A file encryption method and system

Abstract

The application discloses a file encryption method and system, the method comprises the following steps: obtaining and processing face video dynamics and original voice information to obtain face feature information and voiceprint feature information; fusing the face feature information and the voiceprint feature information into identity feature information as a first encryption key; obtaining an initial file and processing the initial file according to the byte importance to obtain a first file and a second file; converting the first file into an initial image, encrypting the initial image through a chaotic iteration encryption algorithm and pixel permutation to obtain an encrypted transmission image; fusing the first encryption key and the initial image to obtain a second encryption key and encrypting the second file to obtain an encrypted transmission file; and encrypting the encrypted transmission image and the encrypted transmission file through the first encryption key and transmitting the encrypted transmission image and the encrypted transmission file to a target end. The key used in the application has strong correlation with plaintext information, can improve the security of file transmission, prevent plaintext attacks, and can reduce the data operation amount and data burden in the encryption process.

Description

Technical Field

[0001] This invention relates to the field of data encryption technology, and in particular to a file encryption method and system. Background Technology

[0002] Data encryption refers to converting plaintext into ciphertext using encryption algorithms and keys, while decryption uses decryption algorithms and keys to recover plaintext. Cryptography is the core of data encryption. Currently, data encryption is widely used in data transmission and computer science, protecting data during transmission and thus achieving information shielding and security. As users' security awareness continues to increase, the demand for file encryption and decryption is also growing. Current file encryption typically uses strings, requiring users to input strings during both encryption and decryption, which is cumbersome, provides a poor user experience, and makes string encryption easy to crack, resulting in low security. Summary of the Invention

[0003] The purpose of this invention is to provide a file encryption method and system that performs different encryption operations on different parts of a file according to the importance of bytes, thereby improving the security of encrypted files and reducing the data burden during the encryption process.

[0004] The solution to the technical problem of this invention is as follows: In a first aspect, this application provides a file encryption method, comprising the following steps:

[0005] Collect facial video dynamics and raw audio information from the initiating end, preprocess the facial video dynamics and raw audio information and perform feature extraction to obtain facial feature information and voiceprint feature information;

[0006] The facial feature information and the voiceprint feature information are fused to obtain identity feature information, which is then used as the first encryption key.

[0007] Obtain the initial file and determine its byte importance. Process the initial file according to the byte importance and importance threshold to obtain the first file and the second file. The byte importance of the first file is greater than that of the second file.

[0008] The first file is converted into a grayscale initial image, and the initial image is encrypted using a chaotic iterative encryption algorithm to obtain a first encrypted image. The first encrypted image is then pixel-scrambled using an image scrambling algorithm to obtain an encrypted transmission image.

[0009] The first encryption key and the initial image are combined to form a second encryption key. The second file is then encrypted using the second encryption key to obtain an encrypted transmission file.

[0010] The encrypted transmission image and the encrypted transmission file are concatenated into encrypted transmission data, and the first encryption key and the second encryption key are embedded in the junction of the encrypted transmission image and the encrypted transmission file. Then, the encrypted transmission image is encrypted using the first encryption key and transmitted to the target end.

[0011] Furthermore, the steps of preprocessing the original speech information and extracting its features include:

[0012] The original speech information is pre-emphasized;

[0013] The original speech information after pre-emphasis is subjected to frame segmentation and windowing to obtain the pre-processed original speech information;

[0014] Performing a Fast Fourier Transform on the preprocessed original speech information yields the following amplitude spectrum of the preprocessed original speech information:

[0015]

[0016] Where U(k) is the amplitude spectrum of the preprocessed original speech information, T[n] is the preprocessed original speech information, and N is the window width of the window function when performing fast Fourier transform;

[0017] The amplitude spectrum is moduloed and squared to obtain the discrete power spectrum of the preprocessed original speech information; the discrete power spectrum satisfies the following formula:

[0018]

[0019] Wherein, P(k) is the discrete power spectrum of the preprocessed original speech information;

[0020] The discrete power spectrum is frequency integrated using a Gammatone filter bank; the time-domain impulse response of the Gammatone filter bank satisfies the following formula:

[0021]

[0022] Where c is the scaling factor, n is the order of the Gammatone filter, b is the time attenuation factor, and f0 is the center frequency of the Gammatone filter. The phase of the Gammatone filter;

[0023] Calculate the long-time frame power of the preprocessed original speech information and mask and suppress noise other than human voice; the long-time frame power satisfies the following formula:

[0024]

[0025] Where Q(i,j) is the long-time frame power, and P[i′,j] is the power spectrum of the current frame and a certain frame among the i frames before and after it;

[0026] The time and frequency domains of the original speech information after normalization preprocessing and noise masking;

[0027] The power spectrum after time-frequency normalization is calculated using a nonlinear function power, and then dimensionality is reduced using discrete cosine transform to obtain voiceprint feature information.

[0028] Furthermore, the steps of preprocessing the facial video dynamics and extracting its features include:

[0029] Several image frames of a facial video are captured at a preset sampling time.

[0030] High-pass filtering is applied to each image frame;

[0031] The high-pass filtered image frames were denoised using median filtering to obtain several face images to be extracted.

[0032] Construct a face detection model, perform face detection on the face image to be extracted using the face detection model, and output a face detection image;

[0033] The face feature information is obtained by extracting face features from the face detection image using a multi-task cascaded convolutional neural network.

[0034] Furthermore, the facial feature information and the voiceprint feature information are fused to obtain identity feature information, including:

[0035] Acquire several facial feature information items, and collect the timestamps of the image frames corresponding to the facial feature information items respectively;

[0036] Based on the timestamp, several facial feature information items are numbered in chronological order.

[0037] According to the number, the facial feature information is divided into ten groups of facial feature information. The first and last groups of facial feature information are removed. Four groups of facial feature information are randomly selected from the remaining eight groups of facial feature information and their average value is calculated to obtain the facial feature information to be fused.

[0038] The identity feature information is obtained by fusing the facial feature information to be fused and the voiceprint feature information.

