Road Extraction and Network Building from Synthetic Aperture Radar Images using A-Priori Information
conference paper
This paper describes a method for the extraction of road
networks from radar images. Three phases can be
distinguished: (1) detection of road lines, (2) network
building, and (3) network fusion. The method has been
demonstrated on two radar images, one urban and one rural.
Despite the differences, the procedure was able to cover a
large part of both networks that was not obscured by trees.
This was done using roughly the same parameters.
Topology was preserved, and one of the networks was fused
with a network from a 1:50,000 digital map, and one
extracted from a high-resolution optical image. However,
the network extraction and merging process described in
this paper can not be regarded as optimal. Additional rules
and processing are required. Useful input is expected from
image analysts with respect to the procedure and results.
networks from radar images. Three phases can be
distinguished: (1) detection of road lines, (2) network
building, and (3) network fusion. The method has been
demonstrated on two radar images, one urban and one rural.
Despite the differences, the procedure was able to cover a
large part of both networks that was not obscured by trees.
This was done using roughly the same parameters.
Topology was preserved, and one of the networks was fused
with a network from a 1:50,000 digital map, and one
extracted from a high-resolution optical image. However,
the network extraction and merging process described in
this paper can not be regarded as optimal. Additional rules
and processing are required. Useful input is expected from
image analysts with respect to the procedure and results.
Topics
TNO Identifier
93933
Publisher
IEEE
Source title
2008 IEEE International Geoscience & Remote Sensing Symposium - IGARSS 2008, July 6-11, 2008, Boston, U.S.A.
Place of publication
Piscataway, NJ
Pages
471-474
Files
To receive the publication files, please send an e-mail request to TNO Repository.