Virtual reality game user identity verification system

Abstract

The invention discloses a virtual reality game user identity verification system, which relates to the technical field of virtual reality security, and comprises an account management module, an information acquisition module, a biological feature verification module, a behavior feature verification module, an iris feature verification module and a storage module, the information acquisition module acquires face information, behavior mode information, iris information and voiceprint information of a user, the biological feature verification module comprises a face feature verification unit and a voiceprint feature verification unit, environment risk conditions are analyzed regularly in the game process, and by combining various kinds of biological feature information and behavior mode analysis, the user experience is improved. According to the method, the comprehensiveness and the safety of identity verification are enhanced, an environment risk coefficient and an iris feature verification module are introduced, the verification strictness is improved in a high-risk environment, and the safety and the accuracy of virtual reality user identity verification are improved by adaptively adjusting a verification threshold and flexibly coping with a dynamically changing use environment.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual reality security, and specifically to a virtual reality game user identity authentication system. Background Technique

[0002] Virtual reality technology is a computer simulation system that can create and experience virtual worlds. It uses a computer to generate a simulated environment, enabling users to immerse themselves in this environment. These phenomena can be real objects in reality or substances invisible to our naked eyes, presented through three-dimensional models. Since these phenomena are not directly visible to us but are simulated by computer technology to represent the real world, it is called virtual reality. When users enter the virtual reality scene through virtual reality devices, their operations and information in the virtual reality scene will be recorded by malicious programs, resulting in the leakage of virtual reality users' information. There are security risks for users' information, which may seriously cause property losses to users. Therefore, it is very necessary to ensure the security of virtual reality users' accounts.

[0003] Currently, for virtual reality game users, usually one of password, fingerprint recognition, or face recognition is used for identity authentication. The authentication method is relatively simple. With the development of forgery technology, the above authentication methods may lead to the risk of being bypassed. Moreover, due to the large changes in the virtual reality game environment, virtual reality devices are used in both home network and external public network environments, increasing the security risks of accounts. The impact of environmental factors on identity authentication is not considered. It cannot flexibly respond in a dynamically changing environment, resulting in poor authentication effects in high-risk environments and unable to fully protect the security of users' accounts. Summary of the Invention

[0004] The purpose of the present invention is to provide a virtual reality game user identity authentication system to solve the problems raised in the above background technique.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A virtual reality game user identity authentication system, including an account management module, an information collection module, a biometric authentication module, a behavior feature authentication module, an iris feature authentication module, and a storage module;

[0006] The information collection module collects users' facial information, behavior pattern information, iris information, and voiceprint information, and stores them through the storage module;

[0007] The biometric authentication module includes a facial feature authentication unit and a voiceprint feature authentication unit. The facial feature authentication unit compares the current user's face with the user's facial information collected during registration, and the voiceprint feature authentication unit is used to compare the current user's voice information with the voiceprint information collected during registration, and judges whether the voice is qualified in combination with the environmental risk situation;

[0008] The behavior feature verification module is used to compare the behavior pattern information of the current user with the behavior pattern information collected during registration to determine whether it is qualified;

[0009] During the game process, the environmental risk situation is analyzed regularly. When the environmental risk situation is too poor, the iris feature verification module compares the iris information of the current user with the iris information collected during registration. When the comparison is unqualified, the user needs to log in to the system again.

[0010] Optionally, the verification process of the facial feature verification unit is as follows:

[0011]

[0012] Where D(F u ,F i ) represents the Euclidean distance between the user's stored facial feature vector F u and the input facial feature vector F i ;

[0013] F u (k) represents the k-th element value of the user's facial feature vector;

[0014] F i (k) represents the k-th element value of the input image feature vector;

[0015] n represents the number of elements;

[0016] The Euclidean distance D(F u between the user's facial feature vector F i and the input image feature vector F u ,F i ) is smaller, indicating that the input image matches the facial information obtained by the facial information acquisition unit better. Otherwise, the gap is larger. And according to the experimental setting, the facial threshold of D(F u ,F i ) is Y1. When D(F u ,F i ) is less than the facial threshold Y1, it indicates that the facial verification is qualified. When D(F u ,F i ) is greater than the facial threshold Y1, it indicates that the facial features scanned by the current device have a large gap from the historical user facial information collected, and at this time the facial verification is unqualified.