[0039] Further, the step of processing the initial file according to byte importance and importance threshold to obtain the first file and the second file includes:

[0040] Obtain the initial file and calculate the importance of each byte in the initial file;

[0041] Record the sorting information of each byte and embed the sorting information into the first encryption key; the sorting information does not participate in the encryption operation.

[0042] Reorder all bytes in the initial file according to the importance of each byte to obtain a sorted initial file;

[0043] Based on the importance threshold, the initial files after sorting are divided into the first file and the second file.

[0044] Further, the step of encrypting the initial image using a chaotic iterative encryption algorithm to obtain a first encrypted image includes:

[0045] Let the initial point of the iteration be the function value of the N0th iteration:

[0046]

[0047] The initial image is divided into 1-byte blocks to form sub-message blocks P. j =p lj p lj+1 , ..., p (l+1)j-1 ;

[0048] Generate independent and identically distributed binary random sequences, and iterate the chaotic map D after this round of encryption. j Second-rate;

[0049] The binary random sequence is as follows:

[0050]

[0051] Sub-message block P j Circular left shift D j Bit, sub-message block P j The sequence after the permutation transformation is denoted as P. j ′;

[0052] For sequence P j Y and binary random sequence j Perform an XOR operation: in The XOR operator, C j For P j The corresponding ciphertext;

[0053] C j Expanding by l bits yields the ciphertext c corresponding to the plaintext string. lj c lj+1 c (l+1)j-1 That is, the first encrypted image.

[0054] Further, after generating the independent and identically distributed binary random sequences, the process includes:

[0055] Let Y in the binary random sequence j 1 The decimal value is D j The number of iterative chaotic mappings, D, is calculated. j ; where D j ′ and iteration number D j The relationship satisfies:

[0056]

[0057] Where, p lj+k k = 0, 2, ..., l-1 are sub-message blocks P j The value of each byte.

[0058] Further, the step of scrambling the pixels of the first encrypted image using an image scrambling algorithm to obtain an encrypted transmission image includes:

[0059] A multidimensional Arnold transform model is constructed, and the pixel positions and pixel grayscale values ​​of the first encrypted image are simultaneously scrambled using the multidimensional Arnold transform model to obtain an encrypted transmission image.

[0060] The multidimensional Arnold variational model is represented as follows:

[0061]

[0062] Among them, b x b y and b z Let r be any positive integer. x r y and r z Let Q be any positive integer, and V be the height and width values ​​of the encrypted initial image, respectively. W is the maximum number of binary bits corresponding to the pixel value, and the number of bits is selected as 16.

[0063] Further, the process of fusing the first encryption key and the initial image to form a second encryption key, and encrypting the second file using the second encryption key to obtain an encrypted transmission file, includes:

[0064] The second encryption key is obtained by associating the first encryption key and the initial image using the SHA256 hash algorithm;

[0065] The encrypted transmission file is obtained by encrypting the second file using a symmetric encryption algorithm and the second encryption key.

[0066] On the other hand, this application provides a file encryption system, the system comprising:

[0067] At least one processor:

[0068] At least one memory for storing at least one program;

[0069] When the at least one program is executed by the at least one processor, the at least one processor implements the file encryption method.

[0070] The beneficial effects of this invention are: it provides a file encryption method and system that performs different encryption processes on different parts of the file according to the importance of the file, and the key used is strongly correlated with the plaintext information, which can improve the security of file transmission, prevent statistical and differential attacks, selected plaintext and birthday attacks, avoid the phenomenon of encrypted files being illegally decrypted during transmission, and at the same time reduce the amount of data computation and data burden in the encryption process.

[0071] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the description, claims and drawings. Attached Figure Description

[0072] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0073] Figure 1 A flowchart illustrating a file encryption method provided in this application embodiment;

[0074] Figure 2 A flowchart illustrating a method for processing an initial file based on byte importance and an importance threshold to obtain a first file and a second file, provided in an embodiment of this application.

[0075] Figure 3 This is a schematic diagram of a file encryption method provided in an embodiment of this application. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0077] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0078] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0079] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0080] With the development of information technology, secure file transmission has become increasingly important. To ensure the security and confidentiality of file transmission, an algorithm is used at the sending end to scramble the file data into ciphertext, making it invisible to non-communication parties. The receiving end can then decrypt the ciphertext using a known key to obtain the original file data. Commonly used file encryption algorithms in this field include RSA (asymmetric encryption), MD5 (Message Digest Algorithm), and AES (Advanced Encryption Standard). However, encrypted files are still vulnerable to external attacks during transmission, such as statistical and differential attacks, chosen plaintext attacks, and birthday attacks. Under the influence of these attacks, the ciphertext can be illegally decrypted, leading to the leakage of plaintext information and failing to achieve the effect of secure transmission.

[0081] Image scrambling is a type of information hiding technique. The main purpose of image scrambling is to transform a given image into a jumbled, invisible image. After scrambling, the image becomes unrecognizable, thus hiding and protecting the information. It can be used for encrypted transmission or as a preprocessing step in image processing. Based on its characteristics, image scrambling can be categorized into: spatial domain scrambling, frequency domain scrambling, and hybrid spatial-frequency domain scrambling.

[0082] To address the aforementioned technical problems, this application provides a file encryption method and system that combines image scrambling, referring to... Figure 1 As shown, Figure 1 The diagram shown is a flowchart of a file encryption method provided in an embodiment of this application. An embodiment of this application will be described and explained below. The file encryption method may include, but is not limited to, the following steps.

[0083] S1: Collect facial video dynamics and raw audio information from the initiating end, preprocess the facial video dynamics and raw audio information and extract features to obtain facial feature information and voiceprint feature information.

[0084] It should be noted that the sender's facial video is captured by the camera of the initiating device, and the sender's raw voice information is captured by the microphone of the initiating device. For example, when the sender selects encrypted file transmission at the initiating device, the device can randomly select a word from its pre-configured random word library and prompt the sender to read it aloud. During the sender's reading, the microphone captures the sender's voice, while the camera captures the video information of the sender reading aloud.

