Strong Robustness Watermarking Algorithm Based on Lifting Wavelet Transform and Hessenberg Decomposition

Abstract

Watermark imperceptibility and robustness in the present watermarking algorithm based on discrete wavelet transform (DWT) could be weakened due to data truncation. To solve this problem, a strong robustness watermarking algorithm based on the lifting wavelet transform is proposed. First, the color channels of the original image are separated, and the selected channels are processed through lifting wavelet transform to obtain low-frequency information. The information is then split into blocks, with Hesseneberg decomposition performed on each block. Arnold algorithm is used to scramble the watermark image, and the scrambled watermark is transformed into a binary sequence that is then embedded into the maximum element of Hessenberg decomposed matrix by quantization modulation. The experimental results exhibit a good robustness of this new algorithm in defending against a wide variety of conventional attacks.