An identity authentication method for mobile secure payment
By calculating the characteristic damage index and fingerprint damage index of the fingerprint digital template, and dynamically update the fingerprint data, the problem of identity authentication error recognition caused by fingerprint aging is solved, and the accuracy and security of identity authentication of mobile security payments is improved.
Patent Information
- Application Number
- CN202410774099.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-17
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-06-17
AI Technical Summary
Existing mobile secure payment identity authentication methods rely on fingerprint recognition, however small differences caused by fingerprint aging or wear may lead to misidentification of identity authentication, increasing fraud risks and transaction disputes.
By obtaining the initial fingerprint digital template and the real-time fingerprint digital template, calculate the feature damage index and fingerprint damage index of the feature point matching pair, dynamically update the learning rate and feature vector data, and generate update fingerprint digital templates to improve the accuracy and security of identity authentication. .
Enhanced security and fraud prevention capabilities in the payment process, ensure the use of the latest and accurate fingerprint information, and improve the accuracy and efficiency of identity authentication.
Smart Images

Figure CN118735521B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of digital information transmission, and in particular to an identity authentication method for mobile secure payment. Background Art
[0002] The popularity of e-commerce and online transactions has made mobile payments a part of daily life. Using smart devices for mobile payments has become one of the mainstream means of commodity transactions for the public. This is accompanied by security issues in the transaction payment process. If there is a lack of an effective identity authentication mechanism, fraudsters can easily impersonate others to conduct transactions, increasing the risks of fraud and account theft. At the same time, in the absence of identity authentication, the authenticity and legality of transactions are difficult to confirm, which may lead to an increase in transaction disputes and legal proceedings and the emergence of a large number of illegal transactions, disrupting the normal market order and causing great economic losses to users and merchants.
[0003] Currently, the mainstream authentication methods for mobile secure payment are fingerprint recognition and face recognition. Fingerprint recognition is more widely used. By pre-storing the user's fingerprint template (digital template) in the smart device, when the user attempts to authenticate the identity, the system will collect the fingerprint again and create a new fingerprint template, and compare the two templates to achieve identity authentication; however, in the process of work and life, people's fingerprints may change due to aging or daily wear and tear. When the sensitivity of the authentication system is high, even the person's fingerprint may be rejected due to slight differences, resulting in incorrect identification of the identity. Summary of the invention
[0004] The invention provides an identity authentication method for mobile secure payment to solve the existing problems.
[0005] The identity authentication method for mobile secure payment of the present invention adopts the following technical solution:
[0006] An embodiment of the present invention provides an identity authentication method for mobile secure payment, the method comprising the following steps:
[0007] Acquire several fingerprint digital templates; the fingerprint digital templates include an initial fingerprint digital template and several real-time fingerprint digital templates, each real-time fingerprint digital template corresponds to a collection time, each fingerprint digital template corresponds to several feature points, and each feature point corresponds to several vector dimension data;
[0008] Match the initial fingerprint digital template with the feature points in each real-time fingerprint digital template to obtain a number of feature point matching pairs; according to the differences between all vector dimension data corresponding to the feature points of each feature point matching pair in each real-time fingerprint digital template and the initial fingerprint digital template and the distribution of the feature points in the real-time fingerprint digital template, obtain the feature damage index of the feature points of each feature point matching pair in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template;
[0009] Among all the feature point matching pairs corresponding to the initial fingerprint digital template and all the real-time fingerprint digital templates, the feature point of the real-time fingerprint digital template in the feature point matching pair where each feature point in the initial fingerprint digital template is located is recorded as the matched feature point of each feature point; according to the difference between the feature damage indexes of the matched feature points of each feature point in the initial fingerprint digital template, the stability coefficient of each feature point in the initial fingerprint digital template is obtained;
[0010] According to the stability coefficient of each feature point in the initial fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template, the update learning rate of each feature point in the initial fingerprint digital template is obtained;
[0011] According to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the matched feature point, each vector dimension data of each feature point of the initial fingerprint digital template is updated to obtain an updated fingerprint digital template; the updated fingerprint digital template is transmitted to the identity authentication template library for identity authentication of mobile secure payment.
[0012] Furthermore, the method of obtaining the feature damage index of the feature points of each feature point matching pair in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template respectively according to the difference between all vector dimension data corresponding to the feature points in each real-time fingerprint digital template and the initial fingerprint digital template and the distribution of the feature points in the real-time fingerprint digital template includes the following specific steps:
[0013] Record any real-time fingerprint digital template as the current fingerprint digital template;
[0014] Among all the feature point matching pairs corresponding to the initial fingerprint digital template and the current fingerprint digital template, according to all the vector dimension data of the feature vectors corresponding to the feature points of each feature point matching pair in the current fingerprint digital template and the initial fingerprint digital template, the feature damage index of the feature points of each feature point matching pair in the current fingerprint digital template is obtained; wherein each feature point corresponds to a feature vector, and each feature vector contains a number of vector dimension data;
[0015] The coordinate system is constructed with the lower left corner of the current fingerprint digital template as the coordinate origin, the horizontal rightward direction as the positive direction of the horizontal axis, and the vertical upward direction as the positive direction of the vertical axis;
[0016] All feature points are matched to feature points in the current fingerprint digital template and recorded as matched feature points; feature points in the current fingerprint digital template other than the matched feature points are recorded as unmatched feature points;
[0017] The fingerprint damage index of the current fingerprint digital template is obtained according to the feature damage index of the matched feature points in the current fingerprint digital template and the coordinates of all the unmatched feature points in the current fingerprint digital template in the coordinate system.