[0017] Optionally, the verification process of the voiceprint feature verification unit is as follows:

[0018]

[0019] Where Svioce Indicates the voiceprint score;

[0020] V base Indicates the basic matching rate, obtained through the voiceprint matching algorithm, with a value range of 0 to 1;

[0021] D t Indicates the facial anomaly score, with a value of 0.2 when facial occlusion is detected and 0 when there is no facial occlusion;

[0022] R risk Indicates the environmental risk coefficient, with a value range of 0 to 1;

[0023] C d Indicates the time warping coefficient;

[0024] T e Indicates the device trust score, obtained from the device matching degree and location information, with a value of 0 to 1;

[0025] The time warping coefficient C d The derivation process is as follows:

[0026]

[0027] D d Indicates the DTW distance between the user's current voice and voiceprint information;

[0028] D b Indicates the DTW distance in the user's voiceprint information;

[0029] β represents the dynamic adjustment coefficient, β = 1 + 0.5 × R risk ;

[0030] The environmental risk coefficient R risk The derivation process is as follows:

[0031]

[0032] Where ML represents the network environment score;

[0033] SL represents the threat intelligence score;

[0034] The voiceprint score S vioce Illustrates the degree of consistency between the current user's voice information and the voiceprint information collected during registration. The larger the voiceprint score S vioce the higher the consistency, and vice versa. Set the voiceprint threshold of the voiceprint score S vioce to Y2. When the voiceprint score S vioce < voiceprint threshold Y2, it indicates non - compliance.

[0035] Optionally, the analysis process of the behavioral feature verification module is as follows:

[0036]

[0037]

[0038] Among them, S bvc represents the consistency score of the behavior pattern;

[0039] σ H represents the standard deviation of the historical behavior value, indicating the degree of fluctuation of the user's behavior;

[0040] Z is a positive number, with a value of 10 -5 , used to avoid a zero denominator;

[0041] ΔV represents the average value of the behavior change rate;

[0042] B x represents the x-th behavior index value of the current behavior pattern;

[0043] H x represents the x-th behavior index value in the historical behavior pattern;

[0044] t represents the time difference, used to standardize the behavior change, with the unit of seconds;

[0045] R risk represents the environmental risk coefficient, with a value range of 0 to 1;

[0046] α represents the environmental risk impact weight, with a value range of 0.5 to 1.5, used to adjust the impact of environmental risk on the behavior pattern;

[0047] The behavior pattern consistency score S bvc represents the degree of consistency between the current user's behavior pattern information and the behavior pattern information collected during registration. The behavior pattern consistency score S bvc The larger it is, the more similar the behaviors are. The smaller it is, the greater the behavior change. Set the behavior threshold of the behavior pattern consistency score S bvc to Y3. When the behavior pattern consistency score S bvc > the behavior threshold Y3, it indicates that the verification is qualified. At this time, the user can log in to the system to play the game.

[0048] Optionally, the set environmental risk coefficient R risk has an environmental threshold of 0.6. When the environmental risk coefficient R risk > 0.6, it indicates that the environmental risk is too high. At this time, the user during the game is verified through the iris feature verification module. The verification process is as follows:

[0049]

[0050] Among them, D(A u,A i ) represents the difference between the user's stored iris image and the output iris image features;

[0051] E u (k) represents the k-th texture feature value of the user's stored iris image;

[0052] E i (k) represents the k-th texture feature value of the output iris image;

[0053] W 1 represents the texture feature influence coefficient;

[0054] R u (k) represents the k-th color feature value of the user's stored iris image;

[0055] R i (k) represents the k-th color feature value of the output iris image;

[0056] W 2 represents the color feature influence coefficient;

[0057] y is the number of color features and texture features;

[0058] By performing weighted summation on the differences between the texture features and color features of the iris, the overall difference between two iris images is obtained. The difference D(A u ,A i ) is smaller, indicating that the iris image features are more similar. According to the iris verification standard and experimental setting, the iris threshold of D(A u ,A i ) is 0.2. When D(A u ,A i ) is greater than 0.2, it indicates non-compliance. At this time, the user needs to re-pass the verification of the biometric verification module and the behavioral feature verification module.

