Image encryption and decryption method and system based on quantum cell neural network

Abstract

The invention discloses an image encryption and decryption method based on a quantum cell neural network, and the method comprises the steps: setting network parameters of the quantum cell neural network, and generating a corresponding motion track; generating an encryption key; carrying out image encryption on the input image based on block cyclic shift and position scrambling; based on block cyclic shift and position regression, decryption operation of the encrypted image is carried out; and completing encryption and decryption of the target image based on the quantum cell neural network. The invention also discloses a system for realizing the image encryption and decryption method based on the quantum cell neural network. According to the method, the encryption and decryption process of the target image based on the quantum cell neural network is realized, the reliability is higher, the security is better, and the efficiency is higher.

Description

Technical Field

[0001] This invention belongs to the field of image processing, specifically relating to an image encryption and decryption method and system based on quantum cell neural networks. Background Technology

[0002] Quantum cellular neural networks are innovative computing models that integrate quantum computing and classical neural networks. They overcome the limitations of traditional neural networks that rely on classical bits for information processing, combining the core characteristics of quantum computing (quantum superposition, quantum entanglement, and quantum parallelism) with the core concepts of neural networks such as "hierarchical architecture" and "weight optimization," forming a completely new paradigm for information representation and processing. Due to their high efficiency and high security in high-dimensional data processing and computation, quantum cellular neural networks have experienced rapid development in recent years.

[0003] Image encryption is a technique that processes image data using specific technologies, transforming the original readable image into disordered, unrecognizable ciphertext to ensure that the image is not accessed or tampered with during transmission and storage. However, existing image encryption schemes often suffer from poor reliability and security when faced with today's brute-force attacks and statistical analysis.

[0004] Using quantum cellular neural networks for image encryption is one of the potential encryption schemes for the future. However, research in this area is currently lacking. Summary of the Invention

[0005] One of the objectives of this invention is to provide an image encryption and decryption method based on quantum cell neural networks that is highly reliable, secure, and efficient.

[0006] The second objective of this invention is to provide a system for implementing the aforementioned image encryption and decryption method based on quantum cellular neural networks.

[0007] The image encryption and decryption method based on quantum cellular neural networks provided by this invention includes the following steps:

[0008] Key generation:

[0009] S1. Set the network parameters of the quantum cell neural network and generate the corresponding motion trajectory;

[0010] S2. Generate an encryption key based on the motion trajectory obtained in step S1;

[0011] Image encryption:

[0012] S3. Based on the encryption key generated in step S2, perform image encryption on the input image using block cyclic shifting and position scrambling;

[0013] Image Decryption:

[0014] S4. Based on the encryption process in step S3, and using block cyclic shifting and position regression, perform the decryption operation on the encrypted image;

[0015] S5. Complete the encryption and decryption of the target image based on quantum cell neural network.

[0016] Step S1, which involves setting the network parameters of the quantum cell neural network and generating the corresponding motion trajectory, specifically includes the following steps:

[0017] Setting the network parameters of a quantum cell neural network, including the tunneling energy between quantum dots. The first parameter is proportional Tunneling energy between quantum dots The second parameter is proportional The first weighting coefficient reflecting the difference in polarization rate between adjacent cells The second weighting coefficient reflects the difference in polarization rate between adjacent cells. The polarizability of the first quantum cell unit The polarizability of the second quantum cell unit The phase angle of the quantum state of the first quantum cell unit Phase angle with the quantum state of the second quantum cell unit ; and Responsible for regulating tunnel strength, and Represents the state of electron distribution;

[0018] The generated motion trajectory is calculated using the following formula:

[0019] In the formula The trajectory of the first quantum cell unit; t is time; The trajectory of the second quantum cell unit; The trajectory of the first quantum cell unit's quantum state, defined by the phase angle. The trajectory of the second quantum cell unit's quantum state, representing the phase angle of its motion.

