Image encryption method based on three-dimensional chaotic mapping and Josephh replacement

Abstract

The invention discloses an image encryption method based on three-dimensional chaotic mapping and Josephh replacement, which comprises the following steps of: taking an image as input of a hash function to obtain a hash value, updating a secret key by using the hash value, inputting the secret key into a three-dimensional chaotic system for iteration to obtain chaotic sequences X, Y and Z, and performing rotation operation on the image according to the value of the chaotic sequence X, then wavelet transformation is carried out on each image block, and multiple Josephh permutation of dynamic step length is carried out on the matrix after multi-scale decomposition according to the value of the chaos sequence Y; dynamically generating a mean value and a variance of Gaussian noise by using the value of the chaos sequence Z, and generating the Gaussian noise; and performing dynamic cross diffusion on the image block, controlling the diffusion direction and intensity through the chaos sequence Z, and adding the generated Gaussian noise into the image block. According to the method, the encryption process is more efficient and has better security, the anti-attack capability is improved, the visual security is better, privacy and integrity are protected, and the practical value is feasible.

Description

Technical Field

[0001] The present invention relates to the field of digital image encryption, and in particular to an image encryption method based on three-dimensional chaotic mapping and Joseph permutation in the field of information security technology. Background Art

[0002] With the continuous development of intelligent medical technology, the digitization and networking of medical images have made the transmission and storage of images more convenient. However, this also brings challenges to image privacy and data security. Medical images involve sensitive information of patients, such as disease diagnosis, condition analysis, etc. If they are tampered with or leaked without authorization, it will have a serious impact on patient privacy. Therefore, how to protect patient privacy while ensuring the safe transmission and storage of medical images has become a key issue waiting to be solved. Although traditional image encryption algorithms, such as AES and DES, can provide good security, their complex structure and high computational overhead make it difficult to meet the requirements of real-time and high efficiency in modern intelligent medical environments. In addition, these algorithms make insufficient use of the high redundancy and correlation of images. Therefore, more and more studies have begun to explore the combination of chaos theory, image permutation and other technologies to develop lighter and more efficient image encryption schemes.

[0003] Chaotic image encryption technology is a technology that uses the randomness and initial value sensitivity of chaotic systems to encrypt images. Chaotic systems have behaviors similar to random numbers and are extremely sensitive to initial conditions, so they are difficult to crack through reverse engineering. In image encryption, chaotic mapping (such as Logistic mapping or Henon mapping) is usually used to scramble the positions and grayscale values ​​of image pixels, making them unrecognizable in appearance. The advantages of this technology are that it is efficient, unpredictable, and can resist various cryptanalysis attacks to a certain extent.

[0004] Joseph permutation is an image encryption technique based on the Joseph problem, which arranges the image pixels in a circular manner in a certain order. The core idea of ​​this method is to use a fixed step size and starting position to delete elements in the image data according to a certain rule and rearrange the remaining pixels. This permutation process has a high degree of obfuscation effect, and combined with other encryption methods (such as chaotic systems or diffusion algorithms), it can further enhance the security of encrypted images. Summary of the invention

[0005] Purpose of the invention: In response to the problems raised in the background technology, the present invention proposes an image encryption method based on three-dimensional chaotic mapping and Joseph permutation, which generates a chaotic sequence by using a three-dimensional chaotic mapping combined with an Ernon mapping and a Chebyshev mapping, and combines the chaotic sequence to rotate the image, perform multiple Joseph permutations with dynamic steps, and add a cross-diffusion operation with Gaussian noise. This not only improves the security of image encryption, but also utilizes the high redundancy and structural characteristics of image data for efficient encryption, ensuring that while ensuring the security of image data, massive medical image data can be quickly processed to meet the needs of low latency and high throughput.

[0006] Technical solution: The purpose of the present invention is to provide an image encryption method based on three-dimensional chaotic mapping and Joseph permutation, comprising the following steps:

[0007] Step 1: Split the original image I of size M×N into image blocks I of size m×n ij , number of blocks

[0008] Step 2: For the original image, use the hash value obtained by the hash function SHA-256 to update the key, and iterate the three-dimensional chaotic system with the updated keys x1, y1 and z1 to obtain the chaotic sequences X, Y and Z;

[0009] Step 3: For image block I ij Perform a rotation operation and determine the rotation angle of the image block according to the value of the chaotic sequence X. If the corresponding chaotic sequence value X is less than 0.5, the image block I ij Rotate 90° clockwise; if the corresponding chaotic sequence value is 0.5≤X<1, then the image block I ij Rotate 90° counterclockwise, and the rotated image block is recorded as I′ ij ;

