scientific article; zbMATH DE number 1140120

The paper is concerned with existence of minimizing pairs \((u,K)\) for the generalized Mumford-Shah functional \[ \int_{\Omega\setminus K}F(\nabla u)+\alpha(u-g)^2 dx +\beta{\mathcal H}^{n-1}(K\cap\Omega). \] The idea of the proof is to estabilish a density lower bound \[ {\mathcal H}^{n-1}(S_u\cap B_\rho(x))\geq\theta\rho^{n-1} \qquad\forall x\in \overline{S}_u,\quad B_\rho(x)\subset \Omega, \] for minimizers \(u\) of the weak formulation of the problem in the space of special functions with bounded variation. The density lower bound implies that \({\mathcal H}^{n-1}(\Omega\cap\overline{S}_u\setminus S_u)=0\), and hence that \((u,\overline{S}_u)\) is a minimizing pair. The proof of the density lower bound is related to the rate of decay in \(\rho\) of the energy \[ \int_{B_\rho(x)}F(\nabla v(y)) dy \] for local minimizers \(v\) of \(\int F(\nabla v)\) in a Sobolev space. A general decay property is proved under a strict convexity assumption on \(F\) (no smoothness is required). By a perturbation argument, also some existence results in the vectorial problem are presented.

Recommendations

Cites work

Cited in

(60)

This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4385189)