Multi-dimensional credible authentication system based on data processing

Abstract

The invention relates to the technical field of data processing, in particular to a multi-dimensional credible authentication system based on data processing, which is characterized in that environment characteristic information, position information, action track information and identity characteristic information of all police and insurance personnel are obtained by performing real-time monitoring in a preset range of a place; generating multi-dimensional credible input information, encrypting and transmitting the multi-dimensional credible input information, performing preprocessing and anomaly detection on the multi-dimensional credible input information to eliminate abnormal multi-dimensional credible input information, and constructing an authentication model based on the abnormal multi-dimensional credible input information to output a response decision; the authentication duration is counted, the authentication process is judged, the reason for disqualification of the authentication process is determined, and an instruction for adjusting the output power of a light supplementing device in the data collection process, or the updating time of a biological feature recognition model or the key updating time in the data transmission process is generated; therefore, the interference to the process of acquiring the multi-dimensional credible input information is reduced, the credibility of the multi-dimensional credible authentication process is improved, and the security and efficiency of credible authentication are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a multi-dimensional trusted authentication system based on data processing. Background Art

[0002] Multi-dimensional trusted authentication is an identity security system that integrates multiple verification dimensions. By combining different types of trusted factors for cross-verification, it significantly improves authentication reliability and attack prevention capabilities. In the joint police-security mechanism, "multi-dimensional trusted authentication" refers to the cross-verification of personnel, equipment, behavior, and data involved in joint operations through multiple technical means and institutional designs to ensure identity authenticity, operational compliance, and data trustworthiness, thereby improving the security and efficiency of joint prevention and control.

[0003] Chinese patent application publication number CN118821029B discloses a method and system for trusted identification of video network asset identities. This technical solution includes the following steps: S1, video asset acquisition and preprocessing: acquiring video data of the target video asset; S2, multimodal data acquisition: acquiring multimodal data related to the video asset; S3, multidimensional feature aggregation: performing multidimensional feature aggregation on the video data and multimodal data; S4, identity authentication feature generation: generating an identity authentication feature code for the video asset; S5, model construction and optimization: constructing a trusted identification model for video network asset identities; and S6, online identification and verification: comparing the authentication feature code with the trusted identification model for video network asset identities. While this technical solution ensures accurate identification of video assets in complex and changing environments by aggregating multidimensional features of data, it lacks identification, detection, and analysis during the data acquisition process, which reduces the credibility of the multidimensional trusted authentication process and fails to improve the security and efficiency of trusted authentication. Summary of the Invention

[0004] To this end, the present invention provides a multi-dimensional trusted authentication system based on data processing to solve the problem that the existing technology lacks identification, detection and analysis processes for the multi-dimensional trusted authentication process, which in turn leads to reduced credibility of the multi-dimensional trusted authentication process and inability to improve the security and efficiency of trusted authentication.

[0005] To achieve the above objectives, the present invention provides a multi-dimensional trusted authentication system based on data processing, comprising: The data collection layer is used to monitor the preset range of the venue in real time through the collection terminal to obtain environmental feature information, the location information and movement trajectory information of each police and security personnel, and the identity feature information obtained through the equipment worn by each police and security personnel, and to generate multi-dimensional trusted input information through weighted fusion; A data transmission layer, connected to the data acquisition layer, for encrypting and transmitting the generated multi-dimensional trusted input information; a data processing layer connected to the data transmission layer and configured to pre-process the multi-dimensional trusted input information; a data detection layer connected to the data processing layer, configured to perform anomaly detection on the pre-processed multi-dimensional credible input information to eliminate abnormal multi-dimensional credible input information; A data decision layer, connected to the data detection layer, is used to build an authentication model based on multi-dimensional trusted input information that eliminates anomalies and output a response decision, which includes establishing multi-level emergency actions; An authentication statistics layer, which is connected to the data collection layer and the data decision layer respectively, and is used to count the authentication time, wherein the authentication time is the time taken from obtaining each data information to outputting the response decision process; an analysis layer connected to the authentication statistics layer, configured to determine whether the authentication process is qualified based on the authentication duration, determine the reason for failure of the authentication process based on the authentication duration, and generate an instruction for adjusting the output power of a fill light device during data collection, or the update time of a biometric recognition model during data collection, or the key update time during data transmission; The control layer is connected to the analysis layer and the control layer respectively, and is used to determine the adjustment of the output power of the fill light device, the update time of the biometric recognition model, or the key update time based on the instruction.

[0006] Furthermore, the authentication statistics layer is also connected to the data detection layer to calculate the percentage of abnormal elimination; The analysis layer is further configured to determine that the authentication process is qualified if a comparison result between the authentication time and a preset authentication time meets a first determination condition; The analysis layer is further configured to re-determine the authentication process based on the abnormal rejection ratio when the comparison result between the authentication duration and the authentication duration does not meet the first determination condition but meets the second determination condition; The analysis layer is further configured to determine that the authentication process is unqualified if the comparison result between the authentication time and the preset authentication time does not meet the second determination condition, and to determine the reason for the unqualified process and corresponding treatment based on the difference between the authentication time and the preset authentication time; Among them, the preset authentication time includes a first preset authentication time and a second preset authentication time that is greater than the first preset authentication time, the first judgment condition is that the authentication time is less than or equal to the first preset authentication time, and the second judgment condition is that the authentication time is greater than the first preset authentication time and less than or equal to the second preset authentication time.

[0007] Furthermore, the analysis layer is further configured to re-determine the authentication process based on a comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio; The analysis layer determines that there is an abnormality in the light when the data collection layer performs real-time monitoring based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio, and adjusts the fill light device of the terminal or determining that the iris matching error rate increases when the data acquisition layer performs real-time monitoring, and adjusting the updating time of the biometric recognition model; The analysis layer is also used to determine the cause of the failure and corresponding treatment based on the authentication time offset when determining that the authentication process fails; The authentication duration offset is the difference between the authentication duration and the second preset authentication duration.