[0018] Furthermore, the specific calculation formula for obtaining the feature damage index of the feature point of each feature point matching pair in the current fingerprint digital template according to all vector dimension data of the feature vector corresponding to the feature point of each feature point matching pair in the current fingerprint digital template and the initial fingerprint digital template is:
[0019]
[0020] In the formula, S i_1 represents the feature damage index of the feature point of the i-th feature point matching pair in the current fingerprint digital template; m represents the number of vector dimension data contained in the feature vector corresponding to each feature point; h a_i_j Indicates the jth vector dimension data of the feature vector corresponding to the feature point of the i-th feature point matching pair in the current fingerprint digital template; h b_i_j Represents the j-th vector dimension data of the feature vector corresponding to the feature point of the i-th feature point matching pair in the initial fingerprint digital template; || is the absolute value function; norm() is the linear normalization function.
[0021] Furthermore, the specific calculation formula for obtaining the fingerprint damage index of the current fingerprint digital template according to the feature damage index of the matched feature points in the current fingerprint digital template and the coordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system is:
[0022]
[0023] Where S represents the fingerprint damage index of the current fingerprint digital template; n1 represents the number of matched feature points in the current fingerprint digital template; n2 represents the number of unmatched feature points in the current fingerprint digital template; h x_k Indicates the horizontal coordinate of the kth unmatched feature point in the current fingerprint digital template in the coordinate system; h y_k Indicates the ordinate of the kth unmatched feature point in the current fingerprint digital template in the coordinate system; μ xRepresents the mean value of the horizontal coordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system; μ y It represents the mean value of the ordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system; S z ′ It represents the feature damage index of the zth matched feature point in the current fingerprint digital template; norm() is a linear normalization function.
[0024] Furthermore, the method of obtaining the stability coefficient of each feature point in the initial fingerprint digital template according to the difference between the feature damage indexes of the matched feature points of each feature point in the initial fingerprint digital template includes the following specific steps:
[0025] The ratio of the number of matched feature points of each feature point to the number of all real-time fingerprint digital templates is recorded as the appearance frequency of each feature point in the initial fingerprint digital template;
[0026] According to the occurrence frequency of each feature point in the initial fingerprint digital template and the feature damage index of the matched feature point of each feature point in the initial fingerprint digital template, the stability coefficient of each feature point in the initial fingerprint digital template is obtained.
[0027] Furthermore, the specific calculation formula for obtaining the stability coefficient of each feature point in the initial fingerprint digital template according to the occurrence frequency of each feature point in the initial fingerprint digital template and the feature damage index of the matched feature point of each feature point in the initial fingerprint digital template is:
[0028]
[0029] Where W b_v represents the stability coefficient of the vth feature point in the initial fingerprint digital template; f b_v represents the frequency of occurrence of the vth feature point in the initial fingerprint digital template; q v represents the number of matched feature points of the vth feature point; S b_v_(x+1) represents the feature damage index of the x+1th matched feature point of the vth feature point in the initial fingerprint template; S b_v_x represents the feature damage index of the xth matched feature point of the vth feature point in the initial fingerprint template; t v It represents the mean of the time intervals between the acquisition times of the real-time fingerprint digital templates corresponding to all two matched feature points of the vth feature point in the initial fingerprint digital template; || is the absolute value function; exp() is an exponential function with a natural constant as the base; norm() is a linear normalization function.
[0030] Furthermore, the updating learning rate of each feature point in the initial fingerprint digital template is obtained according to the stability coefficient of each feature point in the initial fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template, and the specific steps include the following:
[0031] Matching the initial fingerprint digital template with all real-time fingerprint digital templates respectively to obtain the matching result of each real-time fingerprint digital template, wherein the matching result includes a successful match and a failed match;
[0032] The ratio of the number of real-time fingerprint digital templates that failed to match to the number of all real-time fingerprint digital templates is recorded as the matching failure rate of all real-time fingerprint digital templates;
[0033] The ratio of the number of successfully matched real-time fingerprint digital templates in the real-time fingerprint digital template where the matched feature point of the vth feature point in the initial fingerprint digital template is located to the number of all real-time fingerprint digital templates is recorded as the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates;
[0034] According to the matching failure rate of all real-time fingerprint digital templates, the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates, the fingerprint damage index of all real-time fingerprint digital templates and the stability coefficient of each feature point in the initial fingerprint digital template, the update learning rate of each feature point in the initial fingerprint digital template is obtained.