[0059] Optionally, when D(A u ,A i ) is greater than 0.2, the game connection is automatically disconnected, and the user needs to re-login. At the same time, the face threshold Y1 is reduced, and the voiceprint threshold Y2 and the behavior threshold Y3 are increased to improve the identity verification strength.

[0060] Optionally, when the storage module performs data storage, it encrypts the data by combining a symmetric encryption algorithm and an asymmetric encryption algorithm, and configures a key management system to manage the encryption and decryption keys to ensure the security of the encrypted data.

[0061] Optionally, the account management module is used to manage user accounts, including user account registration, login, and binding of user accounts to virtual reality hardware IDs.

[0062] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0063] 1. By collecting various biometric information and behavior pattern information of users, during subsequent user login verification, verification is performed through the biometric verification module and the behavior feature verification module. Through the cooperation of different verification methods, user identity verification is more comprehensive, improving the accuracy and security of identity verification. And an environmental risk coefficient is introduced during the verification process to fully consider the impact of the user's network environment on identity verification. Compared with traditional static verification methods, the complexity of identity verification is increased, reducing the risk of being bypassed.

[0064] 2. By analyzing the environmental risk situation regularly during the game process and being able to verify the iris information of users at any time through the iris feature verification module, the security of user accounts is improved, and iris verification is only enabled when the environmental risk is too high, which helps to balance security and performance and rationally utilize computing resources. When the iris information verification fails, the system will automatically adjust each verification threshold to increase the strictness of subsequent verification. This way of adaptively adjusting the threshold can effectively increase the difficulty of verification, prevent user accounts from being bypassed by malicious attackers, thus flexibly coping with the dynamically changing virtual reality device usage environment, ensuring the quality of identity verification, and improving the security of the user identity verification system. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 is a block diagram of the system module of the present invention;

[0066] Figure 2 is a flowchart of the identity verification of the present invention;

[0067] Figure 3 is a block diagram of the information collection module of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0069] Embodiment 1:

[0070] Please refer to Figures 1 to 3, this embodiment provides a virtual reality game user authentication system, including an account management module, an information collection module, a biometric verification module, a behavior feature verification module, an iris feature verification module, and a storage module;

[0071] The account management module is used to manage user accounts, including user account registration, login, and binding with the virtual reality hardware ID;

[0072] When the storage module stores data, it uses a combination of symmetric encryption algorithm and asymmetric encryption algorithm to encrypt the data, and configures a key management system to manage encryption and decryption keys to ensure the security of the encrypted data;

[0073] When a user registers an account, the information collection module collects the user's facial information, behavior pattern information, iris information, and voiceprint information, and stores them encrypted through the storage module for subsequent verification;

[0074] The biometric verification module includes a facial feature verification unit and a voiceprint feature verification unit. The facial feature verification unit compares the current user's face with the user's facial information collected during registration. The voiceprint feature verification unit is used to compare the current user's voice information with the voiceprint information collected during registration, and combines the environmental risk situation to determine whether the voice is qualified;

[0075] The behavior feature verification module is used to compare the current user's behavior pattern information with the behavior pattern information collected during registration to determine whether it is qualified;

[0076] During the game process, the environmental risk situation is analyzed regularly. When the environmental risk situation is too poor, the iris feature verification module compares the current user's iris information with the iris information collected during registration. When the comparison is unqualified, the user needs to log in to the system again.

[0077] More specifically, in this embodiment: First, the user account information is bound to the virtual reality hardware ID through the account management module. The system can ensure that each account can only be used on a specific hardware device, effectively preventing account sharing, theft, or malicious login. When the user registers through the information collection module, the system will collect various biometric information and behavior pattern information of the user. When the user logs in for verification later, it is verified through the biometric verification module and the behavior feature verification module. By cooperating with different verification methods, the accuracy and security of identity verification are improved, the complexity of user identity verification is increased, the risk of being bypassed is reduced, and the environmental risk coefficient is introduced during the verification process to fully consider the impact of the user's network environment on identity verification. At the same time, the environmental risk situation is analyzed regularly during the game process, and the user can be verified for iris information at any time through the iris feature verification module. When the iris information verification fails, the system will automatically adjust each verification threshold to improve the strictness of subsequent verification. This way of adaptively adjusting the threshold can effectively increase the difficulty of verification, prevent the user account from being bypassed by malicious attackers, and thus flexibly respond to the dynamically changing virtual reality device usage environment, ensure the identity verification effect in high-risk environments, and improve the security of the user identity verification system.

