Method and apparatus for embedding and extracting watermark as well as processing system

Abstract

The embodiment of the invention relates to a method and a device for embedding a watermark, a method and a device for extracting the watermark, and a processing system, wherein the method for embedding the watermark comprises the following steps: a watermark matrix is generated; an orthogonal coordinate of a DFT coefficient for the watermark image to be embedded is transformed into a logarithmic coordinate; and the watermark information is embedded in the DFT coefficient of the orthogonal coordinate corresponding to the logarithmic coordinate according to the logarithmic coordinate and the watermark matrix. The method and the device for embedding a watermark and the method and the device for extracting the watermark ensure that the video watermark can ensure better visual imperceptibility and resist various geometric attacks, such as adding noise, scaling, intraframe clipping, intraframe parallel movement, blended attacks of scaling and clipping, simultaneously better reduce the computational complexity in the watermark extraction and save the time of the watermark extraction.

Description

technical field [0001] The invention relates to the field of multimedia technology, in particular to a watermark embedding method and device, an extraction method and device, and a processing system. Background technique [0002] With the rapid development of global information technology, digital information products have been widely used. Since digital video products are easy to be copied by pirates, higher requirements are put forward for the copyright protection technology of digital video products. Digital watermarking is a near In the development of digital video watermarking, geometric attacks (including scaling, rotation, shifting, cropping, removing or adding pixel matrix, etc.) are one of the biggest challenges faced by watermarking technology. . [0003] In order to overcome the problem that most watermarks cannot be extracted normally after the geometric transformation of the watermark carrier, someone first proposed a digital watermarking technology that can re...

AI Technical Summary

Problems solved by technology

[0004] However, the inventor found in the process of implementing the above-mentioned scheme that the above-mentioned technical scheme has defects. For the first scheme, since the log-polar mapping is an irreversible lossy operation, when the log-polar inverse mapping is performed after embedding the watermark, it is necessary to The interpolation operation is performed on the image coefficients, so that the quality of the image after the watermark is embedded is drastically reduced; and when the watermark is extracted, the watermark image must be subjected to DFT transformation - log-polar mapping - FFT transformation to obtain the coefficients on the geometric invariant domain Then for detection, the whole process requires a very large amount of calculation, and the logarithmic base in the log-polar mapping in the algorithm is a=e=2.7183, so that the embedded redundancy of different watermark bits is very different , that is, different watermark bits have different robustness, which makes the robustness of watermark information bits inconsistent, which is not suitable for watermark embedding algorithm

Images