Video conference network security management method

Abstract

The invention relates to the technical field of video conferences, in particular to a video conference network security management method, which comprises the following steps: sending a conference request to a video conference participant, and obtaining conference associated information; identity authentication is carried out on the video conference participating objects, and the types of the objects for interaction and conference configuration conditions are determined; performing identity authentication on the video conference participating object to obtain identity authentication information; obtaining video interaction information after identity authentication, and determining a video association authentication score according to the video interaction information and the identity authentication information; according to the video association authentication score, verifying the conference association score of the video conference participant and the conference association; performing anomaly detection analysis on the surrounding environment of the video conference participating object, and setting an anomaly association score; determining a comprehensive evaluation score of the current video conference by taking the video association authentication score, the conference association score and the abnormal association score as retrieval conditions; and the conference management efficiency and security are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of video conferencing, and in particular to a method for managing network security of a video conferencing network. Background Art

[0002] Video conferencing refers to the use of modern communication methods to achieve the purpose of holding meetings across regions. With the development of technology, different users in different places can often exchange information through video conferencing. However, in the existing video conferencing remote interaction system, participants can only attend a specific meeting, and if they want to attend different meetings, they need to reload other video conferencing systems, and the host of the video conferencing system can only understand the progress of the meeting after entering the corresponding conference system. In addition, the video conferencing system is usually implemented by authenticating the device, which cannot prevent unauthorized devices from accessing the communication network, resulting in a large number of illegal devices participating in the video conference, which not only affects the user experience, but may also cause losses to the organizer of the video conference.

[0003] For example, Chinese patent publication number CN114615461A discloses a video conferencing remote interaction system for multiple conferences, including several conference terminals, an identification module, a central control module and a cloud database. The identification module includes a facial acquisition device respectively arranged on each corresponding interactive device. The central control module is used to determine the actual authority of a participant using the interactive device according to the initial authority of the interactive device and the participant's authority of the interactive device, and unlock the corresponding interactive operation according to the actual authority of the participant during the meeting.

[0004] For example, Chinese patent publication number CN113132675A discloses a data management method and system for implementing interactive video conferencing. The present invention can determine and record conference resource configuration information of an interactive video conference scenario based on the video conference association information and the video conference participant category corresponding to the object identity information of each video conference participant obtained. The conference resource configuration information is used to characterize the interaction status and conference resource allocation between different video conference participants, and the conference resource configuration information is constantly changing.

[0005] When managing video conferences, the existing technology tends to focus on how to authenticate the video conference connection and complete the video conference management based on this information authentication; however, this management method ignores the associated problems that exist when the video conference itself is connected, as well as the problems that exist in the conference itself during the video conference, resulting in reduced security of the conference itself and reduced user experience. Summary of the invention

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a video conference network security management method, including: S1, sending a meeting request to the video conference participants, obtaining meeting-related information corresponding to the video conference participants, and the meeting-related information includes link information and identity information of the video conference participants.

[0007] S2, performing identity authentication on the video conference participant according to the identity information and link information of the video conference participant, and determining the object category and conference configuration of the video conference participant to interact in the conference.

[0008] S3, according to the conference configuration and object category, authenticate the identity of the video conference participant to obtain identity authentication information; obtain the video interaction information of the video conference participant after identity authentication, and determine the video association authentication score according to the video interaction information and the identity authentication information.

[0009] S4, according to the video association authentication score, control the resource permissions of the video conference participants, determine the speech status of the video conference participants in the corresponding conference, and verify the conference association score of the video conference participants and the conference participants.

[0010] S5, performing anomaly detection and analysis on the surrounding environment of the video conference participants, determining the abnormal conditions occurring in the video conference, and setting an abnormal correlation score.

[0011] S6, using the video association authentication score, conference association score and abnormal association score as retrieval conditions to determine the comprehensive evaluation score of the current video conference.

[0012] The beneficial effects of the present invention are: 1. The present invention combines link information and identity information to perform multi-level identity authentication to ensure that only legitimate users can join the meeting; introduces the comparison of initial device permissions to enhance the double verification of users and their devices, and improves the security of authentication.