[0078] Further, the verification process of the facial feature verification unit is as follows:

[0079]

[0080] Where D(F u ,F i ) represents the Euclidean distance between the user's facial feature vector F u and the input image feature vector F i . The user's facial feature vector F u is obtained by the facial information collection unit, and the input image feature vector F i is obtained by the facial feature verification unit;

[0081] F u (k) represents the k-th element value of the user's facial feature vector;

[0082] F i (k) is the k-th element value of the input image feature vector;

[0083] n represents the number of elements;

[0084] By calculating [F u (k)-F i (k)] 2, The first purpose is to eliminate positive and negative differences and prevent the same differences from canceling each other out due to different signs. The second purpose is to accentuate the impact of larger differences. If the difference between two features is large, the squaring operation will increase the impact of the difference and strengthen the influence of the difference on the final verification result;

[0085] By summing the squares of the differences across all dimensions, we obtain a total difference measure. The differences in each dimension accumulate during the summation process and ultimately reflect the overall facial feature differences. Summation is to comprehensively consider all dimensions of facial features, avoid focusing only on certain specific features, and ensure the accuracy of verification;

[0086] By taking the square root of the summation result, we obtain the final Euclidean distance. The role of the square root operation is to convert the total difference value into an intuitive distance value rather than continuing to exist in squared form. The square root helps ensure that the calculation result conforms to the concept of actual physical distance.

[0087] Specifically, for the user's facial feature vector F u and the input image feature vector F i , the Euclidean distance D(F u , F i ) is smaller, indicating that the input image matches the facial information obtained by the facial information acquisition unit more closely. Conversely, the gap is larger. According to the experimental setup, the facial threshold for D(F u , F i ) is Y1, and the value of Y1 is 0.4. Different thresholds can be set and the false recognition rate and missed recognition rate of the system can be detected. Then, the facial threshold Y1 can be set according to actual needs. When D(F u , F i ) is less than the facial threshold Y1, it indicates that the facial verification is qualified. When D(F u , F i ) is greater than the facial threshold Y1, it indicates that the facial features scanned by the current device have too large a gap from the user's facial information collected historically, and at this time the facial verification is unqualified.

[0088] Furthermore, the verification process of the voiceprint feature verification unit is as follows:

[0089]

[0090] is a threat correction term, and the threat level is non-linearly amplified through the exponential operation. When the exponential term 1 + R risk increases, the increase in the exponent results in:

[0091] If the voiceprint score is accelerated in decrease. If The voiceprint score is punished in reverse, enabling the voiceprint feature verification unit to focus on the quality of the voiceprint information itself and environmental impacts;

[0092] For the device trust compensation term, the time warping coefficient C d and the device trust score T e When multiplied, it can establish a collaborative verification mode between the device trust score T e and the time warping coefficient C d When the device trust score T e is low, even if the time warping coefficient C d is high, the overall contribution will decrease, thereby increasing security. The size of the denominator can represent the degree of influence on the voiceprint score S vioce By setting the denominator to 3, the influence of the device trust compensation term on the voiceprint feature verification unit is reduced, ensuring more focus on the quality of the voiceprint information itself;

[0093] And add the above two items for multi-modal information fusion;

[0094] Among them, S vioce represents the voiceprint score;

[0095] V base represents the basic matching rate, obtained through the voiceprint matching algorithm, with a value range of 0 to 1, where 1 represents a perfect match. The voiceprint matching algorithm is cosine similarity, and the similarity is measured by calculating the cosine value of the angle between two voiceprint feature vectors;

[0096] D t represents the facial anomaly score. When facial occlusion is detected, the value is 0.2, and when there is no facial occlusion, the value is 0;

[0097] R risk represents the environmental risk coefficient, with a value range of 0 to 1;

[0098] C d represents the time warping coefficient, with a value range of 0 to 1.2. The larger the value, the more the voices match;

[0099] T e represents the device trust score, obtained from the device matching degree and location information, with a value of 0 to 1. The initial value is set to 1, and the trust score will decrease when encountering a new device or an unfamiliar network;

[0100] Specifically, the voiceprint score S vioce indicates the degree of consistency between the current user's voice information and the voiceprint information collected during registration. The larger the voiceprint score S vioce , the higher the consistency, and vice versa. The voiceprint threshold for setting the voiceprint score S vioce is Y2. When it is less than the voiceprint score S vioceIt indicates that the voiceprint verification fails.

