No abstract

Abstract:Watermarking is one of the efficient approaches for digital authentication. An adaptive feature point extraction model is proposed in this paper for robust watermarking. The host image is treated with number of geometric attacks for extracting the consistent feature points form the image. The proposed watermarking scheme follows the adaptive feature point extraction method to retrieve the feature points from the host image. The response value of the feature points is calculated for improving the selection of feature points. The watermark insertion procedure is employed by inserting the watermarking bits with in the place of feature points. The feature point portion is extracted from the image and replaces the portion with watermarking bits. The watermark extraction procedure is used to restore the original image from the watermarked image. The watermarking bits in the image are replaced by the feature point portion for restoring the original image. The simulation experiment is carried with the MATLAB simulator. The proposed algorithm is tested with geometric attacks such as scaling, rotation, noise pollution and JPEG compression. The proposed method proved its efficiency when compared to other remaining algorithms.

Index Terms-Image function, algebraic invariants, polar coordinates, affine groups, information redundancy, fourier-mellin invariants, legendre moment invariants, counter-based shape descriptors, geometric moment invariants.

No abstract

No abstract