A user authentication method and device incorporating biometrics

Abstract

The application relates to a user identification method and device combining biological characteristics, and relates to the field of user identification data processing, which comprises the following steps: identifying a palmprint angle and an image angle, analyzing palmprint angle similarities of the palmprint angle and a plurality of historical palmprint angle sequences, and analyzing image angle similarities of the image angle and a plurality of historical image angle sequences; obtaining a plurality of amount fluctuation similarities, combining the plurality of palmprint angle similarities and the plurality of image angle similarities, compensating and adjusting a plurality of palmprint probabilities and a plurality of image probabilities, obtaining a plurality of user probabilities, and outputting a user with the maximum user probability as a user identification result. The application can solve the technical problems that the traditional user identification method has low adaptability in a complex environment, cannot effectively identify the identity of a user, and has insufficient identification accuracy and reliability; through multi-modal information fusion, the precision, accuracy and reliability of user identity identification in a complex environment can be significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of user identification data processing, and in particular to a user identification method and device combining biological characteristics. BACKGROUND

[0002] With the wide application of intelligent devices and self-service systems, user identity identification technology has become increasingly important in the fields of payment, security, intelligent device control, etc. In these application scenarios, the accuracy and reliability of identity verification are directly related to user experience and system security.

[0003] Traditional user identification methods are mainly based on single biological feature recognition, such as palmprint recognition, fingerprint recognition, face recognition, etc. Although these methods can provide good recognition results in specific environments and conditions, in complex environments, background noise, chaotic visual information or irregular physical environment (such as crowded places, dynamic background) may interfere with the recognition process, resulting in misidentification or missed identification problems, greatly limiting the recognition accuracy and reliability. SUMMARY

[0004] The present application provides a user identification method and device combining biological characteristics to solve the technical problem of low adaptability of traditional user identification methods based on single biological characteristics in complex environments, which cannot effectively identify user identity and have insufficient identification accuracy and reliability.

[0005] The technical solution of the present application to solve the above technical problems is as follows:

[0006] In a first aspect, the present application provides a user identification method combining biological characteristics, comprising: collecting palmprint information of a user through a palmprint collection device, performing user palmprint recognition, obtaining a plurality of pending users and a plurality of palmprint probabilities; when the largest palmprint probability is not greater than a preset probability threshold, collecting image information of the user through an image collection device, performing user image recognition, and obtaining a plurality of image probabilities; identifying the palmprint angle of the palmprint information and the image angle of the image information, obtaining a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of pending users, analyzing a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences, and analyzing a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences; obtaining a current charge amount, and obtaining a plurality of historical amount sequences of the plurality of pending users, respectively analyzing a plurality of amount fluctuation similarities, combining the plurality of palmprint angle similarities and the plurality of image angle similarities, compensating and adjusting the plurality of palmprint probabilities and the plurality of image probabilities, obtaining a plurality of user probabilities, outputting a pending user with the largest user probability as a user identification result.

[0007] In a second aspect, the present application provides a user identification device combining biological characteristics, comprising: a user palmprint identification module, configured to collect palmprint information of a user through a palmprint collection device, perform user palmprint identification, and obtain a plurality of undetermined users and a plurality of palmprint probabilities; a user image identification module, configured to, when the maximum palmprint probability is not greater than a preset probability threshold, collect image information of the user through an image collection device, perform user image identification, and obtain a plurality of image probabilities; an angle similarity analysis module, configured to identify a palmprint angle of the palmprint information and an image angle of the image information, obtain a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of undetermined users, analyze a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences, and analyze a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences; and a user identification result output module, configured to obtain a current charging amount, obtain a plurality of historical amount sequences of the plurality of undetermined users, respectively analyze a plurality of amount fluctuation similarities, combine the plurality of palmprint angle similarities and the plurality of image angle similarities, compensate and adjust the plurality of palmprint probabilities and the plurality of image probabilities, obtain a plurality of user probabilities, and output an undetermined user with the maximum user probability as a user identification result.

[0008] The present application has the following beneficial effects: palmprint information of a user is collected to perform user palmprint identification, a plurality of undetermined users and a plurality of palmprint probabilities are obtained; then when the maximum palmprint probability is not greater than a preset probability threshold, image information of the user is collected to perform user image identification, a plurality of image probabilities are obtained; then a palmprint angle of the palmprint information and an image angle of the image information are identified, a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of undetermined users are obtained; further, a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences are analyzed, and a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences are analyzed; on the other hand, a current charging amount is obtained, a plurality of historical amount sequences of the plurality of undetermined users are obtained, and a plurality of amount fluctuation similarities are respectively analyzed; finally, according to the plurality of amount fluctuation similarities, the plurality of palmprint angle similarities and the plurality of image angle similarities, the plurality of palmprint probabilities and the plurality of image probabilities are compensated and adjusted, a plurality of user probabilities are obtained, and an undetermined user with the maximum user probability is output as a user identification result; that is, through multi-modal information fusion, the identification probability of the user can be compensated and optimized in multiple dimensions, avoiding possible misidentification or missed identification of a single feature, thereby significantly improving the precision, accuracy and reliability of user identity identification in a complex environment. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 A flowchart of a user identification method combining biological characteristics provided by the present application;

[0010] Figure 2 A structural schematic diagram of a user identification device combined with biological characteristics is provided in the present application.

