Audio watermarking scheme with self-synchronization based on psychoacoustic model

Abstract

This paper introduces a robust digital audio watermarking scheme. The features of the scheme are as below:(1)using dual-scrambling mechanisms to ensure the security of the watermark;(2)using psychoacoustic models to determine the watermark embedding strength;(3)using synchronization mechanisms to achieve watermark self-synchronize detection;(4)using DWT domain watermarking to improve anti-attack capability;(5)using efficient embedding algorithm to improve the efficiency of watermark detection. Simulating experimental results show that the scheme not only has excellent imperceptible feature, but also can stably withstand various attaches such as re-sampling, re-quantifying, noise adding, low pass filtering and MP3 compressing.