The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008 
1194 
Figure 14 presents the thinned candidate road with 
morphological algorithm. The comparison between the final 
detected road in Figure 15 with the original microwave radar or 
multi-spectral image shows that the road extracted is quite 
satisfactory. 
5.2 Road Extraction with Fusion of Radarsat-1 SAR and 
Landsat ETM+ 
A RADARSAT-1 SAR data (5.3GHz, HH polarization, spatial 
resolution 6.25m) on the November 18, 2002 and the 
corresponding Landsat ETM+ data with the same characteristics 
as that in Figure 6 are also taken as an example. They are 
shown in Figure 16 and Figure 17. Figure 18 presents the road 
extracted with method of this paper, which is also very 
satisfactory. 6 
Figure 16. Radarsat-1 SAR image of Shanghai, China 
Figure 17. Landsat ETM+531 image of Shanghai, China 
Figure 18. The extracted road image 
6. CONCLUSION 
In this paper, the LS-FM algorithm is developed to fuse the 
multi-spectral and microwave radar remote sensing images to 
extract road from complex urban areas. It is applied to the data 
fusion and road extraction from the ERS-2 SAR image and 
Landsat ETM+ , as well as RADARSAT-1 SAR image and 
ETM+, in Shanghai area, China. 
(1) The iteration difference algorithm is a good way to present 
the spectral difference of different objects in multi-spectral 
images with only several iterations, especially those with big 
difference in reflectivity. 
(2) Fused images from multiple sensors, such as infrared ETM+ 
and microwave radar images, can yield satisfactory road 
extraction of complex terrain surfaces. Using the LS-FM and 
multiple sensor’s fused data, their advantages can be 
synthesized together to make better road extraction 
ACKNOWLEDGEMENTS 
This work was supported by China State Major Basic Research 
Project (2005CB724204), the National Natural Science 
Foundation of China (40701020). 
REFERENCES 
Anys H., Bannari A., He D. C., Morin D.,1994. Texture 
analysis for the mapping of urban areas using airborne MEIS-II 
images. In: Proceedings of the First International Airborne 
Remote Sensing Conference and Exhibition, Strasbourg, France, 
Vol.3, pp: 231-245.' 
Baumgartnerl A., Steger C., Mayerl H., Eckstein W., Ebner 
H.,1999. Automatic road extraction based on multi scale, 
grouping, and context. Photogrammetric Engineering and 
Remote Sensing, Vol. 65, pp: 777-785. 
Bums J., Hanson A., Riseman E.,1986. Extracting straight lines. 
IEEE Transactions on Pattern Analysis and Machine Intelligent, 
pp: 425-455. 
Caselles V., Kimmer R., Sapiro G.,1997. Geodesic active 
contours. International Journal of Computer Vision, 22(1), pp: 
61-79. 
Dell A. F., Gamba P.2001. Detection of urban structures in 
SAR images by robust fuzzy clustering algorithms: the example 
of street tracking. IEEE Transactions on Geoscience and 
Remote Sensing, Vol. 39, pp: 2287-2297. 
Haralick R. M., Shanmugan K., Dinstein I., 1973. Textural 
features for image classification. IEEE Transactions on Systems, 
Man, and Cybernetics, 3(6), pp: 610-621. 
Jeon K., Jang J. H., Hong K. S.,2002. Road detection in 
spacebome SAR images using a genetic algorithm. IEEE 
Transactions on Geoscience and Remote Sensing, Vol. 40, 22- 
29. 
Jin Y. Q.,2005. Theory and Approach of Information Retrievals 
from Electromagnetic Scattering and Remote Sensing. Germany: 
Springer, Chapter 3. 
Rouy E., Tourin A., 1992. A viscosity solutions approach to 
shape-from-shading. SIAM Journal on Numerical Analysis, 
29(3), pp: 867-884.