Qualitative Analysis of Cryptanalysis of Images Encrypted by Border Chaotic Cellular-Automaton-Based Model Using Segmentation Techniques

Data transactions on computers and phones have become increasingly common, with numerous entities exchanging both public and private data across expanding networks. However, for sensitive data, such as private information, transmitting it over public access networks like the Internet may pose security risks. Digital images are a particularly representative and increasingly significant data type in this context. To safeguard such data from leaks and unauthorized alterations during transmission, encryption serves as a crucial measure. Various cryptographic algorithms are available, each with distinct security requirements. Among these, encryption algorithms based on Cellular Automata offer alternative solutions characterized by their highly chaotic evolutionary capabilities, providing a significant degree of obfuscation, a critical aspect for maintaining confidentiality. This study employs image segmentation techniques such as k-means clustering, edge detection, and thresholding as tools for cryptanalysis. These segmentation techniques are applied to digital images generated using the Border Chaotic Cellular Automata (BCCA) cryptographic model. This approach enables the identification of artifacts and the highlighting of original image features within low-quality encryption. Conversely, more robustly encrypted images, achieved through additional encryption steps, remain unaffected by these techniques, underscoring the level of confidentiality inherent in the BCCA model.