## Thesis etd-07242017-173219 |

Link copiato negli appunti

Thesis type

Elaborati finali per laurea triennale

Author

ANTONUCCI, CLARA

URN

etd-07242017-173219

Thesis title

A discrete variational approximation of the Mumford-Shah functional in dimension one

Department

MATEMATICA

Course of study

MATEMATICA

Supervisors

**relatore**Prof. Gobbino, Massimo

Keywords

- metric slope
- discrete approximations
- gamma-convergence
- Mumford-Shah functional

Graduation session start date

14/07/2017

Availability

Full

Summary

In this thesis we consider the one dimensional version of the functional introduced by D. Mumford and J. Shah (CPAM 1989) in the context of vision theory.

We present a discrete variational approximation of the Mumford-Shah functional introduced by A. Chambolle (SIAM 1995) that is related to the weak membrane model (A. Blake and A. Zissermann, 1987), also known as softening effect.

The approximating functionals F_n are the discrete version of the norm of the derivative truncated at level n.

In the first part of the thesis we prove

• the Gamma-convergence of the approximating functionals F_n to MS (the proof we

provide is different from the original one),

• a property of equicoerciveness which yields the convergence of minima and minimizers.

In the second part we recall the definition of descending metric slope and then

• we relate the Gamma-limit of the slopes to the slope of the Gamma-limit in a general metric setting,

• we compute the slope of MS and of a suitable regularization of F_n,

• we show that the Gamma-limit (over sequences with bounded energy) of the slopes of F_n is the slope of MS.

We present a discrete variational approximation of the Mumford-Shah functional introduced by A. Chambolle (SIAM 1995) that is related to the weak membrane model (A. Blake and A. Zissermann, 1987), also known as softening effect.

The approximating functionals F_n are the discrete version of the norm of the derivative truncated at level n.

In the first part of the thesis we prove

• the Gamma-convergence of the approximating functionals F_n to MS (the proof we

provide is different from the original one),

• a property of equicoerciveness which yields the convergence of minima and minimizers.

In the second part we recall the definition of descending metric slope and then

• we relate the Gamma-limit of the slopes to the slope of the Gamma-limit in a general metric setting,

• we compute the slope of MS and of a suitable regularization of F_n,

• we show that the Gamma-limit (over sequences with bounded energy) of the slopes of F_n is the slope of MS.

File

Nome file | Dimensione |
---|---|

Clara_An...nnale.pdf | 697.18 Kb |

Contatta l’autore |