[0035] Furthermore, the specific calculation formula for obtaining the update learning rate of each feature point in the initial fingerprint digital template according to the matching failure rate of all real-time fingerprint digital templates, the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates, the fingerprint damage index of all real-time fingerprint digital templates and the stability coefficient of each feature point in the initial fingerprint digital template is:
[0036]
[0037] In the formula, α b_v represents the update learning rate of the vth feature point in the initial fingerprint digital template; p represents the number of all real-time fingerprint digital templates; D a Indicates the matching failure rate of all real-time fingerprint digital templates; represents the mean value of fingerprint damage index of all real-time fingerprint digital templates; W b_v represents the stability coefficient of the vth feature point in the initial fingerprint digital template; f b_v represents the frequency of occurrence of the vth feature point in the initial fingerprint digital template; P b_vIt represents the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates; exp() is an exponential function with a natural constant as the base; norm[] is a linear normalization function.
[0038] Furthermore, the updating of each vector dimension data of each feature point in the initial fingerprint digital template according to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the matched feature point to obtain the updated fingerprint digital template includes the following specific steps:
[0039] According to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the corresponding matched feature point, each vector dimension data of the feature vector corresponding to each feature point after the initial fingerprint digital template is updated is obtained; wherein each feature point corresponds to a feature vector, and each feature vector contains a number of vector dimension data;
[0040] In the initial fingerprint digital template, each vector dimension data of the feature vector corresponding to each feature point is replaced with each vector dimension data of the feature vector corresponding to each feature point after update, so as to obtain an updated fingerprint digital template.
[0041] Furthermore, the specific calculation formula for obtaining each vector dimension data of the feature vector corresponding to each feature point after the initial fingerprint digital template is updated is as follows based on the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the corresponding matched feature point:
[0042]
[0043] In the formula, h′ b_vu Indicates the u-th vector dimension data of the feature vector corresponding to the v-th feature point after the initial fingerprint digital template is updated; h b_vu Represents the u-th vector dimension data of the feature vector corresponding to the v-th feature point of the initial fingerprint digital template; α b_v represents the update learning rate of the vth feature point in the initial fingerprint digital template; k v Indicates the number of matched feature points of the vth feature point; h a_vu_i Represents the u-th vector dimension data of the feature vector corresponding to the i-th matched feature point of the v-th feature point in the initial fingerprint digital template.
[0044] The beneficial effects of the technical solution of the present invention are:
[0045] In the embodiment of the present invention, a plurality of fingerprint digital templates are obtained; the fingerprint digital templates include an initial fingerprint digital template and a plurality of real-time fingerprint digital templates, each real-time fingerprint digital template corresponds to a collection time, each fingerprint digital template corresponds to a plurality of feature points, and each feature point corresponds to a plurality of vector dimension data, providing a data basis for subsequent analysis; the initial fingerprint digital template is matched with the feature points in each real-time fingerprint digital template to obtain a plurality of feature point matching pairs; according to the difference between all vector dimension data corresponding to the feature points in each real-time fingerprint digital template and the initial fingerprint digital template of each feature point matching pair and the distribution of the feature points in the real-time fingerprint digital template, the feature damage index of the feature points of each feature point matching pair in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template are obtained respectively, and the damage degree of the fingerprint digital template is evaluated; among all the feature point matching pairs corresponding to the initial fingerprint digital template and all the real-time fingerprint digital templates, the feature damage index of each feature point in the initial fingerprint digital template is matched. The feature points at the matching pair are the feature points of the real-time fingerprint digital template, and are recorded as the matched feature points of each feature point; according to the difference between the feature damage indexes of the matched feature points of each feature point in the initial fingerprint digital template, the stability coefficient of each feature point in the initial fingerprint digital template is obtained, providing a data basis for the calculation of the updated learning rate; according to the stability coefficient of each feature point in the initial fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template, the updated learning rate of each feature point in the initial fingerprint digital template is obtained, thereby improving the efficiency and accuracy of the update; according to the updated learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the matched feature point, each vector dimension data of each feature point of the initial fingerprint digital template is updated to obtain an updated fingerprint digital template; the updated fingerprint digital template is transmitted to the identity authentication template library for identity authentication of mobile security payment, ensuring the use of the latest and accurate fingerprint information in the payment process, and enhancing the payment security and the ability to prevent fraud. So far, the present invention extracts the stable fingerprint information in the initial fingerprint digital template, dynamically adjusts the learning rate of different feature points in combination with the user authentication behavior, improves the accuracy of fingerprint information update, and improves the accuracy of subsequent identity authentication. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0047] Figure 1The present invention is a flowchart of the steps of an identity authentication method for mobile secure payment. DETAILED DESCRIPTION
[0048] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the identity authentication method for mobile secure payment proposed by the present invention, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0049] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0050] The following is a detailed description of a specific scheme of an identity authentication method for mobile secure payment provided by the present invention in conjunction with the accompanying drawings.
