Video data analysis method and device and computer equipment

Abstract

The invention discloses a video data analysis method and device and computer equipment. The method comprises the steps that an analysis request sent by a client is responded, encrypted video data are received through a pre-established trusted tunnel, the encrypted video data are obtained after the client encrypts video data to be analyzed through a first secret key, and before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data are transmitted through the trusted tunnel; performing secondary encryption on the encrypted video data by using a second key, the first key being a pre-acquired key, and the second key being a dialogue key corresponding to the trusted tunnel; decrypting the encrypted video data by using a second key to obtain the encrypted video data, and analyzing the encrypted video data by using a neural network model obtained from a cloud server to obtain an analysis result, the analysis result at least comprising: missing frames in the video data to be analyzed; and returning an analysis result to the client.

Description

Technical Field

[0001] The present application relates to the field of edge computing technology, and more specifically, to a method, device and computer equipment for analyzing video data. Background Art

[0002] With the widespread application of intelligent video analysis technology in security, transportation, retail and other fields, intelligent video surveillance systems are becoming an indispensable part of modern society. However, this is accompanied by increasing concerns about data privacy and information security. Traditional video surveillance systems have problems such as data leakage, privacy violation and model credibility during data transmission and analysis. These problems not only pose a threat to personal privacy, but may also lead to a decline in system performance and unnecessary security risks. At the same time, facing the challenges of data privacy and information security, existing technologies still have limitations in solving problems in the field of intelligent video surveillance. Traditional encryption methods may have loopholes in protecting data privacy, and identity authentication that relies solely on digital certificates may not prevent threats such as man-in-the-middle attacks, resulting in low security of video data during transmission and analysis. Summary of the invention

[0003] The embodiments of the present application provide a method, apparatus and computer device for analyzing video data to at least solve the technical problem of low security of video data during transmission and analysis in the related art.

[0004] According to one aspect of an embodiment of the present application, a method for analyzing video data is provided, comprising: responding to an analysis request sent by a client, receiving encrypted video data through a pre-established trusted tunnel, the encrypted video data being obtained after the client encrypts the video data to be analyzed using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted using a second key, the first key being a pre-acquired key, and the second key being a conversation key corresponding to the trusted tunnel; decrypting the encrypted video data using the second key to obtain the encrypted video data, and analyzing the encrypted video data using the neural network model obtained from the cloud server to obtain an analysis result, the analysis result at least including: missing frames in the video data to be analyzed; and returning the analysis result to the client.

[0005] Optionally, the video data to be analyzed is encrypted to obtain encrypted video data, including: receiving the encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed using the first key to obtain initial encrypted data, and then performing differential privacy processing on the initial encrypted data according to the privacy protection strength and the privacy leakage probability.

[0006] Optionally, decrypting the encrypted video data using the second key includes: decrypting the encrypted video data using the second key, wherein the second key includes at least: a client digital certificate, an edge computing node digital certificate, and identity information of a target object.

[0007] Optionally, the encrypted video data is analyzed using the neural network model obtained from the cloud server, including: when the neural network model exists in the edge computing node, directly using the pre-trained neural network model to analyze the encrypted video data to obtain an analysis result; when the neural network model does not exist in the edge computing node, synchronizing the encrypted video data to the cloud server for training the neural network model.

[0008] Optionally, the method also includes: when the neural network model does not exist in the edge computing node, synchronizing the encrypted video data to the cloud server, and training the neural network model using a loss function, wherein the loss function is determined by the sum of a self-supervised loss function value, a long short-term memory loss function value, and a differential privacy loss function value, the self-supervised loss function value is used to represent the difference between missing frames and actual frames in the encrypted video data, the long short-term memory loss function value is used to represent the difference between the timing of predicted frames and the timing of actual frames, and the differential privacy loss function value is used to represent the difference between the output result predicted on the initial encrypted data and the output result predicted on the encrypted video data.

