A secure compression method for DNA transversion operations in remote sensing images

Abstract

A secure compression method for remote sensing images using DNA transversion operations relates to the field of remote sensing image compression and encryption technology. It addresses the security and low transmission efficiency issues inherent in existing remote sensing image encryption and transmission processes. The remote sensing image is first compressed, and then, while ensuring transmission efficiency, a DNA transversion operation is performed on the image to achieve secure transmission. This method uses a chaotic system to generate a scrambled sequence as the key for remote sensing image compression and encryption. The sparse matrix of the remote sensing image is then scrambled to further scramble the sparse matrix. A DNA transversion operation is then performed on the compressed image matrix. The chaotic system then diffuses the DNA matrix to obtain a scrambled image matrix. Finally, the DNA matrix is ​​decoded using the chaotic sequence generated by the chaotic system to obtain the final ciphertext image. This method achieves a balance between information security and transmission efficiency for remote sensing images.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image compression and encryption, and in particular to a remote sensing image DNA transversion operation security compression method. Background Art

[0002] In today's big data era, protecting multimedia data during internet transmission has become a hot topic of research, posing unprecedented challenges to data security. Remote sensing images play a crucial role in data transmission, and remote sensing image compression and encryption technology is an effective means of ensuring image security. In today's complex and diverse society, remote sensing images from certain specialized fields are more valuable than ordinary images. Because they contain more information than ordinary images, they need to be protected from attackers during transmission, particularly for remote environmental monitoring and remote sensing information protection. To this end, researchers have developed numerous techniques, and encryption technology is widely used to protect the content of remote sensing images. While encryption is considered an effective technique for image protection, the inherent characteristics of remote sensing images, such as large data volumes, high redundancy, and strong correlation between adjacent pixels, make traditional encryption methods inefficient. Consequently, algorithms for remote sensing image security protection have been widely developed, including methods such as fractional-order chaotic systems, quantum computing, and compressed sensing. Using these methods to encrypt and compress remote sensing images has become a research hotspot in recent years.

[0003] Remote sensing images contain a wealth of useful information, such as water resources, land resources, and ecological resources. Using these images, we can accurately identify the various areas requiring protection. This information can then be used to conduct production operations, atmospheric monitoring, and environmental protection. Given the strategic importance and absolute value of remote sensing images, many researchers have dedicated themselves to protecting them. In 2021, Manjit et al. proposed a seven-dimensional hyperchaotic mapping image encryption method that uses a designed high-dimensional chaotic system to protect various remote sensing image information. In 2016, Li et al. proposed an image encoding algorithm based on DNA sequences, the Lorenz system, and the SHA-2 hash algorithm. This algorithm protects image information by separating the image into its various RGB components. While existing image encryption technologies protect remote sensing image information to a certain extent, the compressed and encrypted transmission of remote sensing images still fails to strike a balance between security and compression efficiency.

[0004] While existing remote sensing image transmission methods meet the needs of general users, they still face numerous security and compression efficiency challenges for highly confidential remote sensing images. For example, Bonny et al. proposed a symmetric key encryption algorithm based on DNA encoding. However, due to its small key space, it is difficult to resist brute force attacks. In 2015, Tu Zhengwu et al. proposed a color image encryption algorithm based on DNA sequences. This algorithm uses DNA encoding within a plane, resulting in low dimensionality and complexity, a single encoding and decoding method, and weak anti-attack capabilities. Therefore, in order to strike a balance between remote sensing image security and compression efficiency, it is still necessary to continuously seek the best solutions to these existing problems. Summary of the Invention

[0005] The present invention solves the security problems and low transmission efficiency problems existing in the existing remote sensing image encryption transmission process, and provides a remote sensing image DNA transversion operation security compression method. The remote sensing image is first compressed, and then the DNA transversion operation is performed on the remote sensing image on the basis of ensuring the transmission efficiency of the remote sensing image to achieve secure transmission of the remote sensing image.

