A method for managing security of confidentiality operation videos in highly confidential scenarios

Abstract

The present invention relates to a method for secure management of confidentiality operation videos in high-confidentiality scenarios, and relates to the field of security technology. In the confidentiality operation in the high-confidentiality scenario, the present application identifies and obtains the identity information of the person performing the confidentiality operation; in the high-confidentiality scenario, the video records the confidentiality operation performed by the person; processes the video recording the confidentiality operation performed by a pre-trained information steganography model, steganates the confidentiality operation information containing the identity information of the person into the video frame of the video, obtains a traceable video frame containing the confidentiality operation information, replaces the original video frame with the traceable video frame to obtain the traceable video; obtains the identity information of the video processing person and his video processing operation, and steganates them into the video frame of the processed traceable video by the information steganography model. The confidentiality operation information is decoded from the traceable video frame by the pre-trained information steganography model to realize secure management of confidentiality operation videos in high-confidentiality scenarios.

Description

Technical Field

[0001] The present invention relates to the field of video processing technology, and in particular to a method for securely managing videos of confidential operations in highly confidential scenarios. Background Art

[0002] Confidential operations occurring in high-security scenarios require strict monitoring. Currently, surveillance videos of confidential operations in high-security scenarios do not contain any information about the confidential operator, nor do they contain any information about the user who processed the surveillance video. Leaked videos cannot determine the user responsible for the leak, and the source of the leak cannot be traced after the surveillance video is leaked. Existing surveillance videos in high-security scenarios are traced using watermarks. The user ID and operation time are randomly attached to the execution locations of certain video frames using watermarks. When auditors download the videos, the auditor's user ID is randomly attached to specific locations of certain video frames using the video transcoding function. This method of adding user information to certain frames of the confidential operation surveillance video achieves the purpose of video leak tracing. However, the watermark is visible and easily removed by technical personnel, ultimately making it impossible to trace the source.

[0003] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the Invention

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the present invention provides a method for secure management of confidential operation videos in high-confidentiality scenarios.

[0005] The present invention provides a method for managing security of confidentiality operation videos in highly confidential scenarios, comprising:

[0006] In high-confidentiality scenarios, identify the identity of the person performing confidentiality operations and obtain their identity information;

[0007] Use surveillance video to record confidential operations performed by confidentiality operators;

[0008] Through the pre-trained information steganography model, the confidential operation information containing the identity information of the confidential operator is steganographically written into the video frame of the video, and the traceable video frame containing the confidential operation information is obtained. The traceable video frame is replaced with the original video frame to obtain the traceable video.

[0009] For traceable videos, before processing the traceable videos, the traceable video manager obtains the identity information of the video processor and their video processing operations, and incrementally writes them into the video frames of the traceable video being processed through the information steganography model;

[0010] Through the pre-trained information steganography model, confidential operation information is decoded from the traceable video frame to achieve high confidentiality scene confidential operation video security management;

[0011] The information steganography model includes: an information graphic code preprocessing module based on a reversible neural network and an information graphic code steganography module; the forward process of the information graphic code preprocessing module is used to encode the spliced ​​information graphic code, downsample the spliced ​​information graphic code into a reduced form that supports hiding in a video frame and obtain a corresponding first key map; the reverse process of the information graphic code preprocessing module is used to upsample the reduced form of the spliced ​​information graphic code decoded by the information graphic code steganography module in combination with the first key map to restore the decoded spliced ​​information graphic code; the reduced form of the spliced ​​information graphic code and the video frame to be processed are input into the information graphic code steganography module, the information graphic code steganography module performs discrete wavelet transform on the reduced form and video frame to the frequency domain, and then the frequency domain data is input into the reversible neural network of the information graphic code steganography module to steganographically obtain a second key map and frequency domain data of the video frame with the information graphic code, and then undergo inverse frequency domain transform to obtain the second key map and the traceable video frame.

[0012] Furthermore, multiple sampled video frames are extracted from the entire time span of the video recording the confidentiality operation through segmented sampling. The information steganography model performs confidentiality operation information steganography on each sampled video frame to obtain a sampled traceable video frame, and the sampled traceable video frame replaces the original corresponding sampled video frame to obtain a traceable video.

[0013] Furthermore, a QR code generator is used to convert the confidential operation information into a corresponding information graphic code, and at least one information graphic code written into the video frame is spliced ​​together according to a set format to obtain a spliced ​​information graphic code.

[0014] Furthermore, when the number of information graphic codes written into the video frame is less than the number of information graphic codes set in the format, the copy result of the information graphic codes written into the video frame is used to fill the number to meet the format setting.

