الفهرس | Only 14 pages are availabe for public view |
Abstract Speech communication has captured the interest of researches in the last decades. There is a bad need to secure speech communication in addition to the necessity to hide some information in speech signals. This thesis is concerned with a vital branch of speech processing systems, which is speech watermarking. Two approaches for watermark embedding in speech signals are presented in this thesis. The first one depends on the Singular Value Decomposition (SVD). The SVD is selected to embed encrypted watermarks in speech signals due to its robustness to slight variations in the signals. This leads to robustness of the watermark in the presence of different signal manipulations. The second approach depends on the Empirical Mode Decomposition (EMD) for watermark embedding with SVD. The inherent advantage in the EMD is the decomposition of the speech signal into a group of envelopes. Watermark embedding can be performed in the least sensitive envelope to deterioration effects. Another trend that is considered in this thesis is the speech enhancement, especially in the presence of watermarking. The sensitivity of the watermark extraction process to the speech enhancement algorithm is also studied. |