Based on Discrete Diagonal Scan: A New Robust Image Encryption Using Confusion and Diffusion

Image encryption technique has emerged as a remarkable solution to shield the image against adversaries. Despite much advancement, a robust image encryption system is needed to achieve excellent encryption quality. Therefore, a highly-secured encryption algorithm needs impressive random key with excellent expansion method, small initial size and regeneratable capabilities, improved confusion and diffusion techniques. To achieve these goals a method for image encryption containing three major phases was developed. Firstly, an encryption and decryption key were generated and expanded with high level of randomness via Knight Tour (KT) algorithm with small initial size. Secondly, a new method of confusion was developed using switching technique and Discrete Diagonal-SCAN (DD-SCAN) approach for enhancing the level of pixel dispersion. Finally, an innovative method of diffusion was introduced to change the image pixels' value using bit wise operator and pixel circular shifting method. These involved steps such as generating a Ginger bread man image using the initial key, applying Arnold Cat Map (ACM) transform to dissolve flat regions in the ginger bread man image, and performing a bit-wise operation between the outcome from ACM transform, using results from DD-SCAN method, and utilizing the pixels circular shifting based on the generated key. The encryption key was proven to withstand against several types of attacks and tested using National Institute of Standards and Technology (NIST) randomness test. Thus, the pixel's value was altered and the cipher image was produced. The performance of the scheme was validated against forty-four images acquired from Signal and Image Processing Institute (SIPI) for standard image dataset. Shortly, the proposed image encryption scheme has been affirmed to be a superior image encryption scheme development.