Image encryption using enhanced DCT transform for frequency domain applications

Abstract

improving system performance and reliability. Key leaks and inadequate key strength are common issues with numerical global key models and public block cipher algorithms that use pseudo-random numbers. Additionally, because of the spatial correlation between neighboring pixel values, no current crypto scheme is appropriate for picture encryption. An intellectual domain-based image cryptosystem with hierarchical modification structures is proposed to overcome all these obstacles. The suggested framework also aims to maximize confusion metrics and improve security efficiency. In this paper modified DCT transform domain based frequency domain image encryption is proposed along with optimized pixel grouping based cipher key formulation with reduced complexity overhead. The proposed image encryption model can achieve 100 percentage transformations irrespective to the image non-linear characteristics and associated spatial dimensions. To validate the performance metrics of proposed image encryption different biomedical images collected and the superiority of proposed system is verified using quantitative and qualitative analyzes. The proposed implementation helps in offering privacy in crypto system based network security. It also validated that this hybrid combination also extend the applicability of crypto system for EMR applications.