Image Watermarking Using Geometric Moments

Image watermarking is a method that embeds a watermark in the digital image by making small changes in the host data. The embedded watermark must be perceptually invisible and it must be difficult to change or remove it against the intentional or unintentional attacks such as compression, linear and non linear filtering and geometric transformations. In watermarking applications, the robustness of the watermark to the common signal processing and geometric attacks is essential to the system. The paper presented a method, based on scale and rotation information estimated from the edges of the image. Fleet proposed an approach in which the watermark is composed of the sum of sinusoids that appear as peaks in the frequency domain which can be used to determine the geometric distortion. A different approach was presented in which watermark is inserted in rotationally, scale and translation invariant parameters based on Fourier mellintransform. However, this method is based on discrete Fourier transform of the image which is highly sensitive to dropping and it cannot survive aspect ratio and rotationally attacks. In order to survive geometric attacks, it is like to develop a watermarking method using Geometric moments. Geometric moments have been intensively used in the literature for image recognition, template matching etc. This proposed a robust scheme test the method using the geometrically distorted software generated using stir mark and mat lab software’s. In order to overcome some of the disadvantage of existing methods and to obtain an invariant watermark an image watermarking technique using geometric moments invariable is proposed in this watermarking section and detection algorithm is presented in order to verify the proposed approach, simulation results are carried out. It also verified the invariance property of moment’s invariant by considerly standard images. The proposed approach in robust, invariant, computationally efficient and no over head needed to transmit to the decoder side. The proposed approaches have proven to be highly robust to geometric manipulations and image compression.