[0101] Time warping coefficient C d The derivation process is as follows:

[0102]

[0103] D d Indicates the DTW distance between the user's current voice and voiceprint information;

[0104] D b Indicates the DTW distance in the user's voiceprint information. Take the average value of the DTW distances when the information acquisition module collects the user's voiceprint information. For example, if the user's recording is collected 5 times, then calculate and take the average value of the DTW distances of these five times;

[0105] β represents the dynamic adjustment coefficient, β = 1 + 0.5×R risk ;

[0106] Environmental risk coefficient R risk The derivation process is as follows:

[0107]

[0108] Where ML is the network environment score;

[0109]

[0110] SL is the threat intelligence score;

[0111]

[0112] If multiple conditions are triggered for the threat intelligence score SL, take the condition with the largest result value.

[0113] Specifically, by introducing the environmental risk coefficient R risk , it can automatically adjust the strictness of identity verification according to the real-time risk level of the user's environment. The larger the environmental risk coefficient R risk , the higher the risk of the user's current network environment and the less secure the account. It enables full consideration of the impact of the user's network environment on identity verification, so as to flexibly respond to the changing virtual reality device usage environment and improve the security of the user identity verification system.

[0114] Furthermore, the analysis process of the behavior feature verification module is as follows:

[0115]

[0116] Where S bvc Represents the behavior pattern consistency score;

[0117] σ HIt represents the standard deviation of historical behavior values and indicates the degree of fluctuation of user behavior;

[0118] Z is a positive number and takes the value of 10 -5 , which is used to avoid a zero denominator;

[0119] ΔV represents the mean value of the behavior change rate;

[0120] B x represents the x-th behavior index value of the current behavior pattern, such as click frequency, head-turning angle, pace rhythm, etc.;

[0121] H x represents the x-th behavior index value in the historical behavior pattern;

[0122] t represents the time difference, which is used to standardize the behavior change, and the unit is seconds;

[0123] R risk represents the environmental risk coefficient, and its value range is from 0 to 1;

[0124] α represents the environmental risk impact weight, and its value range is from 0.5 to 1.5, which is used to adjust the impact of environmental risk on the behavior pattern;

[0125] The behavior feature verification module verifies the identity by analyzing the user's behavior features. Different from traditional static verification methods such as passwords and fingerprints, behavior pattern verification has the characteristics of dynamicity and difficulty in forgery. Even if an attacker obtains the user's account information or device, the system will reject their login request if the behavior pattern does not match.

[0126] The behavior pattern consistency score S bvc represents the degree of consistency between the current user's behavior pattern information and the behavior pattern information collected during registration. The behavior pattern consistency score S bvc The larger it is, the more similar the behaviors are. The smaller it is, the greater the behavior change. Set the behavior threshold of the behavior pattern consistency score S bvc to be Y3. When the behavior pattern consistency score S bvc > the behavior threshold Y3, it indicates that the verification is qualified. At this time, the user can log in to the system to play games. After passing the verification in the biometric verification module and then passing the verification in the behavior feature verification module, the complexity of user verification is increased, and the environmental risk coefficient is introduced during the verification processes of both the biometric verification module and the behavior feature verification module, ensuring the verification intensity and improving the accuracy of identity verification.

[0127] Furthermore, through multiple experiments and analysis of the user's behavior pattern, set the environmental threshold of the environmental risk coefficient R risk to be 0.6. When the environmental risk coefficient R riskWhen it is greater than 0.6, it indicates that the environmental risk is too high. At this time, the user during the game is verified through the iris feature verification module, and the verification process is as follows:

[0128]

[0129] Among them, D(A u , A i ) represents the difference between the user's stored iris image and the output iris image features. The smaller the difference, the more similar the iris features are;

[0130] E u (k) represents the k-th texture feature value of the user's stored iris image. The texture features include the pupil, iris ring pattern, and cornea, etc.;

[0131] E i (k) represents the k-th texture feature value of the output iris image;

[0132] W 1 represents the texture feature influence coefficient;