[0008] Furthermore, the data collection layer is also used to obtain light values ​​within a preset range of the place; The analysis layer is further configured to adjust the output power of the fill light device based on obtaining the light value and comparing it with a preset light value, wherein the preset light value includes a preset light lower limit value and a preset light upper limit value; The analysis layer determines to increase the output power of the fill light device when the light value is less than the preset light lower limit; The analysis layer determines to reduce the output power of the fill light device when the light value is greater than the preset light upper limit value; The analysis layer determines to maintain the output power of the fill light device when the light value is greater than or equal to the preset light lower limit value and less than or equal to the preset light upper limit value.

[0009] Furthermore, when the analysis layer determines to increase the output power of the fill light device, it is further configured to generate an instruction to increase the output power based on a comparison result of the light offset value with a preset light offset value, and the increase in the output power is positively correlated with the light offset value. The light offset value is the difference between the preset light value and the light value.

[0010] Furthermore, when determining to reduce the output power of the fill light device, the analysis layer is further configured to generate an instruction to reduce the output power based on a comparison result of the light difference value with a preset light difference value, and the reduction range of the output power is positively correlated with the light difference value; The light difference is a difference between the light value and the preset light value.

[0011] Furthermore, the analysis layer is also used to generate an instruction to shorten the update time of the biometric recognition model based on the comparison result of the iris matching error rate and the preset iris matching error rate, and the shortening degree of the update time is positively correlated with the iris matching error rate.

[0012] Further, when the control layer completes the control based on an instruction to increase the output power of the fill light device or based on an instruction to reduce the output power of the fill light device, the analysis layer still determines that there is an abnormality in the light when the data acquisition layer performs real-time monitoring based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio; or when the control layer completes the control based on an instruction to shorten the update time of the biometric recognition model, the analysis layer still determines that the iris matching error rate increases when the data acquisition layer performs real-time monitoring based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio, the analysis layer re-determines the cause of the failure and the corresponding processing based on the authentication time offset.

[0013] Furthermore, the analysis layer is further configured to determine a reason for failure of the authentication process based on a comparison result of the authentication time offset with the preset authentication time offset and generate a corresponding instruction, including: If it is determined that the reason for the failure is that there is a transmission anomaly in the data transmission layer, resulting in the comparison result of the authentication time with the preset authentication time not meeting the second determination condition, the analysis layer determines that the network for transmitting information is abnormal and generates an instruction to issue a network maintenance; If it is determined that the reason for failure is that the authentication model constructed in the data decision layer has an anomaly, resulting in the comparison result between the authentication time and the preset authentication time not meeting the second determination condition, the analysis layer determines that the authentication model has an anomaly and generates an instruction to reconstruct the authentication model; If it is determined that the reason for failure is that there is an external data attack resulting in the comparison result of the authentication duration with the preset authentication duration not meeting the second determination condition, the analysis layer is further configured to adjust the key update time based on the second-order difference of the authentication duration; The second-order difference of the authentication duration is the difference between the authentication duration offset and the preset authentication duration offset.

[0014] Furthermore, the analysis layer is also used to generate an instruction to shorten the key update time based on the comparison result of the second-order difference of the authentication time and the preset second-order difference of the authentication time, and the shortening degree of the key update time is positively correlated with the second-order difference of the authentication time.

[0015] Compared with the prior art, the beneficial effect of the multi-dimensional trusted authentication system based on data processing of the present invention is that the system performs real-time monitoring of the preset range of the venue through the data acquisition layer to obtain environmental feature information, location information and movement trajectory information of each police and security personnel, and identity feature information, based on which multi-dimensional trusted input information is generated and encrypted for transmission, and the multi-dimensional trusted input information is pre-processed; then anomaly detection is performed on the pre-processed multi-dimensional trusted input information to eliminate abnormal multi-dimensional trusted input information, and based on this, an authentication model is constructed, and the authentication model outputs a response decision; the time taken from generating the multi-dimensional trusted input information to constructing the authentication model is counted, and the authentication process is judged based on the authentication time, and the reason for the failure of the authentication process is determined based on the authentication time, and an instruction is generated to adjust the output power of the fill light device in the data acquisition process, or the update time of the biometric recognition model in the data acquisition process, or the key update time in the data transmission process, and the adjustment is performed based on the instruction, thereby reducing interference with the process of obtaining multi-dimensional trusted input information, improving the credibility of the multi-dimensional trusted authentication process, and improving the security and qualified efficiency of the trusted authentication.

[0016] Furthermore, the present invention further uses the abnormal rejection ratio and the preset abnormal rejection ratio to make a secondary judgment on the authentication process based on the comparison between the authentication time and the preset authentication time, so as to reduce misjudgment and improve the judgment accuracy, and adjusts the output power of the fill light device of the data acquisition layer or the update time of the biometric recognition model according to the judgment result to reduce the authentication time, thereby improving the security and efficiency of multi-dimensional trusted authentication.

[0017] Furthermore, the present invention also determines the cause of the failure based on the comparison result of the authentication time offset and the preset authentication time offset when determining that the authentication process is unqualified, and determines the corresponding processing based on the cause, including: when it is determined that there is a transmission abnormality in the data transmission layer, the analysis layer determines that the network abnormality of the transmission information is generated and generates an instruction to issue a network maintenance; when it is determined that there is an abnormality in the authentication model constructed in the data decision layer, the analysis layer determines that there is an abnormality in the authentication model and generates an instruction to rebuild the authentication model; when it is determined that it is caused by an external attack, the analysis layer adjusts the key update time based on the comparison result of the second-order difference of the authentication time and the preset second-order difference of the authentication time; such a setting can accurately correct the authentication process with abnormalities, thereby reducing the authentication time, thereby improving the security and efficiency of multi-dimensional trusted authentication. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a module diagram of a multi-dimensional trusted authentication system based on data processing in the present invention; Figure 2 A flowchart of a multi-dimensional trusted authentication method based on data processing in the present invention; Figure 3 This is a logic decision diagram for determining whether the authentication process is qualified based on the authentication duration in the present invention; Figure 4 A logical decision diagram for re-determining the authentication process and determining the corresponding processing based on the abnormal rejection ratio in the present invention; Figure 5 This is a logical decision diagram for determining the reason for failure of the authentication process and the corresponding processing based on the authentication time offset in the present invention. DETAILED DESCRIPTION