[0085] S2, integrate facial feature information and voiceprint feature information to obtain identity feature information, and use the identity feature information as the first encryption key.

[0086] It should be noted that the first encryption key serves as the private key for the symmetric encryption algorithm used in subsequent transmissions.

[0087] Optionally, in this step, a feature layer fusion algorithm is constructed to fuse facial feature information and voiceprint feature information.

[0088] S3: Obtain the initial file and determine its byte importance. Process the initial file according to the byte importance and importance threshold to obtain the first file and the second file. The byte importance of the first file is greater than that of the second file.

[0089] In this step, the bytes in the initial file to be transmitted by the initiating end have different importance levels. The bytes of the initial file are then reordered according to their importance. The reordered initial file can be divided into a first file and a second file based on an importance threshold. Different encryption operations are then performed on the first and second files respectively. The byte importance of the second file is lower than that of the first file.

[0090] It should be noted that encrypting bytes of high importance is far more complex than encrypting bytes of low importance. Simultaneously encrypting both high-importance and low-importance bytes using the same encryption method would make the encryption process even more cumbersome. This application employs different encryption methods for bytes of varying importance, which not only improves the security of high-importance data transmission but also reduces the computational load of the encryption process.

[0091] S4. Convert the first file into a grayscale initial image, encrypt the initial image using a chaotic iterative encryption algorithm to obtain a first encrypted image, and then scramble the pixels of the first encrypted image using an image scrambling algorithm to obtain an encrypted transmission image.

[0092] This step emphasizes the encryption process for the first file. For the encryption of the first file, this application converts it to a grayscale image and uses image encryption to process the first file.

[0093] S5, the first encryption key and the initial image are combined to form a second encryption key, and the second file is encrypted using the second encryption key to obtain an encrypted transmission file.

[0094] In this step, the first file is plaintext information, and the first encryption key is identity feature information, which is a key. This application obtains a second encryption key by fusing the two, making the generation process of the second encryption key related to the plaintext information, thus spreading the influence of a single plaintext bit to more ciphertext and achieving better encryption results.

[0095] S6, the encrypted transmission image and the encrypted transmission file are concatenated into encrypted transmission data, and the first encryption key is embedded in the junction of the encrypted transmission image and the encrypted transmission file. Then, the encrypted transmission image is encrypted using the first encryption key and transmitted to the target end.

[0096] An embodiment of this application will be further described and explained below for S1. The specific process of preprocessing the raw speech information in S1 may include, but is not limited to, the following steps.

[0097] Pre-emphasis processing of raw speech information:

[0098] α[k]=β[k]-δβ[k-1]

[0099] In the above formula, the pre-emphasized original speech information is represented by α[k], and the k-th and (k-1)-th sampling points of the original speech information are represented by β[k] and β[k-1], respectively. Optionally, β is a parameter value greater than or equal to 0.8 and less than or equal to 1.0.

[0100] The pre-emphasized original speech information is framed and windowed to obtain the pre-processed original speech information:

[0101] T[n] = Y[n] * f[n]

[0102]

[0103] The above formula represents windowing processing. In the formula, the preprocessed original speech information is represented by T[n], and f[n] represents the Hamming window function.

[0104] In this specific embodiment, the original speech information is a type of digital signal. During digital signal processing, each FFT transform can only transform a finite length of time-domain data. Therefore, signal truncation of the time-domain signal is necessary. Even with periodic signals, if the truncation time is not periodic, the truncated signal will have leakage. To minimize this leakage error, a weighting function, also called a window function, is needed. The Hamming window function is a type of window function; its amplitude-frequency characteristic features significant sidelobe attenuation, with the main lobe peak and the first sidelobe peak attenuating by up to 40 dB, effectively reducing frequency leakage of the original speech information.

[0105] In one embodiment of this application, the extraction of feature information from the original speech information in S1 will be described and explained below. In this application, the features of the original speech information are extracted using the PNCC algorithm. The PNCC algorithm is a speech feature extraction algorithm. Compared with the traditional MFCC speech feature extraction algorithm, the PNCC algorithm uses power-law nonlinearity to replace the traditional logarithmic nonlinearity in the MFCC coefficients, and adds a noise suppression algorithm based on asymmetric filtering to suppress background excitation and a time masking module to improve the speech recognition effect in noisy scenes. The process of extracting features from the preprocessed original speech information may include, but is not limited to, the following steps. The preprocessed original speech information is denoted as the first speech information.

[0106] First, a Fast Fourier Transform is performed on the preprocessed original speech information to obtain the amplitude spectrum of the preprocessed original speech information.

[0107] It should be noted that the amplitude spectrum is as follows:

[0108] It is subject to the constraint 0≤k≤N.

[0109] In the above formula: the amplitude spectrum is represented by U(k), the first speech information is represented by T[n], and N can be understood as the window width of the Hamming window function.

[0110] By performing modulo and squaring calculations on the amplitude spectrum, the discrete power spectrum of the first speech information is obtained as follows:

[0111]

[0112] In the above formula, the discrete power spectrum is represented by P(k).

[0113] Then, the discrete power spectrum is frequency integrated using a Gammatone filter bank.

[0114] It should be noted that the time-frequency impulse response of the filter bank is:

[0115]

[0116] In the above formula, the scaling factor is represented by c, the filter order by n, the time attenuation factor by b, the center frequency of the filter by f0, and its phase by... To express.

[0117] Calculate the long-time frame power of the preprocessed original speech information and mask and suppress noise other than human voice.

[0118] It should be noted that the power of a long-time frame can be expressed as:

[0119]

[0120] The long-time frame power is represented by Q(i,j), and the power spectrum of the current frame and a certain frame in each of the i frames before and after it is represented by P[i′,j].

[0121] Then, the time and frequency domains of the original speech information are normalized and preprocessed to mask noise.

