ROBUST AUDIO WATERMARKING USING TIME AND FREQUENCY HOPPING

摘要

The embedding and extracting digital signature into audio without perception and robust to data compression is an important issue. In this paper, a robust and flexible audio watermarking using time and frequency hopping is proposed. The proposed method has the robustness to MP3 compression and the awareness is very low. In the embedding process, DCT is used to transform audio signal into a time and frequency domain. The region for embedding bit data is flexibly selected. Two embedding methods with redundancy are applied. In extracting process, the averaging operation is executed to correct the damaged bit. The performance with respect to bit rate, SNR and detection ratio is evaluated to show the effectiveness of our method in the simulations. The relationship between the detection ratio, the amount of embedding bit, and the redundancy parameters are also examined.