[0019] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0020] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0021] It should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the term "connection" should be understood in a broad sense. For example, it can mean a fixed connection, a detachable connection, or an integral connection; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0022] See also Figure 1As shown, it is a module diagram of a multi-dimensional trusted authentication system based on data processing in this embodiment. The system includes a data acquisition layer, a data transmission layer, a data processing layer, a data detection layer, a data decision layer, an authentication statistics layer, an analysis layer and a control layer. The data acquisition layer is used to monitor the preset range of the venue in real time through the acquisition terminal to obtain environmental feature information, the location information and movement trajectory information of each police and security personnel, and to obtain identity feature information through the equipment worn by each police and security personnel, so as to generate multi-dimensional trusted input information through weighted fusion; the data transmission layer is connected to the data acquisition layer to encrypt and transmit the generated multi-dimensional trusted input information; the data processing layer is connected to the data transmission layer to pre-process the multi-dimensional trusted input information; the data detection layer is connected to the data processing layer to perform anomaly detection on the pre-processed multi-dimensional trusted input information to eliminate abnormal multi-dimensional trusted input information; the data decision layer is connected to the data detection layer to construct an authentication system based on the multi-dimensional trusted input information that has eliminated abnormalities. The authentication model is connected to the data acquisition layer and the data decision layer respectively, and is used to count the authentication time, wherein the authentication time is the time taken from obtaining each data information to outputting the response decision; the analysis layer is connected to the authentication statistics layer, and is used to determine whether the authentication process is qualified based on the authentication time, and determine the reason for the failure of the authentication process based on the authentication time and generate an instruction to adjust the output power of the terminal's fill light device, or the update time of the biometric recognition model, or the key update time; the control layer is connected to the analysis layer and the data decision layer respectively, and is used to determine the adjustment of the output power of the fill light device, or the update time of the biometric recognition model, or the key update time based on the instruction.

[0023] See also Figure 2 As shown in FIG, it is a flow chart of a multi-dimensional trusted authentication method based on data processing in this embodiment. The process includes at least the following steps: S1: Use the data collection terminal to conduct real-time monitoring of the preset range of the venue to obtain environmental feature information, the location information and movement trajectory information of each police and security personnel, and the identity feature information obtained through the equipment worn by each police and security personnel, and then generate multi-dimensional trusted input information through weighted fusion; S2: Multi-dimensional trusted input information generated by encrypted transmission; S3: Preprocessing of multi-dimensional credible input information; S4: performing anomaly detection on the pre-processed multi-dimensional credible input information to eliminate abnormal multi-dimensional credible input information; S5: Build an authentication model based on multi-dimensional trusted input information without anomalies and output a response decision, which includes establishing multi-level emergency actions; S6: Count the authentication time, where the authentication time is the time taken from obtaining each data information to outputting the response decision; S7: Determining whether the authentication process is qualified based on the authentication duration, determining the reason for the authentication process being unqualified based on the authentication duration, and generating an instruction for adjusting the output power of the fill light device during the data acquisition process, or the biometric recognition model update time during the data acquisition process, or the key update time during the data transmission process; S8: Determine, based on the instruction, whether to adjust the output power of the fill light device, or the update time of the biometric recognition model, or the update time of the key.

[0024] Specifically, in this embodiment, taking the example of early warning of emergencies within a preset location, the goal is to improve the security and efficiency of multi-dimensional trusted authentication through data processing during the information exchange between security personnel and the command center. The real-time monitoring terminal can use a drone equipped with a high-definition camera to capture images of each divided area within the preset location. The camera then identifies and analyzes the images to determine environmental characteristic information within the preset location. Environmental characteristic information includes crowd density, air anomaly data, and so on. Areas where crowd density exceeds a certain threshold are monitored, and areas where air anomaly data exceeds a certain threshold are also monitored. The areas where each security personnel within the preset location are monitored to determine their location information and movement trajectory information. In the event of an emergency, the command center can determine the response decision-making and planning process based on the location information, movement trajectory information, and environmental characteristic information of each security personnel. The identity of each officer is also obtained through portable devices worn by the officers (such as iris detection devices or fingerprint recognition devices). This identity authentication distinguishes each officer's unit category, including Class I and Class II emergency units. Class I emergency units are those capable of providing immediate reinforcement, while Class II emergency units are those that remain in their designated area awaiting reinforcement. This also serves to verify the true identity of each officer and prevent the infiltration of unauthorized personnel. The collected important information from various dimensions is weighted and fused to generate multi-dimensional trusted input information. Weighted fusion aims to improve the transmission reliability of more important information through dynamic weight allocation. During data transmission, encryption algorithms with higher weights are selected for more difficult-to-crack information within the multi-dimensional trusted input. This allows for targeted encrypted transmission, reducing transmission load and collaboration costs. Furthermore, the independent use of encryption algorithms allows for strict control over key access rights. Preprocessing is performed at the data processing layer, including data cleansing to address data quality issues and ensure accuracy, and data integration to merge data from multiple sources and eliminate redundancy and conflicts. In the data detection layer, when performing anomaly detection on pre-processed multi-dimensional trusted input information, any abnormal data information collected or generated during transmission is first eliminated to prevent it from affecting the subsequent construction of the authentication model and the generation of response decisions. In the data decision layer, an authentication model is constructed based on the multi-dimensional trusted input information. This model is a multi-factor authentication model. When building this model, a machine learning model is selected and combined with deep learning training data. Because this authentication model has undergone trusted authentication, it should meet security and reliability requirements. Response decisions are generated through the authentication model, including multiple levels of emergency actions, such as first-level emergency actions and second-level emergency actions.The duration of the entire process from data information collection to output response decision is counted and recorded as authentication time, and then the authentication process is judged to determine whether it is qualified based on the comparison between the authentication time and the preset authentication time. If it is qualified, the current process from obtaining data information to outputting response decision will not be changed, and the existing authentication process and response decision will be maintained. If it is unqualified, the reason for the unqualified is determined based on the authentication time and the corresponding adjustment method is generated, and then the various parameters in the entire process are adjusted to ensure that the authentication process is qualified and a new response decision is output; by reducing the interference factors in the process of obtaining multi-dimensional trusted input information, the credibility of the multi-dimensional trusted authentication process is improved and the security and qualified efficiency of the authentication are improved.