[0009] Optionally, the encrypted video data is analyzed using the neural network model obtained from the cloud server, including: receiving the neural network model file, the digital signature and the device identification of the neural network model sent by the cloud server; verifying the digital signature using a public key corresponding to the digital signature of the neural network model, and when the digital signature verification passes, comparing the device identification with its own identification; and when the device identification is the same as its own identification, receiving the neural network model file to complete the model deployment.

[0010] Optionally, after returning the analysis result to the client, the method further includes: sending the encrypted video data to the cloud server and adjusting the neural network model; receiving the adjusted neural network model, and deleting the encrypted video data.

[0011] According to another aspect of an embodiment of the present application, a video data analysis device is also provided, including: a receiving module, used to respond to an analysis request sent by a client, and receive encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained after the client encrypts the video data to be analyzed using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted using a second key, the first key is a pre-acquired key, and the second key is a conversation key corresponding to the trusted tunnel; a decryption module, used to decrypt the encrypted video data using the second key to obtain the encrypted video data, and use the neural network model obtained from the server to analyze the encrypted video data to obtain an analysis result, wherein the analysis result at least includes: missing frames in the video data to be analyzed; an output module, used to return the analysis result to the client.

[0012] According to another aspect of the embodiment of the present application, a computer device is provided, including: a memory and a processor, wherein the memory is used to store program instructions; and the processor is connected to the memory and is used to execute the above-mentioned video data analysis method.

[0013] According to another aspect of the embodiments of the present application, a non-volatile storage medium is provided, which includes a stored computer program, wherein the device where the non-volatile storage medium is located executes the above-mentioned video data analysis method by running the computer program.

[0014] According to another aspect of the embodiments of the present application, a computer program product is provided, including computer instructions, which implement the above-mentioned video data analysis method when executed by a processor.

[0015] In an embodiment of the present application, by responding to an analysis request sent by a client, encrypted video data is received through a pre-established trusted tunnel, the encrypted video data being obtained after the client encrypts the video data to be analyzed using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted using a second key, the first key being a pre-acquired key, and the second key being a session key corresponding to the trusted tunnel; the encrypted video data is decrypted using the second key to obtain the encrypted video data, and the encrypted video data is analyzed using the neural network model obtained from the cloud server to obtain an analysis result, the analysis result at least including: missing frames in the video data to be analyzed; the analysis result is returned to the client, thereby achieving the purpose of re-encrypting the video data to be analyzed and analyzing the data to be analyzed based on the encrypted data, thereby achieving the technical effect of improving the security of video data during transmission and analysis, and thus solving the technical problem of low security of video data during transmission and analysis in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0017] Figure 1 is a hardware structure block diagram of a computer terminal for implementing a method for analyzing video data according to an embodiment of the present application;

[0018] Figure 2 is a flow chart of a method for analyzing video data according to an embodiment of the present application;

[0019] Figure 3 is a flow chart of another method for analyzing video data according to an embodiment of the present application;

[0020] Figure 4 is a structural diagram of a video data analysis system according to an embodiment of the present application;

[0021] Figure 5 It is a structural diagram of a video data analysis device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] The information collected in the embodiments of the present application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or reject automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.

[0025] In the related art, the weighting coefficients of various perception indicators are mainly set manually, which is too simple and has poor interpretability. The weighting coefficients are not set accurately, which may have a great impact on the user's network perception evaluation, or the basic thresholds of indicators are set based on business experience for evaluation, resulting in low accuracy of wireless network quality evaluation. In order to solve the problems existing in the related art, the embodiment of the present application provides a method for analyzing video data, which can be run on Figure 1 In the computer terminal shown, the computer terminal is explained below.

[0026] The video data analysis method embodiment provided in the embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal for implementing a method for analyzing video data. Figure 1As shown, the computer terminal 10 may include one or more (102a, 102b, ..., 102n are used to illustrate) processors (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission module 106 for communication functions connected via a wired and / or wireless network. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0027] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10. As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0028] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the video data analysis method in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned video data analysis method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0029] The transmission module 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission module 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission module 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0030] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 .

[0031] It should be noted that, in some optional embodiments, the above Figure 1 The computer terminal shown may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of hardware elements and software elements. It should be noted that Figure 1 This is merely one example of a particular embodiment and is intended to illustrate the types of components that may be present in the computer terminal described above.