[0006] A remote sensing image DNA transversion operation secure compression method is implemented by the following steps:

[0007] Step 1: Select a remote sensing image of size M×N as the original image Image, and split the original image Image into B1, B2, B3, and B4 band images according to the band;

[0008] Step 2: Calculate the mean of any two bands as the initial values ​​X0 and Y0 of the chaotic system; use the initial key userkey1 set by the user to iterate the two-dimensional chaotic system to generate chaotic sequences X and Y;

[0009] Step 3: Sort the chaotic sequence X and chaotic sequence Y in ascending order to obtain the index matrix X index and index matrix Y index ;

[0010] Step 4: Compress the four bands B1, B2, B3, and B4 described in step 1 respectively to obtain four compression matrices I of size M×N / 4. s , and the compression matrix I s Merge into a compressed matrix I of size M×N cs ; The specific process is:

[0011] Step 4. Perform discrete cosine transform on the B1, B2, B3, and B4 band images obtained in step 1 in sequence to obtain the matrix I transformed from the spatial domain to the frequency domain. csk ;

[0012] Step 42: Matrix I described in step 41 csk Set the threshold T and delete the values ​​whose absolute value is less than T to obtain the sparse matrix I cskr ;

[0013] Step 4.3: Use the index matrix X generated in step 3 index For the sparse matrix I in step 42 cskr Perform scrambling operation to obtain scrambled sparse basis matrix I sp ;

[0014] Step 4: Take M×N / 4 elements from the chaotic sequence X generated in step 3 to form a measurement matrix Φ;

[0015] Step 45: The scrambled sparse basis matrix I obtained in step 43 is sp The compressed sensing operation is performed on the measurement matrix Φ obtained in step 4 to obtain the compressed matrix I with a quantized size of M×N / 4. s ;

[0016] Step 46: until all four band images are compressed, and the compression matrices of the four bands are combined into a combined compression matrix I of size M×N cs ;

[0017] Step 5: Combining the chaotic sequence Y generated in step 3 to compress the matrix I obtained in steps 4 and 6 cs Perform DNA encoding operation to obtain DNA chaotic matrix DNAY;

[0018] Step 6: Use the initial key userkey2 set by the user to iterate the two-dimensional chaotic system to obtain the chaotic sequence Z and the chaotic sequence W;

[0019] Step 7: The index matrix Y obtained in step 3 index Acts as a rule selector for DNA transversions;

[0020] Step 8: Using the chaotic sequence Y generated in step 3 to perform DNA encoding on the chaotic sequence Z obtained in step 6, a DNA chaotic matrix DNAZ is obtained; performing a transversion operation on the chaotic matrix DNAZ according to the DNA transversion rule selector obtained in step 7 to obtain a new chaotic matrix DNAZT;

[0021] Step 9: Perform an XOR operation on the DNA chaotic matrix DNAY obtained in step 5 and the DNA chaotic matrix DNAZT obtained in step 8 to obtain composite DNA chaotic matrices HyberDNA;

[0022] Step 10: DNA decoding is performed on the DNA matrix according to the W chaotic sequence obtained in step 6 to obtain a digital matrix, which is the final ciphertext image CImage.

[0023] The significant effect of the present invention is as follows: the present invention proposes a secure compression method based on a chaotic system and a DNA transversion operation of remote sensing images, in which a chaotic sequence is generated by the chaotic system as a key for compression and encryption of remote sensing images, and the sparse matrix of the remote sensing image is scrambled to achieve further scrambling of the sparse matrix, and then a DNA transversion operation is performed on the compressed image matrix, and then the chaotic system is used to diffuse the DNA matrix to obtain a scrambled image matrix, and finally the DNA matrix is ​​decoded according to the chaotic sequence generated by the chaotic system to obtain the final ciphertext image.

[0024] The compression method described in this invention first compresses remote sensing images, then performs DNA transversion on them while ensuring efficient transmission. This ensures secure transmission of these images while also improving their transmission efficiency. This method achieves a balance between information security and transmission efficiency for remote sensing images.

