No abstract

No abstract

No abstract

Generalization of the Fourier domain watermarking is proposed. It has been done using joint space/spatial-frequency analysis and projections of the Wigner distribution, the two-dimensional Radon-Wigner distribution. A simple and numerically efficient algorithm for watermark embedding and detection is presented. The theory is tested on several test images with various attacks, including filtering, cropping, histogram stretching.

The excitation signal contained in a speech signal and the vocal tract parameters determining the spectral envelope can be extracted by linear prediction (LP). I n order to enhance noisy speech signals, time-varying filtering, based on the S-method as a time-frequency analysis tool, is applied to the excitation signal and to the spectral e w elope in parallel. This approach is tested with noisy speech signals apparent in hands-free telephone systems.