[0015] Furthermore, the forward process of the reversible neural network of the information graphic code preprocessing module is as follows:

[0016] Feature map of the concatenated information graphic code input to any k-th reversible neural module Divided into sub-feature maps along the channel dimension and sub-feature maps , the sub-feature map The transformed and sub-feature maps are calculated through the first residual dense convolutional network Additive combination, the result of the additive combination is used as the input feature map of the subsequent k+1th reversible neural module that constitutes the information graphic code preprocessing module Sub-feature map of ; Sub-feature map In the kth reversible neural module, the scaling rotation factor is learned by the second residual dense convolution network and the scalable Sigmoid activation function, and the translation factor is learned by the third residual dense convolution network. After multiplying with the scaling rotation factor and then adding and combining with the translation factor, the result of the addition and combination is used as the feature map of the subsequent k+1th reversible neural module that constitutes the information graphic code preprocessing module Sub-feature map of .

[0017] Furthermore, the reverse process of the reversible neural network of the information graphic code preprocessing module is as follows:

[0018] Feature map of the information graphic code of the concatenation of any k+1th reversible neural module input to the information graphic code preprocessing module Divided into sub-feature maps along the channel dimension and sub-feature maps , the sub-feature map The scaling rotation factor is learned by the second residual dense convolution network and the scalable Sigmoid activation function, and the translation factor is learned by the third residual dense convolution network. Subtract the translation factor and divide it by the scaling rotation factor to obtain the sub-feature map of the k-th reversible neural module of the information graphic code preprocessing module. ; Sub-feature map Subtract sub-feature map The transformation result calculated by the first residual dense convolutional network is used as the sub-feature map of the k-th reversible neural module of the information graphic code preprocessing module .

[0019] Furthermore, the forward process of the reversible neural network of the information graphic code steganography module is as follows:

[0020] The reduced form of the information graphic code input to the first reversible neural module of the information graphic code steganography module and video frames The reduced form of the information graphic code is converted to the frequency domain through frequency domain transformation and then input into any k-th reversible neural module of the information graphic code steganography module. Feature map The fourth residual dense convolutional network calculates the transformed video frame Feature map Additive combination, the result of the additive combination is used as the feature map of the video frame of the subsequent k+1th reversible neural module that constitutes the information graphic code steganography module ;

[0021] Feature map of video frame In the kth reversible neural module of the information graphic code steganography module, the non-scaling rotation factor is learned by the fifth residual dense convolutional network and the scalable Sigmoid activation function, and the translation factor is learned by the sixth residual dense convolutional network. The feature map of the spliced ​​information graphic code The feature map of the information graphic code is obtained by multiplying the non-scaling rotation factor and then adding the translation factor as the concatenation of the subsequent k+1 reversible neural modules that constitute the information graphic code steganography module. .

[0022] Furthermore, the reverse process in the reversible neural network of the information graphic code steganography module is as follows:

[0023] The second key image and the traceable video frame input to the information graphic code steganography module are converted to the frequency domain through frequency domain transformation. The final feature map of the traceable video frame in the frequency domain received by the K-th reversible neural module is The characteristic map of the second key map in the sum frequency domain ;

[0024] Then input the feature map of the traceable video frame in the frequency domain of any k+1th reversible neural module of the information graphic code steganography module The translation factor is calculated by the sixth residual dense convolution network, and the non-scaled rotation factor is calculated by the fifth residual dense convolution network and the scalable Sigmoid activation function; the feature map of the spliced ​​information graphic code in the frequency domain Subtract the translation factor and divide it by the non-scaling rotation factor, and the result is used as the feature map of the information graphic code spliced ​​by the kth reversible neural module of the information graphic code steganalysis module :

[0025] Feature Map Subtract feature map The feature map of the information graphic code after the transformation calculated by the fourth residual dense convolutional network as the splicing of the k-th reversible neural module of the information graphic code steganalysis module .

[0026] Furthermore, the implementation of the scalable Sigmoid activation function is as follows: the Sigmoid activation function compresses the output of the residual dense convolutional network to the range of (0, 1), and then translates it to the range of (-0.5, 0.5) by subtracting 0.5, so that the final scaled or unscaled rotation factor is symmetric around 0, and then the output value is scaled by the scaling or unscaling coefficient and the exponential function of the natural exponent e is taken to obtain the scaled or unscaled rotation factor.

[0027] Furthermore, during training, the learning objectives of the information steganography model include minimizing the following loss terms:

[0028] The difference between the video frame to be processed and the traceable video frame;

[0029] The difference between the frequency domain of the video frame to be processed and the traceable video frame;

[0030] The difference between the reduced form of the spliced ​​information graphic code encoded by the forward process of the information graphic code steganography module and the predicted reduced form of the spliced ​​information graphic code restored by combining the traceable video frame with the second key graph by the reverse process of the information graphic code steganography module;

[0031] The difference between the spliced ​​information graphic code and the original information graphic code decoded by the reverse process of the information graphic code steganography module.