[0025] This method uses a chaotic system as a key generator. This chaotic system has a large key space and a wide parameter range, making it suitable for encrypting remote sensing images and ensuring their security. Band-by-band compression of remote sensing images ensures transmission efficiency while preserving the integrity of the transmitted data to the greatest extent possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 Flowchart of the encryption process in the secure compression method for DNA transversion operation of remote sensing images according to the present invention;

[0027] Figure 2 This is a diagram showing the encryption and decryption effects of the remote sensing image DNA transversion operation security compression method of the present invention;

[0028] Figure 3 This is a diagram of the DNA transversion operation rules of the present invention;

[0029] Figure 4 This is a rendering of the DNA transversion operation rules of the present invention. DETAILED DESCRIPTION

[0030] Specific implementation method 1. Combination Figures 1 to 3 This embodiment describes a secure compression method for DNA transversion operations in remote sensing images. The method uses the average pixel value of the plaintext image as input to obtain an initial value of a chaotic system and a chaotic sequence of the chaotic system. The method is implemented by the following steps:

[0031] Step 1: Select a remote sensing image of size M×N×3 as the original image;

[0032] Step 2: Split the original image Image described in step 1 into B1, B2, B3, and B4 band images according to the band:

[0033] Step 3: Calculate the mean of any two bands as the initial values ​​X0 and Y0 of the chaotic system;

[0034] Step 4: Use the initial key userkey1 set by the user to iterate the two-dimensional chaotic system to generate two chaotic sequences X and Y, and remove the first 1000 data to reduce the interference of the initial value;

[0035] Step 5: Sort the chaotic sequences X and Y in ascending order to obtain the index matrix X index and index matrix Y index ;

[0036] Step 6: Compress the four bands described in step 2 separately to obtain four compression matrices I of size M×N / 4 s (s=1,2,3,4) and merge them into a compressed image I of size M×N cs ;

[0037] Step 7: Perform discrete cosine transform on the band images B1, B2, B3, and B4 obtained in step 2 in sequence to transform them from the spatial domain to the frequency domain I csk (where csk = 1, 2, 3, 4)

[0038] Step 8: Set the threshold T for the transformed matrix in step 7, delete the values ​​whose absolute value is less than T, and obtain the sparse matrix I cskr .

[0039] Step 9. Use the index matrix X generated in step 5 index Perform scrambling operation on the sparse matrix of step eight to obtain the scrambled sparse basis matrix I sp .

[0040] Step 10: Take M×N / 4 elements from the chaotic sequence X generated in step 4 to form a measurement matrix Φ.

[0041] Step 11: Perform compressed sensing operations on the sparse matrices and measurement matrices generated in steps 9 and 10, and then quantize them to obtain a compressed matrix I of size M×N / 4. s .

[0042] Step 12: Execute the operations from step 7 to step 11 in sequence until all four band images are compressed. Then, combine the compression matrices of the four bands into a combined compression matrix I of size M×N. cs .

[0043] Step 13: Combining the chaotic sequence Y generated in step 4 to compress the matrix I obtained in step 12 cs Perform DNA encoding operation to obtain the DNA chaotic matrix DNAY.

[0044] Step 14: Use the initial key userkey2 set by the user which is different from that in step 4 to iterate the two-dimensional chaotic system and obtain two chaotic sequences Z and W.

[0045] Step 15: The index matrix Y obtained in step 5 index Acts as a rule selector for DNA transversions.

[0046] Step 16: DNA encode the chaotic sequence Z generated in step 14 according to the chaotic sequence Y generated in step 4 to obtain a DNA chaotic matrix DNAZ. Perform a transversion operation on the chaotic matrix DNAZ according to the DNA transversion rule obtained in step 15 to obtain a new chaotic matrix DNAZT.