[0122] It should be noted that the power spectrum is obtained after normalization, and is represented as follows:

[0123]

[0124] In the above formula: F[i,j′] represents the noise figure excluding human voice. V[i,j] represents the normalized power spectrum, and Q(i,j) = Q[i,j′].

[0125] Finally, the power function power of the time-frequency normalized power spectrum is calculated, and the dimension is reduced by discrete cosine transform to obtain the voiceprint feature information.

[0126] In this application, voiceprint feature information is represented by PNCC coefficients.

[0127] In one embodiment of this application, the preprocessing process of the facial video dynamics in S1 will be described and explained below. In this application, the acquired facial video dynamics are videos of the sender reading preset fields, which include several video frames or time frames. Before performing face detection and facial feature extraction, it is necessary to first crop the video and then preprocess the images to improve the accuracy of face detection and facial feature extraction. Preprocessing of the facial video dynamics may include, but is not limited to, the following steps.

[0128] A number of image frames f(x) of facial video motion are captured at a preset sampling time. i y i ), i = 1, 2, 3, ..., n.

[0129] For image frames f(x) respectively i y i Perform high-pass filtering:

[0130]

[0131] The formula above represents the Laplace operator. The Laplace operator is a second-order differential operator in n-dimensional Euclidean space, used to preserve the high-frequency components (rapidly changing parts) of an image and suppress the low-frequency components (slowly changing parts).

[0132] The high-pass filtered image frames were denoised using median filtering to obtain several face images to be extracted.

[0133] It should be noted that median filtering can be expressed as:

[0134]

[0135] Among them, the denoised image frame can be obtained by g(x) i y i The image frame after high-pass filtering is represented by ) Let θ be the median filter template. Additionally, (i, j) is the two-dimensional template for median filtering. Optionally, the two-dimensional template can be 5×5, i.e., θ∈5×5.

[0136] In one embodiment of this application, the process of feature extraction from several face images to be extracted in S1 will be described and explained below. The feature extraction process of this application can be divided into two steps: face detection and face feature extraction. The purpose of face detection is to find the region of interest, i.e., the face region, in the face image to be extracted, so as to facilitate subsequent face feature extraction. Face feature extraction is to extract the features of the face from the region of interest. Specifically, feature extraction from several face images to be extracted may include, but is not limited to, the following steps.

[0137] A face detection model is constructed to perform face detection on the extracted face image and output the face detection image.

[0138] Facial feature information is obtained by extracting facial features from face detection images using a multi-task cascaded convolutional neural network.

[0139] The face detection process is as follows:

[0140] First, a face dataset is created, and the dataset is divided into a first training set and a first test set according to a first ratio.

[0141] Specifically, a number of face images are searched online. Optionally, the number of face images is 4000. Then, the face images are preprocessed and uniformly numbered. Next, the face regions in the face images are labeled using the Labellmg image annotation software, and the location information of the face regions and their corresponding classification information (face or non-face) are saved as an XML file required for network training, i.e., the face dataset. Finally, the face dataset is divided into a 70% training set and a 30% test set.

[0142] Optionally, preprocessing the face image includes:

[0143] Randomly translate, flip, and add Gaussian noise to face images;

[0144] The cutout data augmentation method is used to occlude the face image with a fixed-size rectangle and set the values ​​within the rectangle to 0.

[0145] The steps described above are to expand the sample size in the face dataset to prevent overfitting during network training.

[0146] Next, a YOLOv5 neural network was built, and a CBAM attention mechanism was added to the feature extraction network of the YOLOv5 neural network. The bounding box regression function at its output end was optimized to construct the first initial recognition model. The first training set was then input into the first initial recognition model for training.

[0147] It should be noted that YOLOv5 is the fifth generation of the YOLO neural network series. YOLOv5 has a simple training process, effectively improves detection speed, reduces repetitive gradient information, and gives the network better learning capabilities.

[0148] It should be noted that CBAM (Convolutional Block Attention Module) is a commonly used attention mechanism. CBAM represents the attention mechanism module of the convolutional module, which combines spatial and channel attention modules, enabling the model to prioritize key features while ignoring useless features. This application adds the CBAM attention mechanism to the neural network, which can improve YOLOv5's feature extraction capability during training, enabling it to accurately extract key features and thus improve model performance.

[0149] It should be noted that the optimized bounding box regression function satisfies the CIOU_Loss (Complete IntersectionOver Union_Loss) algorithm. Currently, the loss function for YOLOv5 is GIOU_Loss, but it cannot identify the location of the predicted bounding box when it is within the target ground truth bounding box. To address this, this application optimizes the bounding box regression function to solve the problem of failing to identify the location of the predicted bounding box when it is within the target ground truth bounding box, thereby reducing the model's error rate.

[0150] Then, the performance of the first initial recognition model after training is evaluated using the first test set. When the performance of the first initial recognition model reaches the preset conditions, the face detection model is output.

[0151] In the above steps, if the performance of the first initial recognition model does not meet the preset conditions, the training hyperparameters of the first initial recognition model are modified, and the first initial recognition model is retrained. The performance evaluation includes mean precision and accuracy. Accuracy satisfies:

[0152]

[0153] Where Acc is the accuracy, and T is the precision. p F represents the number of samples in the dataset that actually contain faces and whose faces have been detected. p This represents the number of dataset samples that are not actually human faces but are detected as such.

[0154] The mean and average precision satisfy the following formula:

[0155]

[0156] Where MAP is the mean average precision, J(P,R) k This is the average precision function.

[0157] Finally, the trained face detection model is used to perform face detection on the face image to be extracted, resulting in a face detection image, i.e., the face region.

[0158] The process of facial feature extraction:

[0159] This application employs a multi-task convolutional neural network (MTCNN) for face feature extraction. MTCNN consists of three layers: P-Net (Proposal Network), R-Net (Refine Network), and O-Net (Output Network). P-Net is a region proposal network for faces. After the features are input into three convolutional layers, a face classifier determines whether the region contains a face. Simultaneously, bounding box regression and a facial landmark locator are used for preliminary face region localization. This part ultimately outputs many potential face regions, which are then input into R-Net for further processing. In other words, the role of P-Net is to quickly generate face candidate windows.

