Urban feature extraction on simulated WorldView-2 images

Author

Souza dos Anjos, Camila ; de Almeida, C.M. ; Galvao, L.S. ; Garcia Fonseca, Leila Maria

Author_Institution

Nat. Inst. for Space Res. - INPE, São José dos Campos, Brazil

fYear

2013

fDate

21-23 April 2013

Firstpage

284

Lastpage

287

Abstract

This work presents a procedure for simulating the eight multispectral bands of the WorldView-2 satellite by means of bands available in the airborne sensor HSS (Hyperspectral Scanner System). Both the original and simulated images were classified using the SAM (Spectral Angle Mapper) algorithm, aiming to discriminate intra-urban targets, namely: French tiles, asphalt/concrete, steel and aluminum tiles, bare soil, native arborous vegetation, and grass. The simulated images proved to be suitable for the detection of all aforementioned targets.

Keywords

aluminium; asphalt; concrete; feature extraction; geophysical image processing; hyperspectral imaging; image classification; optical sensors; soil; steel; tiles; vegetation; French tiles; HSS; SAM algorithm; WorldView-2 satellite; airborne sensor; aluminum tiles; asphalt; bare soil; concrete; grass; hyperspectral scanner system; intra-urban target discrimination; multispectral band simulation; native arborous vegetation; simulated WorldView-2 images; simulated image classification; spectral angle mapper algorithm; steel tiles; urban feature extraction; Accuracy; Feature extraction; Hyperspectral imaging; Soil; Spatial resolution; Vegetation mapping;

fLanguage

English

Publisher

ieee

Conference_Titel

Urban Remote Sensing Event (JURSE), 2013 Joint

Conference_Location

Sao Paulo

Print_ISBN

978-1-4799-0213-2

Electronic_ISBN

978-1-4799-0212-5

Type

conf

DOI

10.1109/JURSE.2013.6550720

Filename

6550720