[0032] The above technical solution provided by the embodiment of the present invention has the following advantages compared with the prior art:

[0033] In the information steganography model of the present application, the forward process of the information graphic code preprocessing module is used to encode the spliced ​​information graphic code, downsample the spliced ​​information graphic code into a reduced form that supports hiding in the video frame and obtain the corresponding first key graph; the reverse process of the information graphic code preprocessing module is used to upsample the reduced form of the spliced ​​information graphic code decoded by the information graphic code steganography module in combination with the first key graph to restore the decoded spliced ​​information graphic code; the reduced form of the spliced ​​information graphic code and the video frame to be processed are input into the information graphic code steganography module, and the information graphic code steganography module performs a discrete wavelet transform on the reduced form and video frame to the frequency domain. The frequency domain data is then input into the reversible neural network of the information graphic code steganography module to stegize and obtain the second key graph and the frequency domain data of the video frame with the information graphic code, and then undergoes an inverse frequency domain transform to obtain the second key graph and the traceable video frame. Through the forward and reverse processes of the information steganography model, the confidential operation information is effectively recorded in the video frame of the confidential operation video of the high confidentiality scene, and the confidential operation video of the high confidentiality scene can be implicitly traced. The steganography and decoding process of the confidential operation information depends on the information steganography model, the first key map and the second key map, which is not easy to crack. This application steganographically writes the original confidentiality operation and subsequent video processing operations into the traceable video, making it easier to trace the source of the leak. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0036] Figure 1 The present invention provides a flowchart of a method for managing video security in highly confidential scenarios and confidential operations according to an embodiment of the present invention.

[0037] Figure 2 A schematic diagram of converting confidential operation information provided by the disclosed embodiment of the present invention into an information graphic code.

[0038] Figure 3 A schematic diagram of the overall architecture of the information steganography model provided in the disclosed embodiment of the present invention.

[0039] Figure 4 A schematic diagram of the forward process of the information graphic code preprocessing module provided in the disclosed embodiment of the present invention.

[0040] Figure 5 A schematic diagram of the reverse process of the information graphic code preprocessing module provided in the disclosed embodiment of the present invention.

[0041] Figure 6 A schematic diagram of the forward process of the information graphic code steganography module provided in the disclosed embodiment of the present invention.

[0042] Figure 7 A schematic diagram of the reverse process of the information graphic code steganography module provided in the disclosed embodiment of the present invention.

[0043] Figure 8 Schematic diagram of a residual dense convolutional network provided in an embodiment disclosed in the present invention.

[0044] Figure 9 A schematic diagram of a video security management device for confidentiality operation in a high-confidentiality scenario provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0046] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0047] Example 1

[0048] See Figure 1 As shown, an embodiment of the present invention provides a method for securely managing confidential operation videos in a highly confidential scenario, including:

[0049] S100: In a high-confidentiality scenario, identify the identity of a person performing a confidentiality operation and obtain identity information of the confidentiality operator. The identity information of the confidentiality operator includes the name and work number of the confidentiality operator.

[0050] S200: Record the confidentiality operation performed by the confidentiality operator using the surveillance video.

[0051] S300, processes the video recording the confidential operation through a pre-trained information steganography model, steganates the confidential operation information containing the identity information of the confidential operator into the video frame of the video, obtains a traceable video frame containing the confidential operation information, and replaces the original video frame with the traceable video frame to obtain a traceable video.

[0052] S400, this application supports steganography of secondary processing operations of traceable videos. For traceable videos, before the traceable video manager processes the traceable video, he obtains the identity information of the video processing personnel and his video processing operations, and incrementally steganizes them into the video frame of the traceable video being processed through the information steganography model. In the specific implementation process, before the traceable video manager with authority processes the traceable video, the manager's identity information and operation information are obtained, and the manager's identity information and operation information are incrementally steganographically updated to the video frame of the current traceable video through the pre-trained information steganography model. The traceable video after the updated information replaces the current traceable video and is provided to the manager who initiated the operation. The incremental steganographic update is achieved by adding a newly added information graphic code to the spliced ​​information graphic code.

[0053] In one embodiment, in order to reduce the video frame processing volume of the information steganography model, the video recording the confidentiality operation is sampled in segments, and multiple sampled video frames for steganography of confidentiality operation information are extracted from the entire time span of the video recording the confidentiality operation through segmented sampling. The information steganography model steganizes the confidentiality operation information on each sampled video frame to obtain a sampled traceable video frame, and the sampled traceable video frame after steganography replaces the original corresponding sampled video frame to obtain a traceable video recording the confidentiality operation video.

