Biometric Fingerprint Keys and DNA Encoding Secure Satellite Imaging

A research team has demonstrated an image cryptosystem designed to secure large-scale remote sensing and astronomical imagery by combining biometric key derivation, selective frequency domain encryption, biological computing operations, and chaotic mapping. Traditional symmetric encryption algorithms, including the advanced encryption standard and the data encryption standard, experience severe computational strain when applied to high-resolution earth observation files and remain susceptible to chosen-plaintext and known-plaintext attacks due to natural image correlation. To resolve the conflict between prohibitive processing overhead and cryptographic vulnerability, investigators developed a hybrid architecture that integrates selective two-dimensional Fourier domain scrambling, biometric fingerprint hashing via secure hash algorithms, synthetic deoxyribonucleic acid sequence rules, chaotic dynamical systems, and random modulus decomposition to establish an asymmetric defense boundary while guaranteeing full mathematical reconstruction of original images upon authorized decryption.

The cryptographic workflow begins by decomposing an input color image into its distinct red, green, and blue primary color channels before mapping them from the spatial plane into the frequency domain. In their experimental verification, the investigators processed full-color scenes measuring 945 by 1422 pixels. Rather than executing mathematically taxing transformations across the entire spectral plane, the protocol extracts a constrained square region measuring 400 by 400 pixels centered directly over the low-frequency core of the Fourier transformed spectrum.

Pixel scrambling within the extracted frequency block relies on the classical Arnold transformation, an area-preserving chaotic cat map governed by iterative matrix operations modulo the dimension of the extracted image matrix. Crucially, the exact number of iterations applied during the Arnold transformation is not fixed or transmitted over communication channels as a vulnerable operational parameter. Instead, the iteration count is determined dynamically through the biometric credential of the authorized user. The system applies the secure hash algorithm SHA-256 to a digital scan of the user’s personal fingerprint, outputting a unique 256-bit hexadecimal digest represented as an array of sixty-four hexadecimal values. Three consecutive values from this digest are summed to calculate the specific iteration count for each respective color channel. Because both the sending and receiving parties possess authentic biometric data, the Arnold transformation iteration parameter remains completely internal to each terminus, entirely removing the necessity to transmit the scrambling key across open networks.

To generate the pseudo-random numerical sequences required for subsequent masking, the cryptosystem employs a discrete two-dimensional Henon chaotic map parameterized with canonical coefficients set to 1.4 and 0.3. The initial starting coordinates for this chaotic map are seeded using segment extractions converted from hexadecimal to decimal representations derived directly from the biometric fingerprint digest. These chaotic sequences are formatted into two-dimensional matrices matching the dimensional specifications of the image arrays. In parallel, the choice among the eight standard biological rules of synthetic deoxyribonucleic acid encoding is indirectly steered by the biometric hash digest, ensuring that pixel-to-nucleotide translation follows a dynamic, user-dependent rule selection mechanism that significantly broadens key space and prevents analytical discovery through brute-force enumeration.

Following Arnold scrambling in the frequency plane, the transformed data is converted back into intermediate spatial representations using the inverse two-dimensional Fourier transform and subjected to algebraic bio-molecular computation. The intermediate spatial pixel arrays and the chaotic random matrices are mapped into discrete sequences of biological base letters. The protocol executes formal deoxyribonucleic acid addition operations between the scrambled image sequences and the chaotically generated nucleotide sequences before decoding the combined sequences back into algebraic decimal matrices, establishing a dense symmetric diffusion barrier that shuffles pixel intensities and disperses localized structural features across wide data ranges.

The critical structural transition from symmetric computation to an asymmetric cryptosystem is achieved through the integration of random modulus decomposition. Under the proposed architecture, the deoxyribonucleic acid decoded matrix is multiplied by a synthetic random phase mask derived from the chaotic Henon matrix and passed once more through the two-dimensional Fourier transform. The system then decomposes the resulting transformed complex signal into two distinct complex matrices, designated as the ciphertext image and the asymmetric private decryption key. This geometric decomposition distributes random phase vectors and random amplitude factors governed by two independent random parameter matrices, ensuring that mathematical decryption cannot proceed simply by running the symmetric steps in reverse without holding the precise decomposed key matrix.

