level 3 
  
nels 
  
INTERACTIVE 
o 
= 
3 
= 
£e 
= 
o 
2 
o 
D 
= 
£5 
= 
® 
e 
© 
3 
o 
ure 7: process flow 
  
1a 1996 
   
   
  
  
  
   
   
   
   
   
   
   
    
   
   
   
   
   
   
   
   
  
  
  
  
   
   
   
   
   
   
  
   
   
   
  
   
   
   
  
   
  
   
  
   
  
   
  
   
  
   
   
  
  
   
  
  
  
   
  
  
  
   
   
6. Conclusions 
Shape descriptors and distance-weighted triangulation are 
powerful methods to detect very efficiently non-rigid symbols 
from scanned topographic maps. Further developments to 
detect other symbols (e.g. curved text labels, orchards) seem 
to be promising. It is also planned to use orthophotos from 
aerial photography as raster input to detect natural objects like 
single trees and tree groups. 
Nevertheless, several drawbacks of the chosen ,,bottom-up* 
approach have to be kept in mind: 
* trying to recognise all map symbols at once needs a top- 
down approach 
using shape discriminators to detect line symbols is very 
inefficient and lacks topology 
discrimination task needs image processing knowledge and 
user interaction 
e detectability varies with drawing quality and scanning reso- 
lution 
References 
Bihr H.P, F. Quint, U. Stilla: ,,Modellbasierte Verfahren der 
Luftbildanalyse zur Kartenfortführung", Zeitschrift 
für Photogrammetrie und Fernerkundung, Vol. 6, pp. 
224-233, 1995. 
Ballard D.H.: ,,Generalizing the hough transform to detect 
arbitrary shapes", Pattern recognition letters, Vol. 13, 
pp. 111-122, 1981. 
Ballard D. H., C.M. Brown: ,Computer vision", 523 pp., 
Prentice Hall, Englewood Cliffs, 1982. 
Boesch R.: ,Improved utilization of Arc/INFO image 
catalogs", Proc. 10th European Arc/INFO 
Conference, 2.-4. Oct 1995, Prag, pp. IX2-IXT, 
ESRI, Redlands, CA., 1995. S 
Chang Ji Y., Hanson A.J.: ,, Virtual line segment-based hough 
transform", IEEE Trans. on Pattern Analysis and 
Machine Intelligence, Vol. 16 (1), pp. 57-62, 1994. 
Han Joon H., Koczy L.T., Poston T.: ,,Fuzzy Hough 
transform", Pattern Recognition Letters, Vol. 15, pp. 
649-658, 1994. 
Henricsson O., W. Neuenschwander: ,,Controlling growing 
snakes by using key-points", Int. Conference on 
Pattern Recognition, Jerusalem, Israel, pp. 68-73, 
1994. 
Kass M., A. Witkin, D. Terzopoulos: ,,Snakes: active contour 
models", Proc. of First International Conf. on 
Computer Vision, pp. 259-269, 1987. 
L+T , Podukteinformation Pixelkarten (german)”, 10 pp., 
Swiss Federal Office of Topography 
(Landestopographie), Wabern, Bern, 1994. 
Lai Kok Fung: ,,Deformable contours: modeling, extraction, 
detection and classification", Ph.D. Thesis, 92 pp., 
Electrical Engineering, Univ. Wisconsin-Madison, 
1994. 
Lee H., R.H. Park: ,Relaxation algorithm for shape matching 
of two-dimensional objects", Pattern Recognition 
Letters, Vol. 10, pp. 309-313, 1989. 
Lin C.C., R. Chellappa: ,Classification of partial 2-D shapes 
using fourier descriptors”, IEEE Trans. on Pattern 
Analysis and Machine Intelligence, Vol. 9, pp. 687- 
69 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996 
690, 1987. 
Lin Wei: „Ein Beitrag zur kartographischen Mustererkennung 
mittels Methoden der künstlichen Intelligenz", 
Dissertation TU Darmstadt, Vol. 419, 98 pp., 
Deutsche Geodätische Kommission, Bayerische 
Akademie der Wissenschaften, 1994. 
Lütjen K., H. Füger, H.-J. Greif, K. Jurkiewicz: 
„Auswahlverfahren für die wissensbasierte 
Bildauswertung mit dem blackboard-orientierten 
Produktionssystem BPI", Mustererkennung 1987, 9. 
DAGM-Symposium, pp. 290-294, Springer, Berlin, 
1987. 
Marshall S.: „Review of shape coding techniques", Image 
Vision Computing, Vol. 7 (4), pp. 281-294, 1989. 
Meyer H.: „Automatische wissensbasierte Extraktion von 
semantischer Information aus gescannten Karten", 
Dissertation TU München, Vol. 417, 113 pp., 
Deutsche Geodätische Kommission, Bayerische 
Akademie der Wissenschaften, 1993. 
Palmer PL., M. Petrou, J. Kittler: „A Hough transform 
algorithm with a 2D hypothesis testing kernel", pp. 
221-234, 1993. 
Pavlidis T: ,,A review of algorithms for shape analysis", 218 
pp., Dept. of Elec. Eng. and Comp. Science, 
Princeton Univ., 1976. 
Pitas I.,A.N. Venetsanopoulos: ,,Morphological shape 
representation", Pattern Recognition, Vol. 25 (6), pp. 
555-565, 1992. 
Rauber T W.: ,,Two-dimensional shape descriptions”, 
Technical Report GR UNINOVA-RT-10-94 Ph.D. 
Thesis, Universidade Nova de Lisboa, 38 pp., 1994. 
Reeves A.P, A. Rostampour: ,,Shape analysis of segmented 
objects using moments", Proc. 1981 Conf. Pattern 
Recognition Image Processing, Dallas, pp. 171-174, 
IEEE, New York, 1981. 
Rostampour A.R., Madhvapathy PR.: ,,Shape recognition 
using simple measures of projections”, Proceedings 
of 7th Intl. Phoenix Conference on Computers and 
Communications, 16-18 March 1988, pp. 474-479, 
IEEE, Arizona State Univ., 1988. 
Sedgewick R.: ,,Algorithms in C++", 656 pp., Addison- 
Wesley, Reading, MA, 1992. 
Staib L.H., J. Duncan: ,,Boundary finding with parametrically 
deformable models", IEEE Trans. on Pattern 
Analysis and Machine Intelligence, Vol. 14, pp. 
1061-1075, 1992. 
Stengele Roland: ,,Kartografische Mustererkennung", 
Dissertation ETH Zürich, 1995. 
Trahanias P.E.: „Binary shape recognition using the 
morphological skeleton transform", Pattern 
Recognition, Vol. 25 (11), pp. 1277-1288, 1992. 
Udomkesmalee S.: ,,Shape recognition in the Fourier domain", 
Proceedings of the SPIE, Vol. 1564, pp. 464-472, 
SPIE, 1991. 
Weber W.: „Kartographische Mustererkennung", 
Kartographische Nachrichten, Vol. 3, pp. 113-120, 
1988. 
Zahn C.T, R.Z. Roskies: „Fourier descriptors for plane closed 
curves", JEEE Trans. Computing, Vol. C-21, pp. 269- 
281, 1972.