[0010] Step 4: Perform wavelet transform on each image block and decompose the image block into sub-bands of multiple scales. The four subband matrices are combined into a matrix P of size m×n ij ;

[0011] Step 5: Decompose the matrix P after multi-scale decomposition ij Perform multiple Joseph permutations with dynamic step sizes, use the sequence Y generated by chaotic mapping to control the step size of the Joseph ring, and reassemble the permuted blocks in the order of the original image;

[0012] Step 6: Generate Gaussian noise parameters with mean μ and variance σ using hash value 2 , generate Gaussian noise;

[0013] Step 7: Perform dynamic cross diffusion on the image block, control the direction and intensity of the diffusion through the chaotic sequence Z, and add the generated Gaussian noise to the image block.

[0014] Furthermore, the step 2 includes the following steps:

[0015] S1.1: The information of the original image is used as the input of the hash function SHA-256 to obtain a hash value H (h1, h2, ..., h 64 ), convert the hash value into a floating point number and divide it into H1 (h1, h2, ..., h 21 )、H2(h 22 ,h 23 , ..., h 42 ) and H3(h 43 ,h 44 , ..., h 64 ) three parts, the initial keys are x0, y0 and z0, and H1, H2 and H3 are used to update the initial keys to generate updated keys x1, y1 and z1:

[0016]

[0017] in is an XOR operation;

[0018] The mathematical equation of the three-dimensional chaotic map combining the Ernung map and the Chebyshev map is:

[0019]

[0020] The updated key is used as the initial condition of the chaotic system, and the system is iterated by sum(h1, h2, ..., h 64 )+MN times, generating three chaotic sequences X={x sum+1 , x sum+2 , ..., x sum+MN},Y={y sum+1 ,y sum+2 , ..., y sum+MN} and Z = {z sum+1 , z sum+2 , ..., z sum+MN}.

[0021] Furthermore, the step 4 includes the following steps:

[0022] Step S4.1: Select Haar wavelet function to perform wavelet transform, and perform low-frequency (average) and high-frequency (difference) operations on rows and columns of the image;

[0023] Step S4.2: First, perform a one-dimensional wavelet transform on each row of the image block to obtain a low-frequency component, i.e., an approximate coefficient, and a high-frequency component, i.e., a detail coefficient;

[0024] The formula for the averaging operation is:

[0025]

[0026] The formula for the difference operation is:

[0027]

[0028] in, Produces two lengths The vectors represent the low-frequency L and high-frequency H information respectively;

[0029] Step S4.3: Perform the same operation on the generated low-frequency and high-frequency vectors, that is, perform Haar wavelet transform on the columns. The formula is similar to that of the row transform. For the L vector: generate LL and LH subbands; for the H vector: generate HL and HH subbands;

[0030] The formula for the averaging operation is:

[0031]

[0032] The formula for the difference operation is:

[0033]

[0034] in

[0035] Step S4.4: After the Haar wavelet transform of rows and columns, four sub-bands are obtained: LL sub-band: represents the low-frequency information of the image; LH sub-band: captures the horizontal details of the image; HL sub-band: captures the vertical details of the image; HH sub-band: captures the diagonal details of the image. These sub-bands are large and small. Matrix of

[0036] Step S4.5: Merge the four subband matrices into a matrix P of size m×n ij , which contains all the information of the sub-band.

[0037] Furthermore, the step 5 includes the following steps:

[0038] Step S5.1: Matrix P ij Contains m×n pixels, and uses the sequence Y generated by chaotic mapping to control the step size of the Joseph ring. The formula for the step size S is as follows:

[0039]

[0040] in, is the corresponding P in the chaotic sequence ij The value range is between [0, 1], T ij It is a block P ij The total number of pixels in Indicates rounding up;

[0041] Step S5.2: For each block P ij Perform multiple Joseph permutations and set the initial positions of the pixels in the block to From Location start;

[0042] Step S5.3: Perform Joseph permutation on each block according to the following recursive formula to generate a new pixel arrangement:

[0043]

[0044] in, represents the pixel position in the block after the kth permutation, S ij Indicates the step size of the current block;

[0045] Step S5.4: Perform R rounds of Joseph permutation, using a different chaotic sequence control step length in each round. Suppose the step length of the rth round is:

[0046]

[0047] Each round of chaos sequence The value of is different, so the step length of each round Also different;

[0048] Step S5.5: For each image block P ij The above dynamic step-length Joseph permutation is performed independently, and each block has an independent chaotic order To control the step length;

[0049] Step S5.6: After completing the Joseph permutation of all image blocks, these permuted blocks are reassembled in the order of the original image.