[0025] See also Figure 3 As shown, it is a logical decision diagram for determining whether the authentication process is qualified based on the authentication time and performing corresponding processing based on the abnormal elimination ratio in this embodiment. The authentication statistics layer is also connected to the data detection layer to count the abnormal elimination ratio; the analysis layer is also used to determine that the authentication process is qualified based on the result of the comparison between the authentication time and the preset authentication time, if it meets the first judgment condition; the analysis layer is also used to re-determine the authentication process based on the abnormal elimination ratio if the result of the comparison between the authentication time and the authentication time does not meet the first judgment condition but meets the second judgment condition; the analysis layer is also used to determine that the authentication process is not qualified based on the result of the comparison between the authentication time and the preset authentication time, if it does not meet the second judgment condition. Qualified, and the reasons for failure and corresponding treatment are determined based on the authentication time offset; wherein, the abnormal elimination ratio is the ratio of the number of information eliminated in the multi-dimensional trusted input information to the total number of input information, the preset authentication time includes a first preset authentication time and a second preset authentication time that is greater than the first preset authentication time, the first judgment condition is that the authentication time is less than or equal to the first preset authentication time, the second judgment condition is that the authentication time is greater than the first preset authentication time and less than or equal to the second preset authentication time, and the authentication time offset is the difference between the authentication time and the second preset authentication time.

[0026] Specifically, in this embodiment, the data detection layer is first counted and analyzed by the authentication statistics layer to determine the abnormal rejection ratio Q generated when the data detection layer performs abnormal detection. Then, based on the change of the abnormal rejection ratio Q, the authentication time W is compared with the preset authentication time W0 to re-determine whether there is a problem in the current authentication process, and when it is determined that there is a problem, corresponding adjustments are made based on Q. In order to more accurately analyze and determine the authentication process and refine the corresponding adjustment method, W0 can be divided into a first preset authentication time W1 and a second preset authentication time W2. Based on the parameters that meet the judgment criteria for the historical authentication process and combined with the statistical analysis method, the subsequent preset or critical parameter values ​​can be determined. W1=280ms, W2=310ms can be set. W1 is the ideal authentication time upper limit, which is the reasonable maximum time required for complete authentication. W2 is the risk tolerance authentication time upper limit, which is the absolute maximum time to complete authentication. If it exceeds, the current authentication process is directly judged to be unqualified. The comparison process based on W with W1 and W2 is as follows: If W is less than or equal to W1, the comparison result is determined to be in compliance with the first judgment condition, indicating that the authentication process is within the ideal authentication time. Therefore, the current authentication process is directly determined to be qualified, and the corresponding response decision is directly output. If W is greater than W1 and less than or equal to W2, the comparison result is determined to be in compliance with the first judgment condition, but in compliance with the second judgment condition, indicating that during the authentication process, the current authentication time W is out of the ideal range, but has not reached the dangerous threshold. At this time, there are other low-abnormal factors interfering. A new parameter, the abnormal elimination ratio Q, can be introduced to make further judgments based on Q. The collaborative judgment of the new parameter can avoid relying on a single indicator, thereby improving the accuracy of the judgment process. If the value of Q is too large, it indicates that too much abnormal data information has been eliminated in the current authentication process, indicating that there are large abnormalities in the data information collection and transmission process, especially when there are problems in the data collection process, which will cause a sudden increase in the amount of abnormal data information. If W is greater than W2, the comparison result is determined to meet the second judgment condition, indicating that there is a serious authentication timeout in the authentication process. Therefore, the current authentication process is directly judged to be unqualified, and the corresponding response decision cannot be directly output. At this time, it is necessary to calculate the difference between W and W2 and record it as the authentication time offset K. Based on K, the reason for the failure and the corresponding handling method are determined.

[0027] See also Figure 4As shown, this is a logical decision diagram for re-determining the authentication process and determining the corresponding processing based on the abnormal rejection ratio in this embodiment. The analysis layer is also used to re-determine the authentication process based on the comparison result of the abnormal rejection ratio with the preset abnormal rejection ratio; the analysis layer determines based on the comparison result of the abnormal rejection ratio with the preset abnormal rejection ratio that there is an abnormal light when the data acquisition layer is performing real-time monitoring, and adjusts the terminal's fill light device; or determines that the iris matching error rate increases when the data acquisition layer is performing real-time monitoring, and adjusts the update time of the biometric recognition model.

[0028] Specifically, in this embodiment, when the data acquisition layer is performing image acquisition, excessively strong / weak light within a preset range of the location can cause problems with image acquisition, leading to an abnormal increase in multi-dimensional trusted input information. In this case, the terminal's fill light device can be adjusted to adapt to the changes in ambient light during image acquisition, thereby ensuring better image quality, making the obtained multi-dimensional trusted input information more reliable, and thus reducing the authentication time. When the error rate in the iris matching process of obtaining identity feature information in the data acquisition layer suddenly increases, it will lead to an abnormal increase in multi-dimensional trusted input information. For example, when a police officer's eyes are locally swollen after being on duty (wearing goggles for a long time), it will affect biometric recognition. In this case, the update time of the biometric recognition model can be adjusted to reduce the iris matching error rate, thereby making the obtained multi-dimensional trusted input information more reliable, and thus reducing the authentication time.