[0032] In the above-mentioned operating environment, an embodiment of the present application provides an embodiment of a method for analyzing video data. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] Figure 2 is a flow chart of a method for analyzing video data according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0034] Step S202, in response to the analysis request sent by the client, receiving encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed by using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted by using a second key, wherein the first key is a pre-acquired key, and the second key is a session key corresponding to the trusted tunnel;

[0035] In step S202, the encrypted video data may be included in the analysis request, or the analysis request may be sent by the client, and after the edge computing node responds to the analysis request, the encrypted video data is sent by the client to the edge computing node.

[0036] Step S204, decrypting the encrypted video data using the second key to obtain the encrypted video data, and analyzing the encrypted video data using the neural network model obtained from the cloud server to obtain an analysis result, wherein the analysis result at least includes: missing frames in the video data to be analyzed;

[0037] Step S206: Return the analysis result to the client.

[0038] Through the above steps S202 to S206, by responding to the analysis request sent by the client, the encrypted video data is received through the pre-established trusted tunnel, the encrypted video data is obtained by the client encrypting the video data to be analyzed with the first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is encrypted twice with the second key, the first key is a pre-acquired key, and the second key is a session key corresponding to the trusted tunnel; the encrypted video data is decrypted with the second key to obtain the encrypted video data, and the encrypted video data is analyzed with the neural network model obtained from the cloud server to obtain an analysis result, the analysis result at least includes: missing frames in the video data to be analyzed; the analysis result is returned to the client, thereby achieving the purpose of performing secondary encryption on the video data to be analyzed and analyzing the data to be analyzed on the basis of the encrypted data, thereby achieving the technical effect of improving the security of the video data during transmission and analysis, and thus solving the technical problem of low security of the video data during transmission and analysis in the related technology. The following is a detailed description.

[0039] In some embodiments of the present application, the specific steps of encrypting the video data to be analyzed to obtain encrypted video data include: receiving the encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed using the first key to obtain initial encrypted data, and then performing differential privacy processing on the initial encrypted data according to the privacy protection strength and the probability of privacy leakage.

[0040] Taking the video data to be analyzed on the local client as D as an example, D is encrypted and segmented to ensure that the data is not leaked during transmission and storage. The data is encrypted using the first key, where the first key is K e , the initial encrypted data is E=Encrypt(D,K e ).

[0041] To further enhance data privacy, we further add differential privacy to E and define E ′=Differential_Privacy(E,∈,δ), where ∈ represents the privacy protection strength and δ represents the probability of privacy leakage.

[0042] In an optional approach, adding differential privacy mainly includes three main steps: Step 1, determine the privacy budget, including: privacy protection strength and privacy leakage probability are used to control the degree of privacy protection in the data analysis process;

[0043] Step 2: Select a suitable noise distribution (such as Laplace distribution or Gaussian distribution) based on the privacy budget parameter;

[0044] Step 3: Based on the privacy budget parameter, calculate the noise scale to be added. For the Laplace mechanism, the noise scale is inversely proportional to the privacy protection strength; for the Gaussian mechanism, it is related to the privacy protection strength and privacy leakage;

[0045] Step 4: Add the generated noise to the initial encrypted data to obtain encrypted video data.

[0046] In some embodiments of the present application, decrypting the encrypted video data using the second key includes: decrypting the encrypted video data using the second key, wherein the second key includes at least: a client digital certificate, an edge computing node digital certificate, and identity information of the target object.

[0047] The second key (session key) is K s For example,

[0048] K s =Multi_Factor_Authentication(Cert client ,Cert server ,Biometric,OTP)

[0049] In the formula, Cert client Indicates the client digital certificate, Cert server It represents the digital certificate of the edge computing node. Biometric is the identity information, including but not limited to fingerprint, face, voice and other information. OTP represents one-time password.

[0050] like Figure 3As shown, before using the neural network model obtained from the cloud server to analyze the encrypted video data, it is necessary to determine whether the neural network model is deployed in the edge computing node. If the neural network model exists in the edge computing node, the previously trained neural network model is directly used to analyze the encrypted video data to obtain the analysis result; if the neural network model does not exist in the edge computing node, the encrypted video data is synchronized to the cloud server for training the neural network model.

