Abstract

Raster maps are widely available for many areas around the globe. The road network that commonly exists in raster maps is an important source of road vector data for the areas that do not have the road vectors readily available. In our previous work, we presented a supervised approach to extract the road layers from raster maps [3]. With the extracted road layer, in [2], we utilized the thinning operator to generate one-pixel width road lines and extract accurate road-intersection templates automatically. To further vectorize the whole road network, in this paper, we present an approach that utilizes the identified road-intersection templates to avoid the distortions near the intersections caused by the thinning operator and then trace the one-pixel width road lines to extract accurate road vectors from raster maps.

Date

2009

Authors

Yao-Yi Chiang, Craig A Knoblock

Journal

Proceedings of the Eighth International Workshop on Graphics Recognition