Centre de documentation
scientifique

Road extraction using SVM and image segmentation / M. Song in Photogrammetric Engineering & Remote Sensing, PERS, vol 70 n° 12 (December 2004)  

Public [article]  inPhotogrammetric Engineering & Remote Sensing, PERS > vol 70 n° 12 (December 2004) . - pp 1365 - 1371 Titre : Road extraction using SVM and image segmentation Type de document : Article/Communication Auteurs : M. Song, Auteur ; Daniel L. Civco, Auteur Année de publication : 2004 Article en page(s) : pp 1365 - 1371 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Photogrammétrie numérique [Termes IGN] bande spectrale [Termes IGN] classification orientée objet [Termes IGN] classification par maximum de vraisemblance [Termes IGN] classification par séparateurs à vaste marge [Termes IGN] extraction du réseau routier [Termes IGN] objet géographique linéaire [Termes IGN] pixel [Termes IGN] segment de droite [Termes IGN] segmentation d'image [Termes IGN] seuillage d'image [Termes IGN] vectorisation Résumé : (Auteur) In this paper, a unique approach for road extraction utilizing pixel spectral information for classification and image segmentation-derived object features was developed. In this approach, road extraction was performed in two steps. In the first step, support vector machine (SVM) was employed merely to classify the image into two groups of categories: a road group and a non-road group. For this classification, support vector machine (SVM) achieved higher accuracy than Gaussian maximum likelihood (GML). In the second step, the road group image was segmented into geometrically homogeneous objects using a region growing technique based on a similarity criterion, with higher weighting on shape factors over spectral criteria. A simple thresholding on the shape index and density features derived from these objects was performed to extract road features, which were further processed by thinning and vectorization to obtain road centerlines. The experiment showed the proposed approach worked well with images comprised by both rural and urban area features. Numéro de notice : A2004-500 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.14358/PERS.70.12.1365 En ligne : https://doi.org/10.14358/PERS.70.12.1365 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=27017 [article]