Akademik Veri Yönetim Sistemi
JOURNAL OF SUPERCOMPUTING, cilt.82, sa.4, 2026 (SCI-Expanded, Scopus)
Seafood images, unlike traditional images, contain not only visual content but also high-value supply-chain data, such as species identification and quality inspection results. Leakage of such data can compromise trade secrets, reduce market competitiveness, and even cause food safety risks. To address the risk of sensitive information leakage during the acquisition, transmission, and sharing of seafood images, this paper proposes a reversible image encryption method based on a novel three-dimensional Logistic map (3D-LM). The conventional one-dimensional Logistic map suffers from limited key space, restricted chaotic intervals, and periodicity or degradation in digital implementations. In contrast, the proposed 3D-LM offers a higher-dimensional state space, more complex chaotic behavior, a larger key space, and greater parameter sensitivity. In the proposed method, a hash function first extracts a digest from the original image to generate the cryptosystem's initial key parameters. The 3D-LM then produces keystream sequences that serve as high-quality pseudo-random numbers for encryption. The three color channels of the image are combined into a new pixel matrix, and a synchronous permutation-diffusion structure is applied to encrypt the image content. Experimental results show that the proposed scheme provides high security with low computational cost, enabling real-time, high-throughput encryption of seafood images during acquisition, transmission, and sharing. This method is compatible with parallel and high-performance computing, providing an effective solution for data security in the seafood supply chain.