[0013] 2. The present invention dynamically adjusts the user's resource permission level according to the video-related authentication score to ensure that each participant has only the necessary operating permissions; by analyzing the time interval, length and completion of the speaking behavior, the user's interaction is accurately evaluated, and the permissions are adjusted accordingly to ensure the order of the meeting.

[0014] 3. The present invention conducts comprehensive monitoring from multiple dimensions such as video interruption, delay and background changes to ensure that any abnormality can be captured in time; based on the threshold response rate and the minimum response authority, the most appropriate response measures are automatically selected to reduce manual intervention and improve response speed.

[0015] 4. The present invention provides a comprehensive and quantitative evaluation standard by integrating the video association authentication score, conference association score and abnormal association score into a comprehensive evaluation score, thereby improving the efficiency and quality of conference processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0017] Figure 1 The invention is a flowchart of a video conference network security management method.

[0018] Figure 2 The present invention is a flowchart of step S2 of a video conference network security management method.

[0019] Figure 3 The present invention is a flowchart of step S3 of a video conference network security management method.

[0020] Figure 4 The present invention is a flowchart of step S4 of a video conference network security management method.

[0021] Figure 5 The present invention is a flowchart of step S5 of a video conference network security management method. DETAILED DESCRIPTION

[0022] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. If no specific techniques or conditions are specified in the embodiments, the techniques or conditions described in the literature in the art or the product specifications are used.

[0023] See also Figure 1 , a video conference network security management method, including: S1, sending a conference request to a video conference participant, obtaining conference association information corresponding to the video conference participant, the conference association information including link information and identity information of the video conference participant.

[0024] S2, performing identity authentication on the video conference participant according to the identity information and link information of the video conference participant, and determining the object category and conference configuration of the video conference participant to interact in the conference.

[0025] S3, according to the conference configuration and object category, authenticate the identity of the video conference participant to obtain identity authentication information; obtain the video interaction information of the video conference participant after identity authentication, and determine the video association authentication score according to the video interaction information and the identity authentication information.

[0026] S4, according to the video association authentication score, control the resource permissions of the video conference participants, determine the speech status of the video conference participants in the corresponding conference, and verify the conference association score of the video conference participants and the conference participants.

[0027] S5, performing anomaly detection and analysis on the surrounding environment of the video conference participants, determining the abnormal conditions occurring in the video conference, and setting an abnormal correlation score.

[0028] S6, using the video association authentication score, conference association score and abnormal association score as retrieval conditions to determine the comprehensive evaluation score of the current video conference.

[0029] The identity information in step S1 indicates the current user's participation authority and the specific identity corresponding to the corresponding authority; the link information indicates the connection port and other information of the current video conference participant to join the conference.

[0030] In step S1, a new conference instance is first created. After setting the basic parameters of the conference such as time and topic, an identifier is generated for the conference, and the participants who need to attend the conference are selected. An access token is set for each video conference participant for identity verification. After passing the initial identity verification, the video conference is connected to obtain the image and voice information of the video conference participants in the video conference. At this time, the faces, pronunciations, and environment of the participants are identified to determine whether the current participants are participating in the video conference normally. At the same time, every time a new user joins, a new record will be generated to establish an associated record of the current meeting. The associated record is continuously updated to record the speeches and actions of the participants to determine whether the current video conference is conducted in the expected form. Finally, the communication form of the video conference is adjusted according to the associated information to reduce the interaction delay in the meeting and the situation where the corresponding personnel cannot be accurately identified during the interaction.

[0031] In one embodiment of the present invention, during identity authentication, identity information of video conference participants in each video conference is obtained, and the object category of the current video conference participant is determined, thereby configuring the current conference.

[0032] like Figure 2 As shown, the processing method of step S2 also includes: S21, obtaining the initial device permissions of the video conference participants, the device permissions include multiple levels, the first-level device permissions indicate that the current video conference participant has the permission to initiate a meeting and manage the progress of the video conference as a meeting organizer, the second-level device permissions indicate that the current video conference participant, as a participant of the video conference, has the permission to speak in the video conference, and the third-level device permissions indicate that the video conference participant has the permission as a video conference observer and only has the permission to watch the current video conference.

