On the contrary, the topographic map road network which 
is used in part 6 is very reliable, so it can be used as a 
whole and is registered on the image. But it requires to 
be firstly digitized and the symbolic knowledge is not 
directly available. 
As a conclusion, the external data are very useful for 
image analysis but they have to be used cautiously. The 
interpretation method has to be adapted to the data source 
characteristics. 
8. REFERENCES 
Bordes G., Giraudon G., Jamet O. 1995. Road 
Extraction Guided by a Cartographic Database : Creating 
a Strategy. ICCV'95, in the proceedings of the Workshop 
on Context Based Vision, Cambridge, USA, p 30-37 
Guérin P., Deseilligny M., 1995a. Global analysis of 
scanned maps for an automatic interpretation. In ICIP'95, 
Edinburgh, UK, p. 80-84. 
Guérin P., Maitre H., 1995b. Distortion model in road 
networks from maps: identification and assessment. 
proceedings SPIE. Aerosense'95, Orlando, USA, p. 232- 
243. 
138 
Maitre H., Bloch I., Moissinac H., Gouinaud C. 1995. 
Cooperative use of aerial images and maps for the 
interpretation of urban scenes. In proceedings "Automatic 
extraction of Man-Made Objects from Aerial & Space 
Images", Birkhaüser Verlag, Ascona, Swiss, p. 297-306. 
Roux L., Desachy J. 1995. Information fusion fy 
supervised classification in a satellite image. Proceedings 
Fuzzy-IEEE/IFES'95, Yokohama, Japan. 
Servigne S. 1993. Base de Données géographiques et 
photos aériennes : de l'appariement à la mise à jour. PhD, 
INSA, Lyon, France. 
Strat T.M. & Fischler M. 1995. The Role of Context in 
Computer Vision, ICCV 95, In proceedings of the 
workshop "Context based Vision", Cambridge (MA), 
USA, p. 2-12. 
Yu S. Giraudon G. & Berthod M. 1994. Integrating Map 
Knowledge in Satellite Image Analysis. In « Image and 
Signal Processing for Remote Sensing », EUROPTO- 
SPIE, Rome, Italy, vol. 2315, p. 102-112. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996 
Te 
te 
Bi 
ite