[0011] In the drawings, the components represented by the respective reference numerals are described as follows:

[0012] A user palmprint identification module 01, a user image identification module 02, an angle similarity analysis module 03, and a user identification result output module 04. DETAILED DESCRIPTION

[0013] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0014] In the description of the present application, the terms "first", "second" are used only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0015] In the description of the present application, the term "for example" is used to indicate "as an example, illustration or explanation". Any embodiment described as "for example" in the present application is not necessarily interpreted as more preferred or more advantageous than other embodiments. The following description is given in order to enable any person skilled in the art to implement and use the present application. In the following description, details are listed for the purpose of explanation. It should be understood that those skilled in the art can realize the present application without using these specific details. In other examples, well-known structures and processes will not be described in detail to avoid unnecessary details making the description of the present application obscure. Therefore, the present application is not intended to be limited to the shown embodiments, but is consistent with the broadest scope of principles and features disclosed.

[0016] Embodiment one, as shown in the present application, provides a user identification method combined with biological characteristics, specifically comprising the following steps: Figure 1

[0017] S100: Collecting palmprint information of a user through a palmprint collection device, performing user palmprint identification, and obtaining a plurality of undetermined users and a plurality of palmprint probabilities.

[0018] ​Further, the step S100 of the present application further comprises:

[0019] S110: collecting palmprint information of the user through the palmprint collection device; S120: inputting the palmprint information into a pre-constructed palmprint recognition channel, identifying the similarity of the palmprint information with a plurality of sample palmprint information in the palmprint recognition channel, outputting a plurality of palmprint similarities and corresponding users with the largest palmprint similarity, and obtaining a plurality of pending users, wherein the palmprint recognition channel is obtained by training based on a sample palmprint information set and a sample palmprint similarity set; S130: outputting the palmprint similarity corresponding to the plurality of pending users as a plurality of palmprint probabilities.

[0020] Specifically, in the data collection process, all user information (such as palmprint, face image, etc.) is collected under the premise of legality and explicit authorization of the user. When the user first uses the related service, the system will inform the user of the purpose, scope, and use of data collection through the user agreement or privacy policy, and the user needs to explicitly agree to the data collection. At the same time, the user can withdraw consent at any time, and the system will stop collecting and processing user information. On the other hand, the collected biometric data (such as palmprint image, face image) will be strictly protected by security measures (such as encrypted storage and transmission) to ensure that user information is not misused, leaked, or accessed by unauthorized third parties.

[0021] In the charging scene of a self-service charging machine in a convenience store, using a palmprint collection device to collect the palmprint information of the user as a way of identity verification can improve the security and convenience of payment. First, when the user operates the charging machine, the palm is placed on the sensor area of the palmprint collection device. The device is usually equipped with optical, infrared, or capacitive sensors that can capture the detailed features of the palmprint. Then, through the palmprint collection device, the palmprint information of the user is collected, including biometric information such as palmprint direction.

[0022] A convolutional neural network (CNN) is a kind of feedforward neural network, which is particularly good at processing multi-dimensional data such as two-dimensional images. Through the combination of multiple convolutional layers, pooling layers, and fully connected layers, it can automatically extract the features of the input data and perform classification or regression tasks. Based on the convolutional neural network, a palmprint recognition channel is constructed, which includes an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer. The input data of the input layer is palmprint information, and the output data of the output layer is palmprint similarity. The convolutional layer is used to extract local features such as texture and shape. The pooling layer is used to reduce the dimension of the data and enhance the robustness of the model. The fully connected layer is used to convert local features into global features and perform classification.

[0023] Then, based on the historical palmprint detection log, a sample palmprint information set and a sample palmprint similarity set are collected, and the sample palmprint information (palmprint direction, etc.) is taken as input and the sample palmprint similarity is taken as output. The sample palmprint information set and the sample palmprint similarity set are used to supervise the training of the palmprint recognition channel. First, during training, a group of palmprint image data is input each time. These data enter the network through the input layer of the CNN. The CNN automatically extracts the local features of the image through the convolutional layer and the pooling layer, and fuses these features through the fully connected layer. Then, the CNN outputs the similarity prediction value between the palmprints. The loss function (such as mean square error) is used to measure the difference between the prediction result and the true label. Then, through the calculation of the gradient of the loss function, the error is fed back to each layer in the network using the backpropagation algorithm. The weights and biases of each layer are adjusted according to the gradient descent algorithm to reduce the prediction error. Further, the weights and biases of the network are updated using an optimization algorithm (such as Adam optimizer, SGD, etc.) to gradually minimize the loss function. The sample data is iteratively trained until the loss function converges, and the trained palmprint recognition channel is output.