[0033] S22, comparing the initial device authority and identity information of the video conference participant, reading the user password and session key of the current video conference participant, classifying the video conference participant according to the user password and session key, and obtaining the object category.

[0034] S23, according to the similarity between the initial device permissions and the identity information, the video conference participants are configured to obtain the conference configuration. The similarity between the initial device permissions and the identity information can be calculated in the form of cosine similarity. The initial device permissions and the identity information are converted into vectors and then calculated to determine where the current video conference participants need to be configured in the video conference and how the corresponding conference should be set up.

[0035] In one embodiment of the present invention, step S3 mainly performs identity authentication on the video conference participants according to the conference configuration and the object category, and obtains the video interaction information of the video conference participants after identity authentication. The video interaction information is the corresponding video data of the user in the video conference, and this video will contain the face and environment corresponding to the participant, as well as the voice of the participant during the conference and other data.

[0036] The identity authentication information includes: determining whether the current video conference participant is the entry object of the current video conference based on the conference configuration and object category, and whether the permissions of the current video conference participant correspond to the current video conference. If they correspond, the video conference participant is used as the entry object of the current video conference, and is configured according to the permissions of the video conference participant, and the configured video conference participant is used as the identity authentication information.

[0037] The video interaction information represents the video data of the participants in the video conference collected by the camera when the current user is conducting a video conference. The video data includes the faces, backgrounds and voices of the participants in the current video conference, and these contents are calibrated to determine the specific situation of the current participants in the video conference, and whether there are any problems related to the background and the voice. Generally, in a specific video conference, the face in the current video should meet the same conditions as the preset face when compared, and the background also needs to have specific restrictions. For example, there should not be any background shifting or large changes, or any movement of people. The sound also needs to correspond to the lip shape or other conditions in the video to prevent sound interruptions, delays, etc.

[0038] like Figure 3As shown, when determining the video association authentication score based on the video interaction information and the identity authentication information, step S3 also includes: S31, using the conference configuration and object category to obtain the preset face image of the current video conference participant and the real-time face image in the video interaction information; the preset face image is a basic avatar representing the current conference participant, and the real-time face needs to embed this basic avatar to compare whether the combined face can still correspond to the current face. When the face difference is large, it means that the identity of the current participant is different from the actual entry information, and there may be a risk of information leakage.

[0039] S32, embedding the real-time face image into the preset face image, and calculating the embedded preset face image and the real-time face image to obtain a first similarity.

[0040] At this time, the method of embedding the real-time face image into the preset face image is to extract feature points from the real-time face image and the preset face image, set a hash value for these feature points, and then make one-to-one correspondence between the feature points corresponding to the real-time face image and the preset face image, and then calculate the average of the hash values ​​of the feature points as the hash value on the embedded preset face image, and finally calculate the similarity between the hash value on the embedded preset face image and the hash value of the real-time face image, and finally output the value of this similarity as the first similarity; according to the similarity value calculated at this time, it can be known whether the real-time face image and the preset face image are roughly the same. If the similarity is low after embedding, it means that the person currently participating in the video conference is not the person corresponding to this authority, and an alarm needs to be issued, and the user should be controlled.

[0041] When calculating the first similarity, the hash values ​​of the embedded preset face image and the real-time face image at the corresponding feature points are compared. If the two values ​​are the same, 1 is returned, otherwise 0 is returned. All feature points are compared, and the average value of the returned feature points is taken as the first similarity at this time.

[0042] S33, compare the real-time face images at adjacent time points, calculate the image difference ratio of the real-time face images with the scene light, combine the image difference ratios according to the length of the time points to obtain an image difference sequence, calculate the trend amount of the numerical values ​​in the image difference sequence, and output the calculated trend amount and trend direction as the second similarity.

[0043] For the second similarity, the real-time face images at adjacent time points are compared to determine whether the current video conference has instability or other problems. This not only takes into account the similarity of the image itself, but also adds an analysis of the changing trend in the time series, which can better reflect the impact of scene light changes on face recognition.

[0044] The above-mentioned image difference ratio can be expressed as follows: for each adjacent time point t, the Euclidean distance between the real-time face images at the adjacent time points on the image frame is calculated as the image difference ratio.