[0020] Step S2, which involves generating an encryption key based on the motion trajectory obtained in step S1, specifically includes the following steps:

[0021] The first key is calculated using the following formula:

[0022] In the formula The index is the integer part of the time, and M is the length of the input image, and N is the width of the input image; The index is the integer part of the time, and ; The first key to quantum cell neural networks; The second key for quantum cell neural networks; The third key to quantum cell neural networks; This is the fourth key for quantum cell neural networks; This is the fifth key to the quantum cell neural network;

[0023] The second key is calculated using the following formula:

[0024] In the formula for The corresponding second key, and The values ​​are 1, 2, 3, and 4; It is a modulo operation function, and The remainder after dividing aa by bb; express The absolute value is rounded down;

[0025] The first and second encryption keys are generated using the following formulas:

[0026] In the formula This is the first encryption key generated; The generated second encryption key; This is an XOR operation;

[0027] Set the parameters for the Henon mapping, including parameters that control the degree of nonlinear stretching and folding of the mapping. Parameters that control the degree of linear contraction of the mapping Henon mapping initial iteration value in the x-direction The initial iteration value of the Henon mapping in the y-direction a) determines the strength of chaotic behavior, b) affects the convergence characteristics of the iterative trajectory;

[0028] The first and second sequences are generated using the following formulas:

[0029] In the formula It is the first sequence; It is the second sequence; Let be the sequence number, and MN;

[0030] The first and second encryption sequences are generated using the following formulas:

[0031] In the formula This is the first encrypted sequence; This is the second encryption sequence; ;

[0032] Generate a third encryption key for .

[0033] Step S3, which involves encrypting the input image using the encryption key generated in step S2, based on block cyclic shifting and position scrambling, includes the following steps:

[0034] The matrix corresponding to the input image is divided into four equal sub-matrices;

[0035] According to the generated After shifting each submatrix to the left, they are then merged to obtain the first encryption matrix.

[0036] Expand the first encryption matrix into a one-dimensional vector, and use the first encryption key. The elements in the array are rearranged in a set order, and the positions of the one-dimensional vector expanded from the first encryption matrix are scrambled according to the rearranged sequence to obtain the first encryption vector.

[0037] Use the second encryption key The elements in the array are rearranged in a set order, and the positions of the first encryption vector are scrambled according to the rearranged sequence to obtain the second encryption vector.

[0038] The second encryption vector is XORed with the third encryption key, and the resulting vector is transformed back into a two-dimensional image to obtain the encrypted image, thus completing the image encryption of the input image.

[0039] Step S3 specifically includes the following steps:

[0040] The input image matrix A is divided into four equal-sized submatrices, and each submatrix is ​​expanded into a one-dimensional vector, represented as follows: , , and ;

[0041] According to the generated The following formula is used to shift each submatrix to the left:

[0042] In the formula for A one-dimensional vector shifted to the left; Indicates a left shift operation; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shift; (It deals with grayscale images, where pixel values ​​are between 0 and 255, which can be represented by 8-bit binary numbers. The left shift operation is to represent the image's pixel values ​​using 8-bit binary numbers and then shift them cyclically to the left.)

[0043] Will , , and Reconstruct the submatrices and merge them to obtain the first encryption matrix. ;

[0044] The first encryption matrix Expand into a one-dimensional vector ; the first encryption key The elements in the record are rearranged in ascending order. In the rearranged sequence, each element is in The first rearrangement sequence is obtained by taking the corresponding position in the sequence. ;(like , After rearrangement, it becomes ,but The first element 2 in B represents the rearranged element 1 in... (The position in the middle is 2); the position scrambling operation is performed using the following formula to obtain the first encryption vector. :

[0045] In the formula The first encryption vector The jj-th element; Represents a one-dimensional vector The serial number is Element;

[0046] Use the second encryption key The elements in the record are rearranged in descending order. In the rearranged sequence, each element is in The corresponding positions in the sequence yield the second rearrangement sequence. The second encryption vector is obtained by performing a position scrambling operation using the following formula. :

[0047] In the formula For the second encryption vector The k-th element; Represents the first encryption vector The serial number is Element;

[0048] The second encryption vector is generated using the following formula. With the third encryption key Perform an XOR operation:

[0049] In the formula The one-dimensional vector obtained by the XOR operation The first in One element, For the third encryption key The first in One element;

[0050] a one-dimensional vector The image is then transformed back into a two-dimensional image to obtain the encrypted image.

[0051] Complete the image encryption of the input image.

[0052] The decryption operation of the encrypted image described in step S4, based on the encryption process in step S3 and using block cyclic shifting and position regression, includes the following steps:

[0053] The encrypted image obtained in step S3 is converted into a one-dimensional vector, and an XOR operation is performed to obtain the first decryption vector;

[0054] According to the second encryption key The rearrangement rules are used to perform position regression on the first decryption vector to obtain the second decryption vector;

[0055] According to the first encryption key The rearrangement rules are used to perform another position regression on the second decryption vector to obtain the third decryption vector;

[0056] The third decryption vector is restored to a two-dimensional image, and the corresponding image matrix is ​​obtained;

[0057] According to the generated The image matrix is ​​shifted to the right to obtain the final decrypted image;

[0058] Complete the decryption operation of the encrypted image.