[0024] The multiple sample palmprint information is embedded into the palmprint recognition channel, where the multiple sample palmprint information is the palmprint information registered by the user or the palmprint information shared by other payment platforms, and the sample palmprint information has a unique user identifier. Then, the palmprint information is input into the palmprint recognition channel and compared with the multiple sample palmprint information in the palmprint recognition channel for similarity traversal recognition, and the maximum multiple (such as the top 10) palmprint similarities are output. The number of outputs can be set according to the actual scene. For example, when the interference is less, a relatively small number of palmprint similarities can be output, such as 5; when the interference is more, a relatively large number of palmprint similarities can be output, such as 10. The user identifiers corresponding to the maximum multiple palmprint similarities are obtained to obtain multiple pending users (users with high palmprint similarity). Then, the palmprint similarity is set as the palmprint probability, and the multiple palmprint probabilities corresponding to the multiple pending users are output. For example, the multiple palmprint probabilities are 94%, 92%, 90%, 89%, and 85%, respectively.

[0025] By embedding multiple sample palmprint information (including user registration information and other platform shared data) into the palmprint recognition channel, the similarity between real-time palmprint information and multiple sample information can be accurately identified, and according to the similarity, the system can output multiple most similar pending users and provide a basis for subsequent identity verification. This method not only improves the recognition accuracy, but also adjusts the number of recognition results flexibly according to the interference of the actual scene, thereby enhancing the reliability and adaptability of the system.

[0026] S200: When the maximum palmprint probability is not greater than the preset probability threshold, the image information of the user is collected through the image collection device, the user image recognition is performed, and multiple image probabilities are obtained.

[0027] Further, the step S200 of the present application further comprises:

[0028] S210: determining whether the maximum palmprint probability is greater than a preset probability threshold, if yes, outputting the palmprint user corresponding to the maximum palmprint probability as the user identification result; S220: if no, collecting image information of the user through an image collecting device; S230: inputting the image information into a pre-constructed image recognition channel to identify the similarity with sample image information of the plurality of pending users, and obtaining a plurality of image similarities, wherein the image recognition channel is obtained by training based on a sample image information set and a sample image similarity set; S240: outputting the plurality of image similarities as a plurality of image probabilities.

[0029] Specifically, a preset probability threshold is obtained, which is used to measure the accuracy of identification and can be set according to the identification accuracy requirement (such as 95%). If the maximum palmprint probability exceeds this threshold, it indicates that the current palmprint identification is accurate enough to effectively confirm the user identity. Then, it is determined whether the maximum palmprint probability is greater than the preset probability threshold, if the maximum palmprint probability is greater than the preset probability threshold, it indicates that the current identification result is reliable, and then the palmprint user corresponding to the maximum palmprint probability is outputted as the user identification result. If the maximum palmprint probability is less than or equal to the preset probability threshold, the image information (face image) of the user is collected through an image collecting device (such as a camera), at this time, the image information serves as a supplementary identity feature, which can improve the accuracy of identification.

[0030] An image recognition channel is constructed based on a convolutional neural network, and the image recognition channel includes an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer, wherein the input data of the input layer is image information (a human face image), and the output data of the output layer is image similarity. Then, a sample image information set and a sample image similarity set are collected, and the sample image information is used as input and the sample image similarity is used as output, and the sample image information set and the sample image similarity set are used to supervise the training of the image recognition channel. First, the sample image information is used as input, the sample image similarity is used as output label, the image similarity value is used as label, and the similarity between different images is indicated (usually calculated by a Euclidean distance, a cosine similarity, or the like); then, an error between the network output and the real label (i.e., the actual similarity) is calculated by a loss function (such as a mean square error loss function or a cross-entropy loss function), and the goal of the network is to minimize the loss function so that the image recognition result is as close as possible to the actual similarity value; then, the gradient of each layer is calculated by a back propagation algorithm, the weight in the network is updated, and the performance of the network is optimized, and an optimization algorithm such as a stochastic gradient descent (SGD) or an Adam optimizer is usually used to update the parameters of the model. Finally, the weight parameters of the network are continuously adjusted through multiple iteration training processes, the image recognition capability of the model is optimized, the network can recognize and distinguish the image features of different users, and the trained image recognition channel is obtained until the loss function converges.

[0031] Further, sample image information of the plurality of undetermined users is acquired (which can be obtained by data screening from a user information database); then, the image information and the sample image information of the plurality of undetermined users are respectively combined and input into the image recognition channel, and a plurality of image similarities are output; and the image similarity is set as an image probability, and a plurality of image probabilities are output. Through a supervised learning manner, the collected sample images and similarity information are used as training data, the convolutional neural network is used to extract image features and learn the similarity relationship between user identities, through multiple training, the network can efficiently recognize user identities according to input image information, and the palmprint recognition result is combined to realize accurate user identification.

[0032] S300: recognizing a palmprint angle of the palmprint information and an image angle of the image information, acquiring a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of undetermined users, analyzing a plurality of palmprint angle similarities of the palmprint angle and the plurality of historical palmprint angle sequences, and analyzing a plurality of image angle similarities of the image angle and the plurality of historical image angle sequences.

