Road centreline extraction from high‐resolution imagery based on multiscale structural features and support vector machines
- 20 April 2009
- journal article
- research article
- Published by Informa UK Limited in International Journal of Remote Sensing
- Vol. 30 (8), 1977-1987
- https://doi.org/10.1080/01431160802546837
Abstract
This paper investigates road centreline extraction from high‐resolution imagery. A novel road detection system is proposed based on multiscale structural features and support vector machines (SVMs). The salient aspects of the strategy are: (1) structural features are exploited because road objects are narrow and extensive, with large perimeters and small radii; (2) the object‐based approach is used to extract multiscale information so as to reduce the local spectral variation caused by vehicles, shadows, road markings, etc.; (3) the hybrid spectral–structural features are analysed using the SVM classifier; and (4) multiple object levels are integrated because a multiscale approach can exploit the rich spatial information and detect multiscale road objects. Experiments were conducted on two IKONOS multispectral datasets and the results validated the proposed method.Keywords
This publication has 15 references indexed in Scilit:
- Classification and Extraction of Spatial Features in Urban Areas Using High-Resolution Multispectral ImageryIEEE Geoscience and Remote Sensing Letters, 2007
- The use of small training sets containing mixed pixels for accurate hard image classification: Training on mixed spectral responses for classification by a SVMRemote Sensing of Environment, 2006
- Improving Urban Road Extraction in High-Resolution Images Exploiting Directional Filtering, Perceptual Grouping, and Simple Topological ConceptsIEEE Geoscience and Remote Sensing Letters, 2006
- Urban road extraction from high-resolution optical satellite imagesInternational Journal of Remote Sensing, 2005
- A multiscale texture analysis procedure for improved forest stand classificationInternational Journal of Remote Sensing, 2004
- A cognitive pyramid for contextual classification of remote sensing imagesIEEE Transactions on Geoscience and Remote Sensing, 2003
- A comparison of three image-object methods for the multiscale analysis of landscape structureISPRS Journal of Photogrammetry and Remote Sensing, 2003
- Detection of urban structures in SAR images by robust fuzzy clustering algorithms: the example of street trackingIEEE Transactions on Geoscience and Remote Sensing, 2001
- Remote Sensing Digital Image AnalysisPublished by Springer Science and Business Media LLC ,1999
- Automatic finding of main roads in aerial images by using geometric-stochastic models and estimationIEEE Transactions on Pattern Analysis and Machine Intelligence, 1996