[0050] Furthermore, the step 6 includes the following steps:

[0051] Step S6.1: Calculate the parameters of Gaussian noise using the hash value generated by the input key;

[0052] The mean calculation formula is: where μ max is the preset maximum mean;

[0053] Variance calculation formula: where σ 2 maxis the preset maximum mean;

[0054] Step S6.2: After obtaining the dynamically changing mean μ and variance σ 2 Then, the Gaussian noise is generated using the normal distribution formula:

[0055] G(μ,σ 2 )=μ+σ·W

[0056] in, W is a value randomly drawn from a standard normal distribution.

[0057] Furthermore, the step 7 includes the following steps:

[0058] Step S7.1: Control the direction and intensity of diffusion through the chaotic sequence Z. Assume that the initial value of a pixel in the image block is I(i, j), and the pixel value after diffusion is I′(i, j) calculated by the following formula:

[0059]

[0060] Where I(i-1, j) is the value of pixel (i-1, j), I(i, j-1) is the value of pixel (i, j-1), represents the XOR operation, q is the coefficient that controls the diffusion intensity, and Z n is the current value of the chaotic sequence, which is used to dynamically control the intensity of diffusion. G(μ, σ 2 ) is the added Gaussian noise;

[0061] Step S7.2: According to the chaotic sequence Z n The current value of is used to dynamically select the diffusion direction. The direction selection rules are as follows:

[0062] If Z n <0.25, the diffusion direction is diagonal diffusion; if 0.25≤Z n <0.5, the diffusion direction is horizontal diffusion; if 0.5≤Z n <0.5, then the diffusion direction is vertical diffusion; if Z n ≥0.75, the diffusion direction is counter-diagonal diffusion.

[0063] Beneficial effects: The present invention proposes an image encryption method based on three-dimensional chaotic mapping and Josephus permutation in an intelligent medical environment, which is suitable for intelligent medical treatment. The method uses the high sensitivity and pseudo-randomness of the three-dimensional chaotic system to generate complex and unpredictable encryption sequences. This feature makes the encryption process more secure and difficult to crack by traditional cryptanalysis methods. By confusing and diffusing the image data multiple times, simple statistical attacks and differential attacks can be avoided. The combination of Josephus permutation and diffusion makes the disturbance of the image pixel order and the data mixing effect more significant, thereby improving the anti-attack capability. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a schematic diagram of the overall structural flow of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the protection scope of the present invention.

[0066] The present invention provides an image encryption method based on three-dimensional chaotic mapping and Joseph permutation, which is suitable for intelligent medical treatment. Figure 1 , including the following steps:

[0067] S1: Use the hash value obtained by the hash function SHA-256 to update the key, and iterate the three-dimensional chaotic system with the updated keys x1, y1 and z1 to obtain the chaotic sequences x, Y and Z.

[0068] S1.1: Input the original plaintext image I of size M×N, use the original image information as the input of the hash function SHA-256, and obtain the hash value H (h1, h2, ..., h 64 ), convert the hash value into a floating point number and divide it into H1 (h1, h2, ..., h 21 )、H2(h 22 ,h 23 , ..., h 42 ) and H3(h 43 ,h 44 , ..., h 64 ) three parts. The initial keys are x0, y0 and z0, and H1, H2 and H3 are used to update the initial keys to generate updated keys x1, y1 and z1, as shown in formula (1):

[0069]

[0070] in It is an XOR operation.

[0071] The mathematical equation of the three-dimensional chaotic map combining the Ernung map and the Chebyshev map is:

[0072]

[0073] The updated key is used as the initial condition of the chaotic system, and the system is iterated by sum(h1, h2, ..., h 64 )+MN times, generating three chaotic sequences X={x sum+1 , x sum+2 , ..., x sum+MN},Y={y sum+1 ,y sum+2 , ..., y sum+MN} and Z = {z sum+1 , z sum+2 , ..., z sum+MN}.

