Road extraction from high-resolution remote sensing images based on characteristics

Abstract

In this paper, an automated road extraction from high-resolution remote sensing images method based on road synthetical characteristics is presented. First, spectral property, geometric property and textural property of road class have been considered and a road clustering model has been established. Second, mathematical morphology has been applied to extract road centerlines. Experiments prove that the model of this paper could describe the features of road well and accurately identify the road class from high-resolution remote sensing images.