Representative Image Encryption and Decryption Framework with Block-Level Pixel Analysis for Medical Applications

The importance of encryption algorithms has been massively enhanced in recent times due to the unstoppable evaluation of the internet, cybercrimes, and intruders. However, existing encryption algorithms have some genuine problems like most of the traditional encryption methods use only one data server and thus, is not suitable for the distributed framework. Another problem is the high distortion in the marked image at the time of reconstruction and the complexity also remains massive. Thus, the visual quality of reconstructed images and encryption efficiency needs to be improved. Therefore, a Representative Image Encryption and Decryption (RIED) framework is proposed in this work to improve the performance of distributed frameworks using the multi-server concept. The proposed RIED frameworks improve embedding capacity, visual quality, embedding capacity, and encryption efficiency by reducing pixel noise. A block-level pixel analysis method is utilized to analyze the correlation between pixels of different blocks of an image. The proposed RIED framework is segregated into three stages namely the encryption stage to determine pixel correlation of blocks and encrypt the image segment using ST key, the embedding stage to embed secret data encrypted segments and broadcast it to different servers, and the decryption stage to separate the secret data and original image from the marked encrypted image. The mean performance results in terms of PSNR and SSIM using the proposed RIED framework are 50.77 dB and 0.9985, respectively.