Self Adaptive Color Image Encryption Based on Chaotic Mapping

Abstract

With the increase in multidimensional applications of internet-based multimedia content, newlinesecure transmission of multimedia data including images has become the need of the day. newlineBiometric image-based security is deployed in all the electronic gadgets. Electronic newlinetransactions are involved in day-to-day life. Various multimedia applications include newlinetransmission of variety of images though the computer networks. Image encryption algorithms newlinevary in the way they are developed, the domain in which they encrypt and the inputs they newlineaccept. Since, the algorithms need to satisfy a variety of constraints imposed by different newlineapplications, there is a need for adaptivity. Adaptive encryption algorithms meet the security newlinestandards without compromising the quality of the transmission. Self-adaptive encryption newlinetechniques improve the security of the encryption. Inculcating adaptivity at various stages of newlineencryption algorithm is a promising field of current era. newlineThe research work develops few algorithms with a diverse set of concepts and demonstrates newlinethe usefulness of adaptivity in the image encryption algorithms. Algorithms are developed newlinepertaining to four classes: nature encryption algorithms, transform domain encryption newlinealgorithms, fast partial encryption algorithms and self-adaptive encryption algorithms. The newlineperformance parameters chosen for the analysis include histogram uniformity, correlation newlinecoefficients, number of changing pixel rate (NPCR), unified average change in the intensity, newlineentropy and encryption time. newlineInitial part of the research developed an image encryption algorithm with bio-inspired newlineoperations, confusion and diffusion, which are obtained with the mutation and crossover newlineoperations on rows and columns of the image. Improvement in NPCR, UACI, Correlation newlinecoefficients is achieved by the addition of a scrambler unit at the final stage of encryption. The newlinesecond algorithm is based on fractional Fourier transform. The Fourier domain transformed newlineimage is multiplied by two random chaotic phase ma