[0054] In the case of segmented sampling, the sampled traceable video frame is obtained according to the sampled video frame index, and the information steganography model decodes the confidential operation information of each sampled traceable video frame.

[0055] The information steganography model realizes the steganography and decoding of confidential operation information. The working principle of the information steganography model is as follows:

[0056] This application first uses a QR code generator to convert confidential operation information into corresponding information graphic code. Figure 2 As shown, the confidentiality operation information includes the identity information of the confidentiality operator, the confidentiality operation time and the planned confidentiality operation. The confidentiality operation information including the identity information of the confidentiality operator is generated into a corresponding information graphic code in the form of a QR code by a QR code generator. At least one information graphic code written into the video frame is spliced ​​together according to the set format to obtain a spliced ​​information graphic code. For the convenience of subsequent description, the spliced ​​information graphic code is defined as If the number of information graphics codes written into the video frame is less than the number of information graphics codes set in the format, the information graphics codes written into the video frame will be copied to fill the gap to meet the format setting. For example, if the format setting is that 5×5 information graphics codes form a spliced ​​information graphics code, and the number of information graphics codes written into the video frame is 5, the 5 information graphics codes written into the video frame will be copied 5 times to fill the spliced ​​information graphics code.

[0057] This application constructs and trains the information steganography model. Figure 3 As shown, the information steganography model includes: an information graphic code preprocessing module based on a reversible neural network and an information graphic code steganography module.

[0058] In the specific implementation process, Figure 4 and Figure 5 As shown, the information graphic code preprocessing module is implemented by a reversible neural network formed by stacking reversible neural modules, and the information graphic code preprocessing module includes forward and reverse processes.

[0059] The forward process of the information graphic code preprocessing module is used to encode the spliced ​​information graphic code and downsample the spliced ​​information graphic code into a reduced form that supports hiding in the video frame. The reverse process of the information graphic code preprocessing module is used to combine the first key graph to decode the information graphic code steganography module to obtain the reduced form of the spliced ​​information graphic code. Upsampling is performed to restore the decoded spliced ​​information graphic code.

[0060] The forward process of the reversible neural network of the information graphic code preprocessing module is as follows: Figure 4 As shown, the feature map of the information graphic code input to any k-th reversible neural module is Divided into sub-feature maps along the channel dimension and sub-feature maps , the sub-feature map The transformed and sub-feature maps are calculated through the first residual dense convolutional network Additive combination, the result of the additive combination is used as the feature map of the subsequent k+1th reversible neural module that constitutes the information graphic code preprocessing module Sub-feature map of : ,in, is the first residual dense convolutional network in the k-th reversible neural module. Sub-feature map In the kth reversible neural module, the scaling rotation factor is learned by the second residual dense convolution network and the scalable Sigmoid activation function, and the translation factor is learned by the third residual dense convolution network. After multiplying with the scaling rotation factor and then adding and combining with the translation factor, the result of the addition and combination is used as the feature map of the subsequent k+1th reversible neural module that constitutes the information graphic code preprocessing module Sub-feature map of :

[0061] ;

[0062] in, is the second residual dense convolutional network in the k-th reversible neural module of the information graphic code preprocessing module, The third residual dense convolutional network in the kth reversible neural module of the information graphic code preprocessing module. The scalable Sigmoid activation function is implemented as follows: the output of the second residual dense convolutional network is compressed to the range of (0, 1) by the Sigmoid activation function, and then translated to the range of (-0.5, 0.5) by subtracting 0.5, so that the final scaling rotation factor is symmetrical around 0. Then, the scaling factor is used to calculate the rotation factor. After scaling the output value, an exponential function of the natural exponent e is taken to obtain a scaling rotation factor, where the scaling rotation factor is positive and has a controlled range.

[0063] In the forward process of the reversible neural network of the information graphic code preprocessing module, the first reversible neural module in the information graphic code preprocessing module receives the spliced ​​information graphic code, and the final K-th reversible neural module in the information graphic code preprocessing module outputs the sub-feature map As a reduced form of the final spliced ​​information graphic code , the sub-feature map output by the final K-th reversible neural module in the information graphic code preprocessing module As a corresponding reduced form The first key diagram.

[0064] The reverse process in the reversible neural network of the information graphic code preprocessing module is as follows Figure 5 As shown, the feature map of the information graphic code of the concatenation of any k+1th reversible neural module input to the information graphic code preprocessing module is Divided into sub-feature maps along the channel dimension and sub-feature maps , the sub-feature map The scaling rotation factor is learned by the second residual dense convolution network and the scalable Sigmoid activation function, and the translation factor is learned by the third residual dense convolution network. Subtract the translation factor and divide it by the scaling rotation factor to obtain the sub-feature map of the k-th reversible neural module of the information graphic code preprocessing module. :

[0065] ;

[0066] Sub-feature map Subtract sub-feature map The transformation result calculated by the first residual dense convolutional network is used as the sub-feature map of the k-th reversible neural module of the information graphic code preprocessing module :

[0067] .

