CNR Institutional Research Information System

CLASSIFICATORE GERARCHICO PER L ' ANALISI DI IMMAGINI QUICKBIRD

C Tarantino;P Blonda
2008
2008
Istituto di Studi sui Sistemi Intelligenti per l'Automazione - ISSIA - Sede Bari
ECOLOGIA E GOVERNANCE DEL PAESAGGIO, 2008
Sì, ma tipo non specificato
22 e 23 maggio 2008
BARI
classificazione
alta risoluzione
QuickBird
ABSTRACT: Quickbird images, characterized by very high spatial resolution, offer the possibility to monitor antropic processes at very fine scale. Land use changes, useful in many application fields, can be obtained by comparing thematic maps of different years. The reliability of the detected changes depends on the accuracy of the compared classified maps. When multi-temporal information is not available, the reduced spectral band availability of Quickbird images represents a big problem in distinguishing spectrally similar classes, such as road and buildings, woods and permanent crops with trees, arable land and shrub. In this work, spatial information is combined with spectral information to differentiate classes in a hierarchic classification approach. First, a supervised classifier, based on spectral features only, is used to discriminate the following classes: a) artificial areas, including roads and buildings, b) vegetated areas, including woods, permanent crops with trees, vineyard, shrub, arable land, c) shadows, d) barren land e) sea areas. Then the two macro-classes, i.e artificial areas and vegetated areas, are split into their sub-classes components based on the selection of class-oriented spatial features. In particular, this paper mainly focuses on the introduction of spatial information in the form of length and width for the discrimination of roads and buildings in artificial areas. For the discrimination of vegetated classes, texture measures from co-occurrence matrix have been used. In the last step the map containing the whole set of classes is recomposed. The proposed approach produced an overall percentage accuracy better than 10% in comparison with the use of a more conventional classification approach, based on spectral information. The scene analysed covers part of Salento area in Puglia Region.
1
none
N. Amoruso; C. Tarantino;P. Blonda
273
info:eu-repo/semantics/conferenceObject
04 Contributo in convegno::04.01 Contributo in Atti di convegno
04.01 Contributo in Atti di convegno
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14243/104195
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact