Integration of Spatial and Frequency Domain Encryption for Digital Images

Abstract

Transmission of multimedia data such as images, videos, and audio over the Internet is risky due to cyberattacks. To overcome the security issues, several encryption schemes are proposed over the last few decades which also possess few vulnerabilities such as time inefficiency and weak security. In this research, to provide the highest level of security to the digital data, chaos is incorporated for the scrambling of rows and columns of the plaintext image. Further, a noisy image is generated based on the chaotic logistic map and the suitable initial conditions which are selected based on the analysis performed. For the reduction of the encryption computational time, a Discrete Wavelet Transform (DWT) is used in which only low-frequency bands are encrypted because most of the plaintext information lies in such frequency bands. To gauge the performance of the proposed encryption scheme, several security tests such as entropy, correlation, energy, peak signal to noise ratio, mean square error, keyspace, and key sensitivity analysis, noise-resistant, and cropping attack analyses are performed. From the cropping and noise attack analysis, we have found that the proposed encryption algorithm can decrypt the plaintext image with negligible loss of information but the content of the plaintext image can be visualized.