[0051] See also Figure 1 , which shows a flowchart of a method for identity authentication for mobile secure payment provided by an embodiment of the present invention, the method comprising the following steps:
[0052] Step S001: Acquire several fingerprint digital templates; the fingerprint digital templates include an initial fingerprint digital template and several real-time fingerprint digital templates, each real-time fingerprint digital template corresponds to a collection time, each fingerprint digital template corresponds to several feature points, and each feature point corresponds to several vector dimension data.
[0053] Acquire several fingerprint digital templates; the fingerprint digital templates include an initial fingerprint digital template and several real-time fingerprint digital templates, each real-time fingerprint digital template corresponds to a collection time; the fingerprint digital template corresponds to several feature points, each feature point corresponds to a feature vector, and each feature vector contains several vector dimension data;
[0054] It should be noted that, in the process of device authentication, the fingerprint sensor on the smart mobile device is used to collect the user's real-time fingerprint digital template, and the real-time fingerprint digital template of the user each time the identity authentication is performed in the past week is stored, regardless of whether the match is successful; the initial fingerprint template is collected using a capacitive fingerprint identifier, including all the user's fingerprint feature points (such as ridges, bifurcation points, endpoints, etc.); the initial fingerprint template is converted into an initial fingerprint digital template, which is a well-known technology, specifically, all feature points in the initial fingerprint template are converted into feature vectors, wherein each type of vector dimension data is subjected to standard processing so that they have similar scales and distributions for better comparison, which is a well-known technology.
[0055] Step S002: Match the initial fingerprint digital template with the feature points in each real-time fingerprint digital template to obtain a number of feature point matching pairs; according to the differences between all vector dimension data corresponding to the feature points of each feature point matching pair in each real-time fingerprint digital template and the initial fingerprint digital template and the distribution of the feature points in the real-time fingerprint digital template, obtain the feature damage index of the feature points of each feature point matching pair in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template.
[0056] Based on several feature point matching pairs between the current fingerprint digital template and the initial fingerprint digital template, the damage index of the matched feature points is calculated. When the difference in the feature vectors between the two feature points is greater, the feature damage index of the position is greater. After comparing all the feature points, the fingerprint damage index of the real-time fingerprint digital template is obtained according to the number of unmatched feature points and the feature damage index of each matched feature point. It should be noted that when performing fingerprint recognition, the more unmatched feature points there are in the real-time fingerprint data and the more discrete the distribution of the unmatched feature points is, the greater the fingerprint damage index is.
[0057] Record any real-time fingerprint digital template as the current fingerprint digital template;
[0058] Using fingerprint recognition technology, matching the initial fingerprint digital template with the feature points in the current fingerprint digital template to obtain a number of feature point matching pairs, wherein the fingerprint recognition technology is a well-known technology;
[0059] Among all the feature point matching pairs corresponding to the initial fingerprint digital template and the current fingerprint digital template, according to all the vector dimension data of the feature vectors corresponding to the feature points of each feature point matching pair in the current fingerprint digital template and the initial fingerprint digital template, the corresponding calculation formula for the feature damage index of the feature points of each feature point matching pair in the current fingerprint digital template is obtained as follows:
[0060]
[0061] In the formula, S i_1 represents the feature damage index of the feature point of the i-th feature point matching pair in the current fingerprint digital template; m represents the number of vector dimension data contained in the feature vector corresponding to each feature point; h a_i_j Indicates the jth vector dimension data of the feature vector corresponding to the feature point of the i-th feature point matching pair in the current fingerprint digital template; h b_i_j It represents the j-th vector dimension data of the feature vector corresponding to the feature point of the i-th feature point matching pair in the initial fingerprint digital template; || is the absolute value function; norm() is the linear normalization function, which normalizes the data value to the [0,1] interval.
[0062] The coordinate system is constructed with the lower left corner of the current fingerprint digital template as the coordinate origin, the horizontal rightward direction as the positive direction of the horizontal axis, and the vertical upward direction as the positive direction of the vertical axis;
[0063] All feature points are matched to feature points in the current fingerprint digital template and recorded as matched feature points; feature points in the current fingerprint digital template other than the matched feature points are recorded as unmatched feature points;
[0064] According to the feature damage index of the matched feature points in the current fingerprint digital template and the coordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system, the corresponding calculation formula for the fingerprint damage index of the current fingerprint digital template is obtained as follows:
[0065]
[0066] Where S represents the fingerprint damage index of the current fingerprint digital template; n1 represents the number of matched feature points in the current fingerprint digital template; n2 represents the number of unmatched feature points in the current fingerprint digital template; h x_k Indicates the horizontal coordinate of the kth unmatched feature point in the current fingerprint digital template in the coordinate system; h y_k Indicates the ordinate of the kth unmatched feature point in the current fingerprint digital template in the coordinate system; μ x Represents the mean value of the horizontal coordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system; μ y It represents the mean value of the ordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system; S z ′ It represents the feature damage index of the zth matched feature point in the current fingerprint digital template; norm() is a linear normalization function that normalizes the data value to the interval [0,1].
