33. Istanbul 2004 
VIEN: 
CANES 
Fn 
Ted 
  
  
  
      
  
   
     
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
  
Figure 6: Highway extraction results for Karlsruhe 
depicted in Figure 6(b). The left highway (1) has been extracted 
completely. In the upper part, the crash barrier is clearly visible 
and the high-resolution lines deliver good candidates for the high- 
way. In the lower part, the highway is partly occluded by vege- 
tation (shadow), as is also the left part of highway (2). Here, the 
lines extracted at the lower resolution provide essential hypothe- 
ses, so that the extraction was successful for most parts of high- 
way (2). The gaps in the extraction of highway (2) are caused by 
bridges. Here, an introduction of context information as shown 
above might be useful. Segment (3) is a false alarm. 
5 CONCLUSION 
An approach for automatic road extraction from SAR imagery 
was presented and optimized. First, we showed that the intro- 
duction of explicitly modeled context objects helps to overcome 
many local disturbances. Second, the use of global context by in- 
troducing the outline of urban areas as seed points for rural roads 
improved the completeness of the extracted road network and the 
topological correctness of the network. Third, we modeled and 
outlined an extraction scheme for highways. Both, model and 
extraction rely on multiple scales, which makes the results more 
robust compared to single-scale approaches. 
6 ACKNOWLEDGMENT 
The author would like to thank the German Aerospace Center 
(DLR), Infoterra GmbH, Germany, and Intermap Technologies 
Corp. for providing the SAR data and Stefan Hinz for his aid in 
developing the highway road extraction. 
REFERENCES 
Barsi, A., Heipke, C. and Willrich, F., 2002. unction extraction 
by artificial neural network system — JEANS. In: International 
Archives of Photogrammetry and Remote Sensing, Vol. XXXIV, 
Part 3b, Graz, Austria, pp. 18-21. 
Baumgartner, A., Steger, C., Mayer, H., Eckstein, W. and Ebner, 
H., 1999, Automatic Road Extraction in Rural Areas. In: Inter- 
national Archives of Photogrammetry and Remote Sensing, Vol. 
XXXII, part 3-2W5, pp. 107-112. 
Hinz, S. and Baumgartner, A., 2003. Automatic extraction of ur- 
ban road networks from multi-view aerial imagery. International 
Journal of Photogrammetry and Remote Sensing, 58/1-2, pp. 83- 
98. 
Kirscht, M., 1998. Detection, velocity estimation and imaging 
of moving targets with single-channel SAR. In: Proc. of Eu- 
ropean Conference on Synthetic Aperture Radar, EUSAR 798, 
Friedrichshafen, Germany, pp. 587—590. 
Kirscht, M. and Rinke, C., 1998. 3d reconstruction of buildings 
and vegetation from synthetic aperture rader (sar) images. In: 
Workshop on Machine Vision Applications, Makuhari, Chiba, 
Japan, pp. 228-231. 
Stat, T. and Fischler, M. A., 1995. The role of context in computer 
vision. In: Proc. of the workshop on context-based vision, IEEE, 
pp. 2-12. 
Steger, C., 1998. An unbiased detector of curvilinear structures. 
IEEE Transactions on Pattern Analysis and Machine Intelligence 
20(2), pp. 113-125. 
Wessel, B., Wiedemann, C., Hellwich, O. and Arndt, W.-C., 
2002. Evaluation of automatic road extraction results from SAR 
imagery. In: International Archieves of the Photogrammetry, 
Remote Sensing and Spatial Information Sciences, Vol. XXXIV, 
Part 4, pp. 786—791. 
Wiedemann, C. and Ebner, H., 2000. Automatic completion and 
evaluation of road networks. In: International Archives of Pho- 
togrammetry and Remote Sensing, Vol. 33 (B3), pp. 979-986. 
Wiedemann, C. and Hinz, S., 1999. Automatic extraction and 
evaluation of road networks from satellite imagery. In: Interna- 
tional Archives of Photogrammetry and Remote Sensing, Vol. 32 
(3-2W5), pp. 95-100.