A memetic approach to discrete tomography from noisy projections

Vito Di Gesu', Cesare Fabio Valenti, Giosue' Lo Bosco, Filippo Millonzi, Giosuè Lo Bosco, Cesare Valenti, Vito Di Gesù, Filippo Millonzi

Risultato della ricerca: Article

10 Citazioni (Scopus)

Abstract

Discrete tomography deals with the reconstruction of images from very few projections, which is, in the general case, an NP-hard problem. This paper describes a new memetic reconstruction algorithm. It generates a set of initial images by network flows, related to two of the input projections, and lets them evolve towards a possible solution, by using crossover and mutation. Switch and compactness operators improve the quality of the reconstructed images during each generation, while the selection of the best images addresses the evolution to an optimal result. One of the most important issues in discrete tomography is known as the stability problem and it is tackled here, in the case of noisy projections, along four directions. Extensive experiments have been carried out to evaluate the robustness of the new methodology. A comparison with the output of two other evolutionary algorithms and a generalized version of a deterministic method shows the effectiveness of our new algorithm.
Lingua originaleEnglish
pagine (da-a)3073-3082
Numero di pagine10
RivistaPattern Recognition
Volume43
Stato di pubblicazionePublished - 2010

Fingerprint

Tomography
Evolutionary algorithms
Computational complexity
Switches
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cita questo

A memetic approach to discrete tomography from noisy projections. / Di Gesu', Vito; Valenti, Cesare Fabio; Lo Bosco, Giosue'; Millonzi, Filippo; Lo Bosco, Giosuè; Valenti, Cesare; Di Gesù, Vito; Millonzi, Filippo.

In: Pattern Recognition, Vol. 43, 2010, pag. 3073-3082.

Risultato della ricerca: Article

Di Gesu', V, Valenti, CF, Lo Bosco, G, Millonzi, F, Lo Bosco, G, Valenti, C, Di Gesù, V & Millonzi, F 2010, 'A memetic approach to discrete tomography from noisy projections', Pattern Recognition, vol. 43, pagg. 3073-3082.
Di Gesu', Vito ; Valenti, Cesare Fabio ; Lo Bosco, Giosue' ; Millonzi, Filippo ; Lo Bosco, Giosuè ; Valenti, Cesare ; Di Gesù, Vito ; Millonzi, Filippo. / A memetic approach to discrete tomography from noisy projections. In: Pattern Recognition. 2010 ; Vol. 43. pagg. 3073-3082.
@article{8223cdf7435f4cd7b97fa37490c6a44f,
title = "A memetic approach to discrete tomography from noisy projections",
abstract = "Discrete tomography deals with the reconstruction of images from very few projections, which is, in the general case, an NP-hard problem. This paper describes a new memetic reconstruction algorithm. It generates a set of initial images by network flows, related to two of the input projections, and lets them evolve towards a possible solution, by using crossover and mutation. Switch and compactness operators improve the quality of the reconstructed images during each generation, while the selection of the best images addresses the evolution to an optimal result. One of the most important issues in discrete tomography is known as the stability problem and it is tackled here, in the case of noisy projections, along four directions. Extensive experiments have been carried out to evaluate the robustness of the new methodology. A comparison with the output of two other evolutionary algorithms and a generalized version of a deterministic method shows the effectiveness of our new algorithm.",
keywords = "Discrete tomography; Stability problem; Evolutionary reconstruction.",
author = "{Di Gesu'}, Vito and Valenti, {Cesare Fabio} and {Lo Bosco}, Giosue' and Filippo Millonzi and {Lo Bosco}, Giosu{\`e} and Cesare Valenti and {Di Ges{\`u}}, Vito and Filippo Millonzi",
year = "2010",
language = "English",
volume = "43",
pages = "3073--3082",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - A memetic approach to discrete tomography from noisy projections

AU - Di Gesu', Vito

AU - Valenti, Cesare Fabio

AU - Lo Bosco, Giosue'

AU - Millonzi, Filippo

AU - Lo Bosco, Giosuè

AU - Valenti, Cesare

AU - Di Gesù, Vito

AU - Millonzi, Filippo

PY - 2010

Y1 - 2010

N2 - Discrete tomography deals with the reconstruction of images from very few projections, which is, in the general case, an NP-hard problem. This paper describes a new memetic reconstruction algorithm. It generates a set of initial images by network flows, related to two of the input projections, and lets them evolve towards a possible solution, by using crossover and mutation. Switch and compactness operators improve the quality of the reconstructed images during each generation, while the selection of the best images addresses the evolution to an optimal result. One of the most important issues in discrete tomography is known as the stability problem and it is tackled here, in the case of noisy projections, along four directions. Extensive experiments have been carried out to evaluate the robustness of the new methodology. A comparison with the output of two other evolutionary algorithms and a generalized version of a deterministic method shows the effectiveness of our new algorithm.

AB - Discrete tomography deals with the reconstruction of images from very few projections, which is, in the general case, an NP-hard problem. This paper describes a new memetic reconstruction algorithm. It generates a set of initial images by network flows, related to two of the input projections, and lets them evolve towards a possible solution, by using crossover and mutation. Switch and compactness operators improve the quality of the reconstructed images during each generation, while the selection of the best images addresses the evolution to an optimal result. One of the most important issues in discrete tomography is known as the stability problem and it is tackled here, in the case of noisy projections, along four directions. Extensive experiments have been carried out to evaluate the robustness of the new methodology. A comparison with the output of two other evolutionary algorithms and a generalized version of a deterministic method shows the effectiveness of our new algorithm.

KW - Discrete tomography; Stability problem; Evolutionary reconstruction.

UR - http://hdl.handle.net/10447/50132

M3 - Article

VL - 43

SP - 3073

EP - 3082

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

ER -