[0160] Because P-Net's output only provides a face region with a certain degree of confidence, and this face region may not necessarily contain a face, R-Net refines the input, discarding most of the erroneous inputs, and again uses bounding box regression and facial landmark localizers to perform bounding box regression and landmark localization for the face region. Finally, R-Net outputs a more reliable face region for O-Net to use. Compared to P-Net's 1x1x32 feature output using fully convolutional layers, R-Net uses a 128-bit fully connected layer after the last convolutional layer, preserving more image features and achieving better accuracy than P-Net. In other words, R-Net further selects and adjusts the face candidate window, achieving high-precision filtering and face region optimization. O-Net can retain more face image features while performing face discrimination, face region bounding box regression, and face feature localization. Finally, O-Net outputs the top-left and bottom-right coordinates of the face region and five feature points of the face region.

[0161] It should be noted that before inputting the face image to be extracted into MTCNN, the format of the face image to be extracted needs to be adjusted to conform to the input format of MTCNN.

[0162] According to one embodiment of this application, the process of fusing voiceprint feature information and facial feature information in S2 will be described and explained below. The fusion process may include, but is not limited to, the following steps.

[0163] Acquire several facial feature information and collect the timestamps of the image frames corresponding to the facial feature information.

[0164] In this step, facial feature information can be progressively mapped to its source: facial video dynamics. For video information, each sampled frame corresponds to a timestamp. This application extracts the timestamps corresponding to facial feature information to facilitate subsequent filtering and processing of multiple facial feature pieces.

[0165] Based on the timestamps of the image frames, several facial feature information items are numbered in chronological order;

[0166] Based on the number, several facial feature information is divided into ten groups of facial feature information. The first and last groups of facial feature information are removed. Four groups of facial feature information are randomly selected from the remaining eight groups of facial feature information and their average is calculated to obtain the facial feature information to be fused.

[0167] In the above steps, facial feature information is sorted according to time values ​​and assigned corresponding numbers, dividing several facial feature information into ten groups. For example, if the total duration of a facial video animation is t, then the total duration t can be divided into t1, t2, t3…, t… 10 These ten time periods are categorized, and the facial feature information corresponding to each time period is grouped into a specific facial feature information group for that time period. This application excludes the facial feature information group corresponding to t1 and t... 10 The purpose of the corresponding facial feature information group is that when the video is first recorded, there may be external interference that causes the video signal to be unstable. Therefore, the first and last groups of facial feature information will be removed.

[0168] By fusing the facial features and voiceprint features of the person to be fused, identity feature information is obtained.

[0169] In this step, the two feature information can be fused using a feature layer fusion algorithm based on convolutional neural networks.

[0170] Reference Figure 2 As shown in one embodiment of this application, step S3 of this application may specifically include the following steps:

[0171] S31, Obtain the initial file and calculate the importance of each byte in the initial file;

[0172] S32, record the sorting information of each byte and embed the sorting information into the first encryption key. The sorting information does not participate in the encryption operation.

[0173] It should be noted that the sorting information is used by the target end to decrypt and obtain the first and second files, and then reorder them to obtain the initial file.

[0174] S33, Reorder all bytes in the initial file according to the importance of each byte to obtain the sorted initial file;

[0175] S34. Based on the importance threshold, the sorted initial files are divided into the first file and the second file.

[0176] It should be noted that the byte importance of the second file is less than that of the first file.

[0177] In the above steps, encrypting bytes of high importance is far more complex than encrypting bytes of low importance. If the encryption of both high-importance and low-importance bytes is performed simultaneously, the encryption process becomes even more cumbersome. This application applies different encryption methods to bytes of different importance, which not only improves the security of high-importance data transmission but also reduces the computational load and data burden of encryption processing, thereby improving encryption efficiency and speed.

[0178] In one embodiment of this application, the encryption process of the first file in step S4 will be further described and explained below. The encryption process of the first file can be roughly divided into three steps: conversion to grayscale image; chaotic iterative encryption operation; and pixel scrambling operation.

[0179] Convert to grayscale image:

[0180] All initial files in this application are of string type. First, the parameters of the initial file are obtained from the command line; the parameters of the initial file are stored in a dictionary; the string of the template image is obtained from the dictionary, the string of the template image is decoded into template image bytecode, and the template image bytecode is stored in the initial image in binary form.

[0181] Chaotic iterative encryption operation:

[0182] Chaotic encryption primarily utilizes sequences generated iteratively by chaotic systems as a factor sequence for encryption transformation. The theoretical basis of chaotic encryption lies in the self-similarity of chaos, ensuring that locally selected chaotic key sets exhibit a distribution similar to the overall system. Chaotic systems are highly sensitive to their initial states, exhibit complex dynamics, and their distributions do not conform to probabilistic statistical principles; they are quasi-random sequences with complex structures. These quasi-random sequences possess good randomness, correlation, and complexity, making chaotic systems difficult to reconstruct, analyze, and predict.

[0183] First, perform the preliminary work:

[0184] Construct the Logistic mapping: f(x) = μ x (1-x), x∈[0,1].

[0185] Generate independent and identically distributed binary random sequences, and express the values ​​of x in binary form:

[0186] x = 0.b1(x), b2(x)..., bi (x)...

[0187] Where: x∈(a, b), bi∈{0, 1}.

[0188] The chaotic iterative encryption process is as follows: First, a mapping from a 16-bit message block to different regions of the chaotic map image space is defined. To prevent transient effects, the chaotic map is iterated N0 times in advance. The initial point of the iteration is taken as the function value of the N0th iteration.

[0189] It should be noted that the function value satisfies:

[0190]

[0191] Divide the initial image into blocks of l bytes each; the initial image can be considered as a plaintext string. Let l = 16, and the plaintext string p... lj p lj+1 , ..., p (l+1)j-1 This forms a sub-message block P. j =p lj p lj+1 , ..., p (l+1)j-1 This sub-message block P j =p lj p lj +1, ..., p (l+1)j-1 The number of bits is 16l.