[0074] S2: Split the original image of size M×N into image blocks of size m×n I ij , number of blocks

[0075] S3: Rotate the image block and determine the rotation angle of the image block according to the value of the chaotic sequence X. If the corresponding chaotic sequence value X is less than 0.5, the image block I ij Rotate 90° clockwise; if the corresponding chaotic sequence value is 0.5≤X<1, then the image block I ij Rotate 90° counterclockwise, and the rotated image block is recorded as I′ ij .

[0076] S4: Perform wavelet transform on each image block and decompose the image block into sub-bands of multiple scales. The four subband matrices are combined into a matrix P of size m×n ij .

[0077] Step S4.1: Select the Haar wavelet function to perform wavelet transform. The main steps of the Haar wavelet transform are to perform low-frequency (average) and high-frequency (difference) operations on the rows and columns of the image.

[0078] Step S4.2: First, perform a one-dimensional wavelet transform on each row of the image block to obtain a low-frequency component (approximate coefficient) and a high-frequency component (detail coefficient).

[0079] The formula for the averaging operation is:

[0080]

[0081] The formula for the difference operation is:

[0082]

[0083] in

[0084] The result is two lengths The vectors represent low-frequency (L) and high-frequency (H) information respectively.

[0085] Step S4.3: Perform the same operation on the previously generated low-frequency and high-frequency vectors, that is, perform Haar wavelet transform on the columns. The formula is similar to the row transform. For the L vector: generate LL and LH subbands; for the H vector: generate HL and HH subbands;

[0086] The formula for the averaging operation is:

[0087]

[0088] The formula for the difference operation is:

[0089]

[0090] in

[0091] Step S4.4: After the Haar wavelet transform of rows and columns, four sub-bands are obtained: LL sub-band: represents the low-frequency information of the image; LH sub-band: captures the horizontal details of the image; HL sub-band: captures the vertical details of the image; HH sub-band: captures the diagonal details of the image. These sub-bands are large and small. The matrix of .

[0092] Step S4.5: Merge the four subband matrices into a matrix P of size m×n ij , which contains all the information of the sub-band.

[0093] S5: Matrix P after multi-scale decomposition ij Multiple Joseph permutations with dynamic step sizes are performed, using different chaotic sequences to control the step sizes in each round, and the permuted blocks are reassembled in the order of the original image.

[0094] Step S5.1: Matrix P ij Contains m×n pixels, and uses the sequence Y generated by chaotic mapping to control the step size of the Joseph ring. The formula for the step size S is as follows:

[0095]

[0096] in is the corresponding P in the chaotic sequence ij The value range is between [0, 1], T ij It is a block P ij The total number of pixels in Indicates rounding up.

[0097] Step S5.2: For each block P ij Perform multiple Joseph permutations and set the initial positions of the pixels in the block to From Location start.

[0098] Step S5.3: Perform Joseph permutation on each block according to the following recursive formula to generate a new pixel arrangement:

[0099]

[0100] in represents the pixel position in the block after the kth permutation, S ij Indicates the stride of the current block.

[0101] Step S5.4: Perform R rounds of Joseph permutation, using a different chaotic sequence to control the step length in each round. Assume that the step length of the rth round is:

[0102]

[0103] Each round of chaos sequence The value of is different, so the step length of each round The replacement process is more complicated.

[0104] Step S5.5: For each image block P ij The above dynamic step-length Joseph permutation is performed independently, and each block has an independent chaotic order This dynamic step size control makes the pixel arrangement within each block highly unpredictable, further increasing the complexity of encryption.

[0105] Step S5.6: After completing the Joseph permutation of all image blocks, these permuted blocks are reassembled in the order of the original image. Since the pixel arrangement in each block has been disrupted, the combined image is an encrypted version with strong randomness and security.

[0106] S6: Generate Gaussian noise parameters mean σ and variance σ 2 , generate a Gaussian noise matrix for the image patch.

[0107] Step S6.1: Calculate the parameters of Gaussian noise using the hash value generated by the input key.

[0108] The mean calculation formula is: where μ max is the preset maximum mean. Variance calculation formula: where σ 2 maxis the preset maximum mean value.

[0109] Step S6.2: After obtaining the dynamically changing mean μ and variance σ 2 Then, the Gaussian noise is generated using the normal distribution formula:

[0110] G(μ,σ 2 )=μ+σ·W

[0111] in W is a value randomly drawn from a standard normal distribution.

[0112] S7: Perform dynamic cross diffusion on the image block, use the hash value to generate the mean and variance of Gaussian noise, and add the generated Gaussian noise to the image block.