[0051] In the case that the neural network model does not exist in the edge computing node, the encrypted video data is synchronized to the cloud server, and the neural network model is trained using a loss function, wherein the loss function is determined by the sum of a self-supervised loss function value, a long short-term memory loss function value, and a differential privacy loss function value, the self-supervised loss function value is used to represent the difference between missing frames and actual frames in the encrypted video data, the long short-term memory loss function value is used to represent the difference between the timing of predicted frames and the timing of actual frames, and the differential privacy loss function value is used to represent the difference between an output result predicted on the initial encrypted data and an output result predicted on the encrypted video data.

[0052] It is understandable that encrypted video data and pre-acquired training data sets are used when training the neural network model, avoiding data leakage during the training and analysis process.

[0053] During the training process, a long short-term memory model can be used to ensure the temporal nature of the video data;

[0054] Self-supervised learning is introduced during the training process. By predicting the missing parts (missing frames) in the video, the model is trained on unsupervised data to improve data utilization efficiency. Specifically, the self-supervised loss can be calculated using the following formula: L self = Cross_Entropy(f(x t ), x t+k ), where f(·) represents the neural network model, x t represents the tth frame, x t+k Indicates the t+kth frame

[0055] Step 6 trains the model, where the overall loss function L taotal Expressed as

[0056] L total =L self +L LSTM +L Differential

[0057]

[0058] Where M(E) represents the output of the privacy mechanism or query running on data E. M(E′) represents the output of the same privacy mechanism or query running on dataset E′, and Pr[M(E)∈S] represents the probability that the query output executed on dataset D falls into a certain set S.

[0059] It can be understood that the missing frame is a missing image frame.

[0060] To ensure that the trained model (neural network model) will be digitally signed Model To ensure its integrity and credibility, for example: the edge computing node receives the neural network model file, the digital signature and the device identification of the neural network model sent by the cloud server; the digital signature is verified using the public key corresponding to the digital signature of the neural network model; if the digital signature verification passes, the device identification is compared with its own identification; if the device identification is the same as its own identification, the neural network model file is received to complete the model deployment.

[0061] Specifically, the trained model is sent to the edge computing node, where the device id is recorded as Device i , through Device i and Sign Model Verify the model; split the verified model into two parts: Model f , Model p , where Model f Feature extraction for models, Model p Used for prediction of different downstream task models, both of which perform corresponding calculations in a trusted environment; the client sends a request and sends E′ to the edge node. At the same time, the edge node temporarily stores the data in the storage module. The edge node returns the final analysis results to the client through inference calculation.

[0062] In some embodiments of the present application, after returning the analysis results to the client, the edge computing node will also send the encrypted video data to the cloud server to adjust the neural network model; receive the adjusted neural network model, and delete the encrypted video data.

[0063] For example: the edge computing node periodically synchronizes E′ with the cloud server and deletes temporary data; the cloud server re-optimizes the model based on the data and sends the optimized model back to the edge node.

[0064] The embodiment of the present application also provides another video data analysis method, including: the user's local client encrypts and performs differential privacy processing on the video data D to generate encrypted data E'; establishes a trusted channel and multi-factor identity authentication for information transmission; performs time series modeling on the video in the cloud to train the encrypted data and improve the data utilization efficiency through self-supervision, and integrates the loss function, including self-supervision, LSTM and differential privacy loss, and trains the model through methods such as gradient descent; divides the verified model into two parts, feature extraction and prediction, both of which are executed in a trusted TEE environment, and the terminal user sends a request to E ′ It is sent to the edge node for inference tasks. At the same time, the edge node stores data and synchronizes data with the cloud. For the cloud, the model is digitally signed based on the training of the encryption model to ensure its integrity and credibility.

[0065] Figure 4 is a structural diagram of a video data analysis system according to an embodiment of the present application, the system comprising:

[0066] Cloud servers, edge computing nodes, and clients.