[0133] R u (k) represents the k-th color feature value of the user's stored iris image, such as the spots and color differences of the iris, etc.;

[0134] R i (k) represents the k-th color feature value of the output iris image;

[0135] W 2 represents the color feature influence coefficient;

[0136] y is the number of color features and texture features;

[0137] Specifically, by performing a weighted sum of the differences in the texture features and color features of the iris, the overall difference between the two iris images is obtained. The smaller the difference D(A u , A i ) between the user's stored iris image and the output iris image features, the more similar the iris image features are. According to the iris verification standard and the experimental setting, the iris threshold of D(A u , A i ) is 0.2. When D(A u , A i ) is greater than 0.2, it indicates unqualified. At this time, the user needs to re-pass the verification of the biometric feature verification module and the behavior feature verification module.

[0138] Since the virtual reality game environment is dynamic, the user's network conditions, devices, and external threat intelligence change at any time. The system determines whether to initiate more stringent identity verification based on the real-time environmental risk value. When the environmental risk is too high, performing real-time iris verification on the users in the game can improve the security of the user accounts, and iris verification is only enabled when the environmental risk is too high, which helps to balance security and performance and rationally utilize computing resources.

[0139] Further, when D(A u ,A i ) is greater than 0.2, the game connection is automatically disconnected and re-login is required. At the same time, the facial threshold Y1 is reduced, and the voiceprint threshold Y2 and behavior threshold Y3 are increased to enhance the identity verification strength;

[0140] The specific adjustment process is as follows:

[0141] The adjustment process of the voiceprint threshold Y2 is:

[0142]

[0143] Where NY1 is the new facial threshold;

[0144] NY2 is the new voiceprint threshold;

[0145] NY3 is the new behavior threshold;

[0146] The denominator of 3 represents the adjustment intensity of the facial threshold Y1. The smaller the denominator, the greater the adjustment strength, and vice versa. It can be set according to requirements in actual use.

[0147] Specifically, since the failure of the user's iris verification also indicates that the environmental risk is too high at this time, in order to prevent the leakage of game account information, it is necessary to enhance the security verification strength. Therefore, by adjusting each threshold, the system can adapt to the changes in the user's game environment, ensure the identity verification effect in a high-risk environment, and ensure the security of the account.

[0148] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A virtual reality game user identity authentication system, characterized in that: It includes account management module, information collection module, biometric verification module, behavioral feature verification module, iris feature verification module and storage module; The information collection module collects user facial information, behavior pattern information, iris information and voiceprint information, and stores them through the storage module; The biometric verification module includes a facial feature verification unit and a voiceprint feature verification unit. The facial feature verification unit compares the current user's face with the user's facial information collected during registration. The voiceprint feature verification unit is used to compare the current user's voice information with the voiceprint information collected during registration, and judge whether the voice is qualified in combination with the environmental risk situation; The behavior feature verification module is used to compare the behavior pattern information of the current user with the behavior pattern information collected during registration to determine whether it is qualified; During the game, the environmental risk situation is analyzed regularly. When the environmental risk situation is too bad, the iris information of the current user is compared with the iris information collected during registration through the iris feature verification module. If the comparison fails, the user is required to log in to the system again.

2. The virtual reality game user identity authentication system according to claim 1, characterized in that: The facial feature verification unit verification process is as follows: Where D(F u ,F i ) indicates that the user stores the facial feature vector F u With the input facial feature vector F i The Euclidean distance between F u (k) represents the kth element value of the user's facial feature vector; F i (k) represents the kth element value of the input image feature vector; n represents the number of elements; User's facial feature vector F u With the input image feature vector F i The Euclidean distance D(F u ,F i ) is smaller, the more the input image matches the facial information obtained by the facial information acquisition unit, otherwise the gap is larger. According to the experimental setting D(F u ,F i ) is Y1, when D(F u ,F i ) is less than the face threshold Y1, indicating that the face verification is qualified. u ,F i ) is greater than the facial threshold Y1, it means that the facial features scanned by the current device are too different from the user facial information collected historically, and the facial verification fails.

