Difficulties were experienced in areas still under completion 
where the road network was very dense. In parts where also 
houses were under construction the exact delineation was hard to 
determine from the image. Field completion of the map should 
therefore be necessary for accurate mapping. It was found that 
during the analysis, the roads that needed checking could be 
reliably identified. This would be an important means to improve 
the efficiency of the field work. 
From the experience in the study, it was not considered possible 
to create a completely new road map only from the satellite data. 
The small neighbourhood roads were to a large extent impossible 
to distinguish in older areas where the road network is dense and 
vegetation obscure the ground. 
Satellite data are suitable for revision of road information because 
of the high contrast of new roads. In the study area, this higher 
contrast was still noticeable in five year old roads. If the map to 
be revised is too old, the contrast of the roads to be added may 
already have faded and the vegetation cover may preclude 
detection. 
The multispectral image was considerably less valuable than the 
panchromatic data. It was possible to detect the new roads, but 
difficult to determine the delineation where the road network was 
dense. 
As during the land-cover interpretation, the possibilities of map 
overlaying, interactive zooming, and contrast enhancement in the 
image processing/GIS system were useful for the analysis. 
Disadvantages included the somewhat slow progress when 
interpreting and digitizing directly on the monitor, difficulties in 
getting good hard-copy images for field work, and the high cost 
of equipment, training and service. As it should not be 
impossible to produce similar maps using images in photographic 
form, the disadvantages of computer-based processing may 
outweigh the advantages for many operational users. 
CONCLUSIONS 
SPOT data proved to be valuable for accurate mapping and 
revision of roads and land-cover information at 1:25 000 scale. 
Zones of similar pattern, such as residential areas of different 
densities, can be delineated in the image. It is often not possible 
to identify the type of land-cover in detail directly from the 
image. For this, familiarity with the area and reference data in 
the form of field work, aerial photography, and maps are needed. 
Road information was updated with good results. Field 
completion is necessary to achieve a completely accurate map, 
but the roads that needed checking could be reliably identified 
during the analysis. 
As very little reference data was available and as the interpreter 
was not familiar with the area, the study could be taken as a 
measure of the amount of information possible to readily extract 
from the imagery with the help of limited field work. In 
operational applications, the addition of more reference data 
should make possible more detailed and accurate mapping and 
revision. 
357 
ACKNOWLEDGEMENTS 
The author wishes to thank J. van Woerden for fruitful 
collaboration in the work leading up to this paper, L. Eklundh for 
his helpful comments, and the staff at HUDCC for kind 
assistance in the field. 
REFERENCES 
Bertaud, M.-A., 1989. The Use of Satellite Images for Urban Planning: 
A Case Study from Karachi, Pakistan. Report INU 42, World Bank, 
Washington, D.C., USA. 
Brouwer, H. de, C.R. Valenzuela, L.M. Valencia and K. Sijmons, 
1990. Rapid assessment of urban growth using GIS-RS techniques. ITC 
Journal, 1990(3):233-235. 
Gamba, A., 1990. La télédétection satellitaire au service d'un 
observatoire urbain de l'aire metropolitaine de Buenos Aires. IAURIF, 
Paris, France. 
Gastellu-Etchegorry, J.P., 1990. An assessment of SPOT XS and 
Landsat MSS data for digital classification. of near-urban land cover. 
Int. J. Remote Sens., 11(2):225-235. 
Gong, P., and P.J. Howarth, 1990. The Use of Structural Information 
for Improving Land-Cover Classification Accuracies at the Rural-Urban 
Fringe. Photogramm. Eng. Remote Sens. , 56(1):67-73. 
Manning, J. and M. Evans, 1988. Revision of Medium Scale 
Topographic Maps Using Space Imagery. In: Int. Arch. Photogramm. 
Remote Sens., 27(B4):233-245. 
Moller-Jensen, L., 1990. Knowledge-Based Classification of an Urban 
Area Using Texture and Context Information in Landsat-TM Imagery. 
Photogramm. Eng. Remote Sens., 56(6):899-904. 
Paulsson, B., 1992. Urban applications of satellite remote sensing and 
GIS analysis. UMP Report, The World Bank and UNCHS, Washington, 
D.C., USA and Nairobi, Kenya. 
Salgé, F. and M.J. Roos-Josserand, 1988. Apport des images satellites 
à la base de données cartographiques de l’IGN: Résultats de l'expérience 
Franco-Canadienne sur le contenu des images. In: Proc. Int. Symp. on 
Topographic Applications of SPOT Data, Oct. 13-14, Sherbrooke, 
Canada. 
Thibault, C., M. Belliot and R. Delavigne, 1991. Suivre le 
développement urbain du grand Caire par satellite - Premiers résultats. 
La lettre d'information, Réseau A.D.O.C, No 6, ORSTOM, Bondy, 
France. 
Van Cleynenbreugel, J., F. Fierens, P. Suetens and A. Oosterlinck, 
1990. Delineating Road Structures on Satellite Imagery by a GIS-Guided 
Technique. Photogramm. Eng. Remote Sensing, 56(6):893-898.