[0045] ;in, represents the image difference ratio at time point t, represents the image frame of the real-time face image at time point t, represents the image frame of the real-time face image at time point t+1, The Euclidean distance metric between the image frame of the real-time face image at time point t+1 and the image frame of the real-time face image at time point t is represented. The Euclidean distance metric is calculated by summing the distance values ​​between all pixels between the two image frames. It represents the norm of the image frame of the real-time face image at time point t, which is used to standardize the difference. The norm at this time is the sum of the absolute values ​​of all pixels in the image frame.

[0046] The trend volume is then expressed as: ;in, Represents the trend volume, Indicates the length value of a time point. represents the average value at a time point, Represents the average value of the image difference ratio, and the value of t ranges from 1 to T.

[0047] The trend direction is set according to the size of the trend amount. For example, if the trend amount is greater than 0, it means that the image difference increases over time, which means that the scene light may become brighter or more complex; if the trend amount is less than 0, it means that the image difference decreases over time, which means that the scene light may become darker or more stable; if the trend amount is approximately equal to 0, it means that there is no obvious trend of change and the scene light remains relatively constant.

[0048] At this time, the second similarity is calculated as follows: ;in, represents the second similarity, Indicates the adjustment coefficient, which is used to adjust the conversion speed. For example, the adjustment coefficient can be set to a value from 1 to 5. Represents an exponential constant.

[0049] S34, evaluating the current video conference participant according to the obtained first similarity and second similarity to obtain a video association authentication score.

[0050] The video association authentication score can be set to obtain the average value, standard deviation and covariance of the pixels in the real-time face image and the preset face image, and calculate the video association authentication score.

[0051] ;in, represents the video association authentication score, Represents the average value of pixels in the real-time face image. Indicates the standard deviation of pixels in a real-time face image. When calculating the standard deviation and average value of a real-time face image, the images in the image difference sequence are used to calculate to obtain a relative average value and standard deviation. Indicates the covariance of the pixels in the real-time face image and the preset face image. This covariance will make the real-time face image and the preset face image correspond one to one and then be calculated. For example, an image obtained by weighted averaging all the real-time face images in the image difference sequence and the preset face image are used to calculate the covariance value at this time. Represents the average value of pixels in the preset face image. Represents the standard deviation of pixels in the preset face image. and Represents a constant used to prevent the denominator from being zero. Indicates the first similarity. At this time, the first similarity and the second similarity serve as the adjustment part of the pixel value in the current real-time face image and the preset face image to adjust the comprehensive value of these values ​​under the corresponding similarity, so as to indicate the correlation between the current video interaction information and the originally preset information. , , Represents the weight coefficient, and the values ​​of the weight coefficient are set to 0.3, 0.3, and 0.4 respectively.

[0052] In one embodiment of the present invention, step S4 mainly controls the resource permissions of the video conference participants according to the video association authentication score, and verifies the corresponding speech in the video conference, and the score obtained by combining these situations is used as the conference association score.

[0053] The conference association score essentially obtains the operation status information under these scores, and determines whether the video conference participants of the operation status information are speaking normally, as well as the information of the corresponding audience objects, to complete the identification of the corresponding situation in the current meeting.

[0054] like Figure 4 As shown, the implementation method of step S4 also includes: S41, performing permission mapping according to the video association authentication score, determining the resource permission level of the video conference participants, at which time the resource permission level indicates the content that the user can operate in a targeted manner, defining different levels of permissions such as host, speaker, audience, etc., and clarifying the operable resources corresponding to each permission, such as speaking, screen sharing, file uploading, etc., and mapping the video association authentication score obtained at this time to these contents.

[0055] S42, determine the speech behavior of the corresponding video conference participant under the resource authority level, extract the speech time interval, speech time length and speech completion number of the speech behavior, and determine the speech completion degree of the current speech behavior. The speech completion degree is used to evaluate whether the current person has made a normal speech and whether the speech is judged as a valid behavior; at the same time, the speech also needs to be checked to see if it conforms to the preset communication mode to determine whether the current authority can correspond.

[0056] S43, determining the security policy of the speech completion degree relative to the video conference based on the acquired speech completion degree, and calculating the conference relevance score.

[0057] The speech completion degree is expressed as: ;in, Indicates the completion of the speech. represents the speaking time interval of the ith speaker, Indicates the number of completed speeches, i ranges from 1 to n; represents the length of the i-th speech, Indicates the number of valid speeches. The valid number of speeches here means that there is no network fluctuation, interruption, or voice delay in the video conference during the current speech. These normal speeches are regarded as valid speeches. Indicates the total duration of the preset speech interval, which indicates the normal time interval in a normal speech. If there is a corresponding network fluctuation or the current video conference is abnormal, this preset time will be affected, resulting in problems with the final calculated time; Indicates the adjustment coefficient, which is used to adjust the impact of the speech interval. In this case, the adjustment coefficient is set to a value range of 0 to 1; Indicates the initial value of the length of the speech time. Indicates the initial value of the speech time interval; at this time, the initial value of the speech length can be the average value of the corresponding situation in the historical data, and the speech time interval is also set using the corresponding average value in the historical data. This speech completion degree tends to describe the progress of the overall video conference under normal procedures. After completing all port or video authentication, whether the data transmission and receiving capabilities of the current video conference can meet the corresponding speech situation requirements while maintaining the corresponding security requirements, thereby improving the efficiency of the corresponding security policy.

[0058] According to the acquired speech completion degree, the security policy of the speech completion degree relative to the video conference is determined, and the conference association score is calculated.

[0059] In step S43, the implementation method of the security policy is that, when the resource authority level and the speech completion degree are matched, the audio code stream of the current video conference is parsed to determine the hash message authentication value in the audio code stream; the hash message authentication value is an authentication code set by the audio code stream for the transmitted video interaction information. At this time, it is necessary to verify whether the current process and the sent data are the corresponding hash message authentication values, so as to determine whether the video conference participants are normal participants in the meeting; according to the hash message authentication value of the audio code stream, it is verified whether the resource authority level of the video conference participants changes when the audio code stream is converted, and a security policy is selected according to the resource authority level and the speech completion degree. The method of selecting the security policy is to select the security policy corresponding to the maximum support and confidence value according to the support and confidence value of the resource authority level and the speech completion degree appearing in the historical data.

[0060] At this time, after obtaining the security policy, the conference association score will be comprehensively set according to the occurrence probability of the security policy and the common occurrence probability of the speech completion degree under the corresponding security policy to obtain the conference association score.

[0061] The conference relevance score is expressed as: the occurrence probability of the security policy, the occurrence probability of the speech completion, and the co-occurrence probability of the security policy and the speech completion are obtained, and the conference relevance score is calculated.

[0062] ;in, represents the conference association score, Indicates security policy. represents the occurrence probability of the security policy, represents the probability of speech completion, represents the co-occurrence probability of security strategy and speech completion, Indicates the standard value of the occurrence probability of speech completion; this standard value will be set according to the average value of the occurrence probability of speech completion in historical data.

[0063] In one embodiment of the present invention, step S5 performs anomaly detection and analysis on the surrounding environment of the video conference participants to determine abnormal conditions that occur in the video conference. The abnormal conditions can be set to video interruption, video delay, complete inconsistency between sound and lip shape, large differences in background changes of participants, inconsistency between the behavior of participants and permissions, etc. These situations will be regarded as abnormal conditions in the video conference. At this time, it is necessary to perform anomaly detection on the conference association, and according to the results of the anomaly detection, the degree of security management of the video conference is achieved; an abnormal association score is set.

[0064] like Figure 5As shown, step S5 also includes the following processing methods when identifying abnormal conditions: S51, obtaining the security policy of the current video conference participants, configuring the connection endpoints of the video conference participants based on the protocol communication mechanism of the security policy, transmitting abnormal detection information according to the connection endpoint, and determining the abnormal condition occurring in the video conference according to the abnormal detection status uploaded by the abnormal detection information within a preset time period.

[0065] S52, according to the appearance position of the abnormality detection information at the corresponding connection endpoint, the corresponding abnormality detection information is arranged in the order of time points to distribute the abnormality detection information.