[0029] The authentication process can be re-evaluated based on the comparison result of the abnormal rejection ratio Q and the preset abnormal rejection ratio Q0. At this time, Q0 can be divided into the first preset abnormal rejection ratio Q1 and the second preset abnormal rejection ratio Q2, and Q1=2% and Q2=3% are set. The values ​​here are calculated based on the total amount of data information obtained by the data acquisition layer and the number of abnormal data eliminated by the data detection layer in unit time. The specific comparison process based on Q with Q1 and Q2 is as follows. If Q is less than or equal to Q1, it means that the number of abnormal data currently eliminated is within an acceptable range. At this time, even if the comparison result of W with W1 and W2 does not meet the first judgment condition but meets the second judgment condition, it means that the abnormal data in the total amount of data information obtained during the acquisition process or the transmission process is relatively small. Therefore, it is also acceptable for the current authentication process and there is no need to adjust the authentication process. If Q is greater than Q1 and less than or equal to Q2, it means that the number of abnormal data eliminated during the current data detection process exceeds the acceptable range, but is still within a relatively acceptable and reasonable range. Since the identity feature information in the multi-dimensional trusted input information accounts for a relatively small proportion, it can be determined that the increase in the iris matching error rate during the data collection process causes the Q value to increase. At this time, the update time of the biometric recognition model in the wearable device in the data collection layer can be adjusted to reduce abnormal data information. If Q is greater than Q2, it means that the number of abnormal data eliminated during the current data detection process is too large. It can be determined that the abnormal light in the data collection process causes many problems with the collected image information, thereby causing the Q value to increase. At this time, the fill light device of the collection terminal in the data collection layer can be adjusted to reduce abnormal data information.

[0030] Furthermore, the data acquisition layer is also used to obtain the light value within a preset range of the venue; the analysis layer is also used to adjust the output power of the fill light device based on the obtained light value and comparing it with the preset light value, wherein the preset light value includes a preset light lower limit and a preset light upper limit; the analysis layer determines to increase the output power of the fill light device when the light value is less than the preset light lower limit; the analysis layer determines to reduce the output power of the fill light device when the light value is greater than the preset light upper limit; the analysis layer determines to maintain the output power of the fill light device when the light value is greater than or equal to the preset light lower limit and less than or equal to the preset light upper limit.

[0031] Specifically, in this embodiment, when it is determined that the fill light device in the data collection layer needs to be adjusted, the data collection layer captures the light within the preset range of the venue to determine the corresponding light value L. The light value L can be compared with the preset light value L0 to determine the impact of the light change within the preset range of the venue on data collection, and then determine whether to increase or decrease the output power of the fill light device. The preset light value L0 can be divided into a preset light lower limit value L1 and a preset light upper limit value L2. L1 is the minimum illumination requirement for no degradation of image quality, and L2 is the maximum illumination tolerance value to avoid overexposure. L1 is set to 700 lux and L2 is set to 900 lux. L1 and L2 can be set to other appropriate values ​​according to different preset ranges of the venue. The specific comparison process based on L, L1 and L2 is as follows: If L is less than L1, it means that the light value of the current area known by the current data acquisition layer is relatively small, which exceeds the minimum illumination requirement for image acquisition quality not to deteriorate during the current multi-dimensional authentication process. Therefore, you can choose to increase the output power of the fill light device. If L is greater than L2, it means that the light value L of the current area known by the current data acquisition layer is relatively large, which exceeds the maximum illumination tolerance value for overexposure required for image acquisition during the current multi-dimensional authentication process. Therefore, you can choose to reduce the output power of the fill light device. If L is greater than or equal to L1 and less than or equal to L2, it means that the light value L of the current area known by the current data acquisition layer is relatively appropriate, and you can choose to maintain the output power of the current fill light device.

[0032] In other embodiments, the analysis layer may also compare the acquired light value L with the preset light value L0 to determine whether to process the light noise of the captured image. For example, even if the fill light power is adjusted, the fill light source may be uneven, have color cast, or produce strong reflections / glare at certain angles. These may also be regarded as some kind of "noise" or interference information, and the image light noise needs to be processed at this time; the analysis layer may determine whether to trigger the light noise processing module for adjusting the image signal processing in the system based on the comparison result of the detected light value L with the preset light value L0. When L is less than L1, the current image light noise is generated in low light conditions. The primary concern is excessive random noise caused by weak signals. The light noise processing module can be used to reduce image noise using a time-domain noise reduction method. Based on the principle that darker light conditions result in higher gain and greater noise, the light noise processing module should adjust the noise reduction power to a higher level. When L is greater than L2, the current image light noise is caused by strong reflections or direct light in bright light conditions. In this case, the scene illumination chromaticity can be estimated to correct for color shifts in highlight areas. Based on the principle that brighter light conditions result in greater color shifts, the light noise processing module should adjust the color correction to a higher level. When L is greater than or equal to L1 and less than or equal to L2, the current image light noise processing method is maintained.

[0033] Furthermore, when determining to increase the output power of the fill light device, the analysis layer is also used to determine and generate an instruction to increase the output power based on the comparison result of the light offset value and the preset light offset value, and the increase in the output power is also positively correlated with the light offset value; wherein, the light offset value is the difference between the preset light value and the light value.

