Image watermarking using slantlet transform

Abstract

Watermarking robustness is one of the major characteristics that influence the performance and applications of digital image watermarking. For copyright protection, the watermarking robustness must be increased without significantly degrading the visual quality of the host image. For this purpose, a watermarking algorithm is proposed in this paper. The proposed algorithm is based on wavelet-like transform, known as Slantlet Transform (SLT). The basic idea is to decompose the original image using SLT; a binary watermark is then embedded in the high frequency sub-bands. The embedding process is done by modifying horizontal and vertical high frequency coefficients in a content-based manner. A comparison is made between the proposed embedding method and other embedding methods in terms of Peak Signal to Noise Ratio (PSNR). A second comparison is made for the proposed embedding method based on different transforms (i.e. SLT and DWT) in terms of PSNR. The Normalized Correlation (NC) value between the original watermark and the extracted watermark after applying different attacks is calculated to test the robustness of the proposed method. From the first comparison, the results show that the proposed embedding method yields better imperceptibility. The results of the second comparison demonstrate that the proposed algorithm can achieve better robustness and imperceptibility.