Quantifying the Visual Content Perceived in DCT Coefficients and PKI Based Semi-fragile Watermarking for Visual Content Authentication of Images

Abstract

Multimedia data is always exposed for the security threats, because these data can be easily alter by using multimedia data processing tools. Therefore content owners always endure of copyright protection and ownership verification of their digital assets. Digital watermarking is the technology introduces to overcome these limitations. There are different watermarking technologies which focusing different security aspects. Proving the authenticity of digital multimedia data, the watermarking concept was commenced called fragile watermarking. However, evolution of the fragile watermark derives a prospective watermarking concept called semi-fragile watermark, which overcomes several limitations of fragile watermarking such as fragility for trusted image manipulations. Multimedia data consisted huge amount of information; however, most of them are not in the perceptible region of Human Visual System (HVS). In this paper, we presented an experiment that quantifies the perceptible information perceived in DCT domain and a new PKI (Public Key Infrastructure) based semi-fragile watermarking technique that quantitatively authenticates perceptible content of the digital images. Additionally, our definition localizes the compromised regions of the image. The proposed method inherits all security features in public key authentication system. In practice, visual distortions are acceptable up to some extent in digital multimedia applications. Therefore, distortions due to the watermark should be minimized. Moreover, this paper presents a critical analysis of the parameterization of watermark embedding technique and its analysis matrices (graphs & tables), which can be used to optimize the watermarking method according to the userpsilas requirements.