[0066] S53, compare and analyze the anomaly detection information with the security policy, determine the threshold response rate and minimum response authority corresponding to the current video conference participant when the anomaly detection information appears, and calculate the anomaly association score according to the distribution probability of the threshold response rate and the minimum response authority. The threshold response rate indicates the probability that the system will respond immediately when a specific anomaly occurs; for example, when a video interruption is detected, the system should have an 80% probability of responding immediately.

[0067] The minimum response permission is the minimum permission level response measure that the system needs to take when an abnormality is detected; for example, when a slight video delay occurs, the system will issue a response warning message to indicate that the current network may have fluctuations, but if a user has multiple background changes in a normal video conference, the speaking permission may need to be restricted.

[0068] The method of comparing and analyzing the anomaly detection information with the security policy is to compare the specific content of the anomaly detection information with the content in the security policy, determine the specific abnormal situation that occurred at this time, and identify the minimum response authority and threshold response rate for this abnormal situation to determine whether the current abnormal situation needs to be handled.

[0069] Therefore, the calculation method of abnormal correlation score is, ;in, represents the abnormal association score, represents the distribution probability of the threshold response rate, represents the distribution probability of the minimum response authority; represents the weight coefficient of the threshold reaction rate, Indicates the weight coefficient of the minimum response authority. The weight coefficient is set to 0.5 and 0.5 in the order of threshold response rate and minimum response authority.

[0070] In one embodiment of the present invention, the comprehensive evaluation score is expressed as: ;in, represents the comprehensive evaluation score, represents the weight of the video association authentication score, represents the standard value of the abnormal association score, represents the weight of the anomaly association score, represents the standard value of the conference relevance score, Indicates the weight of the conference relevance score. The weights of the video relevance authentication score, conference relevance score, and anomaly relevance score are set to 0.3, 0.4, and 0.3 respectively; the standard value of the anomaly relevance score is expressed as the standard deviation of the anomaly relevance score in the historical data, and the standard value of the conference relevance score is expressed as the standard deviation of the conference relevance score in the historical data.

[0071] After obtaining the comprehensive evaluation score, the specific situation of the current video conference and whether there are corresponding problems in the video conference corresponding to the situation can be understood according to the comprehensive evaluation score, so as to ensure that the video conference can be carried out normally without being affected by external devices.

[0072] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention and they are still covered by the protection scope of the present invention.

Claims

1. A video conference network security management method, characterized in that: include: S1, sending a conference request to a video conference participant, obtaining conference-related information corresponding to the video conference participant, where the conference-related information includes link information and identity information of the video conference participant; S2, authenticating the video conference participants according to their identity information and link information, and determining the object category and conference configuration of the video conference participants to interact with in the conference; S3, according to the conference configuration and object category, authenticate the identity of the video conference participant to obtain identity authentication information; obtain the video interaction information of the video conference participant after identity authentication, and determine the video association authentication score according to the video interaction information and identity authentication information; S4, according to the video association authentication score, control the resource permissions of the video conference participant, determine the speech status of the video conference participant in the corresponding meeting, and verify the meeting association score of the video conference participant and the meeting association; S5, performing anomaly detection and analysis on the surrounding environment of the video conference participants, determining the abnormal conditions occurring in the video conference, and setting an abnormal correlation score; S6, using the video association authentication score, conference association score and abnormal association score as retrieval conditions to determine the comprehensive evaluation score of the current video conference.

2. A video conference network security management method according to claim 1, characterized in that: Step S2 also includes: S21, obtaining the initial device permissions of the video conference participants; S22, comparing the initial device permissions and identity information of the video conference participant, reading the user password and session key of the current video conference participant, and classifying the video conference participant according to the user password and session key to obtain the object category; S23, configuring the video conference participants according to the similarity between the initial device authority and the identity information to obtain the conference configuration status.