[0033] Further, the step S300 of the present application further includes:

[0034] S310: According to the user authentication data in the historical time, a sample palmprint information set and a sample image information set are collected, and the palmprint angle and the image angle of each sample palmprint information and each sample image information are labeled to obtain a sample palmprint angle set and a sample image angle set; S320: The sample palmprint information set and the sample palmprint angle set are used as supervised training data to train a palmprint angle recognition branch; S330: The sample image information set and the sample image angle set are used as supervised training data to train an image angle recognition branch; S340: The palmprint angle recognition branch and the image angle recognition branch are combined to obtain a biological feature angle recognition channel; S350: The palmprint information and the image information are input into the palmprint angle recognition branch and the image angle recognition branch in the biological feature angle recognition channel, and the palmprint angle and the image angle are obtained by identifying the output.

[0035] Specifically, user authentication data in a historical time (such as in the last month) is obtained, and a sample palmprint information set and a sample image information set are collected based on the user authentication data; then, the palmprint angle and the image angle of each sample palmprint information and each sample image information are labeled, wherein the angle of the palmprint refers to the angle of the user's palm in front of the collection device, and this angle affects the accuracy of palmprint recognition. The purpose of angle labeling is to describe the palmprint information of the user in different postures, which can be extracted by image analysis method, and the angle of each palmprint image is labeled. Exemplarily, the palmprint angle can include a horizontal angle of the palmprint, a pitch angle of 0 degrees, i.e., the user's palm is at a horizontal angle and is perpendicular to the user's body plane, and can also include a horizontal angle of 15 degrees to the left relative to the user's direction, a pitch angle of 15 degrees upward to the left, a pitch angle of 15 degrees downward to the right, or a horizontal angle of 15 degrees to the right, etc.

[0036] Similarly, the face image of the user will also be affected by the angle when collected. The image angle refers to the angle of the face image in front of the camera, which may affect the effect of face recognition. By image processing technology or deep learning model (such as face recognition algorithm), the facial features in the image are extracted and the angle of the image is calculated, for example, by recognizing the relative positions of the eyes, nose and mouth, the angle between the device is calculated; a sample palmprint angle set and a sample image angle set are obtained. Exemplarily, the image angle of the face image can be 0 degrees, i.e., directly facing the camera, or can be 15 degrees left or 15 degrees right along the Y-axis relative to the user's direction, and the Y-axis is the coordinate axis perpendicular to the horizontal plane.

[0037] The palmprint angle recognition branch is constructed based on a convolutional neural network, an input of the palmprint angle recognition branch is palmprint information, and output data is a palmprint angle; then, sample palmprint information is taken as an input, sample palmprint angles are taken as supervision, the sample palmprint information set and the sample palmprint angle set are taken as supervised training data, and the palmprint angle recognition branch is supervised training, first, a loss function suitable for a regression task is selected, such as a mean square error (MSE) loss function, the loss function measures the difference between the predicted angle and the true angle; then, by using the sample palmprint information set and the sample palmprint angle set as training data, each input palmprint image has a corresponding angle label, by comparing the difference (that is, the loss) between the predicted angle and the true angle, the weights in the network are adjusted; further, the gradient of the loss function with respect to each weight is calculated using a back propagation algorithm, and the weights in the network are adjusted by a gradient descent method, gradually reducing the error between the predicted angle and the true angle. The network is updated through multiple iterations, and the parameters of the network are constantly optimized until the loss function converges or a preset stopping condition is reached, and the trained palmprint angle recognition branch is obtained.

[0038] On the other hand, the image angle recognition branch is constructed based on a convolutional neural network, an input of the image angle recognition branch is image information, and an output is an image angle; then, sample image information is taken as an input, sample image angles are taken as supervision, the sample image information set and the sample image angle set are taken as supervised training data to train the image angle recognition branch until convergence, and the trained image angle recognition branch is obtained, wherein the training method of the image angle recognition branch is the same as the training method of the above-mentioned palmprint angle recognition branch, and will not be expanded here.

[0039] Then, the palmprint angle recognition branch and the image angle recognition branch are combined to obtain a biometric feature angle recognition channel; then, the palmprint information and the image information are respectively input into the palmprint angle recognition branch and the image angle recognition branch in the biometric feature angle recognition channel for angle recognition, and a palmprint angle and an image angle are output.

[0040] Further, the step S300 of the present application further comprises:

[0041] S360: collecting a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences according to historical authentication data records of the plurality of pending users; S370: calculating a first user average palmprint angle of a first historical palmprint angle sequence of a first pending user; S380: calculating a palmprint angle deviation ratio of the palmprint angle and the first user average palmprint angle, and obtaining a first palmprint angle similarity by subtracting the palmprint angle deviation ratio from 1; S390: continuing to calculate a plurality of palmprint angle similarities of the palmprint angle and the plurality of historical palmprint angle sequences, and calculating a plurality of image angle similarities of the image angle and the plurality of historical image angle sequences.