[0113] Step S7.1: Control the direction and intensity of diffusion through the chaotic sequence Z, so that the diffusion process between pixels is highly random and unpredictable. Assuming that the initial value of a pixel in the image block is I(i, j), the pixel value after diffusion is I′(i, j) calculated by the following formula:

[0114]

[0115] Where I(i-1, j) is the value of pixel (i-1, j), I(i, j-1) is the value of pixel (i, j-1), represents the XOR operation, q is the coefficient that controls the diffusion intensity, and Z n is the current value of the chaotic sequence, which is used to dynamically control the intensity of diffusion. G(μ, σ 2 ) is the added Gaussian noise.

[0116] Step S7.2: According to the chaotic sequence Z n The current value of is used to dynamically select the diffusion direction. The direction selection rules are as follows:

[0117] If Z n <0.25, the diffusion direction is diagonal diffusion; if 0.25≤Z n <0.5, the diffusion direction is horizontal diffusion; if 0.5≤Z n <0.5, then the diffusion direction is vertical diffusion; if Z n ≥0.75, the diffusion direction is the anti-diagonal diffusion. In this way, different pixels will perform diffusion operations in different directions according to the current value of the chaotic sequence, and Gaussian noise will be injected into each pixel.

[0118] Image encryption methods based on three-dimensional chaotic mapping and Joseph permutation have broad application prospects in the medical industry. Medical images, such as X-rays, MRI scans, etc., contain a large amount of sensitive information about patients. The use of this encryption method can effectively protect these data from being accessed or stolen by unauthorized personnel. Electronic medical records contain not only text information, but also a large amount of image data, such as surgical photos, pathological sections, etc. By encrypting the image part of the medical record, the security of the entire electronic medical record can be ensured. Gaussian noise generated by chaotic sequences is added to the image blocks to further enhance the confidentiality of the image. In telemedicine, doctors need to share medical images of patients for diagnosis and discussion. This encryption method can protect the security of images during data transmission and prevent data from being intercepted or tampered with during transmission.

[0119] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. An image encryption method based on three-dimensional chaotic mapping and Joseph permutation, characterized in that: The following steps are involved: Step 1: Split the original image I of size M×N into image blocks I of size m×n ij , number of blocks Step 2: Use the hash function SHA-256 to update the key of the original image I, and iterate the three-dimensional chaotic system with the updated keys x1, y1 and z1 to obtain the chaotic sequences X, Y and Z; Step 3: For image block I ij The rotation operation is performed, and the rotation angle of the image block is determined according to the value of the chaotic sequence X. If the corresponding chaotic sequence value X<0.5, the image block I ij Rotate 90° clockwise; if the corresponding chaotic sequence value is 0.5≤X<1, then the image block I ij Rotate 90° counterclockwise, and the rotated image block is recorded as I' ij ; Step 4: Perform wavelet transform on each image block and decompose the image block into sub-bands of multiple scales. The four subband matrices are combined into a matrix P of size m×n ij ; Step 5: Decompose the matrix P after multi-scale decomposition ij Perform multiple Joseph permutations with dynamic step sizes, use the sequence Y generated by chaotic mapping to control the step size of the Joseph ring, and reassemble the permuted blocks in the order of the original image; Step 6: Generate Gaussian noise parameters with mean μ and variance σ using hash value 2 , generate Gaussian noise; Step 7: Perform dynamic cross diffusion on the image block, control the direction and intensity of the diffusion through the chaotic sequence Z, and add the generated Gaussian noise to the image block.

2. The image encryption method based on three-dimensional chaotic mapping and Joseph permutation according to claim 1 is characterized in that: The step 2 includes the following steps: S1.1: The information of the original image I is used as the input of the hash function SHA-256 to obtain a hash value H (h1, h2, ..., h 64 ), convert the hash value into a floating point number and divide it into H1 (h1, h2, ..., h 21 )、H2(h 22 ,h 23 , ..., h 42 ) and H3(h 43 ,h 44 , ..., h 64 ) three parts, the initial keys are x0, y0 and z0, and H1, H2 and H3 are used to update the initial keys to generate updated keys x1, y1 and z1: in is an XOR operation; The mathematical equation of the three-dimensional chaotic map combining the Ernung map and the Chebyshev map is: The updated key is used as the initial condition of the chaotic system, and the system is iterated by sum(h1, h2, ..., h 64 )+MN times, generating three chaotic sequences X={x sum+1 , x sum+2 , ..., x sum+MN },Y={y sum+1 ,y sum+2 , ..., y sum+MN } and Z = {z sum+1 , z sum+2 , ..., z sum+MN }.