[0068] In the reverse process of the reversible neural network of the information graphic code preprocessing module, the final K-th reversible neural module in the information graphic code preprocessing module receives the reduced form of the spliced ​​information graphic code decoded by the information graphic code steganography module. , which is used as the sub-feature map of the feature map of the K-th reversible neural module of the information graphic code preprocessing module , the first key graph is used as a sub-feature graph of the feature graph of the K-th reversible neural module of the information graphic code preprocessing module The first reversible neural module in the information graphic code preprocessing module outputs the decoded spliced ​​information graphic code.

[0069] A reduced form of the spliced ​​information graphic code converted by the information graphic code preprocessing module and the video frames to be processed The information graphic code is input to the information graphic code steganography module, and the information graphic code steganography module first converts the spliced ​​information graphic code into a reduced form. and video frames Discrete wavelet transformation is performed to the frequency domain, and then the frequency domain data is input into the reversible neural network of the information graphic code steganography module to obtain the frequency domain data of the second key image and the video frame with the information graphic code. Then, the second key image and the traceable video frame are obtained by inverse frequency domain transformation respectively.

[0070] like Figure 8 As shown in the figure, the structure of the residual dense convolutional network involved in this application is consistent. It is implemented by convolution blocks, each of which contains a convolution layer, a LeakyReLU activation function, and a concatenation operation of the residual and dense concatenation of the convolution block output.

[0071] In the specific implementation process, the information graphic code steganography module is realized by a reversible neural network formed by stacking reversible neural modules. The information graphic code steganography module includes forward and reverse processes. The forward process of the information graphic code steganography module is used to encode the reduced form of the spliced ​​information graphic code. Steganography into video frames The reverse process of the information graphic code steganography module is used to restore the traceable video frame to the predicted spliced ​​information graphic code in a reduced form by combining the second key graph. and the corresponding video frames .

[0072] The forward process of the reversible neural network of the information graphic code steganography module is as follows Figure 6 As shown, the reduced form of the information graphic code of the first reversible neural module input to the information graphic code steganography module and video frames The reduced form of the information graphic code is converted to the frequency domain through frequency domain transformation and then input into any k-th reversible neural module of the information graphic code steganography module. Feature map The fourth residual dense convolutional network calculates the transformed video frame Feature map Additive combination of feature maps of video frames as the subsequent k+1th reversible neural module that constitutes the information graphic code steganography module : ,in, It is the fourth residual dense convolutional network in the kth reversible neural module of the information graphic code steganography module. Feature map of the video frame In the kth reversible neural module of the information graphic code steganography module, the non-scaling rotation factor is learned by the fifth residual dense convolutional network and the scalable Sigmoid activation function, and the translation factor is learned by the sixth residual dense convolutional network. The feature map of the information graphic code is obtained by multiplying the non-scaling rotation factor and then adding the translation factor as the concatenation of the subsequent k+1 reversible neural modules that constitute the information graphic code steganography module. :

[0073] ;

[0074] in, is the fifth residual dense convolutional network in the k-th reversible neural module of the information graphic code steganalysis module, The sixth residual dense convolutional network in the kth reversible neural module of the information graphic code steganography module is implemented as follows: the output of the fifth residual dense convolutional network is compressed to the range of (0, 1) by the Sigmoid activation function, and then translated to the range of (-0.5, 0.5) by subtracting 0.5, so that the final rotation factor is symmetrical around 0. Then, the non-scaling coefficient is used. After scaling the output value, the exponential function of the natural exponent e is taken to obtain the rotation factor.

[0075] In the forward process of the information graphic code steganography module, the first reversible neural module in the information graphic code steganography module receives the reduced form of the spliced ​​information graphic code. Feature map and video frames Feature map , the feature map output by the final K-th reversible neural module in the information graphic code steganography module As the frequency domain form of the final second key graph, the feature map output by the final K-th reversible neural module in the information graphic code steganography module As the frequency domain form of the traceable video frame, the second key map and the frequency domain form of the traceable video frame are converted into the spatiotemporal domain through an inverse frequency domain transform.

[0076] The reverse process in the reversible neural network of the information graphic code steganography module is as follows Figure 7 As shown, the second key image and the traceable video frame input to the information graphic code steganography module are converted to the frequency domain through frequency domain transformation processing, and the final feature map of the traceable video frame in the frequency domain received by the K-th reversible neural module is The characteristic map of the second key map in the sum frequency domain .