[0059] Step S4 specifically includes the following steps:

[0060] The encrypted image obtained in step S3 is converted into a one-dimensional vector G; the first decryption vector is obtained by performing an XOR operation using the following formula. :

[0061] In the formula The first decryption vector The mm-th element in;

[0062] According to the second rearrangement sequence The second decryption vector is obtained by performing position regression on the first decryption vector using the following formula. :

[0063] In the formula For the second decryption vector The nth element in; Indicates the second decryption vector The serial number is Element;

[0064] According to the first rearrangement sequence The third decryption vector is obtained by performing position regression on the second decryption vector using the following formula. :

[0065] In the formula The third decryption vector The ooth element in; Represents the third decryption vector The serial number is Element;

[0066] The third decryption vector The image is restored to a two-dimensional image, divided into four equal sub-matrices, and each sub-matrix is ​​expanded into a one-dimensional vector, represented as follows: , , and ;

[0067] According to the generated The following formula is used to calculate... , , and Perform a right shift operation:

[0068] In the formula for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; This is a right shift operation;

[0069] Will , , and The submatrices are then reconstructed and combined to obtain the final decrypted image;

[0070] Complete the decryption operation of the encrypted image.

[0071] This invention also provides a system for implementing the image encryption and decryption method based on quantum cell neural networks, comprising a parameter setting module, a key generation module, an image encryption module, an image decryption module, and an encryption / decryption module; the parameter setting module, key generation module, image encryption module, image decryption module, and encryption / decryption module are connected in series; the parameter setting module is used to set the network parameters of the quantum cell neural network, generate the corresponding motion trajectory, and upload the data information to the key generation module; the key generation module is used to generate an encryption key based on the received data information and the obtained motion trajectory, and upload the data information to the image encryption module; the image encryption module is used to encrypt the input image based on the received data information and the generated encryption key, using block cyclic shift and position scrambling, and upload the data information to the image decryption module; the image decryption module is used to decrypt the encrypted image based on the received data information and the encryption process, using block cyclic shift and position regression, and upload the data information to the encryption / decryption module; the encryption / decryption module is used to complete the encryption and decryption of the target image based on the quantum cell neural network based on the received data information.

[0072] The image encryption and decryption method and system based on quantum cell neural networks provided by this invention generates encryption keys through quantum cell neural networks, and realizes the encryption and decryption process of target images based on quantum cell neural networks based on block cyclic shift, position scrambling and position regression, with higher reliability, better security and higher efficiency. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0074] Figure 2 This is a schematic diagram comparing the encryption effects of embodiments of the method of the present invention.

[0075] Figure 3 This is a schematic diagram of histogram analysis in an embodiment of the method of the present invention.

[0076] Figure 4 This is a schematic diagram of correlation analysis in an embodiment of the method of the present invention.

[0077] Figure 5 This is a schematic diagram of the functional modules of the system of the present invention. Detailed Implementation

[0078] like Figure 1 The diagram shown is a flowchart of the method of the present invention: The image encryption and decryption method based on quantum cellular neural networks disclosed in this invention includes the following steps:

[0079] Key generation:

[0080] S1. Set the network parameters of the quantum cell neural network and generate the corresponding motion trajectory; specifically including the following steps:

[0081] Setting the network parameters of a quantum cell neural network, including the tunneling energy between quantum dots. The first parameter is proportional Tunneling energy between quantum dots The second parameter is proportional The first weighting coefficient reflecting the difference in polarization rate between adjacent cells The second weighting coefficient reflects the difference in polarization rate between adjacent cells. The polarizability of the first quantum cell unit The polarizability of the second quantum cell unit The phase angle of the quantum state of the first quantum cell unit Phase angle with the quantum state of the second quantum cell unit ; and Responsible for regulating tunnel strength, and Represents the state of electron distribution;

[0082] The generated motion trajectory is calculated using the following formula:

[0083] In the formula The trajectory of the first quantum cell unit; t is time; The trajectory of the second quantum cell unit; The trajectory of the first quantum cell unit's quantum state, defined by the phase angle. The trajectory of the second quantum cell unit's quantum state, defined by the phase angle.