[0034] Specifically, in this embodiment, the light offset value H is the difference between L0 and L, more specifically, the difference between L1 and L. When the light offset value H is larger, the light value L is smaller, and the light intensity required during image acquisition is higher to more easily restore the true layering. In this case, the output power of the fill light device needs to be larger. Therefore, the increase in the output power of the fill light device is positively correlated with the light offset value. In order to more accurately determine the increase in the output power of the fill light device, the preset light offset value H0 can be divided into a first preset light offset value H1 and a second preset light offset value H2, with H1 set to 50 lux and H2 set to 100 lux. The comparison process based on H with H1 and H2 is as follows: If H is less than or equal to H1, the analysis layer determines to generate a corresponding first fill light device output power adjustment coefficient instruction. Based on the instruction, the control layer controls the output power of the fill light device of the acquisition terminal in the data acquisition layer to increase by 20% based on the original output power. If H is greater than H1 and less than or equal to H2, the analysis layer determines to generate a corresponding second fill light device output power adjustment coefficient instruction. Based on the instruction, the control layer controls the output power of the fill light device of the acquisition terminal in the data acquisition layer to increase by 30% based on the original output power. If H is greater than H2, the analysis layer determines to generate a corresponding third fill light device output power adjustment coefficient instruction. Based on the instruction, the control layer controls the output power of the fill light device of the acquisition terminal in the data acquisition layer to increase by 40% based on the original output power. It is understood that the increase in the output power of the fill light device can also be set to other values ​​that meet the requirements, for example, when H is less than or equal to H1, the output power is increased by 25% based on the original output power. It should be noted that the increase in the output power of the fill light device is limited to a value that does not negatively impact the image acquisition process.

[0035] Furthermore, when the analysis layer determines to reduce the output power of the fill light device, it is also used to generate an instruction to reduce the output power based on the comparison result of the light difference value and the preset light difference value, and the reduction amplitude of the output power is also positively correlated with the light difference value; wherein, the light difference value is the difference between the light value and the preset light value.

[0036] Specifically, in this embodiment, the light difference F is the difference between L and L0, more specifically, the difference between L and L2. When the light difference F is larger, the light value L is larger, and the light intensity required during image acquisition is lower to prevent exposure. At this time, the output power of the fill light device needs to be smaller. Therefore, the reduction range of the output power of the fill light device is positively correlated with the light difference. In order to more accurately determine the reduction range of the output power of the fill light device, the preset light difference F0 can be divided into a first preset light difference F1 and a second preset light difference F2. Set F1=70lux and F2=110lux. The comparison process based on F with F1 and F2 is as follows: If F is less than or equal to F1, the analysis layer determines and generates a corresponding fourth fill light device output power adjustment coefficient instruction. Based on the instruction, the control layer controls the output power of the fill light device of the acquisition terminal in the data acquisition layer to decrease by 15% from the original output power. If F is greater than F1 and less than or equal to F2, the analysis layer determines and generates a corresponding fifth fill light device output power adjustment coefficient instruction. Based on the instruction, the control layer controls the output power of the fill light device of the acquisition terminal in the data acquisition layer to decrease by 25% from the original output power. If F is greater than F2, the analysis layer determines and generates a corresponding sixth fill light device output power adjustment coefficient instruction. Based on the instruction, the control layer controls the output power of the fill light device of the acquisition terminal in the data acquisition layer to decrease by 35% from the original output power. It is understood that the reduction in the output power of the fill light device can also be set to other values ​​that meet requirements, for example, when F is greater than F2, the output power is reduced by 40% from the original output power. It should be noted that the reduction in the output power of the fill light device is limited to a value that does not negatively impact the image acquisition process.

[0037] Furthermore, the analysis layer is also used to generate an instruction to shorten the update time of the biometric recognition model based on the comparison result of the iris matching error rate and the preset iris matching error rate, and the shortening degree of the update time is positively correlated with the iris matching error rate.

[0038] Specifically, in this embodiment, the iris matching error rate D is based on the ratio of the number of first authentication failures to the total number of authentications when each security guard within a preset range performs iris authentication at the current location. When the iris matching error rate D is larger, the biometric recognition model needs to be updated more frequently. By shortening the update time of the biometric recognition model, it is more suitable for the identity authentication of the security guard. In order to more accurately determine the extent of the reduction in the update time of the biometric recognition model, the preset iris matching error rate D0 can be divided into a first preset iris matching error rate D1 and a second preset iris matching error rate D2, with D1 set to 5% and D2 set to 10%. The specific process of comparing D with D1 and D2 is as follows: If D is less than or equal to D1, the analysis layer determines and generates a corresponding first model update time adjustment coefficient instruction, shortening the original update time by 5%. If D is greater than D1 and less than or equal to D2, the analysis layer determines and generates a corresponding second model update time adjustment coefficient instruction, shortening the original update time by 15%. If D is greater than D2, the analysis layer determines and generates a corresponding third model update time adjustment coefficient instruction, shortening the original update time by 30%. It is understood that the reduction in the update time of the biometric recognition model can also be set to other values ​​that meet the requirements, for example, when D is greater than D2, the original output power is reduced by 25%. It should be noted that the reduction in the update time of the biometric recognition model is limited to not negatively impacting the identity authentication process.

[0039] Furthermore, when the control layer completes the control based on an instruction to increase the output power of the fill light device or based on an instruction to reduce the output power of the fill light device, the analysis layer still determines that there is an abnormality in the light when the data acquisition layer performs real-time monitoring based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio; or when the control layer completes the control based on an instruction to shorten the update time of the biometric recognition model, the analysis layer still determines that the iris matching error rate increases when the data acquisition layer performs real-time monitoring based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio, the analysis layer re-determines the cause of the failure and the corresponding processing based on the authentication time offset.

[0040] Specifically, in this embodiment, if it is determined that the output power of the fill light device is adjusted accordingly, and the abnormal rejection ratio Q re-acquired by the analysis layer is still greater than the second preset abnormal rejection ratio Q2; or after the update time of the biometric recognition model is adjusted accordingly, the abnormal rejection ratio Q re-acquired by the analysis layer is still greater than the first abnormal rejection ratio Q1 and less than or equal to the second abnormal rejection ratio Q2, it means that after the above two methods of adjustment, there is still a situation where W is greater than W1 and less than or equal to W2. At this time, it is also necessary to re-determine the cause of failure and corresponding processing based on the authentication time offset K.

