Face image encryption and decryption method, system and device based on target detection technology
Through the combination of object detection technology and ultra-chaotic Chen system, the face area is detected and encrypted, which solves the problem of large calculation volume and low efficiency of traditional encryption methods, and realizes efficient protection of facial information, which is suitable for a variety of practical application scenarios .
Patent Information
- Application Number
- CN202210490733.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-05-07
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-05-07
AI Technical Summary
The prior art is difficult to effectively protect the privacy of facial information, especially when the risk of facial image leakage increases, traditional encryption methods are large in computing and low in efficiency, making it difficult to meet the security needs of biometric information.
Using the face image encryption and decryption method based on object detection technology, the face area is detected through the MTCNN model, combined with the super chaotic Chen system to generate chaotic sequences, local encryption of the face area, reduce the calculation amount, improve the encryption efficiency, and encrypt the person The face area is fused with the non-face area and embedded in the ciphertext image.
It realizes efficient encryption of facial information, reduces the risk of facial data leakage, improves the efficiency of encryption algorithms, and is suitable for access control systems, identity verification and other scenarios, effectively protecting personal privacy.
Smart Images

Figure CN115102683B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of face privacy protection, and in particular to a face image encryption and decryption method, system and device based on target detection technology. Background Art
[0002] In recent years, biometric recognition technology has been widely used in various fields due to its uniqueness and stability. Biometrics include fingerprints, irises, faces and other inherent characteristics of each person. Among them, face information is widely used in daily scenarios such as mobile payment and smart access control due to its unique convenience.
[0003] With the widespread application of biometric technology, the security of biometric information has received great attention. If facial information is leaked, the encryption system bound to the face will be permanently cracked. With the development of the Internet and multimedia technology, people have become accustomed to transmitting photos on social networks, which increases the risk of facial image leakage. In order to solve the security problem of facial privacy leakage, the technical research on encryption protection of facial images is very important. Summary of the invention
[0004] In order to solve the technical problems existing in the prior art, the present invention provides a facial image encryption and decryption method and system based on target detection technology. The present invention detects and extracts the face area in the image by combining the target detection technology, and performs local encryption on the face area in the image by combining the hyperchaotic Chen system. While realizing the face privacy protection function, the calculation amount of the encryption process is reduced, and the efficiency of the encryption algorithm is improved.
[0005] The first object of the present invention is to provide a face image encryption and decryption method based on target detection technology.
[0006] The second object of the present invention is to provide a face image encryption and decryption system based on target detection technology.
[0007] A third object of the present invention is to provide a computer device.
[0008] The first object of the present invention can be achieved by adopting the following technical solutions:
[0009] A facial image encryption and decryption method based on target detection technology, the method comprising:
[0010] S1. Use the MTCNN model to detect the input plaintext image, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image;
[0011] S2, setting an initial key, calculating a first set of initial values of the hyperchaotic Chen system according to the initial key and the average gray value of the face image, and iterating the hyperchaotic Chen system to generate a first set of chaotic sequences;
[0012] S3, using the average gray value of the face image and the first group of chaotic sequences to perform calculations, updating and obtaining the second group of initial values of the hyperchaotic Chen system, and iterating the hyperchaotic Chen system to generate the second group of chaotic sequences;
[0013] S4, preprocessing the first group of chaotic sequences and the second group of chaotic sequences, converting them into integer random sequences and integer index sequences of preset lengths, and using the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face region image to obtain an encrypted face image;
[0014] S5. The encrypted face image and the non-face area in the plaintext image are merged into one image, and the face position coordinates are XOR-encrypted, and the encrypted face position coordinates are embedded into the merged image to obtain the final ciphertext image.
