Robust Image Watermarking Based on Feature Regions

Abstract

In image watermarking, binding the watermark synchronization with the local features has been widely used to provide robustness against geometric distortions as well as common image processing operations. However, in the existing schemes, the problems with random bending attack, nonisotropic scaling, general affine transformation, and combined attacks still remain difficult. In this chapter, we present and discuss the framework of the extraction and selection of the scale-space feature points. We then propose two robust image watermarking algorithms through synchronizing watermarking with the invariant local feature regions centered at feature points. The first algorithm conducts watermark embedding and detection in the affine covariant regions (ACRs). The second algorithm is combining the local circular regions (LCRs) with Tchebichef moments, and local Tchebichef moments (LTMs) are used to embed and detect watermark. These proposed algorithms are evaluated theoretically and experimentally, and are compared with two representative schemes. Experiments are carried out on a set of standard test images, and the preliminary results demonstrate that the developed algorithms improve the performance over these two representative image watermarking schemes in terms of robustness. Towards the overall robustness against geometric distortions and common image processing operations, the LTMs-based method has an advantage over the ACRs-based method.KeywordsFeature PointImage WatermarkWatermark SchemeGeometric DistortionWatermark EmbeddingThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.