[0041] See also Figure 5 As shown, it is a logical decision diagram for determining the reason for authentication failure based on the authentication time offset and the corresponding processing in this embodiment. The analysis layer is further used to determine the reason for authentication failure based on the comparison result of the authentication time offset and the preset authentication time offset and generate a corresponding instruction, including: if it is determined that the reason for failure is a transmission anomaly in the data transmission layer, resulting in the comparison result of the authentication time with the preset authentication time not meeting the second determination condition, the analysis layer determines that the network for transmitting information is abnormal and generates an instruction to issue a network maintenance; if it is determined that the reason for failure is a transmission anomaly in the authentication model constructed in the data decision layer, resulting in the comparison result of the authentication time with the preset authentication time not meeting the second determination condition, the analysis layer determines that the authentication model is abnormal and generates an instruction to rebuild the authentication model; if it is determined that the reason for failure is an external data attack, resulting in the comparison result of the authentication time with the preset authentication time not meeting the second determination condition, the analysis layer is further used to adjust the key update time based on the second-order difference of the authentication time; wherein the second-order difference of the authentication time is the difference between the authentication time offset and the preset authentication time offset.

[0042] Specifically, in this embodiment, the reason for the failure of the authentication process is analyzed from the perspective of comparing the authentication time offset K with the preset authentication time offset K0. It is assumed that other situations will not affect the analysis process. In order to more accurately determine the cause, the preset authentication time offset K0 can be divided into a first preset authentication time offset K1 and a second preset authentication time offset K2. K1=40ms and K2=80ms can be set. The comparison process based on K with K1 and K2 is as follows: If K is less than or equal to K1, the difference between the current W and W2 is relatively small, indicating a relatively small impact on the authentication process. The analysis layer can determine that a delay in the information transmission network has slowed the authentication process, affecting the actual W. In this case, the analysis layer can generate a network maintenance instruction to reduce the impact of network delays, thereby reducing W. An anomaly in the information transmission network only prolongs a single authentication process and does not affect the system's core logic or data integrity. Therefore, the impact on the authentication process is relatively low. If K is greater than K1 and less than or equal to K2, the difference between the current W and W2 is relatively moderate, indicating a moderate impact on the authentication process. The analysis layer can determine that an anomaly in the constructed authentication model, such as improper model parameter settings, has slowed the authentication process and affected the actual W. In this case, the analysis layer can generate an instruction to rebuild the authentication model to reduce the impact of the anomaly, thereby reducing W. An anomaly in the authentication model affects all requests, leading to a continuous decrease in efficiency or misjudgment. Therefore, the impact on the authentication process is moderate. If K is greater than K2, it means that the difference between the current W and W2 is relatively large, which indicates that the impact on the authentication process is relatively large. The analysis layer can determine that there is an external attack in the transmission process, such as a hacker attack, DDoS attack, or malicious data injection, which makes the data information become abnormal data, interfering with the authentication process, thereby affecting the actual W. At this time, the analysis layer can generate an instruction to adjust the key update time, and adjust the key update time based on the second-order difference N of the authentication time, where N is the difference between K and K0, more specifically the difference between K and K2. By shortening the key update time to reduce the impact of external attacks, W is reduced; external attacks may paralyze the entire authentication service, and therefore have a high degree of impact on the authentication process.

[0043] Furthermore, the analysis layer is also used to generate an instruction to shorten the key update time based on the comparison result of the second-order difference of the authentication time and the preset second-order difference of the authentication time, and the shortening degree of the key update time is positively correlated with the second-order difference of the authentication time.

[0044] Specifically, in this embodiment, in order to more accurately determine the extent of shortening the key update time, the preset authentication time second-order difference N0 can be divided into a first preset authentication time second-order difference N1 and a second preset authentication time second-order difference N2. N1 can be set to 25ms and N2 to 40ms. The comparison process based on the authentication time second-order difference N with N1 and N2 is as follows: If N is less than or equal to N1, the analysis layer determines and generates a corresponding first key update time adjustment coefficient instruction, shortening the original key update time by 10%. If N is greater than N1 and less than or equal to N2, the analysis layer determines and generates a corresponding second key update time adjustment coefficient instruction, shortening the original key update time by 15%. If N is greater than N2, the analysis layer determines and generates a corresponding third key update time adjustment coefficient instruction, shortening the original key update time by 20%. It is understood that the reduction in the key update time can also be set to other values ​​that meet the requirements, for example, when N is greater than N2, the reduction in the original key update time is 25%. It should be noted that the reduction in the key update time is limited to not negatively impact the authentication process.

[0045] It is understandable that in the embodiments of the present invention, no specific limitation is imposed on any preset parameter or critical parameter, and the above values ​​are not limited thereto. Those skilled in the art may adjust the preset parameters or critical parameters accordingly based on actual needs, analysis of historical data, or equipment usage.

[0046] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0047] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A multi-dimensional trusted authentication system based on data processing, characterized in that: include: The data collection layer is used to monitor the preset range of the venue in real time through the collection terminal to obtain environmental feature information, the location information and movement trajectory information of each police and security personnel, and the identity feature information obtained through the equipment worn by each police and security personnel, and to generate multi-dimensional trusted input information through weighted fusion; A data transmission layer, connected to the data acquisition layer, for encrypting and transmitting the generated multi-dimensional trusted input information; a data processing layer connected to the data transmission layer and configured to pre-process the multi-dimensional trusted input information; a data detection layer connected to the data processing layer, configured to perform anomaly detection on the pre-processed multi-dimensional credible input information to eliminate abnormal multi-dimensional credible input information; A data decision layer, connected to the data detection layer, is used to build an authentication model based on multi-dimensional trusted input information that eliminates anomalies and output a response decision, which includes establishing multi-level emergency actions; An authentication statistics layer, which is connected to the data collection layer and the data decision layer respectively, and is used to count the authentication time, wherein the authentication time is the time taken from obtaining each data information to outputting the response decision process; an analysis layer connected to the authentication statistics layer, configured to determine whether the authentication process is qualified based on the authentication duration, determine the reason for failure of the authentication process based on the authentication duration, and generate an instruction for adjusting the output power of a fill light device during data collection, or the update time of a biometric recognition model during data collection, or the key update time during data transmission; A control layer is connected to the analysis layer, the data acquisition layer and the data transmission layer respectively, and is used to determine the adjustment of the output power of the fill light device, the update time of the biometric recognition model, or the key update time based on the instruction.