[0067] Figure 5 is a structural diagram of a video data analysis device according to an embodiment of the present application, the device comprising:

[0068] A receiving module 50, configured to respond to an analysis request sent by a client, and receive encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted using a second key, wherein the first key is a pre-acquired key, and the second key is a session key corresponding to the trusted tunnel;

[0069] A decryption module 52 is used to decrypt the encrypted video data using the second key to obtain the encrypted video data, and analyze the encrypted video data using the neural network model obtained from the server to obtain an analysis result, wherein the analysis result at least includes: missing frames in the video data to be analyzed;

[0070] The output module 54 is used to return the analysis result to the client.

[0071] By means of the above-mentioned video data analysis device, in response to the analysis request sent by the client, the encrypted video data is received through a pre-established trusted tunnel, wherein the encrypted video data is obtained after the client encrypts the video data to be analyzed by using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted by using a second key, the first key is a pre-acquired key, and the second key is a session key corresponding to the trusted tunnel; the encrypted video data is decrypted by using the second key to obtain the encrypted video data, and the encrypted video data is analyzed by using the neural network model obtained from the cloud server to obtain an analysis result, the analysis result at least includes: missing frames in the video data to be analyzed; the analysis result is returned to the client, thereby achieving the purpose of re-encrypting the video data to be analyzed and analyzing the data to be analyzed on the basis of the encrypted data, thereby achieving the technical effect of improving the security of video data during transmission and analysis, and thus solving the technical problem of low security of video data during transmission and analysis in related technologies.

[0072] The receiving module 50 includes: a receiving submodule, which is used to receive encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed using the first key, and then performing differential privacy processing on the initial encrypted data according to the privacy protection strength and the privacy leakage probability.

[0073] The receiving submodule includes a decryption unit, which is used to decrypt the encrypted video data using the second key, including: decrypting the encrypted video data using the second key, wherein the second key at least includes: a client digital certificate, an edge computing node digital certificate and identity information of the target object.

[0074] The decryption module 52 includes: an analysis submodule, which is used to analyze the encrypted video data using the neural network model obtained from the cloud server, including: when the neural network model exists in the edge computing node, directly using the pre-trained neural network model to analyze the encrypted video data to obtain an analysis result; when the neural network model does not exist in the edge computing node, synchronizing the encrypted video data to the cloud server for training the neural network model.

[0075] The analysis submodule includes: a training unit, which is used to synchronize the encrypted video data to the cloud server when the neural network model does not exist in the edge computing node, and train the neural network model using a loss function, wherein the loss function is determined by the sum of a self-supervised loss function value, a long short-term memory loss function value, and a differential privacy loss function value, the self-supervised loss function value is used to represent the difference between the missing frames and the actual frames in the encrypted video data, the long short-term memory loss function value is used to represent the difference between the timing of the predicted frame and the timing of the actual frame, and the differential privacy loss function value is used to represent the difference between the output result predicted on the initial encrypted data and the output result predicted on the encrypted video data.

[0076] The training unit includes: a verification subunit, which is used to analyze the encrypted video data using the neural network model obtained from the cloud server, including: receiving the neural network model file, the digital signature and the device identification of the neural network model sent by the cloud server; using the public key corresponding to the digital signature of the neural network model to verify the digital signature, and when the digital signature verification passes, comparing the device identification with its own identification, and when the device identification is the same as its own identification, receiving the neural network model file to complete the model deployment.

[0077] The video data analysis device also includes: an adjustment submodule, which is used to send the encrypted video data to the cloud server and adjust the neural network model; receive the adjusted neural network model and delete the encrypted video data.

[0078] It should be noted that Figure 5 The video data analysis device shown is used to perform Figure 2 The video data analysis method shown in the figure, therefore the relevant explanations in the above-mentioned video data analysis method are also applicable to the video data analysis device, and will not be repeated here.

[0079] An embodiment of the present application also provides a computer device, including: a memory and a processor, wherein the memory is used to store program instructions; the processor is connected to the memory and is used to execute the above-mentioned video data analysis method.