[0015] The second object of the invention can be achieved by adopting the following technical solutions:
[0016] The face image encryption and decryption system based on target detection technology includes:
[0017] The detection module is used to detect the input plaintext image using the MTCNN model, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image;
[0018] The first generation module is used to set an initial key, calculate a first group of initial values of the hyperchaotic Chen system according to the average gray value of the face image, and iterate the hyperchaotic Chen system to generate a first group of chaotic sequences;
[0019] The second generation module is used to use the average gray value of the face image and the first group of chaotic sequences to perform calculations, update and obtain the second group of initial values of the hyperchaotic Chen system, and iterate the hyperchaotic Chen system to generate the second group of chaotic sequences;
[0020] An encryption module is used to pre-process the first group of chaotic sequences and the second group of chaotic sequences, convert them into integer random sequences and integer index sequences of preset lengths, and use the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face area image to obtain an encrypted face image;
[0021] The fusion output module is used to fuse the encrypted face image and the non-face area in the plaintext image into one image, perform XOR encryption on the face position coordinates, embed the encrypted face position coordinates into the fused image, and obtain the final ciphertext image.
[0022] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0023] 1. The present invention adopts a local area encryption method. It first uses a target detection algorithm to detect the face area in the image, and then encrypts the face area, which reduces the computational complexity of the encryption process, improves the efficiency of the encryption algorithm, realizes the encryption of the face area in the image, and effectively reduces the risk of face data leakage.
[0024] 2. The present invention can be applied to practical scenarios such as access control systems, airport identity verification, and secure payment. It can effectively protect personal privacy and prevent facial data leakage, and has high practical application value. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.
[0026] Figure 1 This is a flow chart of implementing face image encryption based on target detection technology in Example 1 of the present invention;
[0027] Figure 2 This is a basic flow chart for realizing the generation of two-level plaintext associated keys in Embodiment 1 of the present invention;
[0028] Figure 3 This is a schematic diagram of the overall architecture of the face image encryption algorithm in Example 1 of the present invention;
[0029] Figure 4 It is a schematic diagram of the encryption effect of the face image encryption method in Example 1 of the present invention. DETAILED DESCRIPTION
[0030] The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is obvious that the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and the implementation of the present invention is not limited to this. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0031] Embodiment 1:
[0032] like Figure 1 As shown, the face image encryption and decryption algorithm based on target detection technology described in the present invention includes the following steps:
[0033] Step 1: Use the MTCNN model to detect the input plaintext image, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image.
[0034] Specifically, the plaintext image is input into the MTCNN face detection model trained by a high-performance GPU. The model operation obtains the face area position coordinates (x1, y1) and (x2, y2). The face area position coordinates include the upper left corner coordinates of the face area and the lower right corner coordinates of the face area; (x1, y1) represents the upper left corner coordinates of the face area, and (x2, y2) represents the lower right corner coordinates of the face area; the original image is cropped according to the obtained face position coordinates to obtain a face image P of size M×N, and the average grayscale value P of the face image is calculated ave .
[0035] First, the target detection algorithm is used to detect the face area in the image, and then the face area is encrypted, which reduces the amount of calculation in the encryption process and improves the efficiency of the encryption algorithm.
[0036] Step 2: Set the initial key, calculate the first set of initial values of the hyperchaotic Chen system according to the initial key and the average gray value of the face image, and iterate the hyperchaotic Chen system to generate the first set of chaotic sequences.
[0037] In this embodiment, the initial key is set to x0, y0, z0, h0, and the change of the four initial values is calculated according to the average gray value of the face image. ave The value of each decimal place in is represented by C i (i=1,2,L,16), calculate the changes of the four initial values of Δα, Δβ, Δθ, and Δγ:
[0038]
[0039] The initial key x0, y0, z0, h0 is calculated with the change of the four initial values Δα, Δβ, Δθ, Δγ to generate the first group of initial values x associated with the plaintext 01 ,y 01 ,z 01 ,h 01 The specific operation is:
[0040]
[0041] Among them, x0, y0, z0, h0 are the initial keys, and the initial values of Δα, Δβ, Δθ, Δγ are x 01 ,y 01 ,z 01 ,h 01 The amount of change.
