Adaptive digital image watermarking based on predictive embedding and a dynamic fuzzy inference system model

Abstract

A novel image watermarking scheme is introduced that consists of an adaptive watermarking algorithm based on a model of the Human Visual System (HVS) and a Dynamic Fuzzy Inference System (DFIS). This scheme relies on the DFIS to extract the human eye sensitivity knowledge using the HVS model. The DFIS and the HVS combined are used to adjust and select the appropriate watermark length as well as the watermarking strength for each pixel in an image. The main goal of the algorithm is to provide a more robust and imperceptible watermark. The primary contribution of this work is to present a unique approach to perform image-adaptive watermarking by introducing a dynamic fuzzy logic technique. This logic relies on the statistical distributions of the HVS-generated data in order to accurately approximate the relationship found between all properties of the human perceptual system. In addition, we introduce a Predictive Watermark Embedding (PWE) algorithm that exploits the average power of the frequency-domain image coefficients as well as the aforementioned DFIS model in order to insure that the watermark insertion process is adaptively and accurately performed in lower-frequency components and significant DCT coefficients. (Abstract shortened by UMI.)