3. The virtual reality game user identity authentication system according to claim 2, characterized in that: The verification process of the voiceprint feature verification unit is as follows: Where S vioce Indicates the voiceprint score; V base Indicates the basic matching rate, which is obtained through the voiceprint matching algorithm and ranges from 0 to 1; D t It represents the facial abnormality score, where the value of detected facial occlusion is 0.2 and the value of unobstructed face is 0; R risk It represents the environmental risk factor, ranging from 0 to 1; C d represents the time warping coefficient; T e Indicates the device trust score, which is derived from the device matching degree and location information, and the value ranges from 0 to 1; Time warping coefficient C d The process is as follows: D d Indicates the DTW distance between the user's current voice and voiceprint information; D b Indicates the DTW distance in the user's voiceprint information; β represents the dynamic adjustment coefficient, β=1+0.5×R risk ; Environmental risk factor R risk The process is as follows: Where ML represents the network environment score; SL stands for threat intelligence score; Voiceprint score S vioce Indicates the consistency between the current user's voice information and the voiceprint information collected during registration. The voiceprint score S vioce The larger the value, the higher the consistency, and vice versa. Set the voiceprint score to S vioce The voiceprint threshold is Y2. When the voiceprint score S vioce When the value is less than the voiceprint threshold Y2, it indicates failure.

4. The virtual reality game user identity authentication system according to claim 3, characterized in that: The behavior feature verification module analysis process is as follows: Where S bvc represents the behavioral pattern consistency score; σ H Indicates the standard deviation of historical behavior values, indicating the degree of fluctuation of user behavior; Z is a positive number, and its value is 10 -5 , used to avoid the denominator being zero; ΔV represents the mean rate of change of behavior; B x The xth behavior indicator value representing the current behavior mode; H x Indicates the xth behavior indicator value in the historical behavior pattern; t represents the time difference, which is used to standardize the behavior change, in seconds; R risk It represents the environmental risk factor, ranging from 0 to 1; α represents the environmental risk impact weight, ranging from 0.5 to 1.5, which is used to adjust the impact of environmental risks on behavioral patterns; Behavior pattern consistency score S bvc It indicates the consistency between the current user's behavior pattern information and the behavior pattern information collected during registration. The behavior pattern consistency score S bvc The larger the value, the more similar the behaviors are, and the smaller the value, the greater the behavioral changes are. Set the behavioral pattern consistency score S bvc The behavioral threshold is Y3, when the behavioral pattern consistency score S bvc >When the behavior threshold is Y3, it means the verification is qualified and the user can log in to the system to play the game.

5. The virtual reality game user identity authentication system according to claim 4, characterized in that: Set the environmental risk factor R risk The environmental threshold value is 0.6, when the environmental risk factor R risk When the value is greater than 0.6, it indicates that the environmental risk is too high. At this time, the user in the game is verified through the iris feature verification module. The verification process is as follows: Where D(A u ,A i ) represents the difference between the user's stored iris image and the output iris image features; E u (k) represents the kth texture feature value of the iris image stored by the user; E i (k) represents the kth texture feature value of the output iris image; W1 represents the texture feature influence coefficient; R u (k) represents the kth color feature value of the iris image stored by the user; R i (k) represents the kth color feature value of the output iris image; W2 represents the color feature influence coefficient; y is the number of color features and texture features; The overall difference between the two iris images is obtained by weighted summing of the differences in the texture and color features of the iris. The difference D(A) between the features of the user-stored iris image and the output iris image is u ,A i ) is smaller, the more similar the iris image features are, and according to the iris verification standard and experimental setting D(A u ,A i ) has an iris threshold of 0.

2. u ,A i ) is greater than 0.2, indicating failure, in which case the user needs to re-verify the biometric verification module and the behavioral feature verification module.

6. The virtual reality game user identity authentication system according to claim 5, characterized in that: When D(A u ,A i ) is greater than 0.2, the game connection is automatically disconnected and re-login is required. At the same time, the facial threshold Y1 is lowered, and the voiceprint threshold Y2 and behavior threshold Y3 are increased to improve the identity authentication strength.

7. The virtual reality game user identity authentication system according to claim 1, characterized in that: The storage module encrypts data by combining a symmetric encryption algorithm and an asymmetric encryption algorithm when storing data, and configures a key management system for managing encryption and decryption keys to ensure the security of encrypted data.

8. The virtual reality game user identity authentication system according to claim 1, characterized in that: The account management module is used to manage user accounts, including registration and login of user accounts, and binding user accounts with virtual reality hardware IDs.

Images