[0042] Substitute the initial value into the hyperchaotic Chen system and iteratively generate the first chaotic sequence X1, Y1, Z1 and H1 with a length of 2MN+1. The iterative equation of the hyperchaotic Chen system is:
[0043]
[0044] Among them, m, q, p, n, r are the parameters of the system; i = 0, 1, 2, 3...; x, y, z, h are the state variables that generate the chaotic sequence.
[0045] Step 3: Use the average gray value of the face image and the first set of chaotic sequences to calculate, update and obtain the second set of initial values of the hyperchaotic Chen system, iterate the hyperchaotic Chen system again, and generate the second set of chaotic sequences.
[0046] In this embodiment, the elements X1(2MN+1), Y1(2MN+1), Z1(2MN+1) and H1(2MN+1) of the chaotic sequences X1, Y1, Z1 and H1 are used to calculate with the four variation values obtained in step 2 to update the second set of initial values x 02 ,y 02 ,z 02 ,h 02 :
[0047]
[0048] The second set of initial values x 02 ,y 02 ,z 02 ,h 02 Substitute it into the hyperchaotic Chen system and perform a second iteration on the hyperchaotic Chen system to generate four chaotic sequences X2, Y2, Z2 and H2 with a length of 2MN.
[0049] like Figure 2 As shown in the figure, the basic flow chart of two-level plaintext association key generation is as follows: first, the average gray value of the face image is calculated, and the initial key of the hyperchaotic Chen system is set. The first set of initial values of the hyperchaotic Chen system is obtained by calculating and updating the average gray value of the face image, and the hyperchaotic Chen system is iterated to generate a chaotic sequence. The second set of initial values is calculated again by the chaotic sequence and the average gray value of the face image, and the hyperchaotic Chen system is iterated again to generate a second set of chaotic sequences. Finally, the two sets of chaotic sequences are preprocessed to obtain eight integer random sequences.
[0050] Step 4: Preprocess the first and second chaotic sequences and convert them into integer random sequences and integer index sequences of preset lengths. Use the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face area image to obtain an encrypted face image.
[0051] In this embodiment, the first group of chaotic sequences and the second group of chaotic sequences are preprocessed and converted into integer random sequences and integer index sequences of preset lengths, where S1-S3, S5-S7 are integer random sequences, S4 and S8 are chaotic sequences, and S4 is used to generate the integer index sequence Ind S4 , S8 is used to generate the integer index sequence Ind M , integer index sequence Ind N and integer random sequence S pos , as follows:
[0052]
[0053]
[0054] Among them, round(x) is a function that rounds x to the nearest integer towards zero; mod(a, b) represents the modulo operation of a on b.
[0055] Sort S4 to get the integer index sequence Ind S4 :
[0056] Ind S4 =sort(S4)
[0057] Among them, Sort(S) is a function that obtains the sorting index of elements in S in descending order.
[0058] Get integer index sequence Ind through S8 M 、Ind N and integer random sequence S pos :
[0059]
[0060] Among them, M Used to scramble image rows, Ind N Used to shuffle the image columns, S pos Used to encrypt the face position coordinates.
[0061] Preferably, using an integer random sequence and an integer index sequence to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face region image to obtain an encrypted face image includes:
[0062] The face image is pixel-scrambled and decomposed into three channel images PR, PG and PB. Then PR, PG and PB are reshaped into one-dimensional vectors VR, VG and VB of 1×MN.
[0063] All pixels in the three one-dimensional vectors VR, VG, and VB are indexed according to the integer index sequence IndS4 Reorder the elements in the university:
[0064]
[0065] Where i=1,2,L,N,M; temp is a temporary variable.
[0066] VR, VG, and VB are reshaped into three matrices of size M×N to obtain scrambled channel images SR, SG, and SB.