[0067] It should be noted that when the vector dimension data difference of the feature vectors corresponding to the two feature points in the feature point matching pair |h a_i_j -hb_i_j | is larger, the greater the fingerprint damage index at that position; when performing fingerprint recognition, the more unmatched feature points there are in the current fingerprint digital template, and the more discrete the distribution of the unmatched feature points is, the greater the fingerprint damage index is.
[0068] The initial fingerprint digital template is matched with the feature points in all the real-time fingerprint digital templates respectively. According to the above method, the feature damage index of the feature points in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template are obtained.
[0069] Step S003: In all feature point matching pairs corresponding to the initial fingerprint digital template and all real-time fingerprint digital templates, the feature point of the real-time fingerprint digital template in the feature point matching pair where each feature point in the initial fingerprint digital template is located is recorded as the matched feature point of each feature point; based on the difference between the feature damage indexes of the matched feature points of each feature point in the initial fingerprint digital template, the stability coefficient of each feature point in the initial fingerprint digital template is obtained.
[0070] Due to the differences in work characteristics, living habits, individual differences, etc. of different users, the wear and recovery of their finger fingerprints will show different characteristics. The fingerprint feature points in some areas of their fingers may always maintain low damage or high damage, while the fingerprint feature points in some areas may be constantly worn and recovered, which is specifically manifested as the fluctuation of the fingerprint damage degree; the wear of user's finger fingerprint is a long process, and while the fingerprint is worn, the fingerprint will also recover to a certain extent. Therefore, there is no need to frequently update the fingerprint digital template stored in the smart device. Therefore, it is stipulated that the fingerprint digital template of the smart device should be updated every week, and the initial fingerprint digital template should be updated in combination with all real-time fingerprint data of the current week.
[0071] Among all the feature point matching pairs corresponding to the initial fingerprint digital template and all the real-time fingerprint digital templates, the feature point of the real-time fingerprint digital template in the feature point matching pair where the vth feature point in the initial fingerprint digital template is located is recorded as the matched feature point of the vth feature point;
[0072] The ratio of the number of matched feature points of the vth feature point to the number of all real-time fingerprint digital templates is recorded as the appearance frequency of the vth feature point in the initial fingerprint digital template;
[0073] According to the above method, the occurrence frequency of each feature point in the initial fingerprint digital template is obtained;
[0074] According to the occurrence frequency of each feature point in the initial fingerprint digital template and the feature damage index of the matched feature point of each feature point in the initial fingerprint digital template, the corresponding calculation formula of the stability coefficient of each feature point in the initial fingerprint digital template is obtained as follows:
[0075]
[0076] Where W b_v represents the stability coefficient of the vth feature point in the initial fingerprint digital template; f b_v represents the frequency of occurrence of the vth feature point in the initial fingerprint digital template; q v represents the number of matched feature points of the vth feature point; S b_v_(x+1) represents the feature damage index of the x+1th matched feature point of the vth feature point in the initial fingerprint template; S b_v_x represents the feature damage index of the xth matched feature point of the vth feature point in the initial fingerprint template; t v It represents the mean of the time intervals of the acquisition times corresponding to the real-time fingerprint digital templates where all two arbitrary matched feature points of the vth feature point in the initial fingerprint digital template are located; || is the absolute value function; exp() is an exponential function with a natural constant as the base. This embodiment uses exp(-) to present the inverse proportional relationship and normalization processing. The implementer can set the inverse proportional function and normalization function according to the actual situation; norm() is a linear normalization function, which normalizes the data value to the interval [0,1].
[0077] It should be noted that when a feature point appears stably in the real-time fingerprint digital template, the higher the stability coefficient of the feature point; It represents the change rate of the fingerprint damage index of the matched feature point of the vth feature point in the initial fingerprint digital template. The smaller the change rate of the feature damage coefficient of the matched feature point of the feature point, the higher the stability coefficient of the feature point.
[0078] Step S004: according to the stability coefficient of each feature point in the initial fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template, the update learning rate of each feature point in the initial fingerprint digital template is obtained.
[0079] When performing fingerprint recognition identity authentication, the position of the user's finger each time will be different from the initial position, resulting in certain differences in the feature points of each real-time fingerprint digital template, that is, the frequency of occurrence of the feature points is different. The higher the frequency of occurrence, the higher the recognition value, that is, the higher the amplitude when adjusting the learning rate. At the same time, for feature points with high frequency of occurrence in the real-time fingerprint digital template, the higher the matching success rate during the matching process with the initial fingerprint digital template, the lower the learning rate when the feature point is updated.
[0080] In addition, the learning rate is adjusted in combination with the user's authentication behavior. If the user authenticates frequently, the system may collect more data to evaluate the accuracy and stability of the fingerprint digital template. In this case, the learning rate can be set lower to avoid over-adjustment. At the same time, if the user often fails in authentication, this may indicate that the current fingerprint template is not accurate enough or needs to be updated. The learning rate is increased to adapt to changes in user fingerprints more quickly. The larger the fingerprint damage mean of the real-time fingerprint digital template in the past week, the greater the degree of change in the user's fingerprint features in the recent period. In this case, the update learning rate of the feature points is increased as a whole to facilitate faster acquisition of the user's fingerprint features.