[0042] Specifically, according to historical authentication data records of the plurality of pending users, a plurality of historical palmprint angle sequences (palmprint angle information of a user in the past period of time) and a plurality of historical image angle sequences (face image angle information of a user in the past period of time) in a historical time period (such as the last month) are collected; then, a first pending user is randomly selected from the plurality of pending users, a first historical palmprint angle sequence of the first pending user is obtained, and a plurality of first historical palmprint angles in the first historical palmprint angle sequence are averaged to obtain a first user average palmprint angle. Further, a palmprint angle deviation ratio of the palmprint angle and the first user average palmprint angle is calculated, wherein the palmprint angle deviation ratio is a ratio of a first angle difference between the palmprint angle and the first user average palmprint angle to the first user average palmprint angle; then, 1 is subtracted from the palmprint angle deviation ratio, and a difference value between the two is taken as a first palmprint angle similarity, wherein the greater the palmprint angle similarity, the higher the consistency of the palmprint angle of the current user with the historical record. Then, a plurality of palmprint angle similarities of the palmprint angle and the plurality of historical palmprint angle sequences are calculated by using the same method.

[0043] On the other hand, a plurality of average image angles of the plurality of historical image angle sequences are calculated, and a plurality of image angle similarities are calculated according to the image angle and the plurality of average image angles.

[0044] S400: obtaining a current charge amount, and obtaining a plurality of historical amount sequences of the plurality of pending users, respectively analyzing to obtain a plurality of amount fluctuation similarities, combining the plurality of palmprint angle similarities and the plurality of image angle similarities to compensate and adjust the plurality of palmprint probabilities and the plurality of image probabilities, obtaining a plurality of user probabilities, and outputting a pending user with the largest user probability as a user authentication result.

[0045] Further, the step S400 of the present application further comprises:

[0046] S410: obtaining a current charging amount; S420: collecting a plurality of historical amount sequences according to historical consumption data records of the plurality of pending users; S430: calculating a plurality of user average amounts of the plurality of historical amount sequences; S440: calculating a plurality of amount deviation ratios of the charging amount and the plurality of user average amounts, respectively adopting 1 minus the plurality of amount deviation ratios to obtain a plurality of amount fluctuation similarities.

[0047] Specifically, a current charging amount is obtained, such as a real-time recorded consumption amount when a user purchases goods at a self-service charging machine of a convenience store; then a plurality of historical amount sequences in a historical time period (such as the last month) are collected according to historical consumption data records of the plurality of pending users. The plurality of historical amount sequences are respectively subjected to mean value calculation to obtain a plurality of user average amounts. Then a plurality of amount deviation ratios of the charging amount and the plurality of user average amounts are respectively calculated, wherein the amount deviation ratio is a ratio of an absolute value of a difference between the charging amount and the user average amount to the user average amount, for example, assuming that the charging amount is 60 yuan and the user average amount is 50 yuan, then the deviation ratio is (60-50) / 50=0.2. Further, 1 minus the plurality of amount deviation ratios is respectively adopted to obtain a plurality of amount fluctuation similarities.

[0048] By combining the current charging amount and the historical consumption amount of the plurality of pending users, the deviation ratio and the amount fluctuation similarity are calculated, which can effectively improve the identification accuracy of the user identity. In a complex scene, the amount fluctuation similarity provides another effective matching dimension, together with other biological feature information (such as palmprint and face angle similarity), to provide more accurate and reliable user identity identification.

[0049] Further, the step S400 of the present application further comprises:

[0050] S450: respectively calculating a plurality of compensation adjustment coefficients using the plurality of palmprint angle similarities, the plurality of image angle similarities and the plurality of amount fluctuation similarities; S460: multiplying the plurality of compensation palmprint probabilities and the plurality of compensation image probabilities by the plurality of compensation adjustment coefficients respectively to obtain the plurality of compensation palmprint probabilities and the plurality of compensation image probabilities; S470: performing weighted calculation on the plurality of compensation palmprint probabilities and the plurality of compensation image probabilities to obtain a plurality of user probabilities, and outputting a pending user with the largest user probability as a user identification result.

[0051] Specifically, the plurality of palmprint angle similarities, the plurality of image angle similarities and the plurality of amount fluctuation similarities are integrated, and the palmprint angle similarity, the image angle similarity and the amount fluctuation similarity of the first undetermined user are randomly selected; the weight proportions of the palmprint angle, the image angle and the amount fluctuation index are set respectively, wherein the sum of the weights of the three is 1, the influence degree of the index on the accuracy of user identification is set, and the greater the influence degree, the greater the corresponding weight; the palmprint angle similarity, the image angle similarity and the amount fluctuation similarity are weighted and calculated according to the weight proportions, and the first compensation adjustment coefficient of the first undetermined user is obtained, for example, assuming that the palmprint angle similarity, the image angle similarity and the amount fluctuation similarity of the first user are 0.9, 0.8 and 0.8 respectively, and the weight proportions of the palmprint angle, the image angle and the amount fluctuation index are 0.5, 0.4 and 0.1 respectively, then the compensation adjustment coefficient of the first undetermined user is 0.9*0.5+0.8*0.4+0.8*0.1=0.85, and a plurality of compensation adjustment coefficients of a plurality of undetermined users are calculated.