3. A video conference network security management method according to claim 1, characterized in that: Step S3 also includes: S31, using the conference configuration and object category, obtaining a preset face image of the current video conference participant and a real-time face image in the video interaction information; S32, embedding the real-time face image into the preset face image, and calculating the embedded preset face image and the real-time face image to obtain a first similarity; S33, comparing the real-time face images at adjacent time points, calculating the image difference ratio of the real-time face images along with the scene light, combining the image difference ratios according to the length of the time points to obtain an image difference sequence, calculating a trend amount from the values ​​in the image difference sequence, and outputting the calculated trend amount and trend direction as a second similarity; S34, evaluating the current video conference participant according to the obtained first similarity and second similarity to obtain a video association authentication score.

4. A video conference network security management method according to claim 3, characterized in that: The video association authentication score is expressed as follows: the average value, standard deviation and covariance of the pixels in the real-time face image and the preset face image are obtained, and the video association authentication score is calculated; ； in, represents the video association authentication score, Represents the average value of pixels in the real-time face image. Represents the standard deviation of pixels in real-time face images, Represents the covariance of pixels in the real-time face image and the preset face image; Represents the average value of pixels in the preset face image. Represents the standard deviation of pixels in the preset face image. and represents a constant, represents the first similarity, represents the second similarity, represents the exponential constant; , , Represents the weight coefficient.

5. A video conference network security management method according to claim 1, characterized in that: The implementation of step S4 also includes: S41, performing permission mapping according to the video association authentication score to determine the resource permission level of the video conference participant; S42, determining the speech behavior of the corresponding video conference participant under the resource authority level, extracting the speech time interval, speech time length and speech completion number of the speech behavior, and determining the speech completion degree of the current speech behavior; S43, determining the security policy of the speech completion degree relative to the video conference based on the acquired speech completion degree, and calculating the conference relevance score.

6. A video conference network security management method according to claim 5, characterized in that: The speech completion degree is expressed as: ； in, Indicates the completion of the speech. represents the speaking time interval of the i-th speaker, Indicates the number of completed speeches, i ranges from 1 to n; represents the length of the i-th speech, Indicates the number of valid speeches; Indicates the total duration of the preset speaking interval; represents the adjustment coefficient; Indicates the initial value of the length of the speech time. Indicates the initial value of the speaking time interval.

7. A video conference network security management method according to claim 6, characterized in that: The conference relevance score is expressed as: the occurrence probability of the security policy, the occurrence probability of the speech completion, and the co-occurrence probability of the security policy and the speech completion are obtained, and the conference relevance score is calculated; ； in, represents the conference association score, Indicates security policy. represents the occurrence probability of the security policy, represents the probability of speech completion, represents the co-occurrence probability of security strategy and speech completion, This value indicates the standard value of the occurrence probability of speech completion.

8. A video conference network security management method according to claim 1, characterized in that: Step S5 also includes the following processing methods: S51, obtaining the security policy of the current video conference participant, configuring the connection endpoint of the video conference participant based on the protocol communication mechanism of the security policy, transmitting anomaly detection information according to the connection endpoint, and determining the abnormality that occurs in the video conference according to the abnormality detection status uploaded by the anomaly detection information within a preset time period; S52, according to the appearance position of the abnormality detection information at the corresponding connection endpoint, setting the distribution position of the corresponding abnormality detection information in the order of time points; S53, comparing and analyzing the anomaly detection information with the security policy, determining the threshold reaction rate and the minimum reaction authority corresponding to the current video conference participant when the anomaly detection information appears, and calculating the anomaly association score according to the distribution probability of the threshold reaction rate and the minimum reaction authority.

9. A video conference network security management method according to claim 8, characterized in that: The calculation method of abnormal correlation score is: ； in, represents the abnormal association score, represents the distribution probability of the threshold response rate, represents the distribution probability of the minimum response authority; represents the weight coefficient of the threshold reaction rate, The weight coefficient representing the minimum response authority.

10. A video conference network security management method according to claim 1, characterized in that: The comprehensive evaluation score is expressed as: ； in, represents the comprehensive evaluation score, represents the weight of the video association authentication score, represents the standard value of the abnormal association score, represents the weight of the anomaly association score, represents the standard value of the conference relevance score, represents the weight of the conference relevance score, represents the abnormal association score, represents the conference association score, Represents the video association authentication score.

Images