Multiscale Statistical Signal Processing PDF Download

Abstract: "In many applications (e.g. recognition of geophysical and biomedical signals and multiscale analysis of images), it is of interest to analyse and recognize phenomena occurring at different scales. The recently introduced wavelet transforms provide a time-and-scale decomposition of signals. But no corresponding statistical framework is currently available. Our purpose is to fill this gap. Our starting point is the theory of stationary processes on homogeneous trees which is shown to be the convenient framework for such a statistical multiscale signal processing. Here we introduce the corresponding system theory by investigating AR processes. This paper is a very short abstract of the report [7] intended to promote a widespread use of french."