On the localization of discontinuities of the first kind for a function of bounded variation

From MaRDI portal
Publication:2446993





The localization of singularities is a nonlinear ill-posed problem which can be solved by regularizing algorithms. Let \(x\) be a piecewise continuous real function of bounded variation. In this paper, the authors detect the points \(s_j\) \((j=1,2,\dots)\) of jump discontinuities of \(x\), where instead of \(x\), an approximate function \(x^{\delta}\in L^2({\mathbb R})\) with \(\|x- x^{\delta}\|\leq \delta\) is known. Using the continuous function \[ x_{\lambda}^{\delta}(s) = \int_{-\infty}^{\infty} x^{\delta}(t)\, \frac{\partial}{\partial s} \Phi (\frac{s-t}{\lambda})\, dt\,, \] where \(\Phi\) is a convenient window function and where \(\lambda > 0\) is a regularization parameter, the singularities \(s_j\) (with sufficiently large jump magnitudes) of \(x\) are localized. If \(s_j^{\delta}\) is a point of local maximum of \(|x_{\lambda}^{\delta}|\), then it is shown that \(|s_j - s_j^{\delta}| = {\mathcal O}(\delta^2)\). The optimality of this localization method is proven for certain classes of functions.











This page was built for publication: On the localization of discontinuities of the first kind for a function of bounded variation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2446993)