Digital watermarking method against geometric attack based on phase consistency and pre-cancellation

Abstract

The invention discloses an anti-geometric attack digital watermarking method based on phase consistency and pre-cancellation. In the embedding process, first calculate and extract the points with high carrier phase consistency to form a feature point set; calculate the two-dimensional convex hull of the feature point set, number and divide the boundary points of the convex hull to form a triangular network, and each triangle It is called the carrier cell; the carrier cell is transformed into a standard triangle by affine inverse transformation; before embedding the watermark signal, it is pre-offset, and the projection of the carrier signal on the watermark signal is removed from the watermark signal; finally, the The watermark is superimposed on the standard triangle of the carrier, and the carrier cell containing the watermark is obtained through affine transformation. In the watermark extraction process, first analyze and extract the feature points with high phase consistency, calculate and divide the convex hull, generate a triangular network, transform it into a standard triangle, decode it with the watermark template, and extract the watermark in the carrier cell cavity. Even if the watermarked image is attacked by local geometry, the watermark can be extracted correctly.

Description

Technical field [0001] The invention relates to an anti-geometric attack digital watermark embedding method based on phase consistency and pre-offset, and an anti-geometric attack digital watermark extraction method based on phase consistency and pre-offset. Background technique [0002] With the development of the network and the widespread application of multimedia technology, the problem of network infringement is also increasing, and the problem of digital image copyright has become a prominent problem. As an effective method to solve this problem, digital image watermarking technology has achieved good development in recent years. Current watermarking algorithms can effectively resist general signal processing such as compression, noise addition and filtering, but they lack robustness against geometric attacks. After the attacker performs geometric transformations such as rotation and scaling of the target image, most of the watermarking algorithms lose synchronization unde...

AI Technical Summary

Problems solved by technology

Current watermarking algorithms are effective against general signal processing such as compression, noise addition, and filtering, but lack robustness against geometric attacks

After the attacker performs geometric transformations such as rotation and scaling on the target image, most of the watermarking algorithms lose synchronization under this attack, which makes the extraction of the watermark fail, and this transformation does not affect the visual effect, so how to effectively resist Attack is a major difficulty in current research

Images