[0081] Using fingerprint recognition technology, the initial fingerprint digital template is matched with all real-time fingerprint digital templates respectively to obtain the matching result of each real-time fingerprint digital template, wherein the matching result includes matching success and matching failure;
[0082] The ratio of the number of real-time fingerprint digital templates that failed to match to the number of all real-time fingerprint digital templates is recorded as the matching failure rate of all real-time fingerprint digital templates;
[0083] The ratio of the number of successfully matched real-time fingerprint digital templates in the real-time fingerprint digital template where the matched feature point of the vth feature point in the initial fingerprint digital template is located to the number of all real-time fingerprint digital templates is recorded as the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates;
[0084] According to the matching failure rate of all real-time fingerprint digital templates, the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates, the fingerprint damage index of all real-time fingerprint digital templates and the stability coefficient of each feature point in the initial fingerprint digital template, the corresponding calculation formula for the update learning rate of each feature point in the initial fingerprint digital template is obtained as follows:
[0085]
[0086] In the formula, α b_v represents the update learning rate of the vth feature point in the initial fingerprint digital template; p represents the number of all real-time fingerprint digital templates; D a Indicates the matching failure rate of all real-time fingerprint digital templates; represents the mean value of fingerprint damage index of all real-time fingerprint digital templates; W b_v represents the stability coefficient of the vth feature point in the initial fingerprint digital template; f b_v represents the frequency of occurrence of the vth feature point in the initial fingerprint digital template; P b_vIt represents the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates; exp() is an exponential function with a natural constant as the base. This embodiment uses exp(-) to present the inverse proportional relationship and normalization processing. The implementer can set the inverse proportional function and normalization function according to the actual situation; norm[] is a linear normalization function, which normalizes the data value to the interval [0,1].
[0087] Step S005: According to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the matched feature point, each vector dimension data of each feature point in the initial fingerprint digital template is updated to obtain an updated fingerprint digital template; the updated fingerprint digital template is transmitted to the identity authentication template library for identity authentication of mobile secure payment.
[0088] Combined with the update learning rate of each feature point in the initial fingerprint digital template obtained above, the feature vector of the feature point in the initial fingerprint digital template is updated and adjusted. The greater the update learning rate of the feature point, the greater the degree of adjustment of the data in the corresponding feature vector. At the same time, repeat the update step to update the feature vectors corresponding to all feature points in the initial fingerprint digital template to obtain a new initial fingerprint digital template after update.
[0089] According to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the corresponding matched feature point, the corresponding calculation formula for each vector dimension data of the feature vector corresponding to each feature point after the initial fingerprint digital template is updated is:
[0090]
[0091] In the formula, h′ b_vu Indicates the u-th vector dimension data of the feature vector corresponding to the v-th feature point after the initial fingerprint digital template is updated; h b_vu Represents the u-th vector dimension data of the feature vector corresponding to the v-th feature point of the initial fingerprint digital template; α b_v represents the update learning rate of the vth feature point in the initial fingerprint digital template; k v Indicates the number of matched feature points of the vth feature point; h a_vu_i Represents the u-th vector dimension data of the feature vector corresponding to the i-th matched feature point of the v-th feature point in the initial fingerprint digital template.
[0092] In the initial fingerprint digital template, each vector dimension data of the feature vector corresponding to each feature point is replaced with each vector dimension data of the feature vector corresponding to each feature point after update, so as to obtain an updated fingerprint digital template; the updated fingerprint digital template is transmitted to the identity authentication template library, and all real-time fingerprint digital templates are deleted at the same time;
[0093] The user of the smart device is authenticated according to the updated fingerprint digital template, thereby improving the attention of the existing fingerprint recognition algorithm to the local points of the real-time fingerprint digital template and completing the user's identity authentication.
[0094] So far, the present invention is completed.
[0095] In summary, in an embodiment of the present invention, an initial fingerprint digital template and several real-time fingerprint digital templates are obtained, and the fingerprint damage index of each real-time fingerprint digital template is calculated according to several feature point matching pairs obtained by matching the initial fingerprint digital template with the real-time fingerprint digital template, and the stability coefficient of each feature point in the initial fingerprint digital template is further obtained, thereby obtaining the update learning rate of each feature point in the initial fingerprint digital template, and the initial fingerprint digital template is updated to obtain the updated fingerprint digital template, and the identity authentication of mobile security payment is completed. The present invention extracts stable fingerprint information from the initial fingerprint digital template, and dynamically adjusts the learning rate of different feature points in combination with user authentication behavior, thereby improving the accuracy of fingerprint information update and the accuracy of subsequent identity authentication.