[0192] The method described above for generating binary random sequences produces independent and identically distributed binary random sequences, and the chaotic map D is iterated after this round of encryption. j Second-rate.

[0193] Specifically, based on the theoretical formula for binary random sequences, binary random sequences can be generated. Then, use D j ' to represent Y j 1 The decimal value, through D j D was calculated j D j ′ and D j The following relationship exists:

[0194]

[0195] In the above formula: p lj+k k = 0, 2, ..., l-1 is P j The value of each byte.

[0196] In traditional chaotic iterative encryption algorithms, for a given ciphertext string, the state transition function in each round of encryption is independent of the input plaintext symbols; that is, the plaintext does not undergo state transitions. Therefore, traditional algorithms are vulnerable to plaintext attacks. In other words, the key used by the encryption algorithm is actually random and has no relation to the plaintext. To address this problem, this application changes the relationship between the plaintext, the key, and the ordinary binary sequence by altering the number of iterations in the chaotic mapping. This makes the binary sequence generation process correlated with the plaintext, spreading the influence of a single plaintext bit to more ciphertext bits, achieving better diffusion and confusion effects. Furthermore, it can prevent plaintext attacks.

[0197] Then, sub-message block P j Circular left shift D j 1 bit, will sub-message block P j Perform a scrambling transformation to obtain sequence P. j ′, and for sequence P j and Y j Perform an XOR operation.

[0198] It should be noted that the XOR operation satisfies: In this formula: P j The corresponding ciphertext is C j To express. It is the XOR operator.

[0199] Finally, C j Expanding by l bits, we get the plaintext string p. lj p lj+1 , ..., p (l+1)j-1 The corresponding ciphertext c lj， c lj+1 c (l+1)j-1 That is, c lj c lj+1 c (l+1)j-1 This is the first encrypted image.

[0200] Pixel scrambling operation:

[0201] Traditional chaotic iterative encryption algorithms do not consider the scrambling of pixel positions and grayscale values. That is, the modification of pixel positions does not embed information about the pixel grayscale value, resulting in unsatisfactory scrambling effects. Furthermore, the pixel positions (0,0) or (0,0,0) remain unchanged, making them vulnerable to plaintext attacks. Attackers can easily analyze specific images to uncover the encryption rules, leading to the decryption and leakage of plaintext information. The first encrypted image obtained in this application also suffers from poor scrambling effects due to the lack of consideration for pixel grayscale values. To address this, this application performs a secondary permutation on the first encrypted image and modifies both pixel positions and pixel grayscale values ​​using multidimensional Arnold algorithm. The pixel position information influences the modification of pixel grayscale values, and vice versa, thus improving the scrambling effect and enhancing the security of the scrambling result.

[0202] The two-dimensional unequal-length Arnold transform can be described as follows:

[0203]

[0204] Using the aforementioned two-dimensional unequal-length Arnold transform formula, the z, x, and y variables in three-dimensional space are fixed respectively, resulting in three two-dimensional unequal-length Arnold transforms corresponding to the xy plane, yz plane, and zx plane. These are then connected in a specific order to obtain the multidimensional unequal-length Arnold transform.

[0205] Therefore, the pixel scrambling process is as follows:

[0206] A multidimensional Arnold transform model is constructed, and the pixel positions and pixel grayscale values ​​of the first encrypted image are simultaneously scrambled using the multidimensional Arnold transform model to obtain the encrypted transmission image.

[0207] It should be noted that the multidimensional Arnold transform model can be described as follows:

[0208]

[0209] in:

[0210] In the above formula: b x b y and b z and r x r y and r z All are arbitrary positive integers. The height of the first encrypted image is represented by Q, and its width is represented by y. The maximum number of binary bits corresponding to a pixel value is represented by W, with a value of 16.

[0211] Specifically, firstly, the number of times the image is scrambled is selected as n1, the scrambling number control variable is initialized, and the three-dimensional binarized image S corresponding to the scrambled grayscale image is calculated.

[0212] The following formula is used to scramble the 3D binarized image S to obtain a new 3D binarized image S. z :

[0213]

[0214] The following formula is used to perform binarization of the three-dimensional image S. z Scrambling is performed to obtain a new 3D binarized image S. x :

[0215]

[0216] The following formula is used to perform binarization of the three-dimensional image S. x Scrambling is performed to obtain a new 3D binarized image S. y And the three-dimensional binarized image S y Assign the value to the 3D binarized image S:

[0217]

[0218] Then, the two-dimensional stereo binary image S is used as the encrypted transmission image.

[0219] In one embodiment of this application, S5 is the process of encrypting the second file. The first encryption key and the initial image are fused, that is, identity feature information and plaintext information are fused, to obtain the second encryption key. This application obtains the second encryption key by fusing the two, making the generation process of the second encryption key related to the plaintext information, thus spreading the influence of a single plaintext bit to more ciphertext and achieving a better encryption effect. The second encryption key is no longer a randomly generated key unrelated to the plaintext information. It should be noted that the encrypted transmission file is obtained after encryption using a symmetric encryption algorithm and the second encryption key.

[0220] Optionally, the first encryption key and the initial image are associated with the hash algorithm SHA256 to obtain the second encryption key; the second file is then encrypted using a symmetric encryption algorithm in combination with the second encryption key to obtain the encrypted transmission file.

[0221] Optionally, identity information and plaintext information can be fused through the feature layers of a neural network. The neural network can be VGG19.

[0222] In one embodiment of this application, in step S6, the encrypted transmission data consists of an encrypted transmission image obtained by encrypting a first file and an encrypted transmission file obtained by encrypting a second file. In this step, the first encryption key is the key used for the last encryption before transmission. The encrypted transmission image carries both the first and second encryption keys to facilitate decryption of the received encrypted data by the target end. Furthermore, before transmission, the encrypted transmission image is encrypted using the first encryption key in conjunction with a symmetric encryption algorithm. This significantly enhances the security of the ciphertext during transmission and effectively prevents various plaintext attacks.