[0077] Then input the feature map of the traceable video frame in the frequency domain of any k+1th reversible neural module of the information graphic code steganography module The translation factor is calculated by the sixth residual dense convolution network, and the non-scaled rotation factor is calculated by the fifth residual dense convolution network and the scalable Sigmoid activation function; the feature map of the spliced ​​information graphic code in the frequency domain Subtract the translation factor and divide it by the non-scaling rotation factor, and the result is used as the feature map of the information graphic code spliced ​​by the kth reversible neural module of the information graphic code steganalysis module :

[0078] .

[0079] Feature Map Subtract feature map The feature map of the kth reversible neural module as the information graphic code steganography module after transformation calculated by the fourth residual dense convolutional network :

[0080] .

[0081] In order to make the information steganography model have the above effect, the information steganography model is trained as follows:

[0082] During training, the learning objectives of the information steganography model include minimizing the following loss terms:

[0083] The difference between the video frame to be processed and the traceable video frame;

[0084] The difference between the frequency domain of the video frame to be processed and the traceable video frame;

[0085] The reduced form of the spliced ​​information graphic code encoded by the forward process of the information graphic code steganography module The reverse process of the information graphic code steganography module is combined with the second key map to restore the traceable video frame to the predicted spliced ​​information graphic code in a reduced form. the differences between;

[0086] The difference between the spliced ​​information graphic code and the original information graphic code decoded by the reverse process of the information graphic code steganography module.

[0087] S500: Processing traceable video frames in a traceable video using the pre-trained information steganography model, and decoding confidential operation information from the traceable video frames.

[0088] Example 2

[0089] See Figure 9 As shown, the present invention provides a highly confidential scene confidentiality operation video security management device, comprising: at least one processing unit, the processing unit being connected to a storage unit via a bus unit, the storage unit storing a computer program, and the computer program, when executed by the processing unit, implementing the highly confidential scene confidentiality operation video security management method, comprising:

[0090] During confidentiality operations in highly confidential scenarios, identify the identity of the person performing the confidentiality operation and obtain the identity information of the confidentiality operator;

[0091] During confidentiality operations in highly confidential scenarios, the video records the confidentiality operations of the confidentiality operators;

[0092] Through the pre-trained information steganography model, the confidential operation information containing the identified confidential operator identity information is steganographically written into the video frame of the video, and the traceable video frame containing the confidential operation information is obtained. The traceable video frame is replaced with the original video frame to obtain the traceable video.

[0093] Through the pre-trained information steganography model, confidential operation information is decoded from the traceable video frame to achieve high confidentiality scene confidential operation video security management;

[0094] The information steganography model includes: an information graphic code preprocessing module based on a reversible neural network and an information graphic code steganography module; the forward process of the information graphic code preprocessing module is used to encode the spliced ​​information graphic code, downsample the spliced ​​information graphic code into a reduced form that supports hiding in a video frame and obtain a corresponding first key map; the reverse process of the information graphic code preprocessing module is used to upsample the reduced form of the spliced ​​information graphic code decoded by the information graphic code steganography module in combination with the first key map to restore the decoded spliced ​​information graphic code; the reduced form of the spliced ​​information graphic code and the video frame to be processed are input into the information graphic code steganography module, the information graphic code steganography module performs discrete wavelet transform on the reduced form and video frame to the frequency domain, and then the frequency domain data is input into the reversible neural network of the information graphic code steganography module to steganographically obtain a second key map and frequency domain data of the video frame with the information graphic code, and then undergo inverse frequency domain transform to obtain the second key map and the traceable video frame.

[0095] Of course, the computer program stored in the storage unit of the high-confidentiality scene confidentiality operation video security management device provided by an embodiment of the present invention is not limited to the method operations described above, and can also execute related operations in the high-confidentiality scene confidentiality operation video security management method provided by any embodiment of the present invention.

[0096] Example 3

[0097] An embodiment of the present invention provides a computer-readable storage medium storing computer instructions. When the computer instructions are executed by a processor, the method for managing video security in highly confidential scenarios and confidential operations is implemented, including:

[0098] During confidentiality operations in highly confidential scenarios, identify the identity of the person performing the confidentiality operation and obtain the identity information of the confidentiality operator;

[0099] During confidentiality operations in highly confidential scenarios, the video records the confidentiality operations of the confidentiality operators;

[0100] Through the pre-trained information steganography model, the confidential operation information containing the identified confidential operator identity information is steganographically written into the video frame of the video, and the traceable video frame containing the confidential operation information is obtained. The traceable video frame is replaced with the original video frame to obtain the traceable video.