[0096] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. An identity authentication method for mobile secure payment, It is characterized in that The method comprises the following steps: Acquire several fingerprint digital templates; the fingerprint digital templates include an initial fingerprint digital template and several real-time fingerprint digital templates, each real-time fingerprint digital template corresponds to a collection time, each fingerprint digital template corresponds to several feature points, and each feature point corresponds to several vector dimension data; Match the initial fingerprint digital template with the feature points in each real-time fingerprint digital template to obtain a number of feature point matching pairs; according to the differences between all vector dimension data corresponding to the feature points of each feature point matching pair in each real-time fingerprint digital template and the initial fingerprint digital template and the distribution of the feature points in the real-time fingerprint digital template, obtain the feature damage index of the feature points of each feature point matching pair in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template; Among all the feature point matching pairs corresponding to the initial fingerprint digital template and all the real-time fingerprint digital templates, the feature point of the real-time fingerprint digital template in the feature point matching pair where each feature point in the initial fingerprint digital template is located is recorded as the matched feature point of each feature point; according to the difference between the feature damage indexes of the matched feature points of each feature point in the initial fingerprint digital template, the stability coefficient of each feature point in the initial fingerprint digital template is obtained; According to the stability coefficient of each feature point in the initial fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template, the update learning rate of each feature point in the initial fingerprint digital template is obtained; According to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the matched feature point, each vector dimension data of each feature point of the initial fingerprint digital template is updated to obtain an updated fingerprint digital template; the updated fingerprint digital template is transmitted to the identity authentication template library for identity authentication of mobile secure payment.
2. According to claim 1, an identity authentication method for mobile secure payment, It is characterized in that The method of obtaining the feature damage index of the feature points of each feature point matching pair in each real-time fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template respectively according to the difference between all vector dimension data corresponding to the feature points in each real-time fingerprint digital template and the initial fingerprint digital template and the distribution of the feature points in the real-time fingerprint digital template includes the following specific steps: Record any real-time fingerprint digital template as the current fingerprint digital template; Among all the feature point matching pairs corresponding to the initial fingerprint digital template and the current fingerprint digital template, according to all the vector dimension data of the feature vectors corresponding to the feature points of each feature point matching pair in the current fingerprint digital template and the initial fingerprint digital template, the feature damage index of the feature points of each feature point matching pair in the current fingerprint digital template is obtained; wherein each feature point corresponds to a feature vector, and each feature vector contains a number of vector dimension data; The coordinate system is constructed with the lower left corner of the current fingerprint digital template as the coordinate origin, the horizontal rightward direction as the positive direction of the horizontal axis, and the vertical upward direction as the positive direction of the vertical axis; All feature points are matched to feature points in the current fingerprint digital template and recorded as matched feature points; feature points in the current fingerprint digital template other than the matched feature points are recorded as unmatched feature points; The fingerprint damage index of the current fingerprint digital template is obtained according to the feature damage index of the matched feature points in the current fingerprint digital template and the coordinates of all the unmatched feature points in the current fingerprint digital template in the coordinate system.
3. According to claim 2, an identity authentication method for mobile secure payment, It is characterized in that The specific calculation formula for obtaining the feature damage index of the feature point of each feature point matching pair in the current fingerprint digital template according to all vector dimension data of the feature vector corresponding to the feature point of each feature point matching pair in the current fingerprint digital template and the initial fingerprint digital template is: In the formula, S i_1 Represents the feature damage index of the feature point of the i-th feature point matching pair in the current fingerprint digital template; m represents the number of vector dimension data contained in the feature vector corresponding to each feature point; h a_i_j Indicates the jth vector dimension data of the feature vector corresponding to the feature point of the i-th feature point matching pair in the current fingerprint digital template; h b_i_j Represents the j-th vector dimension data of the feature vector corresponding to the feature point of the i-th feature point matching pair in the initial fingerprint digital template; || is the absolute value function; norm() is the linear normalization function.
4. According to claim 2, an identity authentication method for mobile secure payment, It is characterized in that The specific calculation formula for obtaining the fingerprint damage index of the current fingerprint digital template according to the feature damage index of the matched feature points in the current fingerprint digital template and the coordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system is: Where S represents the fingerprint damage index of the current fingerprint digital template; n 1 Indicates the number of matched feature points in the current fingerprint digital template; n 2 Indicates the number of unmatched feature points in the current fingerprint digital template; h x_k Indicates the horizontal coordinate of the kth unmatched feature point in the current fingerprint digital template in the coordinate system; h y_k Indicates the ordinate of the kth unmatched feature point in the current fingerprint digital template in the coordinate system; μ x Represents the mean value of the horizontal coordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system; μ y Represents the mean value of the ordinates of all unmatched feature points in the current fingerprint digital template in the coordinate system; S z ′ It represents the feature damage index of the zth matched feature point in the current fingerprint digital template; norm() is a linear normalization function.
5. The identity authentication method for mobile secure payment according to claim 1, It is characterized in that The specific steps of obtaining the stability coefficient of each feature point in the initial fingerprint digital template according to the difference between the feature damage indexes of the matched feature points of each feature point in the initial fingerprint digital template are as follows: The ratio of the number of matched feature points of each feature point to the number of all real-time fingerprint digital templates is recorded as the appearance frequency of each feature point in the initial fingerprint digital template; According to the occurrence frequency of each feature point in the initial fingerprint digital template and the feature damage index of the matched feature point of each feature point in the initial fingerprint digital template, the stability coefficient of each feature point in the initial fingerprint digital template is obtained.