[0067] The face image is diffused pixel by pixel. The process of pixel cyclic diffusion is divided into two rounds. In the first round, the three channel images of the face image are diffused separately using the integer random sequence S1-S3. The pixels of each channel are XOR-encrypted with the elements in S1-S3 from left to right and from top to bottom, and then XOR-operated with the pixel in the previous position. Specifically, the first pixel is XOR-encrypted using the following formula:
[0068]
[0069] in Represents a bitwise XOR operation. The remaining pixels are diffused iteratively using the following formula:
[0070]
[0071] Where i = 1, 2, L, M; j = 1, 2, L, N; n1 = 2, 3, L, MN; last i and last j Determined by the following formula:
[0072]
[0073] In the second round of diffusion, the chaotic sequences S5-S7 are used to diffuse the three channel images of the face image separately. The pixels of each channel image are XOR-encrypted with the chaotic random number in the order from bottom to top and from right to left, and then XOR-operated with the pixel at the next position. The second round of diffusion algorithm is specifically to iterate the following formula to diffuse and encrypt the pixels:
[0074]
[0075] Where i = M, M-1, L, 1; j = N, N-1, L, 1; n2 = 2, 3, L, MN; last i and last j Determined by the following formula:
[0076]
[0077] And assign the values of the elements of matrices SR, SG and SB to images DR, DG and DB.
[0078] To shuffle the rows and columns of the face image, first use the integer index sequence Ind M The image is scrambled, specifically:
[0079]
[0080] Where i=1,2,L,M; DR(i,:) represents the i-th row of DR.
[0081] Then use the integer index sequence Ind N Scramble the columns of an image:
[0082]
[0083] Where j=1,2,L,N; DR(:,j) represents the j-th column of DR.
[0084] Finally, the three scrambled channel images are fused to obtain the color encrypted face image.
[0085] like Figure 3 As shown in the figure, the overall architecture diagram of the face image encryption algorithm is shown. After the plaintext image is processed by the face detection algorithm, the face image, face position coordinates and non-face area image are obtained. The face image is encrypted by pixel scrambling, pixel diffusion and row and column scrambling to obtain the encrypted face image, and the face position coordinates are encrypted by XOR encryption to obtain the encrypted face position coordinates. The encrypted face image is fused with the non-face area in the plaintext image, and the encrypted face position coordinates are embedded in the fused image to obtain the ciphertext image.
[0086] Step 5: Merge the encrypted face image and the non-face area in the plaintext image into one image, XOR-encrypt the face position coordinates, and embed the encrypted face position coordinates into the fused image to obtain the final ciphertext image.
[0087] Use a random sequence S of integers pos The face position is encrypted by XOR, and the encrypted face position coordinates are obtained, that is, the encrypted watermark is obtained. The XOR encryption is performed by the following formula:
[0088]
[0089] The encrypted face image is fused with the non-face area in the plaintext image, and the encrypted face area is used to replace the original unencrypted face area, so that the encrypted face area and the non-face area are merged into one image; the encrypted face position coordinates are embedded into the first four pixels in the R channel of the fused image to hide the face position information and obtain the final ciphertext image. Figure 4 As shown, it is a schematic diagram of the face image encryption effect. The face area in the image is encrypted, and the non-face area retains the original information.
[0090] Embodiment 2:
[0091] This embodiment provides a face image encryption and decryption system based on target detection technology, which includes a detection module, a first generation module, a second generation module, an encryption module and a fusion output module. The specific functions of each module are as follows:
[0092] The detection module is used to detect the input plaintext image using the MTCNN model, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image;
[0093] The first generation module is used to set an initial key, calculate a first group of initial values of the hyperchaotic Chen system according to the average gray value of the face image, and iterate the hyperchaotic Chen system to generate a first group of chaotic sequences;
[0094] The second generation module is used to use the average gray value of the face image and the first group of chaotic sequences to perform calculations, update and obtain the second group of initial values of the hyperchaotic Chen system, and iterate the hyperchaotic Chen system to generate the second group of chaotic sequences;
[0095] An encryption module is used to pre-process the first group of chaotic sequences and the second group of chaotic sequences, convert them into integer random sequences and integer index sequences of preset lengths, and use the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face area image to obtain an encrypted face image;
[0096] The fusion output module is used to fuse the encrypted face image and the non-face area in the plaintext image into one image, perform XOR encryption on the face position coordinates, embed the encrypted face position coordinates into the fused image, and obtain the final ciphertext image.