[0047] Step 17: Perform an XOR operation on the DNA chaotic matrix DNAY generated in step 13 and the chaotic matrix DNAZT generated in step 16 to obtain a composite DNA chaotic matrix HyberDNA.

[0048] Step 18: DNA decoding is performed on the DNA matrix according to the W chaotic sequence generated in step 14 to obtain a digital matrix, which is also the final ciphertext image CImage.

[0049] Specific implementation method 2: Figures 1 to 4 This embodiment is an example of the method for securely compressing DNA transversion operations in remote sensing images described in Specific Embodiment 1. The method is implemented by the following steps:

[0050] Step 1: Select the color remote sensing image "city" with a size of 256×256×3 as the original image Image, as shown in the attached image. Figure 2 As shown, (a) represents the original image of the color remote sensing image "city", (b) represents the encrypted image of the color remote sensing image "city", and (c) represents the decrypted image of the color remote sensing image "city";

[0051] Step 2: Split the original image Image described in step 1 into four bands, namely B1, B2, B3, and B4, using the method shown in formula (1);

[0052]

[0053] Among them, the shape() function means dividing the original image Image into four bands; B1, B2, B3, and B4 represent the red, green, blue, and near-infrared bands respectively.

[0054] Step 3: Calculate the mean of any two bands as the initial values ​​X0 and Y0 of the chaotic system. The calculation method is shown in formula (2):

[0055]

[0056] Among them, mod() is the remainder function, sum() is the sum function, b1 and b2 are any two bands, and b1, b2∈{B1, B2, B3, B4}.

[0057] Step 4: Taking Henon mapping as an example, use the initial key userkey1 set by the user to iterate the Henon chaotic system and generate two chaotic sequences X and Y. The two-dimensional Henon chaotic system is shown in formula (3):

[0058]

[0059] Where X and Y are the state variables of the two-dimensional cubic iterative chaotic system; a and b are the parameters of the chaotic system. When a∈(1.07, 1.4) and b=0.3, a chaotic attractor exists in the Henon map. In this embodiment, the initial value of the chaotic system is obtained by formula (2). The first 1000 data points in the chaotic sequence are removed, and then a chaotic sequence of size M×N is selected.

[0060] Step 5: Sort the chaotic sequence in step 4 in ascending order to obtain the index matrix X index and index matrix Y index As shown in formula (4):

[0061]

[0062] Step 6: Compress the four bands described in step 2 separately to obtain four compression matrices I of size M×N / 4 s (s=1,2,3,4) and merge them into a compressed matrix I of size M×N cs The following steps describe the compression process in detail.

[0063] Step 7: Perform discrete cosine transform on the band images B1, B2, B3, and B4 obtained in step 2 to transform them from the spatial domain to the frequency domain matrix I csk (where csk = 1, 2, 3, 4).

[0064] I csk =dct(B)(5)

[0065] Among them, B represents each band of the remote sensing image, dct represents discrete cosine transform, I csk Represents the matrix transformed from the spatial domain to the frequency domain.

[0066] Step 8: The matrix I after the transformation in step 7 csk Set the threshold T and delete the values ​​whose absolute value is less than T to obtain the sparse matrix I cskr .

[0067]

[0068] Among them, T = 0.085, reshape means setting the data in the frequency domain to 256 × 256, abs(c1) means taking the absolute value of c1, find() function means returning the subscript of the data that meets the conditions, which means returning the subscript of all data with an absolute value less than 0.085, c1(thd) means setting the value with an absolute value less than 0.085 to 0, I cskr Represents the final sparse matrix after threshold filtering.

[0069] Step 9. Use the index matrix X generated in step 5 index Perform scrambling operation on the sparse matrix of step eight to obtain the scrambled sparse basis matrix I sp .

[0070] I sp =scramble(X index , I sp ) (7)

[0071] Among them, the scramble() function represents the sparse basis matrix I sp Follow X index Scramble the sequence.

[0072] Step 10: Take M×N / 4 elements from the chaotic sequence X generated in step 4 to form a measurement matrix Φ.

