometric vector. The acquisition timing sequence is imported into the standard original biological feature vector to obtain the synchronous original biological feature vector.

9. A tablet dynamic permission management system based on biometrics and behavioral patterns, used to implement the method described in any one of claims 1 to 8, characterized in that, The system includes: The acquisition module is used to acquire fingerprint texture data, facial contour data, voiceprint spectrum signal, behavior sequence, user permission data matrix and multiple tablet computer signal data, and simultaneously construct a synchronous original biometric vector from fingerprint texture data, facial contour data and voiceprint spectrum signal. A correction module is used to construct an extended initial behavior vector and obtain a corrected behavior vector after interference filtering. The confirmation module is used to perform overlap detection on multiple tablet computer signal data to determine the identity feature vector, and associate the identity feature vector with the corresponding correction behavior vector; A classification module is used to classify identity feature vectors and obtain the weights for permission updates. An adjustment module is used to adjust the user permission data matrix according to the permission update basis weight and confirm the final permission result. An allocation module is used to adjust the permissions of the tablet computer based on the final permission result.

Images