[0097] Embodiment 3:
[0098] This embodiment provides a computer device, which can be a server, a computer, etc., which includes a processor, a memory, an input device, a display, and a network interface connected through a system bus. The processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. When the processor executes the computer program stored in the memory, the face image encryption and decryption method based on the target detection technology of the above-mentioned embodiment 1 is implemented as follows:
[0099] Use the MTCNN model to detect the input plaintext image, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image;
[0100] An initial key is set, and the first set of initial values of the hyperchaotic Chen system is calculated according to the average gray value of the face image. The hyperchaotic Chen system is iterated to generate the first set of chaotic sequences.
[0101] The average gray value of the face image is used to calculate with the first group of chaotic sequences, the second group of initial values of the hyperchaotic Chen system is updated, and the hyperchaotic Chen system is iterated to generate the second group of chaotic sequences;
[0102] Preprocessing the first group of chaotic sequences and the second group of chaotic sequences into integer random sequences and integer index sequences of preset lengths, using the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face region image to obtain an encrypted face image;
[0103] The encrypted face image and the non-face area in the plaintext image are fused into one image, and the face position coordinates are XOR-encrypted. The encrypted face position coordinates are embedded into the fused image to obtain the final ciphertext image.
[0104] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.
Claims
1. A face image encryption and decryption method based on target detection technology, characterized in that: The following steps are involved: S1. Use the MTCNN model to detect the input plaintext image, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image; the MTCNN model is an MTCNN face detection model trained by GPU. The plaintext image is input into the MTCNN face detection model to obtain the face area position coordinates, and the original plaintext image is cropped according to the obtained face area position coordinates to obtain a size of M X N face image, calculate the average gray value of the face image; The face area position coordinates include the upper left corner coordinates of the face area and the lower right corner coordinates of the face area; S2, setting an initial key, calculating a first set of initial values of the hyperchaotic Chen system according to the initial key and the average gray value of the face image, and iterating the hyperchaotic Chen system to generate a first set of chaotic sequences; The step S2 comprises: setting an initial key, calculating the variation of four initial values according to the average gray value of the face image; operating the initial key with the variation of the four initial values to generate a first group of initial values associated with plain text; substituting the first group of initial values into the hyperchaotic Chen system, iterating the hyperchaotic Chen system, and obtaining four chaotic sequences; The calculation formula of the first set of initial values is: in, is the first set of initial values, is the initial key, , , , The initial values are The amount of change; The iterative equation of the hyperchaotic Chen system is: Among them, m, q, p, n, r are the parameters of the system, , x, y, z, h are the state variables that generate the chaotic sequence; S3, using the average gray value of the face image and the first group of chaotic sequences to calculate, update and obtain the second group of initial values of the hyperchaotic Chen system, iterate the hyperchaotic Chen system to generate the second group of chaotic sequences; the calculation formula of the second group of initial values is: in, is the second set of initial values, , , , is the element of the chaotic sequence, , , , The initial values are The amount of change; S4, preprocessing the first group of chaotic sequences and the second group of chaotic sequences, converting them into integer random sequences and integer index sequences of preset lengths, and using the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face region image to obtain an encrypted face image; S5. The encrypted face image and the non-face area in the plaintext image are merged into one image, and the face position coordinates are XOR-encrypted, and the encrypted face position coordinates are embedded into the merged image to obtain the final ciphertext image.
2. The facial image encryption and decryption method based on target detection technology as claimed in claim 1, characterized in that: The step S3 includes: using the elements of the four chaotic sequences and the variation of the first set of initial values to calculate and update the second set of initial values; substituting the second set of initial values into the hyperchaotic Chen system, iterating the hyperchaotic Chen system for the second time and obtaining four chaotic sequences.