[0223] Alternatively, the symmetric encryption algorithm can be any one of DES, 3DES, TDEA, Blowfish, RC5, and IDEA.

[0224] Reference Figure 3 As shown, Figure 3 The diagram shown illustrates the principle of the file encryption method provided in this application. The principles and implementation process of the technical solution of this application will be summarized and explained below.

[0225] First, it is necessary to extract the sender's identity information, and use this identity information as the private key for the symmetric encryption algorithm used in subsequent transmissions, as well as as the encryption key for part of the initial file.

[0226] The initial files are reordered based on byte importance and divided into a first file and a second file. The first file has a higher importance than the second file. This means the first file requires more complex encryption to prevent plaintext decryption, while the second file does not require overly complex encryption. This approach reduces the data burden on the initiating end during encryption while ensuring file security.

[0227] The next step is the encryption of the first file: first, the first file is converted into an image, which facilitates subsequent image encryption. After conversion, a chaotic iterative encryption algorithm is first used for image encryption, and then an image scrambling algorithm is used to scramble the pixels of the image, forming an encrypted transmission image. This application changes the relationship between plaintext, key, and ordinary binary sequence by changing the number of iterations of the chaotic mapping, making the binary sequence generation process related to the plaintext, spreading the influence of a single plaintext bit to more ciphertext bits, and achieving better diffusion and confusion effects. Furthermore, it can prevent plaintext attacks.

[0228] The next step is the encryption of the second file: the first encryption key and the first file are combined, ensuring a strong correlation between the resulting key and the plaintext information, effectively reducing the success rate of plaintext attacks. The second file is then encrypted using a symmetric encryption algorithm. This application improves the security of the second file simply by changing the form of the key, while simultaneously reducing the data burden of the encryption process.

[0229] Finally, the encrypted transmitted image and the encrypted transmitted file are concatenated and then encrypted using the first encryption key before being transmitted.

[0230] The method and system provided in this application perform different encryption processes on different parts of the file according to the importance of the file. The key used is strongly correlated with the plaintext information, which can improve the security of file transmission, prevent statistical and differential attacks, selected plaintext and birthday attacks, avoid the phenomenon of encrypted files being illegally decrypted during transmission, and reduce the amount of data computation and data burden in the encryption process.

[0231] Regarding the above encryption process, one embodiment of this application provides a decryption method for the above encryption method. The encryption method of this application can be briefly divided into four steps: the initial file is divided into a first file and a second file; the first file is subjected to chaotic encryption and pixel scrambling; the second file is subjected to symmetric encryption; and the concatenated file of the first and second files is subjected to symmetric encryption. Therefore, the decryption method can be divided into the following four steps:

[0232] Step one: The target receives the encrypted transmitted image, extracts the first encryption key from it, and verifies whether the identity information matches that of the initiator. If so, proceed to step two. If the identity information does not match that of the initiator, the ciphertext is rejected.

[0233] Step two: Decrypt the encrypted transmitted data using the first encryption key to obtain the encrypted transmitted image and encrypted transmitted file;

[0234] Step 3: Decrypt the encrypted file using the second encryption key to obtain the second file;

[0235] Step four: Construct a multidimensional unequal-length Arnold inverse transform model, and use the inverse transform model to recover the pixel positions of the encrypted transmission image to obtain the first encrypted image.

[0236] Specifically, the new three-dimensional stereo binary image S is recovered using the following formula.

[0237] Volume binarized image S y :

[0238]

[0239] The following formula is used to evaluate the three-dimensional binary image S. y After restoration, a three-dimensional binarized image S is obtained. x :

[0240]

[0241] The following formula is used to convert the three-dimensional binary image S. x After restoration, a three-dimensional binarized image S is obtained. z And assign it the value of the first encrypted image:

[0242]

[0243] Step four: Decrypt the first encrypted image using the chaotic iterative encryption algorithm to obtain the initial image. Then, de-grayscale the initial image to obtain the color initial image.

[0244] Specifically, for sequence C j and Y j Perform an XOR operation:

[0245]

[0246] Obtain the replaced message block P j Then according to D j The initial grayscale image can be obtained by performing an inverse permutation transformation on the value.

[0247] Step 5: Convert the initial color image into the first file using image-to-string conversion code;

[0248] Step 6: Extract the sorting information from the first encryption key, reorder the first and second files, and obtain the initial file.

[0249] In addition, this application also provides a file encryption system, the system comprising:

[0250] At least one processor:

[0251] At least one memory for storing at least one program;

[0252] When at least one program is executed by at least one processor, the at least one processor implements one of the above-described file encryption methods.

[0253] Furthermore, this application also provides a file decryption system, the system comprising:

[0254] At least one processor:

[0255] At least one memory for storing at least one program;

[0256] When at least one program is executed by at least one processor, the at least one processor implements one of the file decryption methods described above.

[0257] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatuses.

[0258] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0259] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, apparatuses, or units, and may be electrical, mechanical, or other forms.

[0260] The units described as separate components may or may not be physically separate. 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 the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0261] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0262] If the integrated unit is implemented as 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 this 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. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0263] The step numbers in the above method embodiments are set only for ease of explanation and do not limit the order of the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

Claims

1. A file encryption method, characterized in that, The steps include the following: Collect facial video dynamics and raw audio information from the initiating end, preprocess the facial video dynamics and raw audio information and perform feature extraction to obtain facial feature information and voiceprint feature information; The facial feature information and the voiceprint feature information are fused to obtain identity feature information, which is then used as the first encryption key. Obtain the initial file and determine its byte importance. Process the initial file according to the byte importance and importance threshold to obtain the first file and the second file. The byte importance of the first file is greater than that of the second file. The first file is converted into a grayscale initial image, and the initial image is encrypted using a chaotic iterative encryption algorithm to obtain a first encrypted image. The first encrypted image is then pixel-scrambled using an image scrambling algorithm to obtain an encrypted transmission image. The first encryption key and the initial image are combined to form a second encryption key. The second file is then encrypted using the second encryption key to obtain an encrypted transmission file. The encrypted transmission image and the encrypted transmission file are concatenated into encrypted transmission data, and the first encryption key and the second encryption key are embedded in the junction of the encrypted transmission image and the encrypted transmission file. Then, the encrypted transmission image is encrypted using the first encryption key and transmitted to the target end.

