Abstract:

The security of digital data has become increasingly important as a result of recent advancements in multimedia technology. Researchers frequently focus their efforts on changing the current protocols to solve the flaws of the present security measures. Yet, during the past few decades, several proposed encryption algorithms have been insecure, posing a huge security risk to sensitive data. It is crucial to use the best encryption method to defend against such attacks, but which algorithm is best in a particular situation will depend on the kind of data that has to be secured. However, evaluating various cryptosystems one at a time to determine the optimal choice can consume a significant amount of processing time. We suggest a security-level identification method for picture encryption techniques that uses a support vector machine for quick and precise selection of the correct encryption algorithms (SVM).We also create a dataset using a variety of encryption security criteria, including entropy, contrast, homogeneity, peak signal-to-noise ratio, mean square error, energy, and correlation.. We use these variables as features that are taken from various cipher pictures

Description:

.