[0084] S2. Generate an encryption key based on the motion trajectory obtained in step S1; specifically including the following steps:

[0085] The first key is calculated using the following formula:

[0086] In the formula The index is the integer part of the time, and M is the length of the input image, and N is the width of the input image; The index is the integer part of the time, and ; The first key to quantum cell neural networks; The second key for quantum cell neural networks; The third key to quantum cell neural networks; This is the fourth key for quantum cell neural networks; This is the fifth key to the quantum cell neural network;

[0087] The second key is calculated using the following formula:

[0088] In the formula for The corresponding second key, and The values ​​are 1, 2, 3, and 4; It is a modulo operation function, and The remainder after dividing aa by bb; express The absolute value is rounded down;

[0089] The first and second encryption keys are generated using the following formulas:

[0090] In the formula This is the first encryption key generated; The generated second encryption key; This is an XOR operation;

[0091] Set the parameters for the Henon mapping, including parameters that control the degree of nonlinear stretching and folding of the mapping. Parameters that control the degree of linear contraction of the mapping Henon mapping initial iteration value in the x-direction The initial iteration value of the Henon mapping in the y-direction a) determines the strength of chaotic behavior, b) affects the convergence characteristics of the iterative trajectory;

[0092] The first and second sequences are generated using the following formulas:

[0093] In the formula It is the first sequence; It is the second sequence; Let be the sequence number, and MN;

[0094] The first and second encryption sequences are generated using the following formulas:

[0095] In the formula This is the first encrypted sequence; This is the second encryption sequence; ;

[0096] Generate a third encryption key for ;

[0097] Image encryption:

[0098] S3. Based on the encryption key generated in step S2, perform image encryption on the input image using block cyclic shifting and position scrambling; including the following steps:

[0099] The matrix corresponding to the input image is divided into four equal sub-matrices;

[0100] According to the generated After shifting each submatrix to the left, they are then merged to obtain the first encryption matrix.

[0101] Expand the first encryption matrix into a one-dimensional vector, and use the first encryption key. The elements in the array are rearranged in a set order, and the positions of the one-dimensional vector expanded from the first encryption matrix are scrambled according to the rearranged sequence to obtain the first encryption vector.

[0102] Use the second encryption key The elements in the array are rearranged in a set order, and the positions of the first encryption vector are scrambled according to the rearranged sequence to obtain the second encryption vector.

[0103] The second encryption vector is XORed with the third encryption key, and the resulting vector is transformed back into a two-dimensional image to obtain the encrypted image, thus completing the image encryption of the input image.

[0104] In practice, the following steps can be taken:

[0105] The input image matrix A is divided into four equal-sized submatrices, and each submatrix is ​​expanded into a one-dimensional vector, represented as follows: , , and ;

[0106] According to the generated The following formula is used to shift each submatrix to the left:

[0107] In the formula for A one-dimensional vector shifted to the left; Indicates a left shift operation; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shift; (It deals with grayscale images, where pixel values ​​are between 0 and 255, which can be represented by 8-bit binary numbers. The left shift operation is to represent the image's pixel values ​​using 8-bit binary numbers and then shift them cyclically to the left.)

[0108] Will , , and Reconstruct the submatrices and merge them to obtain the first encryption matrix. ;

[0109] The first encryption matrix Expand into a one-dimensional vector ; the first encryption key The elements in the record are rearranged in ascending order. In the rearranged sequence, each element is in The first rearrangement sequence is obtained by taking the corresponding position in the sequence. ;(like , After rearrangement, it becomes ,but The first element 2 in B represents the rearranged element 1 in... (The position in the middle is 2); the position scrambling operation is performed using the following formula to obtain the first encryption vector. :

[0110] In the formula The first encryption vector The jj-th element; Represents a one-dimensional vector The serial number is Element;

[0111] Use the second encryption key The elements in the record are rearranged in descending order. In the rearranged sequence, each element is in The corresponding positions in the sequence yield the second rearrangement sequence. The second encryption vector is obtained by performing a position scrambling operation using the following formula. :

[0112] In the formula For the second encryption vector The k-th element; Represents the first encryption vector The serial number is Element;

[0113] The second encryption vector is generated using the following formula. With the third encryption key Perform an XOR operation:

[0114] In the formula The one-dimensional vector obtained by the XOR operation The first in One element, For the third encryption key The first in One element;

[0115] a one-dimensional vector The image is then transformed back into a two-dimensional image to obtain the encrypted image.