2. The file encryption method according to claim 1, characterized in that, The steps of preprocessing the raw speech information and extracting its features include: The original speech information is pre-emphasized; The pre-emphasized original speech information is subjected to frame segmentation and windowing to obtain the pre-processed original speech information; Performing a Fast Fourier Transform on the preprocessed original speech information yields the following amplitude spectrum of the preprocessed original speech information: Where U(k) is the amplitude spectrum of the preprocessed original speech information, T[n] is the preprocessed original speech information, and N is the window width of the window function when performing fast Fourier transform; The amplitude spectrum is moduloed and squared to obtain the discrete power spectrum of the preprocessed original speech information; the discrete power spectrum satisfies the following formula: Wherein, P(k) is the discrete power spectrum of the preprocessed original speech information; The discrete power spectrum is frequency integrated using a Gammatone filter bank; the time-domain impulse response of the Gammatone filter bank satisfies the following formula: Where c is the scaling factor, n is the order of the Gammatone filter, b is the time attenuation factor, and f0 is the center frequency of the Gammatone filter. The phase of the Gammatone filter; Calculate the long-time frame power of the preprocessed original speech information and mask and suppress noise other than human voice; the long-time frame power satisfies the following formula: Where Q(i,j) is the long-time frame power, and P[i′,j] is the power spectrum of the current frame and a certain frame among the i frames before and after it; The time and frequency domains of the original speech information after normalization preprocessing and noise masking; The power spectrum after time-frequency normalization is calculated using a nonlinear function power, and then dimensionality is reduced using discrete cosine transform to obtain voiceprint feature information.

3. The file encryption method according to claim 1, characterized in that, The steps of preprocessing the facial video dynamics and extracting its features include: Several image frames of a facial video are captured at a preset sampling time. High-pass filtering is applied to each image frame; The high-pass filtered image frames were denoised using median filtering to obtain several face images to be extracted. Construct a face detection model, perform face detection on the face image to be extracted using the face detection model, and output a face detection image; The face feature information is obtained by extracting face features from the face detection image using a multi-task cascaded convolutional neural network.

4. The file encryption method according to claim 1, characterized in that, The identity feature information is obtained by fusing the facial feature information and the voiceprint feature information, including: Acquire several facial feature information items, and collect the timestamps of the image frames corresponding to the facial feature information items respectively; Based on the timestamp, several facial feature information items are numbered in chronological order. According to the number, the facial feature information is divided into ten groups of facial feature information. The first and last groups of facial feature information are removed. Four groups of facial feature information are randomly selected from the remaining eight groups of facial feature information and their average value is calculated to obtain the facial feature information to be fused. The identity feature information is obtained by fusing the facial feature information to be fused and the voiceprint feature information.

5. A file encryption method according to claim 1, characterized in that, The process of processing the initial file according to byte importance and importance threshold to obtain the first file and the second file includes: Obtain the initial file and calculate the importance of each byte in the initial file; Record the sorting information of each byte and embed the sorting information into the first encryption key; the sorting information does not participate in the encryption operation. Reorder all bytes in the initial file according to the importance of each byte to obtain a sorted initial file; Based on the importance threshold, the initial files after sorting are divided into the first file and the second file.

6. The file encryption method according to claim 1, characterized in that, The process of encrypting the initial image using a chaotic iterative encryption algorithm to obtain a first encrypted image includes: Let the initial point of the iteration be the function value of the N0th iteration: The initial image is blocked according to l bytes to form a sub-message block P j = p lj , p lj+1 ,..., p (l+1)j-1 ; A binary random sequence of independent and identically distributed is generated and iterated with the chaotic map D after encryption in this round j seconds; The binary random sequence is as follows: Sub-message block P j Circular left shift D j Bit, sub-message block P j The sequence after the permutation transformation is denoted as P. j ′; For sequence P j Y and binary random sequence j Perform an XOR operation: in The XOR operator, C j For P j The corresponding ciphertext; C j Expanding by l bits yields the ciphertext c corresponding to the plaintext string. lj c lj +1, ..., c (l+1)j-1 That is, the first encrypted image.

7. A file encryption method according to claim 6, characterized in that, After generating the independent and identically distributed binary random sequence, the process includes: Let Y in the binary random sequence j 1 The decimal value is D j The number of iterative chaotic mappings, D, is calculated. j ; where D j ′ and iteration number D j The relationship satisfies: Among them, P lj+k k = 0, 2, ..., l-1 is the sub-message block P j The value of each byte.

8. A file encryption method according to claim 1, characterized in that, The step of scrambling pixels in the first encrypted image using an image scrambling algorithm to obtain an encrypted transmission image includes: A multidimensional Arnold transform model is constructed, and the pixel positions and pixel grayscale values ​​of the first encrypted image are simultaneously scrambled using the multidimensional Arnold transform model to obtain an encrypted transmission image. The multidimensional Arnold transform model is expressed as follows: Among them, b x b y and b z Let r be any positive integer. x r y and r z Let Q be any positive integer, and V be the height and width values ​​of the encrypted initial image, respectively. W is the maximum number of binary bits corresponding to the pixel value, and the number of bits is selected as 16.

9. A file encryption method according to claim 1, characterized in that, The process of fusing the first encryption key and the initial image to form a second encryption key, and then encrypting the second file using the second encryption key to obtain an encrypted transmission file, includes: The second encryption key is obtained by associating the first encryption key and the initial image using the SHA256 hash algorithm; The encrypted transmission file is obtained by encrypting the second file using a symmetric encryption algorithm and the second encryption key.

10. A file encryption system, characterized in that, The system includes: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a file encryption method as described in any one of claims 1 to 9.

Images