Digital watermarking based on locally linear embedding

Robustness is one of the crucial issues in digital watermarking. Especially the robustness against geometric distortion and JPEG compression at the same time remains challenging. In this paper, a locally linear embedding (LLE) based watermarking algorithm that is robust against affine transformation is proposed. This algorithm improves the robustness via the intrinsic robustness of the LLE. A random generated watermark is embedded in the coefficients of reconstruction weights of the locally linear embedding. In watermark extraction, the watermark can be extracted almost the same process as the watermark embedding. Experimental results have demonstrated that the proposed watermarking scheme is more robust than other watermarking algorithms reported in the literature. Specifically, it is robust against almost all affine transform related testing functions in StirMark 3.1. While the approach is presented for gray-level images, it can be applied to color images and video sequences.