[0101] Through the pre-trained information steganography model, confidential operation information is decoded from the traceable video frame to achieve high confidentiality scene confidential operation video security management;

[0102] The information steganography model includes: an information graphic code preprocessing module based on a reversible neural network and an information graphic code steganography module; the forward process of the information graphic code preprocessing module is used to encode the spliced ​​information graphic code, downsample the spliced ​​information graphic code into a reduced form that supports hiding in a video frame and obtain a corresponding first key map; the reverse process of the information graphic code preprocessing module is used to upsample the reduced form of the spliced ​​information graphic code decoded by the information graphic code steganography module in combination with the first key map to restore the decoded spliced ​​information graphic code; the reduced form of the spliced ​​information graphic code and the video frame to be processed are input into the information graphic code steganography module, the information graphic code steganography module performs discrete wavelet transform on the reduced form and video frame to the frequency domain, and then the frequency domain data is input into the reversible neural network of the information graphic code steganography module to steganographically obtain a second key map and frequency domain data of the video frame with the information graphic code, and then undergo inverse frequency domain transform to obtain the second key map and the traceable video frame.

[0103] Of course, the computer-readable storage medium provided by an embodiment of the present invention stores a computer program which is not limited to the method operations described above, and can also execute related operations in a high-confidentiality scene confidentiality operation video security management method provided by any embodiment of the present invention.

[0104] In the embodiments provided by the present invention, it should be understood that the disclosed structures and methods can be implemented in other ways. For example, the structural embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, structure or unit, which can be electrical, mechanical or other forms.

[0105] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of the solution of this embodiment according to actual needs.

[0106] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0107] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is intended to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for managing security of confidentiality operation videos in high-confidentiality scenarios, characterized in that: include: In high-confidentiality scenarios, identify the identity of the person performing confidentiality operations and obtain their identity information; Use surveillance video to record confidential operations performed by confidentiality operators; Using a pre-trained information steganography model, confidential operation information containing the identity information of the confidential operator is steganographically written into the video frame of the video, a traceable video frame containing the confidential operation information is obtained, and the traceable video frame is replaced with the original video frame to obtain a traceable video; For traceable videos, before processing the traceable videos, the traceable video manager obtains the identity information of the video processor and their video processing operations, and incrementally writes them into the video frames of the traceable video being processed through the information steganography model; Through the pre-trained information steganography model, confidential operation information is decoded from the traceable video frame to achieve high confidentiality scene confidential operation video security management; The information steganography model includes: an information graphic code preprocessing module based on a reversible neural network and an information graphic code steganography module; the forward process of the information graphic code preprocessing module is used to encode the spliced ​​information graphic code, downsample the spliced ​​information graphic code into a reduced form that supports hiding in a video frame and obtain a corresponding first key map; the reverse process of the information graphic code preprocessing module is used to upsample the reduced form of the spliced ​​information graphic code decoded by the information graphic code steganography module in combination with the first key map to restore the decoded spliced ​​information graphic code; the reduced form of the spliced ​​information graphic code and the video frame to be processed are input into the information graphic code steganography module, the information graphic code steganography module performs discrete wavelet transform on the reduced form and video frame to the frequency domain, and then the frequency domain data is input into the reversible neural network of the information graphic code steganography module to steganographically obtain a second key map and frequency domain data of the video frame with the information graphic code, and then undergo inverse frequency domain transform to obtain the second key map and the traceable video frame.

2. The method for managing the security of a video of a highly confidential scene operation according to claim 1, wherein: Through segmented sampling, multiple sampled video frames are extracted from the entire time span of the video recording the confidentiality operation. The information steganography model performs confidentiality operation information steganography on each sampled video frame to obtain a sampled traceable video frame, and the sampled traceable video frame is replaced with the original corresponding sampled video frame to obtain a traceable video.

3. The method for managing the security of a video of a high-confidentiality scene confidentiality operation according to claim 1, characterized in that: A two-dimensional code generator is used to convert the confidential operation information into a corresponding information graphic code, and at least one information graphic code written into the video frame is spliced ​​together according to a set format to obtain a spliced ​​information graphic code.

4. The method for managing the security of a video of a highly confidential scene operation according to claim 3, wherein: When the number of information graphic codes written into the video frame is less than the number of information graphic codes set in the format, the copy result of the information graphic codes written into the video frame is used to fill the number to meet the format setting.

5. The method for managing the security of a video of a high-confidentiality scene confidentiality operation according to claim 1, characterized in that: The forward process of the reversible neural network of the information graphic code preprocessing module is as follows: Feature map of the concatenated information graphic code input to any k-th reversible neural module Divided into sub-feature maps along the channel dimension and sub-feature maps , the sub-feature map The transformed and sub-feature maps are calculated through the first residual dense convolutional network Additive combination, the result of the additive combination is used as the input feature map of the subsequent k+1th reversible neural module that constitutes the information graphic code preprocessing module Sub-feature map of ; Sub-feature map In the kth reversible neural module, the scaling rotation factor is learned by the second residual dense convolution network and the scalable Sigmoid activation function, and the translation factor is learned by the third residual dense convolution network. After multiplying with the scaling rotation factor and then adding and combining with the translation factor, the result of the addition and combination is used as the feature map of the subsequent k+1th reversible neural module that constitutes the information graphic code preprocessing module Sub-feature map of .