[0052] Then the plurality of compensation adjustment coefficients are used to multiply the corresponding palmprint probability and image probability of the undetermined user respectively, and a plurality of compensation palmprint probabilities and a plurality of compensation image probabilities are obtained, for example, assuming that the compensation adjustment coefficient of the first undetermined user is 0.85, and the palmprint probability and the image probability are 0.95 and 0.9 respectively, then the compensation palmprint probability of the first undetermined user is 0.85*0.95=0.8275, and the compensation image probability is 0.85*0.9=0.765. The trusted weights of the palmprint identification and the image identification (face identification) are set respectively, wherein the sum of the weights of the two is 1, the trusted weights can be set according to the historical identification accuracy of the palmprint identification and the image identification, the higher the accuracy, the greater the corresponding trusted weight, and the palmprint identification weight and the image identification weight are obtained. Finally, the plurality of compensation palmprint probabilities and the plurality of compensation image probabilities are weighted and calculated according to the palmprint identification weight and the image identification weight, and a plurality of user probabilities are obtained; and the undetermined user with the maximum user probability in the plurality of user probabilities is output as the user identification result. Optionally, the mean value of the compensation palmprint probability and the compensation image probability of the plurality of undetermined users can also be calculated as the user probability.

[0053] By calculating the compensation adjustment coefficients of the palmprint angle similarity, the image angle similarity and the amount fluctuation similarity respectively, and adjusting the palmprint probability, the image probability and the amount probability according to the compensation adjustment coefficients, and finally obtaining the final user probability through weighted calculation, this method of comprehensively considering multi-dimensional information can effectively improve the accuracy and robustness of user identity identification, and has a significant advantage in accurately identifying the user identity in a complex environment.

[0054] The user identification method provided by the embodiment of the application has at least the following technical effects:

[0055] By collecting palmprint information of a user to perform user palmprint identification, a plurality of pending users and a plurality of palmprint probabilities are obtained; then when the maximum palmprint probability is not greater than a preset probability threshold, image information of the user is collected to perform user image identification, a plurality of image probabilities are obtained; then a palmprint angle of the palmprint information and an image angle of the image information are identified, a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of pending users are obtained; further, a plurality of palmprint angle similarities of the palmprint angle and the plurality of historical palmprint angle sequences are analyzed, and a plurality of image angle similarities of the image angle and the plurality of historical image angle sequences are analyzed; on the other hand, a current charge amount is obtained, and a plurality of historical amount sequences of the plurality of pending users are obtained, a plurality of amount fluctuation similarities are respectively obtained by analysis; finally, according to the plurality of amount fluctuation similarities, the plurality of palmprint angle similarities and the plurality of image angle similarities, the plurality of palmprint probabilities and the plurality of image probabilities are compensated and adjusted, a plurality of user probabilities are obtained, and a pending user with the maximum user probability is output as a user identification result; that is, by multi-modal information fusion, the identification probability of the user can be compensated and optimized in multiple dimensions, avoiding the possible misidentification or missed identification of a single feature, thereby significantly improving the precision, accuracy and reliability of user identity identification in a complex environment.

[0056] Embodiment two, as shown in Figure 2 based on the same inventive concept of the user identification method provided in embodiment one, the present embodiment also provides a user identification device combining biological characteristics, comprising:

[0057] The user palmprint identification module 01 is configured to collect palmprint information of a user through a palmprint collection device, perform user palmprint identification, and obtain a plurality of undetermined users and a plurality of palmprint probabilities; the user image identification module 02 is configured to, when the maximum palmprint probability is not greater than a preset probability threshold, collect image information of the user through an image collection device, perform user image identification, and obtain a plurality of image probabilities; the angle similarity analysis module 03 is configured to identify a palmprint angle of the palmprint information and an image angle of the image information, obtain a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of undetermined users, analyze a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences, and analyze a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences; and the user identification result output module 04 is configured to obtain a current charge amount, obtain a plurality of historical amount sequences of the plurality of undetermined users, respectively analyze a plurality of amount fluctuation similarities, combine the plurality of palmprint angle similarities and the plurality of image angle similarities, compensate and adjust the plurality of palmprint probabilities and the plurality of image probabilities, obtain a plurality of user probabilities, and output an undetermined user with the maximum user probability as a user identification result.

[0058] The user identification device further comprises a palmprint collection device configured to collect palmprint information of a user; an image collection device configured to collect image information of the user; a palmprint recognition channel pre-constructed and configured to identify similarities between the palmprint information and a plurality of sample palmprint information in the palmprint recognition channel, output a plurality of maximum palmprint similarities and corresponding users, and obtain a plurality of undetermined users; and an image recognition channel pre-constructed and configured to identify similarities between the image information and sample image information of the plurality of undetermined users, obtain a plurality of image similarities, and output a plurality of image probabilities.

