Digital watermark production and recognition method used for audio document

Abstract

The invention discloses a digital watermark manufacture method and a recognition method used for audio files. The watermark is added in the audio digital file to be processed and taken as copyright marking information; the method of the invention is characterized in that the watermark is realized by adopting converted voiceprint information and comprises the steps as follows: a section of speech signals with copyright identification is colleted and a voiceprint watermark image is gained; the speech signal is converted into one-dimensional two-value sequence group; the gray-scale value of each matrix unit is presented by adopting an eight-bit binary number; the audio file where the watermark is to be inserted is divided into sections; frequency domain transformation is respectively carried out to each audio section so as to obtain the frequency spectrum file; a bit of one-dimensional two-value sequence group watermark data is embedded into each section of the frequency spectrum file; the embedding method is a method used for quantitatively modifying the frequency spectrum file coefficient; reverse conversion is carried out so as to obtain the audio data of the embedded watermark; the audio data is combined with the audio data which is processed so as to obtain the audio digital file with the digital watermark. Reverse separation operation is carried out during the recognition process. The method of the invention effectively solves the copyright verification problem of the digital audio product.

Description

technical field [0001] The invention relates to a digital audio frequency watermark embedding and detection method, which belongs to the fields of information security and multimedia information processing. Background technique [0002] The development of multimedia technology and network technology has significantly improved the efficiency of information dissemination. But the side effect that comes with it is that malicious individuals and groups may copy, tamper with, and sell copyrighted content at will without the permission of the owner of the work. Digital watermarking technology is proposed to protect multimedia products in the network environment, and is considered to be one of the effective ways to solve copyright issues. It uses the redundancy of the human visual and auditory system to embed secret information related to copyright owners in multimedia files to verify the copyright ownership of the information. From the scope of application, watermarking can be d...

AI Technical Summary

Problems solved by technology

However, in practical applications, to achieve all the above functions, the implementation is complex and time-consuming

[0006] (2) The embedded watermark is abstract, which is not conducive to the court's certification of the copyright of digital audio works

There are also technologies that use some binary images or grayscale images as audio watermarks. Most of these watermark images are simple and easy to forge images, and these watermark images are not unique, so it is impossible to clearly indicate copyright ownership

[0007] Therefore, the existing audio watermarking methods are not practical and abstract, and are not suitable for copyright protection and copyright certification of digital audio works.

Images