Wavelet-multifractal processing of signals

This research develops a new approach in processing multimedia signals (i.e., speech, audio, and images) using wavelets and multifractals. Processing multimedia signals is critical in many modern telecommunication and information systems. Nonstationary behaviours have been the most important characteristics of multimedia signals because they carry signal information. Unfortunately nonstationary parts are difficult to process effectively using traditional tools such as Fourier transforms. In contrast, wavelets and multifractality not only reveal that nonstationary parts are sufficient to represent natural images, but also can be used to reconstruct such signals. The general objectives of this research are to develop the wavelet and multifractal concepts into a new signal-processing framework, to obtain more effective and efficient signal representation. If the objectives can be achieved, the results of the proposed research can be a breakthrough in nonstationary signal processing, as fundamental as the well known Fourier approach in stationary signal processing