3. The image encryption method based on three-dimensional chaotic mapping and Joseph permutation according to claim 1 is characterized in that: The step 4 includes the following steps: Step S4.1: Select Haar wavelet function to perform wavelet transform, and perform low-frequency (average) and high-frequency (difference) operations on rows and columns of the image; Step S4.2: First, perform a one-dimensional wavelet transform on each row of the image block to obtain a low-frequency component, i.e., an approximate coefficient, and a high-frequency component, i.e., a detail coefficient; The formula for the averaging operation is: The formula for the difference operation is: in, Produces two lengths The vectors represent the low-frequency L and high-frequency H information respectively; Step S4.3: Perform the same operation on the generated low-frequency and high-frequency vectors, that is, perform Haar wavelet transform on the columns. The formula is similar to that of the row transform. For the L vector: generate LL and LH subbands; for the H vector: generate HL and HH subbands; The formula for the averaging operation is: The formula for the difference operation is: in Step S4.4: After the Haar wavelet transform of rows and columns, four sub-bands are obtained: LL sub-band: represents the low-frequency information of the image; LH sub-band: captures the horizontal details of the image; HL sub-band: captures the vertical details of the image; HH sub-band: captures the diagonal details of the image. These sub-bands are large and small. Matrix of Step S4.5: Merge the four subband matrices into a matrix P of size m×n ij , which contains all the information of the sub-band.

4. The image encryption method based on three-dimensional chaotic mapping and Joseph permutation according to claim 1 is characterized in that: The step 5 includes the following steps: Step S5.1: Matrix P ij Contains m×n pixels, and uses the sequence Y generated by chaotic mapping to control the step size of the Joseph ring. The formula for the step size S is as follows: in, is the corresponding P in the chaotic sequence ij The value range is between [0, 1], T ij It is a block P ij The total number of pixels in Indicates rounding up; Step S5.2: For each block P ij Perform multiple Joseph permutations and set the initial positions of the pixels in the block to From Location start; Step S5.3: Perform Joseph permutation on each block according to the following recursive formula to generate a new pixel arrangement: in, represents the pixel position in the block after the kth permutation, S ij Indicates the step size of the current block; Step S5.4: Perform R rounds of Joseph permutation, using a different chaotic sequence control step length in each round. Suppose the step length of the rth round is: Each round of chaos sequence The value of is different, so the step length of each round It is also different; Step S5.5: For each image block P ij The above dynamic step-length Joseph permutation is performed independently, and each block has an independent chaotic order To control the step length; Step S5.6: After completing the Joseph permutation of all image blocks, these permuted blocks are reassembled in the order of the original image.

5. The image encryption method based on three-dimensional chaotic mapping and Joseph permutation according to claim 1 is characterized in that: The step 6 includes the following steps: Step S6.1: Calculate the parameters of Gaussian noise using the hash value generated by the input key; The mean calculation formula is: where μ max is the preset maximum mean; Variance calculation formula: where σ 2 max is the preset maximum mean; Step S6.2: After obtaining the dynamically changing mean μ and variance σ 2 Then, the Gaussian noise is generated using the normal distribution formula: G(μ,σ 2 )=μ+σ·W in, W is a value randomly drawn from a standard normal distribution.

6. The image encryption algorithm based on three-dimensional chaotic mapping and Joseph permutation applicable to intelligent medical treatment according to claim 1, characterized in that: The step 7 includes the following steps: Step S7.1: Control the direction and intensity of diffusion through the chaotic sequence Z. Assume that the initial value of a pixel in the image block is I(i, j), and the pixel value after diffusion is I'(i, j) calculated by the following formula: Where I(i-1, j) is the value of pixel (i-1, j), I(i, j-1) is the value of pixel (i, j-1), represents the XOR operation, q is the coefficient for controlling the diffusion intensity, Zn is the current value of the chaotic sequence, which is used to dynamically control the diffusion intensity, G(μ,σ 2 ) is the added Gaussian noise; Step S7.2: According to the chaotic sequence Z n The current value of is used to dynamically select the diffusion direction. The direction selection rules are as follows: If Z n <0.25, the diffusion direction is diagonal diffusion; if 0.25≤Z n <0.5, the diffusion direction is horizontal; if 0.5≤Z n <0.5, the diffusion direction is vertical; if Z n ≥0.75, the diffusion direction is counter-diagonal diffusion.

Images