A Secure Digital Image Encryption Using Hybrid Transformation method with DNA Algorithm
DOI:
https://doi.org/10.7091710.70917/ijcisim-2026-1959Keywords:
Chaos algorithm, DNA method, Fourier Transform, Generalized Arnold Transform, Vector Decomposition, Wavelet TransformAbstract
Due to multiple assaults, securing digital images during transmission and storage is important to maintain data security. Various cryptography techniques have been employed in the past to increase the security of digital images. However, due to significant redundancy and strong correlation between neighboring pixels, conventional encryption algorithms have trouble improving the security level of images. Transform-based encryption approaches are used to secure medical images to deal with these issues. However, traditional wavelet and Fourier transform-based encryption have difficulty maintaining data after processing. The Shearlet transform Integrated Deoxyribonucleic Acid (DNA) method is proposed in this current study to protect medical images with reduced information loss. The supplied image is first transformed into a grayscale image. The input images are then decomposed into 41 sub-images using the Shearlet transform. Following that, this decomposed sub-picture is scrambled using the Generalized Arnold Transform and circularly shifted utilizing Intra-Inter bit level pixel permutation. The deconstructed sub-pictures are assembled using the vector decomposition process into a synthesized image. The final cypher image is then created using the DNA technique. The Chaos algorithm produces the key required for encryption using the DNA method. Finally, the cypher image is created by feeding the result of the DNA method through the gyrator transform. The reverse encryption method is used to recover the original image. This work achieves 0.99 correlation coefficient (CC), 56.75 Peak signal-to-noise ratio (PSNR), 0.0056 mean-square error (MSE), 0.99 Structural Similarity Index Metrics (SSIM), 0.015 Mean Absolute Error (MAE), 0.056 of Root Mean Square Error (RMSE).98.65 of the number of pixels change rate (NPCR) and 32.95 of unified average changing Intensity (UACI). According to experimental data, the algorithm has good robustness and security.