6. An identity authentication method for mobile secure payment according to claim 5, It is characterized in that The specific calculation formula for obtaining the stability coefficient of each feature point in the initial fingerprint digital template according to the occurrence frequency of each feature point in the initial fingerprint digital template and the feature damage index of the matched feature point of each feature point in the initial fingerprint digital template is: Where W b_v represents the stability coefficient of the vth feature point in the initial fingerprint digital template; f b_v represents the frequency of occurrence of the vth feature point in the initial fingerprint digital template; q v represents the number of matched feature points of the vth feature point; S b_v_(x+1) represents the feature damage index of the x+1th matched feature point of the vth feature point in the initial fingerprint template; S b_v_x represents the feature damage index of the xth matched feature point of the vth feature point in the initial fingerprint template; t v It represents the mean of the time intervals between the acquisition times of the real-time fingerprint digital templates corresponding to all two matched feature points of the vth feature point in the initial fingerprint digital template; || is the absolute value function; exp() is an exponential function with a natural constant as the base; norm() is a linear normalization function.
7. The identity authentication method for mobile secure payment according to claim 1, It is characterized in that The specific steps of obtaining the update learning rate of each feature point in the initial fingerprint digital template according to the stability coefficient of each feature point in the initial fingerprint digital template and the fingerprint damage index of each real-time fingerprint digital template are as follows: Matching the initial fingerprint digital template with all real-time fingerprint digital templates respectively to obtain the matching result of each real-time fingerprint digital template, wherein the matching result includes a successful match and a failed match; The ratio of the number of real-time fingerprint digital templates that failed to match to the number of all real-time fingerprint digital templates is recorded as the matching failure rate of all real-time fingerprint digital templates; The ratio of the number of successfully matched real-time fingerprint digital templates in the real-time fingerprint digital template where the matched feature point of the vth feature point in the initial fingerprint digital template is located to the number of all real-time fingerprint digital templates is recorded as the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates; According to the matching failure rate of all real-time fingerprint digital templates, the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates, the fingerprint damage index of all real-time fingerprint digital templates and the stability coefficient of each feature point in the initial fingerprint digital template, the update learning rate of each feature point in the initial fingerprint digital template is obtained.
8. An identity authentication method for mobile secure payment according to claim 7, It is characterized in that The specific calculation formula for obtaining the update learning rate of each feature point in the initial fingerprint digital template according to the matching failure rate of all real-time fingerprint digital templates, the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates, the fingerprint damage index of all real-time fingerprint digital templates and the stability coefficient of each feature point in the initial fingerprint digital template is: In the formula, α b_v represents the update learning rate of the vth feature point in the initial fingerprint digital template; p represents the number of all real-time fingerprint digital templates; D a Indicates the matching failure rate of all real-time fingerprint digital templates; represents the mean value of fingerprint damage index of all real-time fingerprint digital templates; W b_v represents the stability coefficient of the vth feature point in the initial fingerprint digital template; f b_v represents the frequency of occurrence of the vth feature point in the initial fingerprint digital template; P b_v It represents the matching success rate of the vth feature point in the initial fingerprint digital template in all real-time fingerprint digital templates; exp() is an exponential function with a natural constant as the base; norm[] is a linear normalization function.
9. The identity authentication method for mobile secure payment according to claim 1, It is characterized in that The method updates each vector dimension data of each feature point in the initial fingerprint digital template according to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the matched feature point to obtain the updated fingerprint digital template, including the following specific steps: According to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the corresponding matched feature point, each vector dimension data of the feature vector corresponding to each feature point after the initial fingerprint digital template is updated is obtained; wherein each feature point corresponds to a feature vector, and each feature vector contains a number of vector dimension data; In the initial fingerprint digital template, each vector dimension data of the feature vector corresponding to each feature point is replaced with each vector dimension data of the feature vector corresponding to each feature point after update, so as to obtain an updated fingerprint digital template.
10. An identity authentication method for mobile secure payment according to claim 9, It is characterized in that The specific calculation formula for obtaining each vector dimension data of the feature vector corresponding to each feature point after the initial fingerprint digital template is updated according to the update learning rate of each feature point in the initial fingerprint digital template and the difference between the vector dimension data of each feature point and the corresponding matched feature point is: In the formula, h′ b_vu Indicates the u-th vector dimension data of the feature vector corresponding to the v-th feature point after the initial fingerprint digital template is updated; h b_vu Represents the u-th vector dimension data of the feature vector corresponding to the v-th feature point of the initial fingerprint digital template; α b_v represents the update learning rate of the vth feature point in the initial fingerprint digital template; k v Indicates the number of matched feature points of the vth feature point; h a_vu_i Represents the u-th vector dimension data of the feature vector corresponding to the i-th matched feature point of the v-th feature point in the initial fingerprint digital template.