2. The multi-dimensional trusted authentication system based on data processing according to claim 1, characterized in that: The authentication statistics layer is also connected to the data detection layer to calculate the percentage of abnormal rejection; The analysis layer is further configured to determine that the authentication process is qualified if a comparison result between the authentication time and a preset authentication time meets a first determination condition; The analysis layer is further configured to re-determine the authentication process based on the abnormal rejection ratio when the comparison result between the authentication duration and the authentication duration does not meet the first determination condition but meets the second determination condition; The analysis layer is further configured to determine that the authentication process is unqualified if a comparison result between the authentication duration and the preset authentication duration does not meet the second determination condition, and to determine a reason for the unqualified process and a corresponding treatment based on the authentication duration offset; Among them, the abnormal elimination ratio is the ratio of the number of information eliminated in the multi-dimensional trusted input information to the total number of input information, the preset authentication time includes a first preset authentication time and a second preset authentication time that is greater than the first preset authentication time, the first judgment condition is that the authentication time is less than or equal to the first preset authentication time, the second judgment condition is that the authentication time is greater than the first preset authentication time and less than or equal to the second preset authentication time, and the authentication time offset is the difference between the authentication time and the second preset authentication time.

3. The multi-dimensional trusted authentication system based on data processing according to claim 2, characterized in that: The analysis layer is further used to re-determine the authentication process based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio; The analysis layer determines that there is an abnormality in the light when the data acquisition layer performs real-time monitoring based on a comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio, and adjusts the fill light device of the data acquisition layer; Or it is determined that the iris matching error rate increases when the data acquisition layer performs real-time monitoring, and the updating time of the biometric recognition model is adjusted.

4. The multi-dimensional trusted authentication system based on data processing according to claim 3, characterized in that: The data collection layer is also used to obtain light values ​​within a preset range of the place; The analysis layer is further configured to adjust the output power of the fill light device based on obtaining the light value and comparing it with a preset light value, wherein the preset light value includes a preset light lower limit value and a preset light upper limit value; The analysis layer determines to increase the output power of the fill light device when the light value is less than the preset light lower limit; The analysis layer determines to reduce the output power of the fill light device when the light value is greater than the preset light upper limit value; The analysis layer determines to maintain the output power of the fill light device when the light value is greater than or equal to the preset light lower limit value and less than or equal to the preset light upper limit value.

5. The multi-dimensional trusted authentication system based on data processing according to claim 4, characterized in that: When determining to increase the output power of the fill light device, the analysis layer is further configured to generate an instruction to increase the output power based on a comparison result of the light offset value and a preset light offset value, wherein the increase in the output power is positively correlated with the light offset value; The light offset value is the difference between the preset light value and the light value.

6. The multi-dimensional trusted authentication system based on data processing according to claim 4, characterized in that: When determining to reduce the output power of the fill light device, the analysis layer is further configured to generate an instruction to reduce the output power based on a comparison result of the light difference value with a preset light difference value, and the reduction range of the output power is positively correlated with the light difference value; The light difference is a difference between the light value and the preset light value.

7. The multi-dimensional trusted authentication system based on data processing according to claim 3, characterized in that: The analysis layer is further configured to generate an instruction for shortening the update time of the biometric recognition model based on a comparison result between the iris matching error rate and a preset iris matching error rate, wherein the shortening degree of the update time is positively correlated with the iris matching error rate.

8. The multi-dimensional trusted authentication system based on data processing according to any one of claims 5 to 7, characterized in that: In a case where the control layer completes the control based on an instruction to increase the output power of the fill light device or based on an instruction to reduce the output power of the fill light device, the analysis layer still determines that there is an abnormality in the light when the data acquisition layer performs real-time monitoring based on a comparison result of the abnormal rejection ratio with the preset abnormal rejection ratio; Or in the case where the control layer completes the control based on the instruction to shorten the update time of the biometric recognition model, the analysis layer still determines that the iris matching error rate increases when the data acquisition layer performs real-time monitoring based on the comparison result of the abnormal rejection ratio and the preset abnormal rejection ratio, the analysis layer re-determines the cause of the failure and the corresponding processing based on the authentication time offset.

9. The multi-dimensional trusted authentication system based on data processing according to claim 3, characterized in that: The analysis layer is further configured to determine a reason for failure of the authentication process based on a comparison result of the authentication time offset with the preset authentication time offset and generate a corresponding instruction, including: If it is determined that the reason for the failure is that there is a transmission anomaly in the data transmission layer, resulting in the comparison result of the authentication time with the preset authentication time not meeting the second determination condition, the analysis layer determines that the network for transmitting information is abnormal and generates an instruction to issue a network maintenance; If it is determined that the reason for failure is that the authentication model constructed in the data decision layer has an anomaly, resulting in the comparison result between the authentication time and the preset authentication time not meeting the second determination condition, the analysis layer determines that the authentication model has an anomaly and generates an instruction to reconstruct the authentication model; If it is determined that the reason for failure is that there is an external data attack resulting in the comparison result of the authentication duration with the preset authentication duration not meeting the second determination condition, the analysis layer is further configured to adjust the key update time based on the second-order difference of the authentication duration; The second-order difference of the authentication duration is the difference between the authentication duration offset and the preset authentication duration offset.

10. The multi-dimensional trusted authentication system based on data processing according to claim 9, characterized in that: The analysis layer is further configured to generate an instruction to shorten the key update time based on a comparison result of the second-order difference of the authentication time and the preset second-order difference of the authentication time, and the degree of shortening of the key update time is positively correlated with the second-order difference of the authentication time.

Images