Robust curvelet-domain image watermarking based on feature matching

Abstract

Robust image watermarking has abstracted increasing attention in the past decade. For the existing image watermarking approaches, it has been proved that the feature points-based schemes can efficiently resist to geometric distortions. However, the main drawback of such schemes is that their embedding strategy in spatial domain restrains the robustness against common image processing operations. In view of this fact, we present a robust curvelet-domain image watermarking based on feature matching. The proposed scheme consists of three key components: (1) feature points extraction and selection via fuzzy c-means clustering algorithm; (2) matching the selected feature points and estimating the geometric parameters which will be used to restore the distorted watermarked image; and (3) embedding the watermark in the middle-scale curvelet coefficients according to the position relationships. Experimental results obtained using Stirmark confirm that the proposed image watermarking achieve good performance in terms of imperceptibility as well as robustness against many various distortions.