Robust Audio Watermarking for Copyright Protection

This paper introduces an audio watermark encoding system that is robust to commonly used audio compression techniques such as MP3, AAC and OGG. The watermark is embedded by applying MPEG Layer 2 Model 2 psychoacoustic masking in an iterative way. The decoding is performed by using a correlation based blind decoding algorithm supported via a wavelet denoising module. Unlike most of the existing methods, the developed watermarking algorithm controls decoding errors at encoding stage, therefore guarantees 100% decoding accuracy on digital records. It does not require the original audio for decoding and robust to channel noise as well as changes in the sampling rate. Key-Words: -Psychoacoustic masking, data hiding, audio watermarking, MP3.