During decryption, an authorized recipient reverses the pipeline by combining the ciphertext with the specific asymmetric key component, performing an inverse Fourier transformation, and deriving the intermediate nucleotide array. The recipient encodes this array into synthetic biological bases and executes formal deoxyribonucleic acid subtraction utilizing the identical chaotic sequence regenerated locally from the authorized fingerprint hash. After translating the base sequence back into numerical coordinates using the corresponding decoding rule and hash normalization factors, the data is transformed back into the Fourier plane. The system executes the inverse Arnold transformation using the iteration count derived from the recipient’s authenticated fingerprint, finally executing an inverse Fourier transformation to restore the three individual color channels. When recombined, the red, green, and blue components reproduce the full visual information of the original scene, successfully verifying that selective Fourier encryption does not prevent total spatial reconstruction.

The cryptographic efficacy and visual reconstruction fidelity of the architecture were systematically evaluated through numerical simulations conducted in MATLAB R2023b using an open-source photographic test image. Visual fidelity assessments between the original 945 by 1422 pixel scene and the decrypted image confirmed high reconstruction accuracy using standard mathematical benchmarks, including the peak signal-to-noise ratio, the mean square error, and the correlation coefficient calculated via native computational routines. In key sensitivity tests, investigators systematically applied minute mathematical deviations to individual operational parameters, including fractional transform parameters and Arnold iteration counts; every distorted decryption attempt produced completely unintelligible visual noise without recovering identifiable structural details of the source image.

Statistical evaluation further confirmed that the resulting ciphertexts exhibit excellent resistance against eavesdropping, statistical pattern matching, and differential cryptanalysis. Diagonal sampling across four thousand adjacent pixels demonstrated that the severe linear correlation naturally present among adjacent spatial coordinates of the original image components collapsed into an entirely scattered, uncorrelated distribution within the ciphertext. The information entropy of the ciphertext reached approximately 7.75, which approaches the theoretical maximum of 8.0 for eight-bit grayscale representations. Researchers noted that this slight divergence from optimal entropy stems directly from the selective encryption of the central Fourier zone, representing an intentional and advantageous operational trade-off that dramatically relieves computational demands while preserving sufficient structural randomization.

Simulation trials also established remarkable resilience against external channel interference and malicious differential cryptanalysis. When ciphertext was exposed to additive zero-mean Gaussian noise during simulated transmission, the algorithm achieved successful reconstruction with a high correlation coefficient of 0.9285 provided the noise coefficient remained below ten to the negative third power, though higher noise levels altered the synthetic biological base alignments and degraded the output. Intentional data occlusion demonstrated that loss of frequency coefficients prevents unauthorized spatial recovery due to the distributed nature of the frequency plane. Finally, validation on multi-spectral satellite imagery verified that the protocol operates effectively across diverse remote sensing channels, establishing an efficient, highly resilient, and asymmetric security solution for orbital data networks.

Subject of Research: Satellite and remote sensing image encryption using biometric key generation, selective Fourier processing, DNA encoding, and random modulus decomposition

Article Title: Satellite image encryption using biometric information using random modulus decomposition and DNA encoding

Article References: Shikder, A., Rao, S. K., & Nishchal, N. K. (2025). Satellite image encryption using biometric information using random modulus decomposition and DNA encoding. Discover Imaging, 2(1), Article 8. https://doi.org/10.1007/s44352-025-00013-2

Image Credits: AI Generated

DOI: 10.1007/s44352-025-00013-2

Keywords: satellite image encryption, remote sensing security, biometric cryptography, DNA encoding, random modulus decomposition, Fourier transform, Henon chaotic map, Arnold transformation, SHA-256 hashing, differential attack resistance, information entropy, image cryptosystems, multispectral imaging, key sensitivity analysis, selective encryption, asymmetric cryptosystems, correlation coefficient analysis