[0073]

[0074] res9ape() is a matrix transformation function, and res9ape(A, m, n) means returning an m×n matrix, in which the elements are obtained from A column by column.

[0075] Step 11: Perform compressed sensing operations on the sparse matrices and measurement matrices generated in steps 9 and 10, and then quantize them to obtain a compressed matrix I of size M×N / 4. s .

[0076] I s =Φ·I sp ·S (9)

[0077] Among them I sp represents the sparse basis matrix, Φ represents the measurement matrix, and S represents the sparse coefficient.

[0078] Step 12: Execute the operations from step 7 to step 11 in sequence until all four band images are compressed. Then, combine the compression matrices of the four bands into a combined compression matrix I with a size of 256*256. cs .

[0079] Step 13: Combining the chaotic sequence Y generated in step 4 to compress the matrix I obtained in step 12 cs Perform DNA encoding operation to obtain the DNA chaotic matrix DNAY.

[0080] DNAY=DNAENCODE(mod(Y,8)+1,I cs )……(10)

[0081] Among them, DNAENCODE() represents DNA encoding operation, DNAENCODE(Y, I cs ) means returning a DNA matrix encoded according to rule Y, mod(Y, 8)+1 represents the transformation rule of DNA. mod() means performing a modulo operation on the chaotic sequence Y.

[0082] Step 14: Use the initial key userkey2 set by the user which is different from that in step 4 to iterate the Henon chaotic system and obtain two chaotic sequences Z and W.

[0083] Step 15: The index matrix Y obtained in step 5 index As a rule selector for DNA transversion, DNA transversion rules such as Figure 3 As shown in the figure, there are four transversion rules for DNA transversion: AT can be transverted to TA or CG, and GC can be transverted to CG or TA.

[0084] Y trans =DNATrans(mod(Y index ,4)+1,DNAY)......(11)

[0085] Where DNATrans() represents the DNA transversion rule selection function, mod(Y index , 4)+1 means converting the index matrix into integers from 1 to 4, corresponding to the four DNA transversion rules. Figure 4 An example of a DNA transversion rule is shown.

[0086] Step 16: DNA encode the chaotic sequence Z generated in step 14 according to the chaotic sequence Y generated in step 4 to obtain the DNA chaotic matrix DNAZ. trans The chaotic matrix DNAZ is subjected to a transposition operation to obtain a new chaotic matrix DNAZT.

[0087] DNAZ=ENCODE(mod(Y,8)+1,Z) (12)

[0088] Among them, ENCODE() represents the DNA encoding operation, returning a DNA matrix after DNA encoding, mod(Y, 8)+1 represents the DNA transformation rule, and mod() represents the modulo operation of the chaotic sequence Y.

[0089] Step 17: Perform an XOR operation on the DNA chaotic matrix DNAY generated in step 13 and the DNA chaotic matrix DNAZT generated in step 16 to obtain a composite DNA chaotic matrix HyberDNA.

[0090]

[0091] Step 18: DNA decoding is performed on the DNA matrix according to the W chaotic sequence generated in step 14 to obtain a digital matrix, which is also the final ciphertext image CImage.

[0092] CImage=UnEncode(mod(W,8)+1,HyberDNA) (14)

[0093] Wherein, UnEncode() represents the DNA decoding operation, returns a digital matrix after DNA decoding, and mod(W, 8)+1 represents the transformation rule of DNA decoding.

