Enhancing Image Security: A Hybrid Asymmetric Cryptosystem For Robust Watermarking With Comparative Performance Analysis

Abstract

There is a storm of digital data being kept and published on public platforms like LinkedIn, Facebook, Twitter, and others due to the explosion of AI-powered smart devices and the Internet of Things (IoT).  Of them, digital pictures are by far the most popular means of disseminating information.  According to recent statistics, Google Photos stores more than one billion photographs per week.  Furthermore, digital images are vital in many fields and applications, such as business, legal systems, social media, smart healthcare, autonomous vehicles, robotics, the film industry, and industrial automation.  But there are serious privacy and security issues with these pictures being misused.  An increasingly serious problem in the modern day, identity theft is one of the main causes of fraud in India and other nations.  Therefore, protecting digital photos is an important problem that must be solved.  The best way to protect digital photographs without breaking the bank is to use watermarking techniques.  To make digital content more resistant to modification and ensure its longevity, digital image watermarking has become an indispensable tool.  Nevertheless, because to the fact that each method is typically only applicable to certain applications, no digital watermarking solution now available provides total security against all types of attacks.  A recent development in information security has been the incorporation of deep learning methods, which have been beneficial to this area.  To protect the watermarked photographs from intruders, we will present a new and strong Hybrid Asymmetric Image Cryptosystem (HAIC)  method in this paper.  In addition, we have proved that our suggested system outperforms the current systems by comparing our work to similar systems that already exist.