par Carleer, Alexandre ;Debeir, Olivier ;Wolff, Eléonore
Editeur scientifique Bruzzone, Lorenzo
Référence (9-12 September 2003: Barcelona, Spain), Image and Signal Processing for Remote Sensing IX, SPIE, Bellingham, page (532-542)
Publication Publié, 2004-02-02
Publication dans des actes
Résumé : Since 1999, very high spatial resolution data represent the surface of the earth with more details. However, information extraction by computer-assisted classification techniques proves to be very complex owing to the internal variability increase in land-cover units and to the weakness of spectral resolution1, 2, 3. The increase in variability decreases the statistical separability of land-cover classes in the spectral space 4. Per pixel multispectral classification techniques are then insufficient for an extraction of complex categories and spectrally heterogeneous land-cover, like urban areas5. Per region classification was proposed as an alternative approach6, 7. The first step of this approach is the segmentation. A large variety of segmentation algorithms were developed these last 20 years8 and a comparison of their implementation on very high spatial resolution images is necessary. For this study, four algorithms from the two main groups of segmentation algorithms (boundary-based and region-based algorithms) were selected. In order to compare the algorithms, an evaluation of each algorithm was carried out with empirical discrepancy evaluation methods. This evaluation is carried out with a visual segmentation of IKONOS panchromatic images.