6. The method for managing the security of a video of a highly confidential scene operation according to claim 1, wherein: The reverse process of the reversible neural network of the information graphic code preprocessing module is as follows: Feature map of the information graphic code of the concatenation of any k+1th reversible neural module input to the information graphic code preprocessing module Divided into sub-feature maps along the channel dimension and sub-feature maps , the sub-feature map The scaling rotation factor is learned by the second residual dense convolution network and the scalable Sigmoid activation function, and the translation factor is learned by the third residual dense convolution network. Subtract the translation factor and divide it by the scaling rotation factor to obtain the sub-feature map of the k-th reversible neural module of the information graphic code preprocessing module. ; Sub-feature map Subtract sub-feature map The transformation result calculated by the first residual dense convolutional network is used as the sub-feature map of the k-th reversible neural module of the information graphic code preprocessing module .

7. The method for managing the security of a video of a highly confidential scene operation according to claim 1, wherein: The forward process of the reversible neural network of the information graphic code steganography module is as follows: The reduced form of the information graphic code input to the first reversible neural module of the information graphic code steganography module and video frames The reduced form of the information graphic code is converted to the frequency domain through frequency domain transformation and then input into any k-th reversible neural module of the information graphic code steganography module. Feature map The fourth residual dense convolutional network calculates the transformed video frame Feature map Additive combination, the result of the additive combination is used as the feature map of the video frame of the subsequent k+1th reversible neural module that constitutes the information graphic code steganography module ; Feature map of video frame In the kth reversible neural module of the information graphic code steganography module, the non-scaling rotation factor is learned by the fifth residual dense convolutional network and the scalable Sigmoid activation function, and the translation factor is learned by the sixth residual dense convolutional network. The feature map of the spliced ​​information graphic code The feature map of the information graphic code is obtained by multiplying the non-scaling rotation factor and then adding the translation factor as the concatenation of the subsequent k+1 reversible neural modules that constitute the information graphic code steganography module. .

8. The method for managing the security of a video of a highly confidential scene operation according to claim 1, wherein: The reverse process in the reversible neural network of the information graphic code steganography module is as follows: The second key image and the traceable video frame input to the information graphic code steganography module are converted to the frequency domain through frequency domain transformation. The final feature map of the traceable video frame in the frequency domain received by the K-th reversible neural module is The characteristic map of the second key map in the sum frequency domain ; Then input the feature map of the traceable video frame in the frequency domain of any k+1th reversible neural module of the information graphic code steganography module The translation factor is calculated by the sixth residual dense convolution network, and the non-scaled rotation factor is calculated by the fifth residual dense convolution network and the scalable Sigmoid activation function; the feature map of the spliced ​​information graphic code in the frequency domain Subtract the translation factor and divide it by the non-scaling rotation factor, and the result is used as the feature map of the information graphic code spliced ​​by the kth reversible neural module of the information graphic code steganalysis module : Feature Map Subtract feature map The feature map of the information graphic code after the transformation calculated by the fourth residual dense convolutional network as the splicing of the k-th reversible neural module of the information graphic code steganalysis module .

9. The method for securely managing videos of highly confidential scenes and confidential operations according to claim 1, wherein: The implementation of the scalable Sigmoid activation function is as follows: the Sigmoid activation function compresses the output of the residual dense convolutional network to the range of (0,1), and then translates it to the range of (-0.5, 0.5) by subtracting 0.5, so that the final scaled or unscaled rotation factor is symmetric around 0. Then, the output value is scaled by the scaled or unscaled coefficient and the exponential function of the natural exponent e is taken to obtain the scaled or unscaled rotation factor.

10. The method for securely managing videos of highly confidential scenes and confidential operations according to claim 1, wherein: During training, the learning objectives of the information steganography model include minimizing the following loss terms: The difference between the video frame to be processed and the traceable video frame; The difference between the frequency domain of the video frame to be processed and the traceable video frame; The difference between the reduced form of the spliced ​​information graphic code encoded by the forward process of the information graphic code steganography module and the predicted reduced form of the spliced ​​information graphic code restored from the traceable video frame by the reverse process of the information graphic code steganography module combined with the second key graph is minimized; the difference between the spliced ​​information graphic code decoded by the reverse process of the information graphic code steganography module and the original information graphic code is minimized.

Images