[0116] Complete the image encryption of the input image;

[0117] Image Decryption:

[0118] S4. Based on the encryption process in step S3, and using block cyclic shifting and position regression, perform the decryption operation on the encrypted image; including the following steps:

[0119] The encrypted image obtained in step S3 is converted into a one-dimensional vector, and an XOR operation is performed to obtain the first decryption vector;

[0120] According to the second encryption key The rearrangement rules are used to perform position regression on the first decryption vector to obtain the second decryption vector;

[0121] According to the first encryption key The rearrangement rules are used to perform another position regression on the second decryption vector to obtain the third decryption vector;

[0122] The third decryption vector is restored to a two-dimensional image, and the corresponding image matrix is ​​obtained;

[0123] According to the generated The image matrix is ​​shifted to the right to obtain the final decrypted image;

[0124] Complete the decryption operation of the encrypted image.

[0125] In practice, the following steps can be taken:

[0126] The encrypted image obtained in step S3 is converted into a one-dimensional vector G; the first decryption vector is obtained by performing an XOR operation using the following formula. :

[0127] In the formula The first decryption vector The mm-th element in;

[0128] According to the second rearrangement sequence The second decryption vector is obtained by performing position regression on the first decryption vector using the following formula. :

[0129] In the formula For the second decryption vector The nth element in; Indicates the second decryption vector The serial number is Element;

[0130] According to the first rearrangement sequence The third decryption vector is obtained by performing position regression on the second decryption vector using the following formula. :

[0131] In the formula The third decryption vector The ooth element in; Represents the third decryption vector The serial number is Element;

[0132] The third decryption vector The image is restored to a two-dimensional image, divided into four equal sub-matrices, and each sub-matrix is ​​expanded into a one-dimensional vector, represented as follows: , , and ;

[0133] According to the generated The following formula is used to calculate... , , and Perform a right shift operation:

[0134] In the formula for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; This is a right shift operation; (It deals with grayscale images, where pixel values ​​are between 0 and 255, which can be represented by 8-bit binary numbers. The right shift operation represents the image's pixel values ​​using 8-bit binary numbers and then shifts them cyclically to the right. We were encrypting to the left just now, and now we are decrypting to the right.)

[0135] Will , , and The submatrices are then reconstructed and combined to obtain the final decrypted image;

[0136] Complete the decryption operation of the encrypted image;

[0137] S5. Complete the encryption and decryption of the target image based on quantum cell neural network.

[0138] The advantages of the method of the present invention are illustrated below with reference to an embodiment:

[0139] To clearly and rigorously illustrate the superior performance of the quantum cell neural network and Henon mapping fusion encryption scheme proposed in this invention, the following explanation will be provided from three aspects: comparison with the traditional Henon mapping (HM), comparison with the pure quantum cell neural network (QCell) encryption scheme, and comprehensive experimental verification of this invention.

[0140] 1. Comparison with traditional Henon mapping (HM) encryption schemes

[0141] Analysis of theoretical deficiencies:

[0142] Traditional Henon mappings, as a type of two-dimensional discrete chaotic system, suffer from two inherent bottlenecks in image encryption applications, which are widely recognized in existing research. These bottlenecks are as follows:

[0143] The key space is finite: its key typically consists of only two control parameters (a, b) and two initial values ​​(x0, y0). Even considering computational precision, the upper limit of its key space is difficult to meet the minimum requirements of modern cryptography for resisting brute-force attacks (generally considered to be greater than...). ).

[0144] Chaotic behavior degradation: Within a certain parameter range, HM may exhibit periodic windows or weak chaotic behavior, resulting in insufficient randomness and limited complexity of the generated sequences. This "chaotic degradation" phenomenon reduces the sensitivity of the encryption system, making it more vulnerable to statistical analysis and chosen-plaintext attacks.

[0145] Improvements brought by the present invention:

[0146] This invention does not simply use HM, but introduces quantum cell neural network (QCell) as a high-performance hyperchaotic entropy source to make up for the above-mentioned defects of HM.

[0147] Revolutionary expansion of the key space: In the scheme of this invention, the private key of the encryption system is composed of 8 parameters of QCell ( , , , , , , and ) and the two initial values ​​of HM ( , Together they constitute the key space. During key space analysis, in... At an effective precision, its total key space is as high as approximately This far exceeds The security threshold fundamentally solves the problem of insufficient HM key space.

[0148] Significant Improvement in Chaotic Quality and Randomness: QCell, as a model integrating quantum computing and nonlinear dynamics, has been proven to possess hyperchaotic properties (its maximum Lyapunov exponent is always greater than zero), and the sequences it generates exhibit extremely high initial value sensitivity and long-term unpredictability. This invention utilizes the hyperchaotic sequences generated by QCell ( , , , and This method drives the encryption process, completely avoiding the periodic behavior that HM might exhibit and ensuring the strong randomness of the encryption key.

