Fuzzy Inference Rule Based Reversible Watermarking for Digital Images

The last couple of decades have seen rapid growth of research interest in the field of reversible watermarking of multimedia data. The primary aim of reversible watermarking is to restore the original cover data content, with zero residual distortion, after watermark extraction. Such a feature is desirable in industries dealing with highly sensitive data, e.g. in military, medical and legal industries. In this paper we propose a reversible watermarking algorithm for grayscale images, based on fuzzy inference mechanism based pixel prediction method. We apply a thresholding technique on the prediction errors to embed the watermark bits. Our experimental results show that the quality of the watermarked cover data, produced by the proposed method, is considerably high compared to the other state–of–the–art schemes.