3. The facial image encryption and decryption method based on target detection technology as claimed in claim 1, characterized in that: The step S4 comprises: The face image is pixel-scrambled, the face image is decomposed into 3 channel images, and then the 3 channel images are reshaped into 3 one-dimensional vectors, all the pixels in the 3 one-dimensional vectors are reordered according to the elements in the integer index sequence, and the 3 one-dimensional vectors are reshaped into 3 vectors of size M X N , and obtain three scrambled channel images; Perform pixel diffusion on the face image, use an integer random sequence to perform the first round of diffusion on the three channel images of the face image, after the pixels of each channel image are XOR-encrypted with the elements in the integer random sequence from left to right and from top to bottom, they are XOR-operated with the pixels at the previous position; use a chaotic sequence to perform the second round of diffusion on the three channel images of the face image, after the pixels of each channel image are XOR-encrypted with the chaotic random number in the order from bottom to top and from right to left, they are XOR-operated with the pixels at the next position; The face image is scrambled in rows and columns. First, the face image is scrambled in rows using an integer index sequence, and then the face image is scrambled in columns using an integer index sequence. The three scrambled channel images are fused to obtain a color encrypted face image.
4. The facial image encryption and decryption method based on target detection technology as claimed in claim 1, characterized in that: The step S5 includes: fusing the encrypted face image with the non-face area in the plaintext image, replacing the original unencrypted face area with the encrypted face area, and merging the encrypted face area and the non-face area into a fused image; using an integer random sequence to perform XOR encryption on the face position to obtain encrypted face position coordinates, embedding the encrypted face position coordinates into the first four pixels in the R channel of the fused image, hiding the face position information to obtain the final ciphertext image.
5. The face image encryption and decryption system based on target detection technology is characterized by: The system comprises: The detection module is used to detect the input plaintext image using the MTCNN model, obtain the face area image and face position coordinates, and calculate the average grayscale value of the face image; the MTCNN model is an MTCNN face detection model trained by the GPU, the plaintext image is input into the MTCNN face detection model to obtain the face area position coordinates, and the original plaintext image is cropped according to the obtained face area position coordinates to obtain a size of M X N face image, calculate the average gray value of the face image; The face area position coordinates include the upper left corner coordinates of the face area and the lower right corner coordinates of the face area; The first generation module is used to set an initial key, calculate a first group of initial values of the hyperchaotic Chen system according to the initial key and the average gray value of the face image, and iterate the hyperchaotic Chen system to generate a first group of chaotic sequences; the step S2 includes: setting the initial key, calculating the variation of four initial values according to the average gray value of the face image; operating the initial key with the variation of the four initial values to generate a first group of initial values associated with plain text; substituting the first group of initial values into the hyperchaotic Chen system, iterating the hyperchaotic Chen system, and obtaining four chaotic sequences; The calculation formula of the first set of initial values is: in, is the first set of initial values, is the initial key, , , , The initial values are The amount of change; The iterative equation of the hyperchaotic Chen system is: Among them, m, q, p, n, r are the parameters of the system, , x, y, z, h are the state variables that generate the chaotic sequence; The second generation module is used to use the average gray value of the face image and the first group of chaotic sequences to perform calculations, update and obtain the second group of initial values of the hyperchaotic Chen system, and iterate the hyperchaotic Chen system to generate the second group of chaotic sequences; the calculation formula of the second group of initial values is: in, is the second set of initial values, , , , is the element of the chaotic sequence, , , , The initial values are The amount of change; An encryption module is used to pre-process the first group of chaotic sequences and the second group of chaotic sequences, convert them into integer random sequences and integer index sequences of preset lengths, and use the integer random sequences to perform pixel scrambling, pixel diffusion, and row and column scrambling on the face area image to obtain an encrypted face image; The fusion output module is used to fuse the encrypted face image and the non-face area in the plaintext image into one image, perform XOR encryption on the face position coordinates, embed the encrypted face position coordinates into the fused image, and obtain the final ciphertext image.
6. A computer device comprising a processor and a memory for storing a program executable by the processor, characterized in that When the processor executes the program stored in the memory, the face image encryption and decryption method based on target detection technology described in any one of claims 1 to 5 is implemented.