[0094] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0095] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. A secure compression method for DNA transversion operations in remote sensing images, characterized by: The method is implemented by the following steps: Step 1: Select a remote sensing image of size M×N as the original image Image, and split the original image Image into B1, B2, B3, and B4 band images according to the band; Step 2: Calculate the mean of any two bands as the initial values ​​X0 and Y0 of the chaotic system; use the initial key userkey1 set by the user to iterate the two-dimensional chaotic system to generate chaotic sequences X and Y; Step 3: Sort the chaotic sequence X and chaotic sequence Y in ascending order to obtain the index matrix X index and index matrix Y index ; Step 4: Compress the four bands B1, B2, B3, and B4 described in step 1 respectively to obtain four compression matrices I of size M×N / 4. s , and the compression matrix I s Merge into a compressed matrix I of size M×N cs ; The specific process is: Step 4. Perform discrete cosine transform on the B1, B2, B3, and B4 band images obtained in step 1 in sequence to obtain the matrix I transformed from the spatial domain to the frequency domain. csk ; Step 42: Matrix I described in step 41 csk Set the threshold T and delete the values ​​whose absolute value is less than T to obtain the sparse matrix I cskr ; The sparse matrix I is calculated by the following formula cskr ; Where T = 0.085, reshape() sets the data in the frequency domain to 256 × 256, abs(c1) takes the absolute value of c1, find() returns the subscript of the data that meets the conditions, and c1(thd) sets the value with an absolute value less than 0.085 to 0. Step 4.3: Use the index matrix X generated in step 3 index For the sparse matrix I in step 42 cskr Perform scrambling operation to obtain scrambled sparse basis matrix I sp ; Step 4: Take M×N / 4 elements from the chaotic sequence X generated in step 3 to form a measurement matrix Φ; Step 45: The scrambled sparse basis matrix I obtained in step 43 is sp The compressed sensing operation is performed on the measurement matrix Φ obtained in step 4 to obtain the compressed matrix I with a quantized size of M×N / 4. s ; Step 46: Until all the images of bands B1, B2, B3, and B4 are compressed, the compression matrices of the four bands are combined into a combined compression matrix I of size M×N. cs ; Step 5: Combining the chaotic sequence Y generated in step 2 to compress the matrix I obtained in steps 4 and 6 cs Perform DNA encoding operation to obtain DNA chaotic matrix DNAY; Step 6: Use the initial key userkey2 set by the user to iterate the two-dimensional chaotic system to obtain the chaotic sequence Z and the chaotic sequence W; Step 7: The index matrix Y obtained in step 3 index Acts as a rule selector for DNA transversions; Step 8: Using the chaotic sequence Y generated in step 2 to perform DNA encoding on the chaotic sequence Z obtained in step 6, a DNA chaotic matrix DNAZ is obtained; performing a transversion operation on the chaotic matrix DNAZ according to the DNA transversion rule selector obtained in step 7 to obtain a new chaotic matrix DNAZT; Step 9: Perform an XOR operation on the DNA chaotic matrix DNAY obtained in step 5 and the DNA chaotic matrix DNAZT obtained in step 8 to obtain composite DNA chaotic matrices HyberDNA; Step 10: DNA decoding is performed on the chaotic matrix HyberDNA according to the chaotic sequence W obtained in step 6 to obtain a digital matrix, which is the final ciphertext image CImage.

2. The remote sensing image DNA transversion operation secure compression method according to claim 1, characterized in that: Perform discrete cosine transform on the band images B1, B2, B3, and B4 in step 1 to transform them from the spatial domain to the frequency domain matrix I csk ; I csk =dct(B) Where B is the band of the remote sensing image, and dct() is the discrete cosine transform function.

3. The remote sensing image DNA transversion operation secure compression method according to claim 1, characterized in that: In steps 4 and 5, the compression matrix I s It is expressed as follows: I s =Φ·I sp ·S Where, I sp is the sparse basis matrix, Φ is the measurement matrix, and S is the sparse coefficient.

4. The remote sensing image DNA transversion operation secure compression method according to claim 1, characterized in that: In step 5, the chaotic matrix DNAY is expressed as follows: DNAY=DNAENCODE(mod(Y,8)+1,I cs ) Where, DNAENCODE() is the DNA encoding operation, DNAENCODE(Y,I cs ) returns a DNA matrix encoded according to rule Y, mod(Y,8)+1 is the transformation rule of DNA; mod() is the modulo operation on the chaotic sequence Y.

Images