[0059] The user identification device further comprises a palmprint collection device configured to collect palmprint information of a user; an image collection device configured to collect image information of the user; a palmprint recognition channel pre-constructed and configured to identify similarities between the palmprint information and a plurality of sample palmprint information in the palmprint recognition channel, output a plurality of maximum palmprint similarities and corresponding users, and obtain a plurality of undetermined users; and an image recognition channel pre-constructed and configured to identify similarities between the image information and sample image information of the plurality of undetermined users, obtain a plurality of image similarities, and output a plurality of image probabilities.

[0060] The user identification device combined with biological characteristics is also used for: collecting sample palmprint information sets and sample image information sets according to user identification data in a historical time, and labeling the palmprint angle and the image angle of each sample palmprint information and each sample image information to obtain sample palmprint angle sets and sample image angle sets; using the sample palmprint information sets and the sample palmprint angle sets as supervised training data to train a palmprint angle identification branch; using the sample image information sets and the sample image angle sets as supervised training data to train an image angle identification branch; combining the palmprint angle identification branch and the image angle identification branch to obtain a biological characteristic angle identification channel; inputting the palmprint information and the image information into the palmprint angle identification branch and the image angle identification branch in the biological characteristic angle identification channel to identify and output the palmprint angle and the image angle.

[0061] The user identification device combined with biological characteristics is also used for: collecting a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences according to historical identification data records of the plurality of pending users; calculating a first user average palmprint angle of a first historical palmprint angle sequence of a first pending user; calculating a palmprint angle deviation ratio of the palmprint angle and the first user average palmprint angle, and obtaining a first palmprint angle similarity by subtracting the palmprint angle deviation ratio from 1; continuing to calculate a plurality of palmprint angle similarities of the palmprint angle and the plurality of historical palmprint angle sequences, and calculating a plurality of image angle similarities of the image angle and the plurality of historical image angle sequences.

[0062] The user identification device combined with biological characteristics is also used for: obtaining a current charge amount; collecting a plurality of historical amount sequences according to historical consumption data records of the plurality of pending users; calculating a plurality of user average amounts of the plurality of historical amount sequences; calculating a plurality of amount deviation ratios of the charge amount and the plurality of user average amounts, and obtaining a plurality of amount fluctuation similarities by subtracting the plurality of amount deviation ratios from 1 respectively.

[0063] The user identification device combined with biological characteristics is also used for: using the plurality of palmprint angle similarities, the plurality of image angle similarities and the plurality of amount fluctuation similarities to calculate a plurality of compensation adjustment coefficients respectively; using the plurality of compensation adjustment coefficients to multiply the palmprint probability and the image probability of the corresponding pending user respectively to obtain a plurality of compensated palmprint probabilities and a plurality of compensated image probabilities; performing weighted calculation on the plurality of compensated palmprint probabilities and the plurality of compensated image probabilities to obtain a plurality of user probabilities, and outputting a pending user with the largest user probability as a user identification result.

[0064] It should be noted that the above-mentioned embodiments are merely intended to illustrate the technical solutions of the present application, and thus cannot be used to limit the scope of the present application.

[0065] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, device, or computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer-usable program code.

[0066] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (apparatuses), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as a combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate an apparatus for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in one or more blocks.

[0067] These computer program instructions can also be stored in a computer-readable memory that can direct the computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction apparatuses that implement the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in one or more blocks.

[0068] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are performed on the computer or other programmable data processing device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide a process for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in one or more blocks.

[0069] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to these embodiments once they understand the basic inventive concept.

[0070] It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover modifications and variations of this application provided they come within the scope of the application and their equivalent technology.

