Yandong Wang 
  
shown in Figure 2e. At this level, the main structure of the road network was well detected and some minor roads were 
extracted as well. However, some road parts were not extracted at this stage. With the extracted road network as a 
guide, grouping proceeded to the next level and short line segments around the detected roads were connected based on 
their similarities of geometric and radiometric properties. In this way, missing road segments were detected while the 
effects of non-road line features were eliminated effectively. The results of grouping at level 2 and 3 are shown in 
Figure 2f and 2g. As can be seen, the missing road segments were extracted successfully. However, some isolated line 
segments were retained. Some of them are minor roads while others correspond to non-road linear structures. Having 
applied the rules described in Section 3 to them, the isolated line segments were removed and the road network was 
retained, as shown in Figure 2h. There is a short segment missing at the bottom of the road running vertically. This is 
caused by that part of the image being blurred and hence the road was not extracted by the morphological operator. 
5 CONCLUSIONS 
This paper presents a novel approach for automatic road network extraction from low-resolution images. The method 
first extracts line features and generates a hierarchy of line images by image processing. The lines are then grouped 
hierarchically to form a complete road network. One advantage of the method is that it can avoid the problem of 
threshold selection. At the same time, hierarchical grouping of lines, i.e., from sparse to dense distribution of line 
segments, can eliminate most non-road features effectively during the grouping process, and thus can generate a reliable 
road network. Non road features can be removed using the knowledge of road networks. However, some minor roads 
will still be missing as the case in this study. It would be necessary to overcome these problems by a user friendly 
interface allowing an operator to easily determine lines which should be removed or added. 
REFERENCES 
Bajcsy, R. and Tavakoli, M. (1976). Computer recognition of roads from satellite pictures. IEEE Transactions on 
Systems, Man, and Cybernetics, Vol. SMC-6, No. 9, 623-637. 
Bordes, G., Giraudon, G. and Jamet, O. (1997). Road modeling based on a cartographic database for aerial image 
interpretation. In: Semantic Modeling for Acquisition of Topographic Information from Images and Maps (W. Förstner 
and L. Plümer, eds.). Birhäuser, Verlag, Basel, pp. 123-139. 
Dougherty, E.R. (1992). An Introduction to Morphological Image Processing. A Publication of SPIE - The International 
Society for Optical Engineering, Bellingham, Washington. 
Fischler, M.A., Tenenbaum, J.M. and Wolf, H.C. (1981). Detection of roads and linear structures in low-resolution 
aerial imagery using a multisource knowledge integration technique. Computer Graphics and Image Processing, 15, 
201-223. 
Geman, D. and Jedynak, B. (1996). An active testing model for tracking roads in satellite images. IEEE Transactions on 
Pattern Analysis and Machine Intelligence, Vol. 18, No.1, 1-14. 
Grimson, W.E.L. (1990). Object Recognition by Computer: the Role of Geometric Constraints. The MIT Press, 
Cambridge, Massachusetts. 
Grün, A. and Li, H. (1995). Road extraction from aerial and satellite images by dynamic programming. ISPRS Journal 
of Photogrammetry and Remote Sensing, 50(4): 11-20. 
Gunst, M. de. 1996. Knowledge-Based Interpretation of Aerial Images for Updating of Road Maps. Publications on 
Geodesy, No. 44, Netherlands Geodetic Commission. 
Henricsson, O. (1996). Analysis of Image Structures Using Colour Attributes and Similarity Relations. PhD 
Dissertation, ETH, Zurich, Switherland. 
Maillard, P. and Cavayas, F. (1989). Automatic map-guided extraction of roads from SPOT imagery for cartographic 
database updating. Int. J. Remote Sensing, Vol. 10, No. 11, 1775-1787. 
Ramer, U. (1972). An iterative procedure for the polygonal approximation of planar curves. Computer Graphics and 
Image Processing, 1: 244-256. 
  
948 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000.