[0149] Visual comparison of results:

[0150] Encryption effect such as Figure 2 As shown, comparing the original image and the encrypted image reveals that the image encrypted using the method of this invention is completely presented as uniform noise, with no residual visual features. However, if only traditional HM is used, due to chaotic degradation, the encrypted image may still statistically retain some texture information.

[0151] Histogram analysis:

[0152] Histogram pairs, for example Figure 3 As shown ( Figure 3 (a) is the histogram of the original image. Figure 3 (b) Histogram of the encrypted image: The grayscale histogram of the encrypted image obtained by the present invention has an extremely flat and uniform distribution, proving that the pixel values ​​are sufficiently randomized. In contrast, the histogram of a traditional HM encrypted image may not achieve such an ideal uniform distribution, thus leaving an opportunity for statistical attacks.

[0153] Correlation analysis:

[0154] Correlation analysis such as Figure 4 As shown ( Figure 4 (a) is the diagonal method. Figure 4 (b) represents the horizontal direction. Figure 4 (c) Vertical direction: The image shows that the image encrypted by the scheme of this invention has extremely low correlation coefficients between adjacent pixels in the horizontal, vertical, and diagonal directions, and the pixels are completely discretely distributed. This is due to the excellent decorrelation ability of QCell hyperchaotic sequences, while traditional HM usually performs poorly in this respect (its correlation coefficient is often significantly greater than 0).

[0155] Comparison with pure quantum cell neural network (QCell) encryption scheme:

[0156] Potential limitations analysis:

[0157] Quantum Cell Neural Networks (QCells) are powerful hyperchaotic signal generators, but a complete image encryption scheme requires not only high-quality key sources but also an efficient and secure encryption architecture (i.e., diffusion and obfuscation mechanisms) to fully utilize these keys. Directly using QCell sequences for simple permutation or XOR operations may not achieve optimal diffusion and attack robustness in image encryption.

[0158] The innovative architecture and integration advantages of this invention:

[0159] This invention creatively designs a hybrid encryption process that combines the advantages of the HM encryption framework with the chaos advantages of QCell.

[0160] The inherited and enhanced encryption framework: This invention employs multi-stage operations including block cyclic shifting, sort-based dual position scrambling, and XOR diffusion. This framework draws on the design principles of mature chaotic image encryption, ensuring that pixel positions and value ranges are sufficiently and rapidly obfuscated and diffused.

[0161] QCell-driven high-intensity execution: Each of the above operations (such as shift parameter lᵢ, scrambled index sequences B and E, and XOR key K) is dynamically determined by a hyperchaotic sequence generated by QCell. This means that a highly complex chaotic engine is perfectly embedded in a structured encryption process.

[0162] Overall performance improvement: This fusion of "high-quality entropy source + efficient architecture" enables the present invention to inherit the high randomness of QCell while gaining better resistance to pruning and noise attacks than simply using QCell sequences.

[0163] Summary of experimental performance of the present invention:

[0164] Based on the MATLAB platform The key performance indicators of the proposed solution for grayscale images are shown in Table 1 below. These data comprehensively demonstrate its high reliability, high security, and high efficiency.

[0165] In summary, this invention is not a simple replacement or superposition of HM or QCell, but rather a novel encryption system built through deep structural fusion and functional complementarity. It leverages the hyperchaotic properties of QCell to overcome the key space and chaotic degradation problems of HM, while fully utilizing the mature architecture of HM-like encryption to maximize the potential of QCell sequences, ultimately achieving an optimal balance between key space, encryption strength, attack resistance, and execution efficiency. Experimental data fully demonstrate that this scheme can provide a reliable, secure, and efficient quantum-enhanced image encryption solution for high-security scenarios such as robot vision and confidential image transmission.

[0166] like Figure 5 The diagram shows the functional modules of the system of this invention: The system for implementing the image encryption and decryption method based on quantum cell neural networks disclosed in this invention includes a parameter setting module, a key generation module, an image encryption module, an image decryption module, and an encryption / decryption module; the parameter setting module, key generation module, image encryption module, image decryption module, and encryption / decryption module are connected in series; the parameter setting module is used to set the network parameters of the quantum cell neural network, generate the corresponding motion trajectory, and upload the data information to the key generation module; the key generation module is used to generate an encryption key based on the received data information and the obtained motion trajectory, and upload the data information to the image encryption module; the image encryption module is used to encrypt the input image based on the received data information and the generated encryption key, using block cyclic shift and position scrambling, and upload the data information to the image decryption module; the image decryption module is used to decrypt the encrypted image based on the received data information and the encryption process, using block cyclic shift and position regression, and upload the data information to the encryption / decryption module; the encryption / decryption module is used to complete the encryption and decryption of the target image based on the quantum cell neural network based on the received data information.

