Progressive orthogonal wavelets: a review (Q2827615)

This article studies progressive wavelets and collects several results from the literature to provide a unified picture.NEWLINENEWLINEProgressive wavelets are wavelets defined on a half-line. More specifically, they are functions \(\phi \in L_2(\mathbb{R})\) with \(\phi(x) = 0\) for all \(x <0\) and such that {\parindent=0.6cm\begin{itemize}\item[--] All integer translates of \(\phi\) are orthogonal. \item[--] A functional equation: \(\sigma(x) = \phi(x) - \sum_{k=1}^{N-1} \alpha_k \phi(x-k)\), for some real numbers \(\alpha_k\), and \(N \in \mathbb{N}\) holds for some function \(\sigma\). NEWLINENEWLINE\end{itemize}} It is not hard to see that, for given \(\alpha_k\), the functional equation prescribes a one-to-one correspondence between \(\sigma\) and \(\phi\). One thus calls \(\sigma\) the updating function of \(\phi\).NEWLINENEWLINEThe paper collects a couple of results from the literature to provide conditions on \(\alpha_k\) and \(\sigma\) such that there exist associated progressive orthogonal wavelets, i.e., that the integer translates of the corresponding \(\phi\) are orthogonal.NEWLINENEWLINEAs a special example, B-splines are chosen as the update function and precise conditions on the \(\alpha_k\) are given which yield progressive orthogonal wavelets.NEWLINENEWLINEFor the entire collection see [Zbl 1344.00012].