[0080] An embodiment of the present application further provides a non-volatile storage medium, which includes a stored computer program, wherein a device where the non-volatile storage medium is located executes the above-mentioned video data analysis method by running the computer program.

[0081] An embodiment of the present application also provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the video data analysis method in the present application are implemented.

[0082] An embodiment of the present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the video data analysis method in the present application.

[0083] The embodiment of the present application also provides a computer program, which, when executed by a processor, implements the steps of the video data analysis method in the present application.

[0084] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0085] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0086] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0087] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0088] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0089] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0090] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for analyzing video data, characterized in that: include: In response to an analysis request sent by a client, receiving encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed by using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted by using a second key, wherein the first key is a pre-acquired key, and the second key is a session key corresponding to the trusted tunnel; Decrypting the encrypted video data using the second key to obtain the encrypted video data, and analyzing the encrypted video data using the neural network model obtained from the cloud server to obtain an analysis result, wherein the analysis result at least includes: missing frames in the video data to be analyzed; The analysis result is returned to the client.

2. The method according to claim 1, characterized in that The encrypted video data to be analyzed is encrypted to obtain the encrypted video data, including: The encrypted video data is received through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed using the first key to obtain initial encrypted data, and then performing differential privacy processing on the initial encrypted data according to the privacy protection strength and the privacy leakage probability.

3. The method according to claim 2, characterized in that Decrypting the encrypted video data using the second key comprises: The encrypted video data is decrypted using a second key, wherein the second key includes at least: a client digital certificate, an edge computing node digital certificate, and identity information of a target object.

4. The method according to claim 1, characterized in that: The encrypted video data is analyzed using a neural network model obtained from a cloud server, including: When the neural network model exists in the edge computing node, the encrypted video data is directly analyzed using the neural network model to obtain an analysis result; When the neural network model does not exist in the edge computing node, the encrypted video data is synchronized to the cloud server for training the neural network model.

5. The method according to claim 4, characterized in that The method further comprises: In the case that the neural network model does not exist in the edge computing node, the encrypted video data is synchronized to the cloud server, and the neural network model is trained using a loss function, wherein the loss function is determined by the sum of a self-supervised loss function value, a long short-term memory loss function value, and a differential privacy loss function value, the self-supervised loss function value is used to represent the difference between missing frames and actual frames in the encrypted video data, the long short-term memory loss function value is used to represent the difference between the timing of predicted frames and the timing of actual frames, and the differential privacy loss function value is used to represent the difference between an output result predicted on the initial encrypted data and an output result predicted on the encrypted video data.

6. The method according to claim 5, characterized in that The encrypted video data is analyzed using the neural network model obtained from the cloud server, including: Receiving the neural network model file, the digital signature and the device identification of the neural network model sent by the cloud server; The digital signature of the neural network model is verified using the public key corresponding to the digital signature of the neural network model. If the digital signature verification passes, the device identification is compared with its own identification. If the device identification is the same as its own identification, the file of the neural network model is received to complete the model deployment.

7. The method according to claim 1, characterized in that After returning the analysis result to the client, the method further includes: Sending the encrypted video data to the cloud server and adjusting the neural network model; The adjusted neural network model is received, and the encrypted video data is deleted.

8. A video data analysis device, characterized in that: include: A receiving module, configured to respond to an analysis request sent by a client, and receive encrypted video data through a pre-established trusted tunnel, wherein the encrypted video data is obtained by the client encrypting the video data to be analyzed by using a first key, wherein before the client transmits the encrypted video data through the trusted tunnel, the encrypted video data is re-encrypted by using a second key, wherein the first key is a pre-acquired key, and the second key is a session key corresponding to the trusted tunnel; A decryption module, configured to decrypt the encrypted video data using the second key to obtain the encrypted video data, and analyze the encrypted video data using a neural network model obtained from a server to obtain an analysis result, wherein the analysis result at least includes: missing frames in the video data to be analyzed; An output module is used to return the analysis result to the client.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory is used to store program instructions; The processor is connected to the memory, and is used to execute the video data analysis method described in any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the video data analysis method described in any one of claims 1 to 7 is implemented.

Images