Claims

1. A user authentication method incorporating a biometric feature, characterized by, The method comprises: Collecting palmprint information of a user through a palmprint collection device to perform user palmprint identification and obtain a plurality of pending users and a plurality of palmprint probabilities; When the maximum palmprint probability is not greater than a preset probability threshold, collecting image information of the user through an image collection device to perform user image identification and obtain a plurality of image probabilities; Identifying a palmprint angle of the palmprint information and an image angle of the image information, obtaining a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of pending users, analyzing a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences, and analyzing a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences, comprising: According to historical identification data records of the plurality of pending users, collecting a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences; Calculating a first user average palmprint angle of a first historical palmprint angle sequence of a first pending user; Calculating a palmprint angle deviation ratio of the palmprint angle and the first user average palmprint angle, obtaining a first palmprint angle similarity by subtracting the palmprint angle deviation ratio from 1; Continuously calculating a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences, and a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences; Obtaining a current charge amount, obtaining a plurality of historical amount sequences of the plurality of pending users, respectively analyzing a plurality of amount fluctuation similarities, combining the plurality of palmprint angle similarities and the plurality of image angle similarities, compensating and adjusting the plurality of palmprint probabilities and the plurality of image probabilities to obtain a plurality of user probabilities, and outputting a pending user with the maximum user probability as a user identification result; Wherein, combining the plurality of palmprint angle similarities and the plurality of image angle similarities, compensating and adjusting the plurality of palmprint probabilities and the plurality of image probabilities to obtain a plurality of user probabilities, and outputting a pending user with the maximum user probability as a user identification result, comprises: Using the plurality of palmprint angle similarities, the plurality of image angle similarities, and the plurality of amount fluctuation similarities to respectively calculate a plurality of compensation adjustment coefficients; Using the plurality of compensation adjustment coefficients to respectively multiply the palmprint probabilities and the image probabilities of the corresponding pending users to obtain a plurality of compensated palmprint probabilities and a plurality of compensated image probabilities; Weighted calculating the plurality of compensated palmprint probabilities and the plurality of compensated image probabilities to obtain a plurality of user probabilities, and outputting a pending user with the maximum user probability as a user identification result. 2.The user authentication method of claim 1, wherein, Collecting palmprint information of a user through a palmprint collection device to perform user palmprint identification and obtain a plurality of pending users and a plurality of palmprint probabilities, comprising: Collecting palmprint information of a user through a palmprint collection device; Inputting the palmprint information into a pre-constructed palmprint identification channel to identify similarities with a plurality of sample palmprint information in the palmprint identification channel, outputting a plurality of maximum palmprint similarities and corresponding users, and obtaining a plurality of pending users, wherein the palmprint identification channel is obtained by training based on a sample palmprint information set and a sample palmprint similarity set; Output the palm print similarity corresponding to the plurality of pending users as a plurality of palm print probabilities. 3.The user authentication method of claim 1, wherein, When the maximum palm print probability is not greater than a preset probability threshold, collect image information of the user through an image collection device, perform user image recognition, and obtain a plurality of image probabilities, including: Determine whether the maximum palm print probability is greater than the preset probability threshold, and if so, output the palm print user corresponding to the maximum palm print probability as a user identification result; If not, collect image information of the user through an image collection device; Input the image information into a pre-constructed image recognition channel to identify the similarity of the sample image information of the plurality of pending users, and obtain a plurality of image similarities, wherein the image recognition channel is trained based on a sample image information set and a sample image similarity set; Output the plurality of image similarities as a plurality of image probabilities.

4. The user authentication method incorporating a biological feature according to claim 1, characterized by, Identify the palm print angle of the palm print information and the image angle of the image information, including: Collect a sample palm print information set and a sample image information set according to user identification data in a historical time, and label the palm print angle and the image angle of each sample palm print information and each sample image information to obtain a sample palm print angle set and a sample image angle set; Use the sample palm print information set and the sample palm print angle set as supervised training data to train a palm print angle recognition branch; Use the sample image information set and the sample image angle set as supervised training data to train an image angle recognition branch; Combine the palm print angle recognition branch and the image angle recognition branch to obtain a biometric feature angle recognition channel; Input the palm print information and the image information into the palm print angle recognition branch and the image angle recognition branch in the biometric feature angle recognition channel to identify and output the palm print angle and the image angle.

5. The user authentication method incorporating a biological feature according to claim 1, characterized by, Obtain the current charging amount and a plurality of historical amount sequences of the plurality of pending users, and analyze to obtain a plurality of amount fluctuation similarities, including: Obtain the current charging amount; Collect a plurality of historical amount sequences according to historical consumption data records of the plurality of pending users; Calculate a plurality of user average amounts of the plurality of historical amount sequences; Calculate a plurality of amount deviation ratios of the charging amount and the plurality of user average amounts, and obtain a plurality of amount fluctuation similarities by subtracting the plurality of amount deviation ratios from 1 respectively.

6. A user authentication device incorporating a biometric feature, characterized by Steps for implementing the user identification method based on biometric features according to any one of claims 1 to 5, including: A user palm print recognition module for collecting palm print information of a user through a palm print collection device, performing user palm print recognition, and obtaining a plurality of pending users and a plurality of palm print probabilities; A user image recognition module for collecting image information of the user through an image collection device when the maximum palm print probability is not greater than a preset probability threshold, performing user image recognition, and obtaining a plurality of image probabilities; An angle similarity analysis module is configured to identify a palmprint angle of the palmprint information and an image angle of the image information, acquire a plurality of historical palmprint angle sequences and a plurality of historical image angle sequences of the plurality of undetermined users, analyze a plurality of palmprint angle similarities between the palmprint angle and the plurality of historical palmprint angle sequences, and analyze a plurality of image angle similarities between the image angle and the plurality of historical image angle sequences. A user identification result output module is configured to acquire a current charging amount, acquire a plurality of historical amount sequences of the plurality of undetermined users, respectively analyze a plurality of amount fluctuation similarities, combine the plurality of palmprint angle similarities and the plurality of image angle similarities, compensate and adjust the plurality of palmprint probabilities and the plurality of image probabilities, acquire a plurality of user probabilities, and output an undetermined user with the largest user probability as a user identification result.

Images