Chaotic-based image encryption is widely utilized, yet many algorithms relying on low-dimensional chaotic maps suffer from insufficient security margins. To overcome these vulnerabilities, an encryption framework leveraging a six-dimensional (6D) hyperchaotic generator integrated with DNA sequence operations has been developed. Initially, multiple pseudo-random chaotic trajectories drive pixel diffusion and positional scrambling on the source image. Subsequently, at the DNA level, distinct chaotic sequences further permute and diffuse the mapped nucleotide sequences. Finally, the mutated DNA segments are decoded and synthesized to form the final ciphertext image. Evaluations demonstrate that this architecture achieves an information entropy nearing 8, significantly reduced adjacent pixel correlation, and an expansive key space exceeding 2^300. Furthermore, the mechanism exhibits strong resilience against geometric distortions and data loss attacks.

Encryption Mechanism

The core architecture replaces conventional low-dimensional maps with a 6D hyperchaotic system to amplify the pseudo-random sequence complexity, fortifying the encryption against predictive attacks. The implementation follows a structured pipeline:

1. Parameter Initialization: Six initial conditions and system parameters are defined to seed the 6D hyperchaotic generator.
2. Matrix Segmentation: The plain image undergoes binarization and is partitioned into distinct 8x8 spatial blocks.
3. Pixel Diffusion: The generated chaotic vectors alter the numerical values within each spatial block, obscuring statistical patterns.
4. Spatial Permutation: Separate chaotic trajectories shuffle the diffused block positions, disrupting spatial correlations.
5. DNA-Level Transformation: Scrambled blocks are translated into DNA nucleotide sequences. Additional chaotic sequences execute secondary diffusion and permutation directly on the DNA representation.
6. Ciphertext Synthesis: The transformed DNA sequences are recombined and decoded to construct the final encrypted image.

Performance Metrics

Security and robustness assessments validate the efficacy of the proposed method:

• Information Entropy: The ciphertext image achieves entropy values approaching 8, confirming optimal randomness and minimal information leakage.
• Correlation Coefficient: Adjacent pixel correlations in the encrypted output are drastically minimized, demonstrating superior confusion capabilities.
• Key Space Complexity: The algorithm supports a key space greater than 2^300, rendering brute-force decryption computationally infeasible.
• Attack Resilience: The framework maintains structural integrity against both geometric manipulations and cropping attacks, ensuring reliable data recovery.

Entropy Calculation Implementation

The following Matlab function computes the Shannon entropy for each RGB channel of an image, serving as a metric to evaluate the randomness of the encrypted output:

function channel_entropies = calculate_rgb_entropy(input_image)
    channel_entropies = zeros(1, 3);
    
    function e = shannon_entropy(channel_data)
        channel_data = double(channel_data);
        [rows, cols] = size(channel_data);
        flattened = reshape(channel_data, 1, rows * cols);
        freq = zeros(1, 256);
        
        for val = 0:255
            freq(val + 1) = sum(flattened == val) / (rows * cols);
        end
        
        valid_probs = freq(freq > 0);
        e = -sum(valid_probs .* log2(valid_probs));
    end
    
    channel_entropies(1) = shannon_entropy(input_image(:, :, 1));
    channel_entropies(2) = shannon_entropy(input_image(:, :, 2));
    channel_entropies(3) = shannon_entropy(input_image(:, :, 3));
end