DWT and Signal Energy Based Zero-Watermarking Algorithm for Text Image

Abstract

A zero-watermarking algorithm for text image that based on DWT and the signal energy was proposed. We first change the text file into the corresponding binary image, then apply triple-wavelet transform on the binary image, and combine relations of the signal energy to construct zero-watermark. At the watermark detection step we use the similarity to judge the copyright Information. The attack experiments shown that the proposed algorithm were robust to a variety of attacks, and had a good Invisibility. The watermark algorithm obtained a better solution to the embedding capacity and the Invisibility, for the zero-watermark algorithm does not need to modify the carrier data. So it can meet the requirements of practical application.