Claims

1. An image encryption / decryption method based on quantum cellular neural networks, comprising the following steps: Key generation: S1. Set the network parameters of the quantum cell neural network and generate the corresponding motion trajectory; S2. Generate an encryption key based on the motion trajectory obtained in step S1; Image encryption: S3. Based on the encryption key generated in step S2, perform image encryption on the input image using block cyclic shifting and position scrambling; Image Decryption: S4. Based on the encryption process in step S3, and using block cyclic shifting and position regression, perform the decryption operation on the encrypted image; S5. Complete the encryption and decryption of the target image based on quantum cell neural network.

2. The image encryption / decryption method based on quantum cellular neural networks according to claim 1, characterized in that... Step S1, which involves setting the network parameters of the quantum cell neural network and generating the corresponding motion trajectory, specifically includes the following steps: Setting the network parameters of a quantum cell neural network, including the tunneling energy between quantum dots. The first parameter is proportional Tunneling energy between quantum dots The second parameter is proportional The first weighting coefficient reflecting the difference in polarization rate between adjacent cells The second weighting coefficient reflects the difference in polarization rate between adjacent cells. The polarizability of the first quantum cell unit The polarizability of the second quantum cell unit The phase angle of the quantum state of the first quantum cell unit Phase angle with the quantum state of the second quantum cell unit ; The generated motion trajectory is calculated using the following formula: In the formula The trajectory of the first quantum cell unit; t is time; The trajectory of the second quantum cell unit; The trajectory of the first quantum cell unit's quantum state, defined by the phase angle. The trajectory of the second quantum cell unit's quantum state, representing the phase angle of its motion.

3. The image encryption / decryption method based on quantum cellular neural networks according to claim 2, characterized in that... Step S2, which involves generating an encryption key based on the motion trajectory obtained in step S1, specifically includes the following steps: The first key is calculated using the following formula: In the formula The index is the integer part of the time, and M is the length of the input image, and N is the width of the input image; The index is the integer part of the time, and ; The first key to quantum cell neural networks; The second key for quantum cell neural networks; The third key to quantum cell neural networks; This is the fourth key for quantum cell neural networks; This is the fifth key to the quantum cell neural network; The second key is calculated using the following formula: In the formula for The corresponding second key, and The values ​​are 1, 2, 3, and 4; It is a modulo operation function, and The remainder after dividing aa by bb; express The absolute value is rounded down; The first and second encryption keys are generated using the following formulas: In the formula This is the first encryption key generated; The generated second encryption key; This is an XOR operation; Set the parameters for the Henon mapping, including parameters that control the degree of nonlinear stretching and folding of the mapping. Parameters that control the degree of linear contraction of the mapping Henon mapping initial iteration value in the x-direction The initial iteration value of the Henon mapping in the y-direction ; The first and second sequences are generated using the following formulas: In the formula It is the first sequence; It is the second sequence; Let be the sequence number, and MN; The first and second encryption sequences are generated using the following formulas: In the formula This is the first encrypted sequence; This is the second encryption sequence; ; Generate a third encryption key for .

4. The image encryption / decryption method based on quantum cellular neural networks according to claim 3, characterized in that... Step S3, which involves encrypting the input image using the encryption key generated in step S2, based on block cyclic shifting and position scrambling, includes the following steps: The matrix corresponding to the input image is divided into four equal sub-matrices; According to the generated After shifting each submatrix to the left, they are then merged to obtain the first encryption matrix. Expand the first encryption matrix into a one-dimensional vector, and use the first encryption key. The elements in the array are rearranged in a set order, and the positions of the one-dimensional vector expanded from the first encryption matrix are scrambled according to the rearranged sequence to obtain the first encryption vector. Use the second encryption key The elements in the array are rearranged in a set order, and the positions of the first encryption vector are scrambled according to the rearranged sequence to obtain the second encryption vector. The second encryption vector is XORed with the third encryption key, and the resulting vector is transformed back into a two-dimensional image to obtain the encrypted image, thus completing the image encryption of the input image.

5. The image encryption / decryption method based on quantum cellular neural networks according to claim 4, characterized in that... Step S3 specifically includes the following steps: The input image matrix A is divided into four equal-sized submatrices, and each submatrix is ​​expanded into a one-dimensional vector, represented as follows: , , and ; According to the generated The following formula is used to shift each submatrix to the left: In the formula for A one-dimensional vector shifted to the left; Indicates a left shift operation; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; for A one-dimensional vector shifted to the left; for The key used for partial left shifts; Will , , and Reconstruct the submatrices and merge them to obtain the first encryption matrix. ; The first encryption matrix Expand into a one-dimensional vector ; the first encryption key The elements in the record are rearranged in ascending order. In the rearranged sequence, each element is in The first rearrangement sequence is obtained by taking the corresponding position in the sequence. ;(like , After rearrangement, it becomes ,but The first element 2 in B represents the rearranged element 1 in... (The position in the middle is 2); the position scrambling operation is performed using the following formula to obtain the first encryption vector. : In the formula The first encryption vector The jj-th element; Represents a one-dimensional vector The serial number is Element; Use the second encryption key The elements in the record are rearranged in descending order. In the rearranged sequence, each element is in The corresponding positions in the sequence yield the second rearrangement sequence. The second encryption vector is obtained by performing a position scrambling operation using the following formula. : In the formula For the second encryption vector The k-th element; Represents the first encryption vector The serial number is Element; The second encryption vector is generated using the following formula. With the third encryption key Perform an XOR operation: In the formula The one-dimensional vector obtained by the XOR operation The first in One element, For the third encryption key The first in One element; a one-dimensional vector The image is then transformed back into a two-dimensional image to obtain the encrypted image. Complete the image encryption of the input image.

6. The image encryption / decryption method based on quantum cellular neural networks according to claim 5, characterized in that... The decryption operation of the encrypted image described in step S4, based on the encryption process in step S3 and using block cyclic shifting and position regression, includes the following steps: The encrypted image obtained in step S3 is converted into a one-dimensional vector, and an XOR operation is performed to obtain the first decryption vector; According to the second encryption key The rearrangement rules are used to perform position regression on the first decryption vector to obtain the second decryption vector; According to the first encryption key The rearrangement rules are used to perform another position regression on the second decryption vector to obtain the third decryption vector; The third decryption vector is restored to a two-dimensional image, and the corresponding image matrix is ​​obtained; According to the generated The image matrix is ​​shifted to the right to obtain the final decrypted image; Complete the decryption operation of the encrypted image.

7. The image encryption / decryption method based on quantum cellular neural networks according to claim 6, characterized in that... Step S4 specifically includes the following steps: The encrypted image obtained in step S3 is converted into a one-dimensional vector G; the first decryption vector is obtained by performing an XOR operation using the following formula. : In the formula The first decryption vector The mm-th element in; According to the second rearrangement sequence The second decryption vector is obtained by performing position regression on the first decryption vector using the following formula. : In the formula For the second decryption vector The nth element in; Indicates the second decryption vector The serial number is Element; According to the first rearrangement sequence The third decryption vector is obtained by performing position regression on the second decryption vector using the following formula. : In the formula The third decryption vector The ooth element in; Represents the third decryption vector The serial number is Element; The third decryption vector The image is restored to a two-dimensional image, divided into four equal sub-matrices, and each sub-matrix is ​​expanded into a one-dimensional vector, represented as follows: , , and ; According to the generated The following formula is used to calculate... , , and Perform a right shift operation: In the formula for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; for A one-dimensional vector shifted to the right; for The key used for partial right shifts; This is a right shift operation; Will , , and The submatrices are then reconstructed and combined to obtain the final decrypted image; Complete the decryption operation of the encrypted image.

8. A system for implementing the image encryption / decryption method based on quantum cellular neural networks as described in any one of claims 1 to 7, characterized in that... It includes a parameter setting module, a key generation module, an image encryption module, an image decryption module, and an encryption / decryption module; the parameter setting module, key generation module, image encryption module, image decryption module, and encryption / decryption module are connected in series; the parameter setting module is used to set the network parameters of the quantum cell neural network, generate the corresponding motion trajectory, and upload the data information to the key generation module; The key generation module is used to generate an encryption key based on the received data and the obtained motion trajectory, and then upload the data to the image encryption module. The image encryption module is used to encrypt the input image based on the received data information and the generated encryption key, using block cyclic shifting and position scrambling, and then upload the data information to the image decryption module; The image decryption module is used to decrypt the encrypted image based on the received data information and the encryption process, using block cyclic shifting and position regression, and then upload the data information to the encryption / decryption module. The encryption / decryption module is used